Administración de suplementos orales de proteínas y calorías para niños con enfermedades crónicas
Resumen
Antecedentes
El crecimiento y el estado nutricional deficientes son frecuentes en los niños con enfermedades crónicas. La administración de suplementos orales de proteínas y calorías se utiliza para mejorar el estado nutricional de estos niños. Estos productos costosos se pueden asociar con algunos efectos adversos, p.ej. el desarrollo de patrones de hábitos alimentarios inapropiados. Esta es una actualización de una revisión Cochrane actualizada por última vez en 2009.
Objetivos
Examinar la evidencia de que en los niños con enfermedades crónicas, la administración de suplementos orales de proteínas y calorías modifica la ingesta diaria de nutrientes, los índices nutricionales, la supervivencia y la calidad de vida y se asocia con efectos adversos, p.ej. diarrea, vómitos, reducción del apetito, intolerancia a la glucosa, distensión y problemas de hábitos alimentarios.
Métodos de búsqueda
Los ensayos de suplementos orales de proteínas y calorías en niños con enfermedades crónicas se identificaron mediante búsquedas exhaustivas en bases de datos electrónicas, búsquedas manuales en revistas pertinentes y en libros de resúmenes de actas de congresos. También se estableció contacto con las empresas que comercializan estos productos.
Búsqueda más reciente en el registro de ensayos del Grupo: 24 febrero 2015.
Criterios de selección
Ensayos controlados aleatorizados o cuasialeatorizados que compararon la administración de suplementos orales de proteínas y calorías durante al menos un mes para aumentar la ingesta calórica, con el tratamiento convencional existente (incluido el asesoramiento sobre la mejora de la ingesta nutricional a partir de los alimentos, o ninguna intervención específica), en niños con enfermedades crónicas.
Obtención y análisis de los datos
Se evaluaron de forma independiente los resultados: índices de nutrición y crecimiento; medidas antropométricas de la composición corporal; ingesta calórica y de nutrientes (total de suplementos proteicos calóricos orales y de los alimentos); hábitos alimentarios; cumplimiento; calidad de vida; efectos adversos específicos; puntuaciones de la gravedad de la enfermedad, y mortalidad; también se evaluó el riesgo de sesgo en los ensayos incluidos.
Resultados principales
Cuatro estudios (187 niños) cumplieron los criterios de inclusión. Se realizaron tres estudios en niños con fibrosis quística y un estudio incluyó niños con enfermedades malignas pediátricas. En general, hubo un bajo riesgo de sesgo para el cegamiento y los datos de resultados incompletos. Dos estudios tuvieron alto riesgo de sesgo para el ocultamiento de la asignación. Se encontraron pocas diferencias estadísticas en cuanto a los resultados evaluados entre los grupos de tratamiento y control, excepto el cambio en la ingesta energética total a los seis y 12 meses; diferencia de medias 304,86 kcal por día (intervalo de confianza del 95%: 5,62 a 604,10) y diferencia de medias 265,70 kcal por día (intervalo de confianza del 95%: 42,94 a 485,46), respectivamente. Sin embargo, estos hallazgos se basaron en el análisis de solo 58 niños en un solo estudio. Solo se incluyeron en estos análisis dos enfermedades crónicas, la fibrosis quística y la enfermedad maligna pediátrica. No se identificaron otros estudios que evaluaran la efectividad de la administración de suplementos de proteínas y calorías orales en niños con otras enfermedades crónicas.
Conclusiones de los autores
La administración de suplementos orales de proteínas y calorías se utiliza ampliamente para mejorar el estado nutricional de los niños con varias enfermedades crónicas. Se identificó un pequeño número de estudios que evaluaron estos productos en niños con fibrosis quística y enfermedad maligna pediátrica, pero no fue posible establecer conclusiones sobre la base de los limitados datos extraídos. Se recomienda que se realice una serie de ensayos controlados aleatorios grandes que investiguen la administración de estos productos en niños con diferentes enfermedades crónicas. Hasta que se disponga de más datos, se recomienda que estos productos se utilicen con precaución.
PICOs
Resumen en términos sencillos
La administración de suplementos orales de proteínas y calorías en niños con enfermedades crónicas
Antecedentes
La falta de crecimiento y la mala nutrición son frecuentes en los niños con enfermedades crónicas como la fibrosis quística y el cáncer pediátrico. Este hecho se puede deber a la reducción del apetito, a la mala absorción y a la necesidad de calorías adicionales debido a la enfermedad. La administración de suplementos orales de proteínas y calorías, ya sea en forma de leche o jugos, pueden mejorar el estado nutricional y ayudar a los niños a aumentar de peso. Los efectos secundarios de la toma de estos suplementos incluyen el riesgo de que las proteínas y las calorías del suplemento sustituyan a las de la alimentación normal y tengan un efecto negativo sobre el comportamiento alimentario y efectos secundarios físicos (p.ej. distensión, vómitos y diarrea).
Fecha de la búsqueda
La evidencia está actualizada hasta: 24 febrero 2015.
Características de los estudios
Se buscaron ensayos de suplementos orales de proteínas y calorías comparados con el tratamiento habitual o ningún tratamiento alternativo en el que los niños tomaron los suplementos durante al menos un mes. La revisión incluyó cuatro ensayos con 187 niños; en tres de ellos los niños tenían fibrosis quística y en uno tenían cáncer. Los estudios duraron de tres meses a un año. Se registraron los resultados y se valoró si los ensayos tenían riesgo de estar sesgados sobre la base del diseño o a la forma en la que se realizaron. Se observaron resultados como el peso y la talla, la ingesta de calorías, el comportamiento y también los efectos secundarios.
Resultados clave
Un estudio (con 58 niños) mostró aumentos en el total de calorías consumidas a los seis y a los 12 meses. Ninguno de los otros resultados observados mostraron diferencias entre los tratamientos. Esta es una versión actualizada de la revisión, que no encontró evidencia concluyente que apoye la administración de suplementos proteicos orales. Se recomienda que se realice al menos un ensayo de alta calidad. Por lo tanto, se recomienda que estos productos se administren con moderación y precaución.
Calidad de la evidencia
En general, los estudios incluidos tuvieron bajo riesgo de sesgo, excepto dos estudios en los que fue posible que los organizadores supieran en qué grupo de tratamiento se colocarían los niños. Es poco probable que estos problemas cambien los resultados, ya que es poco probable que el hecho de saber qué tratamiento se administra afecte los resultados de las mediciones corporales (p.ej. los resultados de peso y talla). Se informaron todos los resultados previstos, con la excepción de un estudio que no informó sobre el comportamiento alimentario y la ingesta de lipasa que se midieron. La calidad de los resultados de la evaluación del comportamiento alimentario fue cuestionable y muchos de los niños no devolvieron los diarios de alimentos, a partir de los cuales hubiera sido posible calcular la ingesta de lipasa.
