The Links Between Religiosity,Childhood Sexual Abuse,and Subsequent Marijuana Use

Abstract

A number of studies indicate that childhood sexual abuse (CSA) has a negative impact on later psychological well-being. It is well documented that experiences of CSA are associated with depression, self-destructiveness, and subsequent substance abuse or alcohol consumption. Compared with women who experienced no such sexual abuse in childhood, women who were victims of sexual abuse in childhood were more likely to be depressed and use drugs or consume alcohol in later life. Analyzing data of 1,569 females derived from the “Longitudinal Study of Violence Against Women,” this study examines whether the strain caused by sexual victimization leads to a higher level of subsequent marijuana use and whether religiosity moderates the negative effects of CSA. It was found that CSA was associated with an increased level of marijuana use in high school. However, more proximate sexual victimization (victimization in college) seemed to override the impact of CSA on subsequent marijuana use. Religiosity was found to moderate the effect of CSA on marijuana use in high school. Religiosity was negatively associated with marijuana use in high school as well as the second and fourth collegiate years. Policy implications and promising directions for future research are discussed.

Keywords

religiosity religion religious attitude church attendance sexual abuse childhood sexual abuse sexual victimization strain substance use marijuana use college students

Studies have found that sexual abuse in childhood is associated with a number of psychological problems, such as lack of self-esteem, posttraumatic symptoms, higher levels of depression, anxiety, suicidal attempts, delinquency, and other behavioral problems (e.g., Arellano, Kuhn, & Chavez, 1997; Fergusson, Horwood, & Lynskey, 1994; Gall, 2006; Widom & Ames, 1994; Wilsnack, Vogeltanz, Klassen, & Harris, 1997). A number of studies have found childhood sexual abuse (CSA) and related stressors increase the likelihood of substance use and misuse (Dube et al., 2005; Heffernan et al., 2000). According to Rohsenow, Corbett, and Devine (1988), the prevalence rate of reported CSA among women in alcohol or drug treatment programs is as high as 47%. Paone, Chavkin, Willets, Friedman, and Des Jarlais (1992) suggested that one of the possible explanations for the association between CSA and substance use is that victims of CSA may use illicit drugs as a type of self-medication to cope with the trauma of abuse. CSA victims also use illicit substances as a way to deal with abuse-associated negative emotional states, such as depression or anxiety (Jarvis, Copeland, & Walton, 1998).

Having identified a link between CSA and subsequent drug use, research that examines resiliency factors and coping mechanisms thus becomes important as it can help to develop possible avenues for intervention (Gall, 2006; Roche, Runtz, & Hunter, 1999). A number of studies have indicated that CSA and substance use are predictors of adult sexual victimization (e.g., Messman-Moore & Long, 2002). Thus, understanding what disrupts the relationship between sexual victimization and substance use is vital to programming to assist individuals who suffer from CSA and subsequent sexual victimization. Early intervention may reduce the risk factors (e.g., substance abuse) for revictimization.

Most of the studies to date do not concomitantly examine CSA and subsequent sexual victimizations to disentangle how women’s sexual victimizations at different stages in the life course impact their mental health and subsequent substance use. Although religiosity has been found to serve as a coping mechanism in dealing with stress and distressful life events, little is known about whether religiosity exerts a beneficial effect on soothing individuals’ trauma from CSA. Research in psychology and health indicates that females, more than males, deal with depression and stress by emotionally “venting” or looking to religion for spiritual support (Angst et al., 2002). CSA generates trauma and stressors, which impact victims’ emotional states in their later life and can lead to increased levels of substance use.

General strain theory posits that strains or stressors can generate negative emotions, such as anger, frustration, and depression, which in turn may lead to remedial or adaptive actions, such as delinquency (or illicit drug use). In view of the impact of CSA on victims’ emotional states and the subsequent adaptation for victims (e.g., increased substance use), general strain theory is considered appropriate to explain the causes of subsequent substance use.

As stated above, a great body of research has found that religiosity exerts beneficial effects on individuals’ well-being and mental health, such as moderating harmful effects of various stressful events (Ellison, Boardman, Williams, & Jackson, 2001). Nevertheless, none of the existing studies utilize the propositions of Agnew’s (1992) general strain theory to empirically test whether religiosity in some way affects the strain produced by childhood sexual experiences, which influences later substance use. To bridge the gap in the previous studies, this study utilizes the theoretical framework of general strain theory to examine whether the strain caused by CSA is associated with a higher level of substance abuse. It also investigates whether religiosity moderates the negative impacts of CSA and in turn, reduces subsequent substance use. Because of the limited availability of existing data, this study focused only on female samples.

This study aims to answer two research questions: (a) Is CSA associated with higher levels of subsequent marijuana use? and (b) Does religiosity moderate the effect of childhood sexual victimization on subsequent marijuana use? This article starts with a literature review of the relationship between CSA and subsequent drug use. It then provides a brief review of religiosity and its impact on coping with life stress, followed by a presentation of a theoretical framework and methodology along with findings. Finally, directions for future research and policy implications are discussed.

