Alimentación complementaria comunitaria para poblaciones con acceso limitado a los alimentos, en situación de vulnerabilidad y con desnutrición: una revisión global de revisiones sistemáticas
Resumen
Antecedentes
La alimentación complementaria puede ayudar a las personas con acceso limitado a los alimentos y en situación de vulnerabilidad al optimizar el valor nutricional y la adecuación de la dieta, mejorando la calidad de vida y mejorando diversos parámetros de salud de las familias en situación de desventaja. En los países de ingresos bajos y medios (PIBM), los problemas que la alimentación complementaria intenta considerar se encuentran entrelazados con la pobreza y las carencias, los programas son costosos y la provisión es complicada.
Objetivos
1. Resumir la evidencia obtenida de revisiones sistemáticas sobre la alimentación complementaria para poblaciones con acceso limitado a los alimentos, en situación de vulnerabilidad y con desnutrición, incluidos los niños menores de cinco años de edad, los niños en edad escolar, las embarazadas y las mujeres que amamantan, los pacientes con VIH o tuberculosis (o ambos) y las poblaciones de edad avanzada.
2. Describir y explorar los efectos de la alimentación complementaria en los individuos de estos grupos, y describir el rango de resultados entre las revisiones y el rango de efectos en los diferentes grupos.
Métodos
En enero 2017, se hicieron búsquedas en la Base de Datos Cochrane de Revisiones Sistemáticas (Cochrane Database of Systematic Reviews), MEDLINE, Embase y en otras nueve bases de datos. Se incluyeron revisiones sistemáticas que evalúan la alimentación complementaria comunitaria, y que tratan sobre poblaciones con acceso limitado a los alimentos, en situación de vulnerabilidad y con desnutrición. Dos autores de la revisión realizaron de forma independiente la selección de revisiones sistemáticas, la extracción de datos y la evaluación del 'Riesgo de sesgo'. Se evaluó la calidad de la revisión mediante la herramienta AMSTAR, y se utilizaron las tablas de GRADEpro de "Resumen de los resultados" de cada revisión para indicar la certeza de la evidencia para las comparaciones principales. Los hallazgos se resumieron en el texto y los datos para cada resultado se informaron en tablas adicionales. También se usaron los diagramas de bosque para mostrar los resultados de manera gráfica.
Resultados principales
Este resumen incluyó ocho revisiones sistemáticas (con las últimas fechas de búsqueda entre mayo de 2006 y febrero de 2016). Siete eran revisiones Cochrane que evaluaban las intervenciones en mujeres embarazadas; niños (desde el nacimiento hasta los cinco años de edad) de PIBM; neonatos y niños pequeños en situación de desventaja (de tres meses a cinco años de edad); niños con desnutrición moderada aguda (DAM); niños en edad escolar en situación de desventaja; adultos y niños con pruebas positivas para el VIH o con tuberculosis activa (con o sin VIH). Una era una revisión sistemática no Cochrane en personas mayores con enfermedad de Alzheimer. Estas revisiones incluían 95 ensayos relevantes a este resumen y la mayoría (74%) de los participantes provenía de PIBM.
El número de participantes incluidos varió entre 91 y 7940 adultos y entre 271 y más de 12 595 niños. Los ensayos incluían una variedad amplia de intervenciones nutricionales que variaron en cuanto a la duración, la frecuencia y el formato y a menudo se informó la administración de micronutrientes como cointervenciones. El seguimiento varió de seis semanas a dos años; tres ensayos investigaron los resultados a los cuatro a 17 años de edad. Todas las revisiones se calificaron como de alta calidad (puntuación de AMSTAR entre ocho y 11). Las calificaciones de la certeza de la calidad según GRADE variaron de muy bajas a moderadas para las comparaciones individuales y la evidencia a menudo incluyó sólo uno o dos ensayos pequeños, lo que dio lugar a muchos análisis con poco poder estadístico (demasiado pequeños para detectar diferencias pequeñas, pero importantes). Las principales categorías de resultado informadas a través de las revisiones fueron la muerte, la antropometría (adultos y niños) y otros marcadores del estado nutricional, resultados relacionados con las enfermedades, el desarrollo neurocognitivo y los resultados psicosociales, y los eventos adversos.
Los datos de la mortalidad fueron limitados y con poco poder estadístico en el metanálisis en todas las poblaciones (niños con DAM, niños con VIH y adultos con tuberculosis) con la excepción de la administración equilibrada de suplementos energéticos y proteicos en el embarazo, que puede haber reducido el riesgo de mortinato (cociente de riesgos [CR] 0,60; intervalo de confianza [IC] del 95%: 0,39 a 0,94; cinco ensayos, 3408 mujeres). La administración de suplementos en el embarazo también mejoró el peso del neonato al nacer (diferencia de medias [DM] 40,96 g; IC del 95%: 4,66 a 77,26; 11 ensayos, 5385 participantes) y redujo el riesgo de que el neonato presente un tamaño pequeño para la edad gestacional (CR 0,79; IC del 95%: 0,69 a 0,90; siete ensayos, 4408 participantes). Estos efectos no se tradujeron en beneficios a largo plazo demostrables para los niños en cuanto al crecimiento y el desarrollo neurocognitivo en uno o dos ensayos que informaron los resultados a más largo plazo. En un estudio (505 participantes), los suplementos con alto contenido de proteína se asociaron con un aumento en el riesgo de que los recién nacidos presentaran un tamaño pequeño para la edad gestacional.
Los efectos sobre el crecimiento en los niños fueron contradictorios. Una revisión encontró que la alimentación complementaria tuvo poco o ningún efecto sobre el crecimiento de los niños en los menores de cinco años de edad de PIBM; sin embargo, una revisión más reciente en una población similar halló que los que recibieron suplementos nutricionales ganaron un promedio de 0,12 kg más en el peso (DM 0,12 kg; IC del 95%: 0,05 a 0,18; 9 ensayos, 1057 participantes) y 0,27 cm más en la altura (DM 0,27 cm; IC del 95%: 0,07 a 0,48; nueve ensayos, 1463 participantes) que los que no recibieron suplementos. Los suplementos nutricionales en general fueron más efectivos en los niños más pequeños (menores de dos años de edad) y en los que presentaban más desnutrición. En los niños con DAM, la provisión de alimentos especialmente elaborados mejoró el peso, las puntuaciones z del peso para la altura y otros resultados clave como la tasa de recuperación (en un 29%), y también redujo el número de participantes que abandonaron el tratamiento (en un 70%). En los PIBM, las comidas escolares parecieron dar lugar a beneficios pequeños para los niños, incluidas las mejorías en las puntuaciones z del peso, en especial en los niños de países de menores ingresos, las puntuaciones z de la altura, las pruebas cognitivas o del cociente de inteligencia y el rendimiento en matemáticas y al deletrear.
La alimentación complementaria en adultos con pruebas positivas para el VIH aumentó el aporte energético y proteico diario en comparación con la orientación nutricional sola. La administración de suplementos dio lugar a una mejoría inicial en el aumento de peso o el índice de masa corporal, pero no pareció otorgar un beneficio a largo plazo.
En adultos con tuberculosis, un ensayo pequeño encontró un beneficio significativo en la finalización del tratamiento y la tasa de conversión del esputo. También hubo beneficios significativos, pero moderados en cuanto al aumento de peso (hasta 2,60 kg) durante la tuberculosis activa.
El único estudio incluido en la revisión de la enfermedad de Alzheimer halló que la administración de suplementos nutricionales orales diarios durante tres meses mejoró los resultados nutricionales en el grupo de intervención.
Hubo poca o ninguna evidencia con respecto a la calidad de vida de las personas, la adherencia al tratamiento, la asistencia al consultorio o los costes de los programas de alimentación complementaria.
Conclusiones de los autores
Si se considera la base de evidencia actual incluida, los efectos de los suplementos nutricionales son moderados en el mejor de los casos, con evidencia inconsistente y limitada sobre la mortalidad. Los ensayos reflejados en las revisiones principalmente informaron resultados a corto plazo y a través de la totalidad de la bibliografía de los ensayos de la administración de suplementos parece que los resultados importantes, como la calidad de vida y el coste de los programas, no se informan de manera sistemática ni resumida.
Resumen en términos sencillos
Alimentación complementaria para grupos de personas con acceso limitado a los alimentos, en situación de vulnerabilidad y con desnutrición
¿Cuál era el objetivo de esta revisión?
Resumir el efecto de la alimentación complementaria en las poblaciones con acceso limitado a los alimentos, en situación de vulnerabilidad y con desnutrición. Los autores del resumen encontraron ocho revisiones sistemáticas que examinaron la alimentación complementaria en diversas poblaciones.
Mensajes clave
A través de una variedad de poblaciones en situación de vulnerabilidad, los programas de alimentación complementaria a veces muestran un beneficio moderado en los resultados nutricionales. En unos pocos estudios que examinaron la mortalidad (muerte), los efectos fueron pequeños o estuvieron ausentes, y la investigación consideró principalmente los efectos a corto plazo.
¿Qué se estudió en la revisión?
La alimentación complementaria incluye la provisión de alimentos adicionales a las personas o las familias además del régimen alimentario domiciliario y se ha utilizado en poblaciones con acceso limitado a los alimentos (acceso limitado a alimentos adecuados y nutritivos) y en situación de vulnerabilidad (que incluyen mujeres y niños pequeños; niños en edad escolar; pacientes con enfermedades como tuberculosis, VIH y enfermedad de Alzheimer; y personas de edad avanzada) para mejorar su salud y su calidad de vida.
¿Cuáles son los principales resultados de la revisión?
La evidencia presentada aquí está actualizada hasta enero 2017. Se encontraron ocho revisiones sistemáticas para incluir en este resumen. Estas revisiones incluyeron 95 estudios (con hasta 7940 adultos y más de 12 595 niños en unos pocos estudios). La mayoría de los estudios incluidos tuvo una duración desde seis semanas a dos años y sólo tres estudios realizaron el seguimiento de los participantes durante períodos más largos (hasta 17 años). En estas revisiones, hubo una gama amplia de diferentes tipos de alimentación complementaria administrada a grupos en situación de vulnerabilidad durante diferentes períodos, y a menudo en combinación con vitaminas o minerales.
En el embarazo, se encontró que los suplementos energéticos y proteicos equilibrados (es decir, la provisión de cantidades adecuadas de energía y nutrientes, en este caso proteínas) pueden haber reducido la tasa de mortinato (muerte o pérdida de un recién nacido antes o durante el parto), mejoraron el peso del neonato al nacer y redujeron el riesgo de que el neonato sea pequeño para la edad gestacional al nacer (neonatos que son más pequeños que lo esperado). No se observaron beneficios a largo plazo en los niños en cuanto al crecimiento y al desarrollo cognitivo (intelectual) (aunque muy escasos estudios informaron los efectos a largo plazo). Los suplementos con alto contenido de proteína (que contienen cantidades mayores de proteína) se asociaron con riesgos y efectos perjudiciales (mayor riesgo de recién nacidos pequeños para la edad gestacional).
Se encontraron efectos variados de la alimentación complementaria sobre el crecimiento en los niños. En los niños menores de cinco años de edad de países de ingresos bajos y medianos, la alimentación complementaria tuvo una repercusión pequeña sobre el crecimiento de los niños. Se observaron algunos beneficios en cuanto al aumento en el peso y la altura, en especial en los niños más pequeños (menores de dos años de edad) y en los que presentaban más desnutrición. Se pudo observar algún beneficio en los niños con desnutrición aguda moderada en cuanto al aumento de peso, otros factores de crecimiento y la tasa de recuperación. Las comidas escolares parecieron dar lugar a varios beneficios pequeños en los niños en edad escolar (que incluyeron mejorías en el peso, la altura, las pruebas de inteligencia y el rendimiento en matemáticas y al deletrear).
La alimentación complementaria en adultos con pruebas positivas para el VIH aumentó el aporte energético y proteico diario y dio lugar a una mejoría temprana en el aumento de peso o el índice de masa corporal (medida que indica si la persona tiene sobrepeso o bajo peso), o ambos, pero no pareció dar lugar a beneficios a largo plazo (aunque pocos estudios informaron los efectos a largo plazo). En los adultos con tuberculosis (enfermedad pulmonar infecciosa grave), se observaron beneficios pequeños en cuanto al aumento de peso durante la tuberculosis activa.
En la enfermedad de Alzheimer (un tipo de demencia), la provisión de un suplemento nutricional oral diario durante tres meses mejoró los resultados nutricionales (como el peso y el aporte energético).
Hubo poca o ninguna evidencia disponible con respecto a la calidad de vida de las personas, la adherencia al tratamiento, la asistencia al consultorio o los costes de los programas de alimentación complementaria.
Conclusiones de los autores
Antecedentes
Un régimen alimentario adecuado que incluye los macro y micronutrientes necesarios ayuda a asegurar el crecimiento del ser humano, el desarrollo físico y cognitivo y un sistema inmunitario sano. Lo que las personas necesitan en su régimen alimentario varía según la edad, el sexo, la actividad física y el estado de salud (Mahan 2011). El acceso seguro a los alimentos se define como una situación en la cual "todas las personas, en todo momento, tienen acceso físico y económico a alimentos suficientes, seguros y nutritivos para satisfacer sus necesidades alimentarias y las preferencias alimentarias para una vida activa y saludable" (FAO 1996; FAO 2010). La definición de la Food and Agriculture Organization (FAO), refuerza la naturaleza multidimensional y la complejidad del acceso seguro a los alimentos, que incluye la disponibilidad de alimentos, el acceso económico y físico a los alimentos, la utilización de los alimentos y la estabilidad de los suministros con el transcurso del tiempo (FAO 1996; FAO 2013). El acceso seguro a los alimentos es un requisito previo a la nutrición adecuada. Otros factores, incluida la alimentación de los niños y las prácticas de atención, la elección de los alimentos, el conocimiento y el interés en cuanto a la preparación de los alimentos, el agua y un ambiente sanitario adecuado, así como el acceso a la asistencia sanitaria y el estado de salud de la persona, también desempeñan una función importante en si el acceso a los alimentos se traduce en el consumo de un régimen alimentario adecuado y en último término en una nutrición, salud y bienestar adecuados.
La desnutrición, que afecta a una de cada tres personas en todo el mundo, se observa en varias formas que no sólo afectan la salud y el bienestar de las personas, sino que también imponen una gran carga en las familias, las comunidades y los estados (FAO 2017; FAO y WHO 2014; IFPRI 2016). La "carga triple" de la desnutrición incluye la desnutrición, el sobrepeso/obesidad y las deficiencias de micronutrientes, y estas formas pueden coexistir en la misma persona, vivienda y país (FAO 2017). Según los cálculos de la FAO, la prevalencia de la desnutrición sigue siendo alta a pesar de los suministros alimentarios adecuados y el considerable progreso en cuanto a la reducción del hambre en algunas regiones. Más de 795 000 000 de personas todavía se presentaron a la consulta con niveles crónicamente inadecuados de aporte energético alimentario entre 2014 y 2016 (FAO 2015; Sundaram 2015), y las mujeres y los niños son especialmente vulnerables. En 2016 el retraso del crecimiento afectó a un 22,9% o a 154 800 000 de niños y la emaciación continuó amenazando las vidas de un 7,7% o casi 52 000 000 de niños menores de cinco años de edad a nivel global (UNICEF/WHO/World Bank Group 2017). Además, se informó que 108 000 000 de personas a nivel global en 2016 se estaban enfrentando a un acceso limitado a los alimentos a nivel de una crisis o peor, lo que representa un aumento del 35% en comparación con las cifras de 2015 (FSIN 2017). El acceso seguro a los alimentos en el contexto del cambio climático se ha destacado como un riesgo significativo en el Global Risk Report de 2016 (WEF 2016). Los patrones climáticos y el cambio climático podrían amenazar la producción agrícola y el acceso seguro a los alimentos a través de las geografías; el riesgo a un acceso seguro a los alimentos es especialmente grande debido a que la agricultura ya está esforzándose por satisfacer una demanda en rápido crecimiento a partir de una base de recursos finita (WEF 2016).
Se le ha dado prioridad a la posibilidad de acabar con el hambre, al logro de un acceso seguro a los alimentos y a la mejoría en la nutrición como pasos clave hacia un desarrollo sostenible (UN 2016). La comunidad internacional ha reconocido afortunadamente estos retos. La Agenda for Sustainable Development de 2030 estipula una visión sobre cómo se pueden combinar los objetivos múltiples para definir vías de desarrollo nuevas y sostenibles. El segundo Sustainable Development Goal (SDG 2) procura acabar con el hambre, lograr un acceso seguro a los alimentos y a la mejoría en la nutrición y promover una agricultura sostenible en simultáneo para 2030 (FAO 2017; UN 2015a). El 1 de abril de 2016 la General Assembly de las Naciones Unidas aprobó una resolución que proclama un United Nations Decade of Action on Nutrition de 2016 a 2025 (UN 2015b). Este Decade of Action procura movilizar una acción intensificada para acabar con el hambre y erradicar todas las formas de desnutrición en todo el mundo y garantizar el acceso universal a regímenes alimentarios más saludables y más sostenibles para todos (WHO 2016). Además, y en el contexto local, un compromiso político fuerte (incluida la posibilidad de colocar la seguridad del acceso a los alimentos y la nutrición en la parte superior del programa político y crear un contexto favorable) es esencial para la reducción del hambre, que requiere un enfoque integrado, que incluya programas nutricionales específicos.
Muchos factores afectan la vulnerabilidad a la desnutrición y un acceso limitado a los alimentos, y el conocido marco conceptual de la United Nations Children's Fund (UNICEF) indica que las causas de la desnutrición son multisectoriales, teniendo en cuenta los alimentos, la salud y las prácticas de atención. Las causas se categorizan como inmediatas (consumo inadecuado de alimentos y enfermedades), subyacentes (acceso seguro deficiente a los alimentos del hogar, atención materna e infantil inadecuada, acceso limitado a los servicios básicos de salud y un ambiente insalubre [con acceso limitado a agua limpia y a la eliminación segura de desechos]) y básicas (pobreza y falta de recursos), donde los factores a un nivel influyen en otros niveles (UNICEF 1990). Es importante que los responsables de las políticas y los líderes de la comunidad tengan en cuenta las causas de la desnutrición al planificar y dar prioridad a las intervenciones de salud y nutricionales (Black 2013). Como tal, las causas múltiples de la desnutrición requieren un enfoque multisectorial, que incluya enfoques tanto específicos a la nutrición como sensibles a la nutrición (salud; educación básica; agricultura, forestación y pesca; y desarrollo social a nivel local, provincial y nacional) (Bhutta 2013; Garrett 2011; IFPRI 2014; Ruel 2013). Los programas de alimentación complementaria, dirigidos a las viviendas y las personas en situación de vulnerabilidad, no son sino un enfoque para considerar los temas complejos que rodean el acceso seguro a los alimentos y la desnutrición. Estos programas, operados por los gobiernos y los organismos, pueden ser costosos y su administración complicada. Este resumen procura resumir la evidencia de las revisiones sistemáticas existentes sobre los efectos en los grupos destinatarios establecidos.
Descripción de la afección
En 2014; 805 000 000 de personas en el mundo experimentaron hambre crónica, y no tenían alimentos suficientes para asegurar un estado nutricional óptimo y llevar una vida activa y saludable (FAO 2014). Un billón de personas se consideraron vulnerables al "hambre oculta" como resultado de las deficiencias de micronutrientes (MI 2009). Muchos factores interrelacionados afectan la vulnerabilidad al acceso limitado a los alimentos, incluida la pobreza, la falta de tierra y el conflicto, así como otros factores como el sexo, las enfermedades y la edad (FAO 2008).
El acceso seguro a los alimentos en este resumen se definió como una situación en la cual "todas las personas, en todo momento, tienen acceso físico y económico a alimentos suficientes, seguros y nutritivos para satisfacer sus necesidades alimentarias y sus preferencias alimentarias para una vida activa y saludable" (FAO 1996; FAO 2010). Por lo tanto, el acceso limitado a los alimentos existe cuando las personas no tienen acceso físico, social o económico adecuado a los alimentos.
Las poblaciones particularmente en riesgo y en situación de vulnerabilidad en cuanto al acceso limitado a los alimentos incluyen:
-
mujeres y niños pequeños;
-
niños en edad escolar;
-
pacientes con enfermedades infecciosas, en particular tuberculosis y VIH; y
-
poblaciones discapacitadas y mayores.
La desnutrición es un término amplio que abarca tanto la subnutrición como la sobrenutrición. Para este resumen, la desnutrición se refiere a la subnutrición. La desnutrición puede ser resultado de una falta de macronutrientes (carbohidratos, proteínas, grasas), micronutrientes (vitaminas y minerales) o ambos.
El interés de este resumen se centra en los grupos de bajos ingresos y las poblaciones de los países de ingresos bajos y medios (PIBM) en donde el acceso seguro a los alimentos para las familias y las comunidades puede ser amenazado, con repercusiones consiguientes sobre el estado nutricional, en particular en los grupos de la población que están en situación de vulnerabilidad a la desnutrición como las mujeres embarazadas, los niños pequeños, las personas mayores o los pacientes con enfermedades crónicas. La ayuda humanitaria (definida como "ayuda y acción diseñada para salvar vidas, aliviar el sufrimiento y mantener y proteger la dignidad humana durante y después de las crisis producidas por el hombre y los desastres naturales"), incluida la ayuda alimentaria de emergencia a corto plazo, es de importancia crucial (Global Humanitarian Assistance 2014), pero no es el tema de este resumen.
Descripción de las intervenciones
En este resumen, la alimentación complementaria se definió como la provisión de alimentos adicionales a las personas o las familias más allá de la ración normal de sus regímenes alimentarios domiciliarios (Beaton 1982). La alimentación complementaria se usa tanto en situaciones de emergencia como en las que no son de emergencia para considerar el hambre a corto plazo, la escasez de alimentos a más largo plazo y mejorar el estado nutricional (o prevenir el deterioro nutricional) de poblaciones específicas. A veces las autoridades proporcionan suplementos nutricionales para aumentar el uso de los servicios sanitarios, la adherencia a los regímenes de tratamiento o la asistencia (y el rendimiento) a la escuela. En los desastres humanitarios, la asistencia alimentaria procura aliviar la escasez absoluta de alimentos (aunque este resumen no trata con el uso de alimentos en estas circunstancias). Sin embargo, aún hay muchas preguntas acerca de la coste‐efectividad de los programas de alimentación complementaria, su diseño y la combinación apropiada de las actividades complementarias para lograr los resultados concebidos (Morris 2008).
La alimentación complementaria se puede realizar de dos maneras (ver Figura 1):
-
alimentación complementaria general, que procura prevenir la desnutrición o su progresión en poblaciones con acceso limitado a los alimentos, en las que el objetivo se basa en saber que la población es de alto riesgo, y los alimentos se proporcionan a la población en riesgo completa sin una detección previa (UNHCR/WFP 1999); y
-
la alimentación complementaria específica, dirigida a personas seleccionadas que están en riesgo; el tratamiento de la desnutrición leve o moderada detectada mediante la detección de las poblaciones en riesgo y la provisión de suplementos sólo a las personas que se encuentran por debajo de un umbral predeterminado del estado nutricional.
Este resumen incluyó ambas categorías de alimentación complementaria.
Los alimentos a veces se usan para inducir a las personas a comportarse de una manera particular; por ejemplo, la provisión de alimentos a los pacientes con tuberculosis para asegurar que asistan a los consultorios para el tratamiento, o la provisión de alimentos a los niños en la escuela para ayudar a mejorar la asistencia escolar (Devereux 2018). Aunque este resumen no consideró la administración de alimentos en estas circunstancias, cuando los alimentos se podrían haber utilizado para mejorar la nutrición y para mejorar la adherencia al tratamiento (por ejemplo, en pacientes con tuberculosis), los resultados de la adherencia se examinaron como un resultado secundario.
Los suplementos nutricionales se definieron como alimentos adicionales al régimen alimentario normal. Los suplementos nutricionales también incluyen especialmente alimentos elaborados (por ejemplo, combinación de alimentos fortificados) en preparados listos para comer o molidos, que son modificados en cuanto a su densidad energética, las proteínas, las grasas o la composición de micronutrientes para ayudar a satisfacer las necesidades nutricionales de poblaciones específicas (WHO 2012). Los suplementos nutricionales no tienen como objetivo ser la única fuente de nutrientes en una población determinada (WHO 2012). Son diferentes de los suplementos alimentarios que se refieren a los suplementos de vitaminas y minerales en formas de dosis única (como cápsulas, comprimidos, polvos o soluciones), que no son relevantes a este resumen (WHO 2012). La fortificación de los alimentos (con la intención de aumentar el contenido de micronutrientes de la dieta en general) así como las intervenciones de nutrición enteral y parenteral tampoco son parte de este resumen.
De qué manera podría funcionar la intervención
La alimentación complementaria puede tener beneficios directos en la nutrición y la salud (Figura 1). También puede contribuir a una mayor utilización de servicios, con efectos secundarios sobre la mejoría en la salud relacionada con la mayor aceptación de los servicios. Además, puede contribuir a las metas sociales, como los suplementos nutricionales administrados en las escuelas para mejorar la asistencia escolar.
La alimentación complementaria puede tener efectos negativos al aumentar la dependencia, crear expectativas de donativos de alimentos en los consultorios y los servicios, y al repercutir de forma negativa en la asistencia a los consultorios cuando se interrumpe. También es costosa, y necesita buenos sistemas de manejo para asegurar la entrega y disminuir la administración a las personas para las que no fue concebida. Además, los aspectos de seguridad de los alimentos son de importancia crucial para una provisión segura, en especial si se promueve la producción local.
Por qué es importante realizar esta revisión global
Aunque no hay cifras exactas disponibles, una gran parte del financiamiento de asistencia para el desarrollo asignado a la alimentación y la nutrición se usa para los programas de alimentación complementaria, incluida la asistencia de urgencia y la ayuda alimentaria (Morris 2008). Por lo tanto, es importante saber si es efectiva. Además, es importante tratar de identificar las intervenciones más exitosas (o la combinación de las mismas) para la replicación, así como los criterios para mejorar la coste‐efectividad y la eficiencia de las intervenciones.
El público destinatario para este resumen incluye los responsables de políticas y los ejecutores del programa que trabajan en áreas relacionas al acceso seguro a los alimentos y a la nutrición de salud pública. Los encargados del desarrollo también pueden usar este resumen para informar el diseño de las solicitudes de investigación y las propuestas de programas y para ayudar con la evaluación de los programas propuestos de alimentación complementaria actuales. Los médicos que trabajan en regiones en las que existe un acceso limitado a los alimentos o en las que la desnutrición es común también encontrarán que este resumen es útil en sus esfuerzos por defender estrategias de salud pública preventivas y promocionales efectivas en función de los costes.
