Cardiac rehabilitation (CR) is an empirically supported efficacious secondary prevention program aimed at reducing cardiovascular disease (CVD) morbidity and mortality for patients with CVD.1,2 Patients are referred for outpatient CR including those with acute coronary syndrome, coronary artery bypass graft (CABG), percutaneous coronary intervention, stable angina pectoris, valvular repair, heart transplant or heart and lung transplant, congestive heart failure, and peripheral artery disease.3 The goals of outpatient CR are to improve cardiovascular health and achieve optimal physical, psychological, and social functioning in a supervised outpatient setting.4
Outpatient CR sessions typically include monitored cardiovascular exercise (eg, electrocardiography) and possible resistance training under the direct supervision of medical staff, nutritional counseling, and risk factor education and management.4 The American Heart Association and cardiovascular societies worldwide recommend outpatient CR as a class I recommendation for all CHD patients.1 In the United States, outpatient CR typically consists of 36 sessions over the course of 12 wk and is usually covered by private insurance or reimbursed through Medicare.5 Exercise-based outpatient CR is also utilized internationally, similar to outpatient CR in the United States, though the suggested duration and number of sessions vary somewhat by country.4,5
Benefits of completing outpatient CR include improved quality of life and psychosocial well-being,6 reduced depression,7 behavioral risk factor reduction,8,9 and reductions in cardiac mortality and all-cause mortality.10,11 For example, in a sample of 30 161 elderly Medicare beneficiaries, completion of exercised-based outpatient CR was associated with a 50% reduction in subsequent coronary events, a 35% reduction in cardiac mortality, and a 25% reduction in all-cause mortality.10 In addition, a dose-response relationship between number of CR sessions attended and subsequent myocardial infarctions (MI) and all-cause mortality was found.11 Nonetheless, among those who initiate outpatient CR (attend first intake session), attrition is very common. In the United States and Canada, noncompletion rates of 50% or higher are observed.12 Several sociodemographic and medical variables have been investigated regarding CR adherence. For instance, patients who are younger, white, male, more educated, employed, and insured have better CR attendance.13–15 Distance from CR and transportation issues are also associated with CR completion.16 Behaviorally, current smoking, diabetes, and obesity have been studied in relation to outpatient CR completion. Furthermore, medical factors such as no CABG procedure, not having a MI index diagnosis, and history of CVD also appear to be associated with poor completion.15,17
Accumulating research has also been conducted examining the relationship between depression and outpatient CR completion. However, a meta-analysis on this topic has yet to be conducted, which appears warranted because depression is not only common in patients with CVD diagnoses such as CHD18,19 and congestive heart failure20 but also associated with recurrent coronary events and increased cardiac mortality.20–22 In addition, depression in CVD patients may interfere with treatment compliance including reduced adherence to physician recommendations,23 reduced adherence to medical treatments,24 and reduced participation in physical exercise and reduced CR participation.25 Thus, the purpose of the current study is to conduct a systematic literature review and meta-analysis assessing the relationship between depression and outpatient CR program completion among patients who have initiated outpatient CR.
SEARCH STRATEGY AND DATA SOURCES
A systematic literature review and meta-analysis were conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement.26 Databases searched included PsychINFO, MEDLINE, and Dissertation Abstracts International. Thus, published and unpublished studies were included. The umbrella term “depress” was used to capture all studies that had a title related to depression or depressive symptomatology. This was cross-referenced with the terms “cardiac rehabilitation,” “cardiovascular rehabilitation,” “secondary prevention,” and “adherence.” Also, reference lists of located articles were searched by hand for additional research that may not have been identified in the original electronic search.
To be included, studies had to assess the relationship between depression and outpatient CR completion and were required to meet the following criteria: (1) written in English; (2) identified in literature searches up to February 15, 2015; (3) provide a measure of statistical relationship between depression and CR attendance; (4) and provide sufficient information to compute effect sizes. Articles were excluded if they (1) assessed inpatient CR; (2) assessed only outpatient CR initiation, not CR completion; or (3) assessed inpatient and outpatient CR but did not analyze outpatient CR independently. For studies in which reported data were unclear, incomplete, or missing, study authors were contacted to obtain clarification or additional data.
