Do Religion and Spirituality Buffer the Effect of Childhood Trauma on Depressive Symptoms? Examination of a South Asian Cohort from the USA

Abstract

Asian Americans have been identified as a racial group that is disproportionately affected by childhood trauma. The goal of this study was to assess if religion/spirituality moderate the effects of childhood trauma on adult depressive symptoms among a sample of South Asians in the USA. Our analysis drew from the study on stress, spirituality, and health (SSSH) questionnaire fielded in the Mediators of Atherosclerosis in South Asians Living in America (MASALA) study (n = 990) during 2016–2018. A series of regression models with multiplicative interaction terms were conducted. Emotional neglect, emotional abuse, and physical neglect were associated with higher depressive symptoms. Higher religious attendance and negative religious coping techniques were found to exacerbate this relationship. There were two findings conditional on gender. Among men, gratitude and positive religious coping also exacerbated the relationship between childhood trauma and depressive symptoms. Negative religious coping also exacerbated the association between childhood trauma and depressive symptoms for women. This is the first community-based study of US South Asians to consider the association between various forms of childhood trauma and depressive symptom outcomes. South Asians remain an understudied group in the religion and health literature, and this study sheds light on the important differences in the function and effectiveness of religion/spirituality for those faced with early life trauma.

Introduction

Rates of childhood trauma (CT) in the USA are strikingly high, with approximately 50% of adults reporting at least one adverse childhood experience (e.g., neglect and physical abuse) before the age of 18 (Gonzalez et al., 2021). The experience of childhood trauma, however, is not equally distributed in the American populace. Indeed, several studies indicate that people from marginalized or minority communities are disproportionately likely to be exposed to adverse childhood experiences (Assari, 2020). While elevated rates of adverse childhood experiences have been documented among racial groups such as African Americans (Kang & Burton, 2014) followed by Hispanic Americans (Larson et al., 2017), Asian Americans have also been identified as a group that is disproportionately more affected by childhood trauma (Kim et al., 2007; Maker et al., 2005). However, Asian Americans remain a largely understudied population in this domain. Within the broader Asian American population, almost three-quarters of South Asian and Middle Eastern women have experienced at least one form of childhood abuse by the age of 16 (Maker et al., 2005). Moreover, almost half of South Asians in the USA reported having witnessed parental violence as children and approximately one quarter had experienced sexual abuse as a child (Kim et al., 2007).

The high prevalence of childhood trauma among South Asians in the USA raises several broader concerns. First, domestic violence agency staff charged with serving South Asian communities report that perceived stigma may prevent victims of trauma from seeking help and disclosing their abuse (Ragavan et al., 2018). Second, it is well-established that the effects of childhood trauma can leave an indelible imprint over the life course, found to be associated with a range of higher risk of several health conditions at midlife and beyond, including greater depression (Negele et al., 2015), suicidal ideation (Beristianos et al., 2016), and physical health and the onset of chronic disease (Ferraro et al., 2016; Hayward & Gorman, 2004). Such pernicious associations between childhood trauma and lower mental well-being have also been documented among Asian Americans in particular (Hahm et al., 2012; Robertson et al., 2016). Despite this scattering of studies, however, there is a relative dearth of research between multiple forms of childhood trauma and depression in the South Asian population in the USA, which is one gap this study proposes to fill.

Though more work is needed to comprehend the behavioral and biological mechanisms through which adverse childhood experiences impact health, scholars have acknowledged that the effects of childhood trauma could be buffered or mitigated by resources present in adulthood (Brewer-Smyth & Koenig, 2014; Southwick & Charney, 2012). Religion/spirituality (R/S) has been proposed as one mechanism survivors of childhood abuse may use to cope with their traumatic experiences (Jung, 2018; Manning & Miles, 2018; Schafer, 2014; Upenieks, 2021). In general, higher levels of both public and private religiosity tend to mitigate some of the deleterious associations between childhood trauma and mental health in the American population (Jung, 2018; Schafer, 2014; Upenieks, 2021).

Past studies on the use of religious/spiritual coping against childhood trauma, while informative, are all drawn from representative samples of the American population. No existing study has examined such a relationship among South Asian Americans. There are reasons to expect that the religious/spiritual context of South Asians may operate differently than it does for other cultural groups in America, given the generally lower rates of religious service attendance and the higher relative importance of spiritual experiences (Kent et al., 2020; Stroope et al., 2020).

The goal of the present study is to understand if R/S moderates the effects of CT on adult depressive symptoms among a sample of South Asians in the USA and whether there are gender differences in these associations. Drawing on the stress process framework, data were analyzed from the Mediators of Atherosclerosis in South Asians Living in America (MASALA) study (Kanaya et al., 2013). Altogether, this study presents a unique opportunity to study whether a different religious/spiritual context among the South Asian population can be a protective source for reducing mental health outcomes that tend to occur in the aftermath of childhood trauma.

Literature Review

The Stress Process Framework: Childhood Trauma and Religious Resources

The stress process model is a conceptual model within the sociology of mental health that analyzes processes through which stressors exert noxious effects on mental health, and how individuals respond to these stressors based on the social and personal resources available to them (Pearlin, 1989). The stress process model highlights the role of psychosocial resources as moderators. Religiosity has been posited as a resource within the stress process model (Ellison & Henderson, 2011), as it provides cognitive, personal, and social resources that are important in the aftermath of stress. To date, some studies have found that religion exacerbates the harmful effects of stress on mental health, but a growing body of work demonstrates the stress-buffering effects of religion (Schieman et al., 2013).

Before laying out how religiosity may factor into the association between childhood trauma and mental health, we outline the mechanisms by which childhood trauma is proposed to have negative consequences for adult mental health. Childhood trauma is thought to leave an enduring mark on the victim over the life course, interrupting many aspects of adult life including mental health (Greenfield & Marks, 2010). Childhood trauma is known to affect biological systems during the crucial childhood development period, which can have long-standing detrimental effects on mental health over the life span (Horwitz et al., 2001). In addition, childhood trauma can strain family relationships, damage a successful transition to adulthood, which can also jeopardize mental health (Schilling et al., 2008), and increase the risk of subsequent stressors, such as divorce or unemployment (Horwitz et al., 2001). Given these pernicious effects, scholars have called for more work on processes that can blunt the effects of these accumulated risks (Schafer et al., 2011), and the stress process model provides a rich theoretical context for doing so. As we outline below, religion is a multidimensional construct, so we examine how both private and organizational forms of religion may moderate the association between childhood adversity and adult mental health.

Religion/Spirituality in the South Asian Context

Thousands of studies have been conducted on religion/spirituality and health and generally show a positive relationship between religion/spirituality and mental well-being (see Koenig et al., 2012; Page et al., 2020; Schieman et al., 2013 for reviews). However, a significant proportion of these studies typically uses population samples that consist of majority Christian and White respondents. We therefore have a limited understanding of the religious/spiritual practices of other racial groups. To date, a few studies have examined the relationships between religion/spirituality and mental and physical well-being using the MASALA data (Kent et al., 2020; Stroope et al., 2020). Overall, the pattern of findings in this South Asian population shares commonalities with previous research on R/S and health, but also have distinct characteristics.

