Study selection
The search strategy identified 982 records; 970 remained after duplicates were removed. Following title and abstract screening, 247 full text articles were assessed for eligibility and 23 were included. Study selection and reasons for exclusion are presented in Figure 1.
Study characteristics
Study characteristics are presented in Table 1. Nineteen observational studies and 4 randomized controlled trials were included. A total of 5785 individuals were prescribed exercise across all studies (sample sizes ranged from 23-1218 participants) and average age ranged from 66-79 years. Indications for exercise included cardiac rehabilitation (n=6), pulmonary rehabilitation (n=7), and other (n=10; including surgical, medical and neurologic indications). Most (20/23 (87%)) exercise programs were supervised.
Table 1. Study Characteristics
Author
|
Year
|
Design
|
N
|
Average age
|
Medical indication
|
Exercise program
|
Adherence
|
Adherence definition
|
Ades et al.(33)
|
1992
|
OBS
|
226
|
70
|
MI or CABG
|
CRa
|
21%c
|
Entry into the CR
|
Aherne et al.(46)
|
2017
|
OBS
|
98
|
69
|
PVD
|
Othera
|
N/Ab
|
Number of sessions attended
|
Brown et al.(34)
|
2016
|
OBS
|
440
|
66
|
COPD
|
PRa
|
52%c
|
Number of sessions attended
|
Casey et al.(20)
|
2008
|
OBS
|
600
|
66
|
CVD
|
CRa
|
78%c
|
Staff judgement
|
Covey et al.(47)
|
2014
|
RCT
|
113
|
68
|
COPD
|
PRa
|
93%b
|
Percent of exercise completed
|
Cox et al.(35)
|
2013
|
OBS
|
85
|
68
|
Cognitive impairment
|
Other
|
78%b
|
Self-reported
|
Craike et al.(21)
|
2016
|
OBS
|
52
|
67
|
Prostate cancer
|
Othera
|
80%b
|
Number of sessions attended
|
Fan et al.(22)
|
2008
|
OBS
|
1218
|
67
|
COPD
|
PRa
|
79%b
|
Number of sessions attended
|
Gallagher et al.(23)
|
2003
|
OBS
|
196
|
67
|
CVD
|
CRa
|
32%c
|
Number of sessions attended
|
Hogg et al.(24)
|
2012
|
OBS
|
812
|
> 65
|
COPD
|
PRa
|
54%c
|
Number of sessions attended
|
Jensen et al.(25)
|
2016
|
OBS
|
50
|
69
|
Bladder cancer
|
Other
|
66%c
|
Self-reported
|
Mangione et al.(48)
|
2005
|
RCT
|
23
|
79
|
Hip fracture
|
Othera
|
98%b
|
Number of sessions attended
|
Messer et al.(36)
|
2007
|
OBS
|
164
|
66
|
Incontinence
|
Othera
|
70%c
|
Self-reported
|
Mudge et al.(26)
|
2013
|
OBS
|
140
|
> 65
|
CVD, pulmonary disease
|
Othera
|
42%b
|
Number of sessions attended
|
Pakzad et al.(27)
|
2013
|
OBS
|
30
|
66
|
CVD
|
CRa
|
N/Ab
|
Number of sessions attended
|
Pandey et al.(37)
|
2017
|
RCT
|
40
|
67
|
Diabetes
|
Othera
|
70%b
|
Self-reported
|
Pickering et al.(28)
|
2013
|
OBS
|
70
|
73
|
Parkinson's disease
|
Othera
|
79%b
|
Percent of exercise completed
|
Rizk et al.(38)
|
2015
|
RCT
|
35
|
67
|
COPD
|
PRa
|
75%b
|
Percent of exercise completed
|
Selzler et al.(29)
|
2016
|
OBS
|
64
|
69
|
COPD
|
PRa
|
81%b
|
Number of sessions attended
|
Selzler et al.(30)
|
2012
|
OBS
|
814
|
68
|
COPD
|
PRa
|
83%b
|
Number of sessions attended
|
Tiedemann et al.(39)
|
2012
|
OBS
|
76
|
67
|
Stroke
|
Othera
|
60%b
|
Number of sessions attended
|
Tooth et al.(31)
|
1993
|
OBS
|
30
|
66
|
MI
|
CR
|
93%bf, 87%bg
|
Percent of exercise completed
|
van Montfort et al.(32)
|
2016
|
OBS
|
409
|
66
|
PCI
|
CRa
|
N/Ab
|
Number of sessions attended
|