Study characteristics were extracted and coded with a standardized coding form created and pilot tested on 3 randomly selected articles by 2 independent raters (B.E. and A.G.). Studies were coded on the basis of potential moderators obtained from CR literature, including variables related to the independent variable (eg, type of depression measure, dichotomization of depression, reported antidepressant use); the dependent variable (eg, length of CR program, operationalization of CR completion); and additional possible moderators including sociodemographic variables (eg, age, sex, race/ethnicity, income, education level, insurance status, employment, relationship status, distance living from CR clinic); relevant medical variables (eg, index cardiac event [MI vs no MI]; index cardiac diagnoses (all CHD diagnoses vs other CVD diagnoses); CHD surgical or medical intervention (eg, CABG, percutaneous coronary intervention, medication management); number/prevalence of modifiable CHD risk factors (diabetes, hypertension, hyperlipidemia, and current cigarette smoking); and whether the study was peer-reviewed (published vs unpublished). B.E. and A.G. extracted coded study characteristics independently. To assess interrater agreement, Cohen κ was calculated (κ= 0.83, P < .001), which exceeded the recommended value of 0.80.27 Items with disagreement were reviewed and resolved by discussion between raters.
A quality assessment tool was created using recommendations outlined by PRISMA,26 Meta-Analysis of Observational Studies in Epidemiology (MOOSE),28 Quality Assessment Tool for Systematic Reviews of Observational Studies (QATSO),29 and Committee for Strengthening the Reporting of Observational Studies in Epidemiology (STROBE).30 Each study was evaluated on 7 individual items, each scoring 1 point for the following: (1) eligibility criteria defined, type of sampling, and dates of data collection; (2) used a validated measure of current depressive symptomatology; (3) provided rationale for operational definition of completion; (4) reported descriptive data on basic sociodemographic variables (age, sex, and race/ethnicity); (5) reported relevant cardiac variables (diabetes, hypertension, hyperlipidemia, smoking status, and medical intervention/procedures); (6) provided 95% CI; and (7) controlled for confounding variables. Two independent raters (B.E. and S.S.) evaluated each study independently using this tailored quality assessment. A score of 0 to 2 points indicated low study quality, 3 to 4 points indicated moderate quality, and 5 to 7 points indicated high quality. The average study quality rating score of 4.3 points indicated that most of the included studies were of moderate quality and interrater agreement was acceptable (κ= 0.95, P < .001).
DATA SYNTHESIS AND ANALYSIS
To quantify the relationship between depression and CR completion, Hedges' g statistic was used as an effect size measure to calculate group contrasts, representing the difference in CR completion between groups of depressed and nondepressed CR participants. Since 11 of the 19 independent effect size calculations across the 17 included studies reported group differences between CR completers and noncompleters, mean depression scores and standard deviations between these groups were chosen as the preferred method of effect size calculation. For the other studies that did not report depressive score means and standard deviations between CR attenders and nonattenders, a Hedges' g was calculated from a combination of t test, F, and P values; proportions; Pearson r; and odds ratios.
A random-effects model was used to obtain an overall summary effect measure estimating the relationship between depression and outpatient CR completion. Heterogeneity was assessed using the Q statistic, which detects whether variability is significantly greater than by chance alone. In addition, the I2 statistic was used to interpret the proportion of the observed variance representing true effects rather than sampling error. Considering that participation in CR has been shown to be moderated by several sociodemographic and medical variables,13–15 a mixed-effects model was chosen to be used for an accurate and detailed assessment of moderators, in addition to the summary estimate in the population. To reduce statistical error, the following a priori criteria were created for moderator analyses: each variable must have been reported in a minimum of 10 studies (κ≥ 10), with a minimum cell size of 3 (κ≥ 3) for categorical comparisons, consistent with recommended minimum cell sizes.31
Publication bias was assessed using a funnel plot, created by plotting each sample's estimated summary effect against the standard error to examine whether any asymmetry is present. A trim-and-fill analysis was conducted to further explore publication bias by removing the studies that contribute to the asymmetric funnel plot and readjusting the mean summary effect to the remaining studies.32 Cohen κ27 and descriptive data were analyzed on SPSS v22.0 (IBM).33 Meta-analysis, moderator analyses, and publication bias analyses were calculated using Comprehensive Meta-Analysis (Biostat) software, v3.0.34
The literature search identified 1447 titles. After duplicates were removed (i = 278) and records were screened (κ= 1169), 117 full studies containing original data were identified for possible inclusion. One hundred studies were excluded for reasons, such as the study reported on initiation of CR (not CR attendance) or the study did not provide quantitative data for the relationship between depression and CR attendance to allow for calculation of effect size. For the 31 studies in the latter category, corresponding authors were e-mailed for additional data. Ten authors responded and indicated that the requested data were not available; the remaining 21 authors did not respond. Thus, no additional studies were able to be included in the current meta-analysis on the basis of these author contacts. In conclusion, 17 studies in total were identified for inclusion in the current meta-analysis (Figure 1).
