Spiritual struggles and religious cognitive behavioral therapy: a randomized clinical trial in those with depression and chronic medical illness.

Pearce, Michelle J. ; Koenig, Harold G.

OBJECTIVES: Compared the effectiveness of religious cognitive-behavioral therapy (RCBT) versus conventional CBT (CCBT) on reducing spiritual struggles in persons with major depressive disorder (MDD) and medical illness. METHODS: Participants were randomized to receive 10 sessions of RCBT (n = 65) or CCBT (n = 67). Spiritual struggles were assessed at baseline, 12, and 24 weeks using the Brief RCOPE. Mixed effects growth curve models compared the effects of treatment on change in spiritual struggles.

RESULTS: RCBT and CCBT reduced spiritual struggles to a similar degree over time (B = -0.48, SE = 0.31, df = 151, t = -1.53, p = .127, d = 0.25). Similar results were found in those with high religiosity and with high spiritual struggles at baseline. Higher baseline spiritual struggles predicted a slower decline in MDD (B = 0.47, SE = 0.20, t = 2.30, p = .022) independent of treatment group.

CONCLUSIONS: RCBT and CCBT are equally effective in reducing spiritual struggles in persons with MDD and medical illness. High spiritual struggles decrease the response of MDD to both RCBT and CCBT. These findings have treatment implications.

Introduction

The many challenges associated with disabling chronic illness often give rise to spiritual struggles as people try to explain why this is happening to them. Why did God allow this to happen to me? Have I done something wrong to deserve this? Is God punishing me? I've prayed, but the situation doesn't appear to be changing. Does God really love me or have the power to make a difference? No one from my faith community has visited me or offered help. Have they abandoned me? My suffering is so great; I wonder if demonic forces are at work here. These are common questions that persons with physical illness ask, especially when illness is prolonged, suffering is great, and depression or discouragement sets in (Koenig, Pargament, & Nielsen, 1998). Spiritual struggles have been defined as negative thoughts, emotions, concerns, or conflicts regarding a religious/spiritual belief, practice, or experience (Koenig et al., 1998). Research indicates that spiritual struggles of this kind predict poorer mental health (Exline, 2013; McConnell, Pargament, Ellison, & Flannelly, 2006), worse physical functioning (Trevino, Balboni, Zoll-frank, Balboni, & Prigerson, 2014), disordered physiological functions (e.g., increased inflammation; Ai, Seymour, Tice, Kronfol, & Bolling, 2009; Rosmarin, Pargament, & Flannelly, 2009), and greater mortality (independent of baseline physical health, social support, and mental health; Ai, Pargament, Kronfol, Tice, & Appel 2010) in those with chronic medical illness.

Spiritual interventions in small samples have been shown to reduce spiritual struggles in non-medical settings (Murray-Swank & Pargament, 2005; Paragment, Koenig, Tarakeshwar, & Hahn, 2001). Likewise, spiritually-integrated interventions are effective for reducing depression, sometimes more so than secular interventions, among non-medically ill religious participants (Azhar & Varma, 1995; Azhar, Varma, & Dharap, 1994; Berry, 2002; Hodge, 2006; Hook et al., 2010; McCullough, 1999; Pargament, 1997; Propst, 1980; Propst, Ostrom, Watkins, Dean, & Mashburn, 1992; Razali, Hasanah, Aminah, & Subramaniam, 1998; Smith et al., 2007; Tan & Johnson, 2005; Wade, Worthingon, & Vogel, 2007; Worthington & Sandage, 2001). A meta-analytic review of 46 spiritually-integrated intervention studies showed that participants who were more spiritual/religious showed greater improvement in spiritual outcomes and a similar improvement in psychological outcomes when receiving spiritually-integrated, compared to secular, treatments (Worthington et al., 2011). To our knowledge, however, no randomized clinical trials have yet examined the effects of a religiously-integrated therapy versus a conventional secular therapy on changes in spiritual struggles over time in those with chronic medical illness. The present report is intended to fill that gap.

Objectives

We compared the effects of religiously-integrated cognitive behavioral therapy (RCBT) versus conventional CBT (CCBT) on spiritual struggles in persons with major depressive disorder with comorbid chronic medical illness. This analysis was a sub-project of the parent study that examined the effects of RCBT versus CCBT on course of depressive symptoms (Koenig et al., 2015). We hypothesized that (1) RCBT (which addresses spiritual struggles directly) will have a greater effect on decreasing spiritual struggles compared to CCBT ; (2) baseline religiosity will moderate this effect (i.e., RCBT will be more effective in reducing spiritual struggles in those who are more religious); (3) baseline spiritual struggles will moderate the effect of treatment on depressive symptoms (i.e., RCBT will be more effective than CCBT in those with high levels of spiritual struggle); (4) the differences above will be greater in the per-protocol analyses (among those who actually receive the therapy); and (5) baseline spiritual struggles will predict a slower decline in depressive symptoms, independent of treatment group.

Method

A full description of the methods and study design has been reported elsewhere (Koenig et al., 2015), although we briefly summarize them here. This was a multi-site randomized clinical trial that recruited participants from Durham County in North Carolina and from Los Angeles County in southern California. Inclusion criteria were (1) ages 18-85 years; (2) having one or more chronic medical illness; (3) indicating that religion or spirituality was at least somewhat important; (4) meeting DSM-IY criteria for major depressive disorder (MDD) using the MINI Neuropsychiatrie Inventory (Sheehan et al., 1998); (5) having a Beck Depression Inventory (BDI-II; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) score of 10 to 40 (mild to moderately severe MDD); and (6) English speaking. Exclusion criteria were (1) significant cognitive impairment as indicated by the brief Mini-Mental State Exam (< 14; Koenig, 1996); (2) psychotherapy in the past two months; (3) psychotic disorder, alcohol or substance abuse, or post-traumatic stress disorder within the past 12 months; (4) lifetime history of bipolar disorder; (5) active suicidal thoughts; (6) diagnosis of HIV/AIDS, autoimmune diseases, or endocrine disorders, or taking immunosuppressant drugs (excluded because of planned immune and endocrine analyses); and (7) no access to a telephone or the Internet.

Procedures

Potentially eligible participants were screened by telephone and scheduled for an in-person visit when further screening was completed after written informed consent was obtained. If eligible, participants completed a baseline evaluation and were then randomized to either RCBT or CCBT. Follow-up assessments were performed by trained interviewers blind to treatment group at 4, 8, 12, and 24 weeks (BDI was assessed at each time point, and spiritual struggles were assessed at baseline, 12 weeks, and 24 weeks); all outcome assessments were self-rated and required minimal assistance from interviewers. The study was approved by Duke University Medical Center and Glendale Adventist Medical Center institutional review boards.