Objetivos
-
Resumir la evidencia obtenida de revisiones sistemáticas sobre la alimentación complementaria para poblaciones con acceso limitado a los alimentos, en situación de vulnerabilidad y con desnutrición, incluidos los niños menores de cinco años de edad, los niños en edad escolar, las embarazadas y las mujeres que amamantan, los pacientes con VIH o tuberculosis (o ambos) y las poblaciones de edad avanzada.
-
Describir y explorar los efectos de la alimentación complementaria en los individuos de estos grupos, y describir el rango de resultados entre las revisiones y el rango de efectos en los diferentes grupos. Se examinaron las posibles influencias en los efectos entre las revisiones, incluido el estado nutricional inicial y las comorbilidades.
Métodos
Criterios de inclusión de revisiones
Tipos de estudios
Revisiones sistemáticas publicadas (sin restricción en cuanto a la fecha de la última búsqueda) de la alimentación complementaria en grupos en situación de vulnerabilidad.
Los criterios de inclusión para las revisiones no Cochrane fueron:
-
objetivos predeterminados;
-
criterios de elegibilidad predeterminados;
-
búsqueda realizada en al menos dos fuentes de datos, una de las cuales debía haber sido una base de datos electrónica; y
-
extracción de datos y evaluaciones del "riesgo de sesgo" realizadas de forma independiente y por duplicado por los autores de la revisión.
Tipos de participantes
Revisiones sistemáticas que trataban sobre poblaciones en situación de vulnerabilidad (con acceso limitado a los alimentos y con desnutrición), centradas en poblaciones con acceso limitado a los alimentos o en personas identificadas como con desnutrición en estas poblaciones (niños, niños en edad escolar, mujeres embarazadas y que amamantan, pacientes con VIH o tuberculosis [o ambos] y personas mayores). Se incluyeron revisiones que proporcionaban algunas o todas estas condiciones (vulnerabilidad, acceso limitado a los alimentos y desnutrición). Se excluyó a los grupos para los cuales se requirió atención terapéutica especializada (como neonatos prematuros y de bajo peso al nacer).
Tipos de intervenciones
Revisiones que evaluaban la alimentación complementaria comunitaria en grupos en situación de vulnerabilidad (según lo definido en la sección Descripción de las intervenciones).
Los programas de alimentación complementaria comunitaria fueron los que proporcionaban alimentos a las poblaciones o a personas de forma ambulatoria en un contexto no clínico. Por lo tanto, la alimentación complementaria podía tener lugar en el domicilio, en un centro supervisado de alimentación o en otros lugares adaptados para esta finalidad (por ejemplo, centros de asistencia sanitaria y guarderías).
Los programas de alimentación complementaria fueron los que comprendían criterios establecidos que se aplicaron a una población para determinar la elegibilidad para los suplementos nutricionales y los alimentos se proporcionan a dicha población. Los suplementos nutricionales eran macronutrientes (régimen alimentario equilibrado o regímenes alimentarios/alimentos con alto contenido de proteínas, de carbohidratos, o de grasas) administrados como un suplemento además del régimen alimentario habitual (no un reemplazo total del régimen alimentario). Los suplementos nutricionales podían contener micronutrientes agregados (vitaminas y minerales), sin embargo, se excluyeron las revisiones de los micronutrientes solamente. Los suplementos alimentarios se debían haber administrado por vía oral. Las opciones de alimentación complementaria incluyeron la administración de alimentos adicionales, alimentos fortificados, alimentos especialmente elaborados (alimentos combinados fortificados como la mezcla de maíz y soja, alimentos listos para consumir como pastas, barras comprimidas o galletas), o suplementos alimentarios complementarios (como suplementos alimentarios complementarios en polvo que contienen una combinación de micronutrientes, proteína, aminoácidos y enzimas; o complementos nutricionales basados en lípidos (CNL) (120 kcal/día a 250 kcal/día), que con frecuencia contienen leche en polvo, aceite vegetal de alta calidad, pasta de maní, azúcar y micronutrientes (De Pee 2009)). Los alimentos terapéuticos listos para el consumo se usan con frecuencia para tratar desnutrición aguda grave (DAG), pero en algunos casos estas recetas se modifican para el uso en la desnutrición aguda moderada (DAM) y por lo tanto podrían haber sido relevantes a este resumen (De Pee 2009).
Se excluyeron:
-
las revisiones que informaban los efectos de la alimentación complementaria en el ámbito de refugiados u hospitales (después de lesiones o cirugías u otros trastornos médicos agudos);
-
suplementos de vitamina y minerales, productos para la alimentación con sonda enteral o productos para la alimentación parenteral y
-
alimentos terapéuticos para el tratamiento de la desnutrición grave.
Los grupos de comparación fueron los que no recibieron el suplemento o los que recibieron un suplemento diferente.
Tipos de resultados
-
Muerte (incluida la mortalidad perinatal).*
-
Enfermedad (o resultados relacionados con las enfermedades).
-
Crecimiento en los niños (definido como un cambio gradual en comparación con el peso o la altura inicial).
-
Estado nutricional de los niños (evaluado mediante otra antropometría, marcadores bioquímicos e ingesta alimentaria) al final del seguimiento.*
-
Estado nutricional de los adultos (evaluado mediante antropometría, marcadores bioquímicos e ingesta alimentaria) al final del seguimiento.*
-
Adherencia al tratamiento o asistencia al consultorio.
-
Asistencia a la escuela, pruebas cognitivas y logro educacional.*
-
Costes para el proveedor.
-
Costes del propio bolsillo para las personas que recibieron alimentación complementaria.
*Resultados principales
Métodos de búsqueda para la identificación de las revisiones
We first searched the electronic sources listed below on 9 July 2013 and updated the searches on 29 January 2017.
-
Cochrane Database of Systematic Reviews (CDSR; 2017, Issue 1) in the Cochrane Library.
-
MEDLINE Ovid (searched from 1946).
-
MEDLINE In‐Process and Other Non‐Indexed Citations Ovid.
-
MEDLINE Epub Ahead of Print Ovid.
-
Embase Ovid (searched from 1980).
-
Database of Abstracts of Reviews of Effects (DARE; 2015, Issue 2. Final issue), part of the Cochrane Library.
-
Health Technology Assessment Database (HTAD; 2016, Issue 4), part of the Cochrane Library.
-
Campbell Collaboration Online Library of Systematic Reviews (www.campbellcollaboration.org/library.html).
-
Virtual Health Library (bvsalud.org/en).
-
Database of Promoting Health Effectiveness Reviews (DoPHER; eppi.ioe.ac.uk/webdatabases4/Intro.aspx?ID=9).
-
3ie Database of Systematic Reviews (www.3ieimpact.org/en/evidence/systematic‐reviews).
-
PROSPERO (www.crd.york.ac.uk/prospero).
The core search strategy consisted of two concepts: supplementary feeding AND systematic reviews. Each concept was described using controlled vocabulary terms and free‐text terms. The systematic review filter for MEDLINE was adapted from the Scottish Intercollegiate Guidelines Network (SIGN; www.sign.ac.uk/search‐filters.html). We adapted the search terms for each database (Appendix 1). There were no date or language restrictions.
Obtención y análisis de los datos
Selección de las revisiones
For the July 2013 search, one overview author (JV) and one overview reviewer (LN; see Acknowledgements) independently screened titles and abstracts of records yielded by the search for relevance, and rated them as 'for exclusion', 'for inclusion' or 'potentially eligible'. Next, they obtained the full texts of those systematic reviews judged as 'for inclusion' or 'potentially eligible' and independently assessed them against the inclusion criteria (Criteria for considering reviews for inclusion). Two overview authors (JV; PG) and one overview reviewer (LN) resolved any differences of opinion as regards review selection by discussion until a consensus was reached.
For the January 2017 search, one overview author (NM) and two overview reviewers (NH; RW; see Acknowledgements) screened titles, abstracts and full texts using the same methods described above. Two overview authors (NM; JV) and two overview reviewers (NH; RW) resolved differences of opinion by discussion until a consensus was reached. For updated Cochrane Reviews, we included only the most recent publication.
The PRISMA flow diagram in Figure 2 illustrates the study selection process.
Extracción y manejo de los datos
For the July 2013 search, one overview author (JV) extracted the data and one overview reviewer (LN) checked them for accuracy; one overview author (NM) and two overview reviewers (NH; RW) performed this for the January 2017 search. Reviewers resolved any disagreements by discussion, with assistance from the rest of the overview author team, as necessary.
We used a data collection form that was specifically designed by the overview author team to collect data on the key features of the systematic reviews such as objectives; inclusion and exclusion criteria; and information about participants, interventions and comparisons. We assessed whether each included review was up to date (and reported the date of the last search). We described the reviews in relation to background vulnerability, including equity‐related aspects and study characteristics, as available. In this regard, we referred to the relevant economies of the included trials (i.e. LMIC versus high‐income countries (HIC)), any summary of socioeconomic status of participants across trials, setting of the trials (community versus hospital or other) and if methods for targeting the interventions were used.
We present the key characteristics of each included systematic review in 'Characteristics of included systematic reviews' tables, the interventions used in the systematic reviews in a 'Summary of interventions' table, and details of the interventions used in separate 'Details of interventions' tables for each included systematic review. In the case of discordant results of included reviews, we presented all; where necessary, we contacted the authors of the primary reviews for clarification and consulted with a biostatistician experienced in meta‐analysis on statistical issues. We also summarised and presented the target groups of each systematic review by outcome in a 'Results matrix', and the results for each target group across included reviews in tables of comparisons by outcome for which data were available.
Evaluación de la calidad metodológica de las revisiones incluidas
Quality of included reviews
For the July 2013 search, one overview author (JV) and one overview reviewer (LN) independently assessed the quality of the included systematic reviews using AMSTAR: A MeaSurement Tool to Assess Systematic Reviews (Shea 2007a; Shea 2007b; Shea 2009); one overview author (NM) and two overview reviewers (NH; RW) performed this for the January 2017 search. The overview author and reviewer(s) resolved any differences by discussion. We included other members of the author team in the discussion (if needed) until we reached a consensus.
AMSTAR assesses the degree to which review methods avoided bias by evaluating the methods against 11 distinct criteria (listed below).
-
Use of an a priori design.
-
Duplicate study selection and data extraction.
-
Comprehensive searching of the literature.
-
Use of publication status as an exclusion criterion.
-
Provision of (included and excluded) studies.
-
Provision of characteristics of included studies.
-
Assessment of methodological quality of included studies.
-
Appropriate use of quality of included studies in formulating conclusions.
-
Appropriate methods for combining results of studies.
-
Assessment of publication bias.
-
Conflict of interest (both review and included studies) stated.
Review authors rated each AMSTAR item as yes (clearly done), no (clearly not done), cannot answer or not applicable, based on the published and included systematic reviews.
We presented the review quality data in two ways. First, as a narrative report across all studies against the 11 criteria above. Second, using the standard scoring system provided by AMSTAR to classify reviews into three categories: high quality (those achieving scores between eight and 11); medium quality (those achieving scores between four and seven) and low quality (those achieving scores between zero and three). We identified and discussed differences in quality between reviews, and used the quality assessment to interpret the results of reviews when synthesised in this overview. We summarised the AMSTAR scores of each included systematic review in a table.
Certainty of evidence from primary studies in included reviews
We used the GRADEpro 'Summary of findings' tables from each review (if reported in the individual reviews) to indicate the certainty of the evidence for the main comparisons.
Síntesis de los datos
We used a narrative approach to summarise the data, and included: the range of vulnerability assessment criteria; the range of types of food supplements given and the effects of supplementation. Because the different reviews were based on different population groups and the interventions were basically the same across reviews (considering variations in type, quantity and length of time that a food or supplement was provided), we did not explore indirect comparisons. For analyses with differing durations of supplementation and dietary compositions of supplements, we described this within and between reviews. We made reference to statistical heterogeneity indirectly via certainty of evidence ratings for outcomes in 'Summary of findings' tables, as reported for each outcome. We summarised the findings in additional tables ('Characteristics of included systematic reviews', 'Summary of interventions' and 'Details of interventions' per systematic review tables; 'Results matrix'; and tables of comparisons by outcome for which data were available). We also used forest plots to graphically display selected results.
Results
The searches identified 18,069 titles and abstracts. After initial screening of titles and abstracts, we retrieved 190 full texts and assessed them for eligibility against our inclusion criteria (Criteria for considering reviews for inclusion). Of these, we excluded 104 texts that clearly did not meet our inclusion criteria, and formally excluded a further 66 with reasons (see Table 1). We included eight systematic reviews. The reviews by Kramer 1996a, Kramer 1996b, and Kramer 1996c were all withdrawn and replaced by Ota 2015 and are listed in Table 1. We also identified nine potentially relevant protocols for reviews that are currently under way and have listed these in Appendix 2, and a further three reviews that are awaiting assessment (we were unable to locate the published reports for two of these reviews at the time of our last search, despite exhaustive efforts, and the third was published after the date of search of this review) (Appendix 3). See Figure 2.
| Review author and date | Reason for exclusion |
| Methods: stated that Cochrane methods were used but did not specify that 2 reviewers assessed risk of bias. Participants: majority of participants in care homes (1 in hospital); only 3/43 in community | |
| Methods: no predetermined eligibility criteria Participants: mainly residing in long‐term care establishments (75%) | |
| Methods: no 'Risk of bias' assessment | |
| Participants: 5/41 in community, including neurology outpatients and those enrolled at hospital discharge (Protocol published as Gibbs 2012) | |
| Participants: hospital inpatients | |
| Methods: no mention that 'Risk of bias' assessment performed in duplicate. | |
| Methods: only 1 database used; single abstract screening; methods for 'Risk of bias' assessment and data extraction not reported. | |
| Methods: only 1 reviewer assessed trials for inclusion, extracted data and assessed trial quality. | |
| Methods: single screening of abstracts and single data extraction. | |
| Methods: no 'Risk of bias' assessment | |
| Methods: only 1 data source; no 'Risk of bias' assessment | |
| Methods: only 1 data source; no 'Risk of bias' assessment | |
| Methods: no mention that abstract screening or data extraction performed in duplicate. | |
| Methods: no details on screening; no 'Risk of bias' assessment | |
| Methods: no mention that data extraction performed in duplicate. | |
| Methods: did not report methods; no 'Risk of bias' assessment | |
| Methods: no 'Risk of bias' assessment | |
| Methods: no 'Risk of bias' assessment; based on 2 previous reviews | |
| Methods: search terms and databases not listed; no mention that screening or data extraction performed in duplicate. | |
| Methods: AMSTAR score of 5; no mention that data extraction or 'Risk of bias' assessment performed in duplicate, though stated Cochrane methods were used to assess risk of bias. | |
| Methods: no mention that data extraction and 'Risk of bias' assessment performed in duplicate. | |
| Methods: methods not described | |
| Methods: 2 databases searched; review methods not reported | |
| Methods: scoping review Methods: 1 reviewer screened; no 'Risk of bias' assessment | |
| Methods: no mention that screening or data extraction performed in duplicate; no 'Risk of bias' assessment | |
| Methods: not described | |
| Methods: second independent reviewer extracted data from half of the studies. Intervention: dietary intervention, macronutrients, micronutrients; data not reported separately; no information reported on intervention. | |
| Methods: no mention that screening or data extraction performed in duplicate. No 'Risk of bias' assessment. | |
| Methods: no mention that data extraction performed in duplicate. Participants: included both well‐nourished and malnourished participants; mostly elderly (23 participants) with a range of acute and chronic conditions, including those with fractures (4 participants), renal disease (2 participants), cancer (5 participants) and respiratory disease (4 participants) Setting: community and hospital settings | |
| Methods: only 1 database searched; no mention that screening or 'Risk of bias' assessment performed in duplicate. | |
| Methods: no mention that screening or 'Risk of bias' assessment performed in duplicate but did mention that data extraction performed in duplicate. Quote: "Even though we included terms like 'supplementary food' and 'supplementary feed' in our literature search but only those studies were included where the term supplementary food was used for introduction of additional food to a breastfed child at the age of 6 months i.e. complementary feeding." | |
| Methods: only 1 database searched; no mention that screening, data extraction or 'Risk of bias' assessment performed in duplicate. | |
| Intervention: only 7/41 meal supplementation Participants: only 2/41 in community | |
| Status: withdrawn and replaced by Ota 2015 Intervention: nutrition advice (not food or supplements) | |
| Status: withdrawn and replaced by Ota 2015 | |
| Status: withdrawn and replaced by Ota 2015 | |
| Intervention: majority of included studies had micronutrient interventions. | |
| Methods: method of data extraction and screening not described. | |
| Methods: screening and data extraction performed in duplicate but no mentioned that 'Risk of bias' assessment performed in duplicate. | |
| Methods: did not list electronic sources; no mention that screening or data extraction performed in duplicate; no 'Risk of bias' assessment. | |
| Methods: AMSTAR score 7; did not describe that screening, data extraction or 'Risk of bias' assessment performed in duplicate. | |
| Intervention: not community‐based supplementary feeding; strategies/intervention to prevent rickets (4 included studies: 3 micronutrient interventions) Outcomes: few outcomes of interest | |
| Methods: single screening; no 'Risk of bias' assessment | |
| Methods: 'Risk of bias' assessment performed in duplicate but not reported that screening and data extraction performed in duplicate. Participants: acute and tertiary healthcare settings | |
| Methods: 1 database searched | |
| Methods: single screening; second author checked included full texts. | |
| Methods: no mention that 'Risk of bias' assessment performed in duplicate. | |
| Methods: no mention that screening and data extraction performed in duplicate; no 'Risk of bias' assessment. | |
| Methods: screening performed in duplicate but no mentioned that data extraction or 'Risk of bias' assessment performed in duplicate. | |
| Setting: the majority of studies included were in a hospital, long‐care or nursing home setting (66%); only 1 subgroup analysis related to community versus hospital (mortality). | |
| Methods: no mention that 'Risk of bias' assessment performed in duplicate. Participants: hospitalised, residential and aged, care and community dwelling populations (1/6 malnourished studies in community) | |
| Methods: 'Risk of bias' assessment performed in duplicate; no mention that screening or data extraction performed in duplicate. | |
| Methods: 1 database; no mention that double data extraction performed in duplicate; no 'Risk of bias' assessment. Intervention: oral or enteral protein energy supplementation | |
| Type of publication: abstract; no full‐text report available Intervention: mostly micronutrient, but also included balanced protein energy | |
| Methods: data extraction performed in duplicate but no mentioned that screening or 'Risk of bias' assessment performed in duplicate | |
| Methods: single title and abstract screening (Protocol published as Stevens 2013) | |
| Methods: not described | |
| Methods: screening and 'Risk of bias' assessment performed in duplicate but no mention that data extraction performed in duplicate. Participants: community, care homes, rehabilitation/community hospitals Outcome: hospital admissions | |
| Methods: screening and 'Risk of bias' method not described; data extraction by 1 reviewer. | |
| Methods: screening, data extraction and 'Risk of bias' assessment performed by 1 reviewer and checked by a second. Participants: 1/9 studies in community | |
| Type of publication: abstract; full‐text report published as Loveday 2012 and excluded. | |
| Methods: not specified methodology for review; no report of number of reviewers Participants: not specified as vulnerable | |
| Methods: no systematic review methods described other than database search | |
| Methods: no mention that screening, data extraction or 'Risk of bias' assessment performed in duplicate | |
| Methods: screening and data extraction performed by 1 reviewer | |
| Methods: not described other than search |
AMSTAR: A Measurement Tool to Assess Systematic Reviews.
Description of included reviews
This overview of reviews included eight systematic reviews with 128 studies, of which 95 studies were relevant to this overview. All but two reviews included only randomised controlled trials (RCTs): Kristjansson 2007 also included controlled before‐and‐after (CBA) studies and interrupted time series (ITS) studies, and Kristjansson 2015a also included CBA studies. Five RCTs appeared in both Kristjansson 2015a and Sguassero 2012.
Studies included in the reviews were published between 1926 and 2015. The last date of search of these reviews varied between May 2006 (Kristjansson 2007) and February 2016 (Grobler 2016).
The reviews included:
-
pregnant women (Ota 2015);
-
children (aged birth to five years) from LMIC born at term (37 weeks or greater) (Sguassero 2012);
-
disadvantaged infants and young children (aged three months to five years) (Kristjansson 2015a, which was also published as Kristjansson 2015b and Kristjansson 2015c; however, for the purposes of this review we use Kristjansson 2015a);
-
children (aged six to 60 months) in LMIC with MAM (Lazzerini 2013);
-
disadvantaged children and adolescents (aged five to 19 years) attending primary or high school (Kristjansson 2007);
-
HIV‐positive adults and children (Grobler 2013);
-
adults and children with active tuberculosis (with or without HIV) (Grobler 2016); and
-
people with Alzheimer's disease (Droogsma 2014).
The number of participants (relevant to this overview) varied across reviews, ranging from 91 (Droogsma 2014) to 7940 (Ota 2015) adults, and 271 (Grobler 2013) to more than 12,595 (Kristjansson 2007) children. See Table 2 and Table 3 for a summary of the characteristics of included reviews.
| Review | Vulnerability | Last search date | Population | Included studies (relevant to this overview) | Types of studies included | Participants (relevant to this overview) |
| Alzheimer's disease | April 2013 | Community‐dwelling people with Alzheimer's disease | 1 (1) | All RCTs | 91 adults (all relevant) | |
| HIV positive | August 2011 | Adults and children who were HIV positive | 14 (14) | All RCTs | 1725 adults (all relevant) | |
| February 2012 | 271 children (all relevant) | |||||
| TB | February 2016 | Adults and children with active TB (with/without HIV) | 35 (7)a | All RCTs | 7491 adults (986 relevant) | |
| 792 children (none relevant) | ||||||
| Disadvantaged school children | May 2006 | Children and adolescents (aged 5–19 years) attending primary or high school | 18 (18) | 7 RCTs | > 12,595 children (not accurately reported) (all relevant) | |
| 9 CBAs | ||||||
| 2 ITSs | ||||||
| Disadvantaged infants and young children | January 2014 | Infants and children aged 3 months to 5 years | 32 (32) | 21 RCTs (individual and cluster randomised) | 11,602 children (all relevant) | |
| 11 CBAs (individual and cluster randomised) | ||||||
| Children with MAM (< 5 years of age) | October 2012 | Children with MAM (aged 6 to 60 months) in LMIC | 8 (8) | All RCTs (individual and cluster randomised) | 10,037 children (all relevant) | |
| Pregnancy | January 2015 | Pregnant women | 17 (12)b | All RCTs (individual and cluster randomised) onlyc | 9030 adults (7940 relevant) | |
| Children < 5 years of age in LMIC | January 2011 | Children (aged 0–5 years) in LMIC born at term (≥ 37 weeks) | 8 (8) | All RCTs (individual and cluster randomised) onlyc | 1243 children (all relevant) | |
| CBA: controlled before‐and‐after study;ITS: interrupted time series; LMIC: low‐ and middle‐income countries; MAM: moderate acute malnutrition; RCTs: randomised controlled trial; TB: tuberculosis. | ||||||
aOnly seven trials (in adults) assessed macronutrient supplementation.
bOnly 12 trials assessed macronutrient supplementation.
cQuasi‐randomised designs were excluded.
| Review | Vulnerability | Intervention categories (as per the original review) | Duration of intervention | Cointerventions | Associated interventions | Main outcome categories | Length of follow‐up |
| Alzheimer's disease | Oral nutritional supplements (1) | 3 months | — | — | Clinical | 3 months | |
| Nutritional | |||||||
| Biochemical | |||||||
| HIV positive | Supplementary food (2) | 6 weeks to 1 year | Micronutrients | Nutrition counselling | Mortality | 6 weeks to 1 year | |
| Macronutrient formulas providing energy and protein (6) | Anthropometry | ||||||
| Dietary intake | |||||||
| Specific macronutrient supplements (6) | Disease parameters | ||||||
| Adverse events | |||||||
| TB | Supplementary food (5) | 60 days to 6 months (for macronutrient interventions) | Micronutrients | Nutrition counselling | Mortality | 8 weeks to 1 year | |
| Macronutrient formulas providing energy and protein (2) | Anthropometry | ||||||
| Micronutrients (28)a | Disease‐related outcomes | ||||||
| Quality of life | |||||||
| Disadvantaged school children | Supplementary food, snacks and drinks (18) | 20 days to 3 years | — | — | Anthropometry | Not consistently reported | |
| Psychosocial outcomes | |||||||
| Disadvantaged infants and young children | Supplementary food (12) | 3–32 months | Micronutrients | Additional rations for family | Growth | Not consistently reported (up to 8 years) | |
| Cash transfers | Anthropometry | ||||||
| Stimulation | |||||||
| Macronutrient formulas providing energy and protein (20) | Health/nutritional education for mothers | Psychosocial outcomes | |||||
| Healthcare, deworming | Adverse events | ||||||
| Children with MAM (< 5 years of age) | Specially formulated foods, including LNS, blended foods, complementary LNS and blended foods (8) | 8–16 weeks (or upon recovery) | Micronutrients | Nutrition education | Mortality | 8–16 weeks (outcomes in 2 trials: 6 months and 12 months) | |
| Health education | Anthropometry | ||||||
| Medical care | Disease‐related outcomes | ||||||
| Psychosocial stimulation | |||||||
| Pregnancy | Balanced protein energy supplementation (12) | 2.5–9 months + during pregnancy (not consistently reported) | Micronutrients | — | Mortality | Not consistently reported (up to 17 years) | |
| High protein supplementation (1) | Anthropometry | ||||||
| Isocaloric protein supplementation (2) | Neurocognitive development | ||||||
| Nutritional advice (4)a | Adverse events | ||||||
| Children < 5 years of age in LMIC | Supplementary food, snacks and drinks (8) | 2–12 months | Micronutrients | — | Anthropometry | 2–12 months | |
| Adverse events | |||||||
| LMIC: low and middle‐income country; LNS: lipid‐based nutrient supplement; MAM: moderate acute malnutrition; TB: tuberculosis. | |||||||
aExcluded from this overview.