STUDY AND PATIENT CHARACTERISTICS
Baseline characteristics of included studies are listed in the Table. There were 19 independent effect sizes drawn from the 17 included studies,35–51 totaling 30 586 outpatient CR participants, with sample sizes ranging from 35 to 26 957. Two studies46,48 provided and analyzed data for 2 different samples, resulting in a total of 19 samples across the 17 studies. Average overall outpatient CR completion rate was 67.9%. The 17 studies included in the meta-analysis were published between 1982 and 2014. Fourteen were published in peer-reviewed journals, 2 were doctoral dissertations, and 1 was a master's degree thesis. Seventeen studies were conducted in the United States, 1 in the United Kingdom, and 1 in Canada.
Of the 17 studies, all reported sex with the total sample population consisting of 31% females (n = 9460) and a sample population mean age of 61.9 yr. Of the 12 samples that reported ethnicity, patients were 86.5% white (n = 26 450). Other relevant psychosocial variables were reported less frequently: relationship status (κ= 6), patient education level (κ= 3), and employment status data (κ= 3). Relevant medical variables were reported across the 19 samples as follows: index CHD event (κ= 11), index CHD diagnosis (κ= 12), CABG procedure (κ= 10), percutaneous coronary intervention procedure (κ= 4), and CR patient antidepressant medication use (κ= 6). Relatively few studies reported the 4 major independent modifiable risk factors for CHD (hyperlipidemia, κ= 6; hypertension, κ= 5; diabetes, κ= 6; and current smoking, κ= 8).
Across the 17 studies, depression was measured by 5 different assessment tools including Beck Depression Inventory (BDI; κ= 6), Center for Epidemiological Studies—Depression (CES-D; κ= 3), Hospital Anxiety and Depression Scale—Depression (HADS-D; κ= 2), Kellner Symptom Questionnaire (KSQ; κ= 2), and the Minnesota Multiphasic Personality Inventory—Depression Scale (MMPI-D; κ= 1). One study used self-report of historic or current antidepressant use (κ= 1). The operational definition of depression was measured and analyzed continuously in 16 of the samples and measured dichotomously in 3 samples. Outpatient CR completion was assessed and analyzed as a dichotomous variable in 16 samples; however, across these 16 samples, the dichotomized operational definition of CR completion varied considerably, including attending 100% of prescribed sessions plus passing a physical assessment at discharge (κ= 1); attending 100% versus <100% of prescribed sessions (κ= 6); case manager discharge and/or 83% attendance of prescribed sessions (κ= 1); attending a ≥70% or 75% of prescribed sessions (κ= 2); attending ≥66% of prescribed sessions (κ= 1); attending >7 wk (κ= 1); and discharge by medical staff (κ= 4). Outpatient CR completion was measured and analyzed continuously in the other 3 studies, calculated as the total number of sessions attended. The length of CR program was reported in all samples and varied between 6 wk and 1 yr: 6 wk (κ= 1), 12 wk (κ= 12), 24 wk (κ= 1), and 1 yr (κ= 1); the remaining studies used individually tailored CR durations (κ= 4).
The weighted mean summary effect for the random-effects meta-analysis indicated that depressed CR participants were significantly less likely to complete their prescribed program (g = −0.44, P < .001, 95% CI, −0.59 to −0.29) (Figure 2). All but 1 study43 used validated psychometric questionnaires of depressive symptomatology; this 1 study43 operationalized depression as self-report of current or historical antidepressant use. Given this major difference in the operationalization of the independent variable, a decision was made that a sensitivity analysis excluding this 1 study was warranted. Hence, a weighted means summary effect was recalculated on the 18 independent effect sizes derived from 16 individual studies (n = 3036), excluding this 1 study.43 This result revealed a minimal increase in effect size (g = −0.47, P < .001, 95% CI, −0.59 to −0.33).
