Cross-Lagged Relations Between Exercise Capacity and Psychological Distress During Cardiac Rehabilitation

Cross-Lagged Relations Between Exercise Capacity and Psychological Distress During Cardiac... Abstract Background Poorer mental health is associated with lower exercise capacity, above and beyond the effect of other cardiovascular risk factors. However, the directionality of this relationship remains unclear. Purpose The main aim of the present study was to clarify, with a cross-lagged panel design, the relationship between psychological status and exercise capacity among patients in a cardiac rehabilitation (CR) program. Methods A clinical sample of 212 CR patients completed exercise-capacity testing and measures of depression and anxiety (Hospital Anxiety and Depression Scale) pre-CR and post-CR. Demographic and clinical data, including BMI and smoking history, were also collected. Multivariate stepwise regression analysis was performed to identify the best predictors of exercise capacity at discharge. Structural equation modeling was utilized to quantify the cross-lagged effect between exercise capacity and psychological distress. Results Multivariate regression analysis revealed that higher levels of psychological distress pre-CR are predictively associated with less improvement in exercise capacity post-CR, beyond the effects of age, sex, and baseline functional status. Results from structural equation modeling supported a 1-direction association, with psychological distress pre-CR predicting lower exercise capacity post-CR over and above autoregressive effects. Conclusions Study results did not support the hypothesis of a bidirectional relationship between psychological distress and EC. High levels of psychological distress pre-CR appeared to be longitudinally associated with lower exercise capacity post-CR, but not vice versa. This finding highlights the importance of assessing and treating both anxiety and depression in the early phase of secondary prevention programs. Cardiac rehabilitation, Anxiety, Depression, Psychological distress, Exercise capacity Introduction The primary goal of cardiac rehabilitation (CR) is to reduce the cardiac risk profile of patients diagnosed with heart disease. Although CR promotes a multidisciplinary approach to tackle cardiovascular risk factors, structured exercise has been established as the central component of secondary prevention [1]. In fact, exercise-based CR has been found to lower risk of cardiovascular mortality and re-infarction, compared with CR without an exercise component [2, 3]. In this context, the increase in patients’ exercise capacity (EC) triggered by CR plays a central role [4–7]. The influence of psychological factors on cardiovascular improvements among CR patients is well-documented [8, 9], as well as the influence of such factors on the long-term prognosis of cardiac patients [10, 11]. Among these psychological factors, depression and anxiety are among the most widely researched [12, 13]. In a recent meta-analysis, Papasavvas et al. [14] found an association between poorer mental health and lower EC, over and beyond the effect of disease severity, age, and sex. However, the reciprocal relationship between psychological health and EC remains unclear. On the one hand, depression and anxiety seem to predict poorer exercise performance. The Heart and Soul Study [15] on 944 outpatients with stable coronary artery disease found a cross-sectional association between depressive symptoms and poor EC. In another study of CR patients, depressive symptoms predicted slower heart rate recovery after exercise stress testing [16]. Others have found that anxiety and depression are correlated with EC. Egger et al. [17] found that changes in the self-reported levels of depression and anxiety on the HADS from pre- to post-CR were correlated with changes in EC. On the other hand, research suggests that the reverse effect could be true: lower EC may predict symptoms of anxiety and depression. Greater EC predicted lower levels of trait and state anxiety among adolescents [18] and frequent high-intensity exercise has been shown to reduce anxiety sensitivity among adults [19]. Similarly, risk of developing moderate-to-severe symptoms of depression increases significantly as cardiorespiratory fitness decreases [20, 21]. Furthermore, these effects appear to be long-lasting: cardiorespiratory fitness remained a strong predictor of depressive symptoms during a 12-year follow-up [22]. Despite preliminary evidence of the reciprocal association of psychological status and exercise fitness [23], few studies have examined this relationship in the context of cardiovascular disease. Evon and Burns [24] examined a similar research topic with 80 CR patients, although this study mainly focussed on the cross-lagged effects between perceived self-efficacy, patient-staff working alliance, and the change of dietary and exercise habits. The current study evaluated the link between anxiety, depression, and EC among patients completing an exercise-based CR program. In particular, our study aimed (1) to explore the role of multiple risk factors in predicting EC post-CR among cardiac patients and (2) to investigate the bidirectional pathways between psychological distress and EC, using a cross-lagged panel design. The existence of a bidirectional association will be inferred if psychological distress predicts EC over time and vice versa. Support for the following hypotheses would indicate reciprocal associations, whereby (1) decreased symptoms of depression and anxiety will be associated with improvement in EC and (2) lower EC will be associated with increases in depression and anxiety symptoms post-CR. Methods Study Sample The study population consisted of 212 Caucasian patients diagnosed with heart disease, who entered a comprehensive outpatient CR program. Exclusion criteria were: inability to fully participate in the study due to physical (e.g., deafness), cognitive (e.g., dementia), or severe psychiatric (e.g., schizophrenia) impairment. Participants were referred to CR for the following conditions: coronary artery disease (CAD), post-valve repair/replacement, heart failure, or a combination of these diagnoses. Demographic and clinical characteristics of the sample are summarized in Table 1. Participants completed all study measures at admission to the CR program (t1) and at discharge (t2). The study protocol was approved by the Institutional Review Board. Patients were asked to provide written consent to participate, and confidentiality was assured. Table 1 Demographic and clinical characteristics of the sample Variable Sample n = 212 Male sex 80.7% Age mean (SD) 63.11 (11.63) Low education (lower than secondary school) 32.5% Professional status  Retired 49.1%  Full time/part-time working 41%  Not working/Housewives 9.9% Diagnosis  Ischemic heart disease 71%  Heart failure 5%  Valvular disease 16%  Comorbidity 8% Smoking status  Actual smoker 5.2%  Quitted smoking <1 year ago 30.2%  Quitted smoking >1 year ago 34.9%  Never smoked 27.8% BMI-based classification  Underweight 1.4%  Normal weight 45%  Overweight 39.3%  Obese 14.2% Variable Sample n = 212 Male sex 80.7% Age mean (SD) 63.11 (11.63) Low education (lower than secondary school) 32.5% Professional status  Retired 49.1%  Full time/part-time working 41%  Not working/Housewives 9.9% Diagnosis  Ischemic heart disease 71%  Heart failure 5%  Valvular disease 16%  Comorbidity 8% Smoking status  Actual smoker 5.2%  Quitted smoking <1 year ago 30.2%  Quitted smoking >1 year ago 34.9%  Never smoked 27.8% BMI-based classification  Underweight 1.4%  Normal weight 45%  Overweight 39.3%  Obese 14.2% View Large Table 1 Demographic and clinical characteristics of the sample Variable Sample n = 212 Male sex 80.7% Age mean (SD) 63.11 (11.63) Low education (lower than secondary school) 32.5% Professional status  Retired 49.1%  Full time/part-time working 41%  Not working/Housewives 9.9% Diagnosis  Ischemic heart disease 71%  Heart failure 5%  Valvular disease 16%  Comorbidity 8% Smoking status  Actual smoker 5.2%  Quitted smoking <1 year ago 30.2%  Quitted smoking >1 year ago 34.9%  Never smoked 27.8% BMI-based classification  Underweight 1.4%  Normal weight 45%  Overweight 39.3%  Obese 14.2% Variable Sample n = 212 Male sex 80.7% Age mean (SD) 63.11 (11.63) Low education (lower than secondary school) 32.5% Professional status  Retired 49.1%  Full time/part-time working 41%  Not working/Housewives 9.9% Diagnosis  Ischemic heart disease 71%  Heart failure 5%  Valvular disease 16%  Comorbidity 8% Smoking status  Actual smoker 5.2%  Quitted smoking <1 year ago 30.2%  Quitted smoking >1 year ago 34.9%  Never smoked 27.8% BMI-based classification  Underweight 1.4%  Normal weight 45%  Overweight 39.3%  Obese 14.2% View Large CR Program CR was performed according to the protocols followed in our center, according to practice guidelines of the European Society of Cardiology [25]. The program combined physical training and educational sessions and lasted approximately 6 weeks, with a mean of 21.2 (SD = 4.1) exercise sessions performed by the patients. Physical training Each exercise session lasted 90 min, including stretching, calisthenics, and 45 min of aerobic exercise by bicycle or treadmill. All patients were under constant ECG monitoring and cardiologic supervision. The bicycle/treadmill intensity level in Watts was automatically set by a feedback-computerized system (Custo Cardio Concept, CUSTO Med, Ottobrunn, Germany) to obtain the training heart rate (HR). For each patient, training HR was calculated as follows: training HR = rest HR + 60% HR reserve, where HR reserve = peak HR – rest HR at a preliminary bicycle ergometer exercise test. In patients for whom training HR could not be set due to atrial fibrillation or the presence of a pacemaker, training intensity was calculated by means of the perceived effort with a 10-point Borg scale [26] and a score of 4 set as the target exercise intensity. Training intensity was gradually adjusted according to the individual patient’s changes in EC to maintain the same level of HR and to avoid exercising above the initial HR target. Education The CR program included four group-based educational sessions conducted by CR doctors and nurses and focussed on: (1) various cardiovascular diseases and medical interventions (e.g., percutaneous transluminal coronary angioplasty, coronary stenting, and coronary bypass), (2) pharmacological therapies, and (3) the role of modifiable risk factors in cardiovascular health. In addition, trained dieticians conducted two group sessions that provided education on nutrients (e.g., sugars, fats, and proteins), food labels, and the preparation of healthy meals. Outcome Measurements Social-demographic and clinical data Participants provided their age, sex, marital status, educational level, and employment status. Clinical data, including the diagnosis, and body mass index (BMI) were retrieved from the participants’ medical record. Smoking status was assessed by a self-report questionnaire, which distinguished between four categories: (1) current smoker, (2) former smoker who had quit smoking less than 1 year prior to the survey, (3) former smoker who had quit more than 1 year prior to the survey, and (4) non-smoker. Hospital Anxiety and Depression Scale The Hospital Anxiety and Depression Scale (HADS) is a widely used instrument to assess the presence of anxiety and depression symptoms in hospital settings [27] and has recently been utilized as a measure of psychological distress in cardiac patients [28–30]. The HADS consists of 14 items, 7 measuring the anxiety level and 7 measuring depression. Items are scored on a 4-point Likert-scale, from 0 to 3, with a total score of 42, and a maximum score of 21 for each subscale. Higher scores indicate higher psychological distress. Both total score and subscale scores were used in the data analyses. Score interpretations are equivalent for both the Anxiety and Depression subscale; scores between 8 and 10 represent mild symptoms, 11–15 reflects moderate symptoms, and 16 or above represent severe symptoms [31]. A very recent study on 617 Italian cardiovascular patients [32] suggested that an HADS total score of 14 or greater provided high sensitivity and good specificity to identify clinically relevant symptoms of psychological distress among cardiac patients. The authors concluded that the HADS could be considered a good first-step screening tool to identify patients with a relevant level of psychological distress. Internal consistency for this study was good, total score Cronbach alphas was 0.84 at baseline and 0.86 at discharge. Adequate alpha scores for the subscales was found for both the Depression (0.80 at baseline and 0.81 at discharge) and for the Anxiety subscale (0.71 at baseline and 0.75 at discharge). Exercise capacity EC was defined as the maximal power (in Watts) reached during an incremental bicycle stress test, with an increase of 10 Watts every minute. Changes in the workload during the rehabilitation period were considered an index of physical training. Data Analysis Regression analyses To identify other significant predictors of the EC post-CR, a number of univariate regression analyses were performed. Bonferroni correction