Authors' conclusions
Summary of findings
| Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice compared to placebo for children with chronic disease | ||||||
| Patient or population: children with chronic disease | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control (no intervention, placebo or additional nutritional advice) | Oral protein calorie supplements | |||||
| Change in weight in kg at 6 months | The mean change in weight in kg at 6 months ranged across control groups from 1.33 to 6.55 kg | The mean change in weight in kg at 6 months in the intervention group was 0.42 higher (0.12 lower to 0.96 higher) | ‐ | 125 | ⊕⊕⊕ | |
| Change in weight Z score at 6 months | The mean change in weight Z score at 6 months ranged across control groups from 0.06 to 0.83 kg | The mean change in weight Z score at 6 months in the intervention group was 0.03 higher (0.1 lower to 0.16 higher) | ‐ | 109 | ⊕⊕⊕⊕ | |
| Change in height in cm at 6 months | The mean change in height in cm at 6 months ranged across control groups from 3.46 to 6.58 cm | The mean change in height in cm at 6 months in the intervention group was 0.4 lower (1.23 lower to 0.42 higher) | ‐ | 109 | ⊕⊕⊕⊕ | |
| Change in BMI at 6 months | The mean change in BMI at 6 months ranged across control groups from 0.14 to 1.12 kg/m2 | The mean change in BMI at 6 months in the intervention group was 0.18 higher (0.12 lower to 0.47 higher) | ‐ | 109 | ⊕⊕⊕⊕ | |
| Change in total energy intake (kcal/day) at 3 months | The mean change in total energy intake (kcal/day) at 3 months ranged across control groups from ‐56.75 to 190.54 kcal | The mean change in total energy intake (kcal/day) at 3 months in the intervention group was 133.43 higher (102.94 lower to 369.79 higher) | ‐ | 53 | ⊕⊕⊕⊕ | |
| Change in total protein intake (g/day) at 3 months | The mean change in total protein intake (g/day) at 3 months ranged across control groups from ‐1.95 to 7.87 g | The mean change in total protein intake (g/day) at 3 months in the intervention group was 3.45 higher (5.87 lower to 12.76 higher) | ‐ | 53 | ⊕⊕⊕⊕ | |
| Disease severity score: change in FEV1 % predicted at 12 months | The mean disease severity score: change in FEV1 % predicted at 12 months in the control group was ‐1.5 % | The mean disease severity score: change in FEV1 % predicted at 12 months in the intervention group was 4.91 higher (1.75 lower to 11.57 higher) | ‐ | 70 | ⊕⊕⊕⊕ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| BMI: body mass index | ||||||
Background
Description of the condition
Inadequate gains in growth, due to poor nutritional intake, increased nutrient requirements during illness, and reduced appetite are frequently observed features in children with chronic diseases. This has been demonstrated in respiratory disease, chronic renal failure, neuromuscular disease and juvenile chronic arthritis (Hanning 1993; Johansson 1986). Studies also show a prevalence of both acute and chronic malnutrition of children in hospital of between 6% to over 30% in both developed and developing countries globally (Hendricks 1995; Hendrickse 1997; Moy 1990; Pawellek 2008; Joosten 2010; Spagnuolo 2013). Growth failure and poor nutritional status in chronic disease are multi‐factorial in origin. Contributing factors include reduced dietary intake as a result of reduced appetite, malabsorption and increased nutritional requirements associated with some diseases. Poor nutritional status and sub‐optimal growth can have a detrimental effect on both short‐ and long‐term disease outcomes (Corey 1988; CPS 1994; Koscik 2004).
Description of the intervention
Prevention or correction of malnutrition is increasingly recognised as an important component of the management of chronic childhood disease. Several interventions are available to increase nutritional intake thereby improving nutritional status. These include dietetic advice to increase the nutritional content of the child's diet; provision of oral protein calorie supplements; and provision of additional nutrition in the form of tube feeding.
How the intervention might work
Oral protein calorie supplements in the form of either whole protein milk or juice drinks are used to increase the total daily protein and calorie intake in order to improve weight gain and nutritional status. These supplements also contain a range of micronutrients (vitamins and minerals) which may be of benefit to the malnourished child. Supplements which provide only calories with no additional nutrients are also available, however these are infrequently recommended for children with malnutrition as these children require supplementation of protein and other nutrients in addition to calories, rather than calories alone. Provided protein calorie supplements are taken in addition to normal dietary intake from food, then overall nutritional intake should be improved. However, it is possible that these products may replace some of the protein and calories taken as food and their potential effect on overall protein and calorie intake either reduced or eliminated. A further potential adverse consequence of replacing protein and calorie intake from normal food by protein and calories from these supplements may be to have a detrimental effect on eating behaviour which is particularly critical in toddlers and young children who are learning to develop normal eating behaviour. These products may also cause a number of other unpleasant symptoms, for example diarrhoea, vomiting, glucose intolerance and bloating. In addition, these products are expensive especially when compared calorie for calorie to food.
Why it is important to do this review
Given the potential benefits and harms described above it is important to evaluate the effectiveness of oral protein calorie supplements in children with chronic disease. This is an update of a Cochrane review last updated in 2009 (Poustie 2009).
Objectives
To examine the evidence that in children with chronic disease, oral protein calorie supplements:
-
alter nutritional indices, daily calorie, protein and other nutrient intake, survival and quality of life;
-
alter protein and calorie intake from normal food and affect overall nutrient intake;
-
are associated with adverse effects in children with chronic disease which are either important to the child or have long‐term sequelae (these may include diarrhoea, vomiting, reduced appetite, glucose intolerance, bloating and eating behaviour problems).
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (RCTs) and quasi‐randomised controlled trials, published or unpublished.
Types of participants
Children, aged between one year and 16 years, with any defined chronic disease i.e. a disease which requires medical intervention for a period of six months or more, or for the remainder of the person's lifetime. Children with malnutrition resulting from insufficient dietary intake without any causative disease, e.g. during famine, were excluded.