Review of Literature

Links Between CSA and Subsequent Substance Abuse

A number of studies have found a link between CSA and subsequent substance abuse and depression (e.g., Briere & Elliott, 1994; Gold, Sinclair, & Balge, 1999; Luster & Small, 1997; Miller & Downs, 1993). Deriving data from a national representative sample of women aged 21 years or above in the United States, Wilsnack et al. (1997) found that women with histories of CSA were more likely to report recent (12 months prior to their interviews) alcohol consumption and dependence. They suggested that women’s experience of CSA may be an important risk factor for subsequent substance abuse. Similarly, Kilpatrick et al. (2000) found adolescents (12-17 years old) who were victims of sexual abuse were more likely to report current alcohol, marijuana, and hard drug use. Analyzing longitudinal data derived from 150 females, Bailey and McCloskey (2005) revealed that a history of childhood abuse was associated with later substance abuse.

Religiosity, Coping With Life Stress, and Responses to Strain

Although there are no studies that specifically investigate the link between religiosity, CSA, and marijuana use, some studies examine the relationship between religiosity and strain. For example, analyzing the panel data from the National Survey of Black Americans, Jang and Johnson (2003) found that religiosity buffered the effects of strain on negative emotions and weakened the effects of negative emotions on deviance. In a subsequent study, using the same panel data, Jang and Johnson (2005) found that African American females, in times of distress, were less likely to behave aggressively than males because they were more religious. For stressed-out females, it seems that religion has the effect of softening or relieving their impulses to act aggressively.

A number of studies have found that religiosity is a significant predictor that deters individuals from illicit drug use (e.g., Chu, 2007, Hodge, Cardenas, & Montoya, 2001; Jang & Johnson, 2001, 2007). A large body of studies in psychology, health care, and public health indicate that religious involvement has a positive effect on enhancing mental and physical well-being and longevity (e.g., Ellison, 1991; George, Ellison, & Larson, 2002), buffering stress (Bjorck & Thurman, 2007; Hood, Spilka, Hunsberger, & Gorsuch, 1996), and reducing depression and psychological distress (e.g., Williams, Larson, Buckler, Heckman, & Pyle, 1991).

Utilizing data from a probability sample of 78,458 participants from six European countries, Angst et al. (2002) found that females reported a higher number of depressive symptoms than males. With regard to coping with depression, males tended to increase sport activities and alcohol use, whereas females were inclined to seek emotional release (e.g., talking to friends) and religion. Studies on religious coping suggest that a belief in God or a higher power may be a source of support under the most calamitous circumstances or extreme states of trauma or stress (e.g., Gall, Basque, Damasceno-Scott, & Vardy, 2007; Pargament, 1997; Pargament & Brant, 1998). CSA victims can turn to religious behavior, such as praying or attending religious services for support. Similarly, Fiala, Bjorck, and Gorsuch (2002) found that religiosity was associated with lower levels of depression and higher degrees of life satisfaction.

Although relatively and possibly a great many of studies have explored general coping with CSA, there have been relatively few studies examining religious or spiritual coping and survivors’ adaptation. Gall’s study (2006) represented one of the very few studies that examine the relationship between spiritual coping and distress among adult survivors of CSA. For people who accepted religion, God can serve as a source of asylum, which provides security, comfort (Gall et al., 2007) as well as assistance and guidance. Analyzing data derived from 101 women who experienced CSA, Gall found positive coping behavior, such as praying for God’s help and forgiveness, was associated with lower levels of depression and anger. However, some studies (e.g., Gall, 2006; Rossetti, 1995) identified some of the CSA survivors who experienced a disruption of their sense of spirituality and developed negative spiritual coping, such as responding to their current life stress by being angry and discontented with God.

Doxey, Jensen, and Jensen (1997) examined the relationships between CSA, religion, and mental health by analyzing data collected from 5,417 women. They revealed that women with histories of CSA scored lower on positive mental health indicators. However, religion was found to moderate the negative mental health outcomes associated with CSA.

Theoretical Framework

Agnew’s (1992) general strain theory posited that the presentation of negative stressors, such as victimization, can create pressure, anger, and frustration. Broidy and Agnew (1997) argued that although males and females are equally likely to respond to strain with anger, males tend to respond to strain outwardly and to resort to aggressive acts or delinquency, whereas females are more likely to react to strain inwardly and exhibit self-directed emotions, such as anxiety and depression. In a similar vein, Jang (2007) posited that women are likely to respond to strain through self-directed deviant (e.g., drug use) or nondeviant coping behaviors (e.g., avoidance or religious coping).

Agnew (2002) in his general strain theory included several categories of strain, such as the absence of positive stimuli and the appearance of negative stimuli. He found that strain produced by experienced victimization and vicarious victimization (friends or family) of serious assaults was associated with delinquency (Agnew, 2002). The strain of experiencing CSA can also create trauma, frustration, depression, and anger, which leads individuals to use different approaches to cope with these negative affective states. As some of the previous studies identify, variations exist in response to strain between males and females. In view of the positive health outcomes (e.g. lower levels of drug use, personal well-being, and overall adaptation) linked to religiosity (George et al., 2002) and the soothing effect of religion on stressful life events or conditions (Ellison et al., 2001), it is conceptualized that religiosity somehow impacts the strain produced by victimization of CSA, which influences later substance use. Hence, the following hypotheses are formulated.

Hypothesis 1 (H1): Individuals experiencing CSA are more likely to use marijuana compared with individuals who did not experience CSA.

Hypothesis 2 (H2): Religiosity moderates the effect of CSA on subsequent substance use.

Data and Method

Sample

Data derived from the “Longitudinal Study of Violence Against Women (LSVAW): Victimization and Perpetration Among College Students in a State-Supported University in the United States, 1990-1995” (White, Smith, & Humphrey, 2001) were used to assess the links between childhood sexual victimization, religiosity, mental problems, and subsequent drug use.