was applied, which yielded a significance threshold of 0.005. Variables predicting EC with a p-value of <.005 in the univariate regression analyses were included as potential independent predictors for the multivariate linear regression. Data were analyzed using the IBM SPSS (version 22.0, SPSS Inc., Chicago, IL, USA). Cross-lagged panel design A cross-lagged design was used to examine the stability of EC and psychological distress across time as well as potential cross-lagged effects. Model fit was assessed based on Δχ2, the comparative fit index (CFI), and the root mean square error of approximation (RMSEA). Maximum likelihood estimation was used to estimate structural equation models with the Mplus 7 program [33], testing full cross-lagged models, including EC and a composite anxiety and depression score at t1 and t2. All models were estimated using Full Information Maximum Likelihood (FIML) in Mplus. Age was included as covariate in all tested models. Results Descriptive Statistics CR adherence, defined as the percentage of CR exercise sessions completed, was 96.56% (SD = 5.67). Mean EC at CR entry was 91.31 Watts (SD = 27.9), which increased to 108.68 Watts (SD = 34.82) at discharge, showing a significant functional improvement over the course of CR (t194 = −12.93, p < .001; Cohen’s d = 0.55). Average BMI did not change (t208 = −1.05, p = .294; Cohen’s d =0.03) from t1 (M = 25.68, SD = 4.29) to t2 (M = 25.79, SD = 4.17). Average HADS anxiety score was 6.33 (SD = 3.76) at t1 and 5.41 (SD = 3.72) at t2, which marked a significant improvement (t209 = 4.72, p < .001; Cohen’s d = 0.25). Average HADS depression score was 4.32 (SD = 3.39) at t1 and 4.04 (SD = 3.43) at t2 (t207 = 1.54, p = .13; Cohen’s d = 0.08). The mean HADS total score was 10.64 (SD = 6.58) at t1 and 9.42 (SD = 6.57) at t2 (t207 = 3.89, p < .001; Cohen’s d = 0.1). Overall, the HADS total and subscale scores observed in our sample were similar to that reported in a larger Italian sample of cardiac patients [32]. On the basis of the cut-off value of 14 on the HADS total score [32], in the present sample 31.7% of the participants reported a clinically significant level of psychological distress pre-CR. These patients demonstrated significantly lower levels of EC post-CR (M = 97.05; SD = 30.30) than patients with an HADS total score < 14 (M = 113.85; SD = 35.55, t193 = 3.18, p < .002; Cohen’s d = 0.51) Communality and Regression Analyses The results of univariate regression analyses are reported in Table 2. As expected, the most influential predictor of EC post-CR was EC pre-CR (β = .840, p < .001). Younger age (β = −.462, p < .001) and male sex (β = .417, p < .001) predicted better EC post-CR, and a diagnosis of heart failure predicted worse EC post-CR (β = −.237, p = .001). No associations were found for self-reported smoking history (β = −.027, p = .71) and BMI (β = .13, p = .076). Anxiety (β = −.227, p < .001) and depression (β = −.233, p <.001), and the psychological distress measured by the HADS total score (β = −.250, p <.001), negatively predicted EC at post-CR. Table 2 Prediction of exercise capacity at t2: univariate regression models B CI 95% SE b β p-value Adjusted R2 EC at t1  Costant 12.97 3.77, 22.16 5.09 .70  EC at t1 1.05 .952, 1.14 .054 .84 <.001 Age  Costant 197.38 172.83, 21.93 4.66 .21  Age −1.41 −1.80, 1.03 .15 −.46 <.001 Sex  Costant 78.28 67.84, 88.71 5.29 .17  Sex 37.29 25.73, 48.84 5.86 .42 <.001 BMI at t1  Costant 79.4 52.99, 111.62 13.53 .01  BMI at t1 1.09 −.106, 2.13 .52 .13 .08 Smoking status at t1  Costant 109.11 104.08, 114.14 2.55 .001  Smoking at t1 −.38 −2.37, 1.62 1.01 −.03 .71 HADS anxiety at t1  Costant 121.37 112.09, 130.65 4.70 .05  HADS anxiety −2.06 −3.31, −.80 .64 −.23 <.001 HADS depression at t1  Costant 118.62 110.81, 126.42 3.96 .05  HADS depression −2.38 −3.81,−.96 .72 −.23 <.001 HADS total score  Costant 122.12 112.30, 128.97 4.59 .062  HADS total score −.13 −1.9, .36 .37 −.25 <.001 Diagnosis  Costant 113.93 108.29, 119.58 2.86 .07  Valvulopathy versus CAD −10.60 −23.45, 2.25 6.51 −.11 .10  Hearth failure versus CAD −41.43 −65.55, −17.32 12.22 −.24 .001  Multiple diagnoses −21.43 −38.95, −3.92 8.88 −.17 .02 B CI 95% SE b β p-value Adjusted R2 EC at t1  Costant 12.97 3.77, 22.16 5.09 .70  EC at t1 1.05 .952, 1.14 .054 .84 <.001 Age  Costant 197.38 172.83, 21.93 4.66 .21  Age −1.41 −1.80, 1.03 .15 −.46 <.001 Sex  Costant 78.28 67.84, 88.71 5.29 .17  Sex 37.29 25.73, 48.84 5.86 .42 <.001 BMI at t1  Costant 79.4 52.99, 111.62 13.53 .01  BMI at t1 1.09 −.106, 2.13 .52 .13 .08 Smoking status at t1  Costant 109.11 104.08, 114.14 2.55 .001  Smoking at t1 −.38 −2.37, 1.62 1.01 −.03 .71 HADS anxiety at t1  Costant 121.37 112.09, 130.65 4.70 .05  HADS anxiety −2.06 −3.31, −.80 .64 −.23 <.001 HADS depression at t1  Costant 118.62 110.81, 126.42 3.96 .05  HADS depression −2.38 −3.81,−.96 .72 −.23 <.001 HADS total score  Costant 122.12 112.30, 128.97 4.59 .062  HADS total score −.13 −1.9, .36 .37 −.25 <.001 Diagnosis  Costant 113.93 108.29, 119.58 2.86 .07  Valvulopathy versus CAD −10.60 −23.45, 2.25 6.51 −.11 .10  Hearth failure versus CAD −41.43 −65.55, −17.32 12.22 −.24 .001  Multiple diagnoses −21.43 −38.95, −3.92 8.88 −.17 .02 View Large Table 2 Prediction of exercise capacity at t2: univariate regression models B CI 95% SE b β p-value Adjusted R2 EC at t1  Costant 12.97 3.77, 22.16 5.09 .70  EC at t1 1.05 .952, 1.14 .054 .84 <.001 Age  Costant 197.38 172.83, 21.93 4.66 .21  Age −1.41 −1.80, 1.03 .15 −.46 <.001 Sex  Costant 78.28 67.84, 88.71 5.29 .17  Sex 37.29 25.73, 48.84 5.86 .42 <.001 BMI at t1  Costant 79.4 52.99, 111.62 13.53 .01  BMI at t1 1.09 −.106, 2.13 .52 .13 .08 Smoking status at t1  Costant 109.11 104.08, 114.14 2.55 .001  Smoking at t1 −.38 −2.37, 1.62 1.01 −.03 .71 HADS anxiety at t1  Costant 121.37 112.09, 130.65 4.70 .05  HADS anxiety −2.06 −3.31, −.80 .64 −.23 <.001 HADS depression at t1  Costant 118.62 110.81, 126.42 3.96 .05  HADS depression −2.38 −3.81,−.96 .72 −.23 <.001 HADS total score  Costant 122.12 112.30, 128.97 4.59 .062  HADS total score −.13 −1.9, .36 .37 −.25 <.001 Diagnosis  Costant 113.93 108.29, 119.58 2.86 .07  Valvulopathy versus CAD −10.60 −23.45, 2.25 6.51 −.11 .10  Hearth failure versus CAD −41.43 −65.55, −17.32 12.22 −.24 .001  Multiple diagnoses −21.43 −38.95, −3.92 8.88 −.17 .02 B CI 95% SE b β p-value Adjusted R2 EC at t1  Costant 12.97 3.77, 22.16 5.09 .70  EC at t1 1.05 .952, 1.14 .054 .84 <.001 Age  Costant 197.38 172.83, 21.93 4.66 .21  Age −1.41 −1.80, 1.03 .15 −.46 <.001 Sex  Costant 78.28 67.84, 88.71 5.29 .17  Sex 37.29 25.73, 48.84 5.86 .42 <.001 BMI at t1  Costant 79.4 52.99, 111.62 13.53 .01  BMI at t1 1.09 −.106, 2.13 .52 .13 .08 Smoking status at t1  Costant 109.11 104.08, 114.14 2.55 .001  Smoking at t1 −.38 −2.37, 1.62 1.01 −.03 .71 HADS anxiety at t1  Costant 121.37 112.09, 130.65 4.70 .05  HADS anxiety −2.06 −3.31, −.80 .64 −.23 <.001 HADS depression at t1  Costant 118.62 110.81, 126.42 3.96 .05  HADS depression −2.38 −3.81,−.96 .72 −.23 <.001 HADS total score  Costant 122.12 112.30, 128.97 4.59 .062  HADS total score −.13 −1.9, .36 .37 −.25 <.001 Diagnosis  Costant 113.93 108.29, 119.58 2.86 .07  Valvulopathy versus CAD −10.60 −23.45, 2.25 6.51 −.11 .10  Hearth failure versus CAD −41.43 −65.55, −17.32 12.22 −.24 .001  Multiple diagnoses −21.43 −38.95, −3.92 8.88 −.17 .02 View Large Anxiety and depression were highly correlated in the present sample (r = .692, p < .001). This result suggests a considerable overlap between anxiety and depression subscales when simultaneously considered as predictors of EC in a multivariate regression analysis [34]. To distinguish between the unique effect of anxiety, the unique effect of depression, and the overlapping effect of both on EC, a commonality analysis was performed [35]. Results are shown in Fig. 1. In particular, unique effects were computed as a squared semi-partial correlation (sr2) between depression, anxiety, and EC; communality was calculated by subtracting the two squared semi-partial correlation from the total R2 [36]. Anxiety and depression together significantly predicted EC (F2, 192 = 6.270, p = .002), explaining 6.2% of its total variance (R2 = .062). The portion of EC variance explained by the unique effect of depression and anxiety was 0.8% (sr2 = .008) and 1.1% (sr2 = .011), respectively, while 4.3% (sr2 = .043) was explained by the common effect of the two predictors. In conclusion, these results showed the existence of a substantial overlap between anxiety and depression in explaining EC, while the unique variances of the single components were negligible. Fig. 1 View largeDownload slide The proportion of variance shared by Anxiety, Depression and EC. The overlapping areas in the figure reflect the relative amount of variance shared by the three dimensions. a = proportion of EC variance explained by the unique effect of anxiety (1.1%); b = proportion of EC variance explained by the unique effect of depression (0.8%); c = proportion of EC variance explained by the common effect of anxiety and depression (4.3%) Fig. 1 View largeDownload slide The proportion of variance shared by Anxiety, Depression and EC. The overlapping areas in the figure reflect the relative amount of variance shared by the three dimensions. a = proportion of EC variance explained by the unique effect of anxiety (1.1%); b = proportion of EC variance explained by the unique effect of depression (0.8%); c = proportion of EC variance explained by the common effect of anxiety and depression (4.3%) Consequently, in the subsequent multivariate regression analyses, the total HADS score was utilized as a global indicator of psychological distress, in line with that suggested by Herrmann et al. [37]. The possible impact of the attendance of CR sessions on the observed association between functional status and emotional distress was assessed by computing the partial correlation between EC and the total HADS score, controlling for the absolute number of sessions attended by each patient. The resulting partial correlation (r = −.252; p < .001) did not substantially differ from the zero-order correlation (r = −.248; p < .001), showing that the number of sessions did not significantly influence the relationship between the other two variables. Another potential confounder in the relationship between EC and psychological distress is adherence to the CR program, calculated as the percentage of attended sessions over prescribed sessions. However, even after controlling for this variable, the difference from the zero-order correlation is minimal (r = −.227; p = .002). The results of the multivariate stepwise regression analysis are reported in Table 3. Data were verified for assumptions required for regression. Unstandardized residual scores were normally distributed and uncorrelated, with the value of Durbin–Watson diagnostic test of 2.01. The model of best-fit had four predictors: baseline EC, age, HADS total score, and sex. The model was statistically significant (F4, 192 = 134.76, p < .001) and accounted for 74.6% of the variability in the dependent variable. Table 3 Prediction of exercise capacity at t2: multivariate regression models B CI 95% SE B β p-value Adjusted R2 Step 1  Costant 12.96 3.65, 22.27 4.72 .70  EC at t1 1.05 .95, 1.14 .05 .84 <.001 Step 2  Costant 50.08 28.79, 71.38 11.54 .72  EC at t1 .97 .87, 1.07 .05 .77 <.001  Age −.48 −.73, −.23 .13 −.16 .001 Step 3  Costant 65.03 41.75, 88.31 12.30 .73  EC at t1 .93 .82, 1.03 .05 .74 <.001  Age −.56 −.81, −.31 .13 −.18 <.001  HADS total score −.59 −.99, −.18 .22 −.11 .001 Step 4  Costant 64.81 41.85, 87.77 12.15 .74  EC at t1 .87 .76, .98 .06 .69 <.001  Age −.60 −.85, −.35 .13 −.20 <.001  HADS total score −.52 −.93, −.12 .27 −.10 .01  Sex 9.30 2.0, 16.6 4.06 .10 .01 B CI 95% SE B β p-value Adjusted R2 Step 1  Costant 12.96 3.65, 22.27 4.72 .70  EC at t1 1.05 .95, 1.14 .05 .84 <.001 Step 2  Costant 50.08 28.79, 71.38 11.54 .72  EC at t1 .97 .87, 1.07 .05 .77 <.001  Age −.48 −.73, −.23 .13 −.16 .001 Step 3  Costant 65.03 41.75, 88.31 12.30 .73  EC at t1 .93 .82, 1.03 .05 .74 <.001  Age −.56 −.81, −.31 .13 −.18 <.001  HADS total score −.59 −.99, −.18 .22 −.11 .001 Step 4  Costant 64.81 41.85, 87.77 12.15 .74  EC at t1 .87 .76, .98 .06 .69 <.001  Age −.60 −.85, −.35 .13 −.20 <.001  HADS total score −.52 −.93, −.12 .27 −.10 .01  Sex 9.30 2.0, 16.6 4.06 .10 .01 View Large Table 3 Prediction of exercise capacity at t2: multivariate regression models B CI 95% SE B β p-value Adjusted R2 Step 1  Costant 12.96 3.65, 22.27 4.72 .70  EC at t1 1.05 .95, 1.14 .05 .84 <.001 Step 2  Costant 50.08 28.79, 71.38 11.54 .72  EC at t1 .97 .87, 1.07 .05 .77 <.001  Age −.48 −.73, −.23 .13 −.16 .001 Step 3  Costant 65.03 41.75, 88.31 12.30 .73  EC at t1 .93 .82, 1.03 .05 .74 <.001  Age −.56 −.81, −.31 .13 −.18 <.001  HADS total score −.59 −.99, −.18 .22 −.11 .001 Step 4  Costant 64.81 41.85, 87.77 12.15 .74  EC at t1 .87 .76, .98 .06 .69 <.001  Age −.60 −.85, −.35 .13 −.20 <.001  HADS total score −.52 −.93, −.12 .27 −.10 .01  Sex 9.30 2.0, 16.6 4.06 .10 .01 B CI 95% SE B β p-value Adjusted R2 Step 1  Costant 12.96 3.65, 22.27 4.72 .70  EC at t1 1.05 .95, 1.14 .05 .84 <.001 Step 2  Costant 50.08 28.79, 71.38 11.54 .72  EC at t1 .97 .87, 1.07 .05 .77 <.001  Age −.48 −.73, −.23 .13 −.16 .001 Step 3  Costant 