مطالعه کنترل شده تصادفی ورزش ترکیبی و آموزش روانی برای زنان با SES پایین: نتایج کوتاه و طولانی مدت در کاهش علائم استرس و افسردگی

Abstract Exercise may have both a preventive and a therapeutic impact on mental health problems. The Exercise without Worries intervention aims to reduce stress and depressive symptoms in low-SES women by means of a group-based program combining physical exercise and psycho-education. Between September 2005 and May 2008, 161 Dutch low-SES women with elevated stress or depressive symptom levels were randomly assigned to the combined exercise/psycho-education intervention (EP), exercise only (E) or a waiting list control condition (WLC). The E condition provided low to moderate intensity stretching, strength, flexibility, and body focused training as well as relaxation, while the EP program integrated the exercise with cognitive-behavioral techniques. Depressive symptoms (CES-D) and perceived stress (PSS) were measured before and immediately after the intervention and at 2, 6 and 12 month follow-up. Multilevel linear mixed-effects models revealed no differential patterns in reduction of CES-D or PSS scores between the EP, E and WLC groups on the short (post-test and 2 month follow-up) or long term (6 and 12 months follow-up). Depressive symptom outcomes were moderated by initial depressive symptom scores: women from the EP and E groups with fewer initial symptoms benefited from participation on the short term. Further, women in the EP and E groups with the lowest educational level reported more stress reduction at post-test than women with higher educational levels. In the overall target population of low-SES women, no indications were found that the Exercise without Worries course reduced depressive symptom and stress levels on the short or long term. The findings do suggest, however, that exercise alone or in combination with psycho-education may be a viable prevention option for certain groups of disadvantaged women. Especially those low-SES women with less severe initial problems or those with low educational attainment should be targeted for future depression prevention practice.

Introduction Compared to women with higher socio-economic status (SES), low-SES women have a high risk (ORs between 2.7 and 5.4) for depressive symptoms, which in turn considerably increases their likelihood of developing major depression (Cuijpers & Smit, 2004; Kahn, Wise, Kennedy, & Kawachi, 2000). Chronic stress and daily hassles are considered strong predictors of depressive symptoms and episodes in this group (Falconnier & Elkin, 2008; McConagle & Kessler, 1990). One commonly used psychological intervention for the indicated prevention of depression is the Coping with Depression course (CWD) (Brown & Lewinsohn, 1984). Several studies have shown that the CWD course is effective in the prevention of new cases of major depressive disorders (Cuijpers, Munoz, Clarke, & Lewinsohn, 2009). Nevertheless, it appears that low-SES participants have more difficulties with the cognitive orientation of the CWD course, due to the high level of required verbal skills and the execution of homework assignments. This results in low-SES participants dropping out of the intervention more often (Allart-van Dam & Hosman, 2002). Based on the observation that low-SES women do often not engage in depression prevention, innovative ways to offer them preventive activities need to be found. Exercise may have both a preventive and a therapeutic impact on mental health problems and stress and has been found to be effective in reducing depressive symptoms (Alderman, Rogers, Johnson, & Landers, 2003; Stathopoulou, Powers, Berry, Smits, & Otto, 2006). Meta-analyses on the effect of exercise on depressive symptoms report effect sizes ranging between 0.35 and 0.59 (Conn, 2010). In addition to its application as a monotherapy, exercise