Interventions

Therapists. Master's level therapists who were experienced practitioners of CBT delivered the therapy. The four CCBT therapists had no experience integrating religious beliefs into therapy, in contrast to the four RCBT therapists who had extensive experience in this regard. All therapists were trained and supervised by faculty in the department of psychology at Duke. To ensure they were qualified, therapists had to score 40 or higher on the Cognitive Therapy Rating Scale as rated by their supervisors (Vallis, Shaw, & Dobson, 1986).

Interventions. All participants were to receive ten 50-minute sessions of either CCBT or RCBT administered over 12 weeks. Because participants had chronic medical illness, treatment sessions were conducted remotely by telephone (94%), Skype (5%), or instant messaging (1%) in order to improve access to treatment.

Conventional CBT was manual-based and followed the CBT approach for depression as described by Beck, Rush, Shaw, and Emery (1979). CCBT focused on identifying and challenging dysfunctional cognitions and encouraging positive behaviors to counter depressive symptoms. When a participant raised religious issues during the therapy, the CCBT therapist gently redirected the conversation to dealing with the problem in more secular ways. Mindfulness meditation was included as part of the CCBT intervention in order to make the two treatments as similar as possible since meditation was also part of the RCBT intervention. Sessions also incorporated positive-psychology exercises, including forgiveness, gratitude, altruism, benefit-finding, and generating hope.

Religious CBT was also manual-based and was developed directly from the manual used by CCBT therapists. The only difference between CCBT and RCBT was that the latter integrated participants' religious beliefs and motivations into the therapy (Pearce et al., 2014). The RCBT intervention was specific to the religious beliefs of participants and was guided by Christian, Jewish, Muslim, Buddhist, and Hindu versions of the RCBT manual (along with therapist and participant workbooks in each of these faith traditions). These RCBT manuals were developed by university faculty who were both from these faith traditions and were experienced with integrating religious beliefs into therapy using a CBT framework. In addition, these individuals helped supervise therapists when participants from their particular faith tradition entered the study.

Integrating religious beliefs into therapy was the only difference between CCBT and RCBT. Both therapies focused on forgiveness, gratitude, and involvement in social activities; both adapted CBT to dysfunctional cognitions specific to chronic physical illness and disability; and both emphasized meditation practices. Only the RCBT intervention used the religious beliefs of clients as the rationale for behavioral activation and for challenging negative cognitions.

Treatment fidelity. To ensure that treatment followed the manuals and was delivered competently, all sessions were audio recorded and a random sample of 10% (front loaded towards the beginning of the study) were transcribed and rated by trained and supervised therapists not involved in the study using an adapted version of the Adherence Rating Scale (ARS; Waltz, Addis, Koerner, & Jacobson, 1993).

Measures

Spiritual struggles. The 7-item negative religious coping (NRC) subscale of the Brief RCOPE was used to measure spiritual/religious struggles (Pargament, Smith, Koenig, & Perez, 1998). This measure has been the traditional way that researchers have measured spiritual struggles for the past 15 years. The items are: "I wondered whether God had abandoned me"; "I felt punished by God for my lack of devotion"; "I wondered what I did for God to punish me"; I questioned God's love for me"; "I questioned the power of God"; "I wondered whether my church had abandoned me"; and "I decided the Devil made this happen." Each item was self-rated on a scale from 0 (not at all, i.e., very untrue) to 3 (a great deal, i.e., very true"). The scale score ranged from 0 to 21. The original validation study administered the 7-item NRC subscale to 296 members of two churches in Oklahoma City at the time of the bombing of the federal building in 1995; internal consistency (Cronbach's alpha) was .78 in that sample. All seven NRC items loaded on a single factor demonstrating construct validity. In the current sample, the alpha was .77 and all items likewise loaded on a single factor (eigenvalue = 2.45) that explained 92% of the variance.

For the seven Buddhists enrolled in the study, a different measure--the BCOPE--was used to assess religious coping, which included a 3-item subscale that assessed NRC (part of a 13-item measure of religious coping; Phillips, Colvin, Hietbrink, & Vonnegut, 2012). The items were: "Felt powerless because karma had caused the events"; "Find Buddhist practices hard to follow"; and "Find I am upset with myself for not remaining mindful of my experience." Factor analysis of the 3-item subscale revealed a single factor (eigenvalue = 1.50) that explained 100% of the variance. In order to include these individuals, the average score on the 3-items was inserted as the missing value for items 4-7 and summed to form a comparable 7-item measure of NRC that could be used in analyses.

In order to assess moderating effect of spiritual struggles on treatment response, individuals were categorized into high and low spiritual struggles. Those who scored 0.5 standard deviations above the NRC scale mean or higher ([greater than or equal to] 5.0) were considered "high" (n = 38) and those who scored less than that were considered "low" (n = 92). All other analyses used the spiritual struggles score as a continuous measure.

Religiosity. Religiosity at baseline was assessed by religious importance (1 item); attendance at religious services (1 item); and frequency of prayer, meditation, and scripture reading (1 item; Koenig, Meador, & Parkerson, 1997); daily spiritual experiences (16 items; Underwood & Teresi, 2002); and intrinsic religiosity (10 items; Hoge, 1972). Religious variables were combined to form a 29-item religiosity measure with scores that ranged from 44 to 153, which demonstrated high reliability (alpha = .95). One participant failed to complete one item, and one participant failed to complete two items; for these individuals, the mean score on items answered was inserted for the missing value prior to combining items. Those who scored 0.5 standard deviations above the mean or higher ([greater than or equal to] 116) were defined as "high" on religiosity (n = 44) and the remaining individuals were defined as "low" (n = 86). The dichotomized religiosity variable was used only in the stratified analyses; otherwise, the variable was left as continuous. Religiosity was assessed only at baseline.

Depression. The presence of major depressive disorder based on DSM-IV criteria was diagnosed using modules from the Mini-International Neuropsychiatric Interview (MINI; Sheehan et al., 1998). Depression severity was assessed using the Beck Depression Inventory (BDI-II) with a possible range of scores from 0 to 63, where mild to moderately severe depression is defined having a score between 10 and 40 (Beck et al., 1961). The BDI-II is widely used to assess depressive symptoms in studies of primary care patients, especially in randomized clinical trials (Serfaty et al., 2009).