The majority of studies (70 studies; 74%) relevant to this overview were conducted in LMIC. Socioeconomic and food security status of participants were poorly reported, and all but one study in one review were conducted in community settings. See Table 4. Studies in four reviews used a combination of blanket (Kristjansson 2007; Kristjansson 2015a) and targeted (Droogsma 2014; Lazzerini 2013) supplementary feeding approaches. See Table 5.
| Review | Population | Included studies (relevant to this overview) | Economies (relevant to this overview) | SES (relevant to this overview) | Setting | |||
| LMIC | HIC | Economically disadvantaged, including undernourished, nutritionally‐at‐risk, rural | Economically advantaged, including well‐nourished | Community (or outpatient setting) | Hospital inpatients and other | |||
| Community‐dwelling people with Alzheimer's disease | 1 | 0 | 1 | 1 | 0 | 1 | 0 | |
| Adults and children who were HIV positive | 14 | 7 | 7 | NR | NR | 14a | 0 | |
| Adults and children with TB | 35 (7) | 33 (6) | 2 (1) | NR | NR | 24 (6) | 11 (1)b | |
| Disadvantaged school children (aged 5–19 years) | 18 | 9 | 9 | 18 | 0 | 18 | 0 | |
| Disadvantaged infants and young children (aged 3 months to 5 years) | 32 | 29 | 3c | 30 | 2 | 32 | 0 | |
| Children with MAM (< 5 years of age) | 8 | 8 | 0 | 8 | 0 | 8 | 0 | |
| Pregnant women | 17 (12) | 10 (8) | 7 (4) | 7 (6) | 10 (6) | 17 (12) | 0 | |
| Children < 5 years of age in LMIC | 8 | 8 | 0 | 6d | ? | 8 | 0 | |
| ?: unknown; HIC: high‐income country; LMIC: low‐ and middle‐income country; MAM: moderate acute malnutrition; NR: not reported; SES: socioeconomic status. | ||||||||
aIn one study, Rollins 2007, children were included that were treated on an inpatient and outpatient basis.
bOf the seven studies on macronutrients, one study recruited and treated people in a hospital setting (Pérez‐Guzmán 2005).
cOne study included Aboriginal children.
dSix studies included nutritionally‐at‐risk children, whereas in two studies there were no trial entry criteria based on child nutritional status.
eFive studies appeared in both Kristjansson 2015a and Sguassero 2012.
| Systematic review | Population | Supplementary feeding | Intervention categories (as per review) | Intervention summary (number of studies) | |||||
| Supplementary food/drinka | Non‐foodb | Other | |||||||
| No added micronutrients | Added micronutrients | No added micronutrients | Added micronutrients | ||||||
| Adults with Alzheimer's disease | — | T | Oral nutritional supplements (1) | 0 | 0 | 0 | 1 | 0 | |
| Adults who were HIV positive | Children who were HIV positive (3 studies) | B and T | Supplementary food (2) | 0 | 2 | 5 | 7 | 0 | |
| Macronutrient supplements providing energy and protein (6) | |||||||||
| Specific macronutrient supplements (6) | |||||||||
| Adults with TB | Children with TB (3 studies)c | B and T | Supplementary food (5) | 3 | 2 | 1 | 1 | 28 (micronutrients only) | |
| Macronutrient formulas providing energy and protein (2) | |||||||||
| Micronutrients (28)d | |||||||||
| — | Disadvantaged school children (aged 5 to 19 years) | B | Supplementary food, snacks and drinks (18) | 15 | 1 | 1 | 1 | 0 | |
| — | Disadvantaged infants and children (aged 3 months to 5 years) | B | Supplementary food, snacks and drinks (32) | 12 | 2 | 17 | 1 | 2 (nutritional counselling) | |
| 1 (health‐sanitation programme) | |||||||||
| 1 (day‐care centre) | |||||||||
| 1 (day‐care centre + vitamin mineral supplement and sanitation) | |||||||||
| 1 (stimulation only) | |||||||||
| — | Children with MAM (< 5 years of age) | T | Supplementary food, including LNS, blended foods, complementary LNS and blended foods (8) | 0 | 0 | 0 | 8e | 0 | |
| Pregnant women | — | B and T | Balanced energy/protein supplementation (12) | 4 | 2 | 1 | 5 | 5 (nutrition advice only) | |
| High protein supplementation (1) | |||||||||
| Isocaloric protein supplementation (2) | |||||||||
| Nutritional advice (4)d | |||||||||
| — | Children < 5 years of age in LMIC | B and T | Supplementary food, snacks and drinks (8) | 2 | 2 | 3 | 1 | 0 | |
| B: blanket;LMIC: low and middle‐income country; LNS: lipid‐based nutrient supplement; MAM: moderate acute malnutrition; T: targeted; TB: tuberculosis. | |||||||||
aSupplementary food/drink: actual food to eat or fluid to drink as would be found in a household.
bNon‐food: any powder, commercially prepared liquid supplement, mixtures.
cNone relevant to this review.
dExcluded from this overview.
eThis review focused on foods developed for the treatment of MAM, including: LNS, blended food supplements, complementary food supplements.
The reviews evaluated a vast array of different nutritional interventions of varying duration, frequency and format, including solids versus liquids, meals, snacks or drinks, specially formulated foods, fortified foods, traditional foods, commercial macronutrient formulas, mixtures and powders, and specific supplements (e.g. L‐glutamine, spirulina). We included the latter macronutrient supplements for completeness, since they were compared to placebo, no supplements or usual diet in the relevant studies. Intervention categories (as per the relevant reviews) are reported in Table 3. Details of the interventions are summarised in Table 6, Table 7, Table 8, Table 9, Table 10, Table 11, Table 12, and Table 13. The duration of the intervention varied between 20 days (Kristjansson 2007) and three years (Kristjansson 2007), and follow‐up ranged between six weeks (Grobler 2013) and 17 years (Ota 2015). In all but two reviews (Droogsma 2014; Kristjansson 2007), micronutrients were reported as cointerventions in various studies, and associated interventions reported across reviews included nutrition education/counselling, health education, standard medical care, psychosocial stimulation and cash transfers (Table 3).
| Review ID | |
| Types of interventions considered | Nutritional intervention (i.e. any intervention that aimed to improve nutritional status (e.g. weight, upper‐arm anthropometry), such as oral nutritional supplements, dietary advice, food fortification, nutritional education programmes) |
| Details regarding interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
| Review ID | |
| Types of interventions considered |
|
| Details regarding the interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
AA: amino acid; AO: antioxidant; CHO: carbohydrate; FA: fatty acid; ID: identifier; kcal: kilocalories; kJ: kilojoules; RDA: recommended dietary allowances; USDA: US Department of Agriculture.
| Review ID | |
| Types of interventions considered |
|
| Details regarding the interventions |
|
| Comments | Only 7 trials of macronutrient supplementation were reported here (as relevant for this overview). We excluded 1 macronutrient trial, Pérez‐Guzmán 2005, as it was based in a hospital setting (inpatients). Information provided as (and if) reported in systematic review. |
BMI: body mass index; CHO: carbohydrate; ID: identifier; kcal: kilocalories; kJ: kilojoules; RNTCP: Revised National TB Control Program; TB: tuberculosis.
| Review ID | |
| Types of interventions considered | Meals (breakfast or lunch) or snacks (including milk) administered in a school setting |
| Details regarding the interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
DRI: daily recommended intake; ID: identifier; kcal: kilocalories; kJ: kilojoules; NR: not reported; RDA: recommended dietary allowances.
| REVIEW ID | |
| Types of interventions considered | Provision of energy and macronutrients through:
|
| Details regarding interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
CSB: corn‐soy blend; DRA: daily recommended amount; DRI: daily recommended intake; ID: identifier; kcal: kilocalories; kJ: kilojoules; LNS: lipid‐based nutrient supplement; mJ: millijoules; NR: not reported; RDA: recommended dietary allowance; RUTF: ready‐to‐use therapeutic feeding; T: time point.
| Review ID | |
| Types of interventions considered | Any type of food used for children with moderate acute malnutrition, including:
|
| Details regarding the interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
CSB: corn‐soy blend; CSB++: corn‐soy blend enriched; ID: identifier; kcal: kilocalories; LNS: lipid‐based nutrient supplement; MN: micronutrient; RUTF: ready‐to‐use therapeutic feeding.
| Review ID | |
| Types of interventions considered |
|
| Details regarding the interventions |
|
| Comments | 4 trials provided nutrition advice only as intervention and are not reported here (see Ota 2015). Information provided as (and if) reported in systematic review. |
CHO: carbohydrate; ID: identifier; kcal: kilocalories; kJ: kilojoules; MMN: multiple micronutrient; N/A: not applicable; RDA: recommended dietary allowance.
| Review ID | |
| Types of interventions considered | Supplementary feeding was defined as the provision of extra food to children or families beyond the normal rations of their home diets. The intervention had to be community based in that young children could consume the supplementary food at home, at a supervised feeding centre or at other places adapted for this purpose such as healthcare centres and crèches. Supplementary feeding could comprise:
|
| Details regarding the interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
CHO: carbohydrate; ID: identifier; kcal: kilocalories.
The main outcome categories reported across reviews were: mortality; anthropometry (adults and children) and other markers of nutritional status assessment; disease‐related outcomes; neurocognitive development and psychosocial outcomes; and adverse events. A summary of the specific outcomes reported per review is presented in Table 14.
| Systematic reviews | Vulnerability | Outcomes | ||||||||||
| Mortality | Disease‐related outcomes | Nutritional status assessment | Cognition tests, educational attainment and school attendance | Behavioural outcomesa | Quality of lifea | Adverse events | Costs | |||||
| Growth (weight and length/height) | Other anthropometry | Biochemistry | Dietary intake | |||||||||
| Alzheimer's disease | — | — | — | — | — | Adultsb | — | — | — | — | — | |
| HIV positive | Adults | Adults | Children | Adults | Adults | Adults | — | — | Adults | Adults | — | |
| Children | Children | Children | ||||||||||
| TB | Adults | Adults | — | Adults | — | — | — | — | Adults | — | — | |
| Disadvantaged school children | — | Children | Children | Children | Children | — | Children | Children | — | — | — | |
| Disadvantaged infants and children | Children | Children | Children | Children | Children | — | Children | Children | — | Children | — | |
| Children with MAM (< 5 years of age) | Children | — | Children | Children | — | — | — | — | — | Children | — | |
| Pregnancy | Children | — | Children | Adults | — | — | Children | — | — | Adults | — | |
| Children | Children | |||||||||||
| Children < 5 years of age in LMIC | — | — | Children | Children | — | — | — | — | — | Children | — | |
| Total | 2 adults, 4 children | 2 adults, 2 children | 6 children | 3 adults, 6 children | 1 adult, 3 children | 2 adults | 3 children | 2 children | 2 adults | 2 adults, 4 children; | 0 children or adults | |
| LMIC: low‐ and middle‐income country; MAM: moderate acute malnutrition; TB: tuberculosis. | ||||||||||||
aOnly reported narratively.
Methodological quality of included reviews
Quality of systematic reviews
We rated the quality of the eight included systematic reviews using the AMSTAR tool, as described previously under Assessment of methodological quality of included reviews. We found that:
-
all reviews prespecified their clinical question and inclusion criteria;
-
all reviews conducted study selection and data extraction in duplicate;
-
all reviews conducted a comprehensive literature search;
-
seven of the reviews included defined searches of grey literature;
-
all reviews listed included and excluded studies;
-
all reviews described the characteristics of the included studies;
-
all reviews assessed study quality;
-
all reviews appropriately used the quality of included studies in formulating conclusions;
-
all reviews combined the studies using appropriate methods;
-
six reviews formally addressed the risk of publication bias, using a statistical test where appropriate; and
-
all reviews addressed the potential for conflict of interest.
Seven of the eight reviews had conducted a literature search between 2011 and 2016; the one remaining review had an older search date (Kristjansson 2007). We rated all reviews as high quality, as all had scores between eight and 11. See Table 15.
| Criteria | ||||||||
| Was an a priori design provided? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was there duplicate study selection and data extraction? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was a comprehensive literature search performed? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was the status of publication (i.e. grey literature) used as an exclusion criterion?a | N | Nb | N | N | Nb | N | N | Y |
| Was a list of studies (included and excluded) provided? | Y | Y | Y | Y | Y | Y | Y | Y |
| Were the characteristics of the included studies provided? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was the scientific quality of the included studies assessed and documented? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was the scientific quality of the included studies used appropriately in formulating conclusions? | Y | Y | Y | Y | Y | Y | Y | Y |
| Were the methods used to combine the findings of studies appropriate? | N/A | Y | Y | Y | Y | Y | Y | Y |
| Was the likelihood of publication bias assessed? (where relevant) | N/A | Nc | Y | Y | Y | Y | Y | Y |
| Was the conflict of interest stated? | Y | Y | Y | Y | Y | Y | Y | Y |
| AMSTAR scores | 8 | 10 | 11 | 11 | 11 | 11 | 11 | 10 |
| Y: yes; N: no; N/A: not applicable. | ||||||||
aFor all items except item 4, a rating of 'yes' was considered adequate. For item 4, a rating of 'no' was considered adequate.
bExtensive handsearches not undertaken by authors, but trials not excluded if found.
cAuthors discussed the risk involved; no formal assessment.
AMSTAR ratings (scores out of 11 criteria)
-
High quality: 8–11.
-
Medium quality: 4–7.
-
Lower: low quality: ≤ 3.
Certainty of evidence from primary studies in included reviews
The included reviews used GRADE methods to rate the certainty of the evidence reported by the primary studies (as reported in 'Summary of findings' tables in the individual reviews). Ratings ranged from very low to moderate for individual comparisons (see Table 16; Table 17; Table 18; Table 19; Table 20; Table 21; Table 22; Table 23; Table 24; Table 25; Table 26; Table 27; Table 28 for details). We reported the certainty of the evidence in the individuals reviews in these tables and in the text, where available; however, ratings may not be directly comparable between reviews due to different approaches. The main reasons for the certainty of the evidence being downgraded across reviews were: inadequate reporting of allocation concealment and randomisation methods; lack of blinding; imprecision and indirectness. The evidence often comprised one or two small trials.
| Review | Target group | Intervention | Outcome | Assumed risk with comparator | Corresponding risk with intervention | Relative effect (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Children with HIV (aged 6–36 months) | Balancedb | Death (8 weeks) | 120 per 1000 | 163 per 1000 | RR 1.42 (0.59 to 3.40) | 169 (1) | NR | |
| Death (26 weeks) | 217 per 1000 | 291 per 1000 | RR 1.48 (0.74 to 2.98) | 169 (1) | NR | |||
| Adults with TB | Balancedc | Death (1 year follow‐up) | 3 per 100 | 1 per 100 (0 to 4) | RR 0.34 (0.10 to 1.20)d | 567 (4) | Very low | |
| Children with MAM (< 5 years of age) | High lipid and balancede | Death | 10 per 1000 | 4 per 1000 | RR 0.44 (0.14 to 1.36) | 1974 (1) | NR | |
| Pregnant women | Balanced | Stillbirth | 30 per 1000 | 18 per 1000 (12 to 28) | RR 0.60 (0.39 to 0.94)f | 3408 (5) | Moderate | |
| Neonatal death | 26 per 1000 | 18 per 1000 (11 to 28) | RR 0.68 (0.43 to 1.07) | 3381 (5) | Low | |||
| High protein | Stillbirth | 33 per 1000 | 27 per 1000 (10 to 72) | RR 0.81 (0.31 to 2.15) | 529 (1) | Low | ||
| Neonatal death | 11 per 1000 | 31 per 1000 (8 to 115) | RR 2.78 (0.75 to 10.36) | 529 (1) | Low | |||
| CI: confidence interval; MAM: moderate acute malnutrition; NR: not reported; RR: risk ratio; TB: tuberculosis. | ||||||||
aAs reported in 'Summary of findings' tables.
bEnhanced diet: modified milk formula providing 150 kcal/kg/day and 15% of calories as protein.
cProvided as a monthly ration.
dNo subgroup differences between people who were HIV positive and people who were HIV negative.
eComplementary lipid‐based nutrient supplement (Plumpy Doz) and blended foods (corn‐soy blended foods enriched) versus counselling (two subgroups, one study).
fRisk of stillbirth significantly reduced in women given balanced energy and protein supplementation (biscuit (containing roasted groundnuts, rice flour, sugar, groundnut oil); supplement with sesame cake, jaggery, oil; fortified food supplement with peanut butter, soy flour, vegetable oil, sugar, micronutrients; supplement with dried skim milk, enriched bread, vegetable oil; oral supplement (beverage)).
Additional comments
-
Stillbirth refers to death after 20 weeks' gestation and before birth.
-
Neonatal death refers to death of a live infant within the first 28 days of life.
-
'Balanced' refer to additional energy or protein supplementation or both in 'balanced' proportions (balanced: carbohydrate: 45% to 65%; protein: 10% to 20%; fat: 25% to 35%).
-
High protein refers to a protein content > 20% to 25% of total energy.
-
Isocaloric balanced protein: a supplement in which the protein content is 'balanced', i.e. provides < 25% of total energy content, but the protein replaced an equal quantity of non‐protein energy in the control group.
-
High lipid/fat refers to a lipid content > 35% of total energy.
-
Adult mortality outcomes in the Grobler 2013 (HIV) review were reported narratively. Neither supplementary food (Sudarsanam 2011) nor daily supplement of spirulina (Yamani 2009) significantly altered the risk of death compared with no supplement or placebo in malnourished, antiretroviral therapy‐naive adults in these two studies.
-
Child mortality outcomes in the Kristjansson 2015a review were reported narratively. One randomised controlled trial reported that there was no significant difference in mortality between children supplemented with ready‐to‐use therapeutic feeding (1671 children) and children who were not supplemented (1862 children; adjusted hazard ratio 0.76, 95% CI 0.51 to 1.13).
| Review | Target group | Intervention | Outcome | Assumed risk with comparator | Corresponding risk with intervention | Relative effect (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Adults with TB | Balanced | Cured (at 6 months) | 48 per 100 | 44 per 100 (28 to 68) | RR 0.91 (0.59 to 1.41)b | 102 (1) | Very low | |
| Balanced and high energy | Treatment completion (at 6 months) | 79 per 100 | 85 per 100 (70 to 100) | Not pooledc | 365 (2) | Very low | ||
| Balanced and high energy | Sputum negative (at 8 weeks) | 76 per 100 | 82 per 100 (65 to 100) | RR 1.08 (0.86 to 1.37) | 222 (3) | Very low | ||
| Children with MAM (< 5 years of age) | High lipid and balancedd | Recovered | 554 per 1000 | 715 per 1000 (664 to 765) | RR 1.29 (1.20 to 1.38)e | 2152 (2) | Moderate | |
| High lipid and balancedf | Not recovered | 111 per 1000 | 107 per 1000 (82 to 141) | RR 0.97 (0.74 to 1.27) | 1974 (1) | Low | ||
| Progression to SAM | 116 per 1000 | 90 per 1000 | RR 0.78 (0.59 to 1.03) | 1974 (1) | NR | |||
| Defaulted | 185 per 1000 | 55 per 1000 (41 to 72) | RR 0.30 (0.22 to 0.30)g | 1974 (1) | Moderate | |||
| CI: confidence interval; MAM: moderate acute malnutrition; NR: not reported; RR: risk ratio; SAM: severe acute malnutrition; TB: tuberculosis. | ||||||||
aAs reported in 'Summary of findings' tables.
bNo subgroup differences between people who were HIV positive and people who were HIV negative.
cSubtotals were only given for people who were HIV negative (RR 1.20, 95% CI 1.04 to 1.37) and people with unknown HIV status (RR 0.98, 95% CI 0.86 to 1.12).
dComplementary foods (Pusti Packet) and lipid‐based nutrient supplements (LNS) (i.e. Plumpy Doz and corn‐soy blend (CSB++)).
eThe provision of complementary foods (Pusti Packet) and LNS (Plumpy Doz, CSB++) versus standard care significantly increased recovery rate by 29%.
fComplementary foods (LNS: Plumpy Doz, CSB++).
gThe provision of food (complementary foods (LNS: Plumpy Doz, CSB++) versus standard care significantly decreased the number dropping out by 70%.
Additional comments
-
Kristjansson 2015a narratively reported morbidity outcomes in the review. Six studies (four randomised controlled trials (RCTs) and two controlled before‐and‐after (CBAs) studies) reported on morbidity. Three RCTs (Bhandari 2001; Iannotti 2014; Isanaka 2009) and two CBAS (Gopalan 1973; Tomedi 2012) found few differences between the supplemented group and the control group in the prevalence of morbidity. Roy 2005 (a CBA) reported mixed results; the prevalence of diarrhoea and fever was higher in the children who received supplementation (99 children), while the prevalence of respiratory infection was higher in the control group (90 children).
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| CD4 (cells/mm3) | ||||||
| Adults with HIV | Balanced | CD4 (12 weeks' follow‐up) | MD –114.48 (–233.20 to 4.23) | 81 (2) | Low | |
| Specific (OKG)b | Mean CD4 count at study endpoint | MD –28.00 (–134.93 to 78.93) | 46 (1) | NR | ||
| Specific (GLN)c | MD 66.00 (–53.39 to 185.39) | 21 (1) | NR | |||
| Viral load (log10 copies/mL) | ||||||
| Adults with HIV | Balanced | Viral load (12 weeks' follow‐up) | MD –3.71 (–12.16 to 4.74) | 66 (1) | Very low | |
| Specific (OKG)b | Mean viral load at study endpoint | MD 0.20 (–0.58 to 0.98) | 46 (1) | NR | ||
| CI: confidence interval;GLN: L‐glutamine; MD: mean difference; NR: not reported; OKG: ornithine alpha‐ketoglutarate. | ||||||
aAs reported in 'Summary of findings' tables.
bMonohydrated L‐ornithine alpha‐ketoglutarate versus placebo.
c L‐glutamine versus placebo.
Additional comments
-
Additional, disease‐related, biochemical parameter outcomes reported narratively in the Kristjansson 2015a review. One controlled before‐and‐after (CBA) study in a low‐ and middle‐income country reported a significant effect of supplementation on the risk of anaemia (P = 0.003; 110 participants at final survey); those who were supplemented had a lower risk of being anaemic (odds ratio (OR) 0.58, 95% CI 0.24 to 0.75) (Lutter 2008). Similarly, another CBA with 250 participants reported that while the prevalence of anaemia decreased by 27% in the intervention group, it decreased by only 13% in the control group (De Romaña 2000). In high‐income countries, one randomised controlled trial with 103 children found no significant difference between the intervention and the control groups in change in haemoglobin (Yeung 2000). One CBA with 116 children reported an increase in the number of Aboriginal children who had low haemoglobin levels in the intervention group and a decrease in the corresponding number in the control group (Coyne 1980).
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| School children (aged 5–19 years) | Balanced | Weight gain (kg) | MD 0.39 (0.11 to 0.67)b,c | 1462 (3) | NR | |
| MD 1.42 (1.19 to 1.65)d,e | 102 (1) | NR | ||||
| Change in weight (kg) | MD 0.13 (–0.23 to 0.49)f | 520 (1) | NR | |||
| Weight gain (adjusted ICC = 0.025) (kg) | MD 0.71 (0.48 to 0.95)g,h | 984 (3) | NR | |||
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingi | Weight gain (kg) | MD 0.12 (0.05 to 0.18)j,k | 1057 (9) | Moderate | |
| Supplementary feedingl | Weight gain (kg) | MD 0.24 (0.09 to 0.39)m | 1784 (7) | NR | ||
| Supplementary foodn | Weight gain (kg) | MD –0.10 (–0.52 to 0.32)o | 45 (1) | NR | ||
| Balanced | Weight gain (kg) | MD 0.95 (0.58 to 1.33)p | 116 (1) | NR | ||
| Children with MAM (< 5 years of age) | Balancedq | Weight gain total (kg) | MD 0.18 (0.04 to 0.33)r | 178 (1) | Low | |
| Pregnant women | Balanced | Child's birth weight (g) | MD 40.96 (4.66 to 77.26)s,t | 5385 (11) | Moderate | |
| Child's weight at 1 year (g) | MD 30.43 (–139.67 to 200.53) | 623 (2) | NR | |||
| Child's weight at 11 to 17 years (kg) | MD 0.46 (–0.77 to 1.69)u | 855 (2) | NR | |||
| High protein | Child's birth weight (g) | MD –73.0 (–171.26 to 25.26) | 504 (1) | Low | ||
| Child's weight at 1 year (g) | MD 61.0 (–184.60 to 306.60) | 409 (1) | NR | |||
| Isocaloric balanced protein | Child's birth weight (g) | MD 108.25 (–220.89 to 437.4) | 184 (2) | Very low | ||
| Children < 5 years of age (< 24 years) | High energy, protein and balancedv | Weight at end of intervention (kg) | MD –0.03 (–0.17 to 0.12)w | 587 (3) | NR | |
| Balanced | Weight gain during the intervention (kg) | MD 0.04 (–0.03 to 0.11)j,x | 795 (2) | NR | ||
| CI: confidence interval;ICC: intracluster correlation; MAM: moderate acute malnutrition; MD: mean difference; NR: not reported. | ||||||
aAs reported in 'Summary of findings' tables.
bDeveloping country/low‐ and middle‐income country (LMIC) randomised controlled trials (RCTs).
cChildren who were fed (milk with calcium; githeri and meat; breakfast (patty with meat, vegetables, milk or banana cake)) at school gained significantly more weight (sensitivity analyses with ICCs at 0.01, 0.05 and 0.10 made little difference) (gain of 0.25 kg/year). In subgroup analyses, findings were significant for undernourished and adequately nourished children, as well as children aged 9 to 10 years specifically.
dDeveloped country/high‐income country (HIC) controlled before‐and‐after study (CBA).
eChildren who received milk at school gained significantly more weight.
fDeveloped country/HIC RCT.
gDeveloping country/LMIC CBAs.
hChildren who were fed (school lunch; green gram and palm sugar; vegetable protein mixture) at school gained significantly more weight (sensitivity analyses with ICCs at 0.01, 0.05 and 0.10 made little difference) (gain of 0.75 kg/year). In subgroup analyses, findings were significant for boys and girls, and children aged 5 to 6, 6 to 8 and 9 to 10 years specifically.
iBalanced (four studies); high energy (two studies); high lipid (one study); supplementary food (two studies).
jAnalyses include the same RCT: Simondon 1996 (multi‐country study).
kLow‐ and middle‐income country (LMIC) RCT.
lBalanced (two studies); high energy (one study); high lipid (one study); high protein (one study); supplementary food (two studies).
mLMIC CBA.
n113 g wet ration fruit cereal, rice cereal with apple sauce, mixed cereal with apple sauce and bananas, and oatmeal with apple sauce and bananas (Gerber Products Company).
oHigh‐income country (HIC) RCT.
pAboriginal children, HIC CBA.
qComplementary foods (Pusti Packet).
rTotal weight gain significantly higher in group receiving complementary foods (Pusti Packet) than versus standard care.
sBalanced energy and protein supplement associated with significant increases in mean birth weight (liquid, chocolate‐flavoured supplement; biscuit; milk; supplement with sesame cake, jaggery, oil; fortified food supplement with peanut butter, soy flour, vegetable oil, sugar, micronutrients; supplement as dry powder providing energy, protein, fat; supplement with dried skim milk, enriched bread, vegetable oil; mixture of beans, maize and micronutrients or porridge and micronutrients; oral supplement (beverage); glucose drink; glucose drink and skim milk powder).
tNo subgroup differences between undernourished and adequately nourished groups (test for subgroup differences: Chi2 = 2.35, degrees of freedom (df) = 1 (P = 0.12), I2 = 57.5%).
uNo subgroup differences between boys and girls (test for subgroup differences: Chi2 = 0.22, df = 1 (P = 0.64), I2 = 0%).
vComparison group: no food or low‐protein, kcal supplementation.
wNo subgroup differences based on age, nutritional status (stunted/wasted versus not) of the children and duration of feeding (< 12 months versus ≥ 12 months).
xNo subgroup difference based on duration of feeding but subgroup difference based on age (test for subgroup differences: Chi2 = 7.24, df = 1 (P = 0.01), I2 = 86%): children > 24 months (MD 0.22, 95% CI 0.07 to 0.37).