Types of interventions
Protein calorie supplements administered orally, in any amount and given for a period of at least one month, compared with no intervention, routine nutritional advice or placebo. A period of at least one month has been selected as we are interested in the longer‐term use of these products to improve nutritional status and growth. These products are sometimes provided in the short term to boost nutritional intake during a period when dietary intake is compromised, for example during an acute infection. Studies with an intervention period of less than one month will therefore be excluded. Oral protein calorie supplements provided as either whole protein milk or juice drinks were considered. Those which provide calories alone were not included.
Types of outcome measures
Primary outcomes
-
Change in nutritional indices
-
weight
-
height
-
body mass index
-
z score
-
other indices of nutrition (including blood biochemistry if appropriate) or growth, including anthropometric measures of body composition
-
-
Calorie and nutritional intake from food measured by diet diary or weighed food intake daily, weekly or over some other time interval
-
total energy intake
-
total fat intake
-
total protein intake
-
-
Calorie and nutritional intake from oral protein calorie supplements measured by diet diary or weighed food intake daily, weekly or over some other time interval
-
total energy intake
-
total fat intake
-
total protein intake
-
Secondary outcomes
-
Measures of eating behaviour
-
Severity scores for individual chronic diseases (e.g. FEV1)
-
Measures of quality of life
-
Measures of compliance to dietary treatment
-
Adverse effects including diarrhoea, vomiting, reduced appetite and abdominal bloating. Other adverse effects, if reported, would also be examined
-
Number of deaths or age at death in each group
Search methods for identification of studies
Electronic searches
Publications describing RCTs of the use of oral protein calorie supplements in children with chronic diseases were identified from the different trials registers held at the editorial base of the Cochrane Cystic Fibrosis and Genetic Disorders Group.
The search terms used to search this Register were:
CF: nutrition AND supplements AND (protein OR calorie supplements)
HAEM & COAG: (nutritional OR dietary) supplements
OTHERS: (diet* OR calorie* OR protein* OR nutrition*) AND supplement* (ti/abstr)
These registers are compiled from electronic searches of the Cochrane Central Register of Controlled Trials (CENTRAL) (updated each new issue of The Cochrane Library), weekly searches of MEDLINE, a search of Embase to 1995 and the prospective handsearching of two journals ‐ Pediatric Pulmonology and the Journal of Cystic Fibrosis. Unpublished work is identified by searching through the abstract books of three major cystic fibrosis conferences:‐ the International Cystic Fibrosis Conference; the European Cystic Fibrosis Conference and the North American Cystic Fibrosis Conference. For full details of all searching activities for the register, please see the relevant sections of the Cystic Fibrosis and Genetic Disorders Group Module. Date of the most recent search of the Group's trials registers: 24 February 2015.
Detailed computerised searches of MEDLINE from 1966 to October 2013 (using the search strategy described in the Cochrane Handbook for Systematic Reviews of Interventions) and of the Cochrane Central Register of Controlled Trials (Issue 9 of 12 2013) were also undertaken for this review. For the full search strategies employed, please see the additional tables attached to this review (Appendix 1; Appendix 2).
Searching other resources
Unpublished work was identified through the searching of the abstract books of the conferences of the European Society of Parenteral and Enteral Nutrition (ESPEN), 1983 to 1999, the American Society of Parenteral and Enteral Nutrition (ASPEN), 1983 and 1985 to 1999, and the British Association of Parenteral and Enteral Nutrition (BAPEN), 1995, 1997, 1998. Additional RCTs were identified from reference lists. The companies which manufacture oral protein calorie supplements were contacted to ask whether they have data on RCTs of oral protein calorie supplements for children with chronic diseases on file.
Data collection and analysis
Selection of studies
For the original review and up until November 2008, the three authors (VP, RS and RW) independently selected the studies to be included in the review according to criteria set out above and ensured they obtained the full reports of these studies if they were published. As from the 2015 update, two authors (JS and DF) independently decided on which studies to include based on the pre‐stated criteria. In both iterations the authors resolved any disagreements by discussion.
Data extraction and management
For the reviews up to 2009, two authors (VP and RS) independently extracted data and one author (VP) extracted and analysed the individual patient data from one study by calculating mean change from baseline with standard deviations (SDs) for relevant outcomes (Kalnins 1996); the authors used a standard data extraction form adapted to suit this review. From the 2015 update, two authors (JS and DF) carried out data extraction and management. The authors resolved any disagreements regarding extracted data or how it should be handled by discussion.
The authors grouped outcome data into those measured at one, three, six, 12 months and annually thereafter. If study investigators had recorded outcome data at other time periods, then the authors considered examining these as well.
Assessment of risk of bias in included studies
The authors assessed the risk of bias (low, high or unclear) for each study using the 'Risk of bias' assessment tool as documented in chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). They considered the following items:
-
sequence generation;
-
allocation concealment;
-
blinding (or masking) of participants, personnel and outcome assessors;
-
incomplete outcome data;
-
selective outcome reporting;
-
other potential sources of bias.
Again, the authors resolved any differences by discussion.
Measures of treatment effect
For binary outcome measures, the authors calculated a pooled estimate of the treatment effect for each outcome across studies using the Peto odds ratio (OR) (the odds of an outcome among treatment allocated participants to the corresponding odds among controls).
For continuous outcomes, the authors recorded either the mean change from baseline for each group or the mean post‐treatment or intervention values and SD or standard error for each group. They also calculated a pooled estimate of treatment effect by calculating the mean difference (MD). If trial reports had presented standard errors instead of SDs, the authors planned to convert these to SDs in order to enter data into RevMan (RevMan 2014).
For all outcome measures the authors also reported the relevant 95% confidence intervals (CIs).
Unit of analysis issues
If the authors include any cross‐over studies in a future update, ideally they will combine the results in a meta‐analysis using the inverse variance methods that are recommended by Elbourne (Elbourne 2002). If there are restrictions on the data available on the included trials, they will either use first‐arm data only or treat the cross‐over study as if it was a parallel study (assuming a correlation of zero as the most conservative estimate). Elbourne says that this approach produces conservative results as it does not take into account within‐patient correlation (Elbourne 2002). Also each participant appears in both the treatment and control group, so the two groups are not independent. In order to combine these results with the data entered from already included parallel studies, the authors will use the methods recommended by Curtin (Curtin 2002a; Curtin 2002b; Curtin 2002c).
Dealing with missing data
The authors planned to seek data on the number of participants with each outcome event, irrespective of compliance and whether or not the participant was later thought to be ineligible or otherwise excluded from treatment or follow up. Where outcome data or details of the study methodology were missing from the papers, the authors contacted the primary investigators for further details. If the primary investigators had not responded, the authors had planned to contact the co‐investigators.