Demographics and health. Age (years), gender, race, education (years), marital status, and religious affiliation were assessed. Physical functioning (Hlatky et al., 1989), severity of illness (Linn, Linn, & Gurel, 1968), and medical comorbidity (Charlson, Pompei, Ales, & Mackenzie, 1987) were measured using standard scales.

Statistical Analyses

The primary outcome (hypothesis #1) was a comparison of trajectories of change in spiritual struggles (NRC subscale scores) from baseline through 24 weeks between RCBT and CCBT using an intention-to-treat (ITT) approach. Secondary outcomes were examination of the moderating effects of baseline religiosity on response of spiritual struggles to RCBT and CCBT (hypothesis #2), the moderating effects of baseline spiritual struggles on decline in depressive symptoms in response to RCBT and CCBT (hypothesis #3), comparison of trajectories of change in spiritual struggles among those completing 5-10 therapy sessions (per-protocol; hypothesis #4), and the effect of baseline spiritual struggles on change in depressive symptoms over time (hypothesis #5).

At baseline, the Student's t-test was used to compare characteristics between treatment groups for continuous variables, and the chi-square statistic was used for categorical variables. The Student's t-test was also used to compare spiritual struggles at baseline, 12 weeks (at the end of treatment), and 24 weeks (12 weeks after end of treatment; Table 1). In the ITT analysis to test the primary hypothesis, a growth curve model using random intercept and slope (mixed effect regression models) was used to compare the effects of RCBT versus CCBT on trajectory of change in spiritual struggles from baseline to 24 weeks (Model 1). Included in this model were treatment group, time, and group by time interaction. This analytic approach allowed for participants with data for at least one time point to be included in the analysis and helped to address the problem of missing data (i.e., not all participants enrolled in the trial completed the trial or follow-up evaluations; 29.5% completed 0-4 sessions and 70.5% completed 5-10 sessions, requiring that we address the issue of missing data).

For testing the moderating effects of baseline religiosity on response of spiritual struggles to RCBT versus CCBT, religiosity was included in Model 1 along with an interaction term between religiosity and treatment group; analyses were then stratified by high (those scoring at or higher than 0.5 standard deviations above the mean) and low religiosity (other participants). For testing the moderating effects of baseline spiritual struggles on response of depressive symptoms to treatment, spiritual struggles along with an interaction term between spiritual struggles and treatment group were included in a growth curve model (Model 2) predicting change in depressive symptoms that already contained time, time-squared, group, and a group by time interaction; analyses were then stratified by high (those scoring at or higher than 0.5 standard deviations above the mean) and low spiritual struggles. All analyses were repeated in those receiving at least five treatment sessions (per-protocol).

To determine whether spiritual struggles influenced the trajectory of change in depressive symptoms over time independent of treatment group, baseline spiritual struggles was entered into Model 2 (ITT). Effect sizes (Cohen's d) were calculated using degrees of freedom (df) and t values from the mixed models. For the primary ITT analysis, significance level was set at p < .05. For secondary analyses, given their exploratory nature, significance level was also set at p < .05 and trend level was set at .05 < p < .10, without adjusting for multiple comparisons. All analyses were performed using SAS (version 9.3; SAS Institute Inc., Cary, NC).

Results

Between July 29, 2011 and May 29, 2013, 132 participants were enrolled in the study. Of those, 130 completed the NRC subscale of the Brief RCOPE at baseline, 91 completed it at 12 weeks, and 63 completed it at 24 weeks. The average age of participants was 51.6 years (SD = 13.5, range 24-84); average education level was 15.1 years (SD = 3.4); 68.5% were female; and 39.2% were married. Overall, 89.2% were Christian (26.2% Catholic, 39.2% Protestant, 23.9% non-traditional), 3.1% Jewish, 0.8% Muslim, 2.3% Hindu, and 4.6% Buddhist. With regard to physical health, number of comorbid medical conditions was 2.6 (SD = 2.6, range 1-18), physical functioning was 28.8 (SD = 5.7, range 17-36), and severity of illness was 6.8 (SD = 5.2, range 0-22). Average depressive symptoms at baseline were 25.4 on the BDI-II (SD = 8.5, range 10-42). Participants were equally divided between the North Carolina and California sites (47.7% and 52.3%, respectively). Among enrolled participants, 70% completed at least 5 of the 10 therapy sessions. A comparison of the baseline characteristics of each treatment group revealed no significant differences on any demographic, social, religious, physical health, or depression characteristic, indicating randomization was successful in forming two similar groups.

Therapists closely followed the manuals and were competent in the administration of the therapy, with an average score by outside raters for session structure of 13.3 [+ or -] 1.6 for RCBT therapists compared to 13.0 [+ or -] 2.3 for CCBT therapists (range 0-15); scores for therapeutic relation were RCBT 5.4 [+ or -] 0.8 versus CCBT 5.4 [+ or -] 0.9 (range 0-6); for adherence to the manual were RCBT 6.8 [+ or -] 1.0 versus CCBT 6.8 [+ or -] 1.1 (range 0-8); for competence were RCBT 3.2 [+ or -] 0.6 versus CCBT 3.3 [+ or -] 0.6 (range 1-4); and for flexibility were RCBT 3.1 [+ or -] 0.7 versus CCBT 3.4 [+ or -] 0.6 (range 1-4; Koenig et al., 2015).

The average spiritual struggles score on the NRC subscale of the RCOPE was 3.35 (SD = 3.31, range 0-14) at baseline, 2.74 (SD = 3.04, range 0-14) at 12 weeks (immediately after completion of therapy), and 2.52 (SD = 2.89, range 0-17) at 24 weeks from baseline. Table 1 presents the responses for each of the seven NRC items, comparing them to responses given by 1,445 persons who responded to the 1998 General Social Survey. Spiritual struggles appeared to be more prevalent in the present sample than in the community sample, at least in terms of feeling abandoned by God.

Hypothesis #1

Is RCBT more effective than CCBT in affecting change in spiritual struggles? Although spiritual struggles decreased over time, there was no indication that RCBT was more effective than CCBT in hastening this decrease in the intention-to-treat (ITT) analysis based on Student i-tests comparing average spiritual struggle scores between treatment groups (Table 2, Figure 1). These results were confirmed in the mixed effect growth curve model (Model 1), which found that while spiritual struggles decreased significantly over time (p = .003), RCBT was not more likely than CCBT to reduce spiritual struggles (time by group interaction B = -0.48, SE = 0.31, df= 151, i = -1.53, p = .127, d = 0.25; Table 3).