Additional comments
-
Grobler 2013 described weight outcomes narratively for one trial: children receiving enhanced nutrition support had significantly more weight gain in the first eight weeks than children receiving standard care (P < 0.0001) (Rollins 2007).
-
Kristjansson 2015a narratively reported two additional RCTs in LMIC. One 14‐month RCT (60 children) found a large and significant effect of feeding on weight gain for boys (end‐of‐study difference 3.91 kg; statistically significant) and girls (end‐of‐study difference 2.55 kg; statistically significant) (Obatolu 2003). One study found that 48 children who received supplementary feeding gained a mean of 39 g more than the 43 children in the control group (six‐month intervention: not significant) (Fauveau 1992).
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| School children (aged 5–19 years) | Balanced | Height gain (cm) | MD 0.38 (–0.32 to 1.08)b | 1462 (3) | NR | |
| Change in height (cm) | MD 0.28 (–0.01 to 0.57)c | 520 (1) | NR | |||
| Height gain (adjusted ICC = 0.0016) (cm) | MD 1.43 (0.46 to 2.41)d,e | 986 (6) | NR | |||
| MD 0.92 (0.16 to 1.67)f,g | 703 (4) | NR | ||||
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingh | Height gain (cm) | MD 0.27 (0.07 to 0.48)i,j | 1463 (9) | Moderate | |
| Supplementary feedingk | Height gain (cm) | MD 0.52 (–0.07 to 1.10)l | 1782 (7) | NR | ||
| Supplementary foodm | Height gain (cm) | MD –1.00 (–2.12 to 0.12)n | 45 (1) | NR | ||
| Balanced | Height gain (cm) | MD 0.61 (–0.31 to 1.54)o | 116 (1) | NR | ||
| Children with MAM (< 5 years of age) | Balancedp | Height gain (total) (mm) | MD 1.54 (–2.07 to 5.15) | 178 (1) | NR | |
| Pregnant women | Balanced | Child's birth length (cm) | MD 0.16 (0.01 to 0.31)q | 3370 (5) | NR | |
| Child's length at 1 year (cm) | MD 0.00 (–5.69 to 5.69) | 428 (1) | NR | |||
| High protein | Child's height at 11–17 years (cm) | MD –0.39 (–1.73 to 0.94)r | 855 (1) | NR | ||
| Child's length at 1 year (cm) | MD 0.20 (–5.59 to 5.99) | 412 (1) | NR | |||
| Children < 5 years of age | High energy, protein and balanceds | Length/height at the end of the intervention (cm) | MD 0.28 (–0.11 to 0.67)t | 587 (3) | NR | |
| Balanced | Length/height gain during the intervention (cm) | MD 0.19 (0.07 to 0.31)i,u | 795 (2) | NR | ||
| CI: confidence interval; ICC: intracluster correlation coefficient; MAM: moderate acute malnutrition; MD: mean difference; NR: not reported. | ||||||
aAs reported in 'Summary of findings' tables.
bLow‐ and middle‐income country (LMIC) randomised controlled trials (RCTs).
cHigh‐income country (HIC) RCT.
dLMIC controlled before‐and‐after studies (CBAs).
eHeight gain significantly increased for children who received school meals (lunch; green gram and sugar; vegetable protein mixture).
fHIC CBAs.
gHeight gain significantly increased for children who received school meals (milk).
hBalanced (five studies); high energy (two studies); high lipid (one study); supplementary food (one study).
iAnalyses include the same RCT: Simondon 1996 (multi‐country study).
jLMIC RCT.
kBalanced (two studies); high energy (one study); high lipid (one study); high protein (one study); supplementary food (two studies).
lLMIC CBA.
m113 g wet ration fruit cereal, rice cereal with apple sauce, mixed cereal with apple sauce and bananas, and oatmeal with apple sauce and bananas (Gerber Products Company).
nHIC RCT
oAboriginal children, HIC CBA.
pComplementary foods (Pusti Packet).
qBirth length significantly increased in newborns of women given balanced energy, protein supplementation (liquid, chocolate‐flavoured supplement; biscuit; milk; supplement with sesame cake, jaggery, oil; fortified food supplement with peanut butter, soy flour, vegetable oil, sugar, micronutrients).
rNo significant differences for boys and girls.
sComparison group: no food or low‐protein, kcal supplement.
tNo subgroup differences based on age, nutritional status (stunted/wasted versus not) of the children and duration of feeding (< 12 months versus > 12 months).
uLength gain significantly increased in children given supplementary feeding (porridge and yogurt).
Additional comments :
-
Kristjansson 2015a narratively reported two additional RCTs in LMIC. Pollitt 2000 studied effectiveness for two age cohorts, 12 and 18 months old. They found that supplementary feeding had a significant effect on height for the younger (12‐month‐old) cohort only. Obatolu 2003 (60 children) found a significant effect for feeding on length for boys (5.12 cm difference between intervention and control groups; end‐of‐study difference of 5.02; statistically significant) and girls (6.95 cm difference; end‐of‐study difference of 5.92 cm; statistically significant).
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Weight‐for‐age z scores (WAZ) | ||||||
| Children with HIV | Specific (spirulina)b | WAZ | MD 0.00 (–0.44 to 0.44) | 84 (1) | NR | |
| School children (aged 5–19 years) | Balanced | WAZ | MD 0.07 (0.04 to 0.10)c,d | 785 (1) | NR | |
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedinge | WAZ | MD 0.15 (0.05 to 0.24)f | 1565 (8) | Moderate | |
| Supplementary feedingg | WAZ | MD 0.27 (–0.13 to 0.68)h | 999 (4) | Very low | ||
| Supplementary foodi | Change in WAZ | MD 0.02 (0.01 to 0.03)j | 103 (1) | NR | ||
| Children < 5 years of age | Balanced | Change in WAZ during intervention | MD 0.12 (0.05 to 0.19)k | 348 (1) | Low | |
| WAZ at end of intervention | MD –0.18 (–0.49 to 0.12) | 195 (2) | NR | |||
| Length/height‐for‐age z scores (HAZ) | ||||||
| School children (aged 5–19 years) | Balanced | HAZ | MD 0.04 (0.02 to 0.06)l,m | 1021 (2) | NR | |
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingn | HAZ | MD 0.15 (0.06 to 0.24)f | 4544 (9) | Moderate | |
| Supplementary feedingo | HAZ | MD 0.01 (–0.10 to 0.12)h | 999 (4) | Very low | ||
| Supplementary foodi | Change in HAZ | MD 0.04 (0.04 to 0.05)j | 103 (1) | NR | ||
| Children with MAM (< 5 years of age) | High lipid and balancedp | HAZ | MD 0.23 (–0.07 to 0.54) | 1546 (2) | Low | |
| Children < 5 years of age | Balanced | HAZ at end of intervention | MD 0.02 (–0.29 to 0.32) | 195 (2) | NR | |
| Weight‐for‐height/length z score (WHZ) | ||||||
| Children with HIV | Specific (spirulina)b | WHZ | MD 0.35 (–0.21 to 0.91) | 84 (1) | NR | |
| School children (aged 5–19 years) | Balancedq | Change in WHZ | MD 0.20 (–0.24, 0.64)l | 236 (1) | NR | |
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingr | WHZ | MD 0.10 (–0.02 to 0.22)f | 4073 (7) | Moderate | |
| Supplementary feedings | WHZ | MD 0.29 (–0.11 to 0.69)h | 999 (4) | Very low | ||
| Supplementary foodi | WHZ | MD –0.06 (–0.07 to –0.05)j | 103 (1) | NR | ||
| Children with MAM (< 5 years of age) | High lipid and balancedl | WHZ (final) | MD 0.20 (0.03 to 0.37)t | 1546 (2) | Moderate | |
| Balancedu | WHZ gain (total) | MD 0.28 (0.06 to 0.49)t | 178 (1) | NR | ||
| Children (< 5 years of age) | Balanced | WHZ at end of intervention | MD 0.10 (–0.33 to 0.13)v | 260 (3) | Moderate | |
| BMI z score | ||||||
| Pregnant women | Balanced | Child's BMI z score at 11–17 years | MD 0.16 (0.01 to 0.31)w | 855 (1) | NR | |
| BMI: body mass index; MAM: moderate acute malnutrition; MD: mean difference; NR: not reported. | ||||||
aAs reported in 'Summary of findings' tables.
bSpirulina supplementation.
cLow‐ and middle‐income country (LMIC) randomised controlled trial (RCT).
dSignificant effect of school breakfast (cheese sandwich or spiced bun and cheese plus milk) versus control on weight‐for‐age z (WAZ) scores.
eBalanced (two studies); high energy (two studies); high lipid (two studies); supplementary food (two studies).
fLMIC RCT.
gBalanced (one study); supplementary food (three studies).
hLMIC controlled before‐and‐after study (CBA).
iPuréed meat, iron‐fortified infant cereal and whole cows' milk.
jHigh‐income country (HIC) RCT.
kChange in WAZ significantly higher in the group supplemented with yoghurt.
lA small, significant effect of school feeding on height‐for‐age z scores (HAZ) scores demonstrated; z score difference = 0.04 (school breakfast: cheese sandwich or spiced bun and cheese plus milk; githeri and meat).
mLMIC RCTs.
nBalanced (two studies); high energy (three studies); high lipid (two studies); supplementary food (two studies).
oBalanced (one study); supplementary food (three studies).
pComplementary foods (Pusti Packet) and lipid‐based nutrient supplement (LNS) (i.e. Plumpy Doz and corn‐soy blend (CSB++)).
qLMIC RCT.
rBalanced (three studies); high energy (two studies); high lipid (one study); supplementary food (one study).
sBalanced (one study); supplementary food (three studies).
tFinal weight‐for‐height/length z score (WHZ) score and WHZ gain significantly higher in the group receiving food than in standard care.
uComplementary foods (Pusti Packet).
vNo subgroup differences based on age, nutritional status of the children (stunted/wasted versus not stunted/wasted) and duration of feeding.
wSmall increase in mean body mass index (BMI) z score for children receiving supplementary biscuits versus children in the control group (no subgroup differences between girls and boys).
Additional comments
-
Kristjansson 2015a narratively reported one additional cluster‐RCT in an LMIC. In the cluster‐RCT with 282 children, Roy 2005 found significant effects of supplementation with maternal nutrition education. Those children in the intervention group gained 0.71 more in WAZ than the children who received no treatment (P < 0.001), and 0.26 more in WAZ than the children who received only maternal nutrition education (not significant). One additional CBA in an LMIC was also reported narratively; De Romaña 2000 (250 participants) found no significant difference between intervention and control groups in change in prevalence of stunting (i.e. HAZ scores).
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Head circumference | ||||||
| Pregnant women | Balanced | Child's birth head circumference (cm) | MD 0.04 (–0.08 to 0.17) | 3352 (5) | NR | |
| Head circumference at 1 year (cm) | MD –0.13 (–0.35 to 0.10) | 627 (2) | NR | |||
| High protein | Head circumference at 1 year (cm) | MD 0.11 (–0.19 to 0.41) | 412 (1) | NR | ||
| Children < 5 years of age (stunted; after 12 months) | Balanced | Head circumference at end of the intervention (cm) | MD 0.40 (–0.21 to 1.01) | 65 (1) | NR | |
| Children < 5 years of age (stunted/wasted; after 12 months) | High energy, proteinb | MD 0.19 (–0.41 to 0.79) | 75 (1) | NR | ||
| Mid‐upper arm circumference (MUAC) | ||||||
| School children (aged 5–19 years) | Balanced | MUAC (mm) | MD 0.31 (0.16 to 0.46)c,d | 236 (1) | NR | |
| Children with MAM (< 5 years of age) | Balancedc,e | MUAC gain (total, mm) | MD 0.62 (–1.38 to 2.61) | 178 (1) | Very low | |
| Children < 5 years of age (stunted; after 12 months) | Balanced | MUAC at end of intervention (cm) | MD 0.20 (–0.29 to 0.69) | 65 (1) | NR | |
| Children < 5 years of age (stunted/wasted; after 12 months) | High energy, proteinb | MD 0.10 (–0.22 to 0.42) | 75 (1) | NR | ||
| Children < 5 years of age (nutritionally at risk; after 9 months) | Balanced | MD –0.08 (–0.31 to 0.15) | 348 (1) | NR | ||
| Triceps skinfold thickness | ||||||
| Children < 5 years of age (stunted; after 12 months) | Balanced | Triceps skinfold thickness at end of intervention (mm) | MD 0.20 (–0.51 to 0.91) | 65 (1) | NR | |
| Subscapular skinfold thickness | ||||||
| Children < 5 years of age (stunted; after 12 months) | Balanced | Subscapular skinfold thickness at end of intervention (mm) | MD 0.20 (–0.34 to 0.74) | 65 (1) | NR | |
| Mid‐upper arm muscle area | ||||||
| School children (aged 5–19 years) | Balanced | Mid‐upper arm muscle area (mm2) | MD 68.22 (39.57 to 96.87)d,f | 236 (1) | NR | |
| Mid‐upper arm fat area | ||||||
| School children (aged 5–19 years) | Balanced | Mid‐upper arm fat area (mm2) | MD –0.31 (–26.12 to 25.50)d | 236 (1) | NR | |
| Percentage body fat | ||||||
| Pregnant women | Balanced | Child's % body fat at 11–17 years | MD 0.06 (–0.41 to 0.52)g | 847 (1) | NR | |
| CI: confidence interval; MAM: moderate acute malnutrition; MD: mean difference; NR: not reported. | ||||||
aAs reported in 'Summary of findings' tables.
bComparison group: no food or low protein (kcal) supplementation.
cSignificant increase in mid‐upper arm circumference (MUAC) in the meat group compared to the control group; school feeding (meat versus control) had a greater effect on MUAC for boys than for girls.
dLow‐ and middle‐income country (LMIC) randomised controlled trial (RCT).
eComplementary foods (Pusti Packet).
fChildren in the intervention group, who were given meat, gained significantly more in the mid‐upper arm muscle area than the control group.
gNo subgroup differences between boys and girls.
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Change in haemoglobin (g/L) | ||||||
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingb | Change in haemoglobin (g/L) | SMD 0.49 (0.07 to 0.91)c | 300 (5) | NR | |
| CI: confidence interval; NR: not reported; SMD: standardised mean difference. | ||||||
aAs reported in 'Summary of findings' tables.
bBalanced (one study); high energy (two studies); high lipid (one study); supplementary food (one study).
cLow‐ and middle‐income country (LMIC) randomised controlled trials (RCTs).
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Adults with HIV | Balancedb | Body weight (6–12 weeks' follow‐up) | MD –0.17 (–1.10 to 0.75) | 233 (4) | Moderate | |
| Balancedc | Body weight at baseline (ART arm) | MD –0.58 (–1.47 to 0.31) | 617 (1) | NR | ||
| Body weight at baseline (pre‐ART arm) | MD 0.60 (–0.60 to 1.80) | 429 (1) | NR | |||
| Body weight at 1 month (ART arm) | MD 0.58 (–0.62 to 1.78) | 366 (1) | NR | |||
| Body weight at 1 month (pre‐ART arm) | MD 1.09 (–0.59 to 2.77) | 261 (1) | NR | |||
| Body weight at 3 months (ART arm) | MD 0.41 (–0.99 to 1.81) | 322 (1) | NR | |||
| Body weight at 3 months (pre‐ART arm) | MD 2.82 (1.02 to 4.62)d | 211 (1) | NR | |||
| Body weight at 6 months (ART arm) | MD 0.17 (–1.50 to 1.84) | 237 (1) | NR | |||
| Body weight at 6 months (pre‐ART arm) | MD 3.67 (1.50 to 5.84)d | 157 (1) | NR | |||
| Body weight at 12 months (ART arm) | MD –1.00 (–3.19 to 1.19) | 180 (1) | NR | |||
| Body weight at 12 months (pre‐ART arm) | MD 2.25 (–0.41 to 4.91) | 118 (1) | NR | |||
| Change in body weight at 1 month (ART arm) (kg) | MD 0.90 (0.40 to 1.41)e | 366 (1) | NR | |||
| Change in body weight at 1 month (pre‐ART arm) (kg) | MD 0.82 (0.28 to 1.36)f | 261 (1) | NR | |||
| Change in body weight at 3 months (ART arm) (kg) | MD 1.12 (0.29 to 1.95)e | 322 (1) | NR | |||
| Change in body weight at 3 months (pre‐ART arm) (kg) | MD 1.22 (0.31 to 2.12)f | 211 (1) | NR | |||
| Change in body weight at 6 months (ART arm) (kg) | MD 0.89 (–0.30 to 2.08) | 237 (1) | NR | |||
| Change in body weight at 6 months (pre‐ART arm) (kg) | MD 2.06 (0.82 to 3.30)f | 157 (1) | NR | |||
| Change in body weight at 12 months (ART arm) (kg) | MD –0.03 (–1.78 to 1.71) | 180 (1) | NR | |||
| Change in body weight at 12 months (pre‐ART arm) (kg) | MD 0.83 (–0.79 to 2.45) | 118 (1) | NR | |||
| Specific (AA mixture)g | Mean change in body weight (baseline to 8 weeks) (kg) | MD 2.63 (0.72 to 4.54)h | 43 (1) | NR | ||
| Specific (OKG)i | Mean weight at study endpoint (kg) | MD –5.00 (–11.68 to 1.68) | 46 (1) | NR | ||
| Specific (GLN)j | Mean weight at study endpoint (kg) | MD –1.30 (–10.18 to 7.58) | 21 (1) | NR | ||
| Adults with TB | Balanced and high energy | Mean weight gain (after 6 weeks) (kg) | MD 1.73 (0.81 to 2.65)k | 34 (1) | NR | |
| Mean weight gain (after 8 weeks) (kg) | MD 0.78 (–0.05 to 1.60) | 689 (3) | NR | |||
| Mean weight gain (after 12 weeks) (kg) | MD 2.60 (1.74 to 3.46)k | 100 (1) | NR | |||
| Mean weight gain (after 20 weeks) (kg) | MD –0.20 (–1.34 to 0.94) | 306 (1) | NR | |||
| Mean weight gain (after 24 weeks) (kg) | MD 1.78 (–0.25 to 3.81) | 26 (1) | NR | |||
| Mean weight gain (after 32 weeks) (kg) | MD 2.60 (0.52 to 4.68)k | 265 (1) | NR | |||
| Mean weight gain (at 8 weeks) (kg)l | Not pooled | 731 (4) | Moderate | |||
| Pregnant women | Balanced | Weekly gestational weight gain (g/week) | MD 18.63 (–1.81 to 39.07) | 2391 (9) | NR | |
| Maternal weight 4 weeks' postpartum (kg) | MD –0.90 (–1.92 to 0.12) | 354 (1) | NR | |||
| High protein | Weekly gestational weight gain (g/week) | MD 4.50 (–33.55 to 42.55) | 486 (1) | NR | ||
| Isocaloric balanced protein | Weekly gestational weight gain (g/week) | MD 110.45 (–82.87 to 303.76) | 184 (2) | Very low | ||
| AA: amino acid; ART: antiretroviral therapy; CI: confidence intervals; GLN: L‐glutamine; MD: mean difference; NR: not reported; OKG: ornithine alpha‐ketoglutarate; TB: tuberculosis. | ||||||
aAs reported in 'Summary of findings' tables.
bAll commercial balanced macronutrient formulas + nutrition counselling versus nutrition counselling in participants with weight loss.
cFortified blended food + nutrition counselling versus nutrition counselling in malnourished adults on ART and pre‐ART.
dAmong participants not receiving antiretroviral therapy (ART), the supplement group had a significantly greater mean body weight than the non‐supplement group at both three months (P = 0.0022) and at six months (P = 0.001).
eAmong participants receiving ART, the supplement group appeared to gain weight more rapidly than the non‐supplement group in the first three months of the trial, as they had a significantly greater change in body weight gain compared to the non‐supplement group at these time points. After this time point, the change in body weight was not significantly different between the groups.
fAmong participants not receiving ART, the supplement group gained significantly more body weight compared with the non‐supplement group in the first three months of the trial and at the six‐month time point. After this time point, the change in body weight was not significantly different between the groups.
gAmino acid mixture (arginine, glutamine, β‐hydroxy‐β‐methylbutyrate versus placebo).
hAfter eight weeks, the arginine‐rich group gained significantly more body weight than the control group.
iOrnithine alpha‐ketoglutarate versus placebo.
jL‐glutamine versus placebo.
kSupplementation did seem to improve weight gain at specific time points during treatment, although one large trial exclusively in people coinfected with HIV found no difference at any time point (PrayGod 2011).
lSupplementation probably increases weight gain during treatment. Four studies reported measures of weight gain but at different time points, which prevented meta‐analysis.
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Body mass index (BMI) | ||||||
| Adults with HIV | Balancedb | BMI at baseline (ART arm) | MD 0.02 (–0.15 to 0.19) | 617 (1) | NR | |
| BMI at baseline (pre‐ART arm) | MD 0.17 (–0.07 to 0.41) | 429 (1) | NR | |||
| BMI at 1 month (ART arm) | MD 0.36 (0.08 to 0.64)c | 366 (1) | NR | |||
| BMI at 1 month (pre‐ART arm) | MD 0.39 (0.05 to 0.74)d | 261 (1) | NR | |||
| BMI at 3 months (ART arm) | MD 0.43 (0.07 to 0.79)c | 322 (1) | NR | |||
| BMI at 3 months (pre‐ART arm) | MD 0.73 (0.31 to 1.15)d | 211 (1) | NR | |||
| BMI at 6 months (ART arm) | MD 0.42 (–0.07 to 0.91) | 237 (1) | NR | |||
| BMI at 6 months (pre‐ART arm) | MD 0.78 (0.22 to 1.34)c | 157 (1) | NR | |||
| BMI at 12 months (ART arm) | MD –0.08 (–0.72 to 0.56) | 180 (1) | NR | |||
| BMI at 12 months (pre‐ART arm) | MD 0.45 (–0.25 to 1.15) | 118 (1) | NR | |||
| Lean body mass (LBM) | ||||||
| Adults with HIV | Balancedb | % LBM at baseline (ART arm) | MD 0.13 (–0.96 to 1.23) | 569 (1) | NR | |
| % LBM at baseline (pre‐ART arm) | MD –0.30 (–1.51 to 0.92) | 394 (1) | NR | |||
| % LBM at 1 month (ART arm) | MD 0.47 (–1.20 to 2.13) | 253 (1) | NR | |||
| % LBM at 1 month (pre‐ART arm) | MD 0.41 (–1.40 to 2.22) | 185 (1) | NR | |||
| % LBM at 3 months (ART arm) | MD –0.53 (–2.13 to 1.07) | 283 (1) | NR | |||
| % LBM at 3 months (pre‐ART arm) | MD 1.14 (–0.70 to 2.98) | 179 (1) | NR | |||
| % LBM at 6 months (ART arm) | MD 0.32 (–1.48 to 2.12) | 202 (1) | NR | |||
| % LBM at 6 months (pre‐ART arm) | MD 1.65 (–0.79 to 4.09) | 129 (1) | NR | |||
| % LBM at 12 months (ART arm) | MD –1.53 (–3.55 to 0.49) | 169 (1) | NR | |||
| % LBM at 12 months (pre‐ART arm) | MD 0.67 (–1.82 to 3.16) | 107 (1) | NR | |||
| Fat mass | ||||||
| Adults with HIV | Balancede | Fat mass measured in % of TBW | MD –1.14 (–2.58 to 0.29) | 233 (4) | Moderate | |
| Specific (AA mixture)f | Change in fat mass (kg) | MD –0.64 (–2.69 to 1.41) | 43 (1) | NR | ||
| Specific (OKG)g | Mean fat mass (kg) at study endpoint | MD 0.00 (–2.00 to 2.00) | 46 (1) | NR | ||
| Specific (GLN)h | Mean fat mass (kg) at study endpoint | MD –1.00 (–32.40 to 30.40) | 21 (1) | NR | ||
| Fat‐free mass | ||||||
| Adults with HIV | Balancede | Fat‐free mass | MD –0.37 (–2.77 to 2.03) | 218 (3) | Low | |
| Specific (AA mixture)f | Change in fat‐free mass | MD 3.25 (1.25 to 5.25)i | 43 (1) | NR | ||
| Specific (OKG)g | Mean fat‐free mass (kg) at study endpoint | MD –5.10 (–11.11 to 0.91) | 46 (1) | NR | ||
| AA: amino acid; ART: antiretroviral therapy; BMI: body mass index; CI: confidence interval; GLN: L‐glutamine; LBW: lean body weight; MD: mean difference; NR: not reported; OKG: ornithine alpha‐ketoglutarate; TBW: total body weight. | ||||||
aAs reported in 'Summary of findings' tables.
bFortified blended food + nutrition counselling versus nutrition counselling in malnourished adults on antiretroviral therapy (ART) and pre‐ART.
cAmong participants receiving ART, mean body mass index (BMI) and change in BMI in the supplement group was significantly higher in the first three months compared to the no supplement group. After three months, there was no significant difference in BMI or BMI gain between the supplement and no supplement groups in the participants receiving ART.
dAmong participants not receiving ART, mean BMI and change in BMI in the supplement group was significantly higher in the first six months compared to the no supplement group. After six months, there was no significant difference in BMI or BMI gain between the supplement and no supplement groups in the participants not receiving ART.
eAll commercial balanced macronutrient formulas + nutrition counselling versus nutrition counselling in participants with weight loss.
fAmino acid mixture (arginine, glutamine, β‐hydroxy‐β‐methylbutyrate versus placebo).
gOrnithine alpha‐ketoglutarate versus placebo.
hL‐glutamine versus placebo.
iThe increase in fat‐free mass was significantly greater in the arginine group compared with the control group.
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Energy intake | ||||||
| Adults with HIV | Balancedb | Energy intake (6–12 weeks' follow‐up) (kcal/kg) | MD 393.57 (224.66 to 562.47)c | 131 (3) | Low | |
| Specific (OKG)d | Mean daily energy intake at study endpoint (kcal/kg) | MD 0.66 (–564.63 to 432.63) | 46 (1) | NR | ||
| Protein intake | ||||||
| Adults with HIV | Balancede | Protein intake (g/day) (6–12 weeks follow‐up) | MD 23.35 (12.68 to 34.01)c | 81 (2) | Low | |
| Specific (OKG)d | Mean daily protein intake at study endpoint | MD –0.70 (–18.71 to 17.31) | 43 (1) | NR | ||
| CI: confidence interval;OKG: ornithine alpha‐ketoglutarate; MD: mean difference; NR: not reported. | ||||||
aAs reported in 'Summary of findings' tables.
bMacronutrient formulas (Meritene, Ensure, range of fortified oral supplements).
cSupplementation with balanced macronutrient formulas significantly improved energy intake and protein intake compared with no nutritional supplementation or nutrition counselling alone in adults with weight loss.
dOrnithine alpha‐ketoglutarate versus placebo.
eMacronutrient formulas (Meritene, Ensure).