65.03 41.75, 88.31 12.30 .73  EC at t1 .93 .82, 1.03 .05 .74 <.001  Age −.56 −.81, −.31 .13 −.18 <.001  HADS total score −.59 −.99, −.18 .22 −.11 .001 Step 4  Costant 64.81 41.85, 87.77 12.15 .74  EC at t1 .87 .76, .98 .06 .69 <.001  Age −.60 −.85, −.35 .13 −.20 <.001  HADS total score −.52 −.93, −.12 .27 −.10 .01  Sex 9.30 2.0, 16.6 4.06 .10 .01 View Large Cross-Lagged Panel Design In a second step, the structural relations between EC and HADS were specified as cross-lagged effects, which allowed the evaluation of time-lagged reciprocal effects between the two variables, while controlling for their stability over time, that is, autoregressive effects [38]. As such, the full model included the stability coefficients between EC at t1–t2 and HADS at t1–t2, the path from EC at t1 to HADS at t2, the path from HADS t1 to EC at t2 among all participants. Associations between HADS and EC at each of the two measurement times were also included in the model. Next, a series of nested models were tested and compared with the saturated model, which tested a hypothesis of mutual causality, and the change in model fit indices was evaluated to identify the best fitting model. Results of these model comparisons are shown in Table 4. When the path from HADS at t1 to EC at t2 was removed, the model fit worsened significantly (Δχ2 = 4.257, Δdf = 1, p = .039). Removal of the path from EC at t1 to HADS at t2 did not show a significant worsening of the chi-square index (Δχ2 = 1.345, Δdf = 1, p = .246). This model, presented in Fig. 1, fitted the data well (χ2 = 1.345; df = 1; p = .246, CFI = .99, RMSEA = .040). Results confirmed that EC pre-CR is a strong predictor of the EC post-CR (β = .820, p < .001). Furthermore, high levels of psychological distress negatively predicted EC post-CR (β = −.082, p < .036), over and above the effect of baseline EC. Therefore, psychological distress at baseline accounted for a significant, albeit small, amount of variance in EC post-CR. Table 4 Cross-lagged models’ comparisons Model df χ2(p) RMSEA CFI AIC Delta model Δχ2(p) Saturated Mutual causality 0 0 0 1 7905.03 − − Model 1 EC -> HADS 1 4.26 (.04) .12 .99 7907.29 1 vs 0 4.26 (.04) Model 2 HADS-> EC 1 1.34 (.25) .04 1 7904.38 2 vs 0 1.34 (.25) Model 3 Auto-regressions only 2 5.75 (.06) .09 .99 7906.78 3 vs 2 4.40 (.04) Model df χ2(p) RMSEA CFI AIC Delta model Δχ2(p) Saturated Mutual causality 0 0 0 1 7905.03 − − Model 1 EC -> HADS 1 4.26 (.04) .12 .99 7907.29 1 vs 0 4.26 (.04) Model 2 HADS-> EC 1 1.34 (.25) .04 1 7904.38 2 vs 0 1.34 (.25) Model 3 Auto-regressions only 2 5.75 (.06) .09 .99 7906.78 3 vs 2 4.40 (.04) The best fitting model is shown in bold type. View Large Table 4 Cross-lagged models’ comparisons Model df χ2(p) RMSEA CFI AIC Delta model Δχ2(p) Saturated Mutual causality 0 0 0 1 7905.03 − − Model 1 EC -> HADS 1 4.26 (.04) .12 .99 7907.29 1 vs 0 4.26 (.04) Model 2 HADS-> EC 1 1.34 (.25) .04 1 7904.38 2 vs 0 1.34 (.25) Model 3 Auto-regressions only 2 5.75 (.06) .09 .99 7906.78 3 vs 2 4.40 (.04) Model df χ2(p) RMSEA CFI AIC Delta model Δχ2(p) Saturated Mutual causality 0 0 0 1 7905.03 − − Model 1 EC -> HADS 1 4.26 (.04) .12 .99 7907.29 1 vs 0 4.26 (.04) Model 2 HADS-> EC 1 1.34 (.25) .04 1 7904.38 2 vs 0 1.34 (.25) Model 3 Auto-regressions only 2 5.75 (.06) .09 .99 7906.78 3 vs 2 4.40 (.04) The best fitting model is shown in bold type. View Large Furthermore, the t2 correlations between the residual variables is low, although statistically significant (r = −.15, p = .04), indicating a small amount of shared variance in HADS and EC scores, beyond that explained by autoregressive and cross-lagged effects. Overall, as shown by the R2 statistics, the model explained 58.4% of individual differences in HADS score at t2 and 70.9% in EC at t2. Age was included in the model as covariate and it was significantly associated with EC at t1 (r = −.393, p < .001) and at t2 (r = −.284, p < .001). Age was not significantly associated with HADS score, neither at t1 (r = −.067, p = .325), nor at t2 (r = −.004, p = .947). Finally, we tested the stability model, encompassing only autoregressive paths. In this case, the model fit worsened significantly (Δχ2 = 4.405, Δdf = 1, p = .0358), confirming that the previous model was the best fit for our data. In conclusion, results indicate that levels of pre-CR psychological distress predicted the EC post-CR, beyond the effect of baseline EC. Moreover, the results indicate that baseline EC does not significantly predict post-CR levels of anxiety and depression. Discussion Although the association between EC and psychological distress has been previously reported in cardiac patients [14, 39], to our knowledge this is the first study to examine the cross-lagged effects of EC and psychological distress among CR-patients. Consistent with previous studies [17, 39], our results showed a cross-sectional association between psychological distress and EC, even controlling for the patients’ adherence to the CR program. Moreover, this association has been showed in the present paper to remain significant at the end of the rehabilitation period. The multivariate regression analysis revealed that a significant, although small, amount of variance in post-CR EC was accounted for by pre-CR levels of psychological distress, beyond the effects of EC pre-CR, age, and sex. Furthermore, exploratory analyses showed a group difference in EC post-CR between those with and without significant levels of psychological distress pre-CR, with a medium effect size. Cross-lagged effects revealed a 1-direction association between greater psychological distress at pre-CR and lower EC post-CR, over and above that of autoregressive effects. Overall, these results lend support to the literature highlighting that pre-CR psychological distress, such as increased symptoms of depression and anxiety, may be prospectively associated with a reduced improvement in functional capacity post-CR. The mechanisms explaining this result may be related to depressive and anxiety symptomatology. In fact, cardiac patients with depressive symptomatology tend to have poorer fitness levels and exercise tolerance, due to lower levels of energy and decreased motivation to engage regularly in physical activity [40]. It has been suggested [41] that EC results obtained from patients with high levels of depression may be not representative of their true fitness level, even when the full diagnostic criteria for major depression have not been met. In these patients exercise testing may not be as reliable in detecting ischemia, potentially leading to false-negative diagnoses [41]. Fig. 2 View largeDownload slide Cross-lagged effects. Covariance between age and the other variables was included in the model but omitted from the figure. Only significant paths are shown in the figure; p-values are shown in parentheses. Fig. 2 View largeDownload slide Cross-lagged effects. Covariance between age and the other variables was included in the model but omitted from the figure. Only significant paths are shown in the figure; p-values are shown in parentheses. As for the role of anxiety, previous research noted that anxious individuals are prone to interpret cardiorespiratory sensations associated with physical activity as threatening or aversive [18, 42, 43], leading to decreased engagement/tolerance during exercise stress testing, and subsequently depict the patient as less fit than their non-anxious counterparts [39]. The data did not support the proposed bidirectional relationship between EC and psychological distress. Unexpectedly, EC at pre-CR was not associated with changes in psychological distress. Cardiorespiratory fitness and subsequent presence of depressive symptomatology have been reported in a number of studies [18, 22, 42]. However, the referenced studies included healthy subjects, and the presence of cardiovascular disease was a common exclusion criterion [20]. Previous researchers have found that mood improves over the course of a 12-week, 36-session, CR program [23, 44]. Since the current CR program lasted approximately 6 weeks, and the change in psychological distress from pre- to post-CR was of small effect, it is possible that the lack of association between EC pre-CR and change in psychological distress was due to the brevity of this CR program. Although most participants in this study were below the HADS clinical threshold, with an average total score of 10.68 (SD = 6.59), 31.7% of the study sample at pre-CR and 25% at post-CR, reported a clinically significant level of psychological distress (HADS score ≥ 14 [32]). This suggests that, even if the distribution of psychological distress is considerably skewed to low levels of symptomatology, there is a notable proportion of patients whose clinically significant score may explain the effects detected in the cross-lagged analysis. Previous research has found that patients who report depression or anxiety symptoms independently [45, 46], or together [47], following CR are at increased risk of mortality, and more likely to show little-to-no improvement in EC post-CR. Taken together, this highlights the importance of identifying patients who could benefit from adjunctive mental health services during CR. Doing so may improve risk stratification and reduce levels of psychological distress and maximize cardiovascular improvements during CR. Considerable research has noted the overlap between depression and anxiety symptoms in cardiac patients [48, 49] and many approaches have been proposed to account for this overlap in the assessment of depression and anxiety in clinical settings [50–54]. In the current study, the two dimensions were not considered independently, and a composite score was used as a measure of general psychological distress. Although this decision was data driven, it prevented the ability to distinguish between the separate effects of anxiety and depression on EC and to test moderation or mediation hypotheses. This study has several limitations. A longitudinal approach of relatively short length was adopted. Future research should evaluate whether these results hold at 6 months post-CR. Although adherence to the scheduled CR program was statistically controlled for in the analyses, telemetry data were not collected during this study and the effect of adherence to target heart rate recommendations on both psychological distress and EC cannot be determined. Furthermore, although the protocol employed in the study is highly standardized, treatment fidelity was not assessed formally, and this limitation may affect on the interpretation of the results. Moreover, although the cross-lagged panel design is useful to explore the bidirectional pathways linking psychological distress and EC, it does not allow for the causal relationship to be inferred [54]. As for the study sample, participants were exclusively Caucasian and a marked imbalance in favor of men was observed, as the percentage of women in the sample is 20% lower than expected based on cardiovascular morbidity data. Although this ratio is similar to that observed in other CR samples in the Italian population [55], the results may not be generalizable to women or men of ethnic minority groups. Moreover, cardiac patients with depression may be less likely to participate in an exercise-based CR program [56, 57], which could explain why the psychological distress data in this study were skewed toward lower scores on the HADS. In conclusion, results from this study suggest that CR patients with increased levels of psychological distress are more likely to demonstrate less EC post-CR. Therefore, we support a renewed attention to psychological risk factors, which should be routinely screened at the beginning of the rehabilitation period, and their evolution monitored. Early identification of distressed patients will provide the opportunity to connect the patient with outpatient mental health providers, and improve patient-centered care by identifying and addressing potential obstacles to maximal cardiorespiratory improvement, which could include extended CR sessions. The findings of the present study support a model integrating both medical and psychological factors, thus promoting a more personalized and holistic approach to the patient. Acknowledgements We are grateful to Matteo Baruffi, Francesco Borgia and Alessandra Meraviglia for their contributions to the data collection and to Emanuele M. Giusti for his contribution to the final revision of the manuscript. Compliance with Ethical Standards Authors’ Statement of Conflict of Interest and Adherence to Ethical Standards Authors Chiara A.M. Spatola, Emanuele A.M. Cappella, Christina L. Goodwin, Gianluca Castelnuovo, Roberto Cattivelli, Giada Rapelli, Gabriella Malfatto, Mario Facchini, Chiara Mollica, and Enrico Molinari declare that they have no conflict of interest. Ethical Approval All procedures were conducted in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000. Informed consent Informed consent was obtained from all individual participants included in the study. References 1. 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 ( 16 ): 1715 – 1733 . Google Scholar CrossRef Search ADS PubMed 2. 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 ( 4 ): 571 – 584.e2 . Google Scholar CrossRef Search ADS PubMed 3. 