Intervenciones de actividad física para la salud física y mental relacionada con la enfermedad durante y después del tratamiento en pacientes con cáncer colorrectal no avanzado
Resumen
Antecedentes
El cáncer colorrectal es el tercer tipo de cáncer diagnosticado con mayor frecuencia en el mundo. El diagnóstico de cáncer colorrectal y su posterior tratamiento pueden afectar negativamente la salud física y mental del individuo. En otras poblaciones con cáncer se han demostrado los beneficios de las intervenciones de la actividad física para aliviar los efectos secundarios del tratamiento. Dado que la actividad física regular puede disminuir el riesgo de cáncer colorrectal, y que el estado físico cardiovascular es un potente factor de pronóstico del riesgo de mortalidad por cáncer y por todas las causas, las intervenciones de actividad física podría cumplir una función durante todo el proceso de control del cáncer colorrectal. La evidencia sobre la eficacia de las intervenciones de actividad física en esta población sigue siendo incierta.
Objetivos
Evaluar la efectividad y la seguridad de las intervenciones de actividad física en la salud física y mental relacionada con la enfermedad de los individuos con diagnóstico de cáncer colorrectal no avanzado, en el estadio T1‐4 N0‐2 M0, con tratamiento quirúrgico o con terapia neoadyuvante o adyuvante (es decir, quimioterapia, radioterapia o quimiorradioterapia), o ambas.
Métodos de búsqueda
Se realizaron búsquedas en el Registro Cochrane Central de Ensayos Controlados (Cochrane Central Register of Controlled Trials) (CENTRAL; 2019, número 6), además de OVID MEDLINE, otras seis bases de datos y cuatro registros de ensayos sin restricciones de idioma ni de fecha. Se examinaron las listas de referencias de las publicaciones de relevancia y se realizaron búsquedas manuales en los resúmenes de las reuniones y en los resúmenes de congresos de las organizaciones pertinentes, en busca de estudios relevantes adicionales. Todas las búsquedas se completaron entre el 6 de junio y el 14 de junio 2019.
Criterios de selección
Se incluyeron ensayos controlados aleatorizados (ECA) y ECA por grupos que compararon las intervenciones de actividad física, con la atención habitual o ninguna intervención de actividad física en adultos con cáncer colorrectal no avanzado.
Obtención y análisis de los datos
Dos autores de la revisión seleccionaron los estudios de forma independiente, realizaron la extracción de datos, evaluaron el riesgo de sesgo y calificaron la calidad de los estudios según los criterios GRADE. Se agruparon los datos para los metanálisis según la duración del seguimiento, informados como diferencias de medias (DM) o diferencias de medias estandarizadas (DME) mediante el uso de los efectos aleatorios, siempre que fuera posible, o según el modelo de efectos fijos, cuando resultó oportuno. Si no fue posible realizar un metanálisis, los estudios se resumieron en forma narrativa.
Resultados principales
Se identificaron 16 ECA con 992 participantes; 524 se asignaron a un grupo de intervención de actividad física y 468 a un grupo control con atención habitual. La media de edad de los participantes varió entre 51 y 69 años. Diez estudios incluyeron a participantes que habían finalizado el tratamiento activo, dos estudios incluyeron a participantes que recibían tratamiento activo, dos estudios incluyeron tanto a quienes todavía recibían tratamiento activo como a quienes lo habían finalizado. En dos estudios, no quedó claro si los participantes recibían el tratamiento o si lo habían finalizado. El tipo, el contexto y la duración de la intervención de actividad física variaron entre los ensayos. Tres estudios eligieron intervenciones supervisadas; cinco, intervenciones domiciliarias autodirigidas y siete estudios eligieron una combinación de programas supervisados y autodirigidos. Un estudio no informó sobre el contexto de la intervención. La duración más frecuente de la intervención fue de 12 semanas (7 estudios). El tipo de actividad física incluyó la caminata, la bicicleta, ejercicios de resistencia, yoga y ejercicios de estabilización del tronco.
La mayor parte de la incertidumbre asociada con la evaluación del sesgo del estudio se debió a la falta de claridad sobre la ocultación de la asignación y el cegamiento de los evaluadores de resultados. No fue posible cegar a los participantes y al personal. En general, la calidad de la evidencia varió de muy baja a moderada. No se agruparon los resultados de las funciones físicas en el seguimiento inmediato debido a la variación importante en los resultados y a las discordancias en la dirección del efecto. No existe certeza de que las intervenciones de actividad física mejoren la funcionalidad física en comparación con la atención habitual. No se encontró evidencia sobre el efecto de las intervenciones de actividad física comparadas con la atención habitual en la salud mental relacionada con la enfermedad (ansiedad: DME ‐0,11; intervalo de confianza [IC] del 95%: ‐0,40 a 0,18; 4 estudios, 198 participantes; I2 = 0%; y depresión: DME ‐0,21; IC del 95%: ‐0,50 a 0,08; 4 estudios, 198 participantes; I2 = 0%; evidencia de calidad moderada) en el seguimiento a corto o medio plazo. Siete estudios informaron sobre los eventos adversos. No se agruparon los eventos adversos debido a las discordancias en los reportes y en la medición. No se encontró evidencia de eventos adversos graves en los grupos de intervención ni de atención habitual. Los eventos adversos menores más frecuentes fueron el dolor de cuello, de espalda y la mialgia. Ningún estudio informó sobre la supervivencia general ni la supervivencia sin recidiva y ningún estudio evaluó los resultados en el seguimiento a largo plazo
Se encontró evidencia de efectos positivos de las intervenciones de actividad física sobre el componente de capacidad aeróbica del estado físico (DME 0,82; IC del 95%: 0,34 a 1,29; 7 estudios, 295; I2 = 68%; evidencia de calidad baja), la fatiga relacionada con el cáncer (DM 2,16; IC del 95%: 0,18 a 4,15; 6 estudios, 230 participantes; I2 = 18%; evidencia de calidad baja) y la calidad de vida relacionada con la salud (DME 0,36; IC del 95%: 0,10 a 0,62; 6 estudios, 230 participantes; I2 = 0%; evidencia de calidad moderada) en el seguimiento inmediato. Estos efectos positivos también se observaron en el seguimiento a corto plazo, pero no en el seguimiento a medio plazo. Solo tres estudios informaron de un seguimiento a medio plazo de la fatiga relacionada con el cáncer y la calidad de vida relacionada con la salud.
Conclusiones de los autores
Los hallazgos de esta revisión deben interpretarse con cautela debido al bajo número de estudios incluidos y a la calidad de la evidencia. No existe certeza de que las intervenciones de actividad física mejoren la funcionalidad física. Es posible que las intervenciones de actividad física no afecten la salud mental relacionada con la enfermedad. Las intervenciones de actividad física podrían ser beneficiosas para la capacidad aeróbica, la fatiga relacionada con el cáncer y la calidad de vida relacionada con la salud hasta los seis meses de seguimiento. Cuando se informó, los eventos adversos en general fueron menores. Para evaluar el efecto de las intervenciones de actividad física en la salud física y mental relacionada con la enfermedad y en la supervivencia de los pacientes con cáncer colorrectal no avanzado, se necesitan ECA de calidad metodológica alta y con un seguimiento a más largo plazo. Los eventos adversos deben informarse de forma adecuada.
PICO
Resumen en términos sencillos
Intervenciones de actividad física para la salud física y mental de los pacientes durante y después del tratamiento del cáncer de colon
Antecedentes
El cáncer de colon es el tercer tipo de cáncer más frecuente diagnosticado en todo el mundo. Tener un diagnóstico positivo y recibir tratamiento para el cáncer de colon puede afectar negativamente la salud física y mental del paciente. Los efectos secundarios incluyen la reducción en los niveles del estado físico y el aumento del cansancio. Los pacientes también corren el riesgo de que el cáncer regrese después del tratamiento y esto puede causar miedo y preocupación. Las investigaciones sobre los programas de actividad física en otras poblaciones con cáncer han demostrado efectos beneficiosos en la reducción de los efectos secundarios del tratamiento. Dado que las personas activas tienen menos posibilidades de desarrollar cáncer de colon, la actividad física podría ser beneficiosa para quienes reciben el diagnóstico de cáncer de colon, pero la investigación aún no es clara.
Pregunta de la revisión
Esta revisión se realizó para determinar si los programas de actividad física son beneficiosos para la salud física y mental de los pacientes con cáncer de colon y si son seguros.
Resultados clave
Se encontraron 16 estudios con 992 participantes, la evidencia está actualizada hasta junio 2019. Los participantes fueron asignados al azar para recibir un programa de actividad física o cuidados habituales (sin programa de actividad física). En los estudios incluidos, no existe certeza de que los programas de actividad física mejoren la funcionalidad física y no se encontraron efectos de los programas de actividad física comparados con la atención habitual en la salud mental relacionada con la enfermedad. No se produjeron eventos adversos graves en los ocho estudios que examinaron los eventos adversos. Hubo discordancias en la notificación y la medición de los eventos adversos. No se sabe si la actividad física mejora la supervivencia en algún punto temporal, ya que ningún estudio analizó este aspecto. Los estudios incluidos indican que los programas de actividad física podrían aumentar la capacidad aeróbica, la calidad de vida relacionada con la salud (bienestar general) y reducir la fatiga (cansancio) en el corto plazo. No existe certeza sobre los efectos a largo plazo de las intervenciones de actividad física en la funcionalidad física, la salud mental relacionada con la enfermedad, los eventos adversos, el estado físico, la fatiga (cansancio), el peso, la calidad de vida relacionada con la salud (bienestar general) y los niveles de actividad física, porque ningún estudio evaluó esto.
Calidad de la evidencia
La calidad de la evidencia se calificó como muy baja a moderada, principalmente debido al pequeño número de estudios y al bajo número de participantes, así como a las limitaciones del estudio.
Conclusión
Los hallazgos de esta revisión deben interpretarse con cautela debido al bajo número de estudios incluidos y a la calidad de la evidencia. Esta revisión muestra la necesidad de que, en el futuro, se realicen investigaciones de alta calidad y con un seguimiento a largo plazo para evaluar los efectos de las intervenciones de actividad física en la salud física y mental de los pacientes con cáncer de colon, especialmente en relación con la seguridad y la supervivencia.
Authors' conclusions
Summary of findings
| Physical activity compared with usual care in adults with non‐advanced colorectal cancer | ||||||
| Population: adults with non‐advanced colorectal cancer treated surgically or with neoadjuvant or adjuvant therapy, or both Settings: all but one study undertaken in high‐income countries. Included home‐based self‐directed and supervised physical activity programmes Intervention: aerobic or resistance training, flexibility or balance training or a combination of these, lasting at least 4 weeks Comparison: control intervention (usual care or no physical activity intervention) | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of Participants | Quality of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Assumed risk | Corresponding risk | |||||
| Usual care | Physical activity | |||||
| Physical function Assessed with: 30‐Second Chair Stand Test Follow‐up: up to 12 weeks (immediate‐term) | We did not pool results due to considerable variability and inconsistency in direction of effect. Two studies observed no difference between the physical activity and usual care group for physical function at immediate‐term follow‐up. Two other studies reported significant improvements in physical function in the physical activity group compared with usual care | 185 (4 RCTs) | ⊕⊕⊝⊝a,b | We are uncertain whether physical activity interventions improve physical function | ||
| Disease‐related mental health: depression Assessed with: HADS, CES‐D Follow‐up: more than 12 weeks to 6 months (short term) | The mean postintervention HADS for depression ranged across control groups from 2.14 to 4.72 | The mean postintervention depression in the intervention group was 0.84 (2 lower to 0.32 higher) points lower than control | 198 (4 RCTs) | ⊕⊕⊕⊝b | Scores estimated using SMD ‐0.21 (‐0.50 to 0.08)g No evidence of difference in depression in the physical activity group compared with usual care group | |
| Disease‐related mental health: anxiety Assessed with: HADS, State‐Trait Anxiety Inventory Follow‐up: more than 12 weeks to 6 months (short term) | The mean postintervention HADS for anxiety ranged across control groups from 2 to 3 | The mean postintervention anxiety in the intervention groups was 0.40 points (1.2 lower to 0.54 higher) lower than control | 198 (4 RCTs) | ⊕⊕⊕⊝b | Scores estimated using SMD ‐0.11 (‐0.40 to 0.18)g No evidence of difference in anxiety in the physical activity group compared with usual care group | |
| Overall survival (time interval between enrolment in the study and death of the person from any cause) Follow‐up: 12 months | See comment | See comment | Not estimable | The included studies did not report on overall survival | ||
| Recurrence‐free survival (time interval between date of enrolment in the study and the date when colorectal cancer recurs or another cancer occurs during the follow‐up) Follow‐up: 12 months | See comment | See comment | Not estimable | The included studies did not report on recurrence‐free survival | ||
| Adverse events Follow‐up: range 8 weeks to 11 months | 4 studies reported no adverse events, 3 other studies reported no serious adverse events with 7 participants experiencing minor adverse events in one study, 101 minor adverse events being reported in another study and 39 and 36 minor adverse events being reported in the intervention and control groups, respectively in another study. 1 study did not differentiate between serious and minor adverse events and reported 9 adverse events in the intervention group and one in the control | 305 (8 RCTs) | ⊕⊕⊝⊝c,d | |||
| Physical fitness: aerobic fitness Assessed with: 6‐minute walk test, Bruce Protocol Treadmill Test, estimated V02 peak Follow‐up: up to 12 weeks (immediate term) | The mean postintervention 6‐minute walk test score ranged across control groups from 293.7 to 588.9 | The mean postintervention physical fitness in the intervention group was 59 metres (24.5 to 93.1) higher than control | 295 (7 RCTs) | ⊕⊕⊝⊝a,e | Scores estimated using a SMD 0.82 (0.34 to 1.29)f Evidence suggests an improvement in aerobic fitness in the physical activity group compared with usual care group | |
| Cancer‐related fatigue Assessed with: FACIT‐F and FACT‐F (scale 0‐52: higher score indicates lower fatigue) Follow‐up: up to 12 weeks (immediate term) | The mean postintervention cancer‐related fatigue score ranged across control groups from 37.1 to 44 | The mean postintervention cancer‐related fatigue score in the intervention groups was | 230 (6 RCTs) | ⊕⊕⊝⊝a,b | Evidence suggests an improvement in cancer‐related fatigue in the physical activity group compared with the usual care group | |
| Health‐related quality of life (HRQoL) Assessed with: FACT‐C, FACT‐G (higher score indicates better quality of life) Follow‐up: up to 12 weeks (immediate term) | The mean postintervention FACT‐C scores ranged across control groups from 99.1 to 110.8 | The mean postintervention HRQoL in the intervention group was 6.64 (1.8 to 11.4) points higher than control | 230 (6 RCTs) | ⊕⊕⊕⊝b | Scores estimated using SMD 0.36 (0.10 to 0.62)h MID 5 to 8 points Evidence suggests an improvement in HRQoL in the physical activity group compared with the usual care group | |
| *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). CES‐D: Centre for Epidemiological Studies Depression Scale; CI: confidence interval; FACIT‐F: Functional Assessment of Chronic Illness Therapy‐Fatigue; FACT‐C: Functional Assessment of Cancer Therapy‐Colorectal; FACT‐F: Functional Assessment of Cancer Therapy‐Fatigue; FACT‐G: Functional Assessment of Cancer Therapy‐General; HADS: Hospital Anxiety and Depression Scale; HRQoL: health‐related quality of life; MID: Minmal important difference, MD: mean difference: RCT: randomised controlled trial; SD: standard deviation; SMD: standardised mean difference (used when studies assess the same outcome but measure it in a variety of ways). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded one level due to indirectness (applicability of results to those undergoing active treatment). | ||||||
Background
Description of the condition
Colorectal cancer is the third most commonly diagnosed cancer and the second leading cause of cancer death worldwide, accounting for an estimated 881,000 deaths in 2018 (GLOBOCAN 2018). Incidence and mortality rates vary globally, with higher incidence and lower mortality rates in higher‐income countries (Arnold 2017; GLOBOCAN 2018; Stewart 2014). In general, incidence is higher in men than women and is strongly linked with age, with highest incidence among people aged 65 to 74 years (Howlader 2016). Incidence is currently stabilising in high‐income countries, however a two‐fold cumulative increase in incidence is expected by 2025, due to increasing incidence in low‐ to middle‐income countries. With development, comes the adoption of more inactive lifestyles and unhealthy dietary habits; established risk factors for colorectal cancer (Stewart 2014). This is expected to increase the global burden of colorectal cancer, which may be compounded by a lack of health service resources in low‐ and middle‐income countries to deal with the escalation in incidence (Stewart 2014).