Assessment of heterogeneity
Heterogeneity was tested using a standard Chi2 test, to assess whether observed differences in results are compatible with chance alone. The authors used the I2 test was used to assess the impact of heterogeneity on the meta‐analysis. It shows the percentage of variability in effect estimates that are due to heterogeneity rather than to chance. Values over 50% indicate a high level of heterogeneity (Higgins 2011).
If a high level of heterogeneity existed, the authors planned to discuss this narratively.
Assessment of reporting biases
In this update, it was the authors intention to investigate publication bias through use of the funnel plot. However, the review did not include the minimum number of studies (n = 10) for any analysis. Furthermore, while funnel plot asymmetry may indicate publication bias, this is not inevitably the case (Egger 1997). This issue will be considered for a future version of this review provided there is a sufficient number of included studies.
Data synthesis
The authors have presented data using MDs and 95% CIs at the pre‐specified time‐points. They have used a fixed‐effect model for the majority of the analyses; however, they analysed data on change in total dietary fat intake using a random‐effects model due to evidence of significant heterogeneity. If the authors are able to add further trials to the meta‐analysis in the future and there is significant heterogeneity between studies, they will use a random‐effects model.
Subgroup analysis and investigation of heterogeneity
To investigate any heterogeneity, if the authors identify sufficient studies (at least four), they plan to perform subgroup analyses stratifying according to type of control group(s) used, age and severity of nutritional status, disease type and type of supplement.
Sensitivity analysis
If, in future, the authors are able to include sufficient studies (at least four) in the review and combine these in a meta‐analysis; they will test the robustness of their results with a sensitivity analysis based on the risk of bias in the studies from randomisation (e.g. randomised controlled trials compared to quasi‐randomised controlled trials).
Results
Description of studies
The included studies are described below.
Results of the search
The original search identified 2105 titles of publications associated with nutrition. From the titles, studies which clearly did not include either children or a nutritional intervention were excluded, leaving 130 publications for which the abstracts were obtained. Two authors then independently selected 39 papers referring to 28 studies, for which the full reports were obtained. These publications were independently checked by both authors to assess eligibility for inclusion in the review. All these references to studies can be obtained from the former lead reviewer (VP).
The same process has been repeated for the additional studies identified in the searches carried out for the annual updates of this review. For the 2015 update of this review, 4891 titles of publication were reviewed (Figure 1) Titles for studies which did not meet the inclusion criteria (children, nutritional intervention) were excluded, as a result 58 abstracts were obtained for further review. Two authors then independently reviewed these abstracts and identified only two studies as needing to obtain full papers.
Study flow diagram for 2015 update.
For the 2015 update, three new studies were identified in the searches which are now listed in the review (Bayram 2009; Botrán 2011; Cox 2014). One of these has been included in the updated review (Bayram 2009) and one has been excluded with reasons (Botrán 2011). The third new study is still ongoing and the authors have been contacted to ascertain when the published data will be available (Cox 2014). Also, one reference that was previously listed as 'Awaiting classification' has now been included under the main study ID with confirmation of the paediatric data from the authors (Kalnins 1996). Five further studies that were previously listed as 'Awaiting classification' have now been excluded (Jain 2006; Johnson 2006; Nielson 2007; Powers 2006; Rollins 2007). One study that was previously listed as 'Awaiting classification' has been identified as part of an already excluded study (Rickard 1989).
There are currently a total of 106 studies excluded from the review.
Included studies
Study Design
Four studies of parallel design fulfilled the criteria for inclusion in this review (Bayram 2009; Hanning 1993; Kalnins 1996; Poustie 2006). The study by Hanning, which was an explanatory study, aimed to investigate the relationship between nutritional status and skeletal muscle strength (Hanning 1993). Three studies were single centre (Bayram 2009; Hanning 1993; Kalnins 1996) and one was multicentre (Poustie 2006). Two studies were conducted in Canada (Hanning 1993; Kalnins 1996), one in the UK (the 'CALICO' trial) (Poustie 2006) and one in Turkey (Bayram 2009). One study explicitly stated an open‐label design (Bayram 2009); however, the remaining three studies compared a supplement to dietary advice alone, so the participants could not be blinded (Hanning 1993; Kalnins 1996; Poustie 2006). The duration of the trials ranged from three months (Kalnins 1996) to 12 months (Poustie 2006).
Participants
Three studies were in children with cystic fibrosis (Hanning 1993; Kalnins 1996; Poustie 2006); one study included children with paediatric malignant disease (Bayram 2009). The four studies included a total of 187 participants (Bayram 2009; Hanning 1993; Kalnins 1996; Poustie 2006). The number of participants in each study ranged from 13 (Kalnins 1996) to 126 (Bayram 2009). In the Hanning study, the treatment group appeared to be in better clinical condition at baseline (Hanning 1993). Three of the studies included only children (Hanning 1993; Poustie 2006; Bayram 2009) and one study included both children and adults ‐ for this study, we obtained the individual patient data on those participants who were 16 years or less at the start of the study from the authors (Kalnins 1996).
Interventions
All four studies compared oral protein calorie supplements in addition to a normal diet to a control group. Three of these specified that the supplements were in the form of drinks (Bayram 2009Kalnins 1996; Poustie 2006) and one used drink powders, milk shakes or tinned puddings (Hanning 1993). Two studies used dietary counselling in the control group (Kalnins 1996; Poustie 2006) and two did not have any additional intervention in the control group (Bayram 2009; Hanning 1993). One study included follow up of intervention group by a nurse specialized in nutrition to check whether supplement was taken regularly (Bayram 2009).
Outcomes
The Bayram study included relevant primary and secondary outcomes for inclusion in the review, including loss in body weight, body mass index (BMI) and negative deviation from normal weight percentile; however, the authors only reported on percentage change of each outcome comparing treatment and control groups (Bayram 2009). All efforts made by the review authors to obtain individual patient data or reanalyzed results were unsuccessful.
The Hanning study examined a number of outcome measures relevant to the review (Hanning 1993). Mean change in weight from baseline for treatment and comparison groups was recorded, but results for all other outcomes were expressed as mean post‐treatment values. Since the groups were not similar at baseline, we have not included these results. We have attempted, as yet unsuccessfully, to obtain further information from the authors.
The Kalnins study did not report the results as mean change from baseline. From the individual patient data provided by the authors, we have calculated the mean change in weight, height, BMI, with z scores and percentiles, and mean change in percentage ideal weight for height (% WFH), nutritional intake and forced expiratory volume at one second (FEV1) (% predicted) from baseline (Kalnins 1996).
