The Influence of Religiosity, Gender, and Language Preference Acculturation on Sexual Activity Among Latino/a Adolescents

Richard J. Fehring
and Lisa M. Edwards, Keyona M. Jarrett, Kristin A. Haglund
Nursing Faculty Research and Publications Nursing, College of
Marquette University
e-Publications@Marquette
11-1-2008
Reproduced with Permission

The purpose of this study was to determine the main and interactive effects of religiosity, gender, and language preference acculturation on sexual activity among 570 Latino/a adolescents from the 2002 National Survey of Family Growth. Results indicated that adolescents who viewed religion as very important, had frequent church attendance, and had more traditional attitudes on sexuality were less likely ever to have sex compared with adolescents who were less religious. Those with frequent church attendance and high traditional attitudes had fewer lifetime and recent sex partners. Unassimilated religious youth were less likely ever to have sex, had fewer lifetime and recent sexual partners, and a later age of sexual debut. Females were less likely to have had sex, had fewer recent and lifetime partners, and had a later age of coital debut than males. Religiosity had a healthy dampening of sexual activity among Latino/a adolescents and, in particular, among those who were less assimilated.

Early sexual activity among adolescents bears variable consequences in physical, psychological, and emotional domains. Consequences such as sexually transmitted infections and teenage pregnancy have been identified as concerns for youths (Forste & Heaton, 1988; Holder et al., 2000), and the effects of childbearing on girls may include poverty, mental illness, under-performance in school, sexual violence and abuse, family dysfunction, and poor self-esteem (Contreras, Narang, Ikhlas, & Teichman, 2002). Even with these potential consequences, adolescents continue to engage in sexual behaviors at early ages. National representative data from the Youth Risk Behavior Survey revealed that 63% of youths have had sex by 12th grade (Eaton et al., 2006).

Latino/a adolescents report high rates of sexual activity. Among high school students, 51% of Latino/a youth reported engaging in sexual activity, compared with 43% of Caucasian youth (Eaton et al., 2006). By the 12th grade, 75% of Latino males and 64% of Latina females reported having had sex. Also, among these high school students, 22% of Latinos and 10% of Latinas reported having had four or more sexual partners (Centers for Disease Control and Prevention, 2007).

The high rates of sexual behavior and related consequences have prompted researchers to explore factors that may influence sexual behavior among Latino/a adolescents, such as religiosity (DuRant, Pendergrast, & Seymore, 1990; Forste & Heaton, 1988; Holder et al., 2000; Liebowitz, Castellano, & Cuellar, 1999), gender (Liebowitz et al., 1999; Upchurch, Aneshensel, Mudgal, & McNeely, 2001; Vlez-Pastrana, Gonzalez-Rodriguez, & Borges-Hernandez, 2005), and acculturation (Afable-Munsuz & Brindis, 2006; Driscoll, Biggs, Brindis, & Yankah, 2001; Upchurch et al., 2001). However, researchers have yet to examine the relationships among all of these factors and the sexual behaviors of Latino/a adolescents in a nationally representative sample.

Religiosity and Sexual Behavior

Religion has been broadly defined as a set of institutionalized beliefs in a higher power and includes statutes or regulations for how to guide ones life (Holder et al., 2000). Previous research has identified relationships between religiosity and sexual behaviors. Adolescents who attended church more frequently had less sexual experience (Forste & Heaton, 1988; Sinha, Cnaan, & Gelles, 2007; Thornton & Camburn, 1989) and had fewer sexual partners (Lefkowitz, Gillen, Shearer, & Boone, 2004) than those who had less frequent church attendance. Youths who reported high levels of religious importance, strong religious influence in their daily lives, and adherence to their religions teachings were less likely to initiate sexual activity than their peers (Holder et al., 2000; Lefkowitz et al., 2004; Thornton & Camburn, 1989). For males and females, each unit increase in religiosity (composite of importance of religion and attendance at religious services and youth activities) reduced the likelihood of sexual initiation by 12% to 16% (Rostosky, Regnerus, & Wright, 2003). Most research supports that adolescents who are more religious tend to delay sexual activity (Rostosky, Wilcox, Wright, & Randall, 2004), with the exception of a handful of studies that report no relationship (Benson & Torpy, 1995; Sheeran, Abrams, Abraham, & Spears, 1993) or a negative relationship (Sinha et al., 2007) between religiosity and sexual activity.