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 ): 1 – 12 . Google Scholar CrossRef Search ADS PubMed 4. Taylor SE , Way BM , Seeman TE . Early adversity and adult health outcomes . Dev Psychopathol . 2011 ; 23 ( 3 ): 939 – 954 . Google Scholar CrossRef Search ADS PubMed 5. Ades PA . Cardiac rehabilitation and secondary prevention of coronary heart disease . N Engl J Med . 2001 ; 345 ( 12 ): 892 – 902 . Google Scholar CrossRef Search ADS PubMed 6. Kokkinos P , Myers J . Exercise and physical activity: Clinical outcomes and applications . Circulation . 2010 ; 122 ( 16 ): 1637 – 1648 . Google Scholar CrossRef Search ADS PubMed 7. Suaya JA , Stason WB , Ades PA , Normand SL , Shepard DS . Cardiac rehabilitation and survival in older coronary patients . J Am Coll Cardiol . 2009 ; 54 ( 1 ): 25 – 33 . Google Scholar CrossRef Search ADS PubMed 8. Glazer KM , Emery CF , Frid DJ , Banyasz RE . Psychological predictors of adherence and outcomes among patients in cardiac rehabilitation . J Cardiopulm Rehabil . 2002 ; 22 ( 1 ): 40 – 46 . Google Scholar CrossRef Search ADS PubMed 9. Manzoni GM , Villa V , Compare A et al. Short-term effects of a multi-disciplinary cardiac rehabilitation programme on psychological well-being, exercise capacity and weight in a sample of obese in-patients with coronary heart disease: a practice-level study . Psychol Health Med . 2011 ; 16 ( 2 ): 178 – 189 . Google Scholar CrossRef Search ADS PubMed 10. Rothenbacher D , Hahmann H , Wüsten B , Koenig W , Brenner H . Symptoms of anxiety and depression in patients with stable coronary heart disease: prognostic value and consideration of pathogenetic links . Eur J Cardiovasc Prev Rehabil . 2007 ; 14 ( 4 ): 547 – 554 . Google Scholar CrossRef Search ADS PubMed 11. Watkins LL , Koch GG , Sherwood A et al. Association of anxiety and depression with all-cause mortality in individuals with coronary heart disease . J Am Heart Assoc . 2013 ; 2 ( 2 ): e000068 . Google Scholar CrossRef Search ADS PubMed 12. Elderon L , Whooley MA . Depression and cardiovascular disease . Prog Cardiovasc Dis . 2013 ; 55 ( 6 ): 511 – 523 . Google Scholar CrossRef Search ADS PubMed 13. Roest AM , Martens EJ , Denollet J , de Jonge P . Prognostic association of anxiety post myocardial infarction with mortality and new cardiac events: A meta-analysis . Psychosom Med . 2010 ; 72 ( 6 ): 563 – 569 . Google Scholar CrossRef Search ADS PubMed 14. Papasavvas T , Alhashemi M , Micklewright D . Association between depressive symptoms and exercise capacity in patients with heart disease: A meta-analysis . J Cardiopulm Rehabil Prev . 2017 ; 37 ( 4 ): 239 – 249 . Google Scholar CrossRef Search ADS PubMed 15. Ruo B , Rumsfeld JS , Pipkin S , Whooley MA . Relation between depressive symptoms and treadmill exercise capacity in the Heart and Soul Study . Am J Cardiol . 2004 ; 94 ( 1 ): 96 – 99 . Google Scholar CrossRef Search ADS PubMed 16. Hughes JW , Casey E , Luyster F et al. Depression symptoms predict heart rate recovery after treadmill stress testing . Am Heart J . 2006 ; 151 ( 5 ): 1122.e1 – 1122.e6 . Google Scholar CrossRef Search ADS 17. Egger E , Schmid JP , Schmid RW , Saner H , von Känel R . Depression and anxiety symptoms affect change in exercise capacity during cardiac rehabilitation . Eur J Cardiovasc Prev Rehabil . 2008 ; 15 ( 6 ): 704 – 708 . Google Scholar CrossRef Search ADS PubMed 18. Williams SE , Carroll D , Veldhuijzen van Zanten JJ , Ginty AT . Anxiety symptom interpretation: A potential mechanism explaining the cardiorespiratory fitness-anxiety relationship . J Affect Disord . 2016 ; 193 : 151 – 156 . Google Scholar CrossRef Search ADS PubMed 19. Broman-Fulks JJ , Berman ME , Rabian BA , Webster MJ . Effects of aerobic exercise on anxiety sensitivity . Behav Res Ther . 2004 ; 42 ( 2 ): 125 – 136 . Google Scholar CrossRef Search ADS PubMed 20. Tolmunen T , Laukkanen JA , Hintikka J et al. Low maximal oxygen uptake is associated with elevated depressive symptoms in middle-aged men . Eur J Epidemiol . 2006 ; 21 ( 9 ): 701 – 706 . Google Scholar CrossRef Search ADS PubMed 21. Galper DI , Trivedi MH , Barlow CE , Dunn AL , Kampert JB . Inverse association between physical inactivity and mental health in men and women . Med Sci Sports Exerc . 2006 ; 38 ( 1 ): 173 – 178 . Google Scholar CrossRef Search ADS PubMed 22. Sui X , Laditka JN , Church TS et al. Prospective study of cardiorespiratory fitness and depressive symptoms in women and men . J Psychiatr Res . 2009 ; 43 ( 5 ): 546 – 552 . Google Scholar CrossRef Search ADS PubMed 23. Milani RV , Lavie CJ . Reducing psychosocial stress: a novel mechanism of improving survival from exercise training . Am J Med . 2009 ; 122 ( 10 ): 931 – 938 . Google Scholar CrossRef Search ADS PubMed 24. Evon DM , Burns JW . Process and outcome in cardiac rehabilitation: an examination of cross-lagged effects . J Consult Clin Psychol . 2004 ; 72 ( 4 ): 605 – 616 . Google Scholar CrossRef Search ADS PubMed 25. Piepoli MF , Corrà U , Adamopoulos S et al. Secondary prevention in the clinical management of patients with cardiovascular diseases. Core components, standards and outcome measures for referral and delivery: a policy statement from the cardiac rehabilitation section of the European Association for Cardiovascular Prevention & Rehabilitation. Endorsed by the Committee for Practice Guidelines of the European Society of Cardiology . Eur J Prev Cardiol . 2014 ; 21 ( 6 ): 664 – 681 . Google Scholar CrossRef Search ADS PubMed 26. Borg G . A category scale with ratio properties for intermodal and interindividual comparisons . In: Geissler H-G and Petzold P , eds. Psychophysical Judgment and the Process of Perception . Berlin : VEB Deutscher Verlag der Wissenschaften 1982 : 25 – 34 . 27. Zigmond AS , Snaith RP . The hospital anxiety and depression scale . Acta Psychiatr Scand . 1983 ; 67 ( 6 ): 361 – 370 . Google Scholar CrossRef Search ADS PubMed 28. Norton S , Cosco T , Doyle F , Done J , Sacker A . The Hospital Anxiety and Depression Scale: A meta confirmatory factor analysis . J Psychosom Res . 2013 ; 74 ( 1 ): 74 – 81 . Google Scholar CrossRef Search ADS PubMed 29. Palacios JE , Khondoker M , Achilla E , Tylee A , Hotopf M . A single, one-off measure of depression and anxiety predicts future symptoms, higher healthcare costs, and lower quality of life in coronary heart disease patients: Analysis from a multi-wave, primary care cohort study . PLoS One . 2016 ; 11 ( 7 ): e0158163 . Google Scholar CrossRef Search ADS PubMed 30. Burns A , Höfer S , Curry P , Sexton E , Doyle F . Revisiting the dimensionality of the Hospital Anxiety and Depression Scale in an international sample of patients with ischaemic heart disease . J Psychosom Res . 2014 ; 77 ( 2 ): 116 – 121 . Google Scholar CrossRef Search ADS PubMed 31. Snaith R , Zigmond A. The Hospital Anxiety and Depression Scale with the Irritability-depression-anxiety Scale and the Leeds Situational Anxiety Scale: Manual . Windsor : Nfer-Nelson ; 1994 . 32. Tesio V , Marra S , Molinaro S , Torta R , Gaita F , Castelli L . Screening of depression in cardiology: A study on 617 cardiovascular patients . Int J Cardiol . 2017 ; 245 : 49 – 51 . Google Scholar CrossRef Search ADS PubMed 33. Muthen L. Mplus User’s Guide , 6th ed .. Los Angeles, CA : Muthén & Muthén ; 1998 . 34. Capraro RM , Capraro MM . Commonality analysis: Understanding variance contributions to overall canonical correlation effects of attitude toward mathematics on geometry achievement . Mult Lin Regr Viewp . 2001 ; 27 ( 2 ): 16 – 23 . 35. Nimon K . Regression commonality analysis: Demonstration of an SPSS solution . Mult Lin Regr Viewp . 2010 ; 36 ( 1 ): 10 – 17 . 36. Reichwein Zientek L , Thompson B . Commonality analysis: Partitioning variance to facilitate better understanding of data . J Early Interv . 2006 ; 28 ( 4 ): 299 – 307 . Google Scholar CrossRef Search ADS 37. Herrmann C . International experiences with the Hospital Anxiety and Depression Scale—a review of validation data and clinical results . J Psychosom Res . 1997 ; 42 ( 1 ): 17 – 41 . Google Scholar CrossRef Search ADS PubMed 38. Cole DA , Maxwell SE . Testing mediational models with longitudinal data: Questions and tips in the use of structural equation modeling . J Abnorm Psychol . 2003 ; 112 ( 4 ): 558 – 577 . Google Scholar CrossRef Search ADS PubMed 39. Pelletier R , Bacon SL , Laurin C , Arsenault A , Fleet RP , Lavoie KL . The impact of anxiety disorders on assessment of myocardial ischemia and exercise stress test performance . J Cardiopulm Rehabil Prev . 2011 ; 31 ( 1 ): 60 – 66 . Google Scholar CrossRef Search ADS PubMed 40. Buys R , Budts W , Delecluse C , Vanhees L . Exercise capacity, physical activity, and obesity in adults with repaired aortic coarctation . J Cardiovasc Nurs . 2013 ; 28 ( 1 ): 66 – 73 . Google Scholar CrossRef Search ADS PubMed 41. Lavoie KL , Fleet RP , Lespérance F et al. Are exercise stress tests appropriate for assessing myocardial ischemia in patients with major depressive disorder ? Am Heart J . 2004 ; 148 ( 4 ): 621 – 627 . Google Scholar CrossRef Search ADS PubMed 42. Goodwin RD . Association between physical activity and mental disorders among adults in the United States . Prev Med . 2003 ; 36 ( 6 ): 698 – 703 . Google Scholar CrossRef Search ADS PubMed 43. Sui X , Laditka JN , Church TS et al. Prospective study of cardiorespiratory fitness and depressive symptoms in women and men . J Psychiatr Res . 2009 ; 43 ( 5 ): 546 – 552 . Google Scholar CrossRef Search ADS PubMed 44. Lavie CJ , Milani RV . Cardiac rehabilitation and exercise training in secondary coronary heart disease prevention . Prog Cardiovasc Dis . 2011 ; 53 ( 6 ): 397 – 403 . Google Scholar CrossRef Search ADS PubMed 45. Milani RV , Lavie CJ . Impact of cardiac rehabilitation on depression and its associated mortality . Am J Med . 2007 ; 120 ( 9 ): 799 – 806 . Google Scholar CrossRef Search ADS PubMed 46. Herrmann C , Brand-Driehorst S , Buss U , Rüger U . Effects of anxiety and depression on 5-year mortality in 5,057 patients referred for exercise testing . J Psychosom Res . 2000 ; 48 ( 4-5 ): 455 – 462 . Google Scholar CrossRef Search ADS PubMed 47. Kachur S , Menezes AR , De Schutter A , Milani RV , Lavie CJ . Significance of comorbid psychological stress and depression on outcomes after cardiac rehabilitation . Am J Med . 2016 ; 129 ( 12 ): 1316 – 1321 . Google Scholar CrossRef Search ADS PubMed 48. Denollet J , Strik JJ , Lousberg R , Honig A . Recognizing increased risk of depressive comorbidity after myocardial infarction: looking for 4 symptoms of anxiety-depression . Psychother Psychosom . 2006 ; 75 ( 6 ): 346 – 352 . Google Scholar CrossRef Search ADS PubMed 49. Frasure-Smith N , Lespérance F . Depression and anxiety as predictors of 2-year cardiac events in patients with stable coronary artery disease . Arch Gen Psychiatry . 2008 ; 65 ( 1 ): 62 – 71 . Google Scholar CrossRef Search ADS PubMed 50. Rutledge T , Linke SE , Krantz DS et al. Comorbid depression and anxiety symptoms as predictors of cardiovascular events: Results from the NHLBI-sponsored Women’s Ischemia Syndrome Evaluation (WISE) study . Psychosom Med . 2009 ; 71 ( 9 ): 958 – 964 . Google Scholar CrossRef Search ADS PubMed 51. Suls J , Bunde J . Anger, anxiety, and depression as risk factors for cardiovascular disease: The problems and implications of overlapping affective dispositions . Psychol Bull . 2005 ; 131 ( 2 ): 260 – 300 . Google Scholar CrossRef Search ADS PubMed 52. Bleil ME , Gianaros PJ , Jennings JR , Flory JD , Manuck SB . Trait negative affect: Toward an integrated model of understanding psychological risk for impairment in cardiac autonomic function . Psychosom Med . 2008 ; 70 ( 3 ): 328 – 337 . Google Scholar CrossRef Search ADS PubMed 53. Frasure-Smith N , Lespérance F . Depression and other psychological risks following myocardial infarction . Arch Gen Psychiatry . 2003 ; 60 ( 6 ): 627 – 636 . Google Scholar CrossRef Search ADS PubMed 54. Kenny DA. Cross‐Lagged Panel Design . Wiley StatsRef : Statistics Reference Online; 2014. http://onlinelibrary.wiley.com/doi/10.1002/9781118445112.stat06464/compoundindex. 55. Giannuzzi P , Temporelli PL , Marchioli R et al. ; GOSPEL Investigators . Global secondary prevention strategies to limit event recurrence after myocardial infarction: results of the GOSPEL study, a multicenter, randomized controlled trial from the Italian Cardiac Rehabilitation Network . Arch Intern Med . 2008 ; 168 ( 20 ): 2194 – 2204 . Google Scholar CrossRef Search ADS PubMed 56. Cooper AF , Jackson G , Weinman J , Horne R . Factors associated with cardiac rehabilitation attendance: A systematic review of the literature . Clin Rehabil . 2002 ; 16 ( 5 ): 541 – 552 . Google Scholar CrossRef Search ADS PubMed 57. Lane D , Carroll D , Ring C , Beevers DG , Lip GY . Predictors of attendance at cardiac rehabilitation after myocardial infarction . J Psychosom Res . 2001 ; 51 ( 3 ): 497 – 501 . Google Scholar CrossRef Search ADS PubMed © Society of Behavioral Medicine 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Behavioral Medicine Oxford University Press

Cross-Lagged Relations Between Exercise Capacity and Psychological Distress During Cardiac Rehabilitation

Annals of Behavioral Medicine , Volume Advance Article – Feb 22, 2018

Loading next page...