Poustie reported on nutritional parameters, energy and macronutrient intake, lung function, eating behaviour and activity levels (Poustie 2006).
Excluded studies
In summary, a total of 106 studies were excluded from the review. Of these, the authors identified 69 studies which were not of oral protein calorie supplements, 19 studies were not of RCT design, 10 were not in children with chronic diseases, three of the studies were of short intervention duration and five were not within a specified age range.
Risk of bias in included studies
Allocation
Generation of sequence
Generation of allocation sequence was assessed as having a high risk of bias in one study, which was described as randomised in the published manuscript, but no details of this were provided to appraise the generation of allocation sequence (Kalnins 1996). We have since been informed by the author of the published manuscript that although generation of allocation sequence was adequate for the first participant, alternation was used for subsequent participants. The study was therefore assessed as being at high risk of bias.
The Bayram study was judged to have an unclear risk of bias as it was reported that the participants were consecutively selected but no further details provided as to how this consecutive sequence was generated (Bayram 2009).
Generation of the allocation sequence was assessed as having a low risk of bias in the remaining two studies (Hanning 1993; Poustie 2006). In these studies the allocation sequence was determined by random number tables (Hanning 1993; Poustie 2006). However, in one of the studies, the treated group appeared to be in better clinical condition at baseline (Hanning 1993). To overcome this, we decided not to include any of the data from outcomes which were expressed as mean post‐treatment values in the analysis.
Allocation concealment
Concealment of allocation was assessed as having a high risk of bias in two studies (Bayram 2009; Kalnins 1996). The Bayram study was an open‐label design and therefore did not maintain concealment of allocation. The Kalnins study provided no details of the concealment of the allocation sequence. We have since been informed by the author of the published manuscript that although generation of allocation sequence was adequate for the first participant, alternation was used for subsequent participants. The person undertaking the randomisation was not blinded to the sequence of allocation.
Concealment of allocation was assessed as having a low risk of bias in the other two studies (Hanning 1993; Poustie 2006). In these studies allocation concealment was achieved using sealed envelopes (Hanning 1993; Poustie 2006).
Blinding
In none of the studies were the participants able to be blinded to the treatment: one study compared supplements to counselling (Kalnins 1996); two studies compared dietary supplements to no supplements (Bayram 2009; Hanning 1993); and the final study states that participants were not blinded as there was no satisfactory placebo available (Poustie 2006). With regards to the study investigators, in the Poustie study the research assistant was not masked to allocation group; however, a masked investigator used a computerised growth package when converting weight and height to body mass index centile (Poustie 2006). One study comparing supplements to counselling stated investigators were blinded to treatment allocation (Kalnins 1996).
We classified all four studies as being at low risk of bias despite different levels of blinding as blinding was not deemed to interfere with the outcomes assessed in the studies.
Incomplete outcome data
Intention‐to‐treat analysis was not described in one study where four out of 20 participants were lost to follow up (the time demands for testing or the travelling distance were found to be excessive).and no final data were collected (Hanning 1993). Although intention‐to‐treat analysis was not directly mentioned in the trial publication for two of the studies, the data show that none of the children were lost to follow up and that the number of participants randomised to each group was the same at the start and at the end of the study (Bayram 2009; Kalnins 1996). Statistical analysis of the CALICO Trial was by intention to treat, and no participants were lost to follow up (Poustie 2006). We consider the risk of bias for all four studies to be low.
Selective reporting
All four studies appear to report on all the outcome measures they state they have taken in the 'Methods' section of their individual papers (Bayram 2009; Hanning 1993; Kalnins 1996; Poustie 2006). In three of the studies, we were unable to compare the published study reports with the original study protocols to identify if any planned outcome measures are not reported in the full papers (Bayram 2009; Hanning 1993; Kalnins 1996); however, Dr Kalnins has confirmed that all outcomes measured in the trial were reported. In the CALICO Trial, a number of secondary outcome measures were assessed but not recorded in the key trial publication (Poustie 2006). These were assessment of eating behaviour, activity levels and lipase intake. The authors have confirmed that this was due to the poor quality of the available data on these outcomes which precluded the reporting of these data. Tricep skin‐fold measurement and mid‐upper arm circumference were assessed, but not reported, but were used to calculate mid‐arm muscle circumference which was reported. Where possible, data on these outcomes have been provided by the authors and included in this review. We consider the risk of bias for all four studies to be low (Bayram 2009; Hanning 1993; Kalnins 1996; Poustie 2006).
Effects of interventions
It should be noted that for studies with such small numbers, it may not be appropriate to use the Review Manager software to analyse the data due to the high variability associated with such small numbers (RevMan 2014). As such, caution should be given to significant results from these findings.
Primary Outcomes
We have listed the outcomes for which we have data to report. Summary statistics for significant results only are provided here.
1. Change nutritional indices
a. Weight
Three studies reported change in weight (kg) (Hanning 1993; Kalnins 1996; Poustie 2006) (Analysis 1.1) and two studies reported change in weight (z score) (Kalnins 1996; Poustie 2006) (Analysis 1.2). Two studies reported change in weight percentile, but as this measure represents the same information as z score but uses different units, and as z score is more commonly used internationally, data on percentile have not been presented (Bayram 2009; Poustie 2006). Additionally, the Bayram study reported the proportion of participants with weight loss, which was significantly higher among the usual dietary intake group compared to the treatment group at six months (P = 0.03) (Bayram 2009). This was the only study to report a significant difference between the treatment and control arms for this outcome, and did not supply data we were able to analyse.
b. Height
Two studies reported both change in height (cm) and change in height (z score) (Kalnins 1996; Poustie 2006). One study reported change in percentile, but as this measure represents the same information as z score but uses different units, and as z score is more commonly used internationally, data on percentile have not been presented (Poustie 2006). No significant differences between the treatment and control arms were identified for this outcome (Analysis 1.3; Analysis 1.4).
c. BMI
Two studies reported change in BMI and change in BMI z score or percentile, or both (Bayram 2009; Poustie 2006). Bayram reported a significantly higher loss in BMI among the usual dietary intake group compared to the treatment group at three months (P = 0.002), but again did not provide data we could analyse (Bayram 2009). For the Poutsie study, no significant difference between the treatment and control arms were identified for this outcome (Analysis 1.5; Analysis 1.6).
e. Mid‐arm muscle circumference
One study reported on mid‐arm muscle circumference (calculated from mid‐upper arm circumference and tricep skin‐fold measurements) (Poustie 2006). No significant difference between the treatment and control arms were identified for this outcome (Analysis 1.8).