Hypothesis #2

Does baseline religiosity moderate the effects of RCBT on spiritual struggles? Comparison of spiritual struggle score means between treatment groups (Table 2) indicated that, even among highly religious clients, the effect of RCBT appeared no greater than that of CCBT in decreasing spiritual struggles. These results were confirmed in the mixed effects model, which revealed only a weak interaction between group and baseline religiosity (B = 0.34, SE = 0.020, df = 151, t = 1.68,p = .095, d = 0.27; Table 3). When the model was run separately in high and low religious clients, results in the high religious group (group by time interaction B = -0.41, SE = 0.50, df = 55, t = -0.81, p = .42, d = 0.22) and low religious group (B = -0.56, SE = 0.41, df = 94, p = -1.38,p = .17, d = 0.28) were similar.

Hypothesis #3

Do spiritual struggles moderate the effect of RCBT on depressive symptoms? In the mixed effects model that compared the effects of RCBT and CCBT on depressive symptoms (from baseline through 24 weeks of follow-up, Model 2), the interaction term between group and baseline spiritual struggles scores did not contribute significantly to the model (B for interaction = 0.05, SE = 0.4l, df = 363, t = 0.13, p = .900, d = 0.01). When spiritual struggle scores were dichotomized into high and low groups (high = spiritual struggle scores greater than or equal to 0.5 standard deviations above the mean) and Model 2 was run in each group, the effects of treatment group on course of depressive symptoms in those with high spiritual struggle scores (group by time interaction B = -0.65, SE = 1.00, df = 102, t = -0.65,p = .52, d = 0.13) was similar to the effects in those with low spiritual struggles (B = 0.27, SE = 0.49, df = 258, t = 0.55, p = .58, d = 0.07).

Hypothesis #4

The benefits of RCBT over CCBT will be particularly evident in those who actually receive the therapy (per-protocol analyses). Among those who received at least five treatment sessions of both CCBT and RCBT, there was no indication that RCBT was more effective than CCBT in decreasing spiritual struggles when comparing means at the 12- and 24-week follow-up assessments (using Student's i-test; Table 2). When examining results from the mixed effects growth curve model, there was a weak trend favoring those receiving CCBT (group by time interaction B = -0.59, SE = 0.35, df = 137, t = 1.71, p = .09, d = 0.29; Table 3). Furthermore, there was no indication in the per protocol analyses that the religiosity of participants at baseline moderated this effect on spiritual struggles, or that spiritual struggles at baseline moderated the effect of treatment on depressive symptoms. The interaction term between religiosity and treatment group in predicting spiritual struggles bordered on, but was not, significant (B = 0.04, SE = 0.02, df= 137, t= 1.67, p = .097, d = 0.29), nor did stratifying analyses by religiosity indicate that RCBT was more effective than CCBT in those with high religiosity (group by time interaction B = -0.54, SE = 0.54, df= 51, t = -0.99, p = .33, d = 0.28) or in those with low religiosity (B = -0.63, SE = 0.45, df= 84, t = -1.39,p = .17,d= 0.30). Likewise, the interaction between spiritual struggles and treatment group on depressive symptoms was not significant (5 = -0.08, SE = 0.47, df= 319, t = 0.18, p = .86, d = 0.02), nor was there any indication that RCBT was more effective than CCBT in treating depression among clients with high spiritual struggles at baseline (time by group interaction B = 0.30, SE = 1.07, df = 98, t = 0.28,p = .78, d = 0.06) or low spiritual struggles (B =-0.21, SE = 0.50, df= 218, t = -0.42, p = .67, d = 0.06).

[FIGURE 1 OMITTED]

Hypothesis #5

Do baseline spiritual struggles predict changes in depression over time regardless of treatment group? In the overall sample, spiritual struggle scores and depressive symptoms were positively correlated at baseline (r = .15, p = .088), 12 weeks (r = .34,p = .0009), and 24 weeks (r = .17, p = .178), especially in those with lower religiosity (baseline, r = .20, p = .064; 12 weeks, r = .383,p = .003; 24 weeks, r = .13,p = .420). Changes in spiritual struggle scores and depressive symptoms between baseline and follow-up were also weakly correlated (r = .19, p = .069, n = 93). When included in Model 2, baseline spiritual struggles predicted a slower decline in depressive symptoms over time independent of treatment group (B = 0.47, SE = 0.20, df - 363, t = 2.30, p = .022, d = 0.24), especially in those with low religiosity (B = 0.57, SE = 0.26, df = 235, t= 2.16,p = .032, d = 0.28); the effect weakened in those with high religiosity (B = 0.37, SE = 0.33, df = 125, t = 1.13, p = .261, d = 0.20).

Discussion

We examined the effects of religiously-integrated CBT versus secular conventional CBT on spiritual struggles reported by persons with major depressive disorder and chronic medical illness. We found that spiritual struggles decreased over time for both those receiving RCBT and CCBT, although RCBT was not superior to CCBT in this regard. There was no indication that baseline religiosity moderated the effects of the RCBT intervention, and CCBT was just as effective as RCBT in those who were highly religious. Likewise, RCBT was no more effective in those with a high level of spiritual struggle at the beginning of treatment--even though the RCBT intervention targeted spiritual struggles in the treatment while CCBT targeted psychological struggles more generally. The results were the same for both the intention-to-treat analyses and the per-protocol analyses that included only those who received at least 5 of 10 treatment sessions (where CCBT was weakly superior to RCBT, p = .09, d = 0.29). Baseline spiritual struggles predicted a significantly slower decline in depressive symptoms over time, independent of treatment group, and this was especially true for those who were less religious.

Interpretation

Most of these findings were unexpected, particularly because RCBT explicitly addressed spiritual struggles whereas CCBT did not. One reason for the lack of difference between treatments might be the prevalence and measurement of spiritual struggles in this sample. The participants here endorsed a low level of spiritual struggles. The possible range of scores on the negative RCOPE (NRC) scale is 0-21, and our sample's mean score was 3.35 at baseline, 2.74 at 12 weeks, and 2.52 at 24 weeks. Greater variability in spiritual struggle scores may have resulted in a greater power to detect differences between treatment groups, although low prevalence of spiritual struggles is not uncommon (Idler et al., 2003). Our measure of spiritual struggle, the NRC subscale of the Brief RCOPE, assesses thoughts and behaviors used to cope but does not assess the direct subjective experience of spiritual struggles (e.g., negative emotions, beliefs, internal or interpersonal conflict). In 2014, a new Religious and Spiritual Struggles (RSS) scale was published (Exline, Pargament, Grubbs, & Yali, 2014). This 26-item measure assesses six domains pertaining to the subjective experience of religious and spiritual struggles: struggle with God, struggle with religious others, the demonic, moral issues, doubt, and ultimate meaning. In contrast, five of the seven items on the NRC subscale of the Brief RCOPE assess struggle with God. The RSS scale may be a more sensitive and informative measure for future studies in this area. Our participants may also have been reluctant to endorse spiritual struggles, particularly because this was a religiously-integrated treatment study. This issue has been reported by others as well (Exline, Kaplan, & Grubbs, 2012).