Before reviewing the findings, scholars have tended to characterize religious involvement among South Asians as relatively communal or family-based (Kandula et al., 2018). Religious organizations in Dharmic faiths (e.g., Hinduism, Judaism, Sikhism, and Buddhism) are not typically congregation-based in South Asia but have shifted toward a congregation model in the USA (Yang & Ebaugh, 2001). Religious/spiritual organizations were the most common form of organizational affiliation for the US South Asians (Min, 2010) which is especially important because three-quarters of this groups are first-generation immigrants and could benefit from the support of a religious group. Islam was already a religion more centered around congregational worship, as roughly half of Muslims report attending mosques or religious centers on a weekly basis (Sciupac, 2017).

To date, the findings on South Asian reflect this uniqueness of this religious context among this group. For instance, Stroope and colleagues (2020) found that group prayer was associated with more favorable self-rated health and mental health, and lower anxiety and anger. Moreover, providing and receiving love and care in the religious congregation was linked to better self-rated health and mental health and lower anxiety. On the other hand, experiences of criticism from congregation members were associated with greater anxiety and anger. As Kent (2020) notes, most of the findings mirror patterns found in studies of other ethnic groups, which largely indicate that religious/spiritual practices are beneficial for health and that congregations are useful for forming friendships and social support, reinforcing cultural norms and beliefs, and experiencing the transcendent with like-minded others.

However, Stroope and colleagues (2020) also documented patterns that go against the grain of most R/S research conducted in the USA. For instance, religious service attendance was associated with higher levels of anxiety, whereas it is typically associated with lower anxiety in other groups. The authors propose a process of “resource mobilization” to explain this finding: when individuals experience stress, they turn to sources of support to find assistance, so increased anxiety could lead to increased religious attendance as a form of coping. These authors also note that Hindus, a majority of the MASALA sample, attend religious services at a lower level compared to members of other religious traditions. Indeed, Hinduism does not emphasize regular temple visitation, so greater anxiety levels are possible due to lower levels of baseline service attendance. Building on this research, Stroope and colleagues (2022) proposed that the positive association between greater attendance and anxiety might be explained by the fact that “South Asians may be more motivated to engage in religious attendance by extrinsic factors [e.g., family or cultural obligation], and thus be more subject to anxiety-inducing dimensions of attendance” (pg.10). Furthermore, Stroope and colleagues (2023) found that congregational neglect, a form of negative communal religious experience where people felt ignored or neglected by members of their congregation, explained roughly one-third of the positive association between religious attendance and anxiety. These authors acknowledged that high levels of cultural diversity within the US South Asian communities might make congregational neglect more likely to occur, and that the external factors that may prompt religious attendance could lead to susceptibility to anxiety-producing aspects of congregational experiences like neglect.

Another study with the MASALA data by Kent and colleagues (2020) focused on private religious beliefs and practices. Among South Asians, yoga, gratitude, non-theistic spiritual experiences (e.g., “I experience a connection to all of life,” “I am touched by the beauty of creation”), closeness to God, and positive religious coping were all correlated with better self-rated health. With the exception of yoga, all of these private measures of religiosity were also associated with better mental health, and negative religious coping to poorer mental health. There is some research to suggest that Hindus and Muslims engage in negative religious coping (e.g., feeling anger at God, feeling punished by God) less frequently than Christian samples (Tarakeshwar et al., 2003), perhaps because of a lower identification with the human-divine relationship in these faiths compared with religiosity. Despite the less frequent use of negative religious coping, struggles with faith have been linked to greater depression in both sample of Muslims (Abu-Raiya et al., 2008) and Hindus (Abu-Raiya & Pargament, 2015), suggesting that engaging in negative religious coping would also be accompanied by deleterious consequences among the US South Asians. Gratitude and non-theistic experiences were also associated with lower anxiety.

Taken together, the small body of existing research on the religious/spiritual lives of South Asians suggests several similarities to the wealth of existing literature in the USA, but striking differences as well, especially with regards to religious attendance. Having identified these key points of similarity or difference, we next outline how these patterns of R/S and health outcomes may affect the ability of R/S to moderate the association between childhood trauma and depressive symptoms drawing on insights from the stress process framework.

Childhood Trauma, Private Religiosity, and Mental Health

To our knowledge, no previous study has examined whether private forms of religiosity (e.g., gratitude and feelings of closeness to God, non-theistic spiritual experiences) moderate the relationship between childhood trauma and mental health in South Asian Americans. We therefore draw from the stress process framework and existing studies which have looked at R/S among survivors of trauma in both US and non-US cultures to form our theoretical basis.

Before addressing the potential stress-buffering role of private religiosity, let us first consider whether religiosity may be prone to change after adverse childhood events. Previous research has found that childhood trauma might serve as a religious/spiritual turning point in national samples of Americans (Bierman, 2005; Schafer, 2014), either drawing individuals to a deeper religious/spiritual life or precipitating a decline in R/S belief or practice. Bierman (2005) argues that victims of childhood abuse might project a negative characteristic of their abuser onto God or a divine being, but this typically leads individuals to withdraw only from the institutional aspects of religion/spirituality. Another body of evidence, however, suggests that victims of childhood abuse tend to experience increases in the spiritual domain over the life course, in both spiritual openness among a sample of 763 college students (Dyslin & Thomsen, 2011) and spirituality (Ryan, 1998). In non-US cultures, a similar pattern of religious/spiritual increases have been documented. One study among a sample of 137 adolescents from India found that those who experienced childhood adversities tended to report greater adolescent religiosity (Santoro et al., 2016). Similarly, among a sample of 200 adults in the Czech Republic, there was an association found between childhood trauma and higher spirituality (Kosarkova et al., 2020).

Considered as a whole, the weight of existing evidence suggests that private forms of R/S may be more prevalent among those who have survived childhood trauma. However, it remains to be seen whether this increase promotes more positive or negative mental health outcomes. The vast majority of South Asians adhere to Dharmic (Hinduism, Sikhism, Jainism, and Buddhism) faiths, which espouse beliefs regarding inner peace, connection to nature, and the emptying of the self (e.g., through Yoga). Private forms of R/S are also important components of Muslim spirituality (Winchester, 2008) and Christian spirituality (Sharp, 2010). There are thus reasons to expect that private religiosity may either buffer or exacerbate the pernicious mental health effects of childhood trauma.