Five‐year survival from colon and rectal cancer has reached 60% or more in 22 countries worldwide (Allemani 2015). Between 1989 and 2011, colorectal cancer mortality rates decreased by more than 25% and 30% in men and women, respectively in high‐income countries in Northern and Western Europe, but increased in most Eastern European countries (Ouakrim 2015). Similar trends are evident globally, with decreasing mortality rates in high‐income countries, including Australia, Canada (Coleman 2011), the USA (Ryerson 2016), and Japan (Arnold 2017), and contrasting increasing mortality rates in low‐ and middle‐income regions, such as Latin America and the Phillipines (Arnold 2017). These disparities are not easily explained and are likely due to differences in access to diagnostic and treatment services (Haggar 2009), with advancements in treatment and early detection contributing to decreasing mortality in high‐income countries (Coleman 2011; Stewart 2014).
Although treatments are advancing, anti‐cancer therapies are associated with a range of adverse physiological and psychological side effects, which affect morbidity and mortality (Devin 2016a). Surgical resection is the primary treatment modality for stage I‐III (T1‐4 N0‐2 M0) colorectal cancer, with systemic chemotherapy or radiotherapy (more often in rectal cancer), or both, given either in the adjuvant or neoadjuvant setting in stage III and high risk stage II patients (El‐Shami 2015; Labianca 2010). Major abdominal surgery alone has been associated with declines in physical function (Schroeder 1991), and fatigue (Christensen 1982). Cancer‐related fatigue affects between 60% to 96% of people with cancer during and following chemotherapy, radiotherapy or surgery (Cramp 2012; Thomas 2014; Wagner 2004). It is a distressing symptom defined as a sense of "physical tiredness or exhaustion related to cancer or cancer treatment" (NCCN 2016), which can interfere with one's ability to carry out daily activities (Curt 2000), and negatively affect mood and quality of life (Stone 2008). Cancer‐related fatigue is present in some colorectal cancer survivors at four years following diagnosis (Schneider 2007). Physical inactivity has been identified as both a risk factor for (Bower 2014), and a consequence of (Lynch 2010) cancer‐related fatigue.
Declines in cardiorespiratory fitness can occur following treatment for colorectal cancer (Devin 2016a; West 2014a). Lower levels of cardiorespiratory fitness are linked with higher rates of cancer‐specific morbidity and mortality (Peel 2009; Schmid 2015), and can predict morbidity after colonic and rectal surgery (West 2014b; West 2014c). Furthermore, people with colorectal cancer may be susceptible to sarcopenic obesity (obesity with depleted muscle mass), which is associated with poorer functional status and poorer survival rates (Prado 2008; Wang 2017). These adverse effects, alone or in combination can impact adversely on a patient's quality of life and subsequent physical activity levels (Cramer 2014a). Colorectal cancer survivors are also at an increased risk of developing second colorectal cancers (Green 2002; Markle 2010), non‐colorectal cancers (Birgisson 2005), and other comorbidities (Denlinger 2011).
Concerns surrounding recurrence are common, affecting over half of cancer patients at one year following diagnosis (Baker 2005). Even at five years following surgery for colorectal cancer, survivors have concerns surrounding recurrence (Custers 2016). A significant minority of colorectal cancer patients and longer‐term survivors of colorectal cancer (2 or more years postdiagnosis) experience clinically meaningful levels of psychological distress, including symptoms of anxiety and depression or reduced mental well‐being (Mosher 2016). Colorectal cancer survivors report high quality of life at five years or longer postdiagnosis, but have higher rates of depression than age‐matched populations (Ramsey 2002). Psychological outcomes vary greatly in this population, poorer psychological outcomes have been linked with the presence of existing comorbidities (Lynch 2008; Ramsey 2002), worse general health (Yost 2008), and lower socioeconomic status (Ramsey 2002). Levels of anxiety and depression are reported to be higher in people who undergo surgery with adjuvant chemotherapy or radiotherapy compared with surgery alone (Pereira 2012).
Description of the intervention
Physical activity interventions were the focus of this review. Physical activity is defined as any bodily movement produced by contraction of skeletal muscle that results in energy expenditure above resting energy expenditure (ACSM 2009; Caspersen 1985). For the purpose of this review the term 'physical activity interventions' included 'exercise interventions'. Exercise is a subset of physical activity that is planned, structured and repetitive, done to improve or maintain one, or more of the components of physical fitness (ACSM 2009; Caspersen 1985). Physical activity interventions may be less structured than exercise interventions and often focus on promoting the integration of activities into daily life (e.g. gardening, walking or active travel). Physical activity interventions may be self‐directed or supervised by a healthcare professional. They can involve aerobic or resistance training, flexibility or balance training, or a combination of these, can take place in any setting and can be individual or group based, or both. No restrictions were made regarding frequency, intensity, time or type of physical activity intervention included. Interventions were included if they lasted a minimum of four weeks, this was to exclude studies on the acute effects of physical activity.
Physical activity interventions are not currently delivered as part of standard practice during or following treatment for colorectal cancer. Early postoperative mobilisation is, however, strongly recommended, as part of the Enhanced Recovery After Surgery (ERAS) guidelines following colorectal surgery, encouraging patients to be out of bed for two hours on the day of surgery and six hours per day, thereafter until discharge (Lassen 2009). The American College of Sports Medicine (Schmitz 2010), the American Cancer Society (Rock 2012), and the British Association of Sport and Exercise Science (BASES 2011) guidelines confirm that exercise can be safely performed during and following cancer treatment in the general cancer population. Specific guidance statements on physical activity interventions during and following treatment for colorectal cancer have not yet been published, due to lack of evidence on adverse effects and lack of safety data (Schmitz 2010). Side effects of treatments (cancer‐related fatigue, peripheral neuropathy, immune suppression, digestion issues, bowel dysfunction (including faecal incontinence) and urinary incontinence) may increase the risk of adverse events during physical activity. These side effects may represent barriers to physical activity participation (Denlinger 2009; Denlinger 2011; Rock 2012; Schmitz 2010). Indeed, chronic diarrhoea is a side effect that has been associated with limitations in activity and negative body image (Schneider 2007). The presence of a stoma is also associated with diminished body image (Hong 2014). These side effects have been highlighted as factors to consider when prescribing physical activity. Existing comorbidities (most commonly cardiovascular disease, musculoskeletal problems and lung or breathing problems), particularly in older people with colorectal cancer have been highlighted as other factors requiring consideration, to reduce the risk of injury and adverse events (Denlinger 2009; Rock 2012; Schmitz 2010).
How the intervention might work
Physical activity has been proposed as non‐pharmacologic intervention to attenuate the negative physiologic and psychologic effects of treatment in people with cancer (Courneya 2007; Schmitz 2005). There is a growing body of evidence from Cochrane and non‐Cochrane systematic reviews demonstrating the positive impact of physical activity both during and following cancer treatment (Galvao 2005; Knols 2005; Schmitz 2005; Speck 2010). Exercise training improves cardiorespiratory fitness and muscle strength (Schmitz 2005; Speck 2010), overall health‐related quality of life (HRQoL) (Knols 2005; Mishra 2012a; Mishra 2012b), and cancer‐related fatigue in the general cancer population during and following cancer treatment (Cramp 2012; Furmaniak 2016; Speck 2010), and physical functioning during treatment (Mishra 2012a). Through improved cardiorespiratory fitness and muscle strength, physical activity may help address the physical deconditioning associated with cancer treatments (Schmitz 2005; Speck 2010), and help manage cancer‐related fatigue (Al‐Majid 2009; Cramp 2012). Physical activity may also help the emotional and mental aspects of cancer‐related fatigue (Al‐Majid 2009; Cramp 2012). Benefits of exercise interventions on psychological well‐being (Knols 2005), anxiety and depression show positive trends but the evidence is not consistent (Cramp 2012; Furmaniak 2016; Mishra 2012a).
Cardiorespiratory fitness has been highlighted as an independent predictor of cancer mortality risk. Higher cardiorespiratory fitness is associated with a significant reduction in total cancer mortality (Schmid 2015), and colorectal cancer mortality (Peel 2009). Peel and colleagues report that men with at least a moderate fitness level had a 42% lower risk of colorectal mortality compared with men with a low cardiorespiratory fitness level. Evidence from observational studies suggest that physical activity is associated with overall and disease‐free survival (Haydon 2006; Meyerhardt 2006; Meyerhardt 2009), in both colon and rectal cancer patients.
There is consistent evidence linking physical activity to reduced colon cancer risk (Leitzmann 2015; Wolin 2009). A meta‐analysis of 52 studies found an inverse association between physical activity and colon cancer, with an overall relative risk reduction of 24% (Wolin 2009). This is consistent with findings of an earlier meta‐analysis of 19 cohort studies, which demonstrated a lower risk of colon cancer of 22% and 29% in physically active men and women, respectively (Samad 2005). Conversely, there appears to be no consistent association between physical activity and rectal cancer risk (Robsahm 2013).
The exact biological mechanisms for the observed benefit of physical activity for the prevention and secondary prevention of colorectal cancer are not fully understood. Various mechanisms have been proposed. Physical activity may reduce carcinogen exposure in the mucosa through decreased gastrointestinal transit time (Quadrilatero 2003; Slattery 2003), may alter prostaglandin levels (prostaglandins are unsaturated, free fatty acids that affect colonic function) (Quadrilatero 2003), and may alter the insulin‐like growth factor (IGF) pathway (Denlinger 2011; Fairey 2003). In people with colorectal cancer, moderate‐intensity exercise has resulted in reduced levels of urinary markers of oxidative damage (Allgayer 2008), and decreased interleukin‐1 receptor agonist (Allgayer 2004a), which may enhance immune function. Oxidative DNA damage is thought to be involved in tumour formation and may be associated with malignant transition and recurrence (Allgayer 2008). IGF‐1 is important for cellular proliferation and survival (Hursting 2010), higher levels of which may be associated with increased risk of colorectal cancer (Giovannucci 2000), but this association remains elusive. Decreases in IGF and increases in IGF‐binding proteins have been observed following exercise training in breast cancer survivors, which may be clinically relevant for the colorectal cancer population (Fairey 2003)
Physical activity may therefore be potentially effective in improving overall and recurrence‐free survival. Indeed, given that regular physical activity can decrease the risk of colon cancer and has improved cardiorespiratory fitness, muscle strength, HRQoL and cancer‐related fatigue in other cancer populations, it may be of clinical relevance for the colorectal cancer control continuum.
Why it is important to do this review
Colorectal cancer is a major public health problem. With the projected increasing incidence of colorectal cancer in low‐ and middle‐income regions, increasing mortality rates in low‐ and middle‐income countries and 3.5 million colorectal cancer survivors worldwide (Stewart 2014), there is a need to develop effective interventions that aid physical and psychological recovery, help alleviate treatment side effects and increase overall and recurrence‐free survival. The Lancet Oncology commission has prioritised the reduction in morbidity and mortality associated with cancer, with a focus on "less toxic", "cost‐effective" interventions (Sullivan 2011). There is, therefore, a need for a greater understanding of the effects of physical activity interventions on the disease‐related physical and mental health of individuals with colorectal cancer, for policy, practice and for consumers.
To date, there are currently two published, non‐Cochrane systematic reviews on exercise interventions for people with colorectal cancer (Cramer 2014b; van Rooijen 2018). In the review by Cramer 2014b no recommendations regarding exercise as a routine intervention for people with colorectal cancer were made due to insufficient evidence and lack of safety data. The review undertaken by Cramer and colleagues was limited to individuals who had completed treatment. The second review by van Rooijen and colleagues was undertaken in participants undergoing treatment, and highlighted the limited evidence of exercise training during treatment for colorectal cancer. Six out of seven studies included mixed‐cancer populations, three of these studies were not RCTs. This review is restricted to RCTs only and includes individuals who are receiving adjuvant therapy in addition to those who have finished treatment; no previous review has included such a population. This review will update current evidence and include emerging evidence in relation to physical activity interventions for individuals with colorectal cancer and so identify current evidence gaps.
Objectives
To assess the effectiveness and safety of physical activity interventions on the disease‐related physical and mental health of individuals diagnosed with non‐advanced colorectal cancer, staged as T1‐4 N0‐2 M0, treated surgically or with neoadjuvant or adjuvant therapy (i.e. chemotherapy, radiotherapy or chemoradiotherapy), or both.
Methods
Criteria for considering studies for this review
Types of studies
We considered randomised control trials (RCTs) and cluster‐RCTs comparing physical activity interventions to usual care or no physical activity intervention for inclusion in this review.
Types of participants
We included studies that evaluated the effect of physical activity interventions, on adults (aged 18 years or over), regardless of gender, diagnosed with non‐advanced colorectal cancer, staged as T1‐4, N0‐2, M0, treated surgically or with neoadjuvant or adjuvant therapy (i.e. chemotherapy, radiotherapy, chemoradiotherapy), or both. We included studies that examined physical activity interventions delivered during adjuvant therapy, following adjuvant therapy or following surgery alone. We excluded studies that included participants with other cancer types (unless outcomes for colorectal cancer were reported separately), and studies that included participants who were more than five years postdiagnosis.
Types of interventions
We compared physical activity interventions separately to no physical activity intervention or to usual care. Participants in both the control and intervention arms received the same usual care. Physical activity sessions could take place in any setting, be supervised, self‐directed or both, could be individual or group based, or a combination of both. Physical activity modalities could include aerobic or resistance training, flexibility and balance training or a combination of these. No restrictions were made regarding frequency, intensity, time or type of exercise or physical activity intervention. We only included studies with interventions that lasted a minimum of four weeks in duration, to exclude studies on the acute effects of physical activity. We excluded studies with a prehabilitation component. We included studies that provided health education materials or seminars only if the physical activity intervention was the main intervention in the study. We recorded specific details on the intervention according to the FITT‐VP (frequency intensity, time, type, volume, progression) principle (ACSM 2014). We classified physical activity intensity as mild, moderate or vigorous based on the rate of perceived exertion, heart rate or metabolic equivalents report (ACSM 2014), and used the author's classification of mild, moderate, or vigorous when a quantitative measure was unavailable.
Types of outcome measures
We extracted information for the primary and secondary outcomes at all available time points. We sought to analyse overall survival and recurrence‐free survival at 12 months, three years and five years. We analysed the other primary and secondary outcomes according to the length of follow‐up: up to 12 weeks after baseline (immediate); more than 12 weeks but less than or equal to six months after baseline (short term); more than six months but less or equal to 12 months after baseline (medium term) and more than 12 months after baseline (long term).
Primary outcomes
-
Physical function (e.g. Karnofsky Performance Status Scale; Eastern Cooperative Oncology Group Scale; timed chair rise test; timed 'Up & Go' test) or other valid instruments
-
Disease‐related mental health (e.g. Hospital Anxiety and Depression Scale (HADS); Beck Depression Index (BDI))
-
Adverse events (participants experiencing at least 1 adverse event, e.g. injury, death, adverse events resulting in discontinuation of the intervention)
Secondary outcomes
-
Overall survival (time interval between enrolment in the study and death of the person from any cause)
-
Recurrence‐free survival (time interval between date of enrolment in the study and the date when colorectal cancer recurs or another cancer occurs during the follow‐up)
-
Physical fitness (e.g. cardiorespiratory endurance (6‐minute walk test; 10‐metre shuttle walk test; V02 peak or muscle strength (dynamometry; 1 repetition maximum; 5 repetition maximum) or another valid instrument
-
Cancer‐related fatigue (e.g. Functional Assessment of Chronic Illness Therapy‐Fatigue (FACIT‐F); Schwartz Cancer Fatigue Scale (SCFS); Brief Fatigue Inventory (BFI); Piper Fatigue Scale (PFS))
-
Anthropometric measurements (e.g. weight, body mass index (BMI), body composition, waist measurement, skin‐fold measurement)
-
HRQoL (e.g. European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ‐C30); Medical Outcomes Study Short Form‐36 General Health Survey (SF‐36); Functional Assessment of Cancer Therapy–Colorectal scale (FACT‐C))
-
Levels of physical activity (e.g. physical activity questionnaires (International Physical Activity Questionnaire (IPAQ), Global Physical Activity Questionnaire (GPAQ) or objective measures of physical activity using pedometers or accelerometers)
Search methods for identification of studies
Electronic searches
We searched the following electronic databases between 6 June 2019 and 14 June 2019 up to the latest issue, with no language or date restrictions to identify relevant RCTs and cluster‐RCTs for this review.
