Introduction

The consequences of teen pregnancy have significant social, health, and economic costs. Teen mothers are more likely to drop out of high school, live in poverty, be unemployed or underemployed, receive public assistance, and enter violent relationships (Holcombe et al. 2009; Hotz et al. 1997; Perper et al. 2010). Children born to teen parents are more likely to have low birth weight and long-term health problems, be born prematurely, drop out of school, become teen parents themselves, and among males, more likely to spend time in prison (Holcombe et al. 2009; Hotz et al. 1997; Perper et al. 2010). The estimated cost for teen pregnancies and subsequent healthcare costs to US taxpayers in 2010 was a staggering $9.4 billion (The National Campaign to Prevent Teen and Unplanned Pregnancy 2011).

US teen births have reached historically low rates since peaking in 1991. The most current data for birth rates for teens aged 15–19 were 24.2 births per 1000 (Hamilton et al. 2015). This rate represents a 61% decline since 1991 (61.8 births per 1000) and a 9% decline since 2013 (26.5 births per 1000) alone (Hamilton et al. 2015). However, teen birth rates are racially and ethnically disparate. Birth rates among African-American and Hispanic youth are twice as high as birth rates among white youth (Hamilton et al. 2015).

Declines in teen pregnancy and birth rates have been attributed to increased use and more effective use of contraceptives (Boonstra 2014). Trend data highlight a positive shift, suggesting that teens are using more effective contraceptives (Centers for Disease Control and Prevention 2012; Romero et al. 2015). Rates of contraceptive use have increased for teens of all races and ethnicities who are using highly effective methods (long-acting reversible contraception, pill, patch, ring, injectable contraception, sterilization) over the past 20 years (Centers for Disease Control and Prevention 2012). Comparing contraceptive rates in 1995 to rates during 2006–2010, white youth displayed the largest relative increase of 34% (48.5–65.7%) (Centers for Disease Control and Prevention 2012). Smaller increases were found for Hispanic (19% relative increase) and African-American (4% relative increase) youth (Centers for Disease Control and Prevention 2012). Therefore, it is critically important for sexually active teens to know that contraception use is an effective strategy for reducing unplanned pregnancies when used consistently and correctly (Hamilton et al. 2015).

Unfortunately, teens are not consistent users of contraceptive methods. Only just more than half of teens (59%) report the use of condoms, one in five (19%) report the use of birth control pills, and very few (2%) report the use of long-acting reversible contraception (LARC) such as the intrauterine device (IUD) and implant during their last sexual encounter (Kann et al. 2014). The use of dual contraceptive methods (simultaneous use of condoms and a hormonal method of birth control) has been recommended as an effective strategy for reducing teen pregnancy and sexually transmitted infections (STIs), but their use is not commonly reported by teens (8.8%) (American College of Obstetricians Gynecologists 2012; Committee on Adolescence 2014; Kann et al. 2014; Martinez et al. 2011; Trussell 2011).

Research suggests there are protective factors and mediators that affect youth sexual behaviors including: familial relationships; parental monitoring and communication; and community and school connectedness (Kirby et al. 2007; Manlove et al. 2008; Oman et al. 2013). In addition to intrapersonal- and community-level factors, individual-level youth assets such as adolescent religiosity, self-confidence, and aspirations for the future have been shown to impact sexual behaviors (Cooksey et al. 1996; Gold et al. 2010; Hardy and Raffaelli 2003; Kirby et al. 2007; Manlove et al. 2006, 2008; Oman et al. 2013; Studer and Thornton 1987). For example, in a longitudinal study Oman et al. (2013) found that youth who possessed a religiosity asset were less likely to participate in sexual risk behaviors. The study of 1111 youth reported that the religiosity asset was significantly protective from initiation of sexual intercourse and pregnancy, but was not associated with birth control use (Oman et al. 2013).