Researchers have reported that religious teens tend to endorse traditional attitudes on sexuality (Koenig, 2004; Lefkowitz et al., 2004; Sheeran et al., 1993; Thornton & Camburn, 1989). However, the relationship between sexual attitudes and sexual behaviors is not clear. It has been reported that traditional sexual values were related to less sexual activity (Thornton & Camburn, 1989) or unrelated to sexual behaviors (Sheeran et al., 1993). In addition to these contradictions in the literature, few studies have focused specifically on Latino/a youth. More research examining the relationship between religiosity and sexual activity among Latino/a adolescents is therefore needed.

Gender and Acculturation Influences on Sexual Behavior

There is some evidence that Latino adolescent males and females differ in their sexual behavior. Several studies have found that Latino adolescents tend to be more sexually active than their Latina counterparts (Liebowitz et al., 1999; Upchurch et al., 2001; Vlez-Pastrana et al., 2005). The tendency for males to be more sexually active than females may be a manifestation of Latino cultural values, such as machismo and marianismo, which prescribe traditional gender roles (Schuster, 2003). Machismo is a concept that describes the male role as dominant, independent, and protective of the family, whereas marianismo refers to the female role as a caregiver, a virgin, and obedient to men (Gloria, Ruiz, & Castillo, 2004; Upchurch et al., 2001). Men are allowed sexual freedom, while women are expected to maintain their virginity (Schuster, 2003).

The influence of machismo and marianismo in the lives of Latino adolescents may depend on their level of acculturation. Acculturation has been defined as the process by which one is influenced by the host culture and ones own culture of membership (Berry, 2003). Measures of acculturation are both attitudinal and behavioral. Common indicators include cultural identification, pride, and acceptance; language preference and proficiency; traditions and customs; styles of communication; and social affiliations (Zane & Mak, 2003). Research suggests that as individuals become more assimilated to White, majority culture, for example, their cultural values may change to reflect those of the dominant culture. Regarding gender roles, some Latino/a adolescents may adhere strongly to these values, whereas others may have assimilated to a degree and abandoned their cultural values of machismo and marianismo for more egalitarian gender roles (Marn & Gamba, 2003).

Researchers have suggested that some traditional cultural values may serve as protective factors against risky sexual behavior (Schuster, 2003). For example, as Latina adolescent females become more assimilated, they may no longer prescribe to marianismo and thus may be more likely to engage in risky sexual behaviors. Multiple studies have indicated that less assimilated Latino male and female adolescents are significantly less likely to have had sex than their more assimilated peers (Afable-Munsuz & Brindis, 2006; Jimenez, Potts, & Jimenez, 2002; Upchurch et al., 2001). When asked about their reasons for abstaining, less assimilated Latinas more frequently reported waiting until marriage than highly assimilated Latinas (Jimenez et al., 2002).

The Current Study

This study was designed to investigate the influence of religiosity, gender, and acculturation on the sexual activity of Latino adolescents in order to provide a more comprehensive understanding of their sexual behavior. There were several ways in which the current study extended existing research. Many previous studies used one-item measures of sexual activity, such as ever had sex (Durant et al., 1990; Holder et al., 2000; Rostosky et al., 2003), age at sexual debut (Forste & Heaton, 1988; Manlove, Terry-Humen, Ikramullah, & Moore, 2006), or frequency of sexual activity (Liebowitz et al., 1999). The current study sought to measure teen sexual activity more comprehensively by including four indicators of sexual activity.

In addition, many previous studies are limited in their measurement of religiosity (Durant et al., 1990; Lefkowitz et al., 2004; Manlove et al., 2006; Vlez-Pastrana et al., 2005). The influence of religiosity on attitudes about sexuality and how those attitudes influence behavior were not often considered. Thus, the current study extended existing research by including traditional attitudes on sexuality as a measure of religiosity. Finally, this study further developed the current literature by exploring the influences of religion, gender, and acculturation on the sexual activity of Latino males and females into their early 20s utilizing a nationally representative sample.

This research was informed by several theoretical perspectives, including reference group theory (Lefkowitz et al., 2004; Thornton & Camburn, 1989) and the protection-risk conceptual framework (Jessor et al., 2003). Reference group theory, which posits that one adheres to the doctrine of ones religion when determining attitudes and behaviors, would suggest that Latino youth will reference their religions principles concerning delaying or abstaining from sexual behavior. A protection-risk conceptual framework suggests that religiosity can be considered a protective factor against early sexual activity, because it provides models of positive behavior, has sanctions against problem behavior, and often fosters a supportive environment (Jessor et al., 2003).