-
The Cochrane Central Register of Controlled Trials (CENTRAL, in the Cochrane Library) (Appendix 1) (inception to present)
-
Ovid MEDLINE(R) Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE(R) Daily and Ovid MEDLINE(R) (1946 to Present) (Appendix 2)
-
Ovid Embase (1974 to present) (Appendix 3)
-
CINAHL (in EBSCOhost 1982 to present)
-
Web of Science (1970 to present)
-
PsycINFO (1806 to present)
-
Open Grey (formerly SIGLE) (1980 to present)
-
PEDro (1999 to present)
The searches were conducted by Cochrane Colorectal Cancer's Information Specialist and a review author (MAT).
Searching other resources
We searched clinical trials registries separately on 6 June 2019 for ongoing studies and study protocols of:
-
Clinical.trials.gov (www.clinicaltrials.gov);
-
the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) (apps.who.int/trialsearch/);
-
the EU Clinical Trials Register (www.clinicaltrialsregister.eu/); and
-
CenterWatch (www.centerwatch.com).
We screened reference lists of all included studies and any relevant systematic reviews identified. We handsearched conference and meeting abstracts of relevant organisations including:
-
American Society of Clinical Oncology (ASCO);
-
European Society for Medical Oncology (ESMO);
-
American College of Sports Medicine (ACSM);
-
BIT's Annual World Cancer Congress;
-
European Multidisciplinary Colorectal Cancer Congress (EMCCC);
-
European Federation for Colorectal Cancer (EFR); and
-
European Cancer Congress (ECC).
We contacted individuals or organisations for information on unpublished or ongoing studies.
Data collection and analysis
Selection of studies
We imported all records retrieved from the searches into EndNote and removed duplicates (Endnote 2016). We exported these records to covidence for screening (Covidence 2018). Two review authors (MMG and MAT) independently examined the studies identified in the literature search and screened all studies based on their titles and abstracts, excluding studies that obviously did not meet the eligibility criteria. We did not exclude studies solely on the basis of reporting outcome data. We obtained the full texts of potentially eligible studies and the two review authors (MMG and MAT) independently examined the studies. In covidence, authors coded the studies as 'include', 'exclude' or 'uncertain' based on the outlined criteria. We resolved any disagreements through discussion, where necessary involving a third review author (CC or MMC), and kept a record of decisions made. We recorded reasons for exclusion of full text articles.
Data extraction and management
Two review authors (MMG and MAT) independently extracted data from the studies that met the inclusion criteria. We recorded extracted data on an excel spreadsheet, predeveloped for this purpose. MMG and MAT piloted the data extraction form in a random sample of three studies to ensure it captured the required information. We revised the form as required. We resolved any disagreements through discussion, and where necessary referred to a third review author (CC or MMC). Extracted data were entered into the Cochrane software Review Manager 5 (RevMan 5) (Review Manager 2014) for analyses. We extracted the following data.
-
Study details; author and year of publication, country of origin, aim, design, funding source, method of randomisation, method of recruitment, trial inclusion and exclusion criteria, duration of participation, conflicts of interest/ethical concerns, risk of bias assessment.
-
Participant details; total number randomised, age, gender, comorbidities, other relevant sociodemographics, cancer stage, type of cancer treatment, ethnicity, time since diagnosis, time beyond active treatment, baseline imbalances.
-
Intervention details; exercise type, intensity, frequency, volume, setting, duration of intervention, supervised or self‐directed, details of control/comparison intervention, withdrawals and exclusion and co‐interventions.
-
Outcomes; primary and secondary relevant to this review, including adverse events, follow‐up time points, measurement tools used for outcomes.
Assessment of risk of bias in included studies
Two review authors (MMG and MAT) independently assessed each included study for risk of bias using the Cochrane 'Risk of bias' tool (Cochrane Handbook for Systematic Reviews of Interventions, Chapter 8.5.d, Higgins 2011a; Higgins 2017; Appendix 4). We assessed random sequence generation, allocation concealment, blinding of personnel and outcome assessment, completeness of outcome data, selective outcome reporting, and any other sources of bias, and judged the risk of bias as 'high', 'low' or 'unclear'. We resolved any disagreements through discussion and where necessary, through involving a third review author (CC or MMC). For each study, we detailed the risk of bias in table form alongside a statement of justification for our judgement. We summarised results in both a 'Risk of bias' summary figure and 'Risk of bias' graph. When interpreting treatment effects, we took into account the risk of bias for studies that contribute to that outcome.
Measures of treatment effect
For continuous outcomes (physical function, disease‐related mental health, physical fitness, cancer‐related fatigue, anthropometric measurements, levels of physical activity and HRQoL) we determined the mean differences (MDs) or standardised mean differences (SMDs) (in cases where different instruments were used to measure the selected outcome), in the intervention group compared with the control with 95% confidence intervals (CIs). We extracted data for final scores and change from baseline scores, if available.
In this version of the review, no outcomes were reported as time‐to‐event and we were unable to report adverse events as a dichotomous outcome. In future versions of this review for time‐to‐event outcomes (overall survival and recurrence‐free survival) we will extract hazard ratios (HRs) with standard errors, assuming that the HR is constant over time to compare the risk of death or recurrence of cancer in the treatment group with that in the control group. Where HRs are not presented, we will estimate them from reported data (e.g. Kaplan‐Meier curves, logrank observed minus expected events and the logrank variance) using methods described by Tierney and colleagues (Tierney 2007). For adverse events, we will calculate the risk ratio (RR) at individual study level by dividing the risk of an event in the intervention group by the risk of the event in the control group. We will define RRs greater than 1.0 as favouring the control group (i.e. fewer adverse events in the control group) and RRs less than 1.0 as favouring the intervention group (Deeks 2017).
Unit of analysis issues
For parallel‐group, individually randomised trials, the colorectal cancer participant was the unit of analysis in each study. No cluster‐RCTs met our inclusion criteria.
For studies reporting multiple follow‐up time points, we conducted separate meta‐analyses to reflect immediate‐, short‐, medium‐ and long‐term periods of follow‐up. For immediate‐term follow‐up we extracted data closest to the 12‐week follow‐up time point. For short‐ and medium‐term follow‐up, we extracted data closest to the six‐ and 12‐month follow‐up time point. For long‐term follow‐up, we extracted the longest time interval.
For studies with multiple arms, we included only relevant intervention arms. We combined all relevant intervention arms into a single group and combined all relevant control arms into a single group, creating a single, pair‐wise comparison.
For future versions of this review, we will extract data from cluster‐RCTs when they report appropriate analyses, adjusting for the sample size in each cluster. Where control of clustering has not been performed we will attempt to correct for the intervention effects of cluster‐RCTs by reducing the size of each study to its 'effective sample size', which is the number of the original sample size divided by the 'design effect'. We will calculate the design effect as 1 + (M‐1)* ICC, where M is the average cluster size and ICC is the intracluster correlation coefficient as described in the Cochrane Handbook for Systematic reviews of interventions, section 16.3.4 (Higgins 2011b). We will use an estimate of the ICC derived from the study (if possible), from a similar study or from a study of a similar population. If we use ICCs from other sources, we will report this and conduct sensitivity analyses to investigate the effect of variation in the ICC. If we identify both cluster‐randomised trials and individually‐randomised trials, we plan to synthesise the relevant information. We will consider it reasonable to combine the results from both if there is little heterogeneity between the study designs and the interaction between the effect of intervention and the choice of randomisation unit is considered unlikely.
Dealing with missing data
We attempted to contact authors of the included studies to request missing data on outcomes, participants and summary data via email. We documented reasons for missing data (missing at random or missing not at random) and how they were addressed. We assessed the extent to which studies analysed data according to the intention‐to‐treat principle. We assessed the level of missing data for included studies by comparing the number of participants included in the final analysis with the proportion of all participants in each study available in Characteristics of included studies. In future versions of this review, for studies at high risk of attrition bias, we will attempt to perform both the best‐case and worst‐case sensitivity analyses to assess the impact of missing data on the estimates of effect.
Assessment of heterogeneity
We evaluated clinical heterogeneity by examining diversity in participant characteristics, physical activity intervention characteristics, colorectal cancer treatment and outcomes among studies. We evaluated methodological heterogeneity by examining diversity in study designs and risk of bias. We did not pool methodologically heterogeneous studies. We visually inspected forest plots and used the Chi2 test to assess statistical heterogeneity (with P < 0.1). We used the I2 statistic to assess the percentage of variation across studies that is due to heterogeneity and not due to chance (Higgins 2003). We tentatively regarded heterogeneity as 'low' if I2 is less than 49%, 'moderate', if I2 is between 50% and 75% and 'high' if I2 is more than 75% (Deeks 2017). We investigated potential sources of statistical heterogeneity by reassessing diversity in characteristics of studies (participant, intervention, treatment and outcomes) and by means of sensitivity analysis.
Assessment of reporting biases
We attempted to control for time‐lag bias, location bias, citation bias and language bias by using a comprehensive search strategy without language or date restriction, that included searching for unpublished studies and searching trials registers. We controlled for multiple publication bias by identifying duplicate publications of the same study and grouping these together, listing them as one study. For studies published after 1 July 2005, we screened the Clinical Trials Register at the WHO ICTRP for the study protocols (apps.who.int/trialsearch) to evaluate whether selective reporting of outcomes was present (outcome reporting bias).
No analysis in this version of the review included more than 10 studies. For future versions of this review, if there are at least 10 studies included in a meta‐analysis, we will visually inspect funnel plots for asymmetry to investigate potential publication bias or small‐study effects following the recommendations in Chapter 10 of the Cochrane Handbook for Systematic Reviews of Interventions for any statistical testing of funnel plot asymmetry (Sterne 2017).
Data synthesis
We pooled results from comparable groups of studies using both fixed‐effect and random‐effects models, when appropriate. Whenever possible, we used a random‐effects model with inverse variance weighting for meta‐analyses (DerSimonian 1986), due to the nature of physical activity as a highly varied intervention. We used a fixed‐effect model when there were few studies or if the studies were small with few events. Where appropriate, we conducted a sensitivity analysis to investigate the effect of the choice of model (fixed‐effect or random‐effects) on the pooled estimate. MMG and CC conducted statistical analysis using RevMan 5 (Review Manager 2014). We considered a two‐sided P value of less than 0.05 as statistically significant. In cases where measurement tools for outcomes had the opposite direction of effect, we multiplied mean scores of a selected measurement by minus one to ensure all scales had the same direction of effect as discussed in the Cochrane Handbook for Systematic Reviews of Interventions section, 9.2.3.2 (Deeks 2017). Where data aggregation was not possible due to heterogeneity, we provided a narrative synthesis of study results. We summarised the findings of the systematic review alongside an assessment of the quality of evidence for each individual outcome using the GRADE approach (GRADE Working Group 2004).
Subgroup analysis and investigation of heterogeneity
No subgroup analyses were conducted in this version of the review due to insufficient data or inclusion of a range of treatment stages, treatment types, participant ages, gender in each individual study, or both. In future versions of this review where there are sufficient data, we will perform subgroup analyses of the effect of the intervention according to:
-
exercise and physical activity intervention characteristics (using frequency, intensity, time, type, volume progression (FITT‐VP ) to calculate metabolic equivalents/hours per week);
-
participant characteristics (gender, age ‐ over 65 years or under 65 years);
-
cancer stage (T1‐2, N0, M0), (T3‐4, N0, M0), (T1‐4, N1‐2, M0);
-
cancer type (colon or rectal);
-
treatment stage (during or post‐treatment);
-
treatment type (laparoscopic or open surgery, neoadjuvant therapy or no neoadjuvant therapy);
-
time since diagnosis (zero to one year, two to three years, four to five years).
Sensitivity analysis
We undertook sensitivity analyses to assess the robustness of results. We reanalysed data after excluding studies with high risk of bias, those that had co‐interventions (when appropriate) and studies that had not performed an intention‐to‐treat analysis. We conducted a sensitivity analysis to investigate heterogeneous results with the identification and removal of heterogeneous studies. We conducted a sensitivity analysis to investigate the effect of the choice of model (fixed‐effect or random‐effects) on the pooled estimate. In future versions of this review, for studies at high risk of attrition bias, we will conduct a best‐case/worst‐case sensitivity analysis to assess the impact of missing data on the estimates of effect. If there are any assumptions for ICC value used in cluster‐RCTs, we will perform a sensitivity analysis.
Summary of findings
We assessed the overall quality of evidence of the main review outcomes using the GRADE approach in 'summary of findings Table 1 (GRADE Working Group 2004). The 'Summary of findings' table highlights the overall quality of the body of evidence for the main review outcomes, using the GRADE criteria (study limitations (i.e. risk of bias), consistency of effect, indirectness, imprecision and publication bias). We used GRADEpro GDT 2015 software to prepare the 'Summary of findings' table. We will also present the results from the prespecified Sensitivity analysis and Subgroup analysis and investigation of heterogeneity when appropriate in 'Summary of findings' tables.
Results
Description of studies
Results of the search
See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies
We conducted the comprehensive database search between 6 June 2019 and 14 June 2019 and found 5061 records. We identified 1092 additional records upon searching clinical trial registries and handsearching references lists, conference and meeting abstracts. After removing duplicates, 3837 potential records remained. We excluded a total of 3640 records based on the title and abstract and retrieved 197 records for more detailed evaluation (Figure 1).
PRISMA flow diagram.
From these, we excluded 99 studies (127 records) as they did not meet the inclusion criteria (see 'Characteristics of excluded studies') and 16 studies (40 records) were appropriate for inclusion in the current review. In addition, nine studies are ongoing and 15 studies (21 records) are awaiting classification; we did not include these studies in the analysis presented below, but will consider them in future updates of this review. We completed all searches by 14 June 2019. Figure 1 illustrates the process of the literature search and study selection for the review based on the PRISMA template (Moher 2009).
Included studies
We included 16 studies in this review (14 journal articles and 2 dissertations) (Bourke 2011; Brown 2017; Cantarero‐Villanueva 2016; Courneya 2003; Courneya 2016; Cramer 2016; Hubbard 2016; Kim 2018; Lee 2017; Lewis 2016; McDermott 2017; Nuri 2016; Pinto 2013; Van Blarigan 2019; Van Vulpen 2016; Waart 2017). We used the main publication as the study reference. We reviewed and included information on study characteristics and outcome related data from an additional 28 publications that were secondary publications of seven of these 16 studies. We contacted 11 study authors for additional information (Bourke 2011; Brown 2017; Cantarero‐Villanueva 2016; Courneya 2003; Cramer 2016; Lee 2017; Lewis 2016; McDermott 2017; Nuri 2016; Van Vulpen 2016; Waart 2017), eight of these authors replied to information requests (Bourke 2011; Courneya 2003; Cramer 2016; Lewis 2016; McDermott 2017; Nuri 2016; Van Vulpen 2016; Waart 2017). For study characteristics and outcomes, see the Characteristics of included studies table.
Study characteristics
All 16 included studies were randomised controlled trials (RCTs). No cluster‐RCTs met our inclusion criteria. All studies except for two randomised participants to either a physical activity or usual care arm (Brown 2017; Waart 2017). These two studies included an additional study arm, that included variations in exercise volume in Brown 2017 and exercise intensity in Waart 2017. Three studies included co‐interventions (Bourke 2011; Courneya 2016; Hubbard 2016), involving healthy eating seminars and a dietary information pack (Bourke 2011), health education materials for the usual care group (Courneya 2016), and weekly education sessions on topics, including physical activity, diet stress management and cardiac specific issues (Hubbard 2016). In all, investigators allocated 992 participants (mean 62, range 18 to 273) to a physical activity intervention group (n = 524, mean 33, range 9 to 136) or a control group (n = 468, mean 29, range 9 to 137).
Participants
Participants enrolled in the studies had a diagnosis of colon or colorectal cancer, six studies included only participants with colon cancer (Bourke 2011; Brown 2017; Cantarero‐Villanueva 2016; Courneya 2016; Van Vulpen 2016; Waart 2017), whilst the remaining 10 studies included participants with colorectal cancer (Courneya 2003; Cramer 2016; Hubbard 2016; Kim 2018; Lee 2017; Lewis 2016; McDermott 2017; Nuri 2016; Pinto 2013; Van Blarigan 2019). Four studies reported the percentage of rectal cancer participants included (Kim 2018; McDermott 2017; Pinto 2013; Van Blarigan 2019). No studies with exclusively rectal cancer participants met our inclusion criteria. The majority of studies included participants with stage I‐III colorectal cancer, however Courneya 2016, Lee 2017 and Waart 2017 excluded participants with stage I cancer. In addition, Waart 2017, Courneya 2003 and Van Blarigan 2019 included a minority of participants with stage IV cancer; when contacted Waart 2017 confirmed that no stage IV participants were included, Courneya 2003 was unable to provide separate data, excluding the four stage IV participants that were included in the study. Van Blarigan 2019 included one stage IV participant (2% of total participants included). Ten studies included participants who had finished active treatment (Bourke 2011; Brown 2017; Cantarero‐Villanueva 2016; Courneya 2016; Kim 2018; Lee 2017; Lewis 2016; McDermott 2017; Pinto 2013; Van Blarigan 2019), the time beyond treatment ranged between two months and five years. Two studies included participants receiving active treatment (Van Vulpen 2016; Waart 2017). Two studies were conducted among participants who were receiving and finished active treatment (Cramer 2016; Hubbard 2016). It was unclear whether participants were receiving or finished treatment in two studies (Courneya 2003; Nuri 2016). The majority of participants had undergone surgery as treatment; chemotherapy was also common across studies with less participants receiving radiotherapy. Mean time since diagnosis was only reported in six studies and ranged between 10 weeks and 2.99 years (Cantarero‐Villanueva 2016; Courneya 2016; Lewis 2016; Pinto 2013; Van Blarigan 2019; Van Vulpen 2016).