The LSVAW study engaged a sample of two undergraduate classes (1990 and 1991 classes) of a medium-size state-supported university in the southeastern region in the United States. Individuals who were born in 1972 and 1973 were first interviewed when they were 18 years and reinterviewed when they were 19, 20, 21, and 22 years old (Humphrey & White, 2000; White et al., 2001). The survey collected information on respondents’ experience of sexual assault, religious attitude, church attendance, dating behavior, sexual experience, and their psychological states.

An outline of the 1,569 participants in this study showed their economic and demographic characteristics were representative of students attending public universities nationally. Among the sample, Whites comprised about 71% of the sampled respondents, whereas African Americans and other ethnic groups made up approximately 25% and 4% of the respondents, respectively (Humphrey & White, 2000). A total of 47.9% of the original 1990 class and 45.4% of the 1991 sampled students retained and completed each of the five-wave interviews. No significant difference was found between the attrition and remaining sample in terms of demographic characteristics (e.g., ethnicity and age; White et al., 2001).

Measures

The main independent variable—CSA—was defined as “any form of a sexual act perpetrated by an adult or any coercive sexual act perpetrated by a similarly aged peer on the respondent before age 14” (Humphrey & White, 2000, p. 421). It was coded as a dummy variable with 1 representing a victim of CSA; 0 representing a nonvictim. The variable of CSA consisted of three items which asked whether respondents experienced any of the following situations: (a) a person fondled you in a sexual way or touched your sex organs or asked you to touch their sex organs, (b) a male attempted intercourse with you (but penetration did not occur), (c) a male had intercourse with you (penetration occurred, ejaculation not necessary; Humphrey & White, 2000). The respondent was considered a victim and coded as 1 if she had experienced any of the above-mentioned inappropriate conduct(s) as a child (before 14 years old). It is coded as 0 if the respondent did not experience any of the above inappropriate sexual act(s).

The other independent variable—religiosity—was measured by two indicators: church attendance and religious attitude. The indicator of church attendance was inversely coded based on the original question in the survey. After recoding, higher values reflect higher frequencies of church attendance: 1 = never, 2 = less than once a month, 3 = 1 to 3 times a month, 4 = once a week, and 5 = more than once a week. Religious attitude was a categorical variable: 1 = no influence, 2 = some influence, 3 = fair amount of influence, and 4 = great deal of influence. The control variables consisted of race, high school marijuana use, prior sexual victimization, and depression. Race was coded as a dummy variable with 1 representing White and 0 representing non-White. To consider that previous marijuana use may have an impact on subsequent drug use, marijuana use in high school was controlled. In addition, to account for other proximate sexual victimization that may impact individuals’ affective states and subsequent marijuana use, the study also controlled for sexual victimizations in high school and each of the collegiate years. The variable of adolescent victimization was based on the responses to questions in which the respondents were asked to indicate whether they had experienced each of the following sexual behaviors: unwanted contact, verbal coercion, attempted rape, or rape since age 14 (Humphrey & White, 2000; White et al., 2001).

On each of the follow-up surveys (from Year 1 to Year 4 in college), respondents were asked to indicate how many times they had experienced coercive sexual contacts in the past year. Sexual victimization at each year in college was coded as a dummy variable with 1 indicating that the respondent experienced any coercive sexual contacts (unwanted contact, verbal coercion, attempted rape, or rape) in the past year and 0 indicating that the respondent did not experience any coercive sexual contacts (White et al., 2001). In addition, depression index (at high school and collegiate years) was also controlled. The composite of depression was measured by the mean score of the following five items: being moody; in low spirits; feeling of stress, strain, or pressure; being depressed; and being blue (1 = not at all like me, 2 = a little like me, 3 = somewhat like me, 4 = mostly like me, and 5 = very much like me). The mean score was only computed when there were at least three items that were not missing (White et al., 2001). A higher score in the depression composite indicates a higher level of depression. The demographic variable “race” was coded as dummy with 1 representing Whites and 0 representing non-Whites.

The dependent variable—marijuana use—(in high school and in the second and fourth collegiate years) was based on the reported marijuana use in the year prior to the survey. There are five categories in the original coding (1 = never, 2 = less than once a month, 3 = 1 to 3 times a month, 4 = 1 to 2 times a week, 5 = more than 2 times a week). The frequency distribution indicated the variable of “marijuana use” was skewed because only a few respondents reported that they used marijuana on a frequent basis. The frequency distribution also indicated the majority of the respondents were nonusers. Thus, the variable of marijuana use was recoded and dichotomized into two categories (0 = marijuana nonuser, 1 = marijuana user). The description of measures of independent and dependent variables are displayed in Table 1.

Table 1.