A significant amount of heterogeneity was found among the effect sizes, Q (18) = 86.99, P < .001, which confirmed our a priori choice to use a random-effects model. Using the I2 statistic, 79.3% of the variability in the overall effect size was due to study-level characteristics, thus warranting moderator analyses. Eight potential categorical moderator variables met the a priori criteria of 10 studies (κ≥ 10) per moderator and a minimum cell size of 3 (κ≥ 3)31; index cardiac diagnosis (CHD only vs all CVD diagnoses); depression measure used (eg, BDI, CES-D); dichotomization of depression (depressed/not depressed) versus a continuous total depression score; length of CR program (ie, 12 wk); operationalization of CR completion (100% vs <100% sessions attended); qualitative versus quantitative CR completion (case manager discharge vs a quantitative measure of number of sessions completed); dichotomization of completion (complete/not complete) versus number of sessions attended; and peer-review status (peer-reviewed vs unpublished studies). None of these categorical moderator analyses were significant.
Five potential continuous moderator variables met the criteria of 10 studies per moderator (κ≥ 10)31: mean age, percent female, percent Caucasian, percent index cardiac event (ie, MI), and percent CABG procedure. Five separate random-effects meta-regression analyses were performed in which each continuous variable was correlated with its corresponding effect size. None of these meta-regression analyses were significant.
A minimal amount of publication bias was detected by visual inspection of the funnel plot. The Duval and Tweedie trim-and-fill adjustment method was calculated.32 After 5 studies were trimmed and imputed, the summary effect size decreased from g = −0.44 (95% CI, −0.59 to −0.29) to g = −0.33 (95% CI, −0.47 to −0.19), suggesting that a minor amount of publication bias was present in the original omnibus model (Figure 3).
Our current study is the first systematic literature review and meta-analysis conducted assessing the relationship between depression and outpatient CR completion. Overall, a moderate effect size was detected, indicating that depressed CR patients are significantly less likely to complete outpatient CR. By comparison, several sociodemographic variables found to be associated with reduced outpatient CR completion in a meta-analysis such as sex, education, and employment status all have small effect sizes.14 Also, the moderate effect size found in the current study is similar in magnitude, if not larger, than those found in prior research on depression and adherence to medical treatment in cardiac patients. For example, depression in CHD patients is associated with lifestyle behaviors such as sedentary behavior and medication nonadherence, both with small-to-medium effect sizes.52,53
A follow-up sensitivity analysis was run excluding a single study43 that operationally defined the presence of depression as patient antidepressant use, historical or current. This sensitivity analysis was conducted because all other studies used a psychometric assessment specifically for current depression and results of this second omnibus analysis showed a minor increase in the summary effect size and a tightening of confidence intervals. Thus, the overall effect size in this sensitivity analysis remained moderate. These findings are important, because patients who complete outpatient CR experience reduced recurrent coronary events and have lower cardiac and all-cause mortality rates.8,10,11,13 In general, our findings add further evidence that implicates depression as an important factor in CVD patients related to poorer patient compliance and worse cardiac outcomes.21,23–25
A high degree of heterogeneity between studies was indicated by the Q and I2 statistics, which would be expected given that most studies were observational with varying methodological designs and populations.54 However, none of the moderator analyses were statistically significant. Results also indicated the possibility of a minor amount of publication bias; nevertheless, the effect size after the trim-and-fill procedure remained moderate. Thus, publication bias did not appear to account for the relationship between depression and CR completion. Finally, our quality assessment indicated that on average, studies were of moderate quality.