Taken from the perspective of a buffering effect, private religiosity can provide access to resources that can promote an adaptive approach to life. Feelings of gratitude toward God, for instance, have been shown to be associated with higher well-being in the aftermath of stressful life experiences among a nationally representative sample of older adults (Upenieks & Ford-Robertson, 2022). R/S outlooks, such as feeling connected to life or appreciating the beauty of creation, can also provide a deeper sense of meaning and purpose that can render past hardships more interpretable. Finally, feelings of closeness or attachment to God have also been previously found to act as stress buffers among national samples of Americans (Ellison et al., 2012; Upenieks, 2022). These forms of spiritual experiences can help to reframe adversities in positive ways, such as seeing negative events as opportunities for personal growth (Pargament & Park, 1995) or feeling comforted by the love of a divine presence. R/S has also been linked to other beneficial coping strategies, known as positive religious coping (Pargament et al., 2000), which can facilitate active problem-solving and using resources within the spiritual realm as a source of comfort and strength. In a large sample of Seventh-Day Adventist adults in the USA, private religiosity, positive coping, forgiveness, and gratitude were all associated with better general mental health (Reinert et al., 2016). Specific to childhood trauma, a study from the USA found that positive religious coping buffered the negative mental health impacts of childhood trauma (Upenieks, 2021). If South Asian victims of abuse engaged in these more positive forms of R/S, then we would predict R/S to play a buffering role and function as a resource that protects mental health.

From the perspective of an exacerbating effect, however, it is possible that childhood trauma may promote negative forms of R/S coping and perhaps offer no protection or even worsen the mental health consequences of childhood trauma (e.g., Gall, 2006; Mosqueiro et al., 2021). Negative forms of coping could include doubting God or having unresolved spiritual struggles. For instance, questioning the nature and goodness of God or feeling angry or betrayed by God for allowing such negative events to occur are known to be associated with lower mental well-being (Ellison et al., 2013; Hill et al., 2021). It is possible that victims of childhood trauma could project negative feelings that they harbor toward their abuser onto God (e.g., Bierman, 2005), perhaps leading them to doubt God or otherwise abandon spiritual experiences (Waldron et al., 2018). Research has also documented the stress-exacerbating effects of R/S struggles among individuals dealing with illness, as the illness experience may lead patients to see God as apathetic or unfair and by shaking core spiritual assumptions central to one’s orienting system (McConnell et al., 2006). If this is the case among South Asians with a history of childhood trauma, we would expect that engagement in negative religious coping would predict worse mental health outcomes in adulthood.

Childhood Trauma, Organizational Religious Involvement, and Mental Health

In this study, we also consider one form of organizational religious involvement—public religious attendance at worship services—as a moderator of childhood trauma on mental health. As with private religiosity, we see reasons within the existing literature to expect either a stress-buffering role or a stress-exacerbating (or null moderation) pattern with respect to religious attendance for the mental health of childhood abuse victims. Again, we draw from research from both US and non-US samples to outline these competing approaches.

The first approach suggests that religious attendance should play a protective role against depressive symptoms in victims of childhood trauma. For instance, a comprehensive review of the literature on spirituality, childhood trauma, and resilience posits that religious attendance (and the social support it engenders) protects against the negative mental health consequences that typically results from childhood trauma (Brewer-Smyth & Koenig, 2014; see also Walker et al., 2009). Participation in a religious congregation can act as a valuable source of social support that empower people to manage both past and present negative life events (Jung, 2018; VanderWeele, 2017). Additionally, religious participation tends to predict higher psychosocial resources such as self-esteem and mastery, which typically buffer the effects of stressors (e.g., George et al., 2002; Schieman, 2008). A recent study by Jung (2018), drawing from a nationally representative sample of midlife Americans, found that the victims of childhood abuse who attended church weekly in adulthood experienced better mental health than their counterparts who attended less regularly. Members of religious congregations tend to have a shared set of narratives surrounding religious suffering (Krause, 2006); thus, if co-congregants have similar adverse childhood experiences to share, this could represent an important source of socioemotional support in the form of healing words or consolation. Taken together, these resources found within religious communities may have psychological benefits for victims of childhood trauma.

However, as we noted earlier, expectations of regular (weekly) religious attendance are not typically emphasized in South Asian religions. It is possible that religious attendance might not have strong potential to act as a stress buffer for victims of childhood trauma. To begin with, despite some evidence to the contrary, religious attendance has not been identified as a robust stress buffer for victims of early life abuse. For instance, two studies using longitudinal data and capturing a national sample of midlife people in the USA and beyond found that religious attendance did not buffer the association between childhood abuse and adulthood mental health problems (Manning & Miles, 2018; Upenieks, 2021). In addition to this potential for a null moderation pattern, it has also been suggested that some South Asians may only seek out formal religious participation under conditions of duress (Kent et al., 2020). Thus, if victims of childhood trauma who are experiencing higher levels of depressive symptoms “select” into formal religious participation, this may obscure any stress-buffering functions that religious participation could serve. Because so little is known about the potential buffering role of communal religious participation in the South Asian population, we leave open the possibility that religious attendance could play a stress-buffering, stress-exacerbating, or null association with mental health when combined with experiences of childhood trauma.

Gender Differences in the Effects of Childhood Trauma and R/S for Mental Health

As a final goal of the current study, we assess whether there are gender differences in the processes outlined above. As noted by Reinert and colleagues (2016), studies examining gender differences of R/S in coping with trauma are scarce. In the context of our study, estimates suggest that roughly 75% of South Asian American women have experienced one form of childhood trauma, (Maker et al., 2005), perhaps because of high acceptance for physical discipline in this culture (Hong & Hong, 1991), patriarchal conventions, and traditional gender-role stereotypes (Lee & Zane, 1998).

These gender disparities in the prevalence of childhood trauma alone make these differences worthy of further consideration. However, a few existing studies have also found R/S differences in the South Asian culture. For instance, the evidence suggests that South Asian women tend to attend religious services less (Loewenthal et al., 2002) but are more spiritual (Loewenthal et al., 2002; Piedmont & Leach, 2002) than their male counterparts. This is in contrast to studies from the USA focused on non-South Asian groups, which have found women to be more religious than men on almost every indicator of religiosity, including private prayer (Maselko & Kubzansky, 2006) and more frequent church attendance (Schnabel, 2018). Several studies have also assessed whether there are gender differences in the effect of religiosity on well-being. Overall, women tend to benefit more from religious participation than men (Bonhag & Upenieks, 2021; Sharp et al., 2016; Taylor et al., 2005), though at least one study has found that men derived greater benefits from religious attendance compared to women (Maselko & Kubzansky, 2006; McFarland, 2010). Stronger feelings of closeness to God also tend to benefit women more, but negative religious coping (viewing God as harsh and judgmental) also tends to undermine the mental health of women more than men (Bonhag & Upenieks, 2021).

More germane to the current study, a few studies have also found that men and women may use religion/spirituality differently to cope with trauma. Research by Zukerman and colleagues (2017) on a sample of Jewish students found that negative religious coping after traumatic stress had a stronger association with psychological distress among men than women. A similar finding was observed by Reinert and colleagues (2016) in a sample of Seventh-Day Adventists in the USA, as women survivors of early trauma with higher negative religious coping scores had significantly worse physical health than their male counterparts. These existing studies, while useful in laying a base for understanding gender differences in the relationship between R/S and health, were based on either small or non-inclusive samples with respect to age, and race/ethnicity, which limit the conclusions that could be drawn. Moreover, no existing study, to our knowledge, has examined gender differences in R/S coping and mental health for victims of childhood trauma in the South Asian population. Therefore, in addition to conducting analyses on the full MASALA sample, several analyses below display results on gender-stratified samples to assess any potential differences in how the process of R/S moderation may unfold.