Based on theory and existing empirical research the following hypotheses were formulated: (a) there will be less sexual activity among Latino/a adolescents who are more religious and less assimilated when compared with those who are less religious and more assimilated, (b) Latino male adolescents will score higher on measures of sexual activity than Latinas, and (c) an interaction between gender and acculturation will exist such that unassimilated Latina adolescents will score the lowest on measures of sexual activity, and assimilated Latino male adolescents will score the highest on measures of sexual activity for all groups.

Method

Participants

Participant data were from Cycle 6 of the National Survey of Family Growth (National Center for Health Statistics [NCHS], 2007), a periodic government survey about family, gender, and sexuality. Cycle 6 was conducted in 2002 by the NCHS. Male and female participants were selected from an area probability sample to yield a nationally representative sample, and there was about a 79% return rate. Data were gathered during confidential, in-person interviews, each lasting 60 to 90 minutes.

There were 7,365 female and 4,928 male participants in Cycle 6. Of these, 1,675 females and 1,494 males were between the ages of 15 and 21, of which 322 females (19.2%) and 302 males (20.2%) identified themselves as Hispanic. Fifty-four married participants were removed from the data set as it was unknown whether sexual debut occurred prior to their marriage. The final data set utilized in this study included 570 Latino/a adolescents (287 males and 283 females). The mean age of the total (N = 570) respondents was 17.91 years (SD = 1.98), with an age range of 15-22 years.1 The mean age of the 283 female respondents was 17.88 years (SD = 1.97), with an age range of 15-22 years; the mean age for the male respondents was 17.93 years (SD = 2.00), with an age range of 15-21 years.

Of the 3,456 variables, the authors analyzed several demographic variables (i.e., age, religion, and Hispanic origin) as well as the following dependent variables: (a) ever being sexually active, (b) number of sexual partners in a lifetime, (c) the number of sexual partners in the past 12 months, and (d) age of sexual debut.

The independent variable of religiosity was measured by importance of religion, frequency of attendance at religious services, and by a combination of four variables that reflected traditional attitudes on human sexuality. Importance of religion in the respondents daily life was originally recorded as: 1 (Very important), 2 (Somewhat important), and 3 (Not important). For analysis purposes these responses were dichotomized into two categories: 1 (Very important) and 2 (Not important), which included somewhat and not important. Original responses to frequency in religious attendance included: 1 (More than once a week), 2 (Once a week), 3 (1-3 times per month), 4 (Less than once a month), and 5 (Never). For analysis purposes, the responses were collapsed into two categories: (a) Frequent Church Attendance, which included more than once a week and once a week; and (b) Not Frequent Church Attendance (i.e., 1-3 times per month, less than once a month, and never). The independent variable of Traditional Attitudes on Sexuality (TAS) was measured by combining four items and summing a total score. The items were (a) sexual acts between two consenting adults is OK, (b) it is OK for unmarried 16 year olds to have sex if there is a strong affection, (c) it is OK for unmarried 18-year-olds to have sex if there is a strong affection, and (d) it is OK for an unmarried female to have a child. Response options included strongly agree, agree, not sure, disagree, or strongly disagree. A high total score indicated that the respondent possessed traditional attitudes on sexuality and a low score indicated permissive attitudes.

The variable of language preference acculturation was determined by the language of communication at the time of the interviews. Participants who preferred to be interviewed in English were considered assimilated into U.S. society and those who preferred Spanish were considered unassimilated. Although language preference at time of interview is not a comprehensive or direct indicator of acculturation, language spoken in the home has been used to assess acculturation (Guilamo-Ramos, Jaccard, Pena, & Goldberg, 2005; Nguyen, Messe, & Stollak, 1999).

Statistical Analyses

Means, standard deviations, frequencies, and percentages were used to describe the participants. The sample was grouped according to gender, language preference acculturation, and religiosity (religion is very important vs. religion is less important, frequent church attendance vs. less frequent church attendance, and holding traditional attitudes on sexual activity vs. holding permissive attitudes on sexuality). Chi-square and likelihood odds ratios (ORs) were used to determine the likelihood of ever having had sexual intercourse. Student t tests were used to determine the differences of sexual activity with the continuous dependent variables of number of partners in a lifetime and the number of partners in the past 12 months. To determine the influence of religiosity on the age of sexual debut, survival analysis was used (Kaplan & Meier, 1958). The censored event was sexual debut or not, and the time was the current age of the participant (if still a virgin) or the age of debut if the participant had initiated sexual activity. Interaction effects of gender and religiosity and language preference acculturation and religiosity on sexual activity were determined by a 2 &times; 2 univariate analysis of variance. The minimal level of statistical significance was set at p < .05.