3. Calorie and nutritional intake from food
a. Total energy intake
Two studies reported on total energy intake (Kalnins 1996; Poustie 2006). A significant difference in mean total energy intake at six months, MD 304.86 kcal/day (95% confidence interval (CI) 5.62 to 604.10) and at 12 months, MD 265.70 kcal/day (95% CI 42.94 to 485.46) was found to favour the treatment group (Analysis 1.9). However, the data for these outcomes came from only one study (Poustie 2006).
b. Total fat intake
Two studies reported on total fat intake (Kalnins 1996; Poustie 2006). No significant difference between the treatment and control arms were identified for this outcome (Analysis 1.10).
c. Total protein intake
Two studies reported on total protein intake (Kalnins 1996; Poustie 2006). No significant difference between the treatment and control arms were identified for this outcome (Analysis 1.11).
Secondary Outcomes
Data were reported for just two of these outcomes.
2. Severity scores for individual chronic diseases
We have decided to use lung function data to indicate disease severity in the studies including children with cystic fibrosis. Three studies reported lung function data: Hanning and Kalnins reported absolute forced expiratory volume at one second (FEV1) % predicted (Hanning 1993; Kalnins 1996); and the CALICO study reported change in FEV1 % predicted and change in forced vital capacity (FVC) % predicted (Poustie 2006). No significant difference between the treatment and control arms were identified for this outcome apart from change in FEV1 % predicted at three months in one study with 69 participants, MD ‐7.92 (95% ‐13.89 to ‐1.95) (Poustie 2006) (Analysis 1.12).
In the study by Bayram, the remission rates were significantly lower in the usual dietary care group compared to the treatment group (P = 0.036) (Bayram 2009). There were no significant differences between the usual dietary care group and the treatment group in the distribution of the febrile neutropenia attacks when assessing both the frequency and the per cent of participants who suffer these attacks (Bayram 2009).
5. Adverse effects
One study reported a gastrointestinal symptom score (Poustie 2006). No significant difference between the treatment and control arms were identified for this outcome (Analysis 1.13). Poustie also reported that there was no significant difference between the treatment and control arms for headache, OR 0.30 (95% CI 0.04 to 2.58) (Analysis 1.14).
Discussion
Summary of main results
The 2015 updated version of this review screened 4883 studies (abstract and titles) of which just one was included (Bayram 2009); one is still ongoing and may be included in a future update (Cox 2014). Overall four studies with 187 participants in total and ranging in date from 1996 to 2009, are currently included in this review (Bayram 2009; Hanning 1993; Kalnins 1996; Poustie 2006). The main outcomes of interest are summarized below. One study was not included in any of the meta‐analyses as the primary outcomes were reported in dissimilar units to the other studies (Bayram 2009).
Change in weight
Three studies reported on change in weight (kg) (Hanning 1993; Kalnins 1996; Poustie 2006); combined data from all threes studies at six months showed a mean difference in weight of 0.42 kg in the supplemented group compared to the control group, but this was not significant (Analysis 1.1).
Only two studies reported change in weight z score at six months and they found no significant difference in scores between the supplemented group and the control group (Kalnins 1996; Poustie 2006). Although the data from Bayram were not included in the meta‐analyses, the study reported that overall, loss in body weight was 6.1% in the supplemented group compared to 47.4% in the control group. Similarly, negative deviation in weight percentile was significantly lower in the supplemented group compared to the control group (Bayram 2009).
Change in height
Change in height (cm) was reported by two studies at three months (Kalnins 1996; Poustie 2006) and six months (Kalnins 1996; Poustie 2006) and by just one study at 12 months (Poustie 2006). There was no significant difference in change in height between the supplemented group and the control group at any time‐point (Analysis 1.3).
The same two studies reported mean difference in height z score at six months (109 participants); there was no significant difference in the mean change between groups, though the effect estimate favoured the usual diet group (Analysis 1.4).
Change in body mass index (BMI)
Similarly, two studies reported analysable data on the mean difference for the change in BMI at six months (Kalnins 1996; Poustie 2006); the analysis showed no significant difference between the supplemented and usual diet group (Analysis 1.5). The Bayram study reported that the loss in BMI was 12.1% in the supplemented group and 52.6% in the control group (Bayram 2009).
Analysable data for the change in BMI z score was also reported by the same two studies at three months (Kalnins 1996; Poustie 2006). They reported that there was no significant difference in the change in BMI z score between the supplemented and control group (Analysis 1.6).
Overall completeness and applicability of evidence
Oral protein calorie supplements are widely used in an attempt to improve nutritional status in people with a number of chronic diseases, at some considerable cost. It is therefore concerning that their effectiveness has been assessed by adequate clinical trials in very few chronic diseases of childhood .
In this review, just two chronic diseases were addressed in four clinical trials, which highlights the paucity of evidence for rigorous clinical trials addressing appropriate nutritional interventions for children with chronic diseases. The chronic diseases addressed in the trials included cystic fibrosis (Hanning 1993; Kalnins 1996; Poustie 2006) and paediatric malignant disease (Bayram 2009).
There is a link between cystic fibrosis and nutritional status and the role of nutritional management has long been recognized (Corey 1988). Meeting the demand of increased caloric needs in children with cystic fibrosis and cancer may be important in alleviating the increased malnutrition seen in children suffering from these diseases (Florescu 2014). From its comprehensive search of the literature, this review included an explanatory study which aimed to assess the effect of supplementation on skeletal muscle strength, rather than clinical outcomes, although these were also addressed (Hanning 1993). Two other studies were pragmatic in approach to mimic the clinical practice setting (Kalnins 1996; Poustie 2006). In the 2015 update of this review, only one additional study was included in the synthesis which addressed the potential role of oral protein and energy dense supplements in cancer‐related weight loss in children (Bayram 2009). The interventions differed between studies. In the study by Hanning, the participants in the treatment group appear to have been in better clinical condition than the control group, despite the employment of adequate randomisation methods. This is in contrast to the other studies in which both treatment and control groups were similar at baseline. It would appear, that over the last decade there have been few studies addressing the objectives of the review and fitting the inclusion criteria. Furthermore, other potentially eligible studies were run in an older age group and or focused on metabolic outcomes (such as nitrogen balance) as oppose to clinical endpoints (Botrán 2011; Rollins 2007).