Methods

This analysis was drawn from the study on stress, spirituality, and health (SSSH) questionnaire (Kent et al., 2021; Warner et al., 2021) fielded in the Mediators of Atherosclerosis in South Asians Living in America (MASALA) study (Kanaya et al., 2013). MASALA is the largest study of health and well-being among the US South Asians, drawing its sample from the San Francisco Bay and Greater Chicagoland areas. In order to join, participants must have met the following criteria: age 40–84 years at the baseline examination, of South Asian origin (three of four grand-parents must have been born in India, Pakistan, Nepal, Bangladesh, or Sri Lanka), and free of cardiovascular disease. Recruitment consisted of mailing random batches of 100 letters every 2–4 weeks from an initial list of 10,000 South Asian house-holds in the study regions, along with follow-up telephone calls. Initial data collection occurred from 2010 to 2013 (n = 906), with a second set of participants added from 2017 to 2018 (n = 258). The SSSH questionnaire was circulated to all study participants between 2016 and 2018. In all, 990 MASALA participants completed the SSSH questionnaire. Participants were excluded from the present analysis in several cases, including where religious affiliation was missing (n = 1), the dependent variable was missing (n = 2), all three independent variable subscales were missing (n = 45), and where participants indicated: (a) being atheist or (b) where they responded to the statement “I believe God exists” by selecting “not true at all” (n = 84). Several covariates had some missing data, but after imputation (see Analytic Strategy) the sample consisted of 858 participants.

Dependent Variable

Depressive symptoms was measured via the 20-item Center for Epidemiologic Studies—Depression Scale (CES-D) (Radloff, 1977), which asked questions about how participants felt during the previous week. Sample items included “I felt sad,” “I felt everything I did was an effort,” and “I felt people disliked me.” Response options ranged from “almost never” to “almost always” and were additive on a scale of 0–3 (possible range 0–60).

Independent Variables

Childhood trauma questionnaire (CTQ) is a screening tool measuring histories of abuse and neglect in the home during childhood. MASALA collected three (out of five) CTQ subscales: emotional abuse (e.g., “my family said hurtful or insulting things to me”), emotional neglect (e.g., “my family looked out for each other”—reverse coded), and physical neglect (e.g., “there was not enough to eat”). Each subscale consisted of five items scored from “never true” (= 1) to “very often true” (= 5). Scores were summed for each subscale as well as a total score.

Religion and spirituality (R/S) were assessed with several items from the SSSH questionnaire. Questions were preceded with the following statement: “These questions are being asked of people from different religious backgrounds, and although we use the term ‘God’ in some of the questions below, please substitute your own word for ‘God’ (for example, Bhagwan, All, the Divine, etc.).” In this analysis, we considered questions that could potentially act as moderators between the experience of childhood trauma and severity of adult depressive symptoms. Religious service attendance was coded “never” = 0 to “several times per week” = 6. Positive religious coping was an average of 8 items (e.g., “I saw my situation as part of God’s plan”), and negative religious coping was an average of six items (e.g., “I wondered what I did for God to punish me”) (Pargament et al., 2000). Responses ranged from “not at all” = 1 to “a great deal” = 4. Gratitude was an average of two items (“I have so much in life to be thankful for” and “If I had to list everything I felt thankful for, it would be a very long list”) (McCullough et al., 2002). Response options ranged from “strongly disagree” = 1 to “strongly agree” = 5. Non-theistic daily spiritual experiences (NTDSE) consisted of an average of four items (e.g., “I experience a connection to all of life”) and theistic daily spiritual experiences (TDSE) consisted of an average of two items (e.g., “I desire to be closer to God”). Response options ranged from “never” = 0 to “many times a day” = 5 for NTDSE and from “definitely not true of me” = 0 to “definitely true of me” = 5 for TDSE. A full list of the items comprising each measure of religion/spirituality isprovided in Appendix 1.

Covariates included age (in years), gender (female = 1), percent life living in USA, income (≤ $15,000; $15,001–$30,000; $30,001–$50,000; $50,001–$110,000; > $110,000), education (less than college, college graduate, graduate school and above), marital status (married/live with partner = 1), employment status (full-time = 1), self-rated health (poor, fair, good, very good, excellent), childhood parent home ownership (yes = 1), and anti-depressive medication (yes = 1) in order to ensure that the relationship between childhood trauma and adult mental health is not confounded by the use of anti-depressants (i.e., in cases where medication is effective in reducing depressive symptoms, the trauma/depressive symptoms association could potentially be masked) (see Dew et al., 2010; Needham et al., 2018) for a similar approach). Analyses language spoken at home (only South Asian = 1 to only English = 5) and religious affiliation (Hindu, Muslim, Jain, Sikh, other, multiple religions, and none).

Analysis

Missing or incomplete data were accounted for in several variables. In the dependent variable, CES-D, cases with more than four missing items were excluded from analysis (n = 2). Participants with one to three missing items (n = 173) were adjusted to the full-scale equivalency with the following procedure: the 17, 18, or 19 available items were summed, divided by the number of available items, then multiplied by 20. In the independent variable, CTQ cases were dropped from the analytic sample when more than one item was missing from all three subscales (n = 45). If only one item was missing from one or two of the subscales, an equivalency score was calculated in the same manner as the dependent variable. This occurred in emotional abuse (n = 10), emotional neglect (n = 12), and physical neglect (n = 8). The following four covariates had missing items, which were imputed using the median or mode, depending on level of measurement: age (n = 1), percent life in the US (n = 1), home ownership (n = 19), and income (n = 58). Analyses were conducted using both imputed and non-imputed data, and differences were inconsequential, thus imputed data were retained for inferential analyses. The dependent variable, CES-D, was checked for excessive skewness as depressive symptoms may be non-normally distributed, but skewness (1.74) was within the bounds of normal distribution for linear analysis. Robustness checks with a log-transformed version of the variable showed no change in main effects and consistent patterns in interaction effects, so the non-transformed variable was retained. All key independent variables were zero-centered to decrease cross-product collinearity; then, generalized linear models with robust standard errors were used to assess the association between childhood trauma, adult religion and spirituality, and adult depressive symptoms in SAS 9.4 (SAS Institute, Cary, NC). We first inserted CTQ items plus covariates (unreported), then inserted R/S variables one at a time (religious affiliation remained as a control in all models), resulting in 24 main effect models (six for each CTQ subscale plus total CTQ score). We then evaluated the cross-product of each CTQ subscale and total score x R/S for the full sample.

Finally, since religion and spirituality differ in expression and meaning for men and women (Kent et al., 2021, 2023), we stratified the interaction models by gender. In all models, we report both adjusted and unadjusted p-values using the false discovery rate (FDR) technique, which reduces the reporting of false-positives. Our selected FDR threshold of p ≤ 0.10 means that 10 % or fewer of significant tests (prior to adjustment) are truly false-positives. We only draw conclusions on whether statistically significant interaction terms are present for males and/or females and make no attempt to compare the strengths of interaction coefficients in these two samples (see Reinert et al., 2015 for a similar approach).