Results

The frequency and percentage of the demographic variables by gender can be found in Table 1. The majority (57.1%) reported being of Mexican or Mexican American background, and 39.3% were of another Hispanic group. Most of the 570 participants listed Roman Catholic as their current religion (62.8%), with Protestant as the second most frequent choice (21.6%) and No Religion as the third most frequent choice (12.5%). Table 1 shows the breakdown for frequency of age, religion, Hispanic-origin group, primary language, and education, as all variables were reported in the National Survey of Family Growth data set (NCHS, 2007), by gender.

Religiosity and Sexual Activity

Importance of religion. Adolescents who viewed religion as very important were 23.4% less likely ever to have had sex compared with adolescents who viewed religion as not important (OR = 0.766; 95% confidence interval [CI] = 0.621-0.945). Those who viewed religion as very important had significantly fewer lifetime sexual partners (mean = 1.6; SD = 2.16) than those who viewed religion as not important (mean = 2.3; SD = 3.02), with t = 3.07, p < .002. There was no significant difference in the number of sexual partners in the past 12 months. The cumulative survival of virgins at 21 years of age was 19.8% for those who viewed religion as very important compared with 4.6% for those who viewed religion as not important (see Table 2 for estimated ages of sexual debut).

Church attendance. Adolescents who attended church services at least once a week were 39.5% less likely ever to have had sex compared with adolescents who attended church less frequently or not at all (OR = 0.605; 95% CI = 0.475-0.772). Adolescents who had high church attendance had significantly fewer lifetime sexual partners (mean = 1.4; SD = 2.2 vs. mean = 2.4; SD = 2.4; t = 3.97, p < .001) and fewer sexual partners in the past 12 months (mean = 0.6; SD = 1.1 vs. mean = 1.1; SD = 1.4; t = 3.737, p < .001) compared with those with low attendance. The cumulative survival of virgins at age 21 for the high attendance group was 24% compared with 5.6% for the low attendance group (see Table 2).

Traditional attitudes on sexuality (TAS). Participants who held traditional attitudes on sexuality were 55% less likely to ever have had sex compared with adolescents who had less traditional attitudes (OR = 0.450; 95% CI = 0.349-0.581). Those who held TAS had significantly fewer lifetime sexual partners (mean = 1.4; SD = 2.2 vs. mean = 2.4; SD = 2.9; t = 4.11, p < .001) and fewer sexual partners in the past 12 months (mean = 0.5; SD = 0.9 vs. mean 1.0; SD = 0.9; t = 4.49, p < .001). The cumulative survival of virgins at age 21 for the high TAS group was 45% compared with 7.2% for the low TAS group (see Table 2).

Gender, Religiosity, and Sexual Activity

Latinas were 19.2% less likely ever to have had sex than Latinos (OR = 0.808; 95% CI = 0.686-0.952). Females also had significantly fewer lifetime sexual partners (mean = 1.6; SD = 2.7 vs. mean = 2.5; SD = 2.7; t = 3.978, p < .001) and fewer partners in the past 12 months than their male peers (mean = 0.7; SD = 1 vs. mean = 1; SD = 1.5; t = 2.990, p < .003). Mean age in years at coital debut was 13 for males (N = 124) and 15 for females (N = 149; t = 7.75, p = .00). In addition, the cumulative survival of virgins at the age of 21 was 28.9% for females and 6.1% for males (see Table 2).

There was no significant interaction effect between gender and sexual activity with regard to importance of religion and TAS. There was a significant interaction between gender, sexual activity, and church attendance. Females were more likely to attend church at least once a week and have fewer sexual partners in the past 12 months compared with the male adolescents.

Language Preference Acculturation, Religiosity, and Sexual Activity

There was no difference in the likelihood of ever having had sex between the English language acculturation (assimilated) versus Spanish language acculturation (nonassimilated) adolescents. However, the nonassimilated adolescents had significantly fewer lifetime sexual partners (mean = 1.3; SD = 1.7 vs. mean = 2.2; SD = 2.9; t = 2.812, p < .005) and fewer sexual partners in the past 12 months (mean = 0.6; SD = 0.7 vs. mean = 0.9; SD = 1.4; t = 2.554, p < .011) compared with the assimilated adolescents. The cumulative rate of remaining a virgin at age 21 was 12% for the nonassimilated adolescents and 8% for the assimilated adolescents (see Table 2).

There was a significant interaction effect of language preference acculturation and importance of religion, church attendance, and TAS on the variables of number of sexual partners in a lifetime, number of sexual partners in the past 12 months, and age of sexual debut (see Table 3). Unassimilated adolescents who held religion as very important, attended church at least once a week, and had traditional attitudes on sexuality were less likely to ever have had sex, more likely to have had fewer sexual partners in the past 12 months and in their lifetime, and to have had a later age at coital debut.