CABG = coronary artery bypass graft; COPD = chronic obstructive pulmonary disease; CR = cardiac rehabilitation; CVD = cardiovascular disease; FEV1 = forced expiratory volume in 1 second; HADS = Hospital Anxiety and Depression Scale; HDL = high density lipoprotein; IMD = Index of Multiple Deprivation; MI = myocardial infarction; MRC = Medical Research Council; OBS = observational; PCI = primary coronary intervention; PR = pulmonary rehabilitation; PVD = peripheral vascular disease; RCT = randomized controlled trial; a = supervised exercise program; b = adherence as a continuous measure, c = adherence as a categorical threshold; d = % participation; e = % completion; f = % duration; g = % frequency
Adherence to prescribed exercise rates
Exercise adherence was measured as a continuous variable in 16 studies and as a categorical outcome with a specified cut-off (demarking adherent vs not) in the remaining 7 studies. Overall adherence rate was reported in 20 studies and ranged from 21% to 93% (mean 68%, standard deviation (SD) 23%). Adherence was highest for pulmonary rehabilitation (71%, SD 15%), and other indications (74%, SD 13%); cardiac rehabilitation had lower rates (55%, SD 33%). However, lack of variance measures around adherence estimates limited our ability to perform formal comparative meta-analysis or meta-regression.
Predictors of exercise adherence
Predictors of exercise adherence were grouped into the following clusters: demographic, psychological, program-related, medical condition severity, comorbidities, and other (see Supplementary Tables S1, S2, S3; Additional Files 1, 2, 3). Demographic factors were evaluated by 13 studies(20,21,30–32,22–29) (Table 2), psychological factors by 14 studies(20,21,32–36,22–24,27–31) (Table 3), program-related factors by 2 studies,(37,38) medical condition severity by 11 studies,(21,22,39,23,24,28–32,34) comorbidities by 8 studies(20,25,27,29–31,34,35) and other predictors by 5 studies.(24,26,31,33,39)
Table 2. Demographic Predictors of Exercise Adherence
Study
|
Predictors
|
Direction
|
Theme
|
Casey et al. (2008)(20)
|
Age (years)
|
+
|
Age
|
|
Employed (vs not employed/retired)
|
0
|
Employment
|
|
Gender (male vs female)
|
0
|
Sex
|
Craike et al. (2016)(21)
|
Highest level of education (less than university degree vs university degree or higher)
|
0
|
Education
|
Fan et al. (2008)(49)
|
Age (per 1 year change)
|
0
|
Age
|
|
Female gender
|
0
|
Sex
|
|
Education reference: < high school
|
|
Education
|
|
High school
|
+
|
|
|
Some college
|
+
|
|
|
> College
|
+
|
|
Gallagher et al. (2003)(23)
|
Unemployed or retired (vs employed)
|
-
|
Employment
|
|
Age > 70 (vs 55–70)
|
-
|
Age
|
Hogg et al. (2012)(24)
|
Deprivation quintile (IMD score) reference: IMD 6.86-28.1
|
|
Social status
|
|
IMD 28.11–35.02
|
0
|
|
|
IMD 35.03–39.57
|
0
|
|
|
IMD 39.58–43.85
|
-
|
|
|
IMD 43.86–60.41
|
-
|
|
Jensen et al. (2016)(25)
|
Gender (women vs men)
|
0
|
Sex
|
|
Age (<70 vs ≥70)
|
0
|
Age
|
Mudge et al. (2013)(26)
|
Retired from workforce (vs "working" and "not working")
|
+
|
Employment
|
|
Age <65 vs 65+
|
0
|
Age
|
|
Sex (male vs female)
|
0
|
Sex
|
|
Living alone vs living with family/others
|
0
|
Living status
|
Pakzad et al. (2013)(27)
|
Identity
|
0
|
|
Pickering et al. (2013)(28)
|
Gender (male vs female)
|
0
|
Sex