is increasingly accepted as an adjunct intervention in the management of depressive symptoms and depression (Stathopoulou et al., 2006). The underlying assumption in offering these multidisciplinary interventions is that combining different types of effective interventions will increase and prolong their positive effects (Jané-Llopis, Hosman, Jenkins, & Anderson, 2003). However, research on the combination of exercise with empirically supported psychosocial approaches for depression has been limited (Stathopoulou et al., 2006) and the effects of an exercise regimen in combination with a psychosocial intervention component have not yet been tested within a low-SES population. To address this evidence gap, the Exercise without Worries (EWW) course was developed and evaluated in the Netherlands combining an exercise regimen with a CBT-approach. The development of the intervention was guided by the Stress Process Model of mental health disparities first proposed by Pearlin and colleagues (Pearlin, 1989, 1999; Pearlin, Menaghan, Lieberman, & Mullan, 1981). In short, this model proposes that ongoing stress may diminish the availability of women's psychosocial protective resources to cope with stressful circumstances, thus creating both increased psychological distress and diminished resilience. EWW was developed as a multi-component intervention to counter the negative effects of stressful events and increase psychosocial resources. The intervention has been tailored to the specific needs of adult low-SES women by way of its focus on risk factors for stress and depression relevant for this group. The exercise component has been included because of its potential impact on depressive symptoms. Further, it offers a solution for the low perceived attractiveness of exclusively cognitive-oriented programs among low-SES women. Experience from community programs suggests that exercise-based interventions might also appeal to this study population (Craft, Freund, Culpepper, & Perna, 2007; Lowther, Mutrie, & Scott, 2002). Barriers to participation in mental health interventions by low-SES women are often related to financial costs, transportation and time (Beeber et al., 2007; Weinreb, Perloff, Goldberg, Lessard, & Hosmer, 2006). Even though financial barriers form no, or at most a minor, impediment for disadvantaged populations to access mental health care in the Netherlands (Sareen et al., 2007), the EWW course is offered free of charge, and additional expenses related to child care and public transport are reimbursed. In addition, because the intervention is presented and executed as a course rather than as therapy, it is non-stigmatizing. This is of special importance in low-SES groups since this population is more likely to report stigma concerns for mental health issues, which generally inhibit their use of mental health services (Grote, Zuckoff, Swartz, Bledsoe, & Geibel, 2007). The objective of the current study was to evaluate the effects of this multidisciplinary preventive intervention aimed at reducing stress and depressive symptoms among low-SES women. We directly looked at whether the combination of exercise and psycho-education led to synergetic effects on participants' mental wellbeing in comparison to participants receiving exercise only and to a waiting list condition. Since previous research has shown differential intervention effects relating to socio-economic status (Cuijpers, Van Straten, Warmerdam, & Smits, 2008), baseline symptom levels (Allart-van Dam, Hosman, Hoogduin, & Schaap, 2003), current depression treatment and intervention adherence (Stuart, Perry, Le, & Ialongo, 2008), we further explored whether the intervention itself would lead to differential effects among the groups based on these moderators.