Fewer studies have examined the relationship between adolescent religiosity, religious affiliation, and contraception use (Brewster et al. 1998; Casper 1990; Cooksey et al. 1996; Gold et al. 2010; Manlove et al. 2006, 2008; Nonnemaker et al. 2003; Studer and Thornton 1987). A study of 3949 youth from the National Longitudinal Survey of Youth examined the relationships between religiosity and adolescent contraceptive behavior with several parental factors as mediators. Sexually active teens who used contraceptives every time they had sex and those who were inconsistent or nonusers did not differ on any measure of family religiosity (parental attendance at religious services, prayer, religious beliefs, and familiar religious activities) (Manlove et al. 2008). Family religiosity and contraceptive consistency had no direct association, except among males. Interestingly, among males family religiosity had a direct negative effect on contraceptive consistency (Manlove et al. 2006, 2008). One purpose of the current study is to assess youth religiosity rather than family religiosity because past research has suggested that adolescent religiosity is a protective factor for sexual initiation and in some instances pregnancy (Gold et al. 2010; Hardy and Raffaelli 2003; Oman et al. 2013).

Studies using data from the National Survey of Family Growth concluded that protective factors related to delaying sexual activity (including religiosity) were inversely related to protective factors that increase the use of contraceptives with differences by race and religious affiliation (Brewster et al. 1998; Cooksey et al. 1996). Among African-American adolescents, Protestant fundamentalists were the least likely to use contraception, and among white adolescents, Catholics and Protestant fundamentalists reported high levels of non-contraceptive use (Brewster et al. 1998; Cooksey et al. 1996). Differences were also noted in the method of contraceptive use; white adolescent Protestant fundamentalists were most likely to use the pill, whereas adolescent African-American Protestant fundamentalists were least likely to use the pill (Brewster et al. 1998).

In summary, research has found mixed results for the associations between family religiosity, religious affiliation, and contraceptive use. This study’s purpose was to prospectively investigate associations among youth religiosity, religious denomination, and contraception use among sexually active youth and determine: (1) whether stronger religiosity is a predictor of contraception use or consistent contraception use; (2) whether being affiliated with a religious denomination is a predictor of contraception use, or consistent contraception use; and (3) which religious denominations are associated with contraceptive use.

Methods

Sampling and Data Collection

Data were from the Youth Asset Study (YAS), which is a longitudinal study, intended to examine the associations between youth assets and youth risk behaviors. The Oklahoma City Metropolitan area was stratified by income and race/ethnicity using 2000 census data. Twenty census tracts with diverse populations in regard to race/ethnicity and socioeconomic status were randomly selected. Data collectors canvassed every household in the selected census tracts. One parent and one youth from each household were recruited to participate in the study. Study inclusion criteria were youth between 12 and 17 years of age and living with a parent or guardian; speak English or Spanish; and mentally competent to respond to interview questions. Parents signed a consent and HIPAA forms, and the youth signed an assent form. The Institutional Review Board at the University of Oklahoma Health Sciences Center approved the study (Oman et al. 2009).

Data were collected from 1111 parent and youth pairs in 5 Waves of annual in-home and in-person interviews beginning in 2003–2004 and concluding in 2007–2008. The interviews were conducted using a computer-assisted data entry system. Youth self-administered the risk behavior section of the questionnaire in private by listening to a WAV file with headphones (if reading comprehension was an issue) and entering responses into a secure laptop. The response rate was 61%, and the retention rate was 89% at the completion of the study (Oman et al. 2009). Data from Waves 4 and 5 were used for the current study as this was when the specific religiosity items used for the analyses were administered.

Measures

Demographics

Demographic variables reported by the youth were age, gender, and race or ethnicity. Demographic variables reported by the parent were family structure (one- or two-parent households), parental employment status, and parental income.

Religious Denomination

Youth were asked about their religious affiliation or denomination using the following items. “Regardless of whether you now attend any religious services, do you identify with any particular religious tradition, denomination, or church?” Response options were “yes” or “no.” Youth respondents were asked, “With what religious tradition, denomination, or church do you most closely identify?” The item was open-ended. A majority of the participant responses were Baptist, Catholic, Christian, Protestant, Methodist, and Pentecostal (Table 1). Each youth was placed into one category based on their response: Baptist, Catholic, other, or no religious denomination. The “Other” category included youth identifying as Christian, Protestant, Methodist, or Pentecostal, as well as other Christian denominations and non-Christian religions. Youth indicating they did not identify with a religious tradition were categorized as no religious denomination.