The mean age of participants ranged between 51 and 69 years. Fifteen studies included both males and females, with one study including only males (Nuri 2016). Comorbidities and ethnicity were largely unreported in studies. Only two studies reported comorbidities at baseline (Brown 2017; Waart 2017). Two studies reported on ethnicity of participants (Brown 2017; Pinto 2013), in both studies the majority of participants were white. Eight studies reported on education levels of participants (Cantarero‐Villanueva 2016; Courneya 2003; Cramer 2016; Kim 2018; Lee 2017; Van Blarigan 2019; Van Vulpen 2016; Waart 2017) with Kim 2018, Lee 2017, Van Vulpen 2016 and Waart 2017 further reporting on martial status and McDermott 2017, Van Blarigan 2019 and Waart 2017 reporting on employment status. Five studies reported recruiting those who were currently inactive (Bourke 2011; Courneya 2016; Lee 2017; Lewis 2016; McDermott 2017).
Interventions
Type and setting of interventions varied across studies. Three studies opted for exclusively supervised physical activity interventions (Cantarero‐Villanueva 2016; Cramer 2016; Hubbard 2016), likely due to the modes of exercise which included hatha yoga, core stabilisation exercise and cardiac rehabilitation exercise classes. The settings for these interventions were not clearly reported. Five studies opted for exclusively home‐based self‐directed programmes (Brown 2017; Courneya 2003; McDermott 2017; Pinto 2013; Van Blarigan 2019), which consisted of mainly aerobic physical activity (e.g. treadmill walking, cycling). A combination of aerobic and resistance exercise was prescribed in two studies (McDermott 2017 ; Van Blarigan 2019). The Nuri 2016 study involved aerobic physical activity, whether the physical activity was supervised or self‐directed is unclear. Seven studies opted for a combination of supervised and self‐directed physical activity (Bourke 2011; Courneya 2016; Kim 2018; Lee 2017; Lewis 2016; Van Vulpen 2016; Waart 2017), with physical activity logs in Lee 2017 and telephone support in Courneya 2016 and Kim 2018. A combination of aerobic and resistance physical activity was conducted in four of these studies (Bourke 2011; Lee 2017; Van Vulpen 2016; Waart 2017), with Courneya 2016 encouraging activity based on individual preference, with a walking prescription if individuals had no preference.
The intensity of the physical activity varied slightly between studies. Methods used to measure intensity of the physical activity included relatively objective measures, such as percentage of maximum heart rate, heart rate at ventilatory threshold, percentage of predicted maximum workload and ratings of perceived exertion. The majority of studies opted for moderate‐intensity physical activity (Bourke 2011; Brown 2017; Courneya 2003; Lee 2017; Lewis 2016; McDermott 2017; Nuri 2016; Pinto 2013; Waart 2017), with three studies incorporating vigorous physical activity (Lewis 2016; Van Blarigan 2019; Waart 2017). An arm of the Waart 2017 study participated in low‐intensity exercise. Five studies did not report intensity of the physical activity programme (Cantarero‐Villanueva 2016; Courneya 2016; Cramer 2016; Hubbard 2016; Kim 2018).
The most common duration of physical activity intervention was 12 weeks (Bourke 2011; Kim 2018; Lee 2017; Lewis 2016; McDermott 2017; Pinto 2013; Van Blarigan 2019). In one study, the length of the intervention was determined by duration of chemotherapy, with participants beginning the intervention with the first cycle of chemotherapy and finishing three weeks after the last cycle (Waart 2017). For another study (Hubbard 2016), the length of intervention varied depending on hospital site (6, 10 or 12 weeks). Two studies delivered eight‐week interventions (Cantarero‐Villanueva 2016; Nuri 2016). The duration of the other five studies were 10 weeks (Cramer 2016), 16 weeks (Courneya 2003), 18 weeks (Van Vulpen 2016), six months (Brown 2017), and three years (Courneya 2016). However, the Courneya 2016 study is ongoing and we have extracted one‐year interim data for this review. All studies conducted follow‐up assessments on completion of the exercise intervention. Nine studies conducted a further set of equivalent assessments at a later time point (Cantarero‐Villanueva 2016; Cramer 2016; Hubbard 2016; Lewis 2016; McDermott 2017; Nuri 2016; Pinto 2013; Van Vulpen 2016; Waart 2017). Pinto 2013 was the only study to conduct three postintervention assessments, these took place at three, six and 12 months.
Control groups
The control groups received usual care, were not prescribed physical activity and did not take part in any formal exercise training during the course of the intervention. Five studies provided participants with written information on maintaining a healthy lifestyle (Courneya 2016; Hubbard 2016; Pinto 2013; Van Blarigan 2019), or provided recommendations for a healthy lifestyle (Cantarero‐Villanueva 2016). Three studies reported asking participants in the control arm to maintain their usual daily physical activity levels/lifestyle habits during the intervention (Brown 2017; Lee 2017; Van Vulpen 2016). One study instructed participants not to initiate any structured exercise over the course of the intervention (Courneya 2003). Three studies had a waiting list control group, providing participants the opportunity to take part in the physical activity intervention following completion of the study (Brown 2017; Cramer 2016; Lewis 2016). Van Blarigan 2019 offered a fit bit flex to control group participants following study completion. Pinto 2013 controlled for frequency of contact with participants by having a contact control group.
Outcome measures
The most frequently assessed outcome among the 16 included studies was physical fitness, measured in 12 studies (Bourke 2011; Brown 2017; Cantarero‐Villanueva 2016; Courneya 2003; Courneya 2016; Lee 2017; Lewis 2016; McDermott 2017; Nuri 2016; Pinto 2013; Van Vulpen 2016; Waart 2017), the most commonly used tools to measure physical fitness were the six‐minute walk test (Brown 2017; Cantarero‐Villanueva 2016; Courneya 2016; Lee 2017; Lewis 2016), and treadmill tests using a variety of protocols (Courneya 2003; Lewis 2016; Pinto 2013; Van Vulpen 2016; Waart 2017). Twelve studies measured levels of physical activity, objectively, using accelerometer data (Brown 2017; Hubbard 2016; Lewis 2016; McDermott 2017; Pinto 2013; Van Blarigan 2019), and subjectively, using a variety of self‐report questionnaires (Bourke 2011; Courneya 2003; Courneya 2016; Kim 2018; Lee 2017; Lewis 2016; Pinto 2013; Waart 2017). Lewis 2016 and Pinto 2013 measured levels of physical activity, both objectively and subjectively. Ten studies assessed anthropometric measurements (Bourke 2011; Brown 2017; Cantarero‐Villanueva 2016; Courneya 2003; Courneya 2016; Lee 2017; Lewis 2016; McDermott 2017; Nuri 2016; Van Vulpen 2016). All studies except for Courneya 2003 measured weight and only two did not measure body mass index (BMI) (Courneya 2016; Van Vulpen 2016). Fatigue and health‐related quality of life (HRQoL) were assessed in nine studies, out of the 16 included (Bourke 2011; Courneya 2003; Cramer 2016; Kim 2018; Lewis 2016; McDermott 2017; Pinto 2013; Van Vulpen 2016; Waart 2017), most frequently using the Functional Assessment of Cancer Therapy – Fatigue (FACT‐F) scale and Functional Assessment of Cancer Therapy – Colorectal (FACT‐C) scale, respectively. Adverse events were reported in eight studies, of these the absence of any adverse events were reported in four studies (Hubbard 2016; Lewis 2016; Van Vulpen 2016; Waart 2017), whilst four studies recorded the number of adverse events that occurred (Brown 2017; Cantarero‐Villanueva 2016; Cramer 2016; Van Blarigan 2019). Six studies assessed facets of mental health and well‐being (Courneya 2003; Cramer 2016; Kim 2018; McDermott 2017; Van Vulpen 2016; Waart 2017), mainly anxiety and depression, using the Hospital Anxiety and Depression Scale (HADS) score. Other outcomes including overall and recurrence‐free survival were not reported in any of the included studies. Adherence to Enhanced Recovery After Surgery (ERAS) guidelines and length of hospital stay was not reported in any of the included studies. For detailed information on outcome measures, see the Characteristics of included studies table.
Excluded studies
We excluded 99 trials from the review due to the following reasons.
-
Wrong patient population (n = 19)
-
Wrong study design (n = 25)
-
Wrong comparator (n = 18)
-
Wrong intervention (n = 16)
-
Intervention too short (n = 10)
-
Did not analyse colorectal cancer patients separately (n = 8)
-
Study was not carried out (n = 2)
-
Outcomes were not relevant (n = 1)
See Characteristics of excluded studies for an overview.
Risk of bias in included studies
We assessed the risk of bias for each included study using the ‘Risk of bias’ assessment tool and recommendations for judging risk of bias provided in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). See Figure 2 for an overall assessment of risk of bias presented as percentages across all included studies. In addition, Figure 3 provides a 'Risk of bias' summary for each included study. Due to the nature of the intervention, it was expected that blinding of participants and personnel delivering the interventions would not be possible. We therefore judged risk of performance bias as high in all included studies. See Characteristics of included studies.
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.
Allocation
Random sequence generation
Twelve studies (75%) were at low risk of selection bias owing to adequate generation of a randomised sequence (Bourke 2011; Cantarero‐Villanueva 2016; Courneya 2003; Cramer 2016; Hubbard 2016; Kim 2018; Lee 2017; Lewis 2016; McDermott 2017; Van Blarigan 2019; Van Vulpen 2016; Waart 2017). We considered four studies to have an unclear risk of selection bias as they did not describe the generation of a randomised sequence (Brown 2017; Courneya 2016; Nuri 2016; Pinto 2013).
Allocation concealment
Five studies (36%) were at low risk of selection bias owing to adequate concealment of allocation, so that participants and investigators could not foresee assignment to the study groups (Cantarero‐Villanueva 2016; Cramer 2016; Lewis 2016; McDermott 2017; Van Vulpen 2016). The other 11 studies were considered to have an unclear risk of selection bias owing to allocation concealment as they did not describe the method of concealment.
Blinding
Blinding of participants and personnel
Due to the nature of physical activity interventions, blinding of participants and personnel is not possible. Therefore, we judged all included studies at high risk of performance bias.
Blinding of outcome assessors
Six studies were at low risk of detection bias as outcome assessors were blinded to participants group assignment (Bourke 2011; Brown 2017; Courneya 2003; Cantarero‐Villanueva 2016; Pinto 2013; Van Vulpen 2016). We considered seven studies to have unclear risk for detection bias, as blinding of outcome assessors was not described (Courneya 2016; Cramer 2016; Hubbard 2016; Kim 2018; Lee 2017; Nuri 2016; Waart 2017). Three studies were at high risk for detection bias as the outcome assessor was aware of participants' group allocation (Lewis 2016; McDermott 2017; Van Blarigan 2019).
Incomplete outcome data
All included studies reported on adherence. All studies except for Hubbard 2016 were at low risk of attrition bias due to the amount, nature or handling of incomplete outcome data. Hubbard 2016 was at high risk of attrition bias due to the amount of missing data. Adherence to physical activity interventions in other studies varied between 71% and 97%.
Selective reporting
Ten studies were at low risk of reporting bias (Bourke 2011; Courneya 2003; Courneya 2016; Cramer 2016; Kim 2018; Lee 2017; Lewis 2016; McDermott 2017; Pinto 2013; Van Vulpen 2016). Six studies were at high risk of reporting bias, as study protocols or methods sections included outcomes which were not reported in available publications (Brown 2017; Cantarero‐Villanueva 2016; Hubbard 2016; Nuri 2016; Van Blarigan 2019; Waart 2017).
Other potential sources of bias
Baseline imbalances
Eleven studies were at low risk of selection bias owing to the absence of significant imbalances between group at baseline (Brown 2017; Cantarero‐Villanueva 2016; Courneya 2003; Courneya 2016; Kim 2018; Lee 2017; Lewis 2016; Nuri 2016; Van Vulpen 2016), or in studies were baseline imbalances were present, appropriate allocation concealment was described (Cramer 2016; McDermott 2017). Four studies were at high risk of selection bias because group similarity at baseline was inadequate (Hubbard 2016; Pinto 2013;Van Blarigan 2019; Waart 2017). The risk of selection bias owing to baseline imbalances was unclear in one study as baseline imbalances were not reported (Bourke 2011). All included studies are at risk of participation bias, with the potential for the more motivated participants agreeing to participate.
Effects of interventions
See: summary of findings Table 1. For a summary of sensitivity analyses see: Table 1.
| Outcome | Time point | No. of studies | No. of participants | Statistical method | Effect size |
|---|---|---|---|---|---|
| 1 Physical functiond | |||||
| 1.1 Subjective measure of physical function | Short‐term follow‐up | 2 | 114 | SMD (IV, random, 95% CI) | 0.08 (‐0.31 to 0.47)a |
| 2 Disease‐related mental health | |||||
| 2.1 Anxiety | Short‐term follow‐up | 3 | 177 | SMD (IV, random, 95% CI) | ‐0.29 (‐0.60 to 0.01)a,b |
| 2.2. Depression | Short‐term follow‐up | 3 | 177 | SMD (IV, random, 95% CI) | ‐0.18 (‐0.48 to 0.13)a,b |
| 3 Physical fitness | |||||
| 3.1 Aerobic fitness | Immediate‐term follow‐up Short‐term follow‐up | 4 5 | 207 187 | SMD (IV, random, 95% CI) | 0.38 (0.06 to 0.70)a,b 0.45 (0.15 to 0.75)a,b |
| 4 Cancer‐related fatigued | Immediate‐term follow‐up Short‐term follow‐up | 4 5 | 169 224 | MD (IV, random, 95% CI) SMD (IV, random, 95% CI) | 2.22 (‐0.34 to 4.79)a,b 0.32 (‐0.04 to 0.67)a,b |
| 5 Anthropometric measuresd | |||||
| 5.1 Weight | Immediate‐term follow‐up Change from baseline to 12 weeks follow‐up | 4 2 | 207 64 | MD (IV, random, 95% CI) | 0.27 (‐2.87 to 3.42)a,c ‐1.76 [‐4.06 to 0.54]c |
| 5.2 Waist to hip ratio | Immediate‐term follow‐up | 2 | 44 | MD (IV, random, 95% CI) | 0.04 [‐0.01 to 0.10]a,c |
| 5.3 BMI | Immediate‐term follow‐up Change from baseline to 12 weeks follow‐up | 4 2 | 207 64 | MD (IV, random, 95% CI) | 0.10 [‐0.87 to 1.06]a,c ‐0.42 [‐1.30 to 0.46] |
| 5.4 Body fat % | Immediate‐term follow‐up | 3 | 187 | MD (IV, random, 95% CI) | ‐2.13 [‐4.46 to 0.21]a,d |
| 6 HRQoL | Immediate‐term follow‐up | 4 | 169 | SMD (IV, random, 95% CI) | 0.37 [0.07 to 0.68]a,b,c |
| 7 Levels of physical activity | |||||
| 7.1 Objective measures | Immediate‐term follow‐up | 3 | 80 | MD (IV, random, 95% CI) | ‐2.84 [‐12.40 to 6.73]c |
| 7.2 Subjective measures | Immediate‐term follow‐up | 3 | 138 | SMD (IV, random, 95% CI) | 0.68[0.33 to 1.02]c |
BMI: body mass index; CI: confidence interval; HRQoL: health‐related quality of life; MD: mean difference: SD: standard deviation; SMD: standardised mean difference (used when studies assess the same outcome but measure it in a variety of ways).
a Removal of studies that did not conduct an ITT analysis
b Exclusion of studies at high risk of bias
c Exclusion of studies with an additional intervention component
d Results from choice of model (fixed or random) were consistent
All trial authors reported study results as follow‐up values and six studies also included change in score from baseline to follow‐up (Bourke 2011; Brown 2017; Cantarero‐Villanueva 2016; Courneya 2016; Hubbard 2016; Lewis 2016). We completed meta‐analyses for both types of outcomes separately and for each follow‐up time period where data were available. We categorised follow‐up as: up to 12 weeks after baseline (immediate); more than 12 weeks but less than or equal to six months after baseline (short term); more than six months but less than or equal to 12 months after baseline (medium term); and more than 12 months after baseline (long term). No included studies reported follow‐up of greater than 12 months after baseline. One study reported follow‐up time points to end of chemotherapy treatment and at six months following completion of chemotherapy treatment (Waart 2017). We calculated follow‐up time points in months, using the reported percentage of chemotherapy treatment received. This was as a proportion of the average total planned duration of chemotherapy. Where studies had two intervention arms (Brown 2017; Waart 2017), or reported gender separately (Van Vulpen 2016), we combined these arms in RevMan to form a single pair‐wise comparison (Review Manager 2014).