Results 381 Women were found to be eligible for participation in the course after screening. Of the 275 women who consented to an intake, 161 (58.5%) were enrolled in the study. A further 12 women dropped out before the baseline assessment took place and were excluded from any further analyses. The participants who dropped out after the baseline measurement did not differ from those who completed all assessments (p > 0.05). Table 1 presents the socio-demographic characteristics, as well as the presence of previous depressive episodes, current depression treatment and current exercise behavior at baseline for each of the three study conditions. No significant differences were found among the three conditions for any of these variables, with the exception of nationality. All participants in the control condition were Dutch, versus 91% in the EP condition and 84% in the E-condition (χ2 (2) = 7.47, p = 0.02). Furthermore, the 12-month follow-up showed that women in the E condition had experienced significantly more life events during the study period than women in the other conditions, F (2, 146) = 2.97, p < 0.05. Women attended on average 5.06 (SD = 2.79) sessions, without any significant differences between the two intervention conditions. Current treatment for depressive symptoms consisted of antidepressant medication, psychotherapy or a combination of those. Those women who performed exercise activities at baseline were mostly engaged in transport related physical activity such as walking or cycling or participated in leisure time activities as swimming or fitness training. Table 1. Baseline characteristics 149 participants, means (SD), frequencies (%). EP group (n = 55) E group (n = 46) WLC group (n = 48) Total sample Age (years) 44.68 (SD = 8.43) 43.06 (SD = 8.88) 43.87 (SD = 7.67) 43.90 (SD = 8.31) Nationality (Dutch)* 50 (90.9) 39 (84.8) 48 (100.0) 137 (91.9) Marital status Single, Divorced, Widowed 22 (40.0) 26 (56.5) 17 (35.4) 65 (43.6) Married, cohabitating 33 (60.0) 20 (43.5) 31 (64.6) 84 (56.4) Educational levela Low 27 (49.1) 28 (60.9) 24 (50.0) 79 (53.0) Middle 21 (38.2) 14 (30.4) 22 (45.8 57 (38.3) High 7 (12.7) 4 (8.7) 2 (4.2) 13 (8.7) Employment status Employed 13 (23.6) 13 (28.3) 16 (33.3) 42 (28.2) Housewife 14 (25.5) 8 (17.4) 10 (20.8) 32 (21.5) Unemployed 12 (21.8) 8 (17.4) 4 (8.3) 24 (16.1) Disabled 16 (29.1) 17 (37.0) 18 (37.5) 51 (34.2) Net family income per monthb €0–500 1 (1.8) 3 (6.5) 1 (2.1) 5 (3.4) €500–835 11 (20.0) 14 (30.4) 11 (22.9) 36 (24.2) €835–1100 18 (32.7) 8 (17.4) 9 (18.8) 35 (23.4) €1100–1665 17 (30.9) 15 (32.6) 16 (33.3) 48 (32.2) €1665–2500 5 (9.1) 3 (6.5) 6 (12.5) 14 (9.4) ≥€2500 3 (5.5) 3 (6.5) 5 (10.4) 11 (7.4) Previous depressive symptoms No 7 (12.7) 12 (26.1) 12 (25.0) 31 (20.8) Yes 48 (87.3) 34 (73.9) 36 (75.0) 118 (79.2) Current treatment for symptoms No 34 (61.8) 31 (67.4) 36 (75.0) 101 (67.8) Yes 21 (38.2) 15 (32.6) 12 (25.0) 48 (32.2) Current exercise activities No 36 (65.5) 27 (58.7) 32 (66.7) 95 (63.8) Yes 19 (34.5) 19 (41.3) 16 (33.3) 54 (36.2) No. of life events* 11.35 (SD = 9.59) 16.70 (SD = 16.51) 12.35 (SD = 6.53) 13.32 (SD = 11.63) Intervention adherence (No. of sessions attended) 5.47 (SD = 2.63) 4.57 (SD = 2.94) – 5.06 (SD = 2.79) Continuous data: mean (standard deviation); categorical data: frequency (percentage). EP exercise and psycho-education; E exercise only; WLC waiting-list comparison. *p < 0.05. a Educational level: low ≤9 years; mid = 10–14 years; high ≥15 years. b The average net family monthly income in the Netherlands in 2007 was €1743. Table options Intervention effects No significant differences were found in baseline outcome measures among the three research conditions. Table 2 shows the descriptive study measures from baseline to 12 month follow-up. Multilevel linear mixed-effects analyses revealed that women in the EP and E, but not the control, conditions showed significant (p < 0.05) improvements on their