Table 1 Item information for youth religious denomination, religiosity, and contraceptive behaviors
Full size table

Adolescent Religiosity

The adolescent religiosity scale consisted of four items. For example, youth were asked, “How important is it to you to be able to rely on religious teachings when you have a problem?” Possible responses were 1 (Not important at all); 2 (Fairly unimportant); 3 (Fairly important); and 4 (Very important). The items were summed and divided by the number of items, creating a scale with a range of 1–4 with a higher score indicating stronger religiosity. The religiosity scale was found to be reliable (Cronbach’s alpha = 0.91).

Sexual Intercourse

Initiation of sexual intercourse (ever had sex) was assessed via the item, “Have you ever had sexual intercourse (done it, had sex, made love, gone all the way)?” Possible responses were “Yes” or “No” (Card et al. 1999a). Only youth who indicated “Yes” to this item were administered additional sexual behavior and contraceptive items.

Contraceptive Behaviors

Two behaviors related to youth contraceptive use were assessed: type of contraceptive used at last sex and contraception consistency. Youth self-administered the following items. “The last time you had sexual intercourse, did you or the other person use birth control?” Youth responding “No” were considered to be “no method users.” “The last time you had sexual intercourse, what methods of birth control did you or your partner use?”(Card et al. 1999b). Method options were “Shot,” “Birth Control Pill,” “Patch,” “Ring,” “Condom,” “Withdrawal,” “Rhythm,” and “Other.” The response categories for each method were “Yes” or “No.” Youth responding “Yes” to “Condom” but “No” to other methods were considered condom only users. Youth responding “Yes” to “Birth Control Pill,” “Patch,” “Ring,” or “Shot” but “No” to condom use were considered hormonal birth control users only. Dual method users were those that responded “Yes” to condoms or any of the hormonal birth control method (birth control pill, shot, patch, or ring). Less effective methods were considered “Rhythm” and “Withdrawal.” Each youth was placed into one category based on their response: condom use only, hormonal birth control use only, dual method use, or less effective method.

Consistent Contraceptive Use

This was assessed by an item adapted from the literature “In the last 6 months, how often did you use birth control?” (Harris et al. 2009). Response categories were never (0%), a few times (1–40%), half the time (41–60%), most times (61–99%), or always (100%) used a method. Consistent users were those that indicated they “always” used a method.

Analysis

Sub-Analysis regarding type of contraceptive used was limited to youth reported ever having had sex by Wave 5 (N = 757). Analysis regarding consistent contraceptive use was limited to youth who reported having sex in the last 6 months (N = 635). An alpha of 0.05 was used to determine statistical significance unless otherwise noted. Bivariate analysis was performed to examine the relationship between demographic characteristics and type of contraceptive used and contraceptive consistency. Chi-square analysis was conducted for the categorical variables, and analysis of variance (birth control method) or an independent t test (contraceptive use) was conducted for the continuous variables.

Multinomial logistic regression was performed to determine the relationship between religious denomination or youth religiosity and type of contraceptive used at last sex and contraceptive consistency. “No method” was the reference group for type of contraceptive used. When the relationship between religious denomination and type of contraceptive used was significant, additional logistic regression analysis was performed to determine which religious denominations were associated with the contraceptive behavior.

Binomial logistic regression was conducted to determine the relationship between religious denomination or youth religiosity and contraceptive consistency. “Inconsistent contraceptive use” was the reference group for type of contraceptive used. National data suggest there may be differences in contraceptive behaviors and religious affiliation by important demographic variables (Brewster et al. 1998; Centers for Disease Control and Prevention 2014; Cooksey et al. 1996). Each model was therefore adjusted for demographic variables (youth age, gender, and race, parent income, family structure (one- or two-parent household), and parent employment status) found to be significant in the bivariate analysis. To reduce type I error, potential interactions between religiosity, religious denomination, and demographic variables were assessed with an alpha of .01. There was no evidence of interaction in the models, and therefore, main effects were reported. All analyses were completed in SAS 9.3 (SAS Institute Inc 2011).