Primary outcomes
Physical function
A total of 10 studies reported on physical function, assessed using a variety of measures, including the 30‐Second Chair Stand Test (Bourke 2011; Courneya 2016; Lee 2017; Lewis 2016; McDermott 2017), the Physical Functioning subscale of the Short Form‐36 (SF‐36) (Brown 2017; Pinto 2013), the functional well‐being subscale of the Functional Assessment of Cancer Therapy‐Colorectal (FACT‐C; Courneya 2003), the physical function subscale of the European Organisation for Research and Treatment of Cancer Quality of life Questionnaire‐Core 30 (EORTC QLQ‐C30; Waart 2017), and the Trial Outcome Index‐physical/functional/colorectal (Kim 2018). We conducted separate meta‐analyses for objectively and subjectively measured physical function.
Four studies including 185 participants measured physical function using the 30‐Second Chair Stand Test at immediate‐term follow‐up (Bourke 2011; Lee 2017; Lewis 2016; McDermott 2017), we did not pool these results in a meta‐analysis due to considerable variation in results and inconsistency in direction of effect. Lewis 2016 and McDermott 2017 observed no difference between the physical activity and usual care groups for physical function at immediate‐term follow‐up, whilst, Bourke 2011 and Lee 2017 reported significant improvements in physical function in the physical activity group compared with usual care (P = 0.003 and P = 0.005, respectively). We observed no evidence of difference between groups at short‐term follow‐up when measured objectively (mean difference (MD) 0.76, 95% confidence interval (CI) ‐1.84 to 3.36; 2 studies, 39 participants; I2 = 0%; moderate‐quality evidence; Analysis 1.1; Figure 4), and subjectively at short‐term follow‐up (standardised mean difference (SMD) 0.09, 95% CI ‐0.24 to 0.42; 3 studies, 156 participants; I2 = 0%; low‐quality evidence; Analysis 1.2; Figure 5). There were insufficient data to analyse subjective and objective physical function at medium‐term follow‐up, subjective physical function at immediate‐term follow‐up, and change from baseline results at all four follow‐up time points.
Forest plot of comparison: 1 Physical activity versus usual care for physical function (30‐sec chair sit‐to‐stand test), outcome: 1.2 Objective measures more than 12 weeks to six months follow‐up (short term).
Forest plot of comparison: 1 Physical activity versus usual care for physical function, outcome: 1.3 Subjective measures more than 12 weeks to six months follow‐up (short term).
Courneya 2016 reported that the structured exercise group (n = 99) improved relative to the health education material group (n = 98) for the 30‐Second Chair Stand Test (mean between group difference (+1.6 repetitions, 95% CI +0.6 to +2.7); P < 0.001) at one year. Pinto 2013 reported no significant group differences at immediate‐ or medium‐term follow‐up for self‐reported physical functioning (n = 23 and n = 19 for physical activity and usual care, respectively). Brown 2017 reported improvements in change from baseline scores for physical function at short‐term follow‐up with higher doses of physical activity resulting in greater improvements. Kim 2018 reported no difference between groups at immediate‐term follow‐up for physical function (P = 0.254). However, change from baseline to immediate‐term follow‐up improved in the intervention group (mean 64.1 (11.2) versus 66.3 (11.8), P = 0.035).
We performed sensitivity analyses to investigate the choice of model on the pooled estimate at all included time points. Sensitivity analyses using a fixed‐effect model were consistent with findings from a random‐effects model. The sensitivity analysis revealed no difference in effect when we removed studies that did not use an intention‐to‐treat (ITT) analysis; this was only possible for subjective measures at short‐term follow‐up (Table 1).
Disease‐related mental health
A total of seven studies reported on disease‐related mental health, assessed by the Hospital Anxiety and Depression Scale (HADS) (Cramer 2016; Van Vulpen 2016; Waart 2017), the Positive and Negative Affect Scale (PANAS) (McDermott 2017), the State‐Trait Anxiety Inventory (Courneya 2003), the Centre for Epidemiological Studies Depression Scale (CES‐D) (Courneya 2003), and the Patient Health Questionnaire (Kim 2018). We obtained postintervention data and separate scores for anxiety and depression through email correspondence with authors Van Vulpen 2016 and Waart 2017, respectively. We conducted meta‐analyses for anxiety and depression separately at 12 weeks to six months after baseline and more than six months to 12 months after baseline.
There was no evidence of difference in depression between the physical activity group and usual care group at short‐term follow‐up (SMD ‐0.21, 95% CI ‐0.50 to 0.08; 4 studies, 198 participants; I2 = 0%; moderate‐quality evidence; Analysis 2.1; Figure 6), or medium‐term follow‐up (assessed using HADS) (MD ‐1.20, 95% CI ‐2.72 to 0.31; 2 studies, 48 participants; I2 = 32%; low‐quality evidence; Analysis 2.2; Figure 7).
Forest plot of comparison: 2 Physical activity versus usual care for disease‐related mental health, outcome: depression 2.1 More than 12 weeks to six months follow‐up (short term).
Forest plot of comparison: 2 Physical activity versus usual care for disease‐related mental health (Hospital Anxiety and Depression Scale (HADS)), outcome: depression 2.2 More than six months to 12 months follow‐up (medium term).
There was no evidence of difference in anxiety between the physical activity group and usual care group at short‐term follow‐up (SMD ‐0.11, 95% CI ‐0.40 to 0.18; 4 studies, 198 participants; I2 = 0%; moderate‐quality evidence; Analysis 2.3; Figure 8), or medium‐term follow‐up (assessed using HADS) (anxiety: MD 1.79, 95% CI ‐0.37 to 3.94; 2 studies, 47 participants; I2 = 30%; low‐quality evidence; Analysis 2.4; Figure 9). Data were insufficient for analysis of immediate‐term follow‐up and change scores at all four time points.
Forest plot of comparison: 2 Physical activity versus usual care for disease‐related mental health, outcome: 2.3 Anxiety more than 12 weeks to six months follow‐up.
Forest plot of comparison: 2 Physical activity versus usual care for disease‐related mental health (Hospital Anxiety and Depression Scale (HADS)), outcome: 2.4 Anxiety more than six months to 12 months follow‐up (medium term).
McDermott 2017 (unpublished data) reported no statistically significant differences between intervention (n = 11) and control (n = 9) for PANAS‐positive affect and PANAS‐negative affect at immediate‐ or short‐term follow‐up. There were 10 participants in the control group at 24 weeks follow‐up for PANAS‐negative affect. Kim 2018 reported improvements in change from baseline scores for depression at immediate‐term follow‐up in the intervention group (n = 37) (P = 0.053).
A sensitivity analysis including studies at low risk of bias and those that performed an ITT analysis suggest there may be no effect of physical activity on anxiety (SMD ‐0.29, 95% CI ‐0.60 to 0.01; 3 studies, 177 participants; I2 = 0%) or depression (SMD ‐0.18, CI ‐0.48 to 0.13; 3 studies, 177 participants; I2= 0%) in the physical activity group compared to usual care at short‐term follow‐up (Table 1).
Adverse events
Eight out of the 16 included studies reported on adverse events (Brown 2017; Cantarero‐Villanueva 2016; Cramer 2016; Hubbard 2016; Lewis 2016; Van Blarigan 2019; Van Vulpen 2016; Waart 2017). We did not pool adverse events due to inconsistency in reporting and measurement. Of the eight studies, four studies reported that no adverse events (Hubbard 2016; Lewis 2016; Waart 2017) or serious adverse events (Van Vulpen 2016) occurred during the study period. The method of measurement of adverse events was not described by authors, Van Vulpen 2016 and Waart 2017. Van Vulpen 2016 and Lewis 2016 did not make reference to 'non‐serious' adverse events. Hubbard 2016 recorded adverse events in a participant log and recorded them as "related" or "unrelated" to the study. In the study by Lewis 2016, serious adverse events were recorded by the researcher within 24 hours of becoming aware of the event.
Cramer 2016 recorded all adverse events that occurred during the study period and asked open‐ended questions at weeks 10 and 22 to assess any adverse events not previously mentioned. No serious adverse events occurred and seven participants reported minor adverse events in the intervention group, including transient abdominal pain (n = 1), muscle soreness (n = 3), neck pain (n = 1), minor vertigo (n = 1) and hip pain (n = 1). One patient in the control group experienced a serious adverse event that was reported as "probably not causally related to the study intervention". Brown 2017 assessed adverse events using the Common Terminology Criteria for Assessing Adverse Events (CTCAE). No serious (grade 3) adverse events were reported at six months after baseline. One hundred and one non‐serious (grade 1 and 2) adverse events occurred in the intervention compared to 49 in the usual care group. Common non‐serious adverse events reported in the intervention group were: joint pain, back pain, generalised flu‐like symptoms, foot blisters and myalgia. In the study by Cantarero‐Villanueva 2016 each participant kept a diary to record adverse events. Six participants expressed both neck and abdominal discomfort with some of the exercises in the first sessions. Two participants in the intervention group and one participant in the usual care group experienced postoperative ventral hernias. One participant could not perform the aerobic exercise during one week because he suffered a peripheral neuropathy. The author did not report whether these were recorded as "related" or "unrelated" to the exercise intervention. In the study by Van Blarigan 2019 participants completed an online health check survey which recorded adverse events. Commonly reported non‐serious adverse events included low back, knee, joint, muscle and chest pain, inflammation of joints in both the intervention (n = 39) and control (n = 36) groups.
We graded the evidence of the included studies as low quality (8 studies, 305 participants).
Secondary outcomes
Overall survival
No RCTs reported overall survival.
Recurrence‐free survival
No RCTs reported recurrence‐free survival as an outcome. No studies reported cancer recurrence as reasons for dropout. One study reported lung metastasis in the intervention group as a reason for dropout (Lewis 2016).
Physical fitness
Aerobic fitness
A total of 12 studies reported aerobic fitness, using a variety of different measures including; the six‐minute walk test (Brown 2017; Cantarero‐Villanueva 2016; Courneya 2016; Lee 2017; McDermott 2017), the Bruce Protocol Treadmill Test (Bourke 2011; Lewis 2016), the Modified Balke Treadmill Test (Courneya 2003), V02 peak test (Van Vulpen 2016); Cycle Ergometer Peak Power Output Test (Van Vulpen 2016), Rockport Walk Test (Nuri 2016), treadwalk test (Pinto 2013), and the Steep Ramp Test (Waart 2017). Nuri 2016 and Pinto 2013 estimated V02 peak from submaximal fitness tests; we used this data in our meta‐analysis. No included studies reported change from baseline data at more than 12 months follow‐up.
At immediate‐term follow‐up, we observed an improvement in aerobic fitness in the physical activity group compared with the usual care group (SMD 0.82, 95% CI 0.34 to 1.29; 7 studies, 295 participants; I2 = 68%; low‐quality evidence; Analysis 3.1). This effect was also observed at short‐term follow‐up (SMD 0.56, 95% CI 0.29 to 0.82; 7 studies, 248 participants; I2 = 1%; moderate‐quality evidence; Analysis 3.3), but not at medium‐term follow‐up (SMD 0.44, 95% CI ‐0.04 to 0.92; 4 studies, 272 participants; I2 = 57%; very low‐quality evidence; Analysis 3.5).
Change in aerobic fitness from baseline showed an improvement compared with usual care at immediate‐term follow‐up (SMD 0.89, 95% CI 0.43 to 1.36; 3 studies, 81 participants; I2 = 0%; low‐quality evidence; Analysis 3.2) and short‐term follow‐up (SMD 0.62, 95% CI 0.05 to 1.19; 2 studies, 51 participants; I2 = 0%; moderate‐quality evidence; Analysis 3.4). We were unable to include data from the study by Brown 2017 in the meta‐analysis as the authors reported mean change between groups over time. We emailed to request postintervention and change from baseline data, but no reply was received.
We conducted a sensitivity analysis excluding studies with an additional component (dietary advice), studies that did not conduct an ITT analysis and those at high risk of bias; results suggest improvements in aerobic fitness in the physical activity group compared with usual care (SMD 0.38, CI 0.06 to 0.70; 4 studies, 207 participants; I2= 15%; Table 1) at immediate‐ term follow‐up and short‐term follow‐up (SMD 0.45, CI 0.15 to 0.75; 5 studies, 187 participants; I2= 0%; Table 1). We also conducted a sensitivity analysis at medium‐term follow‐up excluding the study in which the physical activity intervention was ongoing. Results of this suggest no improvement in aerobic fitness in the physical activity group compared with usual care (SMD 0.44, CI 0.41 to 1.29; 3 studies, 86 participants).
Upper body strength
A total of five studies reported upper body strength, including arm strength, assessed by hand grip dynamometry (Lee 2017; Lewis 2016; Waart 2017), the 30 second arm‐curl test (Lewis 2016; Courneya 2016), isometric abdominal strength and isometric back strength assessed by the trunk‐curl test and back dynamometry, respectively (Cantarero‐Villanueva 2016), and the push‐up test to assess upper body strength and endurance (Lee 2017). A meta‐analysis was conducted for hand grip strength only, due to the large variability in measurement of upper body strength. Data were insufficient for analysis of more than six months to 12 months follow‐up and for change scores at all four time points.
We observed no difference in hand grip strength, assessed using hand grip dynamometry in the physical activity group compared with usual care at immediate‐term follow‐up (MD 1.92, 95% CI ‐1.17 to 5.00; 2 studies, 147 participants; I2 = 0%, low‐quality evidence; Analysis 4.1) or at short‐term follow‐up (MD 0.94, 95% CI ‐5.98 to 7.87; 2 studies, 39 participants; I2 = 0%; very low‐quality evidence; Analysis 4.2). In the study by Lee 2017, hand grip strength was reported separately for the left and right hand. We contacted Lee 2017 to request combined data, but we received no response. We inputted the data from the right hand only.
Lee 2017 reported a significantly greater improvement in the push‐up test (P < 0.001) in the intervention group (n = 62) compared with control (n = 61) at 12 weeks. The physical activity group (n = 21) experienced a greater increase in isometric abdominal strength compared with that of the usual care group (n = 19) at eight weeks (P = 0.001) but not at six months follow‐up in the study by Cantarero‐Villanueva 2016. No significant difference was found between groups for isometric back strength. There were no significant group differences in change in arm‐curl repetitions in the study by Lewis 2016 (n = 12 for the physical activity group and usual care group at 12 weeks; n = 9 and n = 11 at 6 months, respectively) or in the study by Courneya 2016 (P = 0.18) (n = 99 and n = 98 for the physical activity and usual care groups, respectively) at one year.
Flexibility
Three studies measured flexibility, assessed by the Sit and Reach Test (Courneya 2003; Courneya 2016), and the Modified Sit and Reach Test (Cantarero‐Villanueva 2016). Cantarero‐Villanueva 2016 reported flexibility of the right and left side separately. We emailed the author, but were unable to obtain combined sit and reach scores. We used data for the right side in our analysis. Mean scores were multiplied by minus one to ensure all scales had the same direction of effect as discussed in the Cochrane Handbook for Systematic Reviews of Interventions section, 9.2.3.2 (Deeks 2017). We observed no difference in flexibility in the physical activity group compared to usual care at short‐term follow‐up (SMD 0.02, 95% CI ‐0.36 to 0.39; 2 studies, 119 participants; I2 = 0%; low‐quality evidence; Analysis 5.1). There were insufficient data for analysis of flexibility at immediate‐ and medium‐term follow‐up and for change scores at all four time points.
Courneya 2016 reported that the structured exercise group (n = 99) improved relative to the health education material group (n = 98) for the Sit and Reach Test (mean between group difference of +2.1 cm; 95% CI ‐0.6 to +4.7; P = 0.08) at one year.