Description of Measure

Variable	Indicator
Independent and control variables
Church attendance	How often did you attend church/synagogue in the past 12 months?	1 = never
		2 = less than once a month
		3 = 1 to 3 times a month
		4 = once a week
		5 = more than once a week
Religious attitude	How did religion influence your life?	1 = no influence
		2 = some influence
		3 = fair amount of influence
		4 = great deal
Childhood sexual abuse	Did you experience any of the following by 14 years old?	0 = no
	Being fondled, attempted intercourse, intercourse	1 = yes
Sexual victimization in high school	Did you experience any of the following sexual acts in high school (since 14 years old)?	0 = no
	Unwanted contact, verbal coercion, attempted rape, or rape	1 = yes
Sexual victimization in college	Did you experience any of the following sexual acts in the past year?	0 = no
	Unwanted contact, verbal coercion, attempted rape, or rape	1 = yes
Marijuana use in high school	Marijuana use in the past 12 months	0 = nonuser
		1 = user
Depression index	Moody: moody, broody about things	1 = not at all like me
	In low spirits: low or very low spiritsStrain:	2 = a little like me
	Strain, stress, pressure	3 = somewhat like me
	Depressed: felt depressed	4 = mostly like me
	Blue: downhearted and blue	5 = very much like me
Race	What is your race?	0 = non-White
		1 = White
Dependent variable
Marijuana use (High school, Collegiate year 2, Collegiate year 4)	Marijuana use in the past 12 months	0 = nonuser
		1 = user

Analysis Plan

A four-step analysis plan was employed. First, Pearson correlations were conducted among independent and dependent variables. Second, mean comparisons were conducted, which focused on the mean differences between CSA victims and nonvictims across all independent and dependent variables. Third, binary logistic regression analyses were conducted to test the effects of CSA as well as religiosity on marijuana use across waves. Binary logistic regression is considered appropriate when the dependent variable is dichotomous. Finally, if CSA was found to be predictive of a higher level of marijuana use at certain waves in the regression models, a two-way ANOVA will be utilized to test for the moderating effect of religiosity on marijuana use (Kim, Kaye, & Wright, 2001).

Results

Bivariate Correlations

The correlation matrix presents a general picture of the relationships between independent and dependent variables. As shown in Table 2, church attendance and religious attitude in high school and the fourth collegiate year were positively correlated. Likewise, religiosity (church attendance and religious attitude) were positively correlated between the second and fourth collegiate years. Marijuana use in high school was also positively correlated with the second and fourth collegiate years’ marijuana use. In addition, childhood sexual victimization was positively correlated with adolescent as well as collegiate sexual victimization. Church attendance in high school and the fourth collegiate years was negatively correlated with depression. Depression (in high school and the second collegiate year) was positively correlated with victimization status and marijuana use. High correlations were observed among some variables (e.g., church attendance Collegiate Year 4 and religious attitude Collegiate Year 4). Variance inflation factor (VIF) measure was further used to detect multicollinearity for the independent variables included in each regression model. It was found that the VIF values ranged from 1.01 to 2.01 among all independent variables. Thus, multicollinearity is not a concern.

Table 2.

Correlation Matrix

Variables		1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
Church attendance HS	1	1.00
Religious attitude HS	2	.58**	1.00
Church attendance Y2	3	−.48**	−.32**	1.00
Religious attitude Y2	4	.53**	.63**	.07*	1.00
Church attendance Y4	5	.66**	.57**	.09*	.58**	1.00
Religious attitude Y4	6	.50**	.65**	.09*	.74**	.69**	1.00
CSA	7	−.03	−.02	−.04	−.02	.04	−.01	1.00
Victimization HS	8	−.06	−.02	−.03	−.08*	−.08*	−.10**	.22**	1.00
Victimization Y1	9	−.01	−.09*	.01	−.10**	−.11**	−.09*	.12**	.32**	1.00
Victimization Y2	10	−.03	−.03	−.04	−.09**	−.06	−.05	.16**	.27**	.30**	1.00
Victimization Y3	11	−.03	−.04	−.03	−.09**	−.11**	−.05	.10**	.20**	.23**	.36**	1.00
Victimization Y4	12	−.01	−.09*	−.01	−.06	−.09*	−.02	.09*	.14**	.24**	.24**	.34**	1.00
Depression HS	13	−.12**	−.03	−.01	.07*	−.15**	−.05	.09**	.15**	.11**	.13**	.13**	.08*	1.00
Depression Y2	14	−.05	−.02	.02	−.06*	−.15**	−.09*	−.01	.20**	.10**	.11**	.13**	.11**	.36**	1.00
Depression Y4	15	−.14**	−.13**	.04	−.13**	−.16**	−.13**	.01	.08*	.12**	.11**	.20**	.15**	.35**	.47**	1.00
Race	16	−.13**	−.10**	.04	−.12**	−.21**	−.16**	−.09**	.02	.02	−.03	−.05	−.03	.10**	.04	−.01	1.00
Marijuana use HS	17	−.24**	−.28**	−.02	−.24**	−.30**	−.28**	.13**	.21**	.10**	.11**	.07*	.01	.14**	.08*	.04	.09**	1.00
Marijuana use Y2	18	−.27**	−.30**	−.01	−.33**	−.35**	−.34**	.05*	.17**	.14**	.17**	.08*	.02	.11**	.12**	.05	.16**	.52**	1.00
Marijuana use Y4	19	−.22**	−.28**	−.04	−.27**	−.31**	−.30**	.04	.09*	.07	.09*	.08*	−.02	.09*	.09*	.06	.12**	.47**	.54**	1.00

Note: CSA = childhood sexual abuse; HS = High school; Y1 = Collegiate Year 1; Y2 = Collegiate Year 2; Y3 = Collegiate Year 3; Y4 = Collegiate Year 4.

p<.05. **p<.01

Mean Comparison

Table 3 presents the results from the one-way ANOVA. CSA victims and nonvictims did not show a significant difference in their religious attitude and church attendance (in high school and in the second and fourth collegiate years). In the dimension of subsequent victimization, the F ratios associated with sexual victimization in high school and throughout the four collegiate years are all statistically significant, suggesting that CSA victims and nonvictims differed in the probability of their subsequent sexual victimization. In comparison with nonvictims, victims of CSA were more likely to be sexually victimized in high school and collegiate years. It was also found that CSA victims differed from nonvictims in the depression index, with victims demonstrating more depressive symptoms. However, the variation in the depression index was only found in high school; no significant difference in depression was found between CSA victims and nonvictims in the second and fourth collegiate years.