Additional comments
-
One systematic review described this outcome narratively (Droogsma 2014). In one study in the review, three months of daily oral nutritional supplements significantly improved nutritional outcomes in the intervention group (Lauque 2004). The nutritional status of the control group also improved after three months, although the intervention group improved significantly more than the control group. There were no significant changes on the clinical and biochemical outcomes.
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) |
| Cognitive tests | |||||
| School children (aged 5–19 years) | Balanced | Full scale IQ (total) (adjusted ICC = 0.15) | MD 3.90 (–2.88 to 10.68)a,b | 231 (1) | |
| Full scale IQ (separated) (cluster size as in analysis) (adjusted ICC = 0.15) | MD 3.80 (0.51 to 7.10)a,c,d | 231 (1) | |||
| Performance IQ (total) (adjusted ICC = 0.15) | MD 5.00 (–2.60 to 12.6)a,b | 231 (1) | |||
| Performance IQ (separated) (cluster size as in analysis) (adjusted ICC = 0.15) | MD 5.74 (1.73 to 9.74)a,c,e | 231 (1) | |||
| Verbal IQ (total) (adjusted ICC = 0.15) | MD 3.10 (–2.99 to 9.19)a,b | 231 (1) | |||
| Verbal IQ (separated) (cluster size as in analysis) (adjusted ICC = 0.15) | MD 3.35 (–0.21 to 6.92)a,c | 231 (1) | |||
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplement food | Cognitive ability | SMD 0.58 (0.17 to 0.98)f,g | 99 (1) | |
| High energy | Change on Bailey Scale of Mental Development (BSMD) | SMD –0.40 (–0.79 to –0.00)h | 113 (1) | ||
| Pregnant women | Balanced | Child's Bailey mental score (1 year) | MD –0.74 (–1.95 to 0.47) | 411 (1) | |
| Child's IQ (5 years) | MD 0.00 (–4.98 to 4.98) | 153 (1) | |||
| High protein | Child's Bailey mental score (1 year) | MD 0.32 (–0.91 to 1.55) | 396 (1) | ||
| Educational attainment | |||||
| School children (aged 5–19 years) | Balanced | Maths change overall (ICC = 0.15) | SMD 0.31 (0.09 to 0.53)a,i | 337 (2) | |
| Change in reading (ICC = 0.15) | MD 0.09 (–0.11 to 0.29)a | 106 (1) | |||
| Change in spelling (ICC = 0.15) | MD 0.24 (0.01 to 0.47)a,h | 106 (1) | |||
| School attendance | |||||
| School children (aged 5–19 years) | Balanced | Change in attendance (ICC = 0.15) | MD 4.95 (–3.56 to 13.46)a | 108 (1) | |
| End of study attendance (ICC = 0.15) | MD –0.23 (–17.93 to 17.47)a | 72 (1) | |||
| CI: confidence interval; ICC: intracluster correlations; IQ: intelligence quotient; MD: mean difference. | |||||
aAll comparisons: low‐ and middle‐income country (LMIC) controlled before‐and‐after studies (CBAs).
bSensitivity analyses made very little difference to either the point estimate or the significance.
cFour subgroups of one study (Agarwal 1989).
dChildren who were given school lunches had an end‐of‐study full‐scale intelligence quotient (IQ) that was 3.8 points higher than children who were not given school lunch. Sensitivity analyses with intracluster correlation (ICCs) at 0.10 and 0.20 still significant.
eChildren who were given school lunches had an end‐of‐study performance IQ that was 5.74 points higher than children who were not given school lunch. Sensitivity analyses with ICCs at 0.10 and 0.20 both significant.
fTrial compared results for time point 4 children (supplemented with stimulation from 42 to 84 months of age) to those of time point 2 children (supplemented from 63 to 84 months of age) at 63 months.
gLMIC randomised controlled trial (RCT).
hChange in spelling achievement significantly greater for children who received school meals (breakfast). Sensitivity analysis with an ICC of 0.10 showed much the same results, however, the sensitivity analysis with an ICC of 0.20 was non‐significant.
iChange in maths achievement significantly greater for children who received school meals (lunch and breakfast); results of an analysis with Agarwal 1989 broken down into four nutritional subgroups were similar (standardised mean difference 0.44, 95% confidence interval 0.22 to 0.67). Sensitivity analyses for ICCs of 0.10 and 0.20 made little difference.
Additional comments:
-
An ICC of 0.15 was used for maths, reading, spelling, attendance and intelligence outcomes, with ICCs of 0.10 and 0.20 used for sensitivity analyses (Kristjansson 2007).
-
Kristjansson 2015a narratively reported one additional cluster‐RCT in an LMIC (Pollitt 2000). The study found no main effects of supplementation on the Bailey Scales of Mental Development but reported positive effects in a contrast over time for the younger cohort but not for the older cohort (P < 0.05; 53 children).
-
Kristjansson 2015a narratively reported long‐term follow‐up of cognitive development. Grantham‐McGregor 1997 followed up 97% (127 children) of the original cohort of stunted children (Grantham‐McGregor 1991; 129 children) after four years and tested them on a battery of cognitive and perceptual tests. A multiple regression found effects on perceptual motor tasks, but not on general cognition or memory. Interestingly, stimulation had a significant effect on later perceptual‐motor skills for all children (P < 0.05), but supplementation only had a significant effect for children whose mothers had higher scores on a test of verbal intelligence (P < 0.05). Grantham‐McGregor 2007 also found that supplemented children had higher mean scores than the control group on 14 out of 15 cognitive tests (P = 0.02). Pollitt 1997 performed a seven‐year follow‐up of Husaini 1991. They found no differences between the intervention (125 children) and control (106 children) groups on the Peabody Picture Vocabulary Test, emotionality, and maths. They found small (15‐second difference), positive effects of supplementation on working memory performance, although these are unlikely to be clinically significant.
| Review | Target group | Intervention | Outcome | Assumed risk with comparator | Corresponding risk with intervention | Relative effect (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Adults with HIV | Specific (OKG)b | GI adverse events | 542 per 1000 | 864 per 1000 | RR 1.59 (1.06 to 2.39)c | 46 (1) | NR | |
| Pregnant women | Balanced | Small‐for‐gestational age | 173 per 1000 | 137 per 1000 (120 to 156) | RR 0.79 (0.69 to 0.90)d | 4408 (7) | Moderate | |
| Preterm birth | 112 per 1000 | 108 per 1000 (90 to 130) | RR 0.96 (0.80 to 1.16) | 3384 (5) | Moderate | |||
| Pre‐eclampsia | 73 per 1000 | 108 per 1000 (60 to 195) | RR 1.48 (0.82 to 2.66) | 463 (2) | Very low | |||
| High protein | Small‐for‐gestational age | 117 per 1000 | 185 per 1000 (121 to 282) | RR 1.58 (1.03 to 2.41)e | 505 (1) | Moderate | ||
| Preterm birth | 219 per 1000 | 249 per 1000 (182 to 341) | RR 1.14 (0.83 to 1.56) | 505 (1) | Low | |||
| Children < 5 years of age | Balanced | Diarrhoea | — | — | OR 1.04 (0.67 to 1.62) | 108 (1) | NR | |
| Vomiting | — | — | OR 0.89 (0.38 to 2.10) | 108 (1) | NR | |||
| CI: confidence interval; GI: gastrointestinal; NR: not reported; OKG: ornithine alpha‐ketoglutarate; OR: odds ratio; RR: risk ratio. | ||||||||
aAs reported in 'Summary of findings' tables.
bMonohydrated L‐ornithine alpha‐ketoglutarate (OKG).
cOKG associated with significantly more people reporting one or more GI adverse events.
dIncidence of small‐for‐gestational age birth significantly reduced in women given balanced energy and protein supplementation (liquid, chocolate‐flavoured supplement; biscuit; milk; supplement with sesame cake, jaggery, oil; fortified food supplement with peanut butter, soy flour, vegetable oil, sugar, micronutrients; supplement with dried skim milk, enriched bread, vegetable oil; oral supplement (beverage)).
eHigh‐protein supplementation associated with a significantly increased risk of small‐for‐gestational age babies (high protein oral supplement (beverage)).
Additional comments
-
Lazzerini 2013 only reported adverse events in relation to lipid‐based nutrient supplements (LNS) versus all blended foods and LNS versus specific blended foods.
-
Grobler 2013 poorly reported adverse effects in the included studies and, in general, they were related to tolerance rather than adverse effects. Keithley 2002 found no significant differences for acceptance and tolerance of the formulas (Ensure plus versus Advera). Rabeneck 1998 noted that one participant discontinued the supplement (lipisorb‐specialised medium chain triglycerides formula) due to nausea and epigastric pain, and one discontinued as he did not like the taste of the supplement.
-
Kristjansson 2015a calculated the net benefit from supplementary feeding for seven studies that provided home‐delivered rations (randomised controlled trials (RCTs): Bhandari 2001; De Romaña 2000; Grantham‐McGregor 1991; Rivera 2004; controlled before‐and‐after studies (CBAs): Lutter 2008; Santos 2005; Tomedi 2012); and three of the day‐care/feeding centre studies (RCTs: Husaini 1991; Pollitt 2000; CBA: Devadas 1971). They found important differences in the number of calories provided by the supplementary food and the number of extra calories that the children actually consumed in addition to their regular food. In the take‐home studies, the net benefit to children was only 36% of the extra calories provide by the supplement. In the day‐care/feeding centres, the net benefit was 85% of the extra calories provided by the supplement.
Effect of interventions
The results reported here focused on the pooled analyses as undertaken in the various systematic reviews. We did not report on outcomes where no pooled analyses could be undertaken, but we did make reference to the individual studies, as reported in the reviews. For the Lazzerini 2013 review, we reported comparisons comparing specially formulated foods versus standard care only. We did not report comparisons of various types of specially formulated foods against each other.
Supplementary feeding versus no supplementary feeding (control, placebo, standard care, dietary advice)
Death
Four reviews reported mortality outcomes (Grobler 2013; Grobler 2016; Lazzerini 2013; Ota 2015). Mortality outcomes in the pregnancy review by Ota 2015 included stillbirth (death after 20 weeks' gestation and before birth) and neonatal death (death of a live infant within the first 28 days of life). Lazzerini 2013 reported death in children (aged less than five years) with MAM, and Grobler 2013 reported death (at eight and 26 weeks after study enrolment) for children who were HIV positive (aged six to 36 months) in one study. Grobler 2016 reported death at one‐year follow‐up in adults with tuberculosis. See Table 16 for details.
Figure 3 displays the meta‐analysis estimates for mortality across the four participant groups in the four reviews reporting data on this outcome. Estimates were all underpowered (too small to detect small but important differences), apart from balanced energy and protein supplementation in pregnancy, which suggested potential effects on reducing the risk of stillbirth (risk ratio (RR) 0.60, 95% confidence interval (CI) 0.39 to 0.94; 5 studies, 3408 women; moderate‐certainty evidence).
Outcome ‐ mortality: supplementary feeding versus no supplementary feeding (control, placebo, standard care, dietary advice).
In the HIV review, Grobler 2013 provided a narrative description of adult mortality outcomes, indicating that neither supplementary food (Sudarsanam 2011) nor daily supplementation of spirulina (Yamani 2009) significantly altered the risk of death compared with no supplement or placebo, in malnourished, antiretroviral therapy (ART)‐naive adults in the two studies that reported data on this outcome.
Illness (or disease‐related outcomes)
Three reviews reported illness‐related treatment outcomes (Grobler 2016; Kristjansson 2015a; Lazzerini 2013). Outcomes in the review of children with MAM included aspects of recovery, progression to SAM and defaulting (Lazzerini 2013). The tuberculosis review reported cure rate, treatment completion/failure and sputum conversion (Grobler 2016). Kristjansson 2015a narratively reported morbidity outcomes. All included comparisons included only one or two studies. See Table 17.
In the Lazzerini 2013 review on MAM, the provision of complementary foods (Pusti Packet) combined with LNS (Plumpy Doz, Corn‐Soy Blend (CSB++)), compared with standard care, increased 'recovery rate' by 29% (RR 1.29, 95% CI 1.20 to 1.38; 2 studies, 2152 children; moderate‐certainty evidence). The provision of complementary foods (LNS: Plumpy Doz, Corn‐Soy Blend (CSB++)) compared with standard care made little difference to the number recovering, or progression to SAM, but did reduce the number defaulting from the programme (RR 0.30, 95% CI 0.22 to 0.30; 1 study, 1974 children; moderate‐certainty evidence).
In the tuberculosis review, Grobler 2016, studies assessing the provision of free food or high energy supplements appeared to make little difference in terms of disease outcomes measured in various ways, although studies were all underpowered. One small study found a significant benefit in terms of treatment completion and sputum conversion, although these findings remain to be confirmed (very low‐certainty evidence). See Table 17.
In the review on disadvantaged infants and young children, Kristjansson 2015a provided a narrative description of the morbidity outcome data from six studies (four RCTs; two CBAs). Three RCTs (Bhandari 2001; Iannotti 2014; Isanaka 2009) and two CBAs (Gopalan 1973; Tomedi 2012) found few differences between the provision of food or high‐energy supplements and regular diet in the prevalence of morbidity. Roy 2005 (a CBA) reported mixed results; the prevalence of diarrhoea and fever was higher in the 99 children who received balanced protein supplementary food while the prevalence of respiratory infection was higher in the regular diet group (90 children).
Disease‐related biochemical parameters
One review reported disease‐related biochemical parameters (Grobler 2013). Outcomes in this HIV review included CD4 count and viral load. All comparisons included only one or two very small studies with fewer than 100 participants. None of the comparisons for this outcome were significant. See Table 18.
Growth in children
Weight
Five reviews reported on aspects of children's weight (Kristjansson 2007; Kristjansson 2015a; Lazzerini 2013; Ota 2015; Sguassero 2012). The pregnancy review, Ota 2015, reported birthweight and weight at two time points (aged one year and 11 to 17 years). The other reviews reported change in weight or weight at the end of the intervention. See Table 19 for details and Figure 4.
Outcome ‐ weight (g) in children: supplementary feeding versus no supplementary feeding (control, placebo, standard care, dietary advice).
Except for balanced energy and protein supplementation in the pregnancy review (Ota 2015), and food supplementation in the disadvantaged infants and young children review (Kristjansson 2015a), all comparisons included only one to three studies.
Balanced energy and protein supplementation in the pregnancy review was associated with increases in mean birth weight (mean difference (MD) 40.96 g, 95% CI 4.66 to 77.26; 11 studies, 5385 participants; moderate‐certainty evidence) (Ota 2015). Overall, the effects on weight in children as a result of pregnancy supplementation were limited, and not sustained.
Effects on growth in children were mixed. In children under five years of age from LMIC, one review found that supplementary feeding had a negligible impact on child growth (Sguassero 2012). However, a more recent review found that disadvantaged infants and young children in LMIC who received food supplementation gained a mean of 0.12 kg more than those who were not supplemented (MD 0.12 kg, 95% CI 0.05 to 0.18; 9 RCTs, 1057 participants; moderate‐certainty evidence) (Kristjansson 2015a). Sensitivity analyses with intracluster correlation coefficient (ICCs) at 0.10 made little difference, and findings from a subgroup analysis were significant for infants younger than 12 months of age and young children aged one to two years, but not in children older than two years of age. Supplementary feeding of undernourished children resulted in significant weight gain of 0.34 kg (95% CI 0.18 to 0.50) relative to controls, while the intervention was ineffective for well‐nourished children in a subgroup analysis of one trial at 0.08 kg (95% CI –0.09 to 0.25) (Thakwalakwa 2010). In further comparisons in CBAs in LMIC, infants and young children who received food supplementation gained a mean of 0.24 kg more than those who were not supplemented (MD 0.24 kg, 95% CI 0.09 to 0.39; 7 CBAs, 1784 participants). Sensitivity analyses with ICCs at 0.10 made little difference and, in subgroup analysis, findings were only significant for children aged two years. In HIC, one RCT assessed weight gain, and found that children who received supplementation in the form of an iron‐fortified cereal showed no additional gain in weight than children who received no supplementation (MD –0.10 kg, 95% CI –0.52 to 0.32; 45 participants) (Ziegler 2009). One CBA in young aboriginal children in Australia found that children receiving hot lunches in day care centres gained a mean of 0.95 kg more than those who did not (MD 0.95 kg, 95% CI 0.58 to 1.33; 116 participants) (Coyne 1980).
There seemed to be some gains in weight in children with MAM and school children, but the analyses were underpowered (Lazzerini 2013). In the MAM review, total weight gain was higher in children receiving complementary foods (Pusti Packet) when compared to those receiving standard care (MD 0.18 kg, 95% CI 0.04 to 0.33; 1 study, 178 participants; low‐certainty evidence) (Lazzerini 2013).
As reported in the Kristjansson 2007 review, disadvantaged school‐aged children who were fed at school gained an average of 0.39 kg more than those who were not supplemented (MD 0.39 kg, 95% CI 0.11 to 0.67; 3 RCTs in LMIC, 1462 participants). This translated to a gain of 0.25 kg per year for these children. Sensitivity analyses with ICCs at 0.01, 0.05 and 0.10 made little difference, and findings from a subgroup analysis were significant for undernourished and adequately nourished children, specifically children aged nine to 10 years. In further comparisons, children who received milk at school gained significantly more weight than those who did not (MD 1.42 kg, 95% CI 1.19 to 1.65; 1 CBA in a HIC, 102 participants). Children who were fed at school gained 0.71 kg more weight than controls (MD 0.71 kg, 95% CI 0.48 to 0.95; 3 CBA studies in LMIC, 984 participants). This translated to a gain of 0.75 kg per year for these children. Sensitivity analyses with ICCs at 0.01, 0.05 and 0.10 made little difference, and findings from a subgroup analysis were significant for boys and girls, and specifically for children aged five to six, six to eight and nine to 10 years of age.
Length/height
Five reviews reported on aspects of length/height in children (Kristjansson 2007; Kristjansson 2015a; Lazzerini 2013; Ota 2015; Sguassero 2012). The pregnancy review by Ota 2015 reported birth length and length/height at two time points: one year and 11 to 17 years. The other reviews reported change in length/height or length/height at the end of the intervention. See Table 20 for details and Figure 5.
Outcome ‐ length/height (mm) in children: supplementary feeding versus no supplementary feeding (control, placebo, standard care, dietary advice).
Some gains in height were reported in children under five years of age (Sguassero 2012) and school children (Kristjansson 2007), but again analyses were underpowered (see Table 20 and Figure 5). For children with MAM, evidence was limited (as reported in one study) and no mean effect demonstrated (Lazzerini 2013).
In the pregnancy review by Ota 2015, birth length was significantly increased in newborns of women given balanced energy protein supplementation (MD 0.16 cm, 95% CI 0.01 to 0.31; 5 studies, 3370 participants). Overall, there was some increase in birth length with pregnancy supplementation but, as with the findings for weight, this was not sustained over time (as reported in one study; Rush 1980).
In the Sguassero 2012 review on children younger than five years of age in LMIC, length/height gain at the end of the intervention was significantly higher in children given supplementary feeding (porridge and yogurt) versus children in the control group (MD 0.19 cm, 95% CI 0.07 to 0.31; 2 studies, 795 participants). Kristjansson 2015a also found that disadvantaged infants and young children who received food supplementation gained a mean of 0.27 cm more in height than children who were not supplemented (MD 0.27 cm, 95% CI 0.07 to 0.48; 9 RCTs in LMIC, 1463 participants; moderate‐certainty evidence). A subgroup analysis revealed that supplementary feeding was effective for the youngest age groups (children younger than 12 months and children aged one to two years); there were no significant gains in height in the oldest age group (older than two years of age). In further comparisons of CBAs in LMIC, infants and young children who received food supplementation did not gain significantly more height than those who did not receive supplementation (MD 0.52 cm, 95% CI –0.07 to 1.10; 7 CBAs, 1782 participants). There were no significant effects of supplementation on height in HIC (MD 0.61, 95% CI ‐0.31 to 1.54; 1 study, 116 participants).
In the Kristjansson 2007 review on disadvantaged school children, compared with children in the control group, height gain was significantly increased in children who received school meals in LMIC (MD 1.43 cm, 95% CI 0.46 to 2.41; 6 CBAs, 986 participants) and HIC (MD 0.92 cm, 95% CI 0.16 to 1.67; 4 CBAs, 703 participants).
Growth z scores
Six of the eight reviews included growth z scores (Grobler 2013; Kristjansson 2007; Kristjansson 2015a; Lazzerini 2013; Ota 2015; Sguassero 2012). Four reviews reported data on weight‐for‐age z scores (WAZ) (Grobler 2013; Kristjansson 2007; Kristjansson 2015a; Sguassero 2012), four on length/height‐for‐age z scores (HAZ) (Kristjansson 2007; Kristjansson 2015a; Lazzerini 2013; Sguassero 2012), five on weight‐for‐height/length z (WHZ) scores (Grobler 2013; Kristjansson 2007; Kristjansson 2015a; Lazzerini 2013; Sguassero 2012), and one on body mass index (BMI) z score (Ota 2015). All comparisons included only one to three studies. See Table 21 for details.
Overall, the systematic reviews found modest improvements in z scores for children under five years of age and school‐aged children, but analyses were mostly underpowered.
Weight‐for‐age z scores
In the Sguassero 2012 review on children younger than five years of age in LMIC, the change in WAZ was higher in the group receiving yoghurt supplementation versus control (MD 0.12, 95% CI 0.05 to 0.19; 1 study, 348 participants; low‐certainty evidence). Similarly, RCTs in the Kristjansson 2015a review of disadvantaged infants and young children showed a significant effect on WAZ (MD 0.15, 95% CI 0.05 to 0.24; 8 RCTs in LMIC, 1565 participants; moderate‐certainty evidence) and change in WAZ (MD 0.02, 95% CI 0.01 to 0.03; 1 RCT in HIC, 103 participants). Further comparisons of CBAs in LMIC showed no significant improvement in WAZ (MD 0.27, 955 CI –0.13 to 0.68; 4 CBAs, 999 participants; very low‐certainty evidence). In the Kristjansson 2007 review on disadvantaged school children, a school breakfast (cheese sandwich or spiced bun and cheese plus milk) versus control demonstrated significant benefit in terms of WAZ (MD 0.07, 95% CI 0.04 to 0.10; 1 RCT in a LMIC, 785 participants). Supplementation with spirulina in children who were HIV positive did not lead to improvements in WAZ scores.
Length/height‐for‐age z scores
RCTs in the Kristjansson 2015a review on disadvantaged infants and young children showed a significant effect on HAZ (MD 0.15, 95% CI 0.06 to 0.24; 9 RCTs in LMIC, 4544 participants; moderate‐certainty evidence) and change in HAZ (MD 0.04, 95% CI 0.04 to 0.05; 1 RCT in HIC, 103 participants). Further comparisons of CBAs in LMIC showed no significant improvement in HAZ (MD 0.01, 95% CI –0.10 to 0.12; 4 CBAs, 999 participants; very low‐certainty evidence). In the Kristjansson 2007 review on disadvantaged school children, balanced school‐feeding interventions (school breakfast: cheese sandwich or spiced bun and cheese plus milk; githeri and meat versus control) demonstrated a small significant effect on HAZ with a z score difference of 0.04 (95% CI 0.02 to 0.06; 2 RCTs in LMIC, 1021 participants). Supplementation in children younger than five years of age and children with MAM did not lead to improvements in HAZ scores.
Weight‐for‐height/length z scores
Lazzerini 2013 found that final WHZ (MD 0.20, 95% CI 0.03 to 0.37; 2 studies, 1546 participants; moderate‐certainty evidence) and WHZ (MD 0.28, 95% CI 0.06 to 0.49; 1 trial, 178 participants) were significantly higher in the MAM group of children (younger than five years of age) receiving food versus those in the standard care group. Supplementary feeding in other groups (disadvantaged infants and young children, children younger than five years of age, school‐aged children and children who were HIV positive) did not lead to improvements in WHZ scores.
BMI z score
In the pregnancy review by Ota 2015, there was a small increase in children's mean BMI z score at 11 to 17 years of age in those receiving supplementary biscuits versus control (MD 0.16, 95% CI 0.01 to 0.31; 1 study, 855 participants). There were no subgroup differences between girls and boys.
Nutritional status of children
Other anthropometry indicators
Four of the included reviews reported on a variety of other anthropometrical indices in children. Two reviews reported on head circumference (Ota 2015; Sguassero 2012); three reviews on mid‐upper arm circumference (MUAC) (Kristjansson 2007; Lazzerini 2013; Sguassero 2012); and one review apiece on triceps and subscapular skinfold thickness (Sguassero 2012), mid‐upper arm muscle area and mid‐upper arm fat area (Kristjansson 2007), and percentage body fat (Ota 2015). Most comparisons included only one small study, except for balanced energy and protein supplementation in the pregnancy review (Ota 2015), which included seven studies in two comparison (five studies in one comparison and two studies in the other). See Table 22 for details.
Across reviews, supplementary feeding appeared to have little impact on the specified anthropometric indices, but estimates were all underpowered (apart from balanced energy and protein supplementation in pregnancy).
The only significant finding was in disadvantaged school children (aged five to 19 years) in the Kristjansson 2007 review, who were given meat and gained significantly more mid‐upper arm muscle area than the children in the control group (MD 68.22 mm2, 95% CI 39.57 to 96.87; 1 RCT in LMIC, 236 participants).
Biochemical parameters
Three systematic reviews provided a narrative description of biochemical parameters in children (Grobler 2013; Kristjansson 2007; Kristjansson 2015a).
A Kenyan study (Neumann 2003) (reported in Kristjansson 2007) assessed various micronutrient status indicators (including haemoglobin, plasma ferritin, serum iron, serum zinc, serum copper, plasma vitamin B12, folate and retinol, and erythrocyte riboflavin). School children receiving meat demonstrated significant increases in plasma vitamin B12 concentrations compared to children in the control group (P < 0.0001) after one year‐long intervention (Neumann 2003). All other findings were insignificant. Tisdall 1951 (reported in Kristjansson 2007) compared 'good attenders', 'poor attenders' and controls on serum ascorbic acid, serum carotene and serum vitamin A and reported "statistically significant differences" favouring children who received a school lunch.