Results

The sample was 48.7% female, averaged 61.2 years of age, and had lived in the US for 48.4% of their lives. Approximately 70% were Hindu, followed by 8.5% Muslim, 5.5% Sikh, 5% Jain, 4% other (Christian, Buddhist, etc.), and 5% none; 8% reported multiple affiliations, usually some combination of Hindu, Sikh, Jain, and/or Buddhist. Education and income were high compared to the general US population, along with marriage rates (90%). Depressive symptoms were relatively low at 7.6 on a scale ranging from 0 to 45, with anti-depressive medication used by 3.2% of the sample. The CTQ subscales and total score were also relatively low. These scores and other sample characteristics are summarized in Table 1.

Table 1.

MASALA study participant characteristics (non-imputed data)

Variable	Mean (std)/n (%)	Range
CES-D depressive symptoms	7.58 (7.19)	0–45.26
CTQ emotional abuse	6.59 (2.73)	5–23
CTQ emotional neglect	7.62 (3.55)	5–25
CTQ physical neglect	6.19 (2.03)	5–14
CTQ total score	20.46 (6.58)	15–48
Religious service attendance
Never	0 (0.00%)	–
Rarely	32 (3.76%)	–
About 1/month	281 (32.98%)	–
2–3/month	305 (35.8%)	–
1/week	174 (20.42%)	–
Several times/week	60 (7.04%)	–
Positive religious coping	2.75 (0.88)	1–4
Negative religious coping	1.45 (0.61)	1–4
Gratitude	4.72 (0.52)	1–5
Non-theistic daily spiritual experiences	3.66 (0.79)	1–5
Theistic daily spiritual experiences	4.04 (0.95)	1–5
Religious affiliation
Hindu	549 (63.99%)	–
Muslim	73 (8.51%)	–
Jain	46 (5.36%)	–
Sikh	47 (5.48%)	–
Other	32 (3.73%)	–
Multiple affiliations	66 (7.69%)	–
None	45 (5.24%)	–
Female	418 (48.72%)	–
Age	61.19 (8.84)	44–89
Percent life in the USA	48.42 (18.93)	1.64–100
Income
≤ $15,000	40 (5.00%)	–
$15,001–$30,000	61 (7.63%)	–
$30,001–$50,000	54 (6.75%)	–
$50,001–$110,000	200 (25.00%)	–
> $110,000	445 (55.63%)	–
Education
Less than college	114 (13.29%)	–
College graduate	275 (32.05%)	–
Graduate school and above	469 (54.66%)	–
Married/live with partner	775 (90.33%)	–
Employed full-time	572 (66.67%)	–
Childhood home ownership	547 (65.2%)	–
Anti-depressive medication use	28 (3.26%)	–
Self-rated health
Poor	4 (0.47%)	–
Fair	69 (8.04%)	–
Good	320 (37.30%)	–
Very good	373 (43.47%)	–
Excellent	92 (10.72%)	–
Language spoken at home
Only South Asian	124 (14.45%)	–
South Asian better than English	153 (17.83%)	–
Both equally	270 (31.47%)	–
English better than South Asian	216 (25.17%)	–
Only English	95 (11.07%)	–

Table 2 summarizes 24 main effect models regressing CES-D on four CTQ scales (Emotional abuse, emotional neglect, physical neglect, and total score) and six religion/spirituality (R/S) predictors. The top half of Table 2 reports coefficients of the four CTQ scales when entered with religious attendance as the R/S variable of interest (Models 1a–d). This was done: (a) in the interest of space, and (b) because results for CTQ scores were nearly identical across all models regardless of which R/S variable was entered. Models 1a-d indicate that CTQ emotional abuse, emotional neglect, physical neglect, and total score all have FDR-adjusted p-values ≤ 0.10 with positive slopes of 0.40, 0.40, 0.45, and 0.23, respectively. CTQ coefficients for the remaining 20 models are not reported as discussed above, though they can be summarized as follows. All are significant at p < 0.10 after FDR adjustment and have the following slope ranges: positive coping (Models 2a–d; b = 0.23–0.46), negative coping (Models 3a–d; b = 0.19–0.36), gratitude (Models 4a–d; b = 0.21–0.41), non-theistic DSE (Models 5a–d; b = 0.22–0.45), and theistic DSE (Models 6a–d; b = 0.23–0.47). This suggests that in highly controlled models, adult recollection of childhood trauma is associated with an increase in depressive symptom scores.

Table 2.

Regression of CES-D on childhood trauma and R/S characteristics in the MASALA study

Model Nos.a		CTQ emotional abuse models (1a-6a)		CTQ emotional neglect models (1b-6b)		CTQ physical neglect models (1c-6c)		CTQ total score models (1d-6d)
	Variable	b (95% CI)	p	b (95% CI)	p	b (95% CI)	p	b (95% CI)	p
1ab	CTQ emotional abuse	.40 (.21, .60)	***	–		–		–
1b	CTQ Emotional Neglect	–		.40 (.23,.57)	***	–		–
1c	CTQ Physical Neglect	–		–		.45 (.17, .73)	***	–
1d	CTQ Total Score	–		–		–		.23 (.14, .32)	***
1a–da	Religious attendance	.25 (−.27, .77)		.25 (−.26,.76)		.17 (−.35,.69)		.23 (−.27,.74)
2a–d	Positive coping	.24 (−.33, .81)		.36 (−.22, .93)		.26 (−.32, .83)		.33 (−.24, .89)
3a–d	Negative coping	2.23 (1.28, 3.19)	***	2.07 (1.19, 2.95)	***	2.26 (1.33, 3.18)	***	2.03 (1.14, 2.92)	***
4a–d	Gratitude	−2.07 (−3.32, −.83)	***	−1.77 (−2.99, −.55)	***	−2.07 (−3.28, −.86)	***	−1.74 (−2.97, −.52)	***
5a–d	Non-theistic DSE	−1.25 (−1.83, −.66)	***	−1.08 (−1.66, −.50)	***	−1.27 (−1.85, −.69)	***	−1.09 (−1.67, −.52)	***
6a–d	Theistic DSE	.00 (−.50, .50)		.09 (−.41, .60)		.02 (−.49, .52)		.10 (−.40, .59)

Covariates include gender, income, education, marital status, employment status, language spoken at home, self-rated health, anti-depressant medication, percent life in the USA, childhood parent home ownership, and religious affiliation

^aTable 2 consists of 24 models, six for each CTQ subscale/total score; n ranges from 828 to 852

^bCTQ coefficients reflect CTQ main effects for the four religious attendance models (Models 1a–d). CTQ coefficients for the remaining 20 models are redundant and not reported, though they can be summarized as follows: all are significant at p < 0.10 after FDR adjustment and have the following slope ranges: positive coping (Models 2a–d; b = 0.23 − 0.46), negative coping (Models 3a–d; b = 0.19 − 0.36), gratitude (Models 4a–d; b = 0.21 − 0.41), non-theistic DSE (Models 5a–d; b = 0.22 − 0.45), and theistic DSE (Models 6a–d; b = 0.23 − 0.47)