STRENGTHS OF CURRENT STUDY
The current study followed PRISMA guidelines for best practices for systematic literature reviews and meta-analyses.26 Interrater reliability for data extraction met or exceeded standard reliability criteria.27 The systematic literature review component of the current project was diligent in seeking out further data on articles that were questionable in terms of meeting inclusion criteria. Specifically, for articles that possibly met basic inclusion criteria but did not appear to provide sufficient data to be included (eg, no inferential statistic provided that quantified the relationship of depression and outpatient CR completion), we contacted authors to obtain the missing data. Furthermore, for 10 studies that met inclusion criteria, but which had ambiguous reporting of relevant data, the current authors contacted study authors by e-mail to confirm that correct data were extracted. Finally, the current meta-analysis followed recommended a priori criteria regarding total number of samples required to conduct moderator analyses (eg, minimum of 10 samples per variable; 3 studies per cell or subgroup) to ensure the preservation of statistical integrity and to avoid statistical error or alpha slippage.31
LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH
One limitation of the current research is that it focused solely on research examining the relationship between depression and outpatient CR completion, that is, completion among patients who had attended an initial intake session. We did not include the related, yet separate, research literature examining depression and outpatient CR initiation among all eligible patients, which would be important area of investigation in a future systematic literature review and meta-analysis. Second, in choosing variables for moderator analyses, variables typically associated with outpatient CR completion or recurrent coronary events (eg, education as an indicator of socioeconomic status, race/ethnicity, modifiable CHD risk factors such as cigarette smoking)13,14,17 could not be analyzed because an insufficient number of studies included these variables. Future research on depression and outpatient CR completion should include such relevant variables to allow for more comprehensive moderator analyses in future meta-analyses.14
Third, studies that were written only in English were included in the current meta-analysis, limiting generalization of results to non-English–speaking countries. Fourth, no studies of depression and home-based CR completion were found in the literature. Because home-based CR has been shown to be equally as efficacious as traditional clinic-based CR,55 it thus remains undetermined whether depression impacts completion of home-based CR. Fifth, the date range of included studies in the current systematic literature review was through February 2015.
Finally, the biggest limitation in the literature reviewed was the disparate operational definitions regarding both depression and outpatient CR completion. While all studies but 1 assessed depression using validated psychometric instruments, 3 studies dichotomized scale scores into depressed/nondepressed, and the remainder reported depression as a continuous variable. Even more troubling, 16 of the 19 samples assessed outpatient CR completion dichotomously (ie, noncompleter/completer) rather than continuously. Dichotomization of continuous metrics may reduce the ability to accurately identify the true effect of depression on CR completion, as statistical artifacts reduce sensitivity and may contribute to biased results.56–58 Moreover, operationalizing depression and outpatient CR completion as dichotomous variables prohibits analysis to determine whether there is a dose-response relationship between depression and number of outpatient CR sessions completed.
Most perplexing in this literature was the variability in the operational definition of outpatient CR completion when this variable was dichotomized. For example, dichotomous completion was operationally defined in 8 different ways across the studies (eg, >66% of sessions attended; >75% of sessions attended; 100% of sessions attended). Generally, no rationale was provided regarding why these cutoffs were chosen. It is recommended that future research on depression and outpatient CR completion report and analyze these variables continuously.
To address these and other issues of transparency in future research in this area, it is recommended that studies use best-practice standards such as that outlined by the STROBE30 or Consolidated Standards of Reporting Trials (CONSORT),57 including explanations of participant eligibility criteria, variables explored, analytical/methodological choices, thorough reporting of participant descriptive data, tables outlining baseline sociodemographic and medical characteristics, and prespecified outcome measures.
Prior research has found that interventions focused on depressive symptomatology in patients with coronary artery disease (eg, cognitive behavioral therapy, antidepressant medications) appear to have a small to medium effect size in reducing depression59,60 and possibly reduce recurrent coronary events60; however, they have demonstrated no impact on mortality.60 In contrast, outpatient CR is not only effective in reducing depression6,61,62 but also has been shown to reduce recurrent coronary events and lower risk of mortality.6,10,13,60,62 Thus, future research should examine the impact of attendance interventions63 for depressed cardiac patients to determine whether such interventions may simultaneously increase CR completion rates and decrease depressive symptomatology. If such interventions are successful, subsequent research could examine whether such CR attendance interventions targeted at depressed patients result in reductions in subsequent coronary events and mortality in these patients. Finally, cardiologists who routinely screen for depression may want to consider a concerted effort to refer their depressed patients for outpatient CR, given its impact on depression, recurrent events, and mortality.64
1. Anderson L, Oldridge N, Thompson DR, et al Exercise-based cardiac rehabilitation
for coronary heart disease: Cochrane systematic review and meta-analysis
. J Am Coll Cardiol. 2016;67:1–12.