Results

Descriptive Data

Demographic data for the youth (N = 757) at Wave 4 are listed in Table 2.

Table 2 Youth and parent demographics and youth religious denomination and religiosity descriptive data at Wave 4 (N = 757)
Full size table

Demographic and Religion Data

The youths’ mean age was 17.5 years (SD = 1.6). Respondents were racially and ethnically diverse (37% white, 28% Hispanic, 26% African-American, and 8% other). The majority of youth lived in two-parent homes, had parents who were employed, and had parents with an annual income of less than $35,000.

Seventy-six percent of the youth identified with a religious denomination. Religious affiliation varied widely across the sample, nearly 24% identified as Baptist, 18% identified as Catholic, and 24% did not identify a specific denomination. Youth reported a 3.3 religiosity asset mean score (range 1–4), indicating that religion was fairly or very important to them.

Among all sexually active youth, 41% did not use any method of protection at their last sexual intercourse (Table 3). Nearly a quarter (23%) of the youth reported using only condoms at their last sexual intercourse, 18% reported using only a hormonal form of birth control, 16% reported dual method use, and 2.5% reported using less effective methods such as withdrawal or the rhythm method. Additionally, of the youth who had sex in the last 6 months, 42% reported consistent contraceptive use (Table 3). As anticipated from the research literature, there were significant demographic differences in regard to the type of contraceptive used (by gender, race/ethnicity, and parent education) and consistent contraceptive use (by race/ethnicity, parental education and income, and family structure) (Table 3). These characteristics were statistically controlled in the regression analyses.

Table 3 Percentage of youth by each contraceptive behavior at last sex at Wave 5 according to selected characteristics
Full size table

Youth Religious Denomination and Contraceptive Use

Table 4 shows the unadjusted and adjusted results of the multinomial logistic regression analysis. The model adjusted for demographic characteristics resulted in a significant AOR of 2.17 (95% CI 1.23–4.00) for dual method use indicating for youth identifying with a religious denomination, the odds for use of the dual method as compared to no method of birth control at last sexual intercourse were two times greater, compared to youth who did not identify with a religious tradition. Identifying with a religious denomination was not significantly associated with condom use (AOR = 1.25, 95% CI 0.79–2.03), hormonal birth control use (AOR = 1.37, 95% CI 0.83–2.31), or less effective methods (AOR = 0.51, 95% CI 0.18–1.52) relative to using no method at all (Table 4).

Table 4 Unadjusted and adjusted odds ratio (AOR) with 95% confidence intervals from individual multinomial and binomial logistic regression models for religiosity and religious denomination at Wave 4 on youth contraceptive use at Wave 5
Full size table

Follow-up analysis suggested that youth identifying as Baptist had an odds 2.5 times greater (AOR = 2.57, 95% CI 1.29–5.17) of using dual methods (relative to using no contraceptive methods) as compared to youth identifying with no religion. Similarly, youth identifying with an “Other” religion had an odds nearly two times greater (AOR = 1.96, 95% CI 1.02–3.75) of reporting using dual methods (relative to youth reporting no contraceptive methods) compared to youth identifying with no religion. No other significant differences were noted for other methods of birth control, and there were no associations between youth religious denomination and consistent contraceptive use (AOR = 1.20; 95% CI 0.80–1.80).