Table 3.

Descriptive Statistics and Mean Comparison by Victimization Status

	CSA victims		Nonvictims
	M	SD	M	SD	F ratio^a
Religious attitude (high school)	2.07	0.99	2.10	0.95	0.21
Church attendance (high school)	2.79	1.34	2.86	1.24	0.50
Religious attitude (Collegiate Year 2)	2.23	.97	2.28	0.97	0.65
Church attendance (Collegiate Year 2)	2.86	1.20	2.97	1.28	1.75
Religious attitude (Collegiate Year 4)	2.27	0.98	2.29	1.01	0.03
Church attendance (Collegiate Year 4)	2.66	1.24	2.55	1.18	1.10
Sexual victimization in high school	0.67	0.47	0.43	0.50	75.68****
Collegiate sexual victimization: Year 1	0.41	.49	.28	.45	20.47****
Collegiate Sexual Victimization: Year 2	0.38	0.49	0.23	0.42	28.18****
Collegiate sexual victimization: Year 3	0.33	0.47	0.23	0.42	9.58***
Collegiate sexual victimization: Year 4	0.32	0.47	0.23	0.42	6.19***
Depression index (high school)	2.51	1.07	2.30	0.99	13.21****
Depression index (Collegiate Year 2)	2.28	0.98	2.29	0.98	0.05
Depression index (Collegiate Year 4)	2.25	0.95	2.23	0.91	0.02
Marijuana use in high school	0.28	0.45	0.17	0.37	24.52****
Marijuana use (Collegiate Year 2)	0.29	0.45	0.24	0.43	3.00*
Marijuana use (Collegiate Year 4)	0.24	0.43	0.21	0.41	1.01

Note: CSA = childhood sexual abuse.

One-way ANOVA was used to compute the ratio.

p < .10. **p < .05. ***p < .01. ****p < .001 (one-tailed).

In addition, CSA victims and nonvictims differed significantly in marijuana use (strongly significant in high school and marginally significant in the second collegiate year but not in the fourth collegiate year).

Multivariate Regression

Table 4 summarizes the results from the multivariate regression analysis for marijuana use at different time periods.

Table 4.

Multivariate Regression of Marijuana Use

	Marijuana use (high school)		Marijuana use (Collegiate Year 2)		Marijuana use (Collegiate Year 4)
Variables	B^a (SE)	Exp(B)	B (SE)	Exp(B)	B (SE)	Exp(B)
Religious attitude (high school)	−0.732**** (.148)	0.481	—	—	—	—
Church attendance (high school)	−0.223** (.096)	0.800	—	—	—	—
Religious attitude (Collegiate Year 2)	—	—	−0.754**** (.108)	0.470	—	—
Church attendance (Collegiate Year 2)	—	—	0.010 (.067)	1.010	—	—
Religious Attitude (Collegiate Year 4)	—	—	—	—	−0.374** (.165)	0.688
Church attendance (Collegiate Year 4)	—	—	—	—	−0.468*** (.146)	0.626
Childhood sexual abuse	0.519** (.206)	1.681	0.088 (.198)	0.916	−0.088 (.269)	0.915
Sexual victimization in high school	0.710**** (.202)	2.034	0.122 (.190)	1.130	−0.239 (.253)	0.787
Collegiate sexual victimization: Year 1	—	—	0.296 (.190)	1.345	0.236 (.254)	1.267
Collegiate sexual victimization: Year 2	—	—	0.644*** (.196)	1.904	0.346 (.269)	1.414
Collegiate sexual victimization: Year 3	—	—	—	—	0.491* (.273)	1.634
Collegiate sexual victimization: Year 4	—	—	—	—	−0.489 (.286)	0.613
Marijuana use in high school	—	—	2.443**** (.200)	11.507	2.078****(.249)	7.991
Depression index (high school)	0.213** (.095)	1.238	—	—	—	—
Depression index (Collegiate Year 2)	—	—	0.173** (.086)	1.188	—	—
Depression index (Collegiate Year 4)	—	—	—	—	−0.025 (.124)	0.976
Race (White)	0.477** (.245)	1.610	0.933**** (.244)	2.542	0.384 (.298)	1.468
Constant	−1.527*** (.493)	0.217	−2.757**** (.552)	0.063	−0.719 (.715)	0.487
n	745		1,089		697
Nagelkerke R²	.214		.421		.367

Entries are unstandardized regression coefficients.

p < .08. **p < .05. ***p < .01. ****p < .001 (one-tailed).

High School Marijuana Use

As displayed in Table 4, Model 1 (marijuana use in high school) showed that both church attendance and religious attitude at intake (the year prior to the entry into college) exerted a significant effect on marijuana use in high school. A one-unit increase in church attendance was associated with a 20% reduce in the odds of being a marijuana user. Likewise, a one-unit increase in perceived importance of religion reduced the odds of being a marijuana user by about 52%. In addition, CSA, as well as sexual victimization in high school, were significantly related to a higher likelihood of marijuana use. Individuals with CSA victimization were 68% more likely to use marijuana compared with those without such victimization experience. Those who experienced sexual victimization in high school were twice more likely to use marijuana. The depression index was significantly and positively associated with marijuana use. A one-unit increase in the depression index increased the odds of being a marijuana user by about 24%. The variable of race had a significant effect on marijuana use. Whites were more likely to use marijuana.