^***p ≤ 0.10 after FDR adjustment

The bottom half of Table 2 reports the main effects of six R/S variables predicting CES-D scores. Patterns were consistent across all four CTQ scales, with three R/S variables emerging as non-significant and three emerging as significant after FDR correction. The three non-significant variables were religious attendance (Models 1a–d), positive religious coping (Models 2a–d), and theistic daily spiritual experiences (Models 6a–d). These results indicate no main effect correlation of these variable with CES-D scores. The three significant variables to emerge were negative religious coping (Models 3a–d), gratitude (Models 4a–d), and non-theistic daily spiritual experiences (Models 5a–d). Negative coping indicated a positive relationship (b = 2.23, CI 1.28, 3.19), suggesting that higher depressive symptom scores were correlated with higher negative religious coping. Gratitude (b = − 2.07, CI − 3.32, − 0.83) and non-theistic daily spiritual experiences (b = − 1.25, CI − 1.83, − 0.66) evinced negative slopes, suggesting an inverse correlation between these variables and depressive symptoms.

Table 3 reports the results of 24 interaction models examining the full MASALA sample. These are comprised of six R/S variables interacted with four CTQ scales. Further, given previous evidence demonstrating gender differences in religious habits as well as appropriation of religious and spiritual resources (Kent, 2020; Kent et al., 2023), we elected to report ancillary analyses stratifying the sample by gender, also reported in Table 3.

Table 3.

CTQ × R/S interaction terms, MASALA study

	Full samplea		Male only		Female only
	b (95% CI)	p	b (95% CI)	p	b (95% CI)	p
CTQ emotional abuse
x Religious attendance	.32 (.13, .52)	***	.23 (−.01,.47)		.37 (.08, .66)	***
x Positive religious coping	−.05 (−.31,.20)		.09 (−.17, .36)		−.12 (−.49, .25)
x Negative religious coping	.15 (−.20, .49)		−.29 (−.68, .09)		.71 (.19, 1.23)	***
x Gratitude	.26 (−.08, .59)		.56 (.24, .87)	***	−.04 (−.62, .54)
x Non-theistic DSE	.04 (−.22, .29)		.10 (−.26, .47)		.05 (−.29, .38)
x Theistic DSE	.02 (−.20, .23)		.05 (−.20, .29)		−.03 (−.33, .28)
CTQ emotional neglect
x Religious attendance	.20 (.04, .36)	***	.21 (.04, .38)	***	.17 (−.09, .43)
x Positive religious coping	.11 (−.05, .28)		.20 (.01, .40)	*	.03 (−.21, .28)
x Negative religious coping	.21 (.02, .40)	***	.11 (−.14, .36)		.40 (.03, .78)	*
x Gratitude	.10 (−.14, .34)		.20 (−.09, .49)		−.08 (−.59, .43)
x Non-theistic DSE	.03 (−.16, .22)		.08 (−.17, .33)		.01 (−.27, .28)
x Theistic DSE	.10 (−.06, .25)		.13 (−.04, .30)		.00 (−.24, .25)
CTQ physical neglect
x Religious attendance	.10 (−.20, .39)		.16 (−.17, .49)		−.01 (−.52, .49)
x Positive religious coping	−.11 (−.40, .19)		.02 (−.32, .37)		−.44 (−.97, .08)
x Negative religious coping	.01 (−.42, .43)		.01 (−.51, .52)		.03 (−.66, .72)
x Gratitude	−.08 (−.39, .23)		.00 (−.33, .33)		−.49 (−1.57, .59)
x Non-theistic DSE	−.12 (−.40, .17)		−.06 (−.44, .32)		−.12 (−.59, .35)
x Theistic DSE	−.14 (−.40, .23)		−.03 (−.33, .27)		−.35 (−.77, .07)
CTQ total score
x Religious attendance	.13 (.05, .21)	***	.12 (.03, .21)	***	.12 (−.02, .26)
x Positive religious coping	.03 (−.07, .13)		.11 (−.01, .22)		−.04 (−.20, .11)
x Negative religious coping	.10 (−.03, .24)		.02 (−.15, .19)		.24 (.03, .46)	*
x Gratitude	.07 (−.06, .20)		.16 (.01, .30)	*	−.04 (−.31, .22)
x Non-theistic DSE	.01 (−.10, .12)		.04 (−.10, .18)		.01 (−.15, .16)
x Theistic DSE	.03 (−.05, .11)		.06 (−.03, .15)		−.03 (−.16, .10)

Covariates include gender, income, education, marital status, employment status, language spoken at home, self-rated health, anti-depressant medication, percent life in the USA, childhood parent home ownership, and religious affiliation. All models include main effect and cross-product terms

^an ranges from 828 to 852 for full sample, 423 to 435 for male sample, and 404 to 417 for female sample

^*p ≤ .05 before FDR adjustment,

^***p ≤ .10 after FDR adjustment

In the full sample, three of the 24 models resulted in a significant interaction after FDR adjustment. Three of these involved religious service attendance, which emerged with a positive, significant interaction in three of the four CTQ scales (emotional abuse, emotional neglect, and total score). Figure 1 visually illustrates the third interaction effect of CTQ total score and religious service attendance, indicating a general increase in CES-D as total CTQ score increases, with higher scores noted among those attending religious services more frequently as adults. The visual plots (not shown) for the other two interactions are very comparable. The fourth significant interaction was also positive: negative religious coping in models of emotional neglect. Figure 2 visually illustrates this relationship, indicating that higher levels of negative coping were associated with higher levels of CES-D, though at −1 standard deviation of negative coping the slope approaches zero.

Fig. 1.

Interaction of CTQ total score and religious service attendance, predicting CES-D

Fit computed at rsage=61.19 harminc=4.24 selfhealth=3.56 educat_cat3=2.411 P12AHOME=2.005 P12PERCUS=48.46 gender=Female marital_cat3v1=Married employ_cat3v1=Employed meddep=No sartrad=None child_pov_home=No

Fig. 2.

Interaction of CTQ emotional neglect and negative religious coping, predicting CES-D

Fit computed at rsage=61.1 harminc=4.243 selfhealth=3.564 educat_cat3=2.409 P12AHOME=2.005 P12PERCUS=48.32 gender=Female marital_cat3v1=Married employ_cat3v1=Employed meddep=No sartrad=None child_pov_home=No

Regarding the gender specific analyses, in the male sample, the religious attendance result was replicated in models of emotional neglect and total score, while three other positive interactions emerged: gratitude in models of emotional abuse and total score, and positive coping in models of emotional neglect. In the female sample, religious attendance was replicated in models of emotional abuse; negative coping was replicated in models of emotional neglect. Negative coping also emerged in models pertaining to emotional abuse and total score, suggesting that women may be at higher risk then men for negative outcomes associated with religious struggle. For both men and women, the plots (not shown) for religious attendance, gratitude, and positive coping are very comparable to Fig. 1; the plots for negative coping are very comparable to Fig. 2. We would also note that the results of our gender-stratified analyses were consistent when we conducted three-way interaction terms between childhood trauma, gender, and religiosity in predicting depressive symptoms.