 
/lp/ou_press/cross-lagged-relations-between-exercise-capacity-and-psychological-jUF3jZGuhX
Copyright
© Society of Behavioral Medicine 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0883-6612
eISSN
1532-4796
D.O.I.
10.1093/abm/kax069
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Poorer mental health is associated with lower exercise capacity, above and beyond the effect of other cardiovascular risk factors. However, the directionality of this relationship remains unclear. Purpose The main aim of the present study was to clarify, with a cross-lagged panel design, the relationship between psychological status and exercise capacity among patients in a cardiac rehabilitation (CR) program. Methods A clinical sample of 212 CR patients completed exercise-capacity testing and measures of depression and anxiety (Hospital Anxiety and Depression Scale) pre-CR and post-CR. Demographic and clinical data, including BMI and smoking history, were also collected. Multivariate stepwise regression analysis was performed to identify the best predictors of exercise capacity at discharge. Structural equation modeling was utilized to quantify the cross-lagged effect between exercise capacity and psychological distress. Results Multivariate regression analysis revealed that higher levels of psychological distress pre-CR are predictively associated with less improvement in exercise capacity post-CR, beyond the effects of age, sex, and baseline functional status. Results from structural equation modeling supported a 1-direction association, with psychological distress pre-CR predicting lower exercise capacity post-CR over and above autoregressive effects. Conclusions Study results did not support the hypothesis of a bidirectional relationship between psychological distress and EC. High levels of psychological distress pre-CR appeared to be longitudinally associated with lower exercise capacity post-CR, but not vice versa. This finding highlights the importance of assessing and treating both anxiety and depression in the early phase of secondary prevention programs. Cardiac rehabilitation, Anxiety, Depression, Psychological distress, Exercise capacity Introduction The primary goal of cardiac rehabilitation (CR) is to reduce the cardiac risk profile of patients diagnosed with heart disease. Although CR promotes a multidisciplinary approach to tackle cardiovascular risk factors, structured exercise has been established as the central component of secondary prevention [1]. In fact, exercise-based CR has been found to lower risk of cardiovascular mortality and re-infarction, compared with CR without an exercise component [2, 3]. In this context, the increase in patients’ exercise capacity (EC) triggered by CR plays a central role [4–7]. The influence of psychological factors on cardiovascular improvements among CR patients is well-documented [8, 9], as well as the influence of such factors on the long-term prognosis of cardiac patients [10, 11]. Among these psychological factors, depression and anxiety are among the most widely researched [12, 13]. In a recent meta-analysis, Papasavvas et al. [14] found an association between poorer mental health and lower EC, over and beyond the effect of disease severity, age, and sex. However, the reciprocal relationship between psychological health and EC remains unclear. On the one hand, depression and anxiety seem to predict poorer exercise performance. The Heart and Soul Study [15] on 944 outpatients with stable coronary artery disease found a cross-sectional association between depressive symptoms and poor EC. In another study of CR patients, depressive symptoms predicted slower heart rate recovery after exercise stress testing [16]. Others have found that anxiety and depression are correlated with EC. Egger et al. [17] found that changes in the self-reported levels of depression and anxiety on the HADS from pre- to post-CR were correlated with changes in EC. On the other hand, research suggests that the reverse effect could be true: lower EC may predict symptoms of anxiety and depression. Greater EC predicted lower levels of trait and state anxiety among adolescents [18] and frequent high-intensity exercise has been shown to reduce anxiety sensitivity among adults [19]. Similarly, risk of developing moderate-to-severe symptoms of depression increases significantly as cardiorespiratory fitness decreases [20, 21]. Furthermore, these effects appear to be long-lasting: cardiorespiratory fitness remained a strong predictor of depressive symptoms during a 12-year follow-up [22]. Despite preliminary evidence of the reciprocal association of psychological status and exercise fitness [23], few studies have examined this relationship in the context of cardiovascular disease. Evon and Burns [24] examined a similar research topic with 80 CR patients, although this study mainly focussed on the cross-lagged effects between perceived self-efficacy, patient-staff working alliance, and the change of dietary and exercise habits. The current study evaluated the link between anxiety, depression, and EC among patients completing an exercise-based CR program. In particular, our study aimed (1) to explore the role of multiple risk factors in predicting EC post-CR among cardiac patients and (2) to investigate the bidirectional pathways between psychological distress and EC, using a cross-lagged panel design. The existence of a bidirectional association will be inferred if psychological distress predicts EC over time and vice versa. Support for the following hypotheses would indicate reciprocal associations, whereby (1) decreased symptoms of depression and anxiety will be associated with improvement in EC and (2) lower EC will be associated with increases in depression and anxiety symptoms post-CR. Methods Study Sample The study population consisted of 212 Caucasian patients diagnosed with heart disease, who entered a comprehensive outpatient CR program. Exclusion criteria were: inability to fully participate in the study due to physical (e.g., deafness), cognitive (e.g., dementia), or severe psychiatric (e.g., schizophrenia) impairment. Participants were referred to CR for the following conditions: coronary artery disease (CAD), post-valve repair/replacement, heart failure, or a combination of these diagnoses. Demographic and clinical characteristics of the sample are summarized in Table 1. Participants completed all study measures at admission to the CR program (t1) and at discharge (t2). The study protocol was approved by the Institutional Review Board. Patients were asked to provide written consent to participate, and confidentiality was assured. Table 1 Demographic and clinical characteristics of the sample Variable Sample n = 212 Male sex 80.7% Age mean (SD) 63.11 (11.63) Low education (lower than secondary school) 32.5% Professional status  Retired 49.1%  Full time/part-time working 41%  Not working/Housewives 9.9% Diagnosis  Ischemic heart disease 71%  Heart failure 5%  Valvular disease 16%  Comorbidity 8% Smoking status  Actual smoker 5.2%  Quitted smoking <1 year ago 30.2%  Quitted smoking >1 year ago 34.9%  Never smoked 27.8% BMI-based classification  Underweight 1.4%  Normal weight 45%  Overweight 39.3%  Obese 14.2% Variable Sample n = 212 Male sex 80.7% Age mean (SD) 63.11 (11.63) Low education (lower than secondary school) 32.5% Professional status  Retired 49.1%  Full time/part-time working 41%  Not working/Housewives 9.9% Diagnosis  Ischemic heart disease 71%  Heart failure 5%  Valvular disease 16%  Comorbidity 8% Smoking status  Actual smoker 5.2%  Quitted smoking <1 year ago 30.2%  Quitted smoking >1 year ago 34.9%  Never smoked 27.8% BMI-based classification  Underweight 1.4%  Normal weight 45%  Overweight 39.3%  Obese 14.2% View Large Table 1 Demographic and clinical characteristics of the sample Variable Sample n = 212 Male sex 80.7% Age mean (SD) 63.11 (11.63) Low education (lower than secondary school) 32.5% Professional status  Retired 49.1%  Full time/part-time working 41%  Not working/Housewives 9.9% Diagnosis  Ischemic heart disease 71%  Heart failure 5%  Valvular disease 16%  Comorbidity 8% Smoking status  Actual smoker 5.2%  Quitted smoking <1 year ago 30.2%  Quitted smoking >1 year ago 34.9%  Never smoked 27.8% BMI-based classification  Underweight 1.4%  Normal weight 45%  Overweight 39.3%  Obese 14.2% Variable Sample n = 212 Male sex 80.7% Age mean (SD) 63.11 (11.63) Low education (lower than secondary school) 32.5% Professional status  Retired 49.1%  Full time/part-time working 41%  Not working/Housewives 9.9% Diagnosis  Ischemic heart disease 71%  Heart failure 5%  Valvular disease 16%  Comorbidity 8% Smoking status  Actual smoker 5.2%  Quitted smoking <1 year ago 30.2%  Quitted smoking >1 year ago 34.9%  Never smoked 27.8% BMI-based classification  Underweight 1.4%  Normal weight 45%  Overweight 39.3%  Obese 14.2% View Large CR Program CR was performed according to the protocols followed in our center, according to practice guidelines of the European Society of Cardiology [25]. The program combined physical training and educational sessions and lasted approximately 6 weeks, with a mean of 21.2 (SD = 4.1) exercise sessions performed by the patients. Physical training Each exercise session lasted 90 min, including stretching, calisthenics, and 45 min of aerobic exercise by bicycle or treadmill. All patients were under constant ECG monitoring and cardiologic supervision. The bicycle/treadmill intensity level in Watts was automatically set by a feedback-computerized system (Custo Cardio Concept, CUSTO Med, Ottobrunn, Germany) to obtain the training heart rate (HR). For each patient, training HR was calculated as follows: training HR = rest HR + 60% HR reserve, where HR reserve = peak HR – rest HR at a preliminary bicycle ergometer exercise test. In patients for whom training HR could not be set due to atrial fibrillation or the presence of a pacemaker, training intensity was calculated by means of the perceived effort with a 10-point Borg scale [26] and a score of 4 set as the target exercise intensity. Training intensity was gradually adjusted according to the individual patient’s changes in EC to maintain the same level of HR and to avoid exercising above the initial HR target. Education The CR program included four group-based educational sessions conducted by CR doctors and nurses and focussed on: (1) various cardiovascular diseases and medical interventions (e.g., percutaneous transluminal coronary angioplasty, coronary stenting, and coronary bypass), (2) pharmacological therapies, and (3) the role of modifiable risk factors in cardiovascular health. In addition, trained dieticians conducted two group sessions that provided education on nutrients (e.g., sugars, fats, and proteins), food labels, and the preparation of healthy meals. Outcome Measurements Social-demographic and clinical data Participants provided their age, sex, marital status, educational level, and employment status. Clinical data, including the diagnosis, and body mass index (BMI) were retrieved from the participants’ medical record. Smoking status was assessed by a self-report questionnaire, which distinguished between four categories: (1) current smoker, (2) former smoker who had quit smoking less than 1 year prior to the survey, (3) former smoker who had quit more than 1 year prior to the survey, and (4) non-smoker. Hospital Anxiety and Depression Scale The Hospital Anxiety and Depression Scale (HADS) is a widely used instrument to assess the presence of anxiety and depression symptoms in hospital settings [27] and has recently been utilized as a measure of psychological distress in cardiac patients [28–30]. The HADS consists of 14 items, 7 measuring the anxiety level and 7 measuring depression. Items are scored on a 4-point Likert-scale, from 0 to 3, with a total score of 42, and a maximum score of 21 for each subscale. Higher scores indicate higher psychological distress. Both total score and subscale scores were used in the data analyses. Score interpretations are equivalent for both the Anxiety and Depression subscale; scores between 8 and 10 represent mild symptoms, 11–15 reflects moderate symptoms, and 16 or above represent severe symptoms [31]. A very recent study on 617 Italian cardiovascular patients [32] suggested that an HADS total score of 14 or greater provided high sensitivity and good specificity to identify clinically relevant symptoms of psychological distress among cardiac patients. The authors concluded that the HADS could be considered a good first-step screening tool to identify patients with a relevant level of psychological distress. Internal consistency for this study was good, total score Cronbach alphas was 0.84 at baseline and 0.86 at discharge. Adequate alpha scores for the subscales was found for both the Depression (0.80 at baseline and 0.81 at discharge) and for the Anxiety subscale (0.71 at baseline and 0.75 at discharge). Exercise capacity EC was defined as the maximal power (in Watts) reached during an incremental bicycle stress test, with an increase of 10 Watts every minute. Changes in the workload during the rehabilitation period were considered an index of physical training. Data Analysis Regression analyses To identify other significant predictors of the EC