Introduction

Adolescence is a key period in the lifecourse for establishing preventive health behaviors; however, it is also a time for increased risk [1]. Social determinants of health can contribute to adolescent health and health disparities [2], in particular, environment and social contexts, including living in a rural area or community. Previous research has found that teens living in rural communities may be at more risk for tobacco use, alcohol use, sexual intercourse, and pregnancy [3].

While U.S. teen birth rates have decreased substantially [4], certain teens may be at higher risk for teen pregnancy, including teens from rural areas. From 1990 to 2010, declines in teen birth rates were more gradual in rural counties (31% decline) compared to metropolitan counties (50% decline) [5]. Moreover, the teen birth rate is approximately one-third higher in rural counties (42.9%) compared to metropolitan counties (32.6%) [6]. These disparities in teen birth rates for those in rural counties are associated with reduced access to health services, lack of health insurance, poverty, and the proportion of female headed households [5]. Such high rates are also directly associated with sexual activity and contraceptive use [6].

Research indicates that rural teens may be at greater risk for teen pregnancy compared to urban teens. National data indicate that, compared to urban teens, rural teen females are more likely to report ever having had sex, yet less likely to report use of contraception at first sex. Moreover, female teens in rural areas often rely on community health centers to receive contraception methods [5].

While national data suggest rural areas may present higher risk for adolescents for sexual health, there is a need to examine the unique state contexts. Florida, the third most populous state [7], is a distinctive location due to the diversity of regions and population densities across the state. Moreover, Florida has abysmal rates for many sexual and reproductive health outcomes. According to 2013 data, Florida had the highest number of HIV diagnoses and was fourth highest in number of syphilis cases among all states [8]. In 2011, Florida had a higher teen pregnancy rate (56 per 1000 women aged 15–19 years) compared to the national rate [9]. Among Florida high school students, 40.3% report ever having sexual intercourse, and 38.3% reported not using a condom during last sexual intercourse [10]. Furthermore, according to the Youth Risk Behavior Survey, Florida youths were more likely to report early sex and less likely to use effective forms of contraception (e.g., oral contraceptives, long-acting reversible contraception) compared to U.S. youths [11].

Thus, the unique context of Florida for sexual and reproductive health issues presents an opportunity to examine how rural settings impact adolescent sexual behaviors. The purpose of this study was to assess differences in sexual behaviors among Florida adolescents by rural–urban community location.

Methods

Study Design

The data analyzed come from a randomized controlled trial (RCT) evaluation of a school-based intervention; the details are reported elsewhere [12]. A pair-matched, cluster RCT was conducted with 2 cohorts of Florida high-schoolers to evaluate whether youth enrolled in the intervention experienced an increase in positive sexual health outcomes compared to youth in the control condition. Cohort 1 youth were enrolled in 28 high schools in 12 counties in fall 2012. After program completion, one school was lost due to lack of interest in the evaluation and that school and its matched pair were removed. Youth in cohort 2 were enrolled in 26 high schools in 10 counties. Cohort 2 enrolled youth in fall 2013. For this paper, data from each cohort at baseline were analyzed. Additionally, the dataset was subset to 9th grade students who were beginning high school.

Counties were eligible for inclusion if they demonstrated poor outcomes on indicators, such as birth rate and repeat birth rate among female youth, chlamydia and gonorrhea rates among female youth, high school dropout and graduation rates, and out-of-school suspension rates. Schools were eligible if they had year-long health or health-related classes, agreed to be randomly assigned to either treatment or control, and were willing to have surveys administered in classrooms at multiple time points. Schools were matched based on a variety of criteria and each school within each matched pair was randomized to either treatment or control.

The paper-and-pencil survey contained questions regarding demographics, attitudes, and behaviors including sexual health behaviors. Parental written consent was obtained through a passive (i.e., opt-out) process. Youth were eligible to participate in the evaluation if the following conditions were met: (1) enrolled in a class included in the evaluation, (2) had parental consent, (3) were proficient in English, and (4) able to take a paper-and-pencil survey independently. Youth were asked to provide written assent prior to survey administration. The larger RCT evaluation was reviewed and approved by the Florida Department of Health Institutional Review Board (Protocol H11180).

Measures

The main outcome variables were sexual behaviors. The entire sample was questioned if they ever had sexual intercourse and if they intend to have sex in the next year. Participants who intend to have sex in the next year then responded whether they intend to use condoms while having sex. Participants who reported being sexually experienced (i.e., have ever had sex), were asked to report: number of sexual partners, ever been pregnant, and having sex in the last 3 months. Among those who reported having sex in the last 3 months, they responded to questions about the number of sexual partners and having sex without a condom. These variables were operationalized as dichotomous variables (e.g., yes or no; 4 or more partners or less than 4 partners). Demographic variables included: gender (male, female), race/ethnicity (Black, Hispanic, White, Other), and age.

The location of the school where the participant was sampled was categorized as either urban or rural. The U.S. Census Bureau categorizes counties as being either metropolitan, micropolitan, or noncore. A metropolitan area contains a core urban area with a population of 50,000 or greater and a micropolitan area possesses an urban core of at least 10,000, but less than 50,000 [13]. Each metropolitan or micropolitan area includes a county that contains a core urban area and may also include adjacent areas that have a high amount of social and economic integration [13]. This integration is measured by residents commuting to a place of work within the urban core [13]. Any county not defined as metropolitan or micropolitan is categorized as noncore. For this analysis, counties that were defined by the U.S. Census Bureau as metropolitan or micropolitan were categorized as urban. All other counties were categorized as rural.