A second reason for these unexpected findings may be the nature of the two treatments. We purposely made CCBT and RCBT as similar as possible, perhaps to the detriment of detecting differences between the two treatments. The CCBT intervention was more of a hybrid of CBT with an emphasis on positive psychology because it included sessions and exercises that explicitly focused on forgiveness, gratitude, altruism, benefit-finding, and generating hope. Each of these areas of focus also could have been categorized as spiritual values. In addition, the CCBT intervention included mindfulness, which can be understood and used as a spiritual practice. Clients may have integrated their religious beliefs into their use of mindfulness practice. Thus, our CCBT intervention may not have been as "secular" as we had intended nor as different from RCBT as traditional CBT. Religious/spiritual individuals completing exercises that developed or targeted spiritual values might not have needed explicit direction from the therapist to activate their own religious motivations for completing these exercises. Again, the only difference between RCBT and CCBT was the explicit use of clients' religious beliefs, practices, and teachings to facilitate cognitive restructuring and behavioral activation. This "activation without explicit prompting" also may help explain why religiosity did not moderate the effects of RCBT on spiritual struggles.

There may also have been certain "ingredients" in both treatments that resulted in a significant change in spiritual struggles over time, as we found. One likely ingredient was forgiveness. Interestingly, these results are consistent with findings from randomized clinical trials comparing religious with secular versions of a forgiveness intervention (Rye et al., 2002; Rye & Pargament, 2002). These researchers also speculated that their interventions were more similar than different. Their data also suggested that participants in the secular condition were using religious forgiveness strategies without being prompted to do so.

Although participants in RCBT and CCBT had a similar overall reduction in spiritual struggles and depression (Koenig et al., 2015), this reduction may have had different ramifications for other domains of well-being, particularly that of spirituality. Those with spiritual struggles receiving RCBT may have found a resolution for these struggles that resulted in a strengthening of their religious faith. In contrast, those with spiritual struggles who received CCBT may have found resolution to these struggles by changing their religious beliefs or decreasing the importance of their religious faith. Research has shown that spiritual struggles predict both spiritual decline and secular growth (Desai & Pargament, 2015). This suggests that spiritual struggles may have greater negative consequences for spiritual outcomes. A treatment like RCBT that addresses spiritual struggles might result in fewer negative consequences for a patient's spirituality. This is an interesting empirical question for future research.

Of particular relevance to clinical practice, spiritual struggles appeared to reduce the effectiveness of both RCBT and CCBT when treating major depressive disorder in the context of medical illness. This effect seemed to be slightly stronger among those who were less religious and contrasts other research that suggests spiritual struggles have the greatest negative impact on the psychopathology (including depression) of individuals for whom religion is a core component of their identity compared to individuals for whom religion is less important (Park, Brooks, & Sussman, 2009). Whether among the most religious or least religious, it appears that one possible factor maintaining depression may be unresolved spiritual struggles. That finding underscores the importance of assessing and addressing spiritual struggles in psychotherapy, whether using a secular or religious approach. Not to do so may result in negative consequences for our patients, as longitudinal research has found that spiritual struggles predict an increase in depression (Park et al., 2009), poor physical health, and increased mortality (Pargament et al., 2001). Recent research on the trajectory of spiritual struggles reveals that attributions about the meaning of the struggle, a high degree of integration/assimilation of religion into one's life, religious support, and positive religious coping predict both resolution of spiritual struggles and positive growth as a result of the struggle (Desai & Pargament, 2015). These factors should be included in both secular and religious approaches of assessment and treatment when spiritual struggles are present.

Finally, a key component of evidence-based practice (EBP) is client choice (American Psychological Association, 2006). This means that clients should have input into the selection of their treatment intervention based on their preferences and values. Indeed, psychotherapy is considered a "collaborative enterprise" (APA, 2006, p. 280) where the success of treatment is often contingent upon an involved and informed patient. CCBT is one of the more effective evidence-based interventions for a wide array of disorders, including depression. This study, along with previous research, demonstrated that RCBT is also an effective intervention for depression. This treatment finding, the importance of religion to many clients, and the central role of client choice in EBP, together suggest that clients should be offered the choice between CCBT and RCBT.

Limitations

Several limitations affect the generalizability and interpretation of results here. First, this randomized clinical trial was not powered to compare the effects of RCBT versus CCBT on spiritual struggles since the original intention was to compare the effects on depressive symptoms. Because the study was not powered as a non-inferiority trial, we cannot say for certain that RCBT and CCBT are equal in their ability to reduce spiritual struggles. Second, an inclusion criterion was that religion or spirituality was at least somewhat important for all participants; therefore, these results do not apply to those for whom religion or spirituality is not important. Note, however, that participants in this study were not particularly religious when compared to the U.S. population in general (56% of whom indicate that religion is "very important" in their lives [Pew, 2007] compared to the 47% of the present sample). Third, given that 89% of our sample was Christian, these results apply primarily to clients from this religious faith. Fourth, this report is one of several from this study comparing the effects of RCBT versus CCBT on mental health outcomes (depressive symptom, optimism, gratitude, etc.), and we did not correct our p value for multiple comparisons; thus, the findings here should be regarded as preliminary. Finally, we used mean score to replace missing values in our religiosity measure rather than using a more sophisticated statistical algorithm. Replacing the missing values with the mean was done to minimize the loss of subject data and utilize the information we had from responses to religious items answered that correlated highly with one another.

This study also has a number of strong points. First, this is the first study to examine the effects of an explicitly religious psychotherapy on spiritual struggles of clients with major depressive disorder in the context of chronic medical illness. Second, this was a multi-site clinical trial that employed manual-based interventions that were carefully followed by therapists in both treatment groups. Third, the measures used to assess client characteristics and treatment outcomes had solid psychometric characteristics established in previous studies, including the structured psychiatric interview (MINI) used to diagnose MDD and to determine eligibility, the measure of spiritual struggles (NRC subscale of the Brief RCOPE) that assessed the primary outcome, and the measure of depressive symptoms (the BDI-II). Finally, state-of-the-art mixed effects growth curve models were used to examine trajectories of change in spiritual struggles and depressive symptoms in response to the interventions.