2. Anderson LJ, Taylor RS. Cardiac rehabilitation
for people with heart disease: an overview of Cochrane systematic reviews. Int J Cardiol. 2014;177:348–361.
3. Price KJ, Gordon BA, Bird SR, Benson AC. A review of guidelines for cardiac rehabilitation
exercise programmes: is there an international consensus? Eur J Prev Cardiol. 2016;23:1715–1733.
4. Dalal HM, Doherty P, Taylor RS. Cardiac rehabilitation
. BMJ. 2015;351:h5000.
5. Mampuya WM. Cardiac rehabilitation
past, present and future. Cardiovasc Diagn Ther. 2012;2:38–49.
6. Lavie CJ, Menezes AR, De Schutter A, Milani RV, Blumenthal JA. Impact of cardiac rehabilitation
and exercise training on psychological risk factors and subsequent prognosis in patients with cardiovascular disease. Can J Cardiol. 2016;32:S365–S373.
7. Gellis ZD, Kang-Yi C. Meta-analysis
of the effect of cardiac rehabilitation
interventions on depressions outcomes in adults 64 years of age and older. J Am Coll Cardiol. 2012;110:1219–1224.
8. Chatziefstratiou AA, Gakoumidakis K, Brokalaki H. Cardiac rehabilitation
outcomes: modifiable risk factors. Br J Nurs. 2013;22:200–207.
9. Silberman A, Banthia R, Estay IS, et al The effectiveness and efficacy of an intensive cardiac rehabilitation
program in 24 sites. Am J Health Promot. 2010;24:260–266.
10. Lawler PR, Filion KB, Eisenberg MJ. Efficacy of exercise-based cardiac rehabilitation
post-myocardial infarction: a systematic review and meta-analysis
of randomized controlled trials. Am Heart J. 2011;162:571–584.
11. Hammill BG, Curtis LH, Schulman KA, Whellan DJ. Relationship between cardiac rehabilitation
and long-term risks of death and myocardial infarction among elderly Medicare beneficiaries. Circulation. 2010;121:63–70.
12. Martin BJ, Hauer T, Arena R, et al Cardiac rehabilitation
attendance and outcomes in coronary artery disease patients. Circulation. 2012;126:677–687.
13. Menezes AR, Lavie CJ, Milani RV, Forman DE, King M, Williams MA. Cardiac rehabilitation
in the United States. Prog Cardiovasc Dis. 2014;56:522–529.
14. Sun EY, Jadotte YT, Halperin W. Disparities in cardiac rehabilitation
participation in the United States. J Cardiopulm Rehabil Prev. 2017;37:2–10.
15. Park LG, Schopfer DW, Zhang N, Shen H, Whooley MA. Participation in cardiac rehabilitation
among patients with heart failure. J Card Fail. 2017;23:427–431.
16. Clark AM, King-Shier KM, Spaling MA, et al Factors influencing participation in cardiac rehabilitation
programmes after referral and initial attendance: qualitative systematic review and meta-analysis
. Clin Rehabil. 2013;27:948–959.
17. Taylor GH, Wilson SL, Sharp J. Medical, psychological, and sociodemographic factors associated with adherence to cardiac rehabilitation
programs: a systematic review. J Cardiovasc Nurs. 2011;26:202–209.
18. Barth J, Schumacher M, Herrmann-Lingen C. Depression
as a risk factor for mortality in patients with coronary heart disease: a meta-analysis
. Psychosom Med. 2004;66:802–813.
19. Van der Kooy K, van Hout H, Marwijk H, Marten H, Stehouwer C, Beekman A. Depression
and the risk for cardiovascular diseases: systematic review and meta-analysis
. Int J Geriatr Psychiatry. 2007;22:613–626.
20. Sokoreli I, de Vries JJ, Pauws SC, Steyerberg EW. Depression
and anxiety as predictors of mortality among heart failure patients: systematic review and meta-analysis
. Heart Fail Rev. 2016;21:49–63.