Lower limb strength
Two studies measured lower body strength, using dynamometry to assess maximum voluntary torque production of the knee extensors (Bourke 2011; Waart 2017). Meta‐analysis was precluded as studies measured lower limb strength at different time points. In the study by Bourke 2011, no difference in maximum voluntary torque was observed between the intervention (n = 9) and control group (n = 9) over the 12‐week intervention period (P = 0.127). Conversely, Waart 2017 observed improvements in lower body strength in the usual care control group (n = 6), the onco‐move group (low‐intensity home‐based exercise; n = 6) and a reduction in strength in the on‐track intervention group (moderate‐ to high‐intensity supervised exercise; n = 7) at the end of chemotherapy (calculated as 4.8 to 5 months follow‐ up). At six months following chemotherapy (calculated as 10 to 11 months follow‐up), the on‐track group (n = 6) recovered to above baseline values (no P values were reported).
Cancer‐related fatigue
A total of 10 studies reported cancer‐related fatigue, assessed by the Functional Assessment of Chronic Illness Therapy ‐ Fatigue (FACIT‐F) (Cramer 2016; Kim 2018; McDermott 2017), the Functional Assessment of Cancer Therapy ‐ Fatigue (FACT‐F) (Bourke 2011; Courneya 2003; Lewis 2016; Pinto 2013), the Multidimensional Fatigue Inventory (MFI) (Van Vulpen 2016; Waart 2017), the Multidimensional Fatigue Symptom Inventory ‐ Short Form (MFSI‐SF) (McDermott 2017), and the Fatigue Symptom Inventory (FSI) (Brown 2017). Van Vulpen 2016 and Waart 2017 reported results separately for subscales of the MFI. We extracted the MFI general subscale as a comparable measure for inclusion in the meta‐analysis. We chose the general subscale score as it includes general statements about fatigue and decreased functioning, designed to encompass both physical and psychological aspects of fatigue (Lin 2009). We multiplied mean scores from the MFI general subscale by minus one to ensure all scales had the same direction of effect, as discussed in the Cochrane Handbook for Systematic Reviews of Interventions section, 9.2.3.2 (Deeks 2017).
Low‐quality evidence suggests an improvement in cancer‐related fatigue in the physical activity group compared with the usual care group at immediate‐term follow‐up (MD 2.16, 95% CI 0.18 to 4.15; 6 studies, 230 participants; I2= 18%; Analysis 6.1), short‐term follow‐up (SMD 0.34, 95% CI 0.08 to 0.60; 7 studies, 277 participants; I2 = 9%; moderate‐quality evidence; Analysis 6.3), but not medium‐term follow‐up (SMD 0.25, 95% CI ‐0.16 to 0.67; 3 studies, 91 participants; I2 = 0%; low‐quality evidence; Analysis 6.4). Change from baseline to 12 weeks suggests no improvement in fatigue in the physical activity group compared with the usual care group (MD 0.41, 95% CI ‐1.33 to 2.14; 3 studies, 113 participants; I2 = 70%; low‐quality evidence; Analysis 6.2). There were insufficient data to conduct change from baseline analysis at all other time points. Brown 2017 reported improvements in change from baseline scores in the high dose intervention group compared with control at short‐term follow‐up (P trend = 0.045).
We conducted a sensitivity analysis with studies at low risk of bias and those that performed an ITT analysis at immediate‐ and short‐term follow‐up. There was no difference in fatigue between groups at immediate‐ or short‐term follow‐up (Table 1). Results from the fixed‐effect sensitivity analysis were consistent with those from the random‐effects model.
Anthropometric measurements
A total of 10 studies reported anthropometric measurements assessed by: weight (Bourke 2011; Cantarero‐Villanueva 2016; Courneya 2016; Lee 2017; Lewis 2016; McDermott 2017; Nuri 2016; Van Vulpen 2016), BMI (Bourke 2011; Brown 2017; Cantarero‐Villanueva 2016; Courneya 2003; Lee 2017; Lewis 2016; McDermott 2017; Nuri 2016), waist measurement (Brown 2017; Cantarero‐Villanueva 2016; Lee 2017; McDermott 2017), waist‐to‐hip ratio (Bourke 2011; Brown 2017; Lewis 2016; McDermott 2017), body fat percentage (Cantarero‐Villanueva 2016; Courneya 2003; Lee 2017; Lewis 2016; Nuri 2016), and hip circumference (Brown 2017; Courneya 2016; McDermott 2017). Brown 2017 also assessed visceral fat, subcutaneous fat and fat mass.
We found no evidence of effect of physical activity compared with usual care on weight (Analysis 7.1: moderate‐quality evidence and Analysis 8.3: low‐quality evidence), BMI (Analysis 11.1: low‐quality evidence and Analysis 11.3: low‐quality evidence), waist measurement (Analysis 8.1: very low‐quality evidence and Analysis 8.2: low‐quality evidence), waist‐to ‐hip ratio (Analysis 9.1: very low‐quality evidence and Analysis 9.3: low‐quality evidence), or body fat percentage (Analysis 10.1: low‐quality evidence and Analysis 10.3: very low‐quality evidence) at immediate‐ or short‐term follow‐up time points.
Change from baseline to 12 weeks follow‐up, but not change from baseline from 12 weeks to six months follow‐up suggests a small reduction in weight (MD ‐1.71, 95% CI ‐2.90 to ‐0.51; 3 studies, 82 participants; I2 = 1%; low‐quality evidence; Analysis 7.2; Analysis 7.3) and body fat percentage (MD ‐1.57, 95% CI ‐3.11 to ‐0.04; 2 studies, 60 participants; I2 = 0%; low‐quality evidence; Analysis 10.2 and Analysis 10.4), but not BMI at either time point (Analysis 11.2; Analysis 11.4; low‐ and very low‐quality evidence, respectively, or change from baseline up to 12 weeks for waist‐to‐hip‐ratio (Analysis 9.2; very low‐quality evidence). Low‐quality evidence suggests a small reduction in waist measurement in change from baseline from 12 weeks to six months follow‐up (MD ‐2.79, 95% CI ‐5.21 to ‐0.36; 2 studies, 70 participants; I2 = 38%; Analysis 8.3). When we conducted a sensitivity analysis using a random‐effects model for waist measurement, the difference between the physical activity and usual care group was no longer significant (MD ‐4.36 cm, ‐11.42 to 2.71; 2 studies, 70 participants; I2= 38%).
There was insufficient postintervention and change from baseline data at six months to 12 months to report on any anthropometric measure. Insufficient data precluded meta‐analysis of change from baseline data to 12 weeks follow‐up for waist measurement and more than 12 weeks to six months follow‐up for waist‐to‐hip‐ratio. We did not pool studies that measured hip circumference from 12 weeks to six months follow‐up (Brown 2017; McDermott 2017), as Brown 2017 reported change from baseline scores whilst McDermott 2017 reported postintervention scores.
Courneya 2016 reported no statistically significant difference between the structured exercise programme (SEP) group and the health education materials (HEM) group for weight (SEP = 115, HEM = 112) (P = 0.38), hip circumference (SEP = 99, HEM = 99) (P = 0.90) or waist circumference (SEP = 99, HEM = 99) (P = 0.31) at one year. Between‐group changes in body weight were not significantly different between intervention (n = 16) and usual care (n =14) for both men and women in the study by Van Vulpen 2016 at 18 and 36 weeks. Changes in waist‐to‐hip ratio were not statistically significance in the study by Lewis 2016 (P = 0.43) or Brown 2017 (P = 0.054). Cantarero‐Villanueva 2016 reported a significant change from baseline scores to eight weeks and six months between the physical activity group (n = 21) and usual care group (n = 19) for waist circumference. Changes in hip circumference up to six months were not significant in the study by Brown 2017 (P = 0.518), whilst McDermott 2017 reported an increase in hip circumference in the control group at week 12 (n = 9) and a decrease at week 24 (n = 9), but not to pre‐intervention levels, whilst the intervention group decreased at week 12 (n = 11) and increased at week 24 (n = 10) but not to pre‐intervention levels (P = 0.012).
We conducted a sensitivity analysis for weight, waist‐to‐hip‐ratio, BMI and body fat percentage at immediate‐term follow‐up, we excluded the study by Bourke 2011 because the intervention contained a dietary advice component and the study by Nuri 2016 as an ITT analysis was not conducted. Results of the sensitivity analysis suggests no difference in effect of physical activity compared with usual care on weight, BMI, waist measurement, waist‐to‐hip ratio or body fat percentage. The choice of model (fixed or random) did not effect the estimates of effect at this time point. A sensitivity analysis for change from baseline to 12 weeks follow‐up suggested no reduction in weight or BMI when the study by Bourke 2011 was removed. A sensitivity analysis with studies at low risk of bias was conducted for weight and BMI for change scores from 12 weeks to six months, results suggest no difference in effect of physical activity compared with usual care on weight or BMI at this time point.
Health‐related quality of life (HRQoL)
A total of 10 studies reported HRQoL; seven of these studies assessed HRQoL using the Functional Assessment of Cancer Therapy‐Colorectal (FACT‐C; Bourke 2011; Brown 2017; Courneya 2003; Cramer 2016; Kim 2018; McDermott 2017; Pinto 2013), the Functional Assessment of Cancer Therapy‐General (FACT‐G; Lewis 2016), the European Organisation for Research and Treatment of Cancer Quality of life Questionnaire‐Core 30 (EORTC QLQ‐C30; Van Vulpen 2016; Waart 2017), the EuroQol‐Visual Analogue Scales (EQ‐VAS; Lewis 2016), and the Satisfaction With Life Scale (SWLS; Courneya 2003).
Moderate‐quality evidence suggests a small positive effect for the physical activity group compared with the usual care group on HRQoL at immediate‐term follow‐up (SMD 0.36, 95% CI 0.10 to 0.62; 6 studies, 230 participants; I2 = 0%; Analysis 12.1) and short‐term follow‐up (SMD 0.45, 95% CI 0.03 to 0.88; 7 studies, 278 participants; I2 = 61%; Analysis 12.3) but no evidence of difference between groups at medium‐term follow‐up (SMD 0.05, 95% CI ‐0.37 to 0.47; 3 studies, 89 participants; I2 = 0%; low‐quality evidence; Analysis 12.5) or change from baseline up to 12 weeks follow‐up (SMD ‐0.10, 95% CI ‐0.47 to 0.28; 3 studies, 113 participants; I2 = 16%; low‐quality evidence; Analysis 12.2). We observed a small positive effect on change from baseline scores at more than 12 weeks to six months follow‐up (SMD 0.70, 95% CI 0.14 to 1.26; 2 studies, 58 participants; I2 = 0%; low‐quality evidence; Analysis 12.4). Data were insufficient to conduct a meta‐analysis of change from baseline scores at more than six months to 12 months follow‐up.
A sensitivity analysis suggests that physical activity interventions may have a positive effect on HRQoL compared to usual care at immediate‐term follow‐up, (excluding the study by Bourke 2011), and including studies at low risk of bias and those that conducted an ITT analysis (SMD 0.37, 95% CI 0.07 to 0.68; 4 studies, 169 participants; I2 = 0%; Table 1).
Levels of physical activity
A total of 12 studies reported on levels of physical activity, assessed objectively and subjectively. Objective physical activity was assessed by accelerometry (Brown 2017; Hubbard 2016; Lewis 2016; McDermott 2017; Van Blarigan 2019). We pooled total minutes of moderate to vigorous physical activity for the meta‐analysis. Subjective physical activity was assessed by the Godin Leisure Time Exercise Questionnaire (Bourke 2011; Courneya 2003; Kim 2018; Lee 2017), the International Physical Activity Questionnaire (IPAQ; Lewis 2016), the Seven‐Day Physical Activity Recall (7‐day PAR) questionnaire (Courneya 2003; Pinto 2013), the Tartu Physical Activity Questionnaire (TPAQ; Courneya 2016), the Community Healthy Activities Model Programme for Seniors (CHAMPS; Courneya 2016), and the Physical Activity Scale for the Elderly (PASE; Waart 2017).
We found no evidence of difference in levels of physical activity in the physical activity group compared to usual care at immediate‐term follow‐up (MD ‐8.34, 95% CI ‐21.05 to 4.37; 4 studies, 94 participants; I2 = 43%; very low‐quality evidence; Analysis 13.1) or short‐term follow‐up (MD 13.50, 95% CI ‐56.73 to 83.74; 2 studies, 36 participants; I2 = 13%; Analysis 13.4) when measured objectively using accelerometry. We found no evidence of difference in change from baseline scores up to 12 weeks between groups (SMD ‐0.13, 95% CI ‐0.77 to 0.52; 2 studies, 37 participants; I2 = 0%; very low‐quality evidence; Analysis 13.2).
Subjective measures of levels of physical activity suggest an increase in levels of physical activity in the physical activity group compared to the usual care group at immediate‐term follow‐up (SMD 0.70, 95% CI 0.38 to 1.03; 4 studies, 156 participants; I2 = 0%; very low‐quality evidence; Analysis 13.3), short‐term follow‐up (SMD 0.39, 95% CI ‐0.05 to 0.82; 4 studies, 176 participants; I2 = 38%; low‐quality evidence; Analysis 13.5) and medium‐term follow‐up (SMD 0.35, 95% CI 0.11 to 0.59; 3 studies, 274 participants; I2 = 21%; very low‐quality evidence; Analysis 13.6).
We were unable to extract data from studies by Brown 2017 and Lee 2017 because Brown 2017 reported mean change in moderate to vigorous physical activity and Lee 2017 reported moderate and vigorous physical activity separately. We emailed authors but received no response. In the study by Brown 2017, both the low dose (n = 14) and high dose (n = 12) exercise group increased their moderate to vigorous physical activity compared to the usual care group (n = 13) (P trend = < 0.001) over six months. Lee 2017 reported an increase in moderate to vigorous physical activity in the exercise group (n = 62) with no change in the usual care group at three months (group x time interaction, P < 0.01). Data were insufficient to conduct a meta‐analysis for change from baseline from more than 12 weeks to six months follow‐up and more than six months to 12 months follow‐up for objective measures and subjective measures at all time points.
We performed a sensitivity analysis, excluding studies with dietary advice co‐interventions up to 12 weeks follow‐up after baseline. Objectively measured physical activity suggests no difference in levels of physical activity in the physical activity group compared to usual care in immediate‐term follow‐up (MD ‐2.84, 95% CI ‐12.40 to 6.73; 3 studies, 80 participants; I2 = 0%; Table 1). Sensitivity analysis of subjective measures of physical activity suggest an increase in levels of physical activity in the physical activity group compared to the usual care group in immediate‐term follow‐up (SMD 0.68, 95% CI 0.33 to 1.02; 3 studies, 138 participants; I2 = 0%; Table 1). At short‐term follow‐up, we included studies at low risk of bias and studies that conducted an ITT analysis; this sensitivity analysis suggests no difference between levels of physical activity in the physical activity group compared to the usual care group.
Subgroup analysis
Subgroup analyses were precluded as we were unable to extract sufficient compatible data to undertake the predefined subgroup analysis
Discussion
Summary of main results
Summary of main results
This review included 16 randomised controlled trials (RCTs) with a total of 992 participants diagnosed with non‐advanced colorectal cancer. We are uncertain whether physical activity interventions improve physical function. We found no evidence of effect of physical activity interventions on disease‐related mental health, anthropometric measures or levels of physical activity (measured objectively). Interestingy, when levels of physical activity were measured subjectively they were higher in the physical activity group compared to the usual care group at immediate‐, short‐ and medium‐term follow‐up. We found no evidence of serious adverse events in the intervention or usual care groups in the eight studies that reported on adverse events. There was inconsistency in the measurement and reporting of adverse events between studies. Where reported, adverse events were generally minor. We found evidence of positive effects of physical activity interventions on physical fitness (aerobic fitness, but not other components of fitness), cancer‐related fatigue and health‐related quality of life (HRQoL) at immediate‐ and short‐term follow‐up but not medium‐term follow‐up. Only three studies reported medium‐term follow‐up for cancer‐related fatigue and HRQoL. No studies assessed outcomes at long‐term follow‐up. One study that assessed fatigue and levels of physical activity at four years is currently awaiting classification (Van Vulpen 2016).
No studies reported on overall survival or recurrence‐free survival. Our findings should be interpreted with caution as we rated the quality of evidence between very low to moderate overall. Further, higher quality studies are required to assess the effectiveness and safety of physical activity interventions for non‐advanced colorectal cancer patients.