Conclusions

In both the intention-to-treat and per-protocol analyses, religiously-integrated CBT was not more effective than conventional CBT in decreasing spiritual struggles during treatment for major depression in those with chronic medical illness. Furthermore, RCBT was not more effective than CCBT in those who were more religious at baseline or in those who had more spiritual struggles. Baseline spiritual struggles predicted a significantly slower decrease in depressive symptoms, regardless of treatment group. Since this study was not designed to detect a difference in treatment on spiritual struggles, future clinical trials with adequately powered sample sizes are needed to confirm these findings. For now, religiously-integrated CBT (at least the version used here) is not more effective than conventional CBT (at least the type that includes positive psychology exercises, spiritual values, and mindfulness meditation) in reducing spiritual struggles among clients with major depression in the context of chronic medical illness. Spiritual struggles, however, do appear to reduce the effectiveness of both RCBT and CCBT when treating major depressive disorder in this setting, and so should be a target for all interventions whether secular or religious.

References

Ai, A. L., Pargament, K., Kronfol, Z., Tice, T. N., & Appel, H. (2010). Pathways to postoperative hostility in cardiac patients mediation of coping, spiritual struggle and interleukin-6. Journal of Health Psychology 15(2), 186-195.

Ai, A. L., Seymour, E. M., Tice, T. N., Kronfol, Z., & Bolling, S. F. (2009). Spiritual struggle related to plasma interleukin-6 prior to cardiac surgery. Psychology of Religion and Spirituality 7 (2), 112-128.

American Psychological Association, Presidential Task Force on Evidence-Based Practice. (2006). Evidence-based practice in psychology. American Psychologist, 67(4), 271-285.

Azhar, M. Z., & Varma, S. L. (1995). Religious psychotherapy as management of bereavement. Acta Psychiatrica Scandinavica, 91, 233-235.

Azhar, M. Z., Varma, S. L., & Dharap, A. S. (1994). Religious psychotherapy in anxiety disorder patients. Acta Psychiatrica Scandinavica, 90,1-3.

Beck A. T., Rush J., Shaw B. F., & Emery G. (1979). Cognitive therapy of depression. New York, NY: Guilford Press.

Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561-571.

Berry, D. (2002). Does religious psychotherapy improve anxiety and depression in religious adults? A review of randomized controlled studies. International Journal of Psychiatric Nursing Research, 8, 875-890.

Charlson, M. E., Pompei, P., Ales, K. L., & Mackenzie, C. R. (1987). A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. Journal of Chronic Disease, 40(5), 373-383.

Desai, K. M., & Pargament, K. I. (2015). Predictors of growth and decline following spiritual struggles. The International Journal for the Psychology of Religion, 25, 42-56.

Exline, J. (2013). Religious and spiritual struggles. In K. I. Pargament, J. J. Exline, & J. W. Jones (Eds.), APA handbook of psychology, religion, and spirituality (Vol. 1: Context, theory, and research, pp. 459-475). Washington, DC: American Psychological Association.

Exline, J., Kaplan, K., & Grubbs, J. (2012). Anger, exit, and assertion: Do people see protest toward God as morally acceptable? Psychology of Religion and Spirituality, 4, 264-277.

Exline, J., Pargament, K., Grubbs, J., & Yali A. M. (2014). The religious and spiritual struggles scale: Development and initial validation. Psychology of Religion and Spirituality, 6(3), 208-222.

Hlatky, M. A., Boineau, R. E., Higginbotham, M. B., Lee, K. L., Mark, D. B., Califf, R. M., ... Pryor, D. B. (1989). A brief self-administered questionnaire to determine functional capacity (The Duke Activity Status Index). American Journal of Cardiology, 64, 651-654.

Hodge, D. R. (2006). Spiritually modified cognitive therapy: A review of the literature. Social Work, 51, 157-166.

Hoge, D. R. (1972). A validated intrinsic religious motivation scale. Journal for the Scientific Study of Religion, 11, 369-376.

Hook, J. N., Worthington, Jr., E. L., Davis, D. E., Jennings II, D. J., Gartner, A. L., & Hook, J. P. (2010). Empirically supported religious and spiritual therapies. Journal of Clinical Psychology, 66, 46-72.

Idler, E. L., Musick, M. A., Ellison, C. G., George, L. K., Krause, N., Ory, M. G., ... Williams, D. R. (2003). Measuring multiple dimensions of religion and spirituality for health research conceptual background and findings from the 1998 General Social Survey. Research on Aging 25(4), 327-365.

Koenig, H. G. (1996). An abbreviated Mini-Mental State Exam for medically ill elders. Journal of the American Geriatrics Society, 44, 215-216.

Koenig, H. G., Meador, K. G., & Parkerson, G. (1997). Religion index for psychiatric research. American Journal of Psychiatry, 154(6), 885-886.

Koenig, H. G., Pargament, K. I., & Nielsen, J. (1998). Religious coping and health status in medically ill hospitalized older adults. Journal of Nervous & Mental Disease, 186(9), 513-521.

Koenig, H. G., Pearce, M. J., Nelson, B., Shaw, S. F., Robins, C. J., Daher, N., ... King, M. B. (2015). Religious vs. conventional cognitive-behavioral therapy for major depression in persons with chronic medical illness. Journal of Nervous and Mental Disease, 203(4), 243-251.

Linn, B., Linn, M., & Gurel, L. (1968). Cumulative Illness Rating Scale. Journal of the American Geriatrics Society, 16, 622-626.

McConnell, K. M., Pargament, K. L, Ellison, C. G., & Flannelly, K. J. (2006). Examining the links between spiritual struggles and symp toms of psychopathology in a national sample. Journal of Clinical Psychology, 62(12), 1469-1484.

Murray-Swank, N. A., & Pargament, K. I. (2005). God, where are you?: Evaluating a spiritually-integrated intervention for sexual abuse. Mental Health, Religion & Culture, 8(3), 191-203.

Pargament, K. I. (1997). The psychology of religion and coping: Theory, research, practice. New York, NY: Guilford Press.

Pargament, K. I., Koenig, H. G., Tarakeshwar, N., & Hahn, J. (2001). Religious struggle as a predictor of mortality among medically ill elderly patients: A 2-year longitudinal study. Archives of Internal Medicine, 161(1S), 1881-1885.