|
|
Living status (alone vs partner vs family/friends vs other)
|
0
|
Living status
|
|
Age multiplicative decrease per 10 years
|
-
|
Age
|
Selzler et al. (2016)(29)
|
Age (years)
|
0
|
Age
|
Selzler et al. (2012)(30)
|
Age (years)
|
+
|
Age
|
Tooth et al. (1992)(31)
|
Scale of Status and Prestige (high score = lower social standing)
|
-
|
Social status
|
|
Age (years)
|
0
|
Age
|
|
Education (years)
|
0
|
Education
|
van Montfort et al. (2016)(32)
|
Female sex (vs male)
|
0
|
Sex
|
|
Age (years)
|
0
|
Age
|
IMD = Index of Multiple Deprivation (0, the least deprived, to 86, the most deprived); Scale of Status and Prestige (1 to 7, where 1 represents occupations of the highest social standing); + = significant positive effect; 0 = no significant effect; - = significant negative effect
Table 3. Psychological Predictors of Exercise Adherence
Study
|
Predictors
|
Direction
|
Theme
|
Ades et al. (1992)(33)
|
Presence of depression before hospitalization
|
-
|
Depression
|
Brown et al. (2016)(34)
|
Beck Depression Index
|
0
|
Depression
|
Casey et al. (2008)(20)
|
Beck Depression Index (high scores, more depressed)
|
-
|
Depression
|
Cox et al. (2013)(35)
|
Baseline self-efficacy (higher)
|
+
|
Self-efficacy
|
Craike et al. (2016)(21)
|
Role functioning (higher)
|
+
|
|
|
Sexual activity
|
0
|
|
Fan et al. (2008)(22)
|
State-Trait Anxiety Index ≥ 36
|
-
|
Anxiety
|
|
Beck Depression Index ≥ 5
|
-
|
Depression
|
Gallagher et al. (2003)(23)
|
Perceived control
|
0
|
Control
|
|
Personal stressful event
|
-
|
|
Hogg et al. (2012)(24)
|
Hospital Anxiety and Depression Score "Not depressed" 0-7
|
reference
|
Depression
|
|
“Risk of depression” 8-10
|
0
|
|
|
“Depressed” 11
|
-
|
|
Messer et al. (2007)(36)
|
Task self-efficacy summary scores (higher)
|
+
|
Self-efficacy
|
|
Regulatory self-efficacy summary scores (higher)
|
+
|
|
|
Knowledge self-efficacy
|
0
|
|
van Montfort et al. (2016)(32)
|
Positive affect
|
0
|
|
Pakzad et al. (2013)(27)
|
State-Trait Anxiety Index (higher)
|
+
|
Anxiety
|
|
Consequences
|
0
|
|
|
Chronology (acute/chronic)
|
0
|
|
|
Treatment control
|
0
|
|
|
Personal control
|
0
|
|
Pickering et al. (2013)(28)
|
EQ-5D state of health thermometer
|
+
|
|
|
EQ-5D No pain/discomfort
|
reference
|
|
|
EQ-5D Moderate pain/discomfort
|
0
|
|
|
EQ-5D Extreme pain/discomfort
|
-
|
|
|
EQ-5D Not anxious/depressed
|
reference
|
Anxiety, Depression
|
|
EQ-5D Moderate anxious/depressed
|
-
|
|
|
EQ-5D Extreme anxious/depressed
|
-
|
|
|
Mental health problem (self-reported)
|
-
|
Mental health
|
Selzler et al. (2012)(30)
|
Social functioning (36-Item Short Form Survey)
|
+
|
|
|
Mental health (36-Item Short Form Survey)
|
+
|
Mental health
|
|
Role emotional (36-Item Short Form Survey)
|
+
|
|
Selzler et al. (2016)(29)
|
Task self-efficacy
|
+
|
Self-efficacy
|
|
Coping self-efficacy
|
0
|
|
|
Scheduling self-efficacy
|
0
|
|
Tooth et al. (1992)(31)
|
Expectations (higher)
|
+
|
|
|
Psychological status (profile of mood states score)
|
0
|
|
+ = significant positive effect; 0 = no significant effect; - = significant negative effect
GRADE recommendations
Prognostic factors, categorized by themes, reported by at least 2 observational studies were assessed using the GRADE framework (Table 4).