The Kenyan study by Neumann 2003 found no differences in haemoglobin increase between the meat and control groups. In Devadas 1979 (reported in Kristjansson 2007), school children receiving a vegetable protein mixture reportedly had a greater increase in haemoglobin than the control group (significance level not reported). Tisdall 1951 found no significant difference in increase in haemoglobin between 'good attenders', 'poor attenders' and controls (statistics not given), while Paige 1976 found that school children receiving a high protein drink as a mid‐morning snack had a larger increase in percentage haematocrit than the control group (P < 0.001) (both studies reported in Kristjansson 2007).
One study in children with rapidly progressing HIV found no significant differences between groups (whey protein concentrate (WPC) versus maltodextrin placebo) for leukocytes, erythrocytes, haemoglobin and platelets (Moreno 2005).
The Kristjansson 2015a review on disadvantaged infants and young children found that children who were supplemented showed positive change in haemoglobin status compared to controls (standardised mean difference (SMD) 0.49 g/L, 95% CI 0.07 to 0.91; 5 RCTs in LMIC, 300 participants). See Table 23 for details. One CBA in an LMIC reported a significant effect of balanced protein supplementation on the risk of anaemia with those who were supplemented having a lower risk of being anaemic (odds ratio (OR) 0.58, 95% CI 0.24 to 0.75; 110 participants) (Lutter 2008). Similarly, another CBA with 250 participants reported that while the prevalence of anaemia decreased by 27% in the intervention group, it decreased by only 13% in the control group (De Romaña 2000). One RCT with 103 children in an HIC found no significant difference between the experimental and control group as regards change in haemoglobin (Yeung 2000); one CBA with 116 children reported an increase in the number of Aboriginal children who had low haemoglobin levels in the experimental group and a decrease in the corresponding number in the control group (Coyne 1980).
Dietary intake
None of the systematic reviews reported dietary intake in children (or individual trials included in these reviews).
Nutritional status of adults
Anthropometry indicators
The pregnancy, HIV and tuberculosis reviews reported weight outcomes in adults. The pregnancy review reported weekly gestational weight gain and maternal weight postpartum (Ota 2015), while the HIV (Grobler 2013) and tuberculosis (Grobler 2016) reviews reported body weight and weight gain at specific time points. See Table 24 for details.
Supplementary feeding had no effect on weight indices in pregnant women, but produced modest increases in weight gain at specific time points in adults with infectious diseases, although it did not seem to convey longer‐term benefits in people who were HIV positive. Estimates were mostly underpowered (apart from balanced energy and protein supplementation in pregnancy).
In one study among participants not receiving ART in the HIV review, the group receiving fortified blended foods had a significantly greater mean body weight than the no supplement group at three months (MD 2.82 kg, 95% CI 1.02 to 4.62; P = 0.0022, 211 participants) and six months (MD 3.67 kg, 95% CI 1.50 to 5.84; P = 0.001, 157 participants) (Grobler 2013). In the same study among participants receiving ART, the supplement group appeared to gain weight more rapidly than the no supplement group at month one (MD 0.90 kg, 95% CI 0.40 to 1.41; 366 participants) and month three (MD 1.12 kg, 95% CI 0.29 to 1.95; 322 participants), as they had a significantly greater change in body weight gain compared to the no supplement group at these time points. After this time point the change in body weight was not significantly different between the groups. Among participants not receiving ART, the supplement group had a significantly greater body weight gain compared with the no supplement group at month one (MD 0.82 kg, 95% CI 0.28 to 1.36; 261 participants), month three (MD 1.22 kg, 95% CI 0.31 to 2.12; 211 participants) and at six‐month time point (MD 2.06 kg, 95% CI 0.82 to 3.30; 157 participants). After this time point the change in body weight was not significantly different between the groups. Supplementation with ornithine alpha‐ketoglutarate (OKG) and L‐glutamine in adults who were HIV positive did not demonstrate any benefit in terms of weight gain.
In one small trial in the HIV review, after eight weeks of receiving an amino acid mixture (including arginine, glutamine and β‐hydroxy‐β‐methylbutyrate (HMB)), the arginine‐rich group gained significantly more body weight than the control group (MD 2.63 kg, 95% CI 0.72 to 4.54; 43 participants) (Grobler 2013).
In the tuberculosis review, balanced and high‐energy supplementation did seem to improve weight gain at specific time points during treatment (at six weeks: MD 1.73 kg, 95% CI 0.81 to 2.65; 1 study, 34 participants; 12 weeks: MD 2.60 kg, 95% CI 1.74 to 3.46; 1 study, 100 participants; 32 weeks: MD 2.60 kg, 95% CI 0.52 to 4.68; 1 study, 265 participants) (Grobler 2016), although one large study, exclusively in people coinfected with HIV, found no difference at any time point (PrayGod 2011). Review authors concluded that supplementation probably produced a modest increase in weight gain during treatment for active tuberculosis, although this was not seen consistently across all trials (data not pooled; 5 trials, 883 participants, moderate‐certainty evidence) (Grobler 2016).
The HIV review by Grobler 2013 reported a few other anthropometrical indices in adults, including BMI, lean body mass (LBM), fat mass and fat‐free mass. See Table 25 for details.
Studies assessing supplementary feeding appeared to have little impact on LBM, fat mass and fat‐free mass (again estimates were underpowered), with some benefit demonstrated in terms of improvements in BMI in the shorter term.
In one trial among participants receiving ART, mean BMI and change in BMI in the supplement group (receiving fortified blended foods) was significantly higher in the first three months compared to the no supplement group (month one: MD 0.36, 95% CI 0.08 to 0.64; 366 participants; month three: MD 0.43, 95% CI 0.07 to 0.79; 322 participants). After three months, there was no significant difference in BMI or BMI gain between the groups. In the same trial among participants not receiving ART, mean BMI and change in BMI in the supplement group was significantly higher in the first six months compared to the no supplement group (month one: MD 0.39, 95% CI 0.05 to 0.74; 261 participants; month three: MD 0.73, 95% CI 0.31 to 1.15; 211 participants; month six: MD 0.78, 95% CI 0.22 to 1.34; 157 participants). After six months, there was no significant difference in BMI or BMI gain between the groups.
In one small study, after eight weeks of receiving an amino acid mixture (including arginine, glutamine and HMB), the arginine‐rich group of participants had a significantly greater increase in fat‐free mass than the control group (MD 3.25 kg, 95% CI 1.25 to 5.25; 1 study, 43 participants).
Biochemical parameters
One systematic review provided a narrative description of biochemical parameters (Grobler 2013).
In one study, haemoglobin values decreased in both groups (spirulina versus green clay) between baseline and six months, with no difference between spirulina and placebo groups at the end of the study (Yamani 2009). In the Castleman 2011 study, ART participants receiving supplements (fortified blended food (Insta Foundation Plus), WPC, micronutrients, nutrition counselling) had a significantly higher increase in haemoglobin levels at month three compared to the no supplement (nutrition counselling only) group (P = 0.05). Supplemented pre‐ART participants had a significantly higher increase in haemoglobin levels at months three (P = 0.01) and six (P = 0.05) compared to the no supplement group.
There was no difference in changes in serum albumin at three months between the supplement and no supplement groups for either group (Castleman 2011).
Serum protein concentrations increased in both study groups (spirulina versus green clay) in the first three months and then decreased in the following three months. There was a significant increase in serum protein in the spirulina group between baseline and three months (P value not shown) and baseline and six months (P < 0.001). At three (P = 0.01) and six (P < 0.001) months, serum protein concentrations were significantly higher in the spirulina group compared to the placebo group (Yamani 2009).
Serum creatinine levels decreased in both study groups (spirulina versus green clay) at months three and six. At three months, serum creatinine levels were significantly higher in the spirulina group compared to the placebo group (P = 0.01) (Yamani 2009).
Dietary intake
The HIV review reported dietary intake (Grobler 2013), and the review on Alzheimer's disease narratively reported dietary intake (Droogsma 2014). See Table 26 for details.
In the HIV review, compared with no nutritional supplementation or nutrition counselling alone, supplementation with balanced macronutrient formulas significantly improved energy intake in adults with weight loss (MD 393.57 kcal/kg, 95% CI 224.66 to 562.47; 3 studies, 131 participants; low‐certainty evidence) and protein intake (MD 23.35 g/day, 95% CI 12.68 to 34.01; 2 studies, 81 participants; low‐certainty evidence) (Grobler 2013). OKG supplementation had no effect on energy and protein intake in one, very small study.
The one study included in the Lauque 2004 review on Alzheimer's disease found that three months of daily oral nutrition supplements (ONS) improved nutritional outcomes in the intervention group. The nutritional status of the control group also improved after three months (with standard care), although the intervention group improved more than the control group. There were no changes in clinical and biochemical outcomes.
School attendance, cognition tests and educational attainment
Three reviews reported outcomes related to school attendance, cognition and educational attainment (Kristjansson 2007; Kristjansson 2015a; Ota 2015). See Table 27 for details.
While supplementary feeding had no effect on cognition tests in the descendants of pregnant women (Ota 2015), there were some small benefits with regards to cognition (intelligence quotient (IQ)) and educational attainment (maths and spelling, but not reading) in disadvantaged school children (Kristjansson 2007), but again analyses were underpowered. Supplementary feeding did not affect school attendance in one small study (Powel 1983) (reported in Kristjansson 2007).
In the Kristjansson 2007 review, disadvantaged school children who were given school lunches had an end‐of‐study, full‐scale IQ (adjusted ICC = 0.15) that was 3.8 points higher than children who were not given school lunches (MD 3.80, 95% CI 0.51 to 7.10; 1 study, 231 participants). Sensitivity analyses with ICCs at 0.10 and 0.20 were still significant. Similarly, in the same study, the intervention group also had an end‐of‐study performance IQ (adjusted ICC = 0.15) that was 5.74 points higher than children who were not given a school lunch (MD 5.74, 95% CI 1.73 to 9.74; 1 study, 231 participants). Again, sensitivity analyses with ICCs at 0.10 and 0.20 were both significant.
Change in maths achievement (ICC = 0.15) was significantly greater for children who received school meals (lunch and breakfast) in two studies (SMD 0.31, 95% CI 0.09 to 0.53; 337 participants). Change in spelling achievement (ICC = 0.15) was greater for children who received a school breakfast compared to children in the control group (MD 0.24, 95% CI 0.01 to 0.47; 1 study, 106 participants). A sensitivity analysis with an ICC of 0.10 showed much the same results, but the sensitivity analysis with an ICC of 0.20 was not significant.
In the Kristjansson 2015a review on disadvantaged infants and young children in LMIC, cognitive ability improved more in children who were supplemented than in children who had not yet received supplementation (SMD 0.58, 95% CI 0.17 to 0.98; 1 RCT, 99 participants; McKay 1978). A further study in an LMIC measured change on the Bailey Scale of Mental Development (BSMD) and found a non‐significant difference (SMD –0.40, 95% CI –0.79 to –0.00; 1 RCT, 113 participants).
Behavioural outcomes
One systematic review provided a narrative description on behavioural outcomes (Kristjansson 2007).
The review presented the results from one study, which showed that playground activity levels, particularly prosocial activity, were higher for children who received school meals (githeri and meat versus no intervention) (P < 0.001; Neumann 2003). Using evidence from three studies, the review found that school feeding may have had positive effects on classroom behaviour in both HIC and LMIC (Bro 1994; Bro 1996; Chang 1996); however, the review authors concluded that effects may have depended on the quality of the educational attainment (Kristjansson 2007). Finally, the review reported one cluster‐RCT of breakfast clubs in the UK, which found no significant changes in hyperactivity levels after the intervention (Shemilt 2004).
Quality of life
Two systematic reviews provided a narrative description of quality of life (Grobler 2013; Grobler 2016). In the HIV review by Grobler 2013, only one study reported data on quality of life (Castleman 2011), where in the initial stages of the trial supplementary food had a significant beneficial effect on the quality of life of pre‐ART participants in particular. These benefits did not seem to persist over longer periods of follow‐up. In the tuberculosis review by Grobler 2016, two studies with 134 participants reported data on quality of life (Jahnavi 2010; Paton 2004). Review authors concluded that supplementation may have increased quality of life scores during the first two months of treatment (low‐certainty evidence).
Adverse events
Four reviews reported on adverse events (Grobler 2013; Kristjansson 2015a; Ota 2015; Sguassero 2012), and included serious adverse events (small‐for‐gestational age (SGA), preterm birth and pre‐eclampsia) in the pregnancy review (Ota 2015), and milder adverse events and discomfort (including diarrhoea, vomiting and general gastrointestinal adverse events) in the three remaining reviews (Grobler 2013; Kristjansson 2015a; Sguassero 2012). See Table 28 for details. Additionally, one review described substitution or leakage (where the family cut home rations for the child who has been fed in order to spread food to other family members, or shared the child's supplementary rations with other family members) as an adverse event (Kristjansson 2015a).
Data on adverse events were generally limited or lacking. One exception was the pregnancy review (Ota 2015). It reported the incidence of SGA birth as significantly reduced in women given balanced energy and protein supplementation (RR 0.79, 95% CI 0.69 to 0.90; 7 studies, 4408 participants; moderate‐certainty evidence). In contrast, high protein supplementation was associated with a significantly increased risk of SGA babies (RR 1.58, 95% CI 1.03 to 2.41; 1 study, 505 participants; moderate‐certainty evidence). Balanced energy and protein supplementation and high protein supplementation had no effect on risk of preterm birth (balanced: moderate‐certainty evidence; high protein: low‐certainty evidence) and pre‐eclampsia (very low‐certainty evidence).
The provision of a multi‐mixture to children younger than five years of age in one small study had no effect on incidence of diarrhoea and vomiting (Sguassero 2012). In one small study, participants who were HIV positive and receiving OKG reported significantly more gastrointestinal adverse events than participants in the placebo group (RR 1.59, 95% CI 1.06 to 2.39; 46 participants) (Grobler 2013).
One review described substitution or leakage by calculating the net benefit from supplementary feeding (Kristjansson 2015a). This review included seven studies that provided home‐delivered rations (RCTs: Bhandari 2001; De Romaña 2000; Grantham‐McGregor 1991; Rivera 2004; CBAs: Lutter 2008; Santos 2005; Tomedi 2012), and three day‐care and feeding centre studies (RCTs: Husaini 1991; Pollitt 2000; CBA: Devadas 1971). It found differences in the number of calories provided by the supplementary food and the number of extra calories that the children had actually consumed in addition to their regular food. In the take‐home studies, the net benefit to children was only 36% of the extra calories provided by the supplement, while in the day‐care and feeding centres the net benefit was 85% of the extra calories provided by the supplement.
None of the included reviews reported on the following outcomes: adherence to treatment or attendance at clinic; costs to the provider and out‐of‐pocket costs to people receiving supplementary feeding.
Discusión
Según los cálculos de la FAO, la prevalencia de la desnutrición sigue siendo alta, a pesar de los suministros alimentarios adecuados y el considerable progreso en la reducción del hambre en algunas regiones. Más de 795 000 000 de personas todavía se presentan a la consulta con niveles crónicamente inadecuados en el aporte energético alimentario (FAO 2015; Sundaram 2015), y las mujeres y los niños son especialmente vulnerables. En 2016; la prevalencia del retraso del crecimiento (22,9%) y la emaciación (7,7%) seguían siendo altos en los niños menores de cinco años de edad a nivel global (UNICEF/WHO/World Bank Group 2017), y se informó que 108 000 000 de personas se enfrentaron a un acceso limitado o peor a los alimentos a nivel de una crisis (FSIN 2017). Se cree que la alimentación complementaria es beneficiosa en los grupos con acceso limitado a los alimentos y en situación de vulnerabilidad debido a la posibilidad de optimizar el valor nutricional y la adecuación del régimen alimentario, lo que mejora la calidad de vida y mejora diversos parámetros de salud de las familias en situación de desventaja. En los PIBM, los problemas que la alimentación complementaria procura considerar están entrelazados con la pobreza y las carencias, y requieren un enfoque multidimensional e integrado. Además, las instalaciones de saneamiento apropiadas y el agua potable son componentes esenciales para asegurar una repercusión efectiva de la alimentación complementaria. Otros factores contextuales relevantes incluyen la disponibilidad de servicios sanitarios básicos y atención médica, educación nutricional y conocimiento y atención parenteral (WHO 1999). Este resumen se realizó para describir y explorar los efectos de la alimentación complementaria, específicamente a través de diversos grupos en situación de vulnerabilidad.
Resumen de los resultados principales
Las ocho revisiones sistemáticas (fechas de búsqueda entre mayo de 2006 y febrero de 2016) evaluaron intervenciones en mujeres embarazadas, niños menores de cinco años de PIBM, neonatos y niños pequeños en situación de desventaja (tres meses a cinco años), niños con DAM, niños en edad escolar en situación de desventaja, adultos y niños con pruebas positivas para el VIH, adultos y niños con tuberculosis activa (con o sin VIH) y personas mayores con enfermedad de Alzheimer. Estas revisiones incluyeron 95 ensayos relevantes a este resumen y la mayoría de los participantes eran de PIBM. Los ensayos incluían una gran variedad de diferentes intervenciones nutricionales de duración, frecuencia y formato variables y a menudo se informó que los micronutrientes fueron las cointervenciones en diversos ensayos. El seguimiento varió de seis semanas a dos años en las revisiones y tres ensayos investigaron los resultados a los cuatro a 17 años de edad.
Mujeres embarazadas: se ha sugerido que la administración de suplementos en las mujeres embarazadas puede tener un efecto sobre el mortinato, el peso del neonato al nacer y el riesgo de neonatos PEG al nacer (evidencia de certeza moderada). Aunque fue clínicamente pequeña, la administración de suplementos aumentó de forma significativa la talla al nacer. Estos efectos demostrados no se tradujeron en beneficios a largo plazo para el niño en cuanto al crecimiento y el desarrollo neurocognitivo en uno o dos ensayos que informaron los resultados a más largo plazo. La administración de suplementos con alto contenido de proteínas se asoció con efectos perjudiciales (mayor riesgo de recién nacidos PEG) (evidencia de certeza moderada), lo que indica que en la actualidad no hay justificación para prescribir estos suplementos a las mujeres embarazadas (considerando que este resultado se basa en resultados informados en un ensayo).
Niños: en los niños menores de cinco años de edad de PIBM, la alimentación complementaria tuvo una repercusión insignificante sobre el crecimiento de los niños y los autores advirtieron que la interpretación de los resultados agrupados se debe realizar con cuidado debido a la presencia de heterogeneidad clínica. Sin embargo, una revisión publicada en 2015 halló que los niños que recibieron suplementos nutricionales ganaron una media de 0,12 kg más en el peso y 0,27 cm más en la altura que los niños que no recibieron suplementos (evidencia de certeza moderada). En los niños menores de cinco años de edad con DAM, la provisión de alimentos especialmente elaborados mejoró de forma significativa el peso (evidencia de certeza baja), las puntuaciones Z del peso para la talla (evidencia de certeza moderada) y otros resultados clave como la tasa de recuperación (en un 29%) (evidencia de certeza moderada) y una disminución en el número de participantes que abandonan el tratamiento (en un 70%) (evidencia de certeza moderada). Otras comparaciones en esta revisión incluyeron tipos de alimentos especialmente elaborados comparados entre sí y los autores establecieron la conclusión de que tanto los CNL como los alimentos combinados (como CSB++, por sus siglas en inglés [alimentos combinados con maíz y soja]) son eficaces para tratar a los niños con DAM (evidencia de certeza moderada a alta). En los niños las comidas en la escuela parecieron dar lugar a varios beneficios pequeños que incluyeron mejorías en el peso (en especial en países de menores ingresos), la altura, las puntuaciones Z del PEG (WAZ, por sus siglas en inglés), las puntuaciones Z de la TEG (HAZ, por sus siglas en inglés), el área mediosuperior muscular del brazo (informada en un estudio pequeño), las pruebas cognitivas (en los PIBM), el rendimiento en matemáticas y al deletrear (en los PIBM) y el comportamiento (descrito de forma narrativa).
Enfermedades infecciosas: la alimentación complementaria en adultos con pruebas positivas para el VIH aumentó el aporte energético y proteico diario en comparación con la orientación nutricional sola (evidencia de certeza baja). La administración de suplementos dio lugar a una mejoría inicial en el aumento de peso/IMC, pero no pareció otorgar un beneficio a largo plazo. No es posible establecer una conclusión sólida con respecto a la administración de suplementos en los niños con diagnóstico de VIH. En adultos con tuberculosis, un ensayo pequeño encontró un beneficio significativo en la finalización del tratamiento y la tasa de conversión del esputo (evidencia de muy baja certeza). Hubo beneficios significativos, pero moderados en cuanto al aumento de peso durante la tuberculosis activa (evidencia de certeza moderada). La administración de suplementos puede haber aumentado las puntuaciones de la calidad de vida durante los dos primeros meses del tratamiento de la tuberculosis (evidencia de baja certeza; descrito de forma narrativa).
enfermedad de Alzheimer: el único estudio incluido en la revisión de la enfermedad de Alzheimer halló que la administración de SNO durante tres meses mejoró de forma significativa los resultados nutricionales en el grupo de intervención. No hubo cambios significativos en los resultados clínicos ni bioquímicos.
Ninguna de las revisiones sistemáticas informó resultados relacionados con los costes.
Conclusión general: los datos de la mortalidad indicaron un efecto de la administración de suplementos sobre el mortinato cuando se los proporcionó a mujeres embarazadas; hubo datos de los ensayos sobre la administración de suplementos para los niños con DAM, los niños con infección por VIH y los adultos con tuberculosis, pero estos estudios fueron pocos y pequeños. Hubo efectos contradictorios sobre el peso y el aumento de peso en los niños. En los niños menores de cinco años de edad de PIBM, una revisión halló que la alimentación complementaria tuvo una repercusión insignificante sobre el crecimiento de los niños, sin embargo, una revisión más reciente halló que los niños que recibieron suplementos alimentarios ganaron una media de 0,12 kg más en el peso y 0,27 cm más en la altura que los niños que no recibieron suplementos. Los suplementos nutricionales en general fueron más efectivos en los niños más pequeños (menores de dos años de edad) y en los que presentaban más desnutrición. En los niños con DAM, hubo beneficios moderados. La administración de suplementos en el embarazo no aportó evidencia de efectos a largo plazo; y en los niños en edad escolar, hubo efectos evidentes más grandes, pero el seguimiento fue de 24 meses en el estudio con la duración más larga. La ganancia en la talla/altura en los niños mostró un modelo similar. Hubo mejorías moderadas en las puntuaciones z para los niños menores de cinco años de edad y en edad escolar, pero la mayoría de los análisis tuvieron poco poder estadístico. Hubo algunos beneficios pequeños con respecto a la cognición (CI) y el logro educacional (matemáticas y deletreo) en los niños, pero una vez más los análisis tuvieron poco poder estadístico. En adultos con enfermedades infecciosas la alimentación complementaria no pareció otorgar beneficios a más largo plazo en cuanto al aumento de peso y otros índices antropométricos. En los escolares menores de cinco años de edad en situación de desventaja, la sustitución o la administración del tratamiento a personas para las que no fue concebido fue un problema apreciable cuando la alimentación se proporcionó en casa (en contraposición con los alimentos proporcionados en centros de asistencia de día/de alimentación).
Se encontró buena cobertura de los grupos en situación de vulnerabilidad por parte de las revisiones sistemáticas y menos ensayos que los esperados por revisión (lo último se debe considerar en el contexto de un probable informe insuficiente, debido a que muchos organismos internacionales y ONG (organizaciones sin fines de lucro) probablemente tienen datos no publicados que podrían haber sido relevantes a las revisiones sistemáticas individuales y, por lo tanto, también este resumen). Además, se encontró evidencia general limitada de un efecto sobre la mortalidad (también considerando que se excluyó a los niños con DAG de este resumen) o el estado nutricional o el rendimiento escolar.
Compleción y aplicabilidad general de las pruebas
Este resumen sintetizó las revisiones sistemáticas publicadas de la alimentación complementaria en grupos en situación de vulnerabilidad. Aunque se planificó incluir a las personas mayores como un grupo vulnerable a priori (Visser 2013), se incluyó sólo una revisión (con un ensayo) que investigó a los participantes mayores de la comunidad con enfermedad de Alzheimer. Se considera que el resumen está completo, aunque también se reconoce que no todas las revisiones sistemáticas incluidas en este resumen estaban actualizadas. Aunque se encontró evidencia dentro de cada una de las categorías de resultado (excepto los costes), a menudo sólo hubo un número pequeño de revisiones y estudios dentro de cada categoría. Algunas de las revisiones no consideraron el nivel socioeconómico ni el estado nutricional al inicio en los criterios de inclusión o los análisis, ni informaron sobre los mismos de forma clara, lo que podría haber tenido una repercusión sobre algunos resultados. Muchos estudios individuales no informaron variables de evaluación clínicas y otras importantes (p.ej., mortalidad, aspectos de la calidad de vida y coste‐efectividad), y para muchas intervenciones de las diversas revisiones hubo muy pocos datos para establecer una conclusión sólida.
Los responsables de políticas y los ejecutores del programa pueden usar la información presentada en este resumen para obtener una visión general de los efectos demostrados (y la certeza de la evidencia que apoya dichos efectos) de la alimentación complementaria a través de las revisiones y los grupos (incluidas las embarazadas, los neonatos, los niños, los adolescentes, los pacientes con infección por VIH/SIDA o tuberculosis, o ambos, y las personas mayores con enfermedad de Alzheimer), con miras a informar los diseños de los programas futuros y ayudar con las evaluaciones de los programas de alimentación complementaria. También se cree que este resumen destaca áreas que necesitan investigación adicional.
Certeza de la evidencia
Se incluyeron ocho revisiones sistemáticas (con 95 ensayos relevantes) de la alimentación complementaria en grupos en situación de vulnerabilidad en este resumen, y el número de participantes relevantes incluidos a través de las revisiones varió entre 91 y 7940 adultos y 271 a más de 12 595 niños. Todas las revisiones sistemáticas incluidas se calificaron como de alta calidad (con puntuaciones de AMSTAR entre ocho y 11). En una revisión, los autores no evaluaron de manera formal el sesgo de publicación (Grobler 2013)
La certeza de la evidencia informada por los estudios primarios en las revisiones incluidas varió de muy baja a moderada para las comparaciones individuales y la evidencia a menudo comprendió uno o dos ensayos pequeños. Se utilizaron las tablas "Resumen de los resultados" de GRADEpro de cada revisión (cuando se informaban en las revisiones individuales para las comparaciones principales). Seis de las ocho revisiones incluyeron dichas tablas.