post-CR, a number of univariate regression analyses were performed. Bonferroni correction was applied, which yielded a significance threshold of 0.005. Variables predicting EC with a p-value of <.005 in the univariate regression analyses were included as potential independent predictors for the multivariate linear regression. Data were analyzed using the IBM SPSS (version 22.0, SPSS Inc., Chicago, IL, USA). Cross-lagged panel design A cross-lagged design was used to examine the stability of EC and psychological distress across time as well as potential cross-lagged effects. Model fit was assessed based on Δχ2, the comparative fit index (CFI), and the root mean square error of approximation (RMSEA). Maximum likelihood estimation was used to estimate structural equation models with the Mplus 7 program [33], testing full cross-lagged models, including EC and a composite anxiety and depression score at t1 and t2. All models were estimated using Full Information Maximum Likelihood (FIML) in Mplus. Age was included as covariate in all tested models. Results Descriptive Statistics CR adherence, defined as the percentage of CR exercise sessions completed, was 96.56% (SD = 5.67). Mean EC at CR entry was 91.31 Watts (SD = 27.9), which increased to 108.68 Watts (SD = 34.82) at discharge, showing a significant functional improvement over the course of CR (t194 = −12.93, p < .001; Cohen’s d = 0.55). Average BMI did not change (t208 = −1.05, p = .294; Cohen’s d =0.03) from t1 (M = 25.68, SD = 4.29) to t2 (M = 25.79, SD = 4.17). Average HADS anxiety score was 6.33 (SD = 3.76) at t1 and 5.41 (SD = 3.72) at t2, which marked a significant improvement (t209 = 4.72, p < .001; Cohen’s d = 0.25). Average HADS depression score was 4.32 (SD = 3.39) at t1 and 4.04 (SD = 3.43) at t2 (t207 = 1.54, p = .13; Cohen’s d = 0.08). The mean HADS total score was 10.64 (SD = 6.58) at t1 and 9.42 (SD = 6.57) at t2 (t207 = 3.89, p < .001; Cohen’s d = 0.1). Overall, the HADS total and subscale scores observed in our sample were similar to that reported in a larger Italian sample of cardiac patients [32]. On the basis of the cut-off value of 14 on the HADS total score [32], in the present sample 31.7% of the participants reported a clinically significant level of psychological distress pre-CR. These patients demonstrated significantly lower levels of EC post-CR (M = 97.05; SD = 30.30) than patients with an HADS total score < 14 (M = 113.85; SD = 35.55, t193 = 3.18, p < .002; Cohen’s d = 0.51) Communality and Regression Analyses The results of univariate regression analyses are reported in Table 2. As expected, the most influential predictor of EC post-CR was EC pre-CR (β = .840, p < .001). Younger age (β = −.462, p < .001) and male sex (β = .417, p < .001) predicted better EC post-CR, and a diagnosis of heart failure predicted worse EC post-CR (β = −.237, p = .001). No associations were found for self-reported smoking history (β = −.027, p = .71) and BMI (β = .13, p = .076). Anxiety (β = −.227, p < .001) and depression (β = −.233, p <.001), and the psychological distress measured by the HADS total score (β = −.250, p <.001), negatively predicted EC at post-CR. Table 2 Prediction of exercise capacity at t2: univariate regression models B CI 95% SE b β p-value Adjusted R2 EC at t1  Costant 12.97 3.77, 22.16 5.09 .70  EC at t1 1.05 .952, 1.14 .054 .84 <.001 Age  Costant 197.38 172.83, 21.93 4.66 .21  Age −1.41 −1.80, 1.03 .15 −.46 <.001 Sex  Costant 78.28 67.84, 88.71 5.29 .17  Sex 37.29 25.73, 48.84 5.86 .42 <.001 BMI at t1  Costant 79.4 52.99, 111.62 13.53 .01  BMI at t1 1.09 −.106, 2.13 .52 .13 .08 Smoking status at t1  Costant 109.11 104.08, 114.14 2.55 .001  Smoking at t1 −.38 −2.37, 1.62 1.01 −.03 .71 HADS anxiety at t1  Costant 121.37 112.09, 130.65 4.70 .05  HADS anxiety −2.06 −3.31, −.80 .64 −.23 <.001 HADS depression at t1  Costant 118.62 110.81, 126.42 3.96 .05  HADS depression −2.38 −3.81,−.96 .72 −.23 <.001 HADS total score  Costant 122.12 112.30, 128.97 4.59 .062  HADS total score −.13 −1.9, .36 .37 −.25 <.001 Diagnosis  Costant 113.93 108.29, 119.58 2.86 .07  Valvulopathy versus CAD −10.60 −23.45, 2.25 6.51 −.11 .10  Hearth failure versus CAD −41.43 −65.55, −17.32 12.22 −.24 .001  Multiple diagnoses −21.43 −38.95, −3.92 8.88 −.17 .02 B CI 95% SE b β p-value Adjusted R2 EC at t1  Costant 12.97 3.77, 22.16 5.09 .70  EC at t1 1.05 .952, 1.14 .054 .84 <.001 Age  Costant 197.38 172.83, 21.93 4.66 .21  Age −1.41 −1.80, 1.03 .15 −.46 <.001 Sex  Costant 78.28 67.84, 88.71 5.29 .17  Sex 37.29 25.73, 48.84 5.86 .42 <.001 BMI at t1  Costant 79.4 52.99, 111.62 13.53 .01  BMI at t1 1.09 −.106, 2.13 .52 .13 .08 Smoking status at t1  Costant 109.11 104.08, 114.14 2.55 .001  Smoking at t1 −.38 −2.37, 1.62 1.01 −.03 .71 HADS anxiety at t1  Costant 121.37 112.09, 130.65 4.70 .05  HADS anxiety −2.06 −3.31, −.80 .64 −.23 <.001 HADS depression at t1  Costant 118.62 110.81, 126.42 3.96 .05  HADS depression −2.38 −3.81,−.96 .72 −.23 <.001 HADS total score  Costant 122.12 112.30, 128.97 4.59 .062  HADS total score −.13 −1.9, .36 .37 −.25 <.001 Diagnosis  Costant 113.93 108.29, 119.58 2.86 .07  Valvulopathy versus CAD −10.60 −23.45, 2.25 6.51 −.11 .10  Hearth failure versus CAD −41.43 −65.55, −17.32 12.22 −.24 .001  Multiple diagnoses −21.43 −38.95, −3.92 8.88 −.17 .02 View Large Table 2 Prediction of exercise capacity at t2: univariate regression models B CI 95% SE b β p-value Adjusted R2 EC at t1  Costant 12.97 3.77, 22.16 5.09 .70  EC at t1 1.05 .952, 1.14 .054 .84 <.001 Age  Costant 197.38 172.83, 21.93 4.66 .21  Age −1.41 −1.80, 1.03 .15 −.46 <.001 Sex  Costant 78.28 67.84, 88.71 5.29 .17  Sex 37.29 25.73, 48.84 5.86 .42 <.001 BMI at t1  Costant 79.4 52.99, 111.62 13.53 .01  BMI at t1 1.09 −.106, 2.13 .52 .13 .08 Smoking status at t1  Costant 109.11 104.08, 114.14 2.55 .001  Smoking at t1 −.38 −2.37, 1.62 1.01 −.03 .71 HADS anxiety at t1  Costant 121.37 112.09, 130.65 4.70 .05  HADS anxiety −2.06 −3.31, −.80 .64 −.23 <.001 HADS depression at t1  Costant 118.62 110.81, 126.42 3.96 .05  HADS depression −2.38 −3.81,−.96 .72 −.23 <.001 HADS total score  Costant 122.12 112.30, 128.97 4.59 .062  HADS total score −.13 −1.9, .36 .37 −.25 <.001 Diagnosis  Costant 113.93 108.29, 119.58 2.86 .07  Valvulopathy versus CAD −10.60 −23.45, 2.25 6.51 −.11 .10  Hearth failure versus CAD −41.43 −65.55, −17.32 12.22 −.24 .001  Multiple diagnoses −21.43 −38.95, −3.92 8.88 −.17 .02 B CI 95% SE b β p-value Adjusted R2 EC at t1  Costant 12.97 3.77, 22.16 5.09 .70  EC at t1 1.05 .952, 1.14 .054 .84 <.001 Age  Costant 197.38 172.83, 21.93 4.66 .21  Age −1.41 −1.80, 1.03 .15 −.46 <.001 Sex  Costant 78.28 67.84, 88.71 5.29 .17  Sex 37.29 25.73, 48.84 5.86 .42 <.001 BMI at t1  Costant 79.4 52.99, 111.62 13.53 .01  BMI at t1 1.09 −.106, 2.13 .52 .13 .08 Smoking status at t1  Costant 109.11 104.08, 114.14 2.55 .001  Smoking at t1 −.38 −2.37, 1.62 1.01 −.03 .71 HADS anxiety at t1  Costant 121.37 112.09, 130.65 4.70 .05  HADS anxiety −2.06 −3.31, −.80 .64 −.23 <.001 HADS depression at t1  Costant 118.62 110.81, 126.42 3.96 .05  HADS depression −2.38 −3.81,−.96 .72 −.23 <.001 HADS total score  Costant 122.12 112.30, 128.97 4.59 .062  HADS total score −.13 −1.9, .36 .37 −.25 <.001 Diagnosis  Costant 113.93 108.29, 119.58 2.86 .07  Valvulopathy versus CAD −10.60 −23.45, 2.25 6.51 −.11 .10  Hearth failure versus CAD −41.43 −65.55, −17.32 12.22 −.24 .001  Multiple diagnoses −21.43 −38.95, −3.92 8.88 −.17 .02 View Large Anxiety and depression were highly correlated in the present sample (r = .692, p < .001). This result suggests a considerable overlap between anxiety and depression subscales when simultaneously considered as predictors of EC in a multivariate regression analysis [34]. To distinguish between the unique effect of anxiety, the unique effect of depression, and the overlapping effect of both on EC, a commonality analysis was performed [35]. Results are shown in Fig. 1. In particular, unique effects were computed as a squared semi-partial correlation (sr2) between depression, anxiety, and EC; communality was calculated by subtracting the two squared semi-partial correlation from the total R2 [36]. Anxiety and depression together significantly predicted EC (F2, 192 = 6.270, p = .002), explaining 6.2% of its total variance (R2 = .062). The portion of EC variance explained by the unique effect of depression and anxiety was 0.8% (sr2 = .008) and 1.1% (sr2 = .011), respectively, while 4.3% (sr2 = .043) was explained by the common effect of the two predictors. In conclusion, these results showed the existence of a substantial overlap between anxiety and depression in explaining EC, while the unique variances of the single components were negligible. Fig. 1 View largeDownload slide The proportion of variance shared by Anxiety, Depression and EC. The overlapping areas in the figure reflect the relative amount of variance shared by the three dimensions. a = proportion of EC variance explained by the unique effect of anxiety (1.1%); b = proportion of EC variance explained by the unique effect of depression (0.8%); c = proportion of EC variance explained by the common effect of anxiety and depression (4.3%) Fig. 1 View largeDownload slide The proportion of variance shared by Anxiety, Depression and EC. The overlapping areas in the figure reflect the relative amount of variance shared by the three dimensions. a = proportion of EC variance explained by the unique effect of anxiety (1.1%); b = proportion of EC variance explained by the unique effect of depression (0.8%); c = proportion of EC variance explained by the common effect of anxiety and depression (4.3%) Consequently, in the subsequent multivariate regression analyses, the total HADS score was utilized as a global indicator of psychological distress, in line with that suggested by Herrmann et al. [37]. The possible impact of the attendance of CR sessions on the observed association between functional status and emotional distress was assessed by computing the partial correlation between EC and the total HADS score, controlling for the absolute number of sessions attended by each patient. The resulting partial correlation (r = −.252; p < .001) did not substantially differ from the zero-order correlation (r = −.248; p < .001), showing that the number of sessions did not significantly influence the relationship between the other two variables. Another potential confounder in the relationship between EC and psychological distress is adherence to the CR program, calculated as the percentage of attended sessions over prescribed sessions. However, even after controlling for this variable, the difference from the zero-order correlation is minimal (r = −.227; p = .002). The results of the multivariate stepwise regression analysis are reported in Table 3. Data were verified for assumptions required for regression. Unstandardized residual scores were normally distributed and uncorrelated, with the value of Durbin–Watson diagnostic test of 2.01. The model of best-fit had four predictors: baseline EC, age, HADS total score, and sex. The model was statistically significant (F4, 192 = 134.76, p < .001) and accounted for 74.6% of the variability in the dependent variable. Table 3 Prediction of exercise capacity at t2: multivariate regression models B CI 95% SE B β p-value Adjusted R2 Step 1  Costant 12.96 3.65, 22.27 4.72 .70  EC at t1 1.05 .95, 1.14 .05 .84 <.001 Step 2  Costant 50.08 28.79, 71.38 11.54 .72  EC at t1 .97 .87, 1.07 .05 .77 <.001  Age −.48 −.73, −.23 .13 −.16 .001 Step 3  Costant 65.03 41.75, 88.31 12.30 .73  EC at t1 .93 .82, 1.03 .05 .74 <.001  Age −.56 −.81, −.31 .13 −.18 <.001  HADS total score −.59 −.99, −.18 .22 −.11 .001 Step 4  Costant 64.81 41.85, 87.77 12.15 .74  EC at t1 .87 .76, .98 .06 .69 <.001  Age −.60 −.85, −.35 .13 −.20 <.001  HADS total score −.52 −.93, −.12 .27 −.10 .01  Sex 9.30 2.0, 16.6 4.06 .10 .01 B CI 95% SE B β p-value Adjusted R2 Step 1  Costant 12.96 3.65, 22.27 4.72 .70  EC at t1 1.05 .95, 1.14 .05 .84 <.001 Step 2  Costant 50.08 28.79, 71.38 11.54 .72  EC at t1 .97 .87, 1.07 .05 .77 <.001  Age −.48 −.73, −.23 .13 −.16 .001 Step 3  Costant 65.03 41.75, 88.31 12.30 .73  EC at t1 .93 .82, 1.03 .05 .74 <.001  Age −.56 −.81, −.31 .13 −.18 <.001  HADS total score −.59 −.99, −.18 .22 −.11 .001 Step 4  Costant 64.81 41.85, 87.77 12.15 .74  EC at t1 .87 .76, .98 .06 .69 <.001  Age −.60 −.85, −.35 .13 −.20 <.001  HADS total score −.52 −.93, −.12 .27 −.10 .01  Sex 9.30 2.0, 16.6 4.06 .10 .01 View Large Table 3 Prediction of exercise capacity at t2: multivariate regression models B CI 95% SE B β p-value Adjusted R2 Step 1  Costant 12.96 3.65, 22.27 4.72 .70  EC at t1 1.05 .95, 1.14 .05 .84 <.001 Step 2  Costant 50.08 28.79, 71.38 11.54 .72  EC at t1 .97 .87, 1.07 .05 .77 <.001  Age −.48 −.73, −.23 .13 −.16 .001 Step 3  Costant 65.03 41.75, 88.31 12.30 .73  EC at t1 .93 .82, 1.03 .05 .74 <.001  Age −.56 −.81, −.31 .13 −.18 <.001  HADS total score −.59 −.99, −.18 .22 −.11 .001 Step 4  Costant 64.81 41.85, 87.77 12.15 .74  EC at t1 .87 .76, .98 .06 .69 <.001  Age −.60 −.85, −.35 .13 −.20 <.001  HADS total score −.52 −.93, −.12 .27 −.10 .01  Sex 9.30 2.0, 16.6 4.06 .10 .01 B CI 95% SE B β p-value Adjusted R2 Step 1  Costant 12.96 3.65, 22.27 4.72 .70  EC at t1 1.05 .95, 1.14 .05 .84 <.001 Step 2  Costant 50.08 28.79, 71.38 