No studies were identified in any other conditions (e.g. cholestatic liver disease) for which oral protein calorie supplements are routinely prescribed. The CALICO study resulted from the previous publication of the Cochrane review 'Oral calorie supplements for cystic fibrosis' (Smyth 2014). The search identified many large studies of protein and calorie supplementation for children in the developing world who are malnourished as a result of persistently low nutritional intakes. However, it would not be appropriate to extrapolate data from these studies to children in whom nutritional status is compromised as a consequence of their chronic disease. Although, knowledge of the methods used in these studies may be useful for those planning to investigate supplementation in children with chronic disease.
For reasons of heterogeneity (diversity of outcome measures) only three studies were included in the meta‐analyses (Hanning 1993; Kalnins 1996; Poustie 2006). Analyses showed no significant effect of oral protein supplementation on weight or height change in children with cystic fibrosis. These studies ranged from three months to one year of follow up. Disease severity remained unchanged but there were moderate increases in nutrient intake in those receiving oral protein calorie supplements. In contrast, the study among children with paediatric malignancies found oral protein supplements to be effective in preventing weight loss and negative deviation in weight percentiles (Bayram 2009). The relative inefficacy of oral protein supplements on the reported outcomes in cystic fibrosis may be related to clinical concomitant complications of this chronic disease and methodological issues of the trials. The complex nature of cystic fibrosis affects multiple organs and especially the nutritional status of individuals through the gastrointestinal tract, which usually results in malnutrition as disease progression worsens. The mechanism of the intervention may have been affected by gastrointestinal complications such as diarrhoea, increased metabolic rate, vomiting, fever, and other underlying nutrient deficiencies as a result of malnourished status.
This raises the question of whether prevention of weight loss or correcting nutrient deficiencies related to growth would have been a more sensitive measure of the efficacy of oral protein calorie supplements. Also, none of the studies reported on the appropriateness of the supplements in terms of palatability. None of the three studies of children with cystic fibrosis measured compliance of consumption, but the study in paediatric malignancies reported having a specialised nurse check whether supplements were taken regularly (Bayram 2009).
Quality of the evidence
There were very few published studies that examined the effect of oral protein calorie supplementation on growth in children between one and 16 years of age with chronic diseases. The four included studies only addressed two chronic diseases, malignant disease and cystic fibrosis, highlighting a current gap in the research evidence. The evidence covered in this review marks a slow but important area of work being investigated.
Judgements on the quality of the evidence for the change in weight and height was assessed to be moderate (summary of findings Table for the main comparison). Three randomised controlled trials (RCTs) (125 participants) assessed the change in weight and overall there was a greater weight change in the intervention groups as compared to the control groups. The quality of the evidence for this outcome was moderate, due to the number of studies included and the number of participants who were involved in these studies. For the change in height, two RCTs (109 participants) were assessed and, based on these results, those participants in the intervention group were shorter than those in the control arms. The quality of evidence for this outcome was also moderate due to the number of studies and participants included. For the outcome measure of BMI, only BMI z score was reported in two RCTs (109 participants). Those participants in the intervention group had a higher BMI than those in the control groups; however, the quality of evidence for this outcome was moderate due to the number of studies included and the number of participants assessed.
Potential biases in the review process
In order to reduce bias careful attention was paid to standard systematic review methodology. For example, at least two review authors were involved in every aspect of identifying potential studies, deciding on inclusion and exclusion of studies, extracting data, and conducting analyses. However, a few potential sources of bias may remain.
Publication bias
A comprehensive search of both published and unpublished sources were undertaken to find relevant studies, including making contact with companies which manufacture oral protein calorie supplements for data on RCTs of these products for children with chronic diseases which they might have had on file. It is, however, still possible that some studies were missed. Furthermore, handsearching of relevant journals was also carried out. We attempted to obtain individual patient data from one study, but were unsuccessful.
Due to the fact that this is an update of a previously conducted review, any potential bias (if it exists) included in the initial review such as screening or selection of included studies could not have been corrected.
The original review was only able to access the study protocol for just one study in order to completely assess the risk of bias of included studies (Poustie 2006). However, from the reported available publications risk of bias was assessed as appropriate for all four studies (Figure 2; Figure 3).
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Agreements and disagreements with other studies or reviews
There are very few studies addressing the efficacy of oral protein calorie supplements in children with chronic diseases and we found no existing review with which to compare our findings.
Study flow diagram for 2015 update.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 1 Change in weight in kg.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 2 Change in weight Z score.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 3 Change in height in cm.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 4 Change in height Z score.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 5 Change in body mass index.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 6 Change in body mass index Z score.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 7 Change in % ideal weight for height.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 8 Change in mid‐arm muscle circumference.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 9 Change in total energy intake (kcal/day).
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 10 Change in total fat intake (g/day).
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 11 Change in total protein intake (g/day).
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 12 Disease severity score.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 13 Gastro‐intestinal symptom score.
Comparison 1 Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice, Outcome 14 Headache.