Hubo evidencia de certeza moderada sobre la administración equilibrada de suplementos de energía proteica con relación a la reducción del mortinato, el aumento del peso del neonato al nacer y la reducción del riesgo de neonatos PEG en la revisión del embarazo. Además, la evidencia que vinculaba la administración de suplementos con alto contenido proteico a un mayor riesgo de recién nacidos PEG fue de certeza moderada. Hubo evidencia de certeza moderada para los resultados del crecimiento en los neonatos en situación de desventaja y los niños pequeños de PIBM. Además, hubo evidencia de certeza moderada para la provisión de alimentos especialmente elaborados con relación a las puntuaciones Z del peso para la talla, la tasa de recuperación y los incumplimientos en la revisión de la DAM. Los beneficios demostrados en cuanto al aumento de peso en la revisión de la tuberculosis también se consideraron de certeza moderada. La evidencia de todos los otros resultados significativos se consideró de certeza baja o muy baja según lo evaluado en las revisiones individuales.
Sesgos potenciales en el proceso de revisión global
Hubo varios sesgos potenciales en el proceso de revisión. Como la extracción de datos estaba limitada a las revisiones sistemáticas y no a los estudios originales, es posible que no se hayan obtenido todos los resultados menores. Un autor del resumen y un revisor del resumen aplicaron de forma independiente los criterios de elegibilidad, evaluaron los estudios para la inclusión, extrajeron los datos y evaluaron la calidad científica de las revisiones (mediante la herramienta AMSTAR), lo que debería haber reducido el riesgo de sesgo en el proceso del resumen. Se debe observar y considerar la superposición en los estudios incluidos entre las revisiones sistemáticas, debido a que cinco ECA aparecieron tanto en Kristjansson 2015a como en Sguassero 2012; lo que podría influir potencialmente en la interpretación general de los resultados relacionados con el peso y la talla/altura en los niños.
Acuerdos y desacuerdos con otros estudios o revisiones
No se conocen otros resúmenes publicados de las revisiones de la alimentación complementaria en grupos en situación de vulnerabilidad.
Conceptual framework of supplementary feeding to improve nutrition (adapted from UNICEF 1990).
PRISMA flow diagram.
Outcome ‐ mortality: supplementary feeding versus no supplementary feeding (control, placebo, standard care, dietary advice).
Outcome ‐ weight (g) in children: supplementary feeding versus no supplementary feeding (control, placebo, standard care, dietary advice).
Outcome ‐ length/height (mm) in children: supplementary feeding versus no supplementary feeding (control, placebo, standard care, dietary advice).
| Review author and date | Reason for exclusion |
| Methods: stated that Cochrane methods were used but did not specify that 2 reviewers assessed risk of bias. Participants: majority of participants in care homes (1 in hospital); only 3/43 in community | |
| Methods: no predetermined eligibility criteria Participants: mainly residing in long‐term care establishments (75%) | |
| Methods: no 'Risk of bias' assessment | |
| Participants: 5/41 in community, including neurology outpatients and those enrolled at hospital discharge (Protocol published as Gibbs 2012) | |
| Participants: hospital inpatients | |
| Methods: no mention that 'Risk of bias' assessment performed in duplicate. | |
| Methods: only 1 database used; single abstract screening; methods for 'Risk of bias' assessment and data extraction not reported. | |
| Methods: only 1 reviewer assessed trials for inclusion, extracted data and assessed trial quality. | |
| Methods: single screening of abstracts and single data extraction. | |
| Methods: no 'Risk of bias' assessment | |
| Methods: only 1 data source; no 'Risk of bias' assessment | |
| Methods: only 1 data source; no 'Risk of bias' assessment | |
| Methods: no mention that abstract screening or data extraction performed in duplicate. | |
| Methods: no details on screening; no 'Risk of bias' assessment | |
| Methods: no mention that data extraction performed in duplicate. | |
| Methods: did not report methods; no 'Risk of bias' assessment | |
| Methods: no 'Risk of bias' assessment | |
| Methods: no 'Risk of bias' assessment; based on 2 previous reviews | |
| Methods: search terms and databases not listed; no mention that screening or data extraction performed in duplicate. | |
| Methods: AMSTAR score of 5; no mention that data extraction or 'Risk of bias' assessment performed in duplicate, though stated Cochrane methods were used to assess risk of bias. | |
| Methods: no mention that data extraction and 'Risk of bias' assessment performed in duplicate. | |
| Methods: methods not described | |
| Methods: 2 databases searched; review methods not reported | |
| Methods: scoping review Methods: 1 reviewer screened; no 'Risk of bias' assessment | |
| Methods: no mention that screening or data extraction performed in duplicate; no 'Risk of bias' assessment | |
| Methods: not described | |
| Methods: second independent reviewer extracted data from half of the studies. Intervention: dietary intervention, macronutrients, micronutrients; data not reported separately; no information reported on intervention. | |
| Methods: no mention that screening or data extraction performed in duplicate. No 'Risk of bias' assessment. | |
| Methods: no mention that data extraction performed in duplicate. Participants: included both well‐nourished and malnourished participants; mostly elderly (23 participants) with a range of acute and chronic conditions, including those with fractures (4 participants), renal disease (2 participants), cancer (5 participants) and respiratory disease (4 participants) Setting: community and hospital settings | |
| Methods: only 1 database searched; no mention that screening or 'Risk of bias' assessment performed in duplicate. | |
| Methods: no mention that screening or 'Risk of bias' assessment performed in duplicate but did mention that data extraction performed in duplicate. Quote: "Even though we included terms like 'supplementary food' and 'supplementary feed' in our literature search but only those studies were included where the term supplementary food was used for introduction of additional food to a breastfed child at the age of 6 months i.e. complementary feeding." | |
| Methods: only 1 database searched; no mention that screening, data extraction or 'Risk of bias' assessment performed in duplicate. | |
| Intervention: only 7/41 meal supplementation Participants: only 2/41 in community | |
| Status: withdrawn and replaced by Ota 2015 Intervention: nutrition advice (not food or supplements) | |
| Status: withdrawn and replaced by Ota 2015 | |
| Status: withdrawn and replaced by Ota 2015 | |
| Intervention: majority of included studies had micronutrient interventions. | |
| Methods: method of data extraction and screening not described. | |
| Methods: screening and data extraction performed in duplicate but no mentioned that 'Risk of bias' assessment performed in duplicate. | |
| Methods: did not list electronic sources; no mention that screening or data extraction performed in duplicate; no 'Risk of bias' assessment. | |
| Methods: AMSTAR score 7; did not describe that screening, data extraction or 'Risk of bias' assessment performed in duplicate. | |
| Intervention: not community‐based supplementary feeding; strategies/intervention to prevent rickets (4 included studies: 3 micronutrient interventions) Outcomes: few outcomes of interest | |
| Methods: single screening; no 'Risk of bias' assessment | |
| Methods: 'Risk of bias' assessment performed in duplicate but not reported that screening and data extraction performed in duplicate. Participants: acute and tertiary healthcare settings | |
| Methods: 1 database searched | |
| Methods: single screening; second author checked included full texts. | |
| Methods: no mention that 'Risk of bias' assessment performed in duplicate. | |
| Methods: no mention that screening and data extraction performed in duplicate; no 'Risk of bias' assessment. | |
| Methods: screening performed in duplicate but no mentioned that data extraction or 'Risk of bias' assessment performed in duplicate. | |
| Setting: the majority of studies included were in a hospital, long‐care or nursing home setting (66%); only 1 subgroup analysis related to community versus hospital (mortality). | |
| Methods: no mention that 'Risk of bias' assessment performed in duplicate. Participants: hospitalised, residential and aged, care and community dwelling populations (1/6 malnourished studies in community) | |
| Methods: 'Risk of bias' assessment performed in duplicate; no mention that screening or data extraction performed in duplicate. | |
| Methods: 1 database; no mention that double data extraction performed in duplicate; no 'Risk of bias' assessment. Intervention: oral or enteral protein energy supplementation | |
| Type of publication: abstract; no full‐text report available Intervention: mostly micronutrient, but also included balanced protein energy | |
| Methods: data extraction performed in duplicate but no mentioned that screening or 'Risk of bias' assessment performed in duplicate | |
| Methods: single title and abstract screening (Protocol published as Stevens 2013) | |
| Methods: not described | |
| Methods: screening and 'Risk of bias' assessment performed in duplicate but no mention that data extraction performed in duplicate. Participants: community, care homes, rehabilitation/community hospitals Outcome: hospital admissions | |
| Methods: screening and 'Risk of bias' method not described; data extraction by 1 reviewer. | |
| Methods: screening, data extraction and 'Risk of bias' assessment performed by 1 reviewer and checked by a second. Participants: 1/9 studies in community | |
| Type of publication: abstract; full‐text report published as Loveday 2012 and excluded. | |
| Methods: not specified methodology for review; no report of number of reviewers Participants: not specified as vulnerable | |
| Methods: no systematic review methods described other than database search | |
| Methods: no mention that screening, data extraction or 'Risk of bias' assessment performed in duplicate | |
| Methods: screening and data extraction performed by 1 reviewer | |
| Methods: not described other than search | |
| AMSTAR: A Measurement Tool to Assess Systematic Reviews. | |
| Review | Vulnerability | Last search date | Population | Included studies (relevant to this overview) | Types of studies included | Participants (relevant to this overview) |
| Alzheimer's disease | April 2013 | Community‐dwelling people with Alzheimer's disease | 1 (1) | All RCTs | 91 adults (all relevant) | |
| HIV positive | August 2011 | Adults and children who were HIV positive | 14 (14) | All RCTs | 1725 adults (all relevant) | |
| February 2012 | 271 children (all relevant) | |||||
| TB | February 2016 | Adults and children with active TB (with/without HIV) | 35 (7)a | All RCTs | 7491 adults (986 relevant) | |
| 792 children (none relevant) | ||||||
| Disadvantaged school children | May 2006 | Children and adolescents (aged 5–19 years) attending primary or high school | 18 (18) | 7 RCTs | > 12,595 children (not accurately reported) (all relevant) | |
| 9 CBAs | ||||||
| 2 ITSs | ||||||
| Disadvantaged infants and young children | January 2014 | Infants and children aged 3 months to 5 years | 32 (32) | 21 RCTs (individual and cluster randomised) | 11,602 children (all relevant) | |
| 11 CBAs (individual and cluster randomised) | ||||||
| Children with MAM (< 5 years of age) | October 2012 | Children with MAM (aged 6 to 60 months) in LMIC | 8 (8) | All RCTs (individual and cluster randomised) | 10,037 children (all relevant) | |
| Pregnancy | January 2015 | Pregnant women | 17 (12)b | All RCTs (individual and cluster randomised) onlyc | 9030 adults (7940 relevant) | |
| Children < 5 years of age in LMIC | January 2011 | Children (aged 0–5 years) in LMIC born at term (≥ 37 weeks) | 8 (8) | All RCTs (individual and cluster randomised) onlyc | 1243 children (all relevant) | |
| CBA: controlled before‐and‐after study;ITS: interrupted time series; LMIC: low‐ and middle‐income countries; MAM: moderate acute malnutrition; RCTs: randomised controlled trial; TB: tuberculosis. | ||||||
| aOnly seven trials (in adults) assessed macronutrient supplementation. | ||||||
| Review | Vulnerability | Intervention categories (as per the original review) | Duration of intervention | Cointerventions | Associated interventions | Main outcome categories | Length of follow‐up |
| Alzheimer's disease | Oral nutritional supplements (1) | 3 months | — | — | Clinical | 3 months | |
| Nutritional | |||||||
| Biochemical | |||||||
| HIV positive | Supplementary food (2) | 6 weeks to 1 year | Micronutrients | Nutrition counselling | Mortality | 6 weeks to 1 year | |
| Macronutrient formulas providing energy and protein (6) | Anthropometry | ||||||
| Dietary intake | |||||||
| Specific macronutrient supplements (6) | Disease parameters | ||||||
| Adverse events | |||||||
| TB | Supplementary food (5) | 60 days to 6 months (for macronutrient interventions) | Micronutrients | Nutrition counselling | Mortality | 8 weeks to 1 year | |
| Macronutrient formulas providing energy and protein (2) | Anthropometry | ||||||
| Micronutrients (28)a | Disease‐related outcomes | ||||||
| Quality of life | |||||||
| Disadvantaged school children | Supplementary food, snacks and drinks (18) | 20 days to 3 years | — | — | Anthropometry | Not consistently reported | |
| Psychosocial outcomes | |||||||
| Disadvantaged infants and young children | Supplementary food (12) | 3–32 months | Micronutrients | Additional rations for family | Growth | Not consistently reported (up to 8 years) | |
| Cash transfers | Anthropometry | ||||||
| Stimulation | |||||||
| Macronutrient formulas providing energy and protein (20) | Health/nutritional education for mothers | Psychosocial outcomes | |||||
| Healthcare, deworming | Adverse events | ||||||
| Children with MAM (< 5 years of age) | Specially formulated foods, including LNS, blended foods, complementary LNS and blended foods (8) | 8–16 weeks (or upon recovery) | Micronutrients | Nutrition education | Mortality | 8–16 weeks (outcomes in 2 trials: 6 months and 12 months) | |
| Health education | Anthropometry | ||||||
| Medical care | Disease‐related outcomes | ||||||
| Psychosocial stimulation | |||||||
| Pregnancy | Balanced protein energy supplementation (12) | 2.5–9 months + during pregnancy (not consistently reported) | Micronutrients | — | Mortality | Not consistently reported (up to 17 years) | |
| High protein supplementation (1) | Anthropometry | ||||||
| Isocaloric protein supplementation (2) | Neurocognitive development | ||||||
| Nutritional advice (4)a | Adverse events | ||||||
| Children < 5 years of age in LMIC | Supplementary food, snacks and drinks (8) | 2–12 months | Micronutrients | — | Anthropometry | 2–12 months | |
| Adverse events | |||||||
| LMIC: low and middle‐income country; LNS: lipid‐based nutrient supplement; MAM: moderate acute malnutrition; TB: tuberculosis. | |||||||
| aExcluded from this overview. | |||||||
| Review | Population | Included studies (relevant to this overview) | Economies (relevant to this overview) | SES (relevant to this overview) | Setting | |||
| LMIC | HIC | Economically disadvantaged, including undernourished, nutritionally‐at‐risk, rural | Economically advantaged, including well‐nourished | Community (or outpatient setting) | Hospital inpatients and other | |||
| Community‐dwelling people with Alzheimer's disease | 1 | 0 | 1 | 1 | 0 | 1 | 0 | |
| Adults and children who were HIV positive | 14 | 7 | 7 | NR | NR | 14a | 0 | |
| Adults and children with TB | 35 (7) | 33 (6) | 2 (1) | NR | NR | 24 (6) | 11 (1)b | |
| Disadvantaged school children (aged 5–19 years) | 18 | 9 | 9 | 18 | 0 | 18 | 0 | |
| Disadvantaged infants and young children (aged 3 months to 5 years) | 32 | 29 | 3c | 30 | 2 | 32 | 0 | |
| Children with MAM (< 5 years of age) | 8 | 8 | 0 | 8 | 0 | 8 | 0 | |
| Pregnant women | 17 (12) | 10 (8) | 7 (4) | 7 (6) | 10 (6) | 17 (12) | 0 | |
| Children < 5 years of age in LMIC | 8 | 8 | 0 | 6d | ? | 8 | 0 | |
| ?: unknown; HIC: high‐income country; LMIC: low‐ and middle‐income country; MAM: moderate acute malnutrition; NR: not reported; SES: socioeconomic status. | ||||||||
| aIn one study, Rollins 2007, children were included that were treated on an inpatient and outpatient basis. | ||||||||
| Systematic review | Population | Supplementary feeding | Intervention categories (as per review) | Intervention summary (number of studies) | |||||
| Supplementary food/drinka | Non‐foodb | Other | |||||||
| No added micronutrients | Added micronutrients | No added micronutrients | Added micronutrients | ||||||
| Adults with Alzheimer's disease | — | T | Oral nutritional supplements (1) | 0 | 0 | 0 | 1 | 0 | |
| Adults who were HIV positive | Children who were HIV positive (3 studies) | B and T | Supplementary food (2) | 0 | 2 | 5 | 7 | 0 | |
| Macronutrient supplements providing energy and protein (6) | |||||||||
| Specific macronutrient supplements (6) | |||||||||
| Adults with TB | Children with TB (3 studies)c | B and T | Supplementary food (5) | 3 | 2 | 1 | 1 | 28 (micronutrients only) | |
| Macronutrient formulas providing energy and protein (2) | |||||||||
| Micronutrients (28)d | |||||||||
| — | Disadvantaged school children (aged 5 to 19 years) | B | Supplementary food, snacks and drinks (18) | 15 | 1 | 1 | 1 | 0 | |
| — | Disadvantaged infants and children (aged 3 months to 5 years) | B | Supplementary food, snacks and drinks (32) | 12 | 2 | 17 | 1 | 2 (nutritional counselling) | |
| 1 (health‐sanitation programme) | |||||||||
| 1 (day‐care centre) | |||||||||
| 1 (day‐care centre + vitamin mineral supplement and sanitation) | |||||||||
| 1 (stimulation only) | |||||||||
| — | Children with MAM (< 5 years of age) | T | Supplementary food, including LNS, blended foods, complementary LNS and blended foods (8) | 0 | 0 | 0 | 8e | 0 | |
| Pregnant women | — | B and T | Balanced energy/protein supplementation (12) | 4 | 2 | 1 | 5 | 5 (nutrition advice only) | |
| High protein supplementation (1) | |||||||||
| Isocaloric protein supplementation (2) | |||||||||
| Nutritional advice (4)d | |||||||||
| — | Children < 5 years of age in LMIC | B and T | Supplementary food, snacks and drinks (8) | 2 | 2 | 3 | 1 | 0 | |
| B: blanket;LMIC: low and middle‐income country; LNS: lipid‐based nutrient supplement; MAM: moderate acute malnutrition; T: targeted; TB: tuberculosis. | |||||||||
| aSupplementary food/drink: actual food to eat or fluid to drink as would be found in a household. | |||||||||
| Review ID | |
| Types of interventions considered | Nutritional intervention (i.e. any intervention that aimed to improve nutritional status (e.g. weight, upper‐arm anthropometry), such as oral nutritional supplements, dietary advice, food fortification, nutritional education programmes) |
| Details regarding interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
| Review ID | |
| Types of interventions considered |
|
| Details regarding the interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
| AA: amino acid; AO: antioxidant; CHO: carbohydrate; FA: fatty acid; ID: identifier; kcal: kilocalories; kJ: kilojoules; RDA: recommended dietary allowances; USDA: US Department of Agriculture. | |
| Review ID | |
| Types of interventions considered |
|
| Details regarding the interventions |
|
| Comments | Only 7 trials of macronutrient supplementation were reported here (as relevant for this overview). We excluded 1 macronutrient trial, Pérez‐Guzmán 2005, as it was based in a hospital setting (inpatients). Information provided as (and if) reported in systematic review. |
| BMI: body mass index; CHO: carbohydrate; ID: identifier; kcal: kilocalories; kJ: kilojoules; RNTCP: Revised National TB Control Program; TB: tuberculosis. | |
| Review ID | |
| Types of interventions considered | Meals (breakfast or lunch) or snacks (including milk) administered in a school setting |
| Details regarding the interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
| DRI: daily recommended intake; ID: identifier; kcal: kilocalories; kJ: kilojoules; NR: not reported; RDA: recommended dietary allowances. | |
| REVIEW ID | |
| Types of interventions considered | Provision of energy and macronutrients through:
|
| Details regarding interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
| CSB: corn‐soy blend; DRA: daily recommended amount; DRI: daily recommended intake; ID: identifier; kcal: kilocalories; kJ: kilojoules; LNS: lipid‐based nutrient supplement; mJ: millijoules; NR: not reported; RDA: recommended dietary allowance; RUTF: ready‐to‐use therapeutic feeding; T: time point. | |
| Review ID | |
| Types of interventions considered | Any type of food used for children with moderate acute malnutrition, including:
|
| Details regarding the interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
| CSB: corn‐soy blend; CSB++: corn‐soy blend enriched; ID: identifier; kcal: kilocalories; LNS: lipid‐based nutrient supplement; MN: micronutrient; RUTF: ready‐to‐use therapeutic feeding. | |
| Review ID | |
| Types of interventions considered |
|
| Details regarding the interventions |
|
| Comments | 4 trials provided nutrition advice only as intervention and are not reported here (see Ota 2015). Information provided as (and if) reported in systematic review. |
| CHO: carbohydrate; ID: identifier; kcal: kilocalories; kJ: kilojoules; MMN: multiple micronutrient; N/A: not applicable; RDA: recommended dietary allowance. | |
| Review ID | |
| Types of interventions considered | Supplementary feeding was defined as the provision of extra food to children or families beyond the normal rations of their home diets. The intervention had to be community based in that young children could consume the supplementary food at home, at a supervised feeding centre or at other places adapted for this purpose such as healthcare centres and crèches. Supplementary feeding could comprise:
|
| Details regarding the interventions |
|
| Comments | Information provided as (and if) reported in systematic review. |
| CHO: carbohydrate; ID: identifier; kcal: kilocalories. | |
| Systematic reviews | Vulnerability | Outcomes | ||||||||||
| Mortality | Disease‐related outcomes | Nutritional status assessment | Cognition tests, educational attainment and school attendance | Behavioural outcomesa | Quality of lifea | Adverse events | Costs | |||||
| Growth (weight and length/height) | Other anthropometry | Biochemistry | Dietary intake | |||||||||
| Alzheimer's disease | — | — | — | — | — | Adultsb | — | — | — | — | — | |
| HIV positive | Adults | Adults | Children | Adults | Adults | Adults | — | — | Adults | Adults | — | |
| Children | Children | Children | ||||||||||
| TB | Adults | Adults | — | Adults | — | — | — | — | Adults | — | — | |
| Disadvantaged school children | — | Children | Children | Children | Children | — | Children | Children | — | — | — | |
| Disadvantaged infants and children | Children | Children | Children | Children | Children | — | Children | Children | — | Children | — | |
| Children with MAM (< 5 years of age) | Children | — | Children | Children | — | — | — | — | — | Children | — | |
| Pregnancy | Children | — | Children | Adults | — | — | Children | — | — | Adults | — | |
| Children | Children | |||||||||||
| Children < 5 years of age in LMIC | — | — | Children | Children | — | — | — | — | — | Children | — | |
| Total | 2 adults, 4 children | 2 adults, 2 children | 6 children | 3 adults, 6 children | 1 adult, 3 children | 2 adults | 3 children | 2 children | 2 adults | 2 adults, 4 children; | 0 children or adults | |
| LMIC: low‐ and middle‐income country; MAM: moderate acute malnutrition; TB: tuberculosis. | ||||||||||||
| aOnly reported narratively. | ||||||||||||
| Criteria | ||||||||
| Was an a priori design provided? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was there duplicate study selection and data extraction? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was a comprehensive literature search performed? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was the status of publication (i.e. grey literature) used as an exclusion criterion?a | N | Nb | N | N | Nb | N | N | Y |