21. Lichtman JH, Froelicher ES, Blumenthal JA, et al Depression
as a risk factor for poor prognosis among patients with acute coronary syndrome: systematic review and recommendations. A scientific statement from the American Heart Association. Circulation. 2014;129:1350–1369.
22. Meijer A, Conradi HJ, Bos EH, Thombs BD, van Melle JP, de Jonge P. Prognostic association of depression
following myocardial infarction with mortality and cardiovascular events: a meta-analysis
of 25 years of research. Gen Hosp Psychiatry. 2011;33:203–216.
23. Zieglestein RC, Fauerbach JA, Stevens SS, Romanelli J, Richter DP, Bush DE. Patients with depression
are less likely to follow recommendations to reduce cardiac risk during recovery from a myocardial infarction. Arch Intern Med. 2000;160:1818–1823.
24. Gehi A, Haas D, Pipkin S, Whooley MA. Depression
and medication adherence in outpatients with coronary heart disease: findings from the Heart and Soul Study. Arch Intern Med. 2005;165:2508–2513.
25. Myers V, Gerber Y, Benyamini Y, Goldbourt U, Drory Y. Post-myocardial infarction depression
: increased hospital admissions and reduced adoption of secondary prevention measures—a longitudinal study. J Psychosom Res. 2012;72:5–10.
26. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: the PRISMA Statement. Ann Intern Med. 2009;151:264–269.
27. Cohen J. Weighted kappa: nominal scale agreement provision for scale disagreement or partial credit. Psychol Bull. 1968;70:213–220.
28. Stroup DF, Berlin JA, Morton SC, et al Meta-analysis
of observational studies in epidemiology: a proposal for reporting. J Am Med Assoc. 2000;283:2008–2012.
29. Wong WC, Cheung CS, Hart GJ. Development of a quality assessment tool for systematic reviews of observational studies (QATSO) of HIV prevalence in men having sex with men and associated risk behaviours. Emerg Themes Epidemiol. 2008;5:1–4.
30. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Prev Med. 2007;45:247–251.
31. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Introduction to Meta-Analysis
. West Sussex, United Kingdom: John Wiley & Sons Ltd; 2009.
32. Duval S, Tweedie R. Trim and fill: a simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis
. Biometrics. 2000;56:455–463.
35. Anderson DR, Emery CF. Irrational health beliefs predict adherence to cardiac rehabilitation
: a pilot study. Health Psychol. 2014;33:1614–1617.
36. Beckie TM, Fletcher G, Groer MW, Kip KE, Ji M. Biopsychosocial health disparities among young women enrolled in cardiac rehabilitation
. J Cardiopulm Rehabil Prev. 2014;34:1–10.
37. Blumenthal JA, Williams RS, Wallace AG, Williams RB Jr, Needles TL. Physiology and psychological variables predict compliance to prescribe exercise therapy in patients recovering from myocardial infarction. Psychosom Med. 1982;44:519–527.
38. Casey E, Hughes J, Waechter D, Josephson R, Rosneck J. Depression
predicts failure to complete phase-II cardiac rehabilitation
. J Behav Med. 2008;31:421–431.
39. Casey M. The Role of Depression
in Cardiac Rehabilitation
Attendance [unpublished master's thesis]. Flagstaff, AZ: Northern Arizona University; 2011.
40. Caulin-Glaser T, Maciejewski PK, Snow R, LaLonde M, Mazure C. Depressive symptoms and sex affect completion rates and clinical outcomes in cardiac rehabilitation
. Prev Cardiol. 2007;10:15–21.
41. DeYoung NJ. The Role of Rumination, Negative Affect, and Fitness on Cardiac Rehabilitation
Program Outcomes Following a Discrete Cardiac Event [doctoral dissertation]. ProQuest LLC, Ann Arbor, Michigan. 2013. Retrieved from PsycINFO. (3606098).
42. Glazer KM, Emery CF, Frid DJ, Banqasz RE. Psychological predictors of adherence and outcomes among patients in cardiac rehabilitation
. J Cardiopulm Rehabil Prev. 2002;22:40–46.
43. Gordon NF, Habib A, Salmon RD, et al Effect of exercise-based cardiac rehabilitation
on multiple atherosclerotic risk factors in patients taking antidepressant medication. J Am Coll Cardiol. 2013;111:346–351.