To further examine whether religiosity moderates the effect of CSA on marijuana use, 2 × 2 ANOVA (two-way ANOVA) was used to test for moderating effects (Kim et al., 2001). The two-way ANOVA provides main effect for the interaction between the predictors (religiosity¹ and CSA). As shown in Figure 1, the plotted graph reveals an interaction effect between religiosity and CSA. It indicated that CSA victims were more likely to use marijuana in high school compared with CSA nonvictims (0 = nonuser, 1 = user). However, as the CSA victims’ religiosity increased, the likelihood for them to use marijuana decreased accordingly. It was found from the two-way ANOVA analysis that the interaction effect term (between religiosity and CSA) is statistically significant (results not shown), which indicated that religiosity moderated the impact of CSA on marijuana use in high school.

Figure 1.

Estimated marginal means of marijuana use in high school

Marijuana Use in College

Model 2 presented the multivariate regression of marijuana use in the second collegiate year. Consistent with some of the findings in the first model, religious attitude, depression, high school marijuana use, and race were significantly associated with marijuana use in the second collegiate year. In addition, high school marijuana use was a significant predictor of the second collegiate year marijuana use. Nevertheless, neither CSA nor adolescent sexual victimization had a significant effect on the second collegiate year marijuana use. Instead, the more proximate sexual victimization in college significantly predicted the concurrent marijuana use. Individuals who experienced sexual victimization in the second collegiate year were 90% more likely to use marijuana compared with those without such victimization.

Model 3 displays the multiple regression of marijuana use in the fourth collegiate year. Consistent with the findings in the second model, the more recent sexual victimization in college overrode the impacts of childhood and adolescent victimization on the fourth collegiate year of marijuana use. The third collegiate year victimization was significantly (marginal significant; at the 0.07 level) and positively associated with marijuana use. With regard to religiosity, both church attendance and religious attitude were significantly predictive of marijuana use in college. A one-unit increase in the frequency of church/synagogue attendance reduced the log odds of being a marijuana user by approximately 37%. Likewise, a one-unit increase in perceived importance of religion reduced the odds of being a marijuana user by about 31%. High school marijuana use remained as a significant predictor of collegiate marijuana use, which is coincident with the first and the second models.

Discussion and Conclusions

The findings partially support the first hypothesis and lend full support to the second hypothesis. First, childhood sexual victimization, as well as adolescent victimization, was associated with a higher level of marijuana use in high school. It is noteworthy that more recent sexual victimization seemed to override the impacts of childhood and adolescent sexual victimization on subsequent marijuana use. Second, religiosity was found to be associated with a lower level of marijuana use. It was found that religiosity, either religious attitude (in the second collegiate year) or both of the religious attitude and church attendance (in high school and the fourth collegiate year), exerted an impact on a reduced level of concurrent marijuana use.

The findings from this study suggest that religiosity moderates the effects of CSA on subsequent marijuana use (marijuana use in high school). It was found that both religious attitude and church attendance in high school and the fourth colligate year were correlated with a reduced level of marijuana use. Religious attitude in the second collegiate year was also associated with a lower level of marijuana use. It is likely that perceived importance of religion may somehow exert a buffering effect of sexual victimization on subsequent marijuana use. Nevertheless, in the long run, religious attitude (belief) and religious behavior (church attendance) are equally important to make the protective effect remain.

Although religiosity is found to be influential to subsequent marijuana use, it is important to look into aspects of how religiosity impacts individuals who experienced the strain of sexual victimization. Is it a relationship with a God with whom they can rely on and gain a feeling of acceptance from, a relationship which helps victims to let go of the painful past by forgiving the perpetrators (Doxey et al., 1997), or the social network through religious context, that provides emotional support? In-depth interviews may further explore how different aspects of religiosity impact individuals who experienced sexual victimization.

Although this study provides valuable information, some of the limitations must be noted. First, the attrition of the respondents in the follow-up may place some constraints on the generalizability of our findings. Although no significant difference was found between the attrition and remaining sample in terms of ethnicity and age (White et al., 2001), there is the possibility that the resulting analyses may underrepresent high-rate marijuana users (Chu & Sung, 2009). Studies indicate that individuals who drop out of follow-up interviews at longitudinal studies disproportionately are higher rate drug users (Brame & Paternoster, 2003; Pogarsky, Kim, & Paternoster, 2005). Second, the current data did not allow for an examination of whether different denominations of religion exert differential effects on individuals’ subsequent substance use. As Amoateng and Bahr (1986) suggested, the impact of religion on drug use might vary by denomination. Future research should assess the interactive effects of sexual victimization and the affiliation with certain denominations that are restrictive on drug and alcohol use. Third, as the sample was taken from a state university in the southeastern region, it should be cautioned to generalize the findings of this study to a larger population. Furthermore, future survey design can include questions that measure different dimensions of religious support as well as in-depth interviews to unravel which dimensions of religious support (the relationship with God, praying, confession, or reading Bible) that may help individuals who experienced sexual victimization to cope with stress and reduce subsequent drug use. In addition, future studies can include male samples to further examine whether religiosity exerts the same effects on males who experienced childhood sexual victimization. Moreover, in the current data set, there are no available questions regarding socioeconomic status, peer associations, and parental monitoring. Future data collection which includes variables of socioeconomic status, peers, and family should help researchers further explore the role of socioeconomic factors and whether peers and parental supervision exert influences on individuals’ subsequent marijuana use.