Youth Religiosity and Contraceptive Use

The results indicate nonsignificant adjusted odds ratio (AOR) for youth religiosity and condom use (AOR = 0.86, 95% CI 0.67–1.12), hormonal birth control (AOR = 1.09, 95% CI 0.82–1.46), dual methods (AOR = 1.25, 95% CI 0.92–1.73), and less effective methods (AOR = 0.78, 95% CI 0.45–1.44). These results indicate there was no association between higher youth religiosity scores and type of birth control used (relative to using no method) at last sexual intercourse (Table 4). The binomial logistic regression analyses produced a nonsignificant AOR of 1.03 (95% CI 0.83–1.28) for youth religiosity and consistent contraceptive use indicating no association between youth religiosity and consistent contraceptive use (Table 4).

Discussion

This study explored prospective associations among youth religiosity and religious denomination, and type of contraceptive used and consistent contraceptive use among sexually active youth. Consistent with the literature there were differences based on demographic factors in regard to the type of contraceptive used (by gender, race/ethnicity, and parent education) and consistent contraceptive use (by race/ethnicity, parental education and income, and family structure). These demographic differences are notable as hormonal birth control use, dual method use, and consistent contraceptive use are effective strategies in reducing teen pregnancy rates (American College of Obstetricians Gynecologists 2012; Committee on Adolescence 2014; Hamilton et al. 2015). Consistent use of highly effective methods of birth control is particularly important for minority youth, and youth from low socioeconomic backgrounds as teen pregnancy rates among these populations are disproportionately high (Hamilton et al. 2015; Penman-Aguilar et al. 2013; Singh et al. 2001). Identifying protective factors, including religiosity, that increase contraceptive use among these high-risk youth can contribute to declines in teen pregnancy rates.

Findings revealed that identifying with a religious denomination was a predictor of dual contraceptive use at last sexual intercourse relative to using no method of birth control identifying with a religious tradition increased the odds of using dual methods (AOR = 2.17) relative to no method. Follow-up analysis suggests youth identifying as Baptist or with “Other” religious traditions such as Methodist, Pentecostal or Protestant were twice as likely to use dual methods (relative to using no method of birth control) as youth not identifying with any religious tradition. This is consistent with the findings of Cooksey et al. and Brewster et al. both of which suggest that contraceptive use varies by religious affiliation (Brewster et al. 1998; Cooksey et al. 1996). These results also generally agree with previous research, suggesting that, among sexually active adolescents, Protestants were more likely than other adolescents without a religious affiliation to use highly effective contraceptives including hormonal methods (Casper 1990; Cooksey et al. 1996; Mauldon and Luker 1996). The variation in contraception use by religious affiliation may be due to the more progressive stance of Mainline Protestant, Methodist and some Baptist denominations supporting family planning and contraceptive use for married women, and in some cases all women (General Board of the American Baptist Churches 1994; Jones and Dreweke 2011; United Methodist Church 2012).

However, there was no significant relationship between youth’s religious affiliation and condom use, hormonal birth control use, or less effective methods such as the rhythm method and withdrawal (relative to using no method of birth control). Additionally, there was no significant association between a youth’s denomination and consistent contraceptive use. Additional qualitative and quantitative research is needed to understand why religious affiliation is associated with dual method use, but not to other contraceptive types. Nonnemaker et al. (2003) suggest that whereas most religious denominations have consistent values about delaying sexual intercourse, the norms regarding contraception use are inconsistent across religions.

Youth religiosity was not found to be a predictor for the use of any type of contraceptive. Youth with higher religiosity scores were no more likely to use hormonal methods, dual methods, condoms, or less effective methods, relative to using no method at all, than youth with lower religiosity scores. Similarly, youth religiosity was not significantly associated with consistent use of contraceptives in the last 6 months. This result agrees with the findings of Manlove et al. Studer et al. and Cooksey et al., suggesting that there is little or no relationship between religiosity and contraceptive behaviors (Cooksey et al. 1996; Gold et al. 2010; Manlove et al. 2006, 2008; Nonnemaker et al. 2003; Studer and Thornton 1987). These results support the notion that religiosity is a focused belief in God or a Deity with little clear connection to contraceptive use behavior for sexually active youth. Rather, the relationship between religiosity and contraceptive behaviors may be influenced by use of time engaged in religious activities or mediated by factors such as positive peer behaviors and parental monitoring (Manlove et al. 2008).