Discussion

The results of this study about religiosity and sexual activity in Latino/a youth support a growing body of research that indicates that increased religiosity decreases sexual risk for youths (Durant et al., 1990; Forste & Heaton, 1988; Holder et al., 2000; Liebowitz et al., 1999). Religiosity had a robust influence on helping youths to avoid sexual activity, thus providing support for the studys first hypothesis. Specifically, the three indicators of religiosity, viewing religion as important, attending church services often, and holding traditional sexual attitudes, were independently related to increased sexual abstinence through age 21 and decreased lifetime partners. Frequent church attendance and holding TAS were also related to decreased partners in the past 12 months. These results may indicate that religious youths are internalizing religious doctrine and translating it into attitudes and behaviors as theorized by reference group theory (Lefkowitz et al., 2004; Thornton & Camburn, 1989). Though not explicitly measured in this study, it is possible that religious doctrine may be shared and reinforced within a church-based social network of adolescent peers. In one study, youths who perceived social support from their spiritual friends were more likely to be virgins (Holder et al., 2000). Similarly, Bazargan and West (2006) found that those Hispanic and African American youths who experienced less peer pressure to engage in sexual activity and perceived a lower level of peer norms for engaging in risky behaviors were more likely to remain sexually inactive.

In the current study, Latina adolescents reported increased rates of abstinence and higher survival rates through age 21 and fewer partners in the past 12 months and in their lifetimes compared with their male counterparts. This provided support for the studys second hypothesis about gender, and was similar to previous research in which Latino males reported increased rates of sexual initiation and more partners than females (Driscoll et al., 2001; Eaton et al., 2006). In the present study, both males and females reported early ages at coitarche, 13 and 15 years, respectively, although males were significantly younger. National data from the Youth Risk Behavior Survey revealed similar gender differences, in that 11% of Latino males and 4% of Latina females reported coitarche before age 13, and by ninth grade, 51% of Latino males and 31% of Latina girls reported having had sex (Centers for Disease Control and Prevention, 2007).

Results in this study provided support for the third hypothesis. Adolescents who indicated Spanish as their preferred language had significantly fewer lifetime sexual partners, fewer sexual partners in the past 12 months, and higher rates of virginity at age 21 compared with those adolescents who preferred to be interviewed in English. These results indicate that less assimilation may also be protective for youth in terms of avoiding sexual risks. Results of previous studies have found that highly assimilated Latino adolescents are more likely to have had sex than less assimilated Latino adolescents (Afable-Munsuz & Brindis, 2006; Jimenez et al., 2002; Upchurch et al., 2001).

Taken together, these findings suggest that providing support for youths to maintain their ethnic cultural heritage may help them to more readily abstain from intercourse and have fewer partners. Researchers have begun to test culturally tailored HIV prevention interventions and have reported significant outcomes including decreased sexual activity, decreased number of partners, increased abstinence, and increased use of condoms (Crepaz et al., 2007; Villarruel, Jemmott, & Jemmott, 2006). These efforts should continue to be explored.

It is important to note limitations in the current study, as well as areas for future research. First, although language preference has been used frequently as a marker of acculturation (Guilamo-Ramos et al., 2005; Nguyen et al., 1999), acculturation is a more complex, multidimensional construct that goes beyond preferred language (Zane & Mak, 2003). It is possible, therefore, that our study did not capture the true level of acculturation of participants, and it is important for future studies to consider using a more comprehensive measure.

In addition, our indicators of religiosity may not have accounted for the complexity of religious faith, attendance, and beliefs. Although it is noteworthy that we included items about importance, attendance, and traditional beliefs, respondents views on religion or the nature of their personal beliefs were not assessed. It would be helpful for future research to consider including a validated measure of religious faith rather than just individual items.

The results of this study contribute to a growing body of knowledge about protective influences in the lives of Latino youth with respect to delaying sexual activity. The findings extend the literature by investigating the influence of religion, gender, and acculturation on the sexual activity of Latino adolescents into their early 20s using a nationally representative sample. Future studies that identify trends in adolescent attitudes and behaviors over time will help to further extend the field, as will continued research about gender and cultural influences that help prevent or delay early and risky sexual behaviors among Latino youth.


Note

1 Although our age limit was 21 years, one of the female participants provided two different ages (21 and 22) during the interview. [Back]

References



Appendix

Table 1


Table 2


Table 3

Top