Overall completeness and applicability of evidence
This systematic review includes studies from nine different countries. All studies except for one were undertaken in higher‐income countries (Nuri 2016). This evidence may therefore limit the applicability to low‐middle income countries. All studies in this version of the review except for one, exclusively included individuals with non‐advanced colorectal cancer; Courneya 2003 included four stage IV participants. The majority of studies included only participants that had finished active treatment, this was apparent across all outcomes, especially at immediate‐term follow‐up, limiting applicability to those undergoing active treatment. Ethnicity was only reported in three studies; these studies included mostly white people. Level of education was reported in nine studies and there was some variation in the percentage of participants educated to degree level. The physical activity interventions varied somewhat in their frequency, intensity, time, mode and duration, and because of this we are unable to ascertain optimal mode, frequency, intensity or duration of physical activity interventions for effects on the primary and secondary outcomes assessed in this review, limiting recommendations for clinical practice. We were unable to undertake subgroup analysis according to physical activity intervention or participant characteristics, cancer characteristics, treatment stages and length of time since diagnosis due to lack of comparable data, which limits applicability of findings. The objective measure of physical fitness that was used most commonly amongst included studies was the 6‐minute walk test. This is not the gold standard of measuring physical fitness and has been shown to underestimate VO2 peak in cancer survivors (Schumacher 2017). This may also limit the applicability of the evidence in this population. The highest incidence of colorectal cancer occurs in people aged between 65 and 74, the mean age of participants included in this review ranged between 51 and 69, this may therefore limit the applicability of this evidence to the population with the highest incidence of colorectal cancer.
For primary outcomes, such as physical function and disease‐related mental health, the greatest number of studies included in a meta‐analysis at any given time point was four. These findings should be interpreted with caution due to the limited number of studies and small sample sizes, especially at longer periods of follow‐up, where fewer studies were included. In addition, some of the outcome data included were not normally distributed which may have affected study results. Therefore, findings should be interpreted with caution. We were unable to undertake a meta‐analysis of adverse events due to inconsistencies in measurement and reporting of this data. Less than half of the included studies reported on adverse events and four of these reported that no adverse events occurred. In future, RCTs should systematically record and report adverse events and define whether these events are 'related' or 'unrelated' to the intervention. We could not provide an analysis of effects of physical activity interventions on important outcomes, such as overall survival or recurrence‐free survival because no included studies reported these outcomes. No studies assessed the long‐term effects (more than 12 months follow‐up) of physical activity interventions on any of the primary and secondary outcomes. One study is currently awaiting classification, which reports on four‐year effects of exercise on fatigue and levels of physical activity (Van Vulpen 2016). In addition, across all outcomes, few studies conducted follow‐up measurements at medium‐term follow‐up; these findings should be interpreted with caution
Quality of the evidence
The quality of the findings are reported in summary of findings Table 1 for the main comparison. We assessed the quality of evidence for each outcome using GRADE methodology (GRADE Working Group 2004). For the immediate‐term follow‐up, GRADE revealed low‐quality evidence for physical function, adverse events, physical fitness (aerobic fitness) and cancer‐related fatigue. Moderate‐quality evidence was revealed for HRQoL (immediate‐term follow‐up) and disease‐related mental health (anxiety and depression) at short‐term follow‐up. The quality of evidence for change from baseline scores and all other time points, ranged between very low to moderate quality. We downgraded the quality of evidence, namely in light of imprecision and indirectness due to the small number of studies/participants included and considering the majority of studies were conducted in participants who had finished active treatment. In addition, three studies included a health education component which may also introduce indirectness, although only one of these studies reported outcomes at immediate‐term follow‐up. Due to the nature of the intervention it was not possible to blind personnel or participants to the intervention, putting all studies at high risk of performance bias, which may be accentuated where participants completed subjective assessments. We found discrepancy in direction of effect of interventions when levels of physical activity were measured subjectively compared to objectively. It has been recently documented that people with cancer self‐report their physical activity levels to be nearly four‐fold higher when compared to objective physical activity monitoring data (Vassbakk‐Brovold 2016). Most of the uncertainty in judging study bias came from a lack of clarity around allocation concealment and blinding of outcome assessors. The 'Risk of bias' table and graph are available in Figure 2 and Figure 3. In addition, some of the data included in the meta‐analysis for measures of mental health, cancer‐related fatigue, aerobic fitness and levels of physical activity were not normally distributed, which may have affected the quality of the evidence.
Potential biases in the review process
This systematic review included comprehensive search strategies of eight electronic databases, and four clinical trial registries. We (MMG and MAT) independently screened reference lists of all included studies and any relevant systematic reviews identified. We handsearched conference and meeting abstracts of relevant organisations and undertook data extraction and 'Risk of bias judgement' independently. We sent emails to seven authors of studies registered in clinical trial registries with recent completion dates and received correspondence from six of these authors. We emailed four authors of unpublished studies, two of these studies are included in this review (Lewis 2016; McDermott 2017). We sent a further 13 emails to study authors that included colorectal cancer participants alongside other cancer cohorts, to request colorectal cancer participant data separately and emailed authors to request missing data,some did not reply and exclusion of these results may also be a source of bias. We were unable to explore publication bias because there were less than 10 studies included in each comparison. The mean age of participants ranged between 51 and 69, which is markedly lower than the incidence of colorectal cancer. The review therefore does not include a cohort of the relevant population, which is a source of selection bias. Furthermore, the mean age range of participants may be indicative of participation bias, where healthier participants are more likely to agree to participate in the physical activity intervention research.
A limitation of this review is that three included studies involved additional components, e.g. healthy eating seminars and a dietary information pack (Bourke 2011), health education materials for the usual care group (Courneya 2016), and weekly education sessions on a range of topics (Hubbard 2016), which could have potential synergistic effects of exercise or physical activity on some outcomes. We investigated their effect on outcomes through exclusion in sensitivity analysis when possible.
Agreements and disagreements with other studies or reviews
To our knowledge only two other systematic reviews have investigated the effectiveness of physical activity interventions in colorectal cancer patients. One if these reviews published in 2014 included 5 RCTs and 238 patients assessing the effect of exercise on HRQoL, fatigue, physical fitness, survival and/or tumour‐associated biomarkers in colorectal cancer patients (Cramer 2014b). The meta‐analysis for physical fitness, fatigue and HRQoL included three studies all of which are included in our review (Bourke 2011; Courneya 2003; Pinto 2013). At short‐term follow‐up (corresponding to our immediate‐term follow‐up) authors observed improvements in physical fitness after aerobic exercise compared with controls which is in agreement with our review. Also in agreement with our review, no studies reported on survival. Conversely, authors found no evidence for short‐term effects on quality of life or cancer‐related fatigue, whereas we found evidence of favouring effects of physical activity interventions on these outcomes at the same time point. We also observed this effect at short‐term follow‐up. Cramer 2014b was unable assess outcomes at this time point due to insufficient data. In addition, authors were unable to report on safety data, whereas we reported adverse events narratively due to the extent of variation in measurement and reporting of this outcome. The differences between these reviews are due to the time of the literature search, indeed all of our other included studies were conducted after 2014. The second review conducted by van Rooijen 2018 highlighted the lack of evidence available on exercise training in those explicitly undergoing active treatment, which was also evident in this review. Six of the seven studies included in the van Rooijen 2018 review included mixed cancer populations, we therefore did not compare results.
PRISMA flow diagram.
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.
Forest plot of comparison: 1 Physical activity versus usual care for physical function (30‐sec chair sit‐to‐stand test), outcome: 1.2 Objective measures more than 12 weeks to six months follow‐up (short term).
Forest plot of comparison: 1 Physical activity versus usual care for physical function, outcome: 1.3 Subjective measures more than 12 weeks to six months follow‐up (short term).
Forest plot of comparison: 2 Physical activity versus usual care for disease‐related mental health, outcome: depression 2.1 More than 12 weeks to six months follow‐up (short term).
Forest plot of comparison: 2 Physical activity versus usual care for disease‐related mental health (Hospital Anxiety and Depression Scale (HADS)), outcome: depression 2.2 More than six months to 12 months follow‐up (medium term).
Forest plot of comparison: 2 Physical activity versus usual care for disease‐related mental health, outcome: 2.3 Anxiety more than 12 weeks to six months follow‐up.
Forest plot of comparison: 2 Physical activity versus usual care for disease‐related mental health (Hospital Anxiety and Depression Scale (HADS)), outcome: 2.4 Anxiety more than six months to 12 months follow‐up (medium term).
Comparison 1: Physical activity versus usual care for physical function, Outcome 1: Objective measures more than 12 weeks to 6 months follow‐up (30‐Second Chair Stand Test)
Comparison 1: Physical activity versus usual care for physical function, Outcome 2: Subjective measures more than 12 weeks to 6 months follow‐up
Comparison 2: Physical activity versus usual care for disease‐related mental health, Outcome 1: Depression: more than 12 weeks to 6 months follow‐up
Comparison 2: Physical activity versus usual care for disease‐related mental health, Outcome 2: Depression: more than 6 months to 12 months follow‐up (HADS)
Comparison 2: Physical activity versus usual care for disease‐related mental health, Outcome 3: Anxiety: more than 12 weeks to 6 months follow‐up
Comparison 2: Physical activity versus usual care for disease‐related mental health, Outcome 4: Anxiety: more than 6 months to 12 months follow‐up (HADS)
Comparison 3: Physical activity versus usual care for physical fitness (aerobic fitness), Outcome 1: Up to 12 weeks follow‐up
Comparison 3: Physical activity versus usual care for physical fitness (aerobic fitness), Outcome 2: Change from baseline up to 12 weeks follow‐up
Comparison 3: Physical activity versus usual care for physical fitness (aerobic fitness), Outcome 3: More than 12 weeks to 6 months follow‐up
Comparison 3: Physical activity versus usual care for physical fitness (aerobic fitness), Outcome 4: Change from baseline more than 12 weeks to 6 months follow‐up
Comparison 3: Physical activity versus usual care for physical fitness (aerobic fitness), Outcome 5: More than 6 months to 12 months follow‐up
Comparison 4: Physical activity versus usual care for physical fitness (hand grip strength), Outcome 1: Up to 12 weeks follow‐up (hand dynamometry)
Comparison 4: Physical activity versus usual care for physical fitness (hand grip strength), Outcome 2: More than 12 weeks to 6 months follow‐up (hand dynamometry)
Comparison 5: Physical activity versus usual care for physical fitness (flexibility), Outcome 1: More than 12 weeks to 6 months follow‐up
Comparison 6: Physical activity versus usual care for cancer‐related fatigue, Outcome 1: Up to 12 weeks follow‐up (FACT‐F and FACIT‐F)
Comparison 6: Physical activity versus usual care for cancer‐related fatigue, Outcome 2: Change from baseline up to 12 weeks follow‐up (FACT‐F and FACIT‐F)
Comparison 6: Physical activity versus usual care for cancer‐related fatigue, Outcome 3: More than 12 weeks to 6 months follow‐up
Comparison 6: Physical activity versus usual care for cancer‐related fatigue, Outcome 4: More than 6 months to 12 months follow‐up
Comparison 7: Physical activity versus usual care for anthropometric measure of weight (kg), Outcome 1: Up to 12 weeks follow‐up
Comparison 7: Physical activity versus usual care for anthropometric measure of weight (kg), Outcome 2: Change from baseline up to 12 weeks follow‐up
Comparison 7: Physical activity versus usual care for anthropometric measure of weight (kg), Outcome 3: Change from baseline more than 12 weeks to 6 months follow‐up
Comparison 7: Physical activity versus usual care for anthropometric measure of weight (kg), Outcome 4: More than 12 weeks to 6 months follow‐up
Comparison 8: Physical activity versus usual care for anthropometric measure of waist circumference, Outcome 1: Up to 12 weeks follow‐up
Comparison 8: Physical activity versus usual care for anthropometric measure of waist circumference, Outcome 2: More than 12 weeks to 6 months follow‐up
Comparison 8: Physical activity versus usual care for anthropometric measure of waist circumference, Outcome 3: Change from baseline more than 12 weeks to 6 months follow‐up
Comparison 9: Physical activity versus usual care anthropometric measure of waist to hip ratio, Outcome 1: Up to 12 weeks follow‐up
Comparison 9: Physical activity versus usual care anthropometric measure of waist to hip ratio, Outcome 2: Change from baseline up to 12 weeks follow‐up
Comparison 9: Physical activity versus usual care anthropometric measure of waist to hip ratio, Outcome 3: More than 12 weeks to 6 months follow‐up
Comparison 10: Physical activity versus usual care for anthropometric measure of body fat (%), Outcome 1: Up to 12 weeks follow‐up
Comparison 10: Physical activity versus usual care for anthropometric measure of body fat (%), Outcome 2: Change from baseline up to 12 weeks follow‐up
Comparison 10: Physical activity versus usual care for anthropometric measure of body fat (%), Outcome 3: More than 12 weeks to 6 months follow‐up
Comparison 10: Physical activity versus usual care for anthropometric measure of body fat (%), Outcome 4: Change from baseline more than 12 weeks to 6 months follow‐up
Comparison 11: Physical activity versus usual care for anthropometric measure of BMI, Outcome 1: Up to 12 weeks follow‐up
Comparison 11: Physical activity versus usual care for anthropometric measure of BMI, Outcome 2: Change from baseline up to 12 weeks follow‐up
Comparison 11: Physical activity versus usual care for anthropometric measure of BMI, Outcome 3: More than 12 weeks to 6 months follow‐up
Comparison 11: Physical activity versus usual care for anthropometric measure of BMI, Outcome 4: Change from baseline more than 12 weeks to 6 months follow‐up
Comparison 12: Physical activity versus usual care for HRQoL, Outcome 1: Up to 12 weeks follow‐up
Comparison 12: Physical activity versus usual care for HRQoL, Outcome 2: Change from baseline up to 12 weeks follow‐up
Comparison 12: Physical activity versus usual care for HRQoL, Outcome 3: More than 12 weeks to 6 months follow‐up
Comparison 12: Physical activity versus usual care for HRQoL, Outcome 4: Change from baseline more than 12 weeks to 6 months follow‐up
Comparison 12: Physical activity versus usual care for HRQoL, Outcome 5: More than 6 months to 12 months follow‐up