With such a high rate of CSA, it is important to identify the variables that could operate as protective factors that may offset the negative impacts of sexual victimization. Although many research efforts have been directed to investigating factors that may lessen the negative psychological effects of CSA, the role of religiosity has seldom been explored. A large body of research in epidemiology, public health, psychology, and behavioral science has found that religiosity has beneficial effects on diverse dimensions of mental health outcomes, such as depression, drug and alcohol use, suicide, psychological distress, personal well-being, and overall adaptation (e.g., Ellison et al., 2001; Payne, Bergin, Bielema, & Jenkins, 1991).

It was found in this study that religiosity moderates the effect of CSA victimization on marijuana use in high school. Religiosity may be a promising dimension that may moderate the negative effects and hurt associated with sexual victimization. The effect of religiosity is especially prominent on deterring subsequent drug use. As Banyard (2003) suggested, spiritual/religious coping in relation to the adjustment of CSA victims is one of the new aspects that has not been investigated (Gall, 2006). The role that religiosity may play in assisting victims of sexual abuse in coping with strains or may be utilized in interventions warrants further examination.

Footnotes

Acknowledgements

The author is grateful to the editor and anonymous reviewers for their helpful comments.

Author’s Notes

An earlier version of this article was presented at the 2010 annual meeting of Academy of Criminal Justice Sciences in San Diego.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Agnew

(1992). Foundation for a general strain theory of crime and delinquency. Criminology, 30, 47-87.

Agnew

(2002). Experienced, vicarious, and anticipated strain: An exploratory study on physical victimization and delinquency. Justice Quarterly, 19, 603-632.

Amoateng

A. Y.

Bahr

(1986). Religion, family, and adolescent drug use. Sociological Perspectives, 29, 53-76.

Angst

Gamma

Gastpar

Lépine

J. P.

Mendlewicz

Tylee

(2002). Gender differences in depression. European Archives of Psychiatry and Clinical Neuroscience, 252, 201-209.

Arellano

C. M.

Kuhn

J. A.

Chavez

E. L.

(1997). Psychosocial correlates of sexual assault among Mexican American and White non-Hispanic adolescent females. Hispanic Journal of Behavioral Sciences, 19, 446-459.

Bailey

J. A.

McCloskey

L. A.

(2005). Pathways to adolescent substance use among sexually abuse girls. Journal of Abnormal Child Psychology, 33, 39-53.

Banyard

V. L.

(2003). Explaining links between sexual abuse and psychological distress: Identifying mediating processes. Child Abuse & Neglect, 16, 101-118.

Bjorck

J. P.

Thurman

J. W.

(2007). Negative life events, patterns of positive and negative religious coping, and psychological functioning. Journal for the Scientific Study of Religion, 46, 159-167.

Brame

Paternoster

(2003). Missing data problems in criminological research: Two case studies. Journal of Quantitative Criminology, 19, 55-78.

10.

Briere

J. N.

Elliott

D. M.

(1994). Immediate and long-term impacts of child sexual abuse victims. Future of Children, 4, 54-69.

11.

Broidy

Agnew

(1997). Gender and crime: A general strain theory perspective. Journal of Research in Crime and Delinquency, 34, 275-306.

12.

Chu

D. C.

(2007). Religiosity and desistance from drug use. Criminal Justice and behavior, 34, 661-679.

13.

Chu

D. C.

Sung

H. E.

(2009). Racial differences in desistance from substance abuse: The impact of religious involvement on recovery. International Journal of Offender Therapy and Comparative Criminology, 53, 696-716.

14.

Doxey

Jensen

(1997). The influence of religion on victims of childhood sexual abuse. International Journal of the Psychology of Religion, 7, 179-186.

15.

Dube

S. R.

Anda

R. F.

Whitefield

C. L.

Brown

D. W.

Felitti

V. J.

Dong

Giles

W. H.

(2005). Long-term consequences of childhood sexual abuse by gender of victim. American Journal of Preventive Medicine, 28, 430-438.

16.

Ellison

C. G.

(1991). Religious involvement and subjective well-being. Journal of Health and Social Behavior, 30, 92-104.

17.

Ellison

C. G.

Boardman

J. D.

Williams

D. R.

Jackson

J. S.

(2001). Religious involvement, stress, and mental health: Findings from the 1995 Detroit Area Study. Social Forces, 80, 215-249.

18.

Fergusson

D. M.

Horwood

L. J.

Lynskey

(1994). The childhoods of multiple problem adolescents: A 15-year longitudinal study. Journal of Child Psychology and Psychiatry, 35, 1123-1140.

19.

Fiala

W. E.

Bjorck

J. P.

Gorsuch

(2002). The religious support scale: Construction, validation, and cross-validation. American Journal of Community Psychology, 30, 761-786.

20.

Gall

T. L.

(2006). Spirituality and coping with life stress among adult survivors of childhood sexual abuse. Child Abuse & Neglect, 30, 829-844.

21.

Gall

T. L.

Basque

Damasceno-Scott

Vardy

(2007). Spiritual and the current adjustment of adult survivors of childhood sexual abuse. Journal for the Scientific Study of Religion, 46, 101-117.

22.

George

L. K.

Ellison

C. G.

Larson

D. B.

(2002). Explaining the relationship between religious involvement and health. Psychological Inquiry, 13, 190-200.

23.