| Was a list of studies (included and excluded) provided? | Y | Y | Y | Y | Y | Y | Y | Y |
| Were the characteristics of the included studies provided? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was the scientific quality of the included studies assessed and documented? | Y | Y | Y | Y | Y | Y | Y | Y |
| Was the scientific quality of the included studies used appropriately in formulating conclusions? | Y | Y | Y | Y | Y | Y | Y | Y |
| Were the methods used to combine the findings of studies appropriate? | N/A | Y | Y | Y | Y | Y | Y | Y |
| Was the likelihood of publication bias assessed? (where relevant) | N/A | Nc | Y | Y | Y | Y | Y | Y |
| Was the conflict of interest stated? | Y | Y | Y | Y | Y | Y | Y | Y |
| AMSTAR scores | 8 | 10 | 11 | 11 | 11 | 11 | 11 | 10 |
| Y: yes; N: no; N/A: not applicable. | ||||||||
| aFor all items except item 4, a rating of 'yes' was considered adequate. For item 4, a rating of 'no' was considered adequate. AMSTAR ratings (scores out of 11 criteria)
| ||||||||
| Review | Target group | Intervention | Outcome | Assumed risk with comparator | Corresponding risk with intervention | Relative effect (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Children with HIV (aged 6–36 months) | Balancedb | Death (8 weeks) | 120 per 1000 | 163 per 1000 | RR 1.42 (0.59 to 3.40) | 169 (1) | NR | |
| Death (26 weeks) | 217 per 1000 | 291 per 1000 | RR 1.48 (0.74 to 2.98) | 169 (1) | NR | |||
| Adults with TB | Balancedc | Death (1 year follow‐up) | 3 per 100 | 1 per 100 (0 to 4) | RR 0.34 (0.10 to 1.20)d | 567 (4) | Very low | |
| Children with MAM (< 5 years of age) | High lipid and balancede | Death | 10 per 1000 | 4 per 1000 | RR 0.44 (0.14 to 1.36) | 1974 (1) | NR | |
| Pregnant women | Balanced | Stillbirth | 30 per 1000 | 18 per 1000 (12 to 28) | RR 0.60 (0.39 to 0.94)f | 3408 (5) | Moderate | |
| Neonatal death | 26 per 1000 | 18 per 1000 (11 to 28) | RR 0.68 (0.43 to 1.07) | 3381 (5) | Low | |||
| High protein | Stillbirth | 33 per 1000 | 27 per 1000 (10 to 72) | RR 0.81 (0.31 to 2.15) | 529 (1) | Low | ||
| Neonatal death | 11 per 1000 | 31 per 1000 (8 to 115) | RR 2.78 (0.75 to 10.36) | 529 (1) | Low | |||
| CI: confidence interval; MAM: moderate acute malnutrition; NR: not reported; RR: risk ratio; TB: tuberculosis. | ||||||||
| aAs reported in 'Summary of findings' tables. Additional comments
| ||||||||
| Review | Target group | Intervention | Outcome | Assumed risk with comparator | Corresponding risk with intervention | Relative effect (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Adults with TB | Balanced | Cured (at 6 months) | 48 per 100 | 44 per 100 (28 to 68) | RR 0.91 (0.59 to 1.41)b | 102 (1) | Very low | |
| Balanced and high energy | Treatment completion (at 6 months) | 79 per 100 | 85 per 100 (70 to 100) | Not pooledc | 365 (2) | Very low | ||
| Balanced and high energy | Sputum negative (at 8 weeks) | 76 per 100 | 82 per 100 (65 to 100) | RR 1.08 (0.86 to 1.37) | 222 (3) | Very low | ||
| Children with MAM (< 5 years of age) | High lipid and balancedd | Recovered | 554 per 1000 | 715 per 1000 (664 to 765) | RR 1.29 (1.20 to 1.38)e | 2152 (2) | Moderate | |
| High lipid and balancedf | Not recovered | 111 per 1000 | 107 per 1000 (82 to 141) | RR 0.97 (0.74 to 1.27) | 1974 (1) | Low | ||
| Progression to SAM | 116 per 1000 | 90 per 1000 | RR 0.78 (0.59 to 1.03) | 1974 (1) | NR | |||
| Defaulted | 185 per 1000 | 55 per 1000 (41 to 72) | RR 0.30 (0.22 to 0.30)g | 1974 (1) | Moderate | |||
| CI: confidence interval; MAM: moderate acute malnutrition; NR: not reported; RR: risk ratio; SAM: severe acute malnutrition; TB: tuberculosis. | ||||||||
| aAs reported in 'Summary of findings' tables. Additional comments
| ||||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| CD4 (cells/mm3) | ||||||
| Adults with HIV | Balanced | CD4 (12 weeks' follow‐up) | MD –114.48 (–233.20 to 4.23) | 81 (2) | Low | |
| Specific (OKG)b | Mean CD4 count at study endpoint | MD –28.00 (–134.93 to 78.93) | 46 (1) | NR | ||
| Specific (GLN)c | MD 66.00 (–53.39 to 185.39) | 21 (1) | NR | |||
| Viral load (log10 copies/mL) | ||||||
| Adults with HIV | Balanced | Viral load (12 weeks' follow‐up) | MD –3.71 (–12.16 to 4.74) | 66 (1) | Very low | |
| Specific (OKG)b | Mean viral load at study endpoint | MD 0.20 (–0.58 to 0.98) | 46 (1) | NR | ||
| CI: confidence interval;GLN: L‐glutamine; MD: mean difference; NR: not reported; OKG: ornithine alpha‐ketoglutarate. | ||||||
| aAs reported in 'Summary of findings' tables. Additional comments
| ||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| School children (aged 5–19 years) | Balanced | Weight gain (kg) | MD 0.39 (0.11 to 0.67)b,c | 1462 (3) | NR | |
| MD 1.42 (1.19 to 1.65)d,e | 102 (1) | NR | ||||
| Change in weight (kg) | MD 0.13 (–0.23 to 0.49)f | 520 (1) | NR | |||
| Weight gain (adjusted ICC = 0.025) (kg) | MD 0.71 (0.48 to 0.95)g,h | 984 (3) | NR | |||
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingi | Weight gain (kg) | MD 0.12 (0.05 to 0.18)j,k | 1057 (9) | Moderate | |
| Supplementary feedingl | Weight gain (kg) | MD 0.24 (0.09 to 0.39)m | 1784 (7) | NR | ||
| Supplementary foodn | Weight gain (kg) | MD –0.10 (–0.52 to 0.32)o | 45 (1) | NR | ||
| Balanced | Weight gain (kg) | MD 0.95 (0.58 to 1.33)p | 116 (1) | NR | ||
| Children with MAM (< 5 years of age) | Balancedq | Weight gain total (kg) | MD 0.18 (0.04 to 0.33)r | 178 (1) | Low | |
| Pregnant women | Balanced | Child's birth weight (g) | MD 40.96 (4.66 to 77.26)s,t | 5385 (11) | Moderate | |
| Child's weight at 1 year (g) | MD 30.43 (–139.67 to 200.53) | 623 (2) | NR | |||
| Child's weight at 11 to 17 years (kg) | MD 0.46 (–0.77 to 1.69)u | 855 (2) | NR | |||
| High protein | Child's birth weight (g) | MD –73.0 (–171.26 to 25.26) | 504 (1) | Low | ||
| Child's weight at 1 year (g) | MD 61.0 (–184.60 to 306.60) | 409 (1) | NR | |||
| Isocaloric balanced protein | Child's birth weight (g) | MD 108.25 (–220.89 to 437.4) | 184 (2) | Very low | ||
| Children < 5 years of age (< 24 years) | High energy, protein and balancedv | Weight at end of intervention (kg) | MD –0.03 (–0.17 to 0.12)w | 587 (3) | NR | |
| Balanced | Weight gain during the intervention (kg) | MD 0.04 (–0.03 to 0.11)j,x | 795 (2) | NR | ||
| CI: confidence interval;ICC: intracluster correlation; MAM: moderate acute malnutrition; MD: mean difference; NR: not reported. | ||||||
| aAs reported in 'Summary of findings' tables. Additional comments
| ||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| School children (aged 5–19 years) | Balanced | Height gain (cm) | MD 0.38 (–0.32 to 1.08)b | 1462 (3) | NR | |
| Change in height (cm) | MD 0.28 (–0.01 to 0.57)c | 520 (1) | NR | |||
| Height gain (adjusted ICC = 0.0016) (cm) | MD 1.43 (0.46 to 2.41)d,e | 986 (6) | NR | |||
| MD 0.92 (0.16 to 1.67)f,g | 703 (4) | NR | ||||
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingh | Height gain (cm) | MD 0.27 (0.07 to 0.48)i,j | 1463 (9) | Moderate | |
| Supplementary feedingk | Height gain (cm) | MD 0.52 (–0.07 to 1.10)l | 1782 (7) | NR | ||
| Supplementary foodm | Height gain (cm) | MD –1.00 (–2.12 to 0.12)n | 45 (1) | NR | ||
| Balanced | Height gain (cm) | MD 0.61 (–0.31 to 1.54)o | 116 (1) | NR | ||
| Children with MAM (< 5 years of age) | Balancedp | Height gain (total) (mm) | MD 1.54 (–2.07 to 5.15) | 178 (1) | NR | |
| Pregnant women | Balanced | Child's birth length (cm) | MD 0.16 (0.01 to 0.31)q | 3370 (5) | NR | |
| Child's length at 1 year (cm) | MD 0.00 (–5.69 to 5.69) | 428 (1) | NR | |||
| High protein | Child's height at 11–17 years (cm) | MD –0.39 (–1.73 to 0.94)r | 855 (1) | NR | ||
| Child's length at 1 year (cm) | MD 0.20 (–5.59 to 5.99) | 412 (1) | NR | |||
| Children < 5 years of age | High energy, protein and balanceds | Length/height at the end of the intervention (cm) | MD 0.28 (–0.11 to 0.67)t | 587 (3) | NR | |
| Balanced | Length/height gain during the intervention (cm) | MD 0.19 (0.07 to 0.31)i,u | 795 (2) | NR | ||
| CI: confidence interval; ICC: intracluster correlation coefficient; MAM: moderate acute malnutrition; MD: mean difference; NR: not reported. | ||||||
| aAs reported in 'Summary of findings' tables. Additional comments :
| ||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Weight‐for‐age z scores (WAZ) | ||||||
| Children with HIV | Specific (spirulina)b | WAZ | MD 0.00 (–0.44 to 0.44) | 84 (1) | NR | |
| School children (aged 5–19 years) | Balanced | WAZ | MD 0.07 (0.04 to 0.10)c,d | 785 (1) | NR | |
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedinge | WAZ | MD 0.15 (0.05 to 0.24)f | 1565 (8) | Moderate | |
| Supplementary feedingg | WAZ | MD 0.27 (–0.13 to 0.68)h | 999 (4) | Very low | ||
| Supplementary foodi | Change in WAZ | MD 0.02 (0.01 to 0.03)j | 103 (1) | NR | ||
| Children < 5 years of age | Balanced | Change in WAZ during intervention | MD 0.12 (0.05 to 0.19)k | 348 (1) | Low | |
| WAZ at end of intervention | MD –0.18 (–0.49 to 0.12) | 195 (2) | NR | |||
| Length/height‐for‐age z scores (HAZ) | ||||||
| School children (aged 5–19 years) | Balanced | HAZ | MD 0.04 (0.02 to 0.06)l,m | 1021 (2) | NR | |
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingn | HAZ | MD 0.15 (0.06 to 0.24)f | 4544 (9) | Moderate | |
| Supplementary feedingo | HAZ | MD 0.01 (–0.10 to 0.12)h | 999 (4) | Very low | ||
| Supplementary foodi | Change in HAZ | MD 0.04 (0.04 to 0.05)j | 103 (1) | NR | ||
| Children with MAM (< 5 years of age) | High lipid and balancedp | HAZ | MD 0.23 (–0.07 to 0.54) | 1546 (2) | Low | |
| Children < 5 years of age | Balanced | HAZ at end of intervention | MD 0.02 (–0.29 to 0.32) | 195 (2) | NR | |
| Weight‐for‐height/length z score (WHZ) | ||||||
| Children with HIV | Specific (spirulina)b | WHZ | MD 0.35 (–0.21 to 0.91) | 84 (1) | NR | |
| School children (aged 5–19 years) | Balancedq | Change in WHZ | MD 0.20 (–0.24, 0.64)l | 236 (1) | NR | |
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingr | WHZ | MD 0.10 (–0.02 to 0.22)f | 4073 (7) | Moderate | |
| Supplementary feedings | WHZ | MD 0.29 (–0.11 to 0.69)h | 999 (4) | Very low | ||
| Supplementary foodi | WHZ | MD –0.06 (–0.07 to –0.05)j | 103 (1) | NR | ||
| Children with MAM (< 5 years of age) | High lipid and balancedl | WHZ (final) | MD 0.20 (0.03 to 0.37)t | 1546 (2) | Moderate | |
| Balancedu | WHZ gain (total) | MD 0.28 (0.06 to 0.49)t | 178 (1) | NR | ||
| Children (< 5 years of age) | Balanced | WHZ at end of intervention | MD 0.10 (–0.33 to 0.13)v | 260 (3) | Moderate | |
| BMI z score | ||||||
| Pregnant women | Balanced | Child's BMI z score at 11–17 years | MD 0.16 (0.01 to 0.31)w | 855 (1) | NR | |
| BMI: body mass index; MAM: moderate acute malnutrition; MD: mean difference; NR: not reported. | ||||||
| aAs reported in 'Summary of findings' tables. Additional comments
| ||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Head circumference | ||||||
| Pregnant women | Balanced | Child's birth head circumference (cm) | MD 0.04 (–0.08 to 0.17) | 3352 (5) | NR | |
| Head circumference at 1 year (cm) | MD –0.13 (–0.35 to 0.10) | 627 (2) | NR | |||
| High protein | Head circumference at 1 year (cm) | MD 0.11 (–0.19 to 0.41) | 412 (1) | NR | ||
| Children < 5 years of age (stunted; after 12 months) | Balanced | Head circumference at end of the intervention (cm) | MD 0.40 (–0.21 to 1.01) | 65 (1) | NR | |
| Children < 5 years of age (stunted/wasted; after 12 months) | High energy, proteinb | MD 0.19 (–0.41 to 0.79) | 75 (1) | NR | ||
| Mid‐upper arm circumference (MUAC) | ||||||
| School children (aged 5–19 years) | Balanced | MUAC (mm) | MD 0.31 (0.16 to 0.46)c,d | 236 (1) | NR | |
| Children with MAM (< 5 years of age) | Balancedc,e | MUAC gain (total, mm) | MD 0.62 (–1.38 to 2.61) | 178 (1) | Very low | |
| Children < 5 years of age (stunted; after 12 months) | Balanced | MUAC at end of intervention (cm) | MD 0.20 (–0.29 to 0.69) | 65 (1) | NR | |
| Children < 5 years of age (stunted/wasted; after 12 months) | High energy, proteinb | MD 0.10 (–0.22 to 0.42) | 75 (1) | NR | ||
| Children < 5 years of age (nutritionally at risk; after 9 months) | Balanced | MD –0.08 (–0.31 to 0.15) | 348 (1) | NR | ||
| Triceps skinfold thickness | ||||||
| Children < 5 years of age (stunted; after 12 months) | Balanced | Triceps skinfold thickness at end of intervention (mm) | MD 0.20 (–0.51 to 0.91) | 65 (1) | NR | |
| Subscapular skinfold thickness | ||||||
| Children < 5 years of age (stunted; after 12 months) | Balanced | Subscapular skinfold thickness at end of intervention (mm) | MD 0.20 (–0.34 to 0.74) | 65 (1) | NR | |
| Mid‐upper arm muscle area | ||||||
| School children (aged 5–19 years) | Balanced | Mid‐upper arm muscle area (mm2) | MD 68.22 (39.57 to 96.87)d,f | 236 (1) | NR | |
| Mid‐upper arm fat area | ||||||
| School children (aged 5–19 years) | Balanced | Mid‐upper arm fat area (mm2) | MD –0.31 (–26.12 to 25.50)d | 236 (1) | NR | |
| Percentage body fat | ||||||
| Pregnant women | Balanced | Child's % body fat at 11–17 years | MD 0.06 (–0.41 to 0.52)g | 847 (1) | NR | |
| CI: confidence interval; MAM: moderate acute malnutrition; MD: mean difference; NR: not reported. | ||||||
| aAs reported in 'Summary of findings' tables. | ||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Change in haemoglobin (g/L) | ||||||
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplementary feedingb | Change in haemoglobin (g/L) | SMD 0.49 (0.07 to 0.91)c | 300 (5) | NR | |
| CI: confidence interval; NR: not reported; SMD: standardised mean difference. | ||||||
| aAs reported in 'Summary of findings' tables. | ||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Adults with HIV | Balancedb | Body weight (6–12 weeks' follow‐up) | MD –0.17 (–1.10 to 0.75) | 233 (4) | Moderate | |
| Balancedc | Body weight at baseline (ART arm) | MD –0.58 (–1.47 to 0.31) | 617 (1) | NR | ||
| Body weight at baseline (pre‐ART arm) | MD 0.60 (–0.60 to 1.80) | 429 (1) | NR | |||
| Body weight at 1 month (ART arm) | MD 0.58 (–0.62 to 1.78) | 366 (1) | NR | |||
| Body weight at 1 month (pre‐ART arm) | MD 1.09 (–0.59 to 2.77) | 261 (1) | NR | |||
| Body weight at 3 months (ART arm) | MD 0.41 (–0.99 to 1.81) | 322 (1) | NR | |||
| Body weight at 3 months (pre‐ART arm) | MD 2.82 (1.02 to 4.62)d | 211 (1) | NR | |||
| Body weight at 6 months (ART arm) | MD 0.17 (–1.50 to 1.84) | 237 (1) | NR | |||
| Body weight at 6 months (pre‐ART arm) | MD 3.67 (1.50 to 5.84)d | 157 (1) | NR | |||
| Body weight at 12 months (ART arm) | MD –1.00 (–3.19 to 1.19) | 180 (1) | NR | |||
| Body weight at 12 months (pre‐ART arm) | MD 2.25 (–0.41 to 4.91) | 118 (1) | NR | |||
| Change in body weight at 1 month (ART arm) (kg) | MD 0.90 (0.40 to 1.41)e | 366 (1) | NR | |||
| Change in body weight at 1 month (pre‐ART arm) (kg) | MD 0.82 (0.28 to 1.36)f | 261 (1) | NR | |||
| Change in body weight at 3 months (ART arm) (kg) | MD 1.12 (0.29 to 1.95)e | 322 (1) | NR | |||
| Change in body weight at 3 months (pre‐ART arm) (kg) | MD 1.22 (0.31 to 2.12)f | 211 (1) | NR | |||
| Change in body weight at 6 months (ART arm) (kg) | MD 0.89 (–0.30 to 2.08) | 237 (1) | NR | |||
| Change in body weight at 6 months (pre‐ART arm) (kg) | MD 2.06 (0.82 to 3.30)f | 157 (1) | NR | |||
| Change in body weight at 12 months (ART arm) (kg) | MD –0.03 (–1.78 to 1.71) | 180 (1) | NR | |||
| Change in body weight at 12 months (pre‐ART arm) (kg) | MD 0.83 (–0.79 to 2.45) | 118 (1) | NR | |||
| Specific (AA mixture)g | Mean change in body weight (baseline to 8 weeks) (kg) | MD 2.63 (0.72 to 4.54)h | 43 (1) | NR | ||
| Specific (OKG)i | Mean weight at study endpoint (kg) | MD –5.00 (–11.68 to 1.68) | 46 (1) | NR | ||
| Specific (GLN)j | Mean weight at study endpoint (kg) | MD –1.30 (–10.18 to 7.58) | 21 (1) | NR | ||
| Adults with TB | Balanced and high energy | Mean weight gain (after 6 weeks) (kg) | MD 1.73 (0.81 to 2.65)k | 34 (1) | NR | |
| Mean weight gain (after 8 weeks) (kg) | MD 0.78 (–0.05 to 1.60) | 689 (3) | NR | |||
| Mean weight gain (after 12 weeks) (kg) | MD 2.60 (1.74 to 3.46)k | 100 (1) | NR | |||
| Mean weight gain (after 20 weeks) (kg) | MD –0.20 (–1.34 to 0.94) | 306 (1) | NR | |||
| Mean weight gain (after 24 weeks) (kg) | MD 1.78 (–0.25 to 3.81) | 26 (1) | NR | |||
| Mean weight gain (after 32 weeks) (kg) | MD 2.60 (0.52 to 4.68)k | 265 (1) | NR | |||
| Mean weight gain (at 8 weeks) (kg)l | Not pooled | 731 (4) | Moderate | |||
| Pregnant women | Balanced | Weekly gestational weight gain (g/week) | MD 18.63 (–1.81 to 39.07) | 2391 (9) | NR | |
| Maternal weight 4 weeks' postpartum (kg) | MD –0.90 (–1.92 to 0.12) | 354 (1) | NR | |||
| High protein | Weekly gestational weight gain (g/week) | MD 4.50 (–33.55 to 42.55) | 486 (1) | NR | ||
| Isocaloric balanced protein | Weekly gestational weight gain (g/week) | MD 110.45 (–82.87 to 303.76) | 184 (2) | Very low | ||
| AA: amino acid; ART: antiretroviral therapy; CI: confidence intervals; GLN: L‐glutamine; MD: mean difference; NR: not reported; OKG: ornithine alpha‐ketoglutarate; TB: tuberculosis. | ||||||
| aAs reported in 'Summary of findings' tables. | ||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Body mass index (BMI) | ||||||
| Adults with HIV | Balancedb | BMI at baseline (ART arm) | MD 0.02 (–0.15 to 0.19) | 617 (1) | NR | |
| BMI at baseline (pre‐ART arm) | MD 0.17 (–0.07 to 0.41) | 429 (1) | NR | |||
| BMI at 1 month (ART arm) | MD 0.36 (0.08 to 0.64)c | 366 (1) | NR | |||
| BMI at 1 month (pre‐ART arm) | MD 0.39 (0.05 to 0.74)d | 261 (1) | NR | |||
| BMI at 3 months (ART arm) | MD 0.43 (0.07 to 0.79)c | 322 (1) | NR | |||
| BMI at 3 months (pre‐ART arm) | MD 0.73 (0.31 to 1.15)d | 211 (1) | NR | |||
| BMI at 6 months (ART arm) | MD 0.42 (–0.07 to 0.91) | 237 (1) | NR | |||
| BMI at 6 months (pre‐ART arm) | MD 0.78 (0.22 to 1.34)c | 157 (1) | NR | |||
| BMI at 12 months (ART arm) | MD –0.08 (–0.72 to 0.56) | 180 (1) | NR | |||
| BMI at 12 months (pre‐ART arm) | MD 0.45 (–0.25 to 1.15) | 118 (1) | NR | |||
| Lean body mass (LBM) | ||||||
| Adults with HIV | Balancedb | % LBM at baseline (ART arm) | MD 0.13 (–0.96 to 1.23) | 569 (1) | NR | |
| % LBM at baseline (pre‐ART arm) | MD –0.30 (–1.51 to 0.92) | 394 (1) | NR | |||
| % LBM at 1 month (ART arm) | MD 0.47 (–1.20 to 2.13) | 253 (1) | NR | |||
| % LBM at 1 month (pre‐ART arm) | MD 0.41 (–1.40 to 2.22) | 185 (1) | NR | |||
| % LBM at 3 months (ART arm) | MD –0.53 (–2.13 to 1.07) | 283 (1) | NR | |||
| % LBM at 3 months (pre‐ART arm) | MD 1.14 (–0.70 to 2.98) | 179 (1) | NR | |||
| % LBM at 6 months (ART arm) | MD 0.32 (–1.48 to 2.12) | 202 (1) | NR | |||
| % LBM at 6 months (pre‐ART arm) | MD 1.65 (–0.79 to 4.09) | 129 (1) | NR | |||
| % LBM at 12 months (ART arm) | MD –1.53 (–3.55 to 0.49) | 169 (1) | NR | |||
| % LBM at 12 months (pre‐ART arm) | MD 0.67 (–1.82 to 3.16) | 107 (1) | NR | |||
| Fat mass | ||||||
| Adults with HIV | Balancede | Fat mass measured in % of TBW | MD –1.14 (–2.58 to 0.29) | 233 (4) | Moderate | |
| Specific (AA mixture)f | Change in fat mass (kg) | MD –0.64 (–2.69 to 1.41) | 43 (1) | NR | ||
| Specific (OKG)g | Mean fat mass (kg) at study endpoint | MD 0.00 (–2.00 to 2.00) | 46 (1) | NR | ||
| Specific (GLN)h | Mean fat mass (kg) at study endpoint | MD –1.00 (–32.40 to 30.40) | 21 (1) | NR | ||
| Fat‐free mass | ||||||
| Adults with HIV | Balancede | Fat‐free mass | MD –0.37 (–2.77 to 2.03) | 218 (3) | Low | |
| Specific (AA mixture)f | Change in fat‐free mass | MD 3.25 (1.25 to 5.25)i | 43 (1) | NR | ||
| Specific (OKG)g | Mean fat‐free mass (kg) at study endpoint | MD –5.10 (–11.11 to 0.91) | 46 (1) | NR | ||
| AA: amino acid; ART: antiretroviral therapy; BMI: body mass index; CI: confidence interval; GLN: L‐glutamine; LBW: lean body weight; MD: mean difference; NR: not reported; OKG: ornithine alpha‐ketoglutarate; TBW: total body weight. | ||||||
| aAs reported in 'Summary of findings' tables. | ||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Energy intake | ||||||
| Adults with HIV | Balancedb | Energy intake (6–12 weeks' follow‐up) (kcal/kg) | MD 393.57 (224.66 to 562.47)c | 131 (3) | Low | |
| Specific (OKG)d | Mean daily energy intake at study endpoint (kcal/kg) | MD 0.66 (–564.63 to 432.63) | 46 (1) | NR | ||
| Protein intake | ||||||
| Adults with HIV | Balancede | Protein intake (g/day) (6–12 weeks follow‐up) | MD 23.35 (12.68 to 34.01)c | 81 (2) | Low | |
| Specific (OKG)d | Mean daily protein intake at study endpoint | MD –0.70 (–18.71 to 17.31) | 43 (1) | NR | ||
| CI: confidence interval;OKG: ornithine alpha‐ketoglutarate; MD: mean difference; NR: not reported. | ||||||
| aAs reported in 'Summary of findings' tables. Additional comments
| ||||||
| Review | Target group | Intervention | Outcome | Corresponding risk with intervention (95% CI) | Number of participants (studies) |
| Cognitive tests | |||||
| School children (aged 5–19 years) | Balanced | Full scale IQ (total) (adjusted ICC = 0.15) | MD 3.90 (–2.88 to 10.68)a,b | 231 (1) | |
| Full scale IQ (separated) (cluster size as in analysis) (adjusted ICC = 0.15) | MD 3.80 (0.51 to 7.10)a,c,d | 231 (1) | |||
| Performance IQ (total) (adjusted ICC = 0.15) | MD 5.00 (–2.60 to 12.6)a,b | 231 (1) | |||
| Performance IQ (separated) (cluster size as in analysis) (adjusted ICC = 0.15) | MD 5.74 (1.73 to 9.74)a,c,e | 231 (1) | |||
| Verbal IQ (total) (adjusted ICC = 0.15) | MD 3.10 (–2.99 to 9.19)a,b | 231 (1) | |||
| Verbal IQ (separated) (cluster size as in analysis) (adjusted ICC = 0.15) | MD 3.35 (–0.21 to 6.92)a,c | 231 (1) | |||
| Disadvantaged infants and young children (aged 3 months to 5 years) | Supplement food | Cognitive ability | SMD 0.58 (0.17 to 0.98)f,g | 99 (1) | |
| High energy | Change on Bailey Scale of Mental Development (BSMD) | SMD –0.40 (–0.79 to –0.00)h | 113 (1) | ||
| Pregnant women | Balanced | Child's Bailey mental score (1 year) | MD –0.74 (–1.95 to 0.47) | 411 (1) | |
| Child's IQ (5 years) | MD 0.00 (–4.98 to 4.98) | 153 (1) | |||
| High protein | Child's Bailey mental score (1 year) | MD 0.32 (–0.91 to 1.55) | 396 (1) | ||
| Educational attainment | |||||
| School children (aged 5–19 years) | Balanced | Maths change overall (ICC = 0.15) | SMD 0.31 (0.09 to 0.53)a,i | 337 (2) | |
| Change in reading (ICC = 0.15) | MD 0.09 (–0.11 to 0.29)a | 106 (1) | |||
| Change in spelling (ICC = 0.15) | MD 0.24 (0.01 to 0.47)a,h | 106 (1) | |||
| School attendance | |||||
| School children (aged 5–19 years) | Balanced | Change in attendance (ICC = 0.15) | MD 4.95 (–3.56 to 13.46)a | 108 (1) | |
| End of study attendance (ICC = 0.15) | MD –0.23 (–17.93 to 17.47)a | 72 (1) | |||
| CI: confidence interval; ICC: intracluster correlations; IQ: intelligence quotient; MD: mean difference. | |||||
| aAll comparisons: low‐ and middle‐income country (LMIC) controlled before‐and‐after studies (CBAs). Additional comments:
| |||||
| Review | Target group | Intervention | Outcome | Assumed risk with comparator | Corresponding risk with intervention | Relative effect (95% CI) | Number of participants (studies) | Certainty of evidence (GRADE)a |
| Adults with HIV | Specific (OKG)b | GI adverse events | 542 per 1000 | 864 per 1000 | RR 1.59 (1.06 to 2.39)c | 46 (1) | NR | |
| Pregnant women | Balanced | Small‐for‐gestational age | 173 per 1000 | 137 per 1000 (120 to 156) | RR 0.79 (0.69 to 0.90)d | 4408 (7) | Moderate | |
| Preterm birth | 112 per 1000 | 108 per 1000 (90 to 130) | RR 0.96 (0.80 to 1.16) | 3384 (5) | Moderate | |||
| Pre‐eclampsia | 73 per 1000 | 108 per 1000 (60 to 195) | RR 1.48 (0.82 to 2.66) | 463 (2) | Very low | |||
| High protein | Small‐for‐gestational age | 117 per 1000 | 185 per 1000 (121 to 282) | RR 1.58 (1.03 to 2.41)e | 505 (1) | Moderate | ||
| Preterm birth | 219 per 1000 | 249 per 1000 (182 to 341) | RR 1.14 (0.83 to 1.56) | 505 (1) | Low | |||
| Children < 5 years of age | Balanced | Diarrhoea | — | — | OR 1.04 (0.67 to 1.62) | 108 (1) | NR | |
| Vomiting | — | — | OR 0.89 (0.38 to 2.10) | 108 (1) | NR | |||
| CI: confidence interval; GI: gastrointestinal; NR: not reported; OKG: ornithine alpha‐ketoglutarate; OR: odds ratio; RR: risk ratio. | ||||||||
| aAs reported in 'Summary of findings' tables. Additional comments
| ||||||||