Comparison 13: Physical activity versus usual care for levels of physical activity, Outcome 1: Objective measures up to 12 weeks follow‐up (accelerometry moderate to vigorous physical activity mins/per day)
Comparison 13: Physical activity versus usual care for levels of physical activity, Outcome 2: Change from baseline in objective measures up to 12 weeks follow‐up (accelerometry moderate to vigorous physical activity)
Comparison 13: Physical activity versus usual care for levels of physical activity, Outcome 3: Subjective measures up to 12 weeks follow‐up
Comparison 13: Physical activity versus usual care for levels of physical activity, Outcome 4: Objective measures more than 12 weeks to 6 months follow‐up (accelerometry moderate to vigorous physical activity mins/week)
Comparison 13: Physical activity versus usual care for levels of physical activity, Outcome 5: Subjective measures more than 12 weeks to 6 months follow‐up
Comparison 13: Physical activity versus usual care for levels of physical activity, Outcome 6: Subjective measures more than 6 months to 12 months follow‐up
| Physical activity compared with usual care in adults with non‐advanced colorectal cancer | ||||||
| Population: adults with non‐advanced colorectal cancer treated surgically or with neoadjuvant or adjuvant therapy, or both Settings: all but one study undertaken in high‐income countries. Included home‐based self‐directed and supervised physical activity programmes Intervention: aerobic or resistance training, flexibility or balance training or a combination of these, lasting at least 4 weeks Comparison: control intervention (usual care or no physical activity intervention) | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of Participants | Quality of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Assumed risk | Corresponding risk | |||||
| Usual care | Physical activity | |||||
| Physical function Assessed with: 30‐Second Chair Stand Test Follow‐up: up to 12 weeks (immediate‐term) | We did not pool results due to considerable variability and inconsistency in direction of effect. Two studies observed no difference between the physical activity and usual care group for physical function at immediate‐term follow‐up. Two other studies reported significant improvements in physical function in the physical activity group compared with usual care | 185 (4 RCTs) | ⊕⊕⊝⊝a,b | We are uncertain whether physical activity interventions improve physical function | ||
| Disease‐related mental health: depression Assessed with: HADS, CES‐D Follow‐up: more than 12 weeks to 6 months (short term) | The mean postintervention HADS for depression ranged across control groups from 2.14 to 4.72 | The mean postintervention depression in the intervention group was 0.84 (2 lower to 0.32 higher) points lower than control | 198 (4 RCTs) | ⊕⊕⊕⊝b | Scores estimated using SMD ‐0.21 (‐0.50 to 0.08)g No evidence of difference in depression in the physical activity group compared with usual care group | |
| Disease‐related mental health: anxiety Assessed with: HADS, State‐Trait Anxiety Inventory Follow‐up: more than 12 weeks to 6 months (short term) | The mean postintervention HADS for anxiety ranged across control groups from 2 to 3 | The mean postintervention anxiety in the intervention groups was 0.40 points (1.2 lower to 0.54 higher) lower than control | 198 (4 RCTs) | ⊕⊕⊕⊝b | Scores estimated using SMD ‐0.11 (‐0.40 to 0.18)g No evidence of difference in anxiety in the physical activity group compared with usual care group | |
| Overall survival (time interval between enrolment in the study and death of the person from any cause) Follow‐up: 12 months | See comment | See comment | Not estimable | The included studies did not report on overall survival | ||
| Recurrence‐free survival (time interval between date of enrolment in the study and the date when colorectal cancer recurs or another cancer occurs during the follow‐up) Follow‐up: 12 months | See comment | See comment | Not estimable | The included studies did not report on recurrence‐free survival | ||
| Adverse events Follow‐up: range 8 weeks to 11 months | 4 studies reported no adverse events, 3 other studies reported no serious adverse events with 7 participants experiencing minor adverse events in one study, 101 minor adverse events being reported in another study and 39 and 36 minor adverse events being reported in the intervention and control groups, respectively in another study. 1 study did not differentiate between serious and minor adverse events and reported 9 adverse events in the intervention group and one in the control | 305 (8 RCTs) | ⊕⊕⊝⊝c,d | |||
| Physical fitness: aerobic fitness Assessed with: 6‐minute walk test, Bruce Protocol Treadmill Test, estimated V02 peak Follow‐up: up to 12 weeks (immediate term) | The mean postintervention 6‐minute walk test score ranged across control groups from 293.7 to 588.9 | The mean postintervention physical fitness in the intervention group was 59 metres (24.5 to 93.1) higher than control | 295 (7 RCTs) | ⊕⊕⊝⊝a,e | Scores estimated using a SMD 0.82 (0.34 to 1.29)f Evidence suggests an improvement in aerobic fitness in the physical activity group compared with usual care group | |
| Cancer‐related fatigue Assessed with: FACIT‐F and FACT‐F (scale 0‐52: higher score indicates lower fatigue) Follow‐up: up to 12 weeks (immediate term) | The mean postintervention cancer‐related fatigue score ranged across control groups from 37.1 to 44 | The mean postintervention cancer‐related fatigue score in the intervention groups was | 230 (6 RCTs) | ⊕⊕⊝⊝a,b | Evidence suggests an improvement in cancer‐related fatigue in the physical activity group compared with the usual care group | |
| Health‐related quality of life (HRQoL) Assessed with: FACT‐C, FACT‐G (higher score indicates better quality of life) Follow‐up: up to 12 weeks (immediate term) | The mean postintervention FACT‐C scores ranged across control groups from 99.1 to 110.8 | The mean postintervention HRQoL in the intervention group was 6.64 (1.8 to 11.4) points higher than control | 230 (6 RCTs) | ⊕⊕⊕⊝b | Scores estimated using SMD 0.36 (0.10 to 0.62)h MID 5 to 8 points Evidence suggests an improvement in HRQoL in the physical activity group compared with the usual care group | |
| *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). CES‐D: Centre for Epidemiological Studies Depression Scale; CI: confidence interval; FACIT‐F: Functional Assessment of Chronic Illness Therapy‐Fatigue; FACT‐C: Functional Assessment of Cancer Therapy‐Colorectal; FACT‐F: Functional Assessment of Cancer Therapy‐Fatigue; FACT‐G: Functional Assessment of Cancer Therapy‐General; HADS: Hospital Anxiety and Depression Scale; HRQoL: health‐related quality of life; MID: Minmal important difference, MD: mean difference: RCT: randomised controlled trial; SD: standard deviation; SMD: standardised mean difference (used when studies assess the same outcome but measure it in a variety of ways). | ||||||
| GRADE Working Group grades of evidence | ||||||
| aDowngraded one level due to indirectness (applicability of results to those undergoing active treatment). | ||||||
| Outcome | Time point | No. of studies | No. of participants | Statistical method | Effect size |
|---|---|---|---|---|---|
| 1 Physical functiond | |||||
| 1.1 Subjective measure of physical function | Short‐term follow‐up | 2 | 114 | SMD (IV, random, 95% CI) | 0.08 (‐0.31 to 0.47)a |
| 2 Disease‐related mental health | |||||
| 2.1 Anxiety | Short‐term follow‐up | 3 | 177 | SMD (IV, random, 95% CI) | ‐0.29 (‐0.60 to 0.01)a,b |
| 2.2. Depression | Short‐term follow‐up | 3 | 177 | SMD (IV, random, 95% CI) | ‐0.18 (‐0.48 to 0.13)a,b |
| 3 Physical fitness | |||||
| 3.1 Aerobic fitness | Immediate‐term follow‐up Short‐term follow‐up | 4 5 | 207 187 | SMD (IV, random, 95% CI) | 0.38 (0.06 to 0.70)a,b 0.45 (0.15 to 0.75)a,b |
| 4 Cancer‐related fatigued | Immediate‐term follow‐up Short‐term follow‐up | 4 5 | 169 224 | MD (IV, random, 95% CI) SMD (IV, random, 95% CI) | 2.22 (‐0.34 to 4.79)a,b 0.32 (‐0.04 to 0.67)a,b |
| 5 Anthropometric measuresd | |||||
| 5.1 Weight | Immediate‐term follow‐up Change from baseline to 12 weeks follow‐up | 4 2 | 207 64 | MD (IV, random, 95% CI) | 0.27 (‐2.87 to 3.42)a,c ‐1.76 [‐4.06 to 0.54]c |
| 5.2 Waist to hip ratio | Immediate‐term follow‐up | 2 | 44 | MD (IV, random, 95% CI) | 0.04 [‐0.01 to 0.10]a,c |
| 5.3 BMI | Immediate‐term follow‐up Change from baseline to 12 weeks follow‐up | 4 2 | 207 64 | MD (IV, random, 95% CI) | 0.10 [‐0.87 to 1.06]a,c ‐0.42 [‐1.30 to 0.46] |
| 5.4 Body fat % | Immediate‐term follow‐up | 3 | 187 | MD (IV, random, 95% CI) | ‐2.13 [‐4.46 to 0.21]a,d |
| 6 HRQoL | Immediate‐term follow‐up | 4 | 169 | SMD (IV, random, 95% CI) | 0.37 [0.07 to 0.68]a,b,c |
| 7 Levels of physical activity | |||||
| 7.1 Objective measures | Immediate‐term follow‐up | 3 | 80 | MD (IV, random, 95% CI) | ‐2.84 [‐12.40 to 6.73]c |
| 7.2 Subjective measures | Immediate‐term follow‐up | 3 | 138 | SMD (IV, random, 95% CI) | 0.68[0.33 to 1.02]c |
| BMI: body mass index; CI: confidence interval; HRQoL: health‐related quality of life; MD: mean difference: SD: standard deviation; SMD: standardised mean difference (used when studies assess the same outcome but measure it in a variety of ways). a Removal of studies that did not conduct an ITT analysis b Exclusion of studies at high risk of bias c Exclusion of studies with an additional intervention component d Results from choice of model (fixed or random) were consistent | |||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Objective measures more than 12 weeks to 6 months follow‐up (30‐Second Chair Stand Test) Show forest plot | 2 | 39 | Mean Difference (IV, Fixed, 95% CI) | 0.76 [‐1.84, 3.36] |
| 1.2 Subjective measures more than 12 weeks to 6 months follow‐up Show forest plot | 3 | 156 | Std. Mean Difference (IV, Random, 95% CI) | 0.09 [‐0.24, 0.42] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 2.1 Depression: more than 12 weeks to 6 months follow‐up Show forest plot | 4 | 198 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.21 [‐0.50, 0.08] |
| 2.2 Depression: more than 6 months to 12 months follow‐up (HADS) Show forest plot | 2 | 48 | Mean Difference (IV, Fixed, 95% CI) | ‐1.20 [‐2.72, 0.31] |
| 2.3 Anxiety: more than 12 weeks to 6 months follow‐up Show forest plot | 4 | 198 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.11 [‐0.40, 0.18] |
| 2.4 Anxiety: more than 6 months to 12 months follow‐up (HADS) Show forest plot | 2 | 47 | Mean Difference (IV, Fixed, 95% CI) | 1.79 [‐0.37, 3.94] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 3.1 Up to 12 weeks follow‐up Show forest plot | 7 | 295 | Std. Mean Difference (IV, Random, 95% CI) | 0.82 [0.34, 1.29] |
| 3.2 Change from baseline up to 12 weeks follow‐up Show forest plot | 3 | 81 | Std. Mean Difference (IV, Random, 95% CI) | 0.89 [0.43, 1.36] |
| 3.3 More than 12 weeks to 6 months follow‐up Show forest plot | 7 | 248 | Std. Mean Difference (IV, Random, 95% CI) | 0.56 [0.29, 0.82] |
| 3.4 Change from baseline more than 12 weeks to 6 months follow‐up Show forest plot | 2 | 51 | Std. Mean Difference (IV, Fixed, 95% CI) | 0.62 [0.05, 1.19] |
| 3.5 More than 6 months to 12 months follow‐up Show forest plot | 4 | 272 | Std. Mean Difference (IV, Random, 95% CI) | 0.44 [‐0.04, 0.92] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 4.1 Up to 12 weeks follow‐up (hand dynamometry) Show forest plot | 2 | 147 | Mean Difference (IV, Fixed, 95% CI) | 1.92 [‐1.17, 5.00] |
| 4.2 More than 12 weeks to 6 months follow‐up (hand dynamometry) Show forest plot | 2 | 39 | Mean Difference (IV, Fixed, 95% CI) | 0.94 [‐5.98, 7.87] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 5.1 More than 12 weeks to 6 months follow‐up Show forest plot | 2 | 119 | Std. Mean Difference (IV, Fixed, 95% CI) | 0.02 [‐0.36, 0.39] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 6.1 Up to 12 weeks follow‐up (FACT‐F and FACIT‐F) Show forest plot | 6 | 230 | Mean Difference (IV, Random, 95% CI) | 2.16 [0.18, 4.15] |
| 6.2 Change from baseline up to 12 weeks follow‐up (FACT‐F and FACIT‐F) Show forest plot | 3 | 113 | Mean Difference (IV, Fixed, 95% CI) | 0.41 [‐1.33, 2.14] |
| 6.3 More than 12 weeks to 6 months follow‐up Show forest plot | 7 | 277 | Std. Mean Difference (IV, Random, 95% CI) | 0.34 [0.08, 0.60] |
| 6.4 More than 6 months to 12 months follow‐up Show forest plot | 3 | 91 | Std. Mean Difference (IV, Fixed, 95% CI) | 0.25 [‐0.16, 0.67] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 7.1 Up to 12 weeks follow‐up Show forest plot | 6 | 252 | Mean Difference (IV, Random, 95% CI) | 0.29 [‐2.55, 3.14] |
| 7.2 Change from baseline up to 12 weeks follow‐up Show forest plot | 3 | 82 | Mean Difference (IV, Random, 95% CI) | ‐1.71 [‐2.90, ‐0.51] |
| 7.3 Change from baseline more than 12 weeks to 6 months follow‐up Show forest plot | 3 | 89 | Mean Difference (IV, Random, 95% CI) | ‐0.73 [‐2.17, 0.72] |
| 7.4 More than 12 weeks to 6 months follow‐up Show forest plot | 3 | 74 | Mean Difference (IV, Random, 95% CI) | 0.59 [‐6.87, 8.04] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 8.1 Up to 12 weeks follow‐up Show forest plot | 3 | 183 | Mean Difference (IV, Random, 95% CI) | 0.02 [‐2.88, 2.93] |
| 8.2 More than 12 weeks to 6 months follow‐up Show forest plot | 2 | 50 | Mean Difference (IV, Fixed, 95% CI) | 1.58 [‐5.58, 8.74] |
| 8.3 Change from baseline more than 12 weeks to 6 months follow‐up Show forest plot | 2 | 70 | Mean Difference (IV, Fixed, 95% CI) | ‐2.79 [‐5.21, ‐0.36] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 9.1 Up to 12 weeks follow‐up Show forest plot | 3 | 62 | Mean Difference (IV, Random, 95% CI) | ‐0.01 [‐0.12, 0.10] |
| 9.2 Change from baseline up to 12 weeks follow‐up Show forest plot | 2 | 42 | Mean Difference (IV, Fixed, 95% CI) | ‐0.00 [‐0.02, 0.02] |
| 9.3 More than 12 weeks to 6 months follow‐up Show forest plot | 2 | 43 | Mean Difference (IV, Fixed, 95% CI) | 0.06 [‐0.03, 0.14] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 10.1 Up to 12 weeks follow‐up Show forest plot | 4 | 214 | Mean Difference (IV, Random, 95% CI) | ‐1.93 [‐4.04, 0.18] |
| 10.2 Change from baseline up to 12 weeks follow‐up Show forest plot | 2 | 60 | Mean Difference (IV, Fixed, 95% CI) | ‐1.57 [‐3.11, ‐0.04] |
| 10.3 More than 12 weeks to 6 months follow‐up Show forest plot | 3 | 139 | Std. Mean Difference (IV, Fixed, 95% CI) | ‐0.08 [‐0.42, 0.27] |
| 10.4 Change from baseline more than 12 weeks to 6 months follow‐up Show forest plot | 2 | 51 | Mean Difference (IV, Fixed, 95% CI) | ‐1.26 [‐3.11, 0.59] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 11.1 Up to 12 weeks follow‐up Show forest plot | 6 | 252 | Mean Difference (IV, Random, 95% CI) | 0.14 [‐0.73, 1.02] |
| 11.2 Change from baseline up to 12 weeks follow‐up Show forest plot | 3 | 82 | Mean Difference (IV, Random, 95% CI) | ‐0.32 [‐0.81, 0.17] |
| 11.3 More than 12 weeks to 6 months follow‐up Show forest plot | 4 | 158 | Std. Mean Difference (IV, Random, 95% CI) | 0.00 [‐0.32, 0.33] |
| 11.4 Change from baseline more than 12 weeks to 6 months follow‐up Show forest plot | 3 | 89 | Mean Difference (IV, Random, 95% CI) | ‐0.26 [‐1.17, 0.66] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 12.1 Up to 12 weeks follow‐up Show forest plot | 6 | 230 | Std. Mean Difference (IV, Random, 95% CI) | 0.36 [0.10, 0.62] |
| 12.2 Change from baseline up to 12 weeks follow‐up Show forest plot | 3 | 113 | Std. Mean Difference (IV, Fixed, 95% CI) | ‐0.10 [‐0.47, 0.28] |
| 12.3 More than 12 weeks to 6 months follow‐up Show forest plot | 7 | 278 | Std. Mean Difference (IV, Random, 95% CI) | 0.45 [0.03, 0.88] |
| 12.4 Change from baseline more than 12 weeks to 6 months follow‐up Show forest plot | 2 | 58 | Std. Mean Difference (IV, Fixed, 95% CI) | 0.70 [0.14, 1.26] |
| 12.5 More than 6 months to 12 months follow‐up Show forest plot | 3 | 89 | Std. Mean Difference (IV, Fixed, 95% CI) | 0.05 [‐0.37, 0.47] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 13.1 Objective measures up to 12 weeks follow‐up (accelerometry moderate to vigorous physical activity mins/per day) Show forest plot | 4 | 94 | Mean Difference (IV, Random, 95% CI) | ‐8.34 [‐21.05, 4.37] |
| 13.2 Change from baseline in objective measures up to 12 weeks follow‐up (accelerometry moderate to vigorous physical activity) Show forest plot | 2 | 37 | Std. Mean Difference (IV, Fixed, 95% CI) | ‐0.13 [‐0.77, 0.52] |
| 13.3 Subjective measures up to 12 weeks follow‐up Show forest plot | 4 | 156 | Std. Mean Difference (IV, Random, 95% CI) | 0.70 [0.38, 1.03] |
| 13.4 Objective measures more than 12 weeks to 6 months follow‐up (accelerometry moderate to vigorous physical activity mins/week) Show forest plot | 2 | 36 | Mean Difference (IV, Fixed, 95% CI) | 13.50 [‐56.73, 83.74] |
| 13.5 Subjective measures more than 12 weeks to 6 months follow‐up Show forest plot | 4 | 176 | Std. Mean Difference (IV, Random, 95% CI) | 0.39 [‐0.05, 0.82] |
| 13.6 Subjective measures more than 6 months to 12 months follow‐up Show forest plot | 3 | 274 | Std. Mean Difference (IV, Fixed, 95% CI) | 0.35 [0.11, 0.59] |