Discussion

Adverse childhood experiences have been recognized as a public health priority (Ganson et al., 2021). The chief goal of this study was to understand the relationship between childhood trauma and depressive symptoms among a South Asian cohort in the USA, and to further assess whether several facets of religiosity moderated this relationship. On the whole, the religion and health literature have tended to neglect South Asian cohorts, but as our results show, there are potential differences in how religiosity may operate in the face of early life adversity to predict depressive symptoms in adulthood.

Several key findings emerged from our study. We found that the recollection of childhood emotional neglect, emotional abuse, and physical neglect were all associated with higher depressive symptoms in this South Asian sample. This is an important finding because Asian Americans are a group thought to be at higher risk of experiencing early life maltreatment (Kim et al., 2007; Maker et al., 2005), with nearly half of South Asians in the USA having witnessed parental violence as children and approximately 25% experiencing some other type of abuse (Kim et al., 2007). Therefore, our study extends empirical research previously documenting pernicious associations between childhood adversity (e.g., emotional and physical) and lower mental well-being in the general population (Negele et al., 2015) and among Asian Americans (Hahm et al., 2012; Robertson et al., 2016) to the specific study of a South Asian group in the USA.

The second key set of findings was related to the direct associations between various dimensions of religion/spirituality and depressive symptoms among our South Asian sample. We found evidence that across the six indicators of R/S assessed in the current study, three bore null associations with depressive symptoms (religious attendance, positive religious coping, and theistic daily spiritual experiences), while negative religious coping was associated with greater depressive symptoms and gratitude and non-theistic daily spiritual experiences were associated with lower depressive symptoms. Our findings were generally consistent with a limited body of work that has examined the relationship between R/S and mental well-being in a South Asian sample. A study by Kent et al. (2020) revealed that it was the private religious beliefs and practices that predicted well-being; indeed, gratitude, non-theistic spiritual experiences, and closeness to God predicted better self-rated mental health, while negative religious coping was linked to worse self-rated mental health, consistent with the findings of our study. As Kent and colleagues noted, the most consistent of the variables in predicting outcomes related to well-being, encapsulating both physical and mental health, were non-theistic daily experiences (e.g., the extent to which a person makes a spiritual connection between themselves and the world), measured by items such as “I am touched by the beauty of creation” or “I experience a connection to all of life.” In addition, measures of religiosity that have been found to be robustly linked to mental well-being in the overall American populace—including religious attendance (Schieman et al., 2013), positive religious coping (Pargament et al., 1988) and closeness to God (Upenieks, 2022)—were not found to significantly predict mental health outcomes in the South Asian sample. Taken as a whole, the findings suggest that elements of R/S that have long-standing associations with well-being in the general US population may operate differently for religious minority groups in the USA, highlighting the importance of future research measuring religion/spirituality as both a complex and multidimensional construct (Bradshaw & Ellison, 2010).

Though these findings with respect to the main associations of religiosity with depressive symptoms in a South Asian sample are undoubtedly important, the main contribution of our study was to show which facets of religious/spiritual life significantly moderated the relationship between early life adversity and depressive symptoms in adulthood. Drawing from the stress process framework, we consider findings related to our total sample first before progressing to the gender-stratified analyses.

Moderation analyses found that religious attendance and negative religious coping exacerbated the relationship between childhood trauma and depressive symptoms. This pattern of findings appeared to extend to more domains of early life trauma for attendance, which saw negative interactions for emotional abuse, emotional neglect, and total CTQ score. To make sense of this somewhat counterintuitive finding—that greater religious attendance was associated with higher depressive symptoms for victims of greater childhood trauma—it is crucial to consider that rates of religious service attendance are significantly lower among South Asians (Kent et al., 2020; Stroope et al., 2020). On the one hand, religious attendance has been conceptualized as a stress buffer, one that should attenuate the adverse effects of a stressor by providing people the opportunity to build networks of support within a congregation (Bradshaw & Ellison, 2010) that can empower management of past and present trauma (Jung, 2018), reducing its harmful effects. On the other hand, recent studies of nationally representative data from the USA have tended to suggest that religious attendance fails to buffer the association between childhood abuse and mental health in adulthood (Manning & Miles, 2018; Upenieks, 2021). This latter set of studies has tended to suggest that, at least within the context of childhood adversity, institutional religiosity may be less efficacious compared to more private dimensions.

Our study, however, found that religious attendance strengthened the already harmful relationship between childhood trauma and depressive symptoms. The uniqueness of the South Asian religious/spiritual context likely underlies why this pattern diverges from previous research. Indeed, work by Stroope and colleagues (2020) suggests that especially in the South Asian context, where weekly religious attendance is neither encouraged nor required, a higher frequency of attendance may be a form of “resource mobilization”—in other words, when confronted with stress and adversity, people may seek out sources of support. Empirically in our data, any experience of childhood trauma and religious attendance were moderately correlated at r = 0.35 (p < 0.05). Though far from a perfect correlation, this may be a case of reverse causation. Indeed, similar explanations have been used to account for the associations between frequent prayer and psychopathology in the USA, where people facing heightened stress and discomfort may turn to religion, especially if other coping resources have been exhausted (Masters & Spielmans, 2007). In our study, we see a possible selection effect at work, whereby increased mental health difficulties (stemming from trauma) could lead to greater religious attendance. As a result, it is possible that an uptick in religious attendance—a pattern not uncommon among victims of childhood trauma (Walker et al., 2009)—might account for this stress exacerbation pattern at higher frequencies of religious attendance. If victims of childhood trauma who are experiencing higher depressive symptoms “select” into communal worship participation, this could potentially obscure any stress-buffering functions that religious participation could serve. We also note that this same study by Stroope and colleagues (2020) found that the experience of criticism from congregation members was also associated with greater anxiety and anger. Though religious communities have been known to offer programs or services that could help victims of childhood trauma rebuild their self-esteem (Chaves, 2004), to the extent that these religious environments are characterized by interpersonal conflict, it might undermine any potential positive effects of worship participation.

A second important moderation pattern that we observed was that negative religious coping exacerbated the association between childhood trauma and depressive symptoms. Both positive and negative religious coping have been identified in the literature following trauma exposure (Pargament et al., 1988). Our results show that engaging in positive religious coping, defined as seeking spiritual support and forgiveness from God or a higher power, did not effectively function as a stress buffer. This finding is inconsistent with research that has reported beneficial links between positive religious coping and lower levels of negative psychological symptoms following trauma, such as depression, withdrawal, or isolation (e.g., Ano & Vasconcelles, 2005; Gerber et al., 2011; Pargament et al., 2000; Upenieks, 2021), but aligns with a recent study by Walker and colleagues (2022) which shows that engaging in positive religious coping failed to moderate the relationship between childhood trauma and resilience. It is possible that in the South Asian context, other factors are more important for cultivating resilience following exposure to childhood trauma, including support from one’s family (Schaefer et al., 2018). Though positive religious coping has been identified as a protective factor following trauma exposure in predominantly Christian samples in the USA, even these studies recognize that the true impact of religious coping is difficult to ascertain due to the influence of negative religious coping (Bradley et al., 2005; Witvliet et al., 2004).