Pargament, K. I., Smith, B. W., Koenig, H. G., Sc Perez, L. (1998). Patterns of positive and negative religious coping with major life stressors. Journal for the Scientific Study of Religion, 37(4), 710-724.

Park, C., Brooks, M., & Sussman, J. (2009). Dimensions of religion and spirituality in psychological adjustment of older adults living with congestive heart failure. In A. L. Ai Sc M. Ardelt (Eds.), Faith and well-being later in life (pp. 41-58). Hauppauge, NY: Nova Science.

Pearce, M. J., Koenig, H. G., Robins, C. J., Nelson, B., Shaw, S. F., Cohen, H. J., Sc King, M. B. (2014). Religiously-integrated cognitive behavioral therapy: A new method of treatment for major depression in patients with chronic medical illness. Psychotherapy. Advance online publication, http://dx.doi.org/10.1037/a0036448

Pew Forum (2007). U.S. religious landscape survey: Beliefs and practices: Importance of religion. Retrieved from http://religions. pewforum.org/ maps

Phillips, R. E., Colvin, S., Hietbrink, L., Sc Vonnegut, E. (2012). A manualfor a scale of Buddhist coping: The BCOPE. St. Joseph, MO: Missouri Western State University.

Propst, L. R. (1980). The comparative efficacy of religious and nonreligious imagery for the treatment of mild depression in religious individuals. Cognitive Therapy and Research, 4, 167-178.

Propst, L. R., Ostrom, R., Watkins, P., Dean, T., Sc Mashburn, D. (1992). Comparative efficacy of religious and nonreligious cognitive-behavior therapy for the treatment of clinical depression in religious individuals. Journal of Consulting and Clinical Psychology, 60, 94-103.

Razali, S. M., Hasanah, C. L, Aminah, K., Sc Subramaniam, M. (1998). Religious-sociocultural psychotherapy in patients with anxiety and depression. Australian & New Zealand Journal of Psychiatry, 32, 867-872.

Rosmarin, D. H., Pargament, K. L, & Flannelly, K. J. (2009). Do spiritual struggles predict poorer physical/mental health among Jews? International Journal for the Psychology of Religion, 19(4), 244-258.

Rye, M. S., Sc Pargament, K. I. (2002). Forgiveness and romantic relationships in college: Can it heal the wounded heart? Journal of Clinical Psychology, 58,419-441.

Rye, M. S., Pargament, K. I., Pan, W., Yingling, D. W., Shorgren, K. A., & Ito, M. (2002). Can group interventions facilitate forgiveness of an ex-spouse? A randomized clinical trial. Journal of Consulting and Clinical Psychology, 73(5), 880-892.

Serfaty, M. A., Haworth, D., Blanchard, M., Buszewicz, M., Murad, S., &c King, M. (2009). Clinical effectiveness of individual cognitive behavioral therapy for depressed older people in primary care: A randomized controlled trial. Archives of General Psychiatry, 66(12), 1332-1340.

Sheehan, B. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., ... Dunbar, G. C. (1998). The Mini International Neuropsychiatrie Interview (MINI): The development and validation of structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59(Suppl 20), 22-33.

Smith, T. B., Bartz, J., & Richards, P. S. (2007). Outcomes of religious and spiritual adaptations in psychotherapy: A meta-analytic review. Psychotherapy Research, 17, 643-655.

Tan, S. Y., & Johnson, W. B. (2005). Spiritually-oriented cognitive-behavioral-therapy. In L. Sperry Sc E. P. Shafranske (Eds.), Spiritually-oriented psychotherapy (pp. 77-103). Washington, DC: American Psychological Association.

Trevino, K. M., Balboni, M., Zollfrank, A., Balboni, T., Sc Prigerson, H. G. (2014). Negative religious coping as a correlate of suicidal ideation in patients with advanced cancer. Psycho-Oncology, 23(8), 936-945.

Underwood, L. G., & Teresi, J. A. (2002). The Daily Spiritual Experiences Scale: Development, theoretical description, reliability, exploratory factor analysis, and preliminary construct validity using health-related data. Annals of Behavioral Medicine, 24,22-33.

Vallis, T. M., Shaw, B. G., Sc Dobson, K. S. (1986). The Cognitive Therapy Scale: Psychometric properties. Journal of Consulting and Clinical Psychology, 54, 381-385.

Wade, N. G., Worthington, Jr., E. L., Sc Vogel, D. L. (2007). Effectiveness of religiously tailored interventions in Christian therapy. Psychotherapy Research, 17, 91-105.

Waltz, J., Addis, M. E., Koerner, K., & Jacobson, N. S. (1993). Testing the integrity of a psychotherapy protocol: Assessment of adherence and competence. Journal of Consulting and Clinical Psychology, 61,620-630.

Worthington, Jr., E. L., Hook, J. N., Davis, D. E., & McDaniel, M. A. (2011). Religion and spirituality .Journal of Clinical Psychology: In Session, 67, 204-214.

Worthington, Jr., E. L., & Sandage, S. J. (2001). Religion and spirituality. Psychotherapy, 38, 473-477.

Michelle J. Pearce

University of Maryland (School of Medicine)

Duke University Medical Center (Department of Psychiatry and Behavioral Sciences)

Duke University (Center for Spirituality, Theology and Health)

Harold G. Koenig

Duke University Medical Center (Department of Psychiatry and Behavioral Sciences)

Duke University (Center for Spirituality, Theology and Health)

Duke University Medical Center (Department of Medicine)

King Abdulaziz University (Department of Medicine)

Ningxia Medical University (School of Public Health)

Author Note: Funding support provided by the John Templeton Foundation. None of the authors have a conflict of interest. Correspondence concerning this article should be addressed to Michelle J. Pearce, Department of Family and Community Medicine, Center for Integrative Medicine, University of Maryland School of Medicine, 520 W. Lombard Street, East Hall, Baltimore, MD 21201. Email: mpearce@som.umaryland.edu

Author Information

PEARCE, MICHELLE J. PhD. Address: Department of Family and Community Medicine, Center for Integrative Medicine, University of Maryland School of Medicine, 520 W. Lombard Street, East Hall, Baltimore, MD 21201. Title: Assistant Professor. Degree: PhD, Yale University. Specializations: Spiritual issues in psychotherapy, cognitive behavioral therapy.