44. Jackson JL. Influence Traits, Coping, Affect, and Illness Knowledge on Adherence Among Patients in Cardiac Rehabilitation
[doctoral dissertation]. ProQuest LLC, Ann Arbor, Michigan. 2010. Retrieved from PsycINFO. (3428666).
45. Josephson EA, Casey EC, Waechter D, Rosneck J, Hughes JW. Gender and depression
symptoms in cardiac rehabilitation
. J Cardiopulm Rehabil Prev. 2006;26:160–163.
46. Lavie CJ, Milani RV. Adverse psychological and coronary risk profiles in young patients with coronary artery disease and benefits from formal cardiac rehabilitation
. Arch Intern Med. 2006;166:1878–1883.
47. McGrady A, McGinnis R, Badenhop D, Bentle M, Rajput M. Effects of depression
and anxiety on adherence to cardiac rehabilitation
. J Cardiopulm Rehabil Prev. 2009;29:358–364.
48. McGrady A, Burkes R, Badenhop D, McGinnis R. Effects of a brief intervention on retention of patients in a cardiac rehabilitation
program. Appl Psychophysiol Biofeedback. 2014;39:163–170.
49. Sanderson BK, Bittner V. Women in cardiac rehabilitation
: outcomes and identifying risk for dropout. Am Heart J. 2005;150:1052–1058.
50. Swardfager W, Herrmann N, Marzolini S, et al Major depressive disorder predicts completion, adherence, and outcomes in cardiac rehabilitation
: a prospective cohort study of 195 patients with coronary artery disease. J Clin Psychiatry. 2011;72:1181–1188.
51. Yohannes A, Yalfani A, Doherty P, Bundy C. Predictors of drop-out from an outpatient cardiac rehabilitation
programme. Clin Rehabil. 2007;21:222–229.
52. Crawshaw J, Auyeung V, Norton S, Weinman J. Identifying psychosocial predictors of medication non-adherence following acute coronary syndrome: a systematic review and meta-analysis
. J Psychosom Res. 2016;90:10–32.
53. Sin NL, Kumar AD, Gehi AK, Whooley MA. Direction of association between depressive symptoms and lifestyle behaviors in patients with coronary heart disease: The Heart and Soul Study. Ann Behav Med. 2016;50:523–532.
54. Huedo-Medina TB, Sánchez-Meca J, Marín-Martínez F, Botella J. Assessing heterogeneity in meta-analysis
: Q statistic or I² index? Psychol Methods. 2006;11:193–206.
55. Dalal HM, Zawada A, Jolly K, Moxham T, Taylor RS. Home based versus centre based cardiac rehabilitation
: Cochrane systematic review and meta-analysis
. BMJ, 2010;340:1–15.
56. Altman DG, Royston P. The cost of dichotomizing continuous variables. BMJ. 2006;332:1080.
57. Lipsey MW, Wilson DB. Practical Meta-Analysis
. Thousand Oaks, CA: SAGE Publications, Inc; 2001.
58. MacCallum RC, Zhang S, Preacher KJ, Rucker DD. On the practice of dichotomization of quantitative variables. Psychol Methods. 2002;7:19–40.
59. Schulz KF, Altman DG, Moher D; CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ. 2010;340:c332.
60. Baumeister H, Hutter N, Bengel J. Psychological and pharmacological interventions for depression
in patients with coronary artery disease. Cochrane Database Syst Rev. 2011;9:CD008012.
61. Rutledge T, Redwine LS, Linke SE, Mills PJ. A meta-analysis
of mental health treatments and cardiac rehabilitation
for improving clinical outcomes and depression
among patients with coronary heart disease. Psychosom Med. 2013;75:335–349.
62. Lavie RV, Milani CJ. Impact of cardiac rehabilitation
and its associated mortality. Am J Med. 2007;120:799–806.
63. Beswick AD, Rees K, West RR, et al Improving uptake and adherence in cardiac rehabilitation
: literature review. J Adv Nurs. 2005;49:535–555.
64. Lavie RV, Milani CJ, Blumenthal JA. Exercise, cardiac rehabilitation
, and post-acute coronary syndrome depression
. JAMA Intern Med. 2014;174:165–166.