11.54 .72  EC at t1 .97 .87, 1.07 .05 .77 <.001  Age −.48 −.73, −.23 .13 −.16 .001 Step 3  Costant 65.03 41.75, 88.31 12.30 .73  EC at t1 .93 .82, 1.03 .05 .74 <.001  Age −.56 −.81, −.31 .13 −.18 <.001  HADS total score −.59 −.99, −.18 .22 −.11 .001 Step 4  Costant 64.81 41.85, 87.77 12.15 .74  EC at t1 .87 .76, .98 .06 .69 <.001  Age −.60 −.85, −.35 .13 −.20 <.001  HADS total score −.52 −.93, −.12 .27 −.10 .01  Sex 9.30 2.0, 16.6 4.06 .10 .01 View Large Cross-Lagged Panel Design In a second step, the structural relations between EC and HADS were specified as cross-lagged effects, which allowed the evaluation of time-lagged reciprocal effects between the two variables, while controlling for their stability over time, that is, autoregressive effects [38]. As such, the full model included the stability coefficients between EC at t1–t2 and HADS at t1–t2, the path from EC at t1 to HADS at t2, the path from HADS t1 to EC at t2 among all participants. Associations between HADS and EC at each of the two measurement times were also included in the model. Next, a series of nested models were tested and compared with the saturated model, which tested a hypothesis of mutual causality, and the change in model fit indices was evaluated to identify the best fitting model. Results of these model comparisons are shown in Table 4. When the path from HADS at t1 to EC at t2 was removed, the model fit worsened significantly (Δχ2 = 4.257, Δdf = 1, p = .039). Removal of the path from EC at t1 to HADS at t2 did not show a significant worsening of the chi-square index (Δχ2 = 1.345, Δdf = 1, p = .246). This model, presented in Fig. 1, fitted the data well (χ2 = 1.345; df = 1; p = .246, CFI = .99, RMSEA = .040). Results confirmed that EC pre-CR is a strong predictor of the EC post-CR (β = .820, p < .001). Furthermore, high levels of psychological distress negatively predicted EC post-CR (β = −.082, p < .036), over and above the effect of baseline EC. Therefore, psychological distress at baseline accounted for a significant, albeit small, amount of variance in EC post-CR. Table 4 Cross-lagged models’ comparisons Model df χ2(p) RMSEA CFI AIC Delta model Δχ2(p) Saturated Mutual causality 0 0 0 1 7905.03 − − Model 1 EC -> HADS 1 4.26 (.04) .12 .99 7907.29 1 vs 0 4.26 (.04) Model 2 HADS-> EC 1 1.34 (.25) .04 1 7904.38 2 vs 0 1.34 (.25) Model 3 Auto-regressions only 2 5.75 (.06) .09 .99 7906.78 3 vs 2 4.40 (.04) Model df χ2(p) RMSEA CFI AIC Delta model Δχ2(p) Saturated Mutual causality 0 0 0 1 7905.03 − − Model 1 EC -> HADS 1 4.26 (.04) .12 .99 7907.29 1 vs 0 4.26 (.04) Model 2 HADS-> EC 1 1.34 (.25) .04 1 7904.38 2 vs 0 1.34 (.25) Model 3 Auto-regressions only 2 5.75 (.06) .09 .99 7906.78 3 vs 2 4.40 (.04) The best fitting model is shown in bold type. View Large Table 4 Cross-lagged models’ comparisons Model df χ2(p) RMSEA CFI AIC Delta model Δχ2(p) Saturated Mutual causality 0 0 0 1 7905.03 − − Model 1 EC -> HADS 1 4.26 (.04) .12 .99 7907.29 1 vs 0 4.26 (.04) Model 2 HADS-> EC 1 1.34 (.25) .04 1 7904.38 2 vs 0 1.34 (.25) Model 3 Auto-regressions only 2 5.75 (.06) .09 .99 7906.78 3 vs 2 4.40 (.04) Model df χ2(p) RMSEA CFI AIC Delta model Δχ2(p) Saturated Mutual causality 0 0 0 1 7905.03 − − Model 1 EC -> HADS 1 4.26 (.04) .12 .99 7907.29 1 vs 0 4.26 (.04) Model 2 HADS-> EC 1 1.34 (.25) .04 1 7904.38 2 vs 0 1.34 (.25) Model 3 Auto-regressions only 2 5.75 (.06) .09 .99 7906.78 3 vs 2 4.40 (.04) The best fitting model is shown in bold type. View Large Furthermore, the t2 correlations between the residual variables is low, although statistically significant (r = −.15, p = .04), indicating a small amount of shared variance in HADS and EC scores, beyond that explained by autoregressive and cross-lagged effects. Overall, as shown by the R2 statistics, the model explained 58.4% of individual differences in HADS score at t2 and 70.9% in EC at t2. Age was included in the model as covariate and it was significantly associated with EC at t1 (r = −.393, p < .001) and at t2 (r = −.284, p < .001). Age was not significantly associated with HADS score, neither at t1 (r = −.067, p = .325), nor at t2 (r = −.004, p = .947). Finally, we tested the stability model, encompassing only autoregressive paths. In this case, the model fit worsened significantly (Δχ2 = 4.405, Δdf = 1, p = .0358), confirming that the previous model was the best fit for our data. In conclusion, results indicate that levels of pre-CR psychological distress predicted the EC post-CR, beyond the effect of baseline EC. Moreover, the results indicate that baseline EC does not significantly predict post-CR levels of anxiety and depression. Discussion Although the association between EC and psychological distress has been previously reported in cardiac patients [14, 39], to our knowledge this is the first study to examine the cross-lagged effects of EC and psychological distress among CR-patients. Consistent with previous studies [17, 39], our results showed a cross-sectional association between psychological distress and EC, even controlling for the patients’ adherence to the CR program. Moreover, this association has been showed in the present paper to remain significant at the end of the rehabilitation period. The multivariate regression analysis revealed that a significant, although small, amount of variance in post-CR EC was accounted for by pre-CR levels of psychological distress, beyond the effects of EC pre-CR, age, and sex. Furthermore, exploratory analyses showed a group difference in EC post-CR between those with and without significant levels of psychological distress pre-CR, with a medium effect size. Cross-lagged effects revealed a 1-direction association between greater psychological distress at pre-CR and lower EC post-CR, over and above that of autoregressive effects. Overall, these results lend support to the literature highlighting that pre-CR psychological distress, such as increased symptoms of depression and anxiety, may be prospectively associated with a reduced improvement in functional capacity post-CR. The mechanisms explaining this result may be related to depressive and anxiety symptomatology. In fact, cardiac patients with depressive symptomatology tend to have poorer fitness levels and exercise tolerance, due to lower levels of energy and decreased motivation to engage regularly in physical activity [40]. It has been suggested [41] that EC results obtained from patients with high levels of depression may be not representative of their true fitness level, even when the full diagnostic criteria for major depression have not been met. In these patients exercise testing may not be as reliable in detecting ischemia, potentially leading to false-negative diagnoses [41]. Fig. 2 View largeDownload slide Cross-lagged effects. Covariance between age and the other variables was included in the model but omitted from the figure. Only significant paths are shown in the figure; p-values are shown in parentheses. Fig. 2 View largeDownload slide Cross-lagged effects. Covariance between age and the other variables was included in the model but omitted from the figure. Only significant paths are shown in the figure; p-values are shown in parentheses. As for the role of anxiety, previous research noted that anxious individuals are prone to interpret cardiorespiratory sensations associated with physical activity as threatening or aversive [18, 42, 43], leading to decreased engagement/tolerance during exercise stress testing, and subsequently depict the patient as less fit than their non-anxious counterparts [39]. The data did not support the proposed bidirectional relationship between EC and psychological distress. Unexpectedly, EC at pre-CR was not associated with changes in psychological distress. Cardiorespiratory fitness and subsequent presence of depressive symptomatology have been reported in a number of studies [18, 22, 42]. However, the referenced studies included healthy subjects, and the presence of cardiovascular disease was a common exclusion criterion [20]. Previous researchers have found that mood improves over the course of a 12-week, 36-session, CR program [23, 44]. Since the current CR program lasted approximately 6 weeks, and the change in psychological distress from pre- to post-CR was of small effect, it is possible that the lack of association between EC pre-CR and change in psychological distress was due to the brevity of this CR program. Although most participants in this study were below the HADS clinical threshold, with an average total score of 10.68 (SD = 6.59), 31.7% of the study sample at pre-CR and 25% at post-CR, reported a clinically significant level of psychological distress (HADS score ≥ 14 [32]). This suggests that, even if the distribution of psychological distress is considerably skewed to low levels of symptomatology, there is a notable proportion of patients whose clinically significant score may explain the effects detected in the cross-lagged analysis. Previous research has found that patients who report depression or anxiety symptoms independently [45, 46], or together [47], following CR are at increased risk of mortality, and more likely to show little-to-no improvement in EC post-CR. Taken together, this highlights the importance of identifying patients who could benefit from adjunctive mental health services during CR. Doing so may improve risk stratification and reduce levels of psychological distress and maximize cardiovascular improvements during CR. Considerable research has noted the overlap between depression and anxiety symptoms in cardiac patients [48, 49] and many approaches have been proposed to account for this overlap in the assessment of depression and anxiety in clinical settings [50–54]. In the current study, the two dimensions were not considered independently, and a composite score was used as a measure of general psychological distress. Although this decision was data driven, it prevented the ability to distinguish between the separate effects of anxiety and depression on EC and to test moderation or mediation hypotheses. This study has several limitations. A longitudinal approach of relatively short length was adopted. Future research should evaluate whether these results hold at 6 months post-CR. Although adherence to the scheduled CR program was statistically controlled for in the analyses, telemetry data were not collected during this study and the effect of adherence to target heart rate recommendations on both psychological distress and EC cannot be determined. Furthermore, although the protocol employed in the study is highly standardized, treatment fidelity was not assessed formally, and this limitation may affect on the interpretation of the results. Moreover, although the cross-lagged panel design is useful to explore the bidirectional pathways linking psychological distress and EC, it does not allow for the causal relationship to be inferred [54]. As for the study sample, participants were exclusively Caucasian and a marked imbalance in favor of men was observed, as the percentage of women in the sample is 20% lower than expected based on cardiovascular morbidity data. Although this ratio is similar to that observed in other CR samples in the Italian population [55], the results may not be generalizable to women or men of ethnic minority groups. Moreover, cardiac patients with depression may be less likely to participate in an exercise-based CR program [56, 57], which could explain why the psychological distress data in this study were skewed toward lower scores on the HADS. In conclusion, results from this study suggest that CR patients with increased levels of psychological distress are more likely to demonstrate less EC post-CR. Therefore, we support a renewed attention to psychological risk factors, which should be routinely screened at the beginning of the rehabilitation period, and their evolution monitored. Early identification of distressed patients will provide the opportunity to connect the patient with outpatient mental health providers, and improve patient-centered care by identifying and addressing potential obstacles to maximal cardiorespiratory improvement, which could include extended CR sessions. The findings of the present study support a model integrating both medical and psychological factors, thus promoting a more personalized and holistic approach to the patient. Acknowledgements We are grateful to Matteo Baruffi, Francesco Borgia and Alessandra Meraviglia for their contributions to the data collection and to Emanuele M. Giusti for his contribution to the final revision of the manuscript. Compliance with Ethical Standards Authors’ Statement of Conflict of Interest and Adherence to Ethical Standards Authors Chiara A.M. Spatola, Emanuele A.M. Cappella, Christina L. Goodwin, Gianluca Castelnuovo, Roberto Cattivelli, Giada Rapelli, Gabriella Malfatto, Mario Facchini, Chiara Mollica, and Enrico Molinari declare that they have no conflict of interest. Ethical Approval All procedures were conducted in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000. Informed consent Informed consent was obtained from all individual participants included in the study. References 1. 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 ( 16 ): 1715 – 1733 . Google Scholar CrossRef Search ADS PubMed 2. 