KOENIG, HAROLD G. MD. MHSc. Address: Center for Spirituality, Theology, and Health, Duke University Medical Center, Box 3400, Busse Building, Suite 0505, Durham, North Carolina 27710. Title: Professor, Psychiatry Sc Behavioral Sciences; Associate Professor, Medicine; Director, Center for Spirituality, Theology and Health; Adjunct Professor, Department of Medicine, King Abdulaziz University; Adjunct Professor, Department of Public Health, Ningxia Medical University. Degree: M.D. University of California at San Francisco; M.H.Sc. Duke University; R.N. San Joaquin Delta College; B.S. Stanford University. Specializations: Psychiatry (general).

TABLE 1
Baseline Responses Regarding Spiritual Struggles (n = 124) (1)

                                       Current        General
                                       Sample      Social Survey

Spiritual Struggle Responses          %     (n)     %      (n)

1. Felt abandoned by God
   --not at all (very untrue)        53.2   (66)   87.6   (1264)
   --somewhat                        29.8   (37)   9.5    (137)
   --quite a bit                     13.7   (17)   1.7     (25)
   --a great deal (very true)        3.2    (4)    1.2     (17)

2. Felt punished by God
   --not at all (very untrue)        73.4   (91)   76.8   (1110)
   --somewhat                        21.8   (27)   17.4   (251)
   --quite a bit                     3.2    (4)    3.8     (54)
   --a great deal (very true)        1.6    (2)    2.0     (29)

3. Wondered why God punishing me
   --not at all (very untrue)        62.1   (77)    --      --
   --somewhat                        32.3   (40)    --      --
   --quite a bit                     2.4    (3)     --      --
   --a great deal (very true)        3.2    (4)     --      --

4. Questioned God's love for me
   --not at all (very untrue)        72.6   (90)    --      --
   --somewhat                        20.2   (25)    --      --
   --quite a bit                     2.4    (3)     --      --
   --a great deal (very true)        4.8    (6)     --      --

5. Questioned power of God
   --not at all (very untrue)        71.8   (89)    --      --
   --somewhat                        22.6   (28)    --      --
   --quite a bit                     2.4    (3)     --      --
   --a great deal (very true)        3.2    (4)     --      --

6. Felt church had abandoned them
   --not at all (very untrue)        71.0   (87)    --      --
   --somewhat                        23.4   (37)    --      --
   --quite a bit                     4.8    (6)     --      --
   --a great deal (very true)        4.8    (6)     --      --

7. Decided Devil made this happen
   --not at all (very untrue)        70.2   (87)    --      --
   --somewhat                        23.4   (29)    --      --
   --quite a bit                     4.0    (5)     -       --
   --a great deal (very true)        2.4    (3)     --      --

Note. (1) Buddhists {n = 8) excluded since asked different questions.

TABLE 2
Average Spiritual Struggle (SS) Scores by Treatment Group

                                CCBT         RCBT        t
                              Mean (SD)    Mean (SD)   value    P
Intention-to-Treat Analysis

All participants
SS at baseline (n = 67/65)   3.37 (3.29)  3.32 (3.35)   0.10  0.924
SS at 12 wk (n = 47/44)      2.77 (2.86)  2.70 (3.25)   0.10  0.924
SS at 24 wk {n = 31 /32)     2.35 (2.06)  2.68 (3.55   -0.44  0.660

High religious participants
SS at baseline In = 24/20)   3.47 (3.45)  2.15 (3.31)   1.29  0.205
SS at 12 wk (n = 18/15)      2.28 (2.47)  1.73 (1.87)   0.70  0.488
SS at 24 wk (n = 13/12)      2.69 (2.14)  2.03 (2.30)   0.75  0.461

Per-Protocol Analysis

All participants
SS at baseline (n = 46/45)   3.83 (3.12)  3.31 (3.43)   0.75  0.455
SS at 12 wk (n = 44/40)      2.89 (2.92)  2.78 (3.31)   0.16  0.870
SS at 24 wk (n = 28/28)      2.46 (2.13)  2.85 (3.71)  -0.47  0.640

High religious participants
SS at baseline (n = 16/17)   3.90 (3.66)  2.18 (3.49)   1.38  0.177
SS at 12 wk (n = 16/15)      2.44 (2.58)  1.73 (1.87)   0.86  0.394
SS at 24 wk (n = 11/12)      2.91 (2.26)  2.03 (2.30)   0.93  0.365

Note. CCBT = conventional cognitive behavioral therapy; RCBT =
religiously-integrated cognitive behavioral therapy; wk = weeks;
SD = standard deviation; p = significance level (Student's t-test).

TABLE 3
Effect of RCBT Versus CCBT on Trajectory of Change in Spiritual
Struggles from Baseline Through 24 Weeks

                                B      SE      t     P (1)

Intention-to-Treat Analysis

All Participants (n = 130)
  Time                        -0.23   0.22   -1.05   0.003
  Main effect of group         0.51   0.73    0.70   0.485
  Time x group interaction    -0.48   0.31   -1.53   0.127

Highly Religious (n = 44)
  Time                        -0.21   0.37   -0.58   0.103
  Main effect of group         1.51   1.14    1.32   0.194
  Time x group interaction    -0.41   0.50   -0.81   0.421

Low Religious (n = 86)
  Time                        -0.22   0.28   -0.77   0.016
  Main effect of group         0.10   0.93    0.10   0.918
  Time x group interaction    -0.56   0.41   -1.38   0.170

Per-protocol Analysis

All Participants (n = 91)
  Time                        -0.25   0.25   -1.03   0.002
  Main effect of group         1.07   0.85    1.25   0.212
  Time x group interaction    -0.59   0.35   -1.71   0.090

Highly Religious (n = 33)
  Time                        -0.22   0.38   -0.57   0.078
  Main effect of group         1.96   1.32    1.48   0.145
  Time x group interaction    -0.54   0.54   -0.99   0.328

Low Religious (n--58)
  Time                        -0.26   0.33   -0.80   0.013
  Main effect of group         0.34   1.11    0.48   0.630
  Time x group interaction    -0.63   0.45   -1.39   0.168

Note. CCBT = conventional cognitive behavioral therapy;
RCBT = religiously-integrated cognitive behavioral therapy;
B = unstandardized coefficient for "time" indicates change
in spiritual struggles during the course of therapy
(independent of group); "main effect of group" = the average
difference between treatment groups (RCBT = 1, CCBT = 0);
"group x time interaction" = whether the two groups changed
at the same rate; B's are from mixed effects growth curve
models; SE = standard error; p = significance level
(based on Type 3 tests of fixed effects).