Table 9.

Details of the quantitative studies included in the review regarding: method, participants, setting, summary of findings pertinent to psychological and social factors, and notes regarding theoretical models.

Study	Methods	Participants	Setting	Findings	Notes
Rimmer et al. [34]	Survey: barriers to physical activity and disability survey	N = 83 54.2 (8.2) 30 M, 70 F not reported	USA Health promotion programme	Psychological: competence: not knowing how and where to exercise main personal barriers to exercise, reported by 46% and 44% of participants. Motivation: lack of motivation ranked 4th personal barrier by (37%) of participants. Too lazy to exercise ranked 6th personal barrier by 33% of participants. Beliefs: 36% reported belief that exercise will not improve condition Social: importance of exercise professionals: personal trainer or exercise instructor unable to help reported by 36% of participants	No theoretical model used to underpin questionnaire or findings
Galvin et al. [33]	Survey: family-mediated exercise survey	N = 73 people with stroke 40 F, 35 M Age not reported N = 100 family members/friends 71 F, 29 M	Ireland Hospital physiotherapy services	Social: 91% of respondents with stroke believed that family member/friend had a role in assisting them with exercise after rehabilitation. 99% of family members/friends reported they would be willing to help with exercise after rehabilitation	No theoretical model used to underpin questionnaire or findings Not clear how questionnaire was developed. Validity and reliability not reported
Shaughnessy et al. [14]	Survey using the Short Self-Efficacy for Exercise Scale and the Short Outcome Expectations for Exercise Scale	N = 321 Age 62.9 (11.7) F 177, M 127 60.2 months after stroke	USA Mailed survey distributed via the National Stroke Association	Psychological: self-efficacy, outcome expectations, exercise behaviour before stroke, and physician advice to exercise significantly associated with exercise behaviour. These variables predicted 33% of variance in exercise behaviour. Older participants and those experiencing fatigue had lower self-efficacy for exercise. Exercise history also significantly predicted exercise behaviour	Low response rate from national survey, attended stroke support groups therefore likely to be motivated
Johnston et al. [40]	Longitudinal survey assessing outcomes at 10–20 days, one month after discharge, and six months after discharge Measures: (i) the recovery locus of control scale (ii) Exercise coping self-rating for frequency and duration (iii) HADS (iv) Barthel Index (v) Observer assessed disability	N = 71 Age 69.4 years 35 F, 36 M 10–20 days after stroke to 6 months after stroke	Scotland Acute inpatients followed up after discharge	Psychological: perceived control predicts recovery from disability after stroke. Frequency of exercise not correlated with one month recovery locus of control (P > 0.05), nor with observer assessed recovery at 6 months (P > 0.05), suggests that exercise not a coping response that medicates between control cognitions and recovery	Amount of explained variance is small Some measures designed for the study—validity and reliability not tested fully
Cardinal et al. [37]	Survey National cross-sectional survey of individuals with disability Stage of change algorithm to assess stage, process of change measure, self-efficacy scale, decisional balance scale, and exercise barriers scale	N = 322 52.5 (13.9) years 62% F Disabled population Stroke: n = 18	USA Recruitment by targeting voluntary organisations and hospitals	Psychological: transtheoretical model used to examine what constructs associated with stages. All major constructs associated with the stages of change. Largest portion of variance derived from behavioural processes of change and self-efficacy. Cognitive processes decrease and behavioural processes increase with stage progression. Cons decrease and pros increase with stage progression, self-efficacy increases with stage progression Social: helping relationships significantly associated with stage	Stroke participants formed only small proportion of sample therefore generalisation difficult Unable to extract specific data about behavioural processes and cognitive processes used
Nosek et al. [36]	Survey Physical activity Social support Environmental factors Impairment and functioning (SF-36 physical functioning and role limitations) Psychological factors (SF36 mental health, role functioning emotional, self-efficacy) Social factors CHART short form and MOS social support survey	N = 386 47.1 (10.1) years 386 F Stroke n = 25 (6.5%)	USA National survey recruitment via disability service organisations and print and broadcast media	Psychological: self-efficacy highest for gentle flexibility, lowest for aerobic exercise 3 × per week. Physical activity significantly correlated with self-efficacy (r = 0.50, P < 0.0001) Model predicted 33.5% of variance. Self-efficacy predicts greater involvement in physical activity, pain and duration of disability predictive of less physical activity (F(7,268) = 19.27; P < 0.0001) Social: physical activity significantly correlated with level of personal assistance (F(6,269) = 2.23, P < 0.05)	Physically active population with 73% engaging in physical activity once a week. Self-report of physical activity Disabled sample, self-selected, and few people with stroke
Kinne et al. [38]	Survey Sickness Impact Scale Stage of change Self-rated abilities for health: self-efficacy and outcome expectancy Barriers to health Activities for disabled persons	N = 83 Age 47 (1.4 ) years Stroke 7% of sample	USA Community dwelling, recruited via disability support groups	Psychological: self-efficacy significantly higher for participants maintaining exercise (P < 0.0001) Motivational barriers scores significantly lower for exercise maintainers (P = 0.0006) People with higher motivational barriers less likely to maintain exercise (P = 0.01). Those with higher maintenance self-efficacy higher probability of maintenance Information about what to do a barrier to exercise behaviour ranked 4th behind impairment, money, and accessible facilities	Small proportion of sample are people with stroke Although some participants indicated that they were less active in previous year, reasons for this not explored External factors, access to facilities, transportation, money, social support, and physical and functional status not predictive of maintenance
Boysen et al. [41]	RCT primary outcome: Physical Activity Scale for the Elderly	N = 314 people with stroke able to walk unassisted Intervention group n = 157 Control group N = 157 Age 69.6 years (59.6–77.7) <90 days after onset	Centres in Denmark, China, Poland, and Estonia stroke units	No significant difference between the groups on PASE at 3,6,9,12, 18 or 24 month followup	Self-reported levels of physical activity—not assessed by accelerometer, pedometer, and so forth, or by other fitness measures. No qualitative evaluation of involvement in PA made