Table 4. Grading of Recommendations Assessment, Development and Evaluation
Predictors
|
Participants
|
Studies
|
+
|
0
|
-
|
Phase
|
Limitations
|
Inconsistency
|
Indirectness
|
Imprecision
|
Publication bias
|
↑ effect size
|
Dose effect
|
Quality
|
Effect
|
Demographic
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Age (older)
|
3591
|
10
|
2
|
6
|
2
|
2
|
✓
|
X
|
✓
|
✓
|
✓
|
Æ
|
Æ
|
M
|
~
|
Sex (male)
|
2487
|
6
|
|
6
|
|
2
|
✓
|
✓
|
✓
|
✓
|
✓
|
Æ
|
Æ
|
H
|
~
|
Employed
|
936
|
3
|
1
|
1
|
1
|
2
|
✓
|
X
|
✓
|
✓
|
X
|
Æ
|
Æ
|
L
|
~
|
More education
|
1300
|
3
|
1
|
2
|
|
2
|
✓
|
X
|
✓
|
X
|
X
|
Æ
|
Æ
|
VL
|
~
|
Living alone
|
210
|
2
|
|
2
|
|
2
|
✓
|
X
|
✓
|
✓
|
X
|
Æ
|
Æ
|
L
|
~
|
Lower SES
|
842
|
2
|
|
|
2
|
2
|
✓
|
✓
|
✓
|
X
|
X
|
Æ
|
Æ
|
L
|
↓
|
Psychological
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Anxiety
|
1318
|
3
|
1
|
|
2
|
2
|
✓
|
X
|
✓
|
✓
|
X
|
Æ
|
Æ
|
L
|
~
|
Depression
|
3366
|
6
|
|
1
|
5
|
2
|
✓
|
✓
|
✓
|
✓
|
✓
|
Æ
|
D
|
H
|
↓
|
Higher self-efficacy
|
313
|
3
|
3
|
|
|
2
|
✓
|
✓
|
✓
|
X
|
X
|
Æ
|
D
|
M
|
↑
|
Higher control
|
226
|
2
|
|
2
|
|
2
|
✓
|
✓
|
✓
|
X
|
X
|
Æ
|
Æ
|
L
|
~
|
Good mental health
|
884
|
2
|
2
|
|
|
2
|
✓
|
✓
|
✓
|
✓
|
X
|
Æ
|
Æ
|
M
|
↑
|
Comorbidities
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
High BMI
|
1848
|
3
|
|
3
|
|
2
|
✓
|
✓
|
✓
|
✓
|
X
|
Æ
|
Æ
|
M
|
~
|
Smoker
|
1446
|
5
|
1
|
2
|
2
|
2
|
✓
|
X
|
✓
|
✓
|
X
|
Æ
|
Æ
|
L
|
~
|
High cholesterol
|
158
|
3
|
1
|
2
|
|
2
|
✓
|
X
|
✓
|
X
|
X
|
Æ
|
Æ
|
VL
|
~
|
Hypertension
|
128
|
2
|
|
2
|
|
2
|
✓
|
✓
|
✓
|
X
|
X
|
Æ
|
Æ
|
L
|
~
|
higher CCI
|
1268
|
2
|
|
2
|
|
2
|
✓
|
✓
|
✓
|
✓
|
X
|
Æ
|
Æ
|
M
|
~
|
Condition severity
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Better respiratory function
|
878
|
2
|
1
|
1
|
|
2
|
✓
|
X
|
✓
|
✓
|
X
|
Æ
|
Æ
|
L
|
~
|
Higer FEV1
|
1658
|
2
|
1
|
1
|
|
2
|
✓
|
X
|
✓
|
✓
|
X
|
Æ
|
Æ
|
L
|
~
|
Program
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Farther distance
|
1444
|
2
|
|
|
2
|
2
|
✓
|
✓
|
✓
|
✓
|
X
|
Æ
|
Æ
|
M
|
↓
|
Continuous exercise (vs intermittent)
|
75
|
2*
|
1
|
|
1
|
2
|
✓
|
X
|
✓
|
✓
|
X
|
Æ
|
Æ
|
M
|
~
|
Other
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Exercise history
|
160
|
2
|
1
|
1
|
|
2
|
✓
|
X
|
✓
|
X
|
X
|
Æ
|
Æ
|
VL
|
~
|
+ = number of studies with a significant positive effect; 0 = number of studies with no significant effect; - = number of studies with a significant negative effect; ✓ = no serious limitations; X = serious limitations; D = present; Æ = not present; * randomized controlled trials
|
Demographics