Gold

S. R.

Sinclair

B. B.

Balge

K. A.

(1999). Risk of sexual revictimization: A theoretical model. Aggression and Violent Behavior, 4, 457-470.

24.

Heffernan

Cloitre

Tardiff

Marzuk

P. M.

Portera

Leon

A. C.

(2000). Childhood trauma as a correlate of lifetime opiate use in psychiatric patients. Addictive Behavior, 25, 797-803.

25.

Hodge

D. R.

Cardenas

Montoya

(2001). Substance use: Spiritual and religious participation as protective factors among rural youths. Social Work Research, 25, 153-161.

26.

Hood

R. W.

Jr. Spilka

Hunsberger

Gorsuch

(1996). The psychology of religion: An empirical approach (2nd ed.). New York, NY: Guilford.

27.

Humphrey

J. A.

White

J. W.

(2000). Women’s vulnerability to sexual assault from adolescence to young adulthood. Journal of adolescent health, 27, 419-424.

28.

Jang

S. J.

(2007). Gender differences in strain, negative emotions, and coping behaviors: A general strain theory approach. Justice Quarterly, 24, 523-553.

29.

Jang

S. J.

Johnson

B. R.

(2001). Neighborhood disorder, individual religiosity, and adolescent use of illicit drugs: A test of multilevel hypotheses. Criminology, 39, 109-143.

30.

Jang

S. J.

Johnson

B. R.

(2003). Strain, negative emotions, and deviant coping among African Americans: A test of general strain theory. Journal of Quantitative Criminology, 19, 79-105.

31.

Jang

S. J.

Johnson

B. R.

(2005). Gender, religiosity, and reactions to strain among African Americans. Sociological Quarterly, 46, 323-357.

32.

Jang

S. J.

Johnson

B. R.

(2007). Neighborhood disorder, individual religiosity, and adolescent use of illicit drugs: A test of multilevel hypothesis. Criminology, 39, 109-143.

33.

Jarvis

T. J.

Copeland

Walton

(1998). Exploring the nature of the relationship between child sexual abuse and substance use among women. Addition, 93, 865-875.

34.

Kilpatrick

D. G.

Acierno

Saunders

B. E.

Resnick

H. S.

Best

C. L.

Schnurr

P. P.

(2000). Risk factors for adolescent substance abuse and dependence: Data from a national sample. Journal of Consulting and Clinical Psychology, 68, 19-30.

35.

Kim

J. S.

Kaye

Wright

L. K.

(2001). Moderating and mediating effects in causal models. Issues in Mental Health Nursing, 22, 63-75.

36.

Luster

Small

S. A.

(1997). Sexual abuse history and problems in adolescence: Exploring effects of moderating variables. Journal of Marriage and the Family, 59, 131-142.

37.

Messman-Moore

T. L.

Long

P. J.

(2002). Alcohol and substance use disorders as predictors of child to adult sexual revictimization in a sample of community women. Violence and Victims, 17, 319-340.

38.

Miller

B. A.

Downs

W. R.

(1993). The impact of family violence on the use of alcohol by women. Alcohol Health and Research World, 17, 137-143.

39.

Paone

Chavkin

Willets

Friedman

Des Jarlais

(1992). The impact of sexual abuse: Implications for drug treatment. Journal of Women’s Health, 1, 149-153.

40.

Pargament

K. I.

(1997). The psychology of religion and coping. New York, NY: Guilford.

41.

Pargament

K. I.

Brant

C. R.

(1998). Religion and coping. In Koenig

H. G.

(Eds.), Handbook of religion and mental health (pp. 111-128). San Diego, CA: Academic Press.

42.

Payne

I. R.

Bergin

A. E.

Bielema

K. A.

Jenkins

P. H.

(1991). Review of religion and mental health: Prevention and the enhancement of psychological functioning. Prevention in Human Services, 9, 11-40.

43.

Pogarsky

Kim

Paternoster

(2005). Perceptual change in the National Youth Survey: Lessons for deterrence theory and offender decision-making. Justice Quarterly, 22, 1-29.

44.

Roche

D. N.

Runtz

M. G.

Hunter

M. A.

(1999). Adult attachment: A mediator between child sexual abuse and later psychological adjustment. Journal of Interpersonal Violence, 14, 184-207.

45.

Rohsenow

D. J.

Corbett

Devine

(1988). Molested as children: A hidden contribution to substance abuse? Journal of Substance Abuse Treatment, 5, 13-18.

46.

Rossetti

S. J.

(1995). The impact of child sexual abuse on attitudes toward God and the Catholic Church. Child Abuse & Neglect, 19, 1469-1481.

47.

White

J. W.

Smith

P. H.

Humphrey

J. A.

(2001). Longitudinal study of violence against women: Victimization and perpetration among college students in a state-supported university in the United States, 1990-1995 [Computer file] ICPSR 03212-v1. Greensboro: University of North Carolina.

48.

Widom

C. S.

Ames

M. A.

(1994). Criminal consequences of childhood sexual victimization. Child Abuse & Neglect, 18, 303-318.

49.

Williams

D. R.

Larson

D. B.

Buckler

R. E.

Heckman

R. C.

Pyle

C. M.

(1991). Religion and psychological distress in a community sample. Social Science & Medicine, 32, 1257-1262.

50.

Wilsnack

S. C.

Vogeltanz

N. D.

Klassen

A. D.

Harris

T. R.

(1997). Childhood sexual abuse and Women’s substance abuse: National Survey Findings. Journal of Study of Alcohol, 58, 264-271.