The present study extends current research involving family or parental religiosity and contraceptive behaviors by examining the association between specific contraceptive methods and adolescent religiosity rather than family religiosity. Adolescent religiosity focuses on characteristics and behaviors of the adolescent and not the family as a unit. This distinction is important, as accumulating research indicates that adolescent religiosity protects from early sexual initiation and pregnancy (Gold et al. 2010; Hardy and Raffaelli 2003; Oman et al. 2013). Additionally, this study extends the research of Oman et al. involving adolescent religiosity and contraceptive use by examining the association between specific adolescent religiosity and specific contraceptive methods rather than general birth control use (Oman et al. 2013).

This study’s results suggest that identifying with a religious denomination may be associated with increased dual method use, but not all contraceptive behaviors. Public health practitioners should work closely with faith leaders in high-risk minority and impoverished communities to develop and implement strategies to prevent teen pregnancy. In particular, faith-based comprehensive sexuality education programs for youth that may already be sexually active are needed. Promising sexual health education programs for communities of faith exist. For example, Our Whole Lives (OWL): A Lifespan Sexuality Education Series, developed by the Unitarian Universalist Association, is a comprehensive sexuality curriculum that meets or exceeds the National Standards for Sexuality Education Core Curriculum (Goldfarb and Casparian 2000). However, there are no comprehensive and rigorously evaluated sexuality education programs that have been developed for communities of faith.

In a survey from the National Campaign to prevent teen pregnancy, 75% of teens and 73% of adults said that religious leaders and groups should be doing more to prevent teen pregnancy (The National Campaign to Prevent Teen and Unplanned Pregnancy 2013b). Although organizations such as the National Campaign to Prevent Teen Pregnancy and Esperanza have worked to develop tips and toolkits to address teen pregnancy in African-American and Latino communities, the results of this study suggest that additional work is needed (The National Campaign to Prevent Teen and Unplanned Pregnancy 2013a, b). For example, experts suggest that, rather than remain silent on the issue of teen pregnancy, youth pastors should engage youth in meaningful culturally appropriate, age-appropriate conversations about “healthy sexuality, the Church’s stance on teen contraceptive use, and community resources … [and] offer young people support and guidance to explore and develop their own values and attitudes on issued related to sex” (The National Campaign to Prevent Teen and Unplanned Pregnancy 2013a).

Limitations

The findings of this study may be limited by the validity of self-reported contraceptive behaviors. Youth may report socially desirable behaviors that indicate responsible sexual activity, resulting in an over-reporting of contraceptive use. To reduce potential bias, youth were able to read (and listen) and respond to all questions pertaining to sexual behaviors on a computer without the interviewer present. Males may not be knowledgeable about a female partner’s hormonal contraceptive use, which could lead to over-reporting. In addition, very small numbers of youth identifying with broad categories of Christianity (such as “Protestant”), less-represented Christian denominations (such as Methodist), and non-Christian religions (such as Islam) were combined into a single religious affiliation category, which may have obscured important differences among them in regard to contraceptive use. Future study is necessary, with larger populations involving youth who belong to other Christian denominations, as well as those belonging to non-Christian religions. Although this was a prospective study that assessed religiosity and religious affiliation 1 year prior to assessing contraceptive use, it is possible that a third unobserved variable significantly influenced the reported associations. Finally, study data were collected from participants living in the Oklahoma City Metropolitan nearly 9 years ago. Therefore, the data and statistical analyses may not reflect current trends in the study variables such as religious denomination or contraceptive use, or perhaps even the associations among the variables.

Conclusion

Overall, this study found that identifying with some religious denominations is positively associated with some but not all contraceptive behaviors. Faith leaders and those working in the field of public health should work together to develop strategies and resources to help religious leaders to engage in conversations with sexually active youth about contraceptive use. Continued research is needed to further develop our understanding of predictors of contraception consistency, dual contraception use, condom use, birth control use, and hormonal method use as well as effective intervention strategies to reduce unplanned pregnancies among teens.