Demographic predictors included age, sex or gender, employment, education, social status and living situation. There was low-quality evidence that lower socioeconomic status predicted lower adherence. High-quality evidence suggested that sex was not predictive of adherence and moderate-quality evidence suggests that age does not predict adherence. Low, low and very low-quality evidence, respectively, suggested a lack of prediction of adherence for employment status, living status and education.
Psychological factors
Psychological predictors included anxiety, depression, self-efficacy, control, and self-rated mental health. High-quality evidence supported a negative association between the presence of depression and adherence. Individuals who had good self-rated mental health and who had good self-efficacy were more likely to be adherent (moderate-quality evidence). Low-quality evidence suggested that anxiety and perception of control did not predict adherence.
Comorbidities
Identified comorbidities reported as predictors of exercise adherence were Body Mass Index (BMI), smoking status, hypercholesterolemia, hypertension, and Charleston Comorbidity Index (CCI). None of these were predictive of exercise adherence, which was supported by moderate-quality evidence for BMI and CCI, low-quality evidence for smoking status and hypertension, and very low-quality evidence for hypercholesterolemia. Frailty was not assessed or reported in any of the studies.
Medical condition severity
Measures of respiratory disease severity were not found to be predictive of adherence, but this was only supported by low-quality evidence.
Program factors
The type of exercise program (continuous vs interval exercise) was evaluated by two randomized controlled trials. Although randomized trials are considered to provide high-quality evidence, we downgraded the evidence of no association to moderate quality, given that trial findings were contradictory (one trial reported better adherence to interval exercise, one reported better adherence with continuous exercise). Moderate-quality evidence suggests that living a further distance from the exercise facility decreased adherence.
Other
Low-quality evidence suggests that a history of exercise participation is not predictive of exercise adherence.
Risk of bias within studies
Nine observational studies were deemed to be at low risk of bias and 10 were at moderate risk of bias; no studies were at high risk of bias (Supplementary Table S5). Importantly, prognostic factor measurement and study confounding components of the tool scored low risk of bias across all studies. All four randomized trials were assessed as high risk of bias due to lack of blinding, however, this is recognizably difficult in exercise interventions (Supplementary Table S6). All other domains were low or unclear risk of bias.