| Oral protein calorie supplements compared with no intervention, placebo or additional nutritional advice compared to placebo for children with chronic disease | ||||||
| Patient or population: children with chronic disease | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control (no intervention, placebo or additional nutritional advice) | Oral protein calorie supplements | |||||
| Change in weight in kg at 6 months | The mean change in weight in kg at 6 months ranged across control groups from 1.33 to 6.55 kg | The mean change in weight in kg at 6 months in the intervention group was 0.42 higher (0.12 lower to 0.96 higher) | ‐ | 125 | ⊕⊕⊕ | |
| Change in weight Z score at 6 months | The mean change in weight Z score at 6 months ranged across control groups from 0.06 to 0.83 kg | The mean change in weight Z score at 6 months in the intervention group was 0.03 higher (0.1 lower to 0.16 higher) | ‐ | 109 | ⊕⊕⊕⊕ | |
| Change in height in cm at 6 months | The mean change in height in cm at 6 months ranged across control groups from 3.46 to 6.58 cm | The mean change in height in cm at 6 months in the intervention group was 0.4 lower (1.23 lower to 0.42 higher) | ‐ | 109 | ⊕⊕⊕⊕ | |
| Change in BMI at 6 months | The mean change in BMI at 6 months ranged across control groups from 0.14 to 1.12 kg/m2 | The mean change in BMI at 6 months in the intervention group was 0.18 higher (0.12 lower to 0.47 higher) | ‐ | 109 | ⊕⊕⊕⊕ | |
| Change in total energy intake (kcal/day) at 3 months | The mean change in total energy intake (kcal/day) at 3 months ranged across control groups from ‐56.75 to 190.54 kcal | The mean change in total energy intake (kcal/day) at 3 months in the intervention group was 133.43 higher (102.94 lower to 369.79 higher) | ‐ | 53 | ⊕⊕⊕⊕ | |
| Change in total protein intake (g/day) at 3 months | The mean change in total protein intake (g/day) at 3 months ranged across control groups from ‐1.95 to 7.87 g | The mean change in total protein intake (g/day) at 3 months in the intervention group was 3.45 higher (5.87 lower to 12.76 higher) | ‐ | 53 | ⊕⊕⊕⊕ | |
| Disease severity score: change in FEV1 % predicted at 12 months | The mean disease severity score: change in FEV1 % predicted at 12 months in the control group was ‐1.5 % | The mean disease severity score: change in FEV1 % predicted at 12 months in the intervention group was 4.91 higher (1.75 lower to 11.57 higher) | ‐ | 70 | ⊕⊕⊕⊕ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| BMI: body mass index | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Change in weight in kg Show forest plot | 3 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 1.1 3 months | 2 | 107 | Mean Difference (IV, Fixed, 95% CI) | 0.24 [‐0.18, 0.66] |
| 1.2 6 months | 3 | 125 | Mean Difference (IV, Fixed, 95% CI) | 0.42 [‐0.12, 0.96] |
| 1.3 12 months | 1 | 102 | Mean Difference (IV, Fixed, 95% CI) | 0.16 [‐0.68, 1.00] |
| 2 Change in weight Z score Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 2.1 3 months | 2 | 107 | Mean Difference (IV, Fixed, 95% CI) | ‐0.02 [‐0.15, 0.11] |
| 2.2 6 months | 2 | 109 | Mean Difference (IV, Fixed, 95% CI) | 0.03 [‐0.10, 0.16] |
| 2.3 12 months | 1 | 102 | Mean Difference (IV, Fixed, 95% CI) | 0.08 [‐0.06, 0.22] |
| 3 Change in height in cm Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 3.1 3 months | 2 | 107 | Mean Difference (IV, Fixed, 95% CI) | ‐0.39 [‐1.15, 0.38] |
| 3.2 6 months | 2 | 109 | Mean Difference (IV, Fixed, 95% CI) | ‐0.40 [‐1.23, 0.42] |
| 3.3 12 months | 1 | 102 | Mean Difference (IV, Fixed, 95% CI) | 0.05 [‐0.67, 0.77] |
| 4 Change in height Z score Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 4.1 3 months | 2 | 107 | Mean Difference (IV, Fixed, 95% CI) | ‐0.09 [‐0.25, 0.08] |
| 4.2 6 months | 2 | 109 | Mean Difference (IV, Fixed, 95% CI) | ‐0.10 [‐0.26, 0.07] |
| 4.3 12 months | 1 | 102 | Mean Difference (IV, Fixed, 95% CI) | 0.02 [‐0.07, 0.11] |
| 5 Change in body mass index Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 5.1 3 months | 2 | 107 | Mean Difference (IV, Fixed, 95% CI) | 0.15 [‐0.08, 0.38] |
| 5.2 6 months | 2 | 109 | Mean Difference (IV, Fixed, 95% CI) | 0.18 [‐0.12, 0.47] |
| 5.3 12 months | 1 | 102 | Mean Difference (IV, Fixed, 95% CI) | 0.08 [‐0.28, 0.44] |
| 6 Change in body mass index Z score Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 6.1 3 months | 2 | 107 | Mean Difference (IV, Fixed, 95% CI) | 0.07 [‐0.08, 0.23] |
| 6.2 6 months | 2 | 10 | Mean Difference (IV, Fixed, 95% CI) | ‐0.46 [‐1.18, 0.26] |
| 6.3 12 months | 1 | 102 | Mean Difference (IV, Fixed, 95% CI) | 0.08 [‐0.12, 0.28] |
| 7 Change in % ideal weight for height Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 7.1 3 months | 1 | 8 | Mean Difference (IV, Fixed, 95% CI) | ‐2.75 [‐9.55, 4.05] |
| 7.2 6 months | 1 | 8 | Mean Difference (IV, Fixed, 95% CI) | ‐5.5 [‐12.73, 1.73] |
| 8 Change in mid‐arm muscle circumference Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 8.1 3 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 8.2 6 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 8.3 12 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 9 Change in total energy intake (kcal/day) Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 9.1 1 month | 1 | 8 | Mean Difference (IV, Fixed, 95% CI) | 70.5 [‐249.69, 390.69] |
| 9.2 3 months | 2 | 53 | Mean Difference (IV, Fixed, 95% CI) | 133.43 [‐102.94, 369.79] |
| 9.3 6 months | 1 | 48 | Mean Difference (IV, Fixed, 95% CI) | 304.86 [5.62, 604.10] |
| 9.4 12 months | 1 | 58 | Mean Difference (IV, Fixed, 95% CI) | 265.70 [42.94, 488.46] |
| 10 Change in total fat intake (g/day) Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 10.1 1 month | 1 | 8 | Mean Difference (IV, Random, 95% CI) | 15.48 [‐12.58, 43.54] |
| 10.2 3 months | 2 | 53 | Mean Difference (IV, Random, 95% CI) | 27.31 [‐42.72, 97.33] |
| 10.3 6 months | 1 | 48 | Mean Difference (IV, Random, 95% CI) | 11.71 [‐2.99, 26.41] |
| 10.4 12 months | 1 | 58 | Mean Difference (IV, Random, 95% CI) | 8.85 [‐4.64, 22.34] |
| 11 Change in total protein intake (g/day) Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 11.1 1 month | 1 | 8 | Mean Difference (IV, Fixed, 95% CI) | 4.15 [‐24.62, 32.92] |
| 11.2 3 months | 2 | 53 | Mean Difference (IV, Fixed, 95% CI) | 3.45 [‐5.87, 12.76] |
| 11.3 6 months | 1 | 48 | Mean Difference (IV, Fixed, 95% CI) | 8.77 [‐1.24, 18.78] |
| 11.4 12 months | 1 | 58 | Mean Difference (IV, Fixed, 95% CI) | 6.82 [‐2.36, 16.00] |
| 12 Disease severity score Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 12.1 FEV1 % predicted at 3 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 12.2 Change in FEV1 % predicted at 3 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 12.3 Change in FEV1 % predicted at 6 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 12.4 Change in FEV1 % predicted at 12 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 12.5 Change in FVC % predicted at 3 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 12.6 Change in FVC % predicted at 6 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 12.7 Change in FVC % predicted at 12 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 13 Gastro‐intestinal symptom score Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 13.1 12 months | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 14 Headache Show forest plot | 1 | Peto Odds Ratio (Peto, Fixed, 95% CI) | Totals not selected | |
| 14.1 3 months | 1 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