Patterns of negative religious coping, characterized by an insecure relationship with God, fear or punishment, or blaming God for not being a protector, have generally been associated with poor psychological outcomes (Ano & Vasconcelles, 2005; Gerber et al., 2011). That negative religious coping strengthened the relationship between childhood trauma and adult depressive symptoms suggests that the stressful experiences associated with childhood abuse could promote a tendency to question the nature of a divine power (Bierman, 2005). There is also some evidence that those who experienced childhood abuse tend to view God as punitive and harsh and may develop negative emotions such as disappointment or frustration that accompany such traumatic experiences (Reinert & Edwards, 2009). If trauma survivors are unable to reconcile or resolve any conflict—either within and/or independent of their R/S beliefs—that may arise due to childhood trauma, they may struggle to develop a coherent global meaning system (Santoro et al., 2016; Walker et al., 2009). Work by McCormick and colleagues (2017) suggests that greater childhood adversity is associated with all six subscales of R/S struggles identified by Exline and colleagues (2014), including divine, demonic, moral, meaning, interpersonal, and doubt struggles. Our findings are thus consistent with past research, suggesting that individuals may experience distress or conflict across multiple domains of R/S after the experience of trauma in childhood, such that engaging in negative religious coping might actually heighten the experience of depressive symptoms. Taken together, this pattern of moderation finding suggests the unique role of negative religious coping in a South Asian sample that, absent norms of weekly religious participation, might be more affected by negative patterns of religious coping for childhood trauma victims and illustrates that, within the stress process framework, religion more often than not functions as a stress-exacerbator for US South Asians.

Our final study objective was to assess whether the moderation patterns observed in the main sample operated in a similar fashion for male and female respondents. Generally, the main pattern of findings was replicated with two main exceptions: among men, gratitude and positive religious coping, in addition to religious attendance, exacerbated the relationship between childhood trauma and depressive symptoms. Meanwhile, the depressive symptoms of women seemed to be more affected by negative religious coping and were also adversely affected at higher levels of religious attendance.

Speaking to the first set of findings regarding men, fewer South Asian men experience childhood trauma compared to their female counterparts (Maker et al., 2005). Though a speculative interpretation, the rarer occurrences of trauma for men may lead them to seek out greater religious/spiritual experiences, representing a selection effect whereby greater mental health problems may foster greater self-reports of religiosity/spirituality. Within the American context, previous research suggests that women tend to benefit more from religious participation than men (Bonhag & Upenieks, 2021; Kent, 2020; Sharp et al., 2016), so men may not be as poised to utilize elements of religion/spirituality in a way that would be efficacious for their mental health. With respect to women, the result that women are more affected by religious/spiritual struggles is consistent with work from the American context. Negative religious coping, including the tendency for women to view God as harsh and judgmental, has been found to undermine the health of women more than men (Bonhag & Upenieks, 2021). Since no previous study has looked at gender differences in R/S coping and mental health for victims of childhood trauma in the South Asian population, more research is needed to test the robustness of the patterns we document here.

Limitations and Future Directions

Several study limitations must be acknowledged. First, the external validity of our sample is affected by MASALA’s sampling frame, which included South Asians from the Chicago and San Francisco area that were 40 years of age or older and comprised a higher proportion of high-SES South Asians. Our results may therefore not reflect all South Asians in the USA.

Second, the cross-sectional nature of our data limits our ability to derive causal conclusions regarding the relationship between childhood trauma, R/S, and depressive symptoms in adulthood. Additional research is also needed to explore potential changes in levels of religiosity and/or religious coping between childhood and adulthood following exposure to childhood trauma (e.g., Walker et al., 2009), and whether any changes across religious/spiritual dimensions are linked with stress-buffering or exacerbating patterns in this South Asian sample. We would also note that retrospective self-reports were used to gauge childhood trauma, so our results could be impacted by recall bias due to people trying to remember events from childhood that in many cases occurred several decades prior.

Finally, we would note that the MASALA data only have three of five CTQ score domains available. A more comprehensive test of the framework that we put forth would be further informed if all dimensions were considered. In addition, further research is needed to expand the scope of childhood trauma or adversity to include other forms of trauma, such as community violence in conjunction with neglect and abuse. Taking a more expansive approach to assessing childhood trauma, including the possibility of polyvictimization (trauma in multiple domains) and the role of religiosity in either combatting or exacerbating such experiences, would be instructive.

Conclusion

Despite these limitations, this is the first community-based study of South Asians, to our knowledge, to consider the association between various forms of trauma and depressive symptom outcomes. Our findings point to the importance of private features of religion and spirituality (e.g., gratitude and non-theistic spiritual experiences) for the mental well-being of South Asians in the USA, with moderation patterns generally suggesting detrimental aspects of religious attendance and negative religious coping among victims of childhood trauma. The US South Asians remain an understudied group in the religion and health literature, and we hope that this study sheds light on the important differences in the function and effectiveness of religion/spirituality for those faced with early life trauma, suggesting that a “one-size-fits-all” approach with respect to religion/spirituality and religious coping across diverse religious samples is likely to be ineffective.

Appendix 1. Full list of SSSH religion/spirituality items

Religious/spirituality measure	Item (if applicable)	Response categories
Religious attendance	How often do you attend religious services?	Never Rarely About once a month 2–3 times per month Once a week Several times per week
Positive religious coping	In facing recent stressful life events… I saw my situation as part of God’s plan I tried to make sense of the situation with God I worked together with God to relieve my worries I did what I could and put the rest in God’s hands I sought God’s love or care I trusted that God would be by my side I trusted that God would help me get through one day at a time I looked to my faith in God for hope about the future	Not at all Somewhat Quite a bit Very much so
Negative religious coping	In facing recent stressful life events… I wondered what I did for God to punish me I wondered if God allowed this to happen because of my wrongdoings I believed the devil or evil spirits were responsible for my situation I felt as though the devil, or an evil spirit was trying to turn me away from God I wondered whether God had abandoned me I questioned God’s love or care for me	Not at all Somewhat Quite a bit Very much so
Gratitude	I have so much in life to be grateful for If I had to list everything that I felt grateful for it would be a very long list	Strongly disagree Somewhat disagree Neutral Somewhat agree Strongly agree
Non-theistic daily spiritual experiences	I experience a connection to all of life I feel deep inner peace or harmony I am touched by the beauty of creation I feel a selfless caring for others	Never Once in a while Some days Every day Many times a day
Theistic daily spiritual experiences	I feel God’s love for me through others I desire to be closer or in union with God	Definitely true of me Tends to be true of me Unsure Tends not to be true of me Definitely not true of me