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 ( 4 ): 571 – 584.e2 . Google Scholar CrossRef Search ADS PubMed 3. 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 ): 1 – 12 . Google Scholar CrossRef Search ADS PubMed 4. Taylor SE , Way BM , Seeman TE . Early adversity and adult health outcomes . Dev Psychopathol . 2011 ; 23 ( 3 ): 939 – 954 . Google Scholar CrossRef Search ADS PubMed 5. Ades PA . Cardiac rehabilitation and secondary prevention of coronary heart disease . N Engl J Med . 2001 ; 345 ( 12 ): 892 – 902 . Google Scholar CrossRef Search ADS PubMed 6. Kokkinos P , Myers J . Exercise and physical activity: Clinical outcomes and applications . Circulation . 2010 ; 122 ( 16 ): 1637 – 1648 . Google Scholar CrossRef Search ADS PubMed 7. Suaya JA , Stason WB , Ades PA , Normand SL , Shepard DS . Cardiac rehabilitation and survival in older coronary patients . J Am Coll Cardiol . 2009 ; 54 ( 1 ): 25 – 33 . Google Scholar CrossRef Search ADS PubMed 8. Glazer KM , Emery CF , Frid DJ , Banyasz RE . Psychological predictors of adherence and outcomes among patients in cardiac rehabilitation . J Cardiopulm Rehabil . 2002 ; 22 ( 1 ): 40 – 46 . Google Scholar CrossRef Search ADS PubMed 9. Manzoni GM , Villa V , Compare A et al. Short-term effects of a multi-disciplinary cardiac rehabilitation programme on psychological well-being, exercise capacity and weight in a sample of obese in-patients with coronary heart disease: a practice-level study . Psychol Health Med . 2011 ; 16 ( 2 ): 178 – 189 . Google Scholar CrossRef Search ADS PubMed 10. Rothenbacher D , Hahmann H , Wüsten B , Koenig W , Brenner H . Symptoms of anxiety and depression in patients with stable coronary heart disease: prognostic value and consideration of pathogenetic links . Eur J Cardiovasc Prev Rehabil . 2007 ; 14 ( 4 ): 547 – 554 . Google Scholar CrossRef Search ADS PubMed 11. Watkins LL , Koch GG , Sherwood A et al. Association of anxiety and depression with all-cause mortality in individuals with coronary heart disease . J Am Heart Assoc . 2013 ; 2 ( 2 ): e000068 . Google Scholar CrossRef Search ADS PubMed 12. Elderon L , Whooley MA . Depression and cardiovascular disease . Prog Cardiovasc Dis . 2013 ; 55 ( 6 ): 511 – 523 . Google Scholar CrossRef Search ADS PubMed 13. Roest AM , Martens EJ , Denollet J , de Jonge P . Prognostic association of anxiety post myocardial infarction with mortality and new cardiac events: A meta-analysis . Psychosom Med . 2010 ; 72 ( 6 ): 563 – 569 . Google Scholar CrossRef Search ADS PubMed 14. Papasavvas T , Alhashemi M , Micklewright D . Association between depressive symptoms and exercise capacity in patients with heart disease: A meta-analysis . J Cardiopulm Rehabil Prev . 2017 ; 37 ( 4 ): 239 – 249 . Google Scholar CrossRef Search ADS PubMed 15. Ruo B , Rumsfeld JS , Pipkin S , Whooley MA . Relation between depressive symptoms and treadmill exercise capacity in the Heart and Soul Study . Am J Cardiol . 2004 ; 94 ( 1 ): 96 – 99 . Google Scholar CrossRef Search ADS PubMed 16. Hughes JW , Casey E , Luyster F et al. Depression symptoms predict heart rate recovery after treadmill stress testing . Am Heart J . 2006 ; 151 ( 5 ): 1122.e1 – 1122.e6 . Google Scholar CrossRef Search ADS 17. Egger E , Schmid JP , Schmid RW , Saner H , von Känel R . Depression and anxiety symptoms affect change in exercise capacity during cardiac rehabilitation . Eur J Cardiovasc Prev Rehabil . 2008 ; 15 ( 6 ): 704 – 708 . Google Scholar CrossRef Search ADS PubMed 18. Williams SE , Carroll D , Veldhuijzen van Zanten JJ , Ginty AT . Anxiety symptom interpretation: A potential mechanism explaining the cardiorespiratory fitness-anxiety relationship . J Affect Disord . 2016 ; 193 : 151 – 156 . Google Scholar CrossRef Search ADS PubMed 19. Broman-Fulks JJ , Berman ME , Rabian BA , Webster MJ . Effects of aerobic exercise on anxiety sensitivity . Behav Res Ther . 2004 ; 42 ( 2 ): 125 – 136 . Google Scholar CrossRef Search ADS PubMed 20. Tolmunen T , Laukkanen JA , Hintikka J et al. Low maximal oxygen uptake is associated with elevated depressive symptoms in middle-aged men . Eur J Epidemiol . 2006 ; 21 ( 9 ): 701 – 706 . Google Scholar CrossRef Search ADS PubMed 21. Galper DI , Trivedi MH , Barlow CE , Dunn AL , Kampert JB . Inverse association between physical inactivity and mental health in men and women . Med Sci Sports Exerc . 2006 ; 38 ( 1 ): 173 – 178 . Google Scholar CrossRef Search ADS PubMed 22. Sui X , Laditka JN , Church TS et al. Prospective study of cardiorespiratory fitness and depressive symptoms in women and men . J Psychiatr Res . 2009 ; 43 ( 5 ): 546 – 552 . Google Scholar CrossRef Search ADS PubMed 23. Milani RV , Lavie CJ . Reducing psychosocial stress: a novel mechanism of improving survival from exercise training . Am J Med . 2009 ; 122 ( 10 ): 931 – 938 . Google Scholar CrossRef Search ADS PubMed 24. Evon DM , Burns JW . Process and outcome in cardiac rehabilitation: an examination of cross-lagged effects . J Consult Clin Psychol . 2004 ; 72 ( 4 ): 605 – 616 . Google Scholar CrossRef Search ADS PubMed 25. Piepoli MF , Corrà U , Adamopoulos S et al. Secondary prevention in the clinical management of patients with cardiovascular diseases. Core components, standards and outcome measures for referral and delivery: a policy statement from the cardiac rehabilitation section of the European Association for Cardiovascular Prevention & Rehabilitation. Endorsed by the Committee for Practice Guidelines of the European Society of Cardiology . Eur J Prev Cardiol . 2014 ; 21 ( 6 ): 664 – 681 . Google Scholar CrossRef Search ADS PubMed 26. Borg G . A category scale with ratio properties for intermodal and interindividual comparisons . In: Geissler H-G and Petzold P , eds. Psychophysical Judgment and the Process of Perception . Berlin : VEB Deutscher Verlag der Wissenschaften 1982 : 25 – 34 . 27. Zigmond AS , Snaith RP . The hospital anxiety and depression scale . Acta Psychiatr Scand . 1983 ; 67 ( 6 ): 361 – 370 . Google Scholar CrossRef Search ADS PubMed 28. Norton S , Cosco T , Doyle F , Done J , Sacker A . The Hospital Anxiety and Depression Scale: A meta confirmatory factor analysis . J Psychosom Res . 2013 ; 74 ( 1 ): 74 – 81 . Google Scholar CrossRef Search ADS PubMed 29. Palacios JE , Khondoker M , Achilla E , Tylee A , Hotopf M . A single, one-off measure of depression and anxiety predicts future symptoms, higher healthcare costs, and lower quality of life in coronary heart disease patients: Analysis from a multi-wave, primary care cohort study . PLoS One . 2016 ; 11 ( 7 ): e0158163 . Google Scholar CrossRef Search ADS PubMed 30. Burns A , Höfer S , Curry P , Sexton E , Doyle F . Revisiting the dimensionality of the Hospital Anxiety and Depression Scale in an international sample of patients with ischaemic heart disease . J Psychosom Res . 2014 ; 77 ( 2 ): 116 – 121 . Google Scholar CrossRef Search ADS PubMed 31. Snaith R , Zigmond A. The Hospital Anxiety and Depression Scale with the Irritability-depression-anxiety Scale and the Leeds Situational Anxiety Scale: Manual . Windsor : Nfer-Nelson ; 1994 . 32. Tesio V , Marra S , Molinaro S , Torta R , Gaita F , Castelli L . Screening of depression in cardiology: A study on 617 cardiovascular patients . Int J Cardiol . 2017 ; 245 : 49 – 51 . Google Scholar CrossRef Search ADS PubMed 33. Muthen L. Mplus User’s Guide , 6th ed .. Los Angeles, CA : Muthén & Muthén ; 1998 . 34. Capraro RM , Capraro MM . Commonality analysis: Understanding variance contributions to overall canonical correlation effects of attitude toward mathematics on geometry achievement . Mult Lin Regr Viewp . 2001 ; 27 ( 2 ): 16 – 23 . 35. Nimon K . Regression commonality analysis: Demonstration of an SPSS solution . Mult Lin Regr Viewp . 2010 ; 36 ( 1 ): 10 – 17 . 36. Reichwein Zientek L , Thompson B . Commonality analysis: Partitioning variance to facilitate better understanding of data . J Early Interv . 2006 ; 28 ( 4 ): 299 – 307 . Google Scholar CrossRef Search ADS 37. Herrmann C . International experiences with the Hospital Anxiety and Depression Scale—a review of validation data and clinical results . J Psychosom Res . 1997 ; 42 ( 1 ): 17 – 41 . Google Scholar CrossRef Search ADS PubMed 38. Cole DA , Maxwell SE . Testing mediational models with longitudinal data: Questions and tips in the use of structural equation modeling . J Abnorm Psychol . 2003 ; 112 ( 4 ): 558 – 577 . Google Scholar CrossRef Search ADS PubMed 39. Pelletier R , Bacon SL , Laurin C , Arsenault A , Fleet RP , Lavoie KL . The impact of anxiety disorders on assessment of myocardial ischemia and exercise stress test performance . J Cardiopulm Rehabil Prev . 2011 ; 31 ( 1 ): 60 – 66 . Google Scholar CrossRef Search ADS PubMed 40. Buys R , Budts W , Delecluse C , Vanhees L . Exercise capacity, physical activity, and obesity in adults with repaired aortic coarctation . J Cardiovasc Nurs . 2013 ; 28 ( 1 ): 66 – 73 . Google Scholar CrossRef Search ADS PubMed 41. Lavoie KL , Fleet RP , Lespérance F et al. Are exercise stress tests appropriate for assessing myocardial ischemia in patients with major depressive disorder ? Am Heart J . 2004 ; 148 ( 4 ): 621 – 627 . Google Scholar CrossRef Search ADS PubMed 42. Goodwin RD . Association between physical activity and mental disorders among adults in the United States . Prev Med . 2003 ; 36 ( 6 ): 698 – 703 . Google Scholar CrossRef Search ADS PubMed 43. Sui X , Laditka JN , Church TS et al. Prospective study of cardiorespiratory fitness and depressive symptoms in women and men . J Psychiatr Res . 2009 ; 43 ( 5 ): 546 – 552 . Google Scholar CrossRef Search ADS PubMed 44. Lavie CJ , Milani RV . Cardiac rehabilitation and exercise training in secondary coronary heart disease prevention . Prog Cardiovasc Dis . 2011 ; 53 ( 6 ): 397 – 403 . Google Scholar CrossRef Search ADS PubMed 45. Milani RV , Lavie CJ . Impact of cardiac rehabilitation on depression and its associated mortality . Am J Med . 2007 ; 120 ( 9 ): 799 – 806 . Google Scholar CrossRef Search ADS PubMed 46. Herrmann C , Brand-Driehorst S , Buss U , Rüger U . Effects of anxiety and depression on 5-year mortality in 5,057 patients referred for exercise testing . J Psychosom Res . 2000 ; 48 ( 4-5 ): 455 – 462 . Google Scholar CrossRef Search ADS PubMed 47. Kachur S , Menezes AR , De Schutter A , Milani RV , Lavie CJ . Significance of comorbid psychological stress and depression on outcomes after cardiac rehabilitation . Am J Med . 2016 ; 129 ( 12 ): 1316 – 1321 . Google Scholar CrossRef Search ADS PubMed 48. Denollet J , Strik JJ , Lousberg R , Honig A . Recognizing increased risk of depressive comorbidity after myocardial infarction: looking for 4 symptoms of anxiety-depression . Psychother Psychosom . 2006 ; 75 ( 6 ): 346 – 352 . Google Scholar CrossRef Search ADS PubMed 49. Frasure-Smith N , Lespérance F . Depression and anxiety as predictors of 2-year cardiac events in patients with stable coronary artery disease . Arch Gen Psychiatry . 2008 ; 65 ( 1 ): 62 – 71 . Google Scholar CrossRef Search ADS PubMed 50. Rutledge T , Linke SE , Krantz DS et al. Comorbid depression and anxiety symptoms as predictors of cardiovascular events: Results from the NHLBI-sponsored Women’s Ischemia Syndrome Evaluation (WISE) study . Psychosom Med . 2009 ; 71 ( 9 ): 958 – 964 . Google Scholar CrossRef Search ADS PubMed 51. Suls J , Bunde J . Anger, anxiety, and depression as risk factors for cardiovascular disease: The problems and implications of overlapping affective dispositions . Psychol Bull . 2005 ; 131 ( 2 ): 260 – 300 . Google Scholar CrossRef Search ADS PubMed 52. Bleil ME , Gianaros PJ , Jennings JR , Flory JD , Manuck SB . Trait negative affect: Toward an integrated model of understanding psychological risk for impairment in cardiac autonomic function . Psychosom Med . 2008 ; 70 ( 3 ): 328 – 337 . Google Scholar CrossRef Search ADS PubMed 53. Frasure-Smith N , Lespérance F . Depression and other psychological risks following myocardial infarction . Arch Gen Psychiatry . 2003 ; 60 ( 6 ): 627 – 636 . Google Scholar CrossRef Search ADS PubMed 54. Kenny DA. Cross‐Lagged Panel Design . Wiley StatsRef : Statistics Reference Online; 2014. http://onlinelibrary.wiley.com/doi/10.1002/9781118445112.stat06464/compoundindex. 55. Giannuzzi P , Temporelli PL , Marchioli R et al. ; GOSPEL Investigators . Global secondary prevention strategies to limit event recurrence after myocardial infarction: results of the GOSPEL study, a multicenter, randomized controlled trial from the Italian Cardiac Rehabilitation Network . Arch Intern Med . 2008 ; 168 ( 20 ): 2194 – 2204 . Google Scholar CrossRef Search ADS PubMed 56. Cooper AF , Jackson G , Weinman J , Horne R . Factors associated with cardiac rehabilitation attendance: A systematic review of the literature . Clin Rehabil . 2002 ; 16 ( 5 ): 541 – 552 . Google Scholar CrossRef Search ADS PubMed 57. Lane D , Carroll D , Ring C , Beevers DG , Lip GY . Predictors of attendance at cardiac rehabilitation after myocardial infarction . J Psychosom Res . 2001 ; 51 ( 3 ): 497 – 501 . Google Scholar CrossRef Search ADS PubMed © Society of Behavioral Medicine 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Annals of Behavioral MedicineOxford University Press

Published: Feb 22, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off