score for depressive symptoms at post-test. For the E condition, the change to the baseline scores was still significant at 6 and 12 month follow-up (p < 0.01). However, this change in depressive symptom scores was not significantly different (p = 0.12) between the EP, E and control conditions on any of the measurement time-points ( Table 3). Table 2. Observed, uncorrected means (SD) for CES-D and PSS scores per time point and group. Baseline* Post-test FU 2 months FU 6 months FU 12 months CES-D EP 24.91 (SD = 11.83) 21.29 (SD = 10.49) 21.52 (SD = 11.75) 22.12 (SD = 11.33) 20.87 (SD = 12.76) E 25.69 (SD = 12.30) 20.44 (SD = 11.47) 24.22 (SD = 13.55) 18.95 (SD = 11.87) 17.30 (SD = 11.08) WLC 23.24 (SD = 12.54) 20.51 (SD = 11.91) 20.95 (SD = 11.48) – – PSS EP 23.25 (SD = 6.76) 20.96 (SD = 6.78) 18.94 (SD = 7.02) 20.11 (SD = 6.16) 20.38 (SD = 6.41) E 24.39 (SD = 6.60) 21.32 (SD = 7.59) 21.97 (SD = 7.97) 21.19 (SD = 6.49) 20.20 (SD = 6.74) WLC 23.02 (SD = 6.27) 21.88 (SD = 6.76) 21.39 (SD = 6.42) – – EP exercise and psycho-education; E exercise only; WLC waiting-list comparison; FU follow-up. *Baseline group differences p > 0.05. Table options Table 3. Corrected intervention effects and effect sizea of group comparisons on CES-D and PSS scores per time point. CES-D EP–E† EP–WLC† E–WLC† B (SE) ES (95%CI) B (SE) ES (95%CI) B (SE) ES (95%CI) Post-test 2.24 (2.37) −0.08 (−0.31–0.47) −1.10 (2.31) 0.07 (−0.32–0.46) −3.34 (2.45) 0.00 (−0.41–0.40) FU 2 months −2.50 (2.40) 0.21 (−0.61–0.18) 0.22 (2.36) −0.05 (−0.34–0.44) 2.76 (2.53) −0.26 (−0.15–0.66) FU 6 months 2.25 (2.45) −0.27 (−0.12–0.66) – – – – FU 12 months 2.92 (2.47) −0.30 (−0.09–0.68) – – – – p-Value overall intervention effect 0.12 PSS Post-test 0.77 (1.37) 0.05 (−0.44–0.34) −1.20 (1.33) 0.13 (−0.52–0.25) −1.97 (1.42) 0.07 (−0.48–0.33) FU 2 months −2.13 (1.39) 0.40 (−0.79 to −0.01) −2.14 (1.37) 0.36 (−0.75–0.03) −0.01 (1.47) −0.08 (−0.32–0.48) FU 6 months −0.82 (1.43) 0.17 (−0.56–0.22) – – – – FU 12 months 0.89 (1.43) −0.02 (−0.36–0.41) – – – – p-Value overall intervention effect 0.34 EP exercise and psycho-education; E exercise only; WLC waiting-list comparison; FU follow-up. B – Regression weight; SE – standard error; ES – Effect size. The second group† is the reference category. a Corrected for baseline differences on the outcome variable and covariates. Table options Outcomes on perceived stress levels demonstrated a similar pattern. Compared to their baseline scores, women in both the EP and E conditions showed a significant decrease in their perceived stress levels at post-test and all follow-up time points (p < 0.01); there was no decrease, however, in the scores of the women in the control condition. Yet, overall intervention effects on decreases in stress levels among the three conditions were not significant at any of the measurement time-points (p = 0.34). Intervention effects in subgroups The moderation analyses showed intervention effects in some subgroups. For depressive symptoms, women in the EP and E conditions with a CES-D baseline score below the 16-point cut-off score reported significantly fewer depressive symptoms at 2 months follow-up (βEP = −7.77, p = 0.01, ES = 0.71; βE = −7.28, p = 0.03, ES = 0.63) than women in the control condition. Conversely, women in the E condition with a CES-D baseline score above the 16-point cut-off had significantly more depressive symptoms at 2 months follow-up than women in the control condition (βE = 8.05, p = 0.009, ES = −0.56). Subgroup analyses for the PSS showed that women from the EP and E conditions with the lowest educational level had lower perceived stress scores at post-test than women from the control condition (βEP = −4.12, p = 0.024, ES = 0.53; βE = −4.13, p = 0.025, ES = 0.35). None of the other condition × time × subgroup variable interactions was significant for either outcome measure.