Data Analysis

All analyses were conducted using SAS/STAT© software, version 9.4 of the SAS System for Windows. Multiple imputation was employed to account for potential bias in estimated treatment effects due to missing outcome data [14]. Fifteen imputation data sets were created for the analysis, which provided approximately 99% efficiency relative to a full sample analysis [15]. Results of individual model fits for all imputed data sets were combined using the SAS MIANALYZE procedure.

We estimated the effect of rural–urban status on risk outcomes after controlling for demographic variables using generalized linear mixed models [16] implemented in the SAS GLIMMIX procedure with a logit link for dichotomous outcomes and a cumulative logit link for ordinal outcomes. The log odds of the outcome variables was modeled as a linear function of rural–urban status, race, age group, gender, and cohort. We included random effects at the school level in order to account for dependence between measures on each student within a school. Using these models, we estimated the odds ratio for dichotomous risk outcomes comparing metropolitan to non-metropolitan schools.

Results

This sample comprised 6,316 Florida 9th graders (Table 1). Among these participants, 74.4% (N = 4696) were classified as urban, and 25.6% (N = 1620) were classified as rural. The sample was almost evenly split between males and females. Most adolescents identified as white (63.4%) and were age 14 (72.2%).

Table 1 Demographic characteristics of Florida adolescents by rural–urban status (N = 6316)
Full size table

Participants described their sexual behaviors and intentions for sexual behaviors (Table 2). More teens from rural areas reported ever having sex (24.0%) compared to urban teens (19.7%). Similarly, among teens who engaged in sex, more rural teens reported four or more sexual partners and ever being pregnant. Approximately one-eighth of the sample who had sex reported having sex in the past 3 months, and 4.6% had four or more sexual partners in the last 3 months. Among adolescents having sex in the last 3 months, more teens from rural areas reported having sex without a condom in the past 3 months (61.2%) compared to urban teens (53.7%). Regarding intentions for sexual behavior, among both groups, the majority of teens said they were not going to have sex in the next year. However, among those intending to have sex, most said they would use a condom.

Table 2 Sexual behaviors of Florida adolescents by rural–urban status (N = 6316)
Full size table

Despite differences in rates of sexual behaviors between urban and rural statuses, there were limited differences in sexual behavior outcomes when adjusting for demographic characteristics. No statistically significant differences were observed for any of the sexual behaviors assessed, aside from one sexual health outcome (Table 3). Urban students were significantly less likely to intend to have sex without a condom in the next year compared to rural students (OR 0.76, 95% CI 0.63, 0.92).

Table 3 Regression model of urban status impact on sexual behaviors among Florida adolescents (N = 6316)
Full size table

Discussion

This study assessed the association of rural and urban community location on sexual health behaviors among adolescents in Florida. Overall, we found no major differences in sexual behaviors between rural and urban adolescents. However, sexual intentions differed between rural and urban adolescents; specifically, rural adolescents were more likely to intend to have sex without a condom in the next year compared to urban adolescents.

Reassuringly, we did not find major disparities in sexual health behavior between rural and urban adolescents in Florida. This is despite national data indicating persistent disparities by urbanicity for teen births [5]. This may be due to our study examining teen pregnancy, rather than teen birth, which do not represent the same phenomenon. Measurement of teen pregnancy does not take into account the potential outcomes of the pregnancy [9], which may vary by location [17, 18].

Yet, there are still opportunities for improvement for prevention behaviors among rural youth in Florida. This study revealed that rural youth are more likely to intend to have sex without a condom in the next year, which not only places adolescents at risk for sexually transmitted infection, but also pregnancy. Potential contributors to this intention for unprotected sex may be associated with the lack of anonymity and confidentiality in rural communities, socioeconomic status, isolation, and stigma associated with sexual activity and contraception use [19]. For example, in a study by Geske et al. [19], rural youth were reluctant to obtain contraception for fear that they would be easily recognized by the clinician, pharmacist, or a community member due to the low population of rural areas [19]. Moreover, challenges exist for addressing health concerns for adolescents in rural communities, including locations of services, confidentiality, isolation, shortage of healthcare professionals, opposition to sex education, distances, and transportation [6]. School-based programs may be an opportunity to address these barriers to sexual health education and prevention in rural communities.

This study’s sample was derived from 9th grade students in rural and urban counties in the state of Florida. Among rural adolescents, 24% reported ever having sex, while 20% of urban students reported ever having sex. The rural adolescents were more similar to the overall state (Florida) Youth Behavior Risk Factor Surveillance System (YRBSS) estimate of 24.1% of 9th graders ever having sex in 2015 [10]. Similarly, 6.4% of rural adolescents reported ever having sex with four or more persons, whereas it was 6.7% according to YRBSS 2015 data [10]. Despite being similar on average to the state of Florida, opportunities still exist to improve the sexual well-being of adolescents. Promoting access to long-acting reversible contraception among adolescents can provide sustained pregnancy prevention over long periods of time [20].

This study should be considered in context of the limitations. First, there are multiple definitions available for operationalizing rural/urban status. This can differ by the agency measuring urbanicity. This study utilized the U.S. Census Bureau’s classifications for urban and rural. Additionally, the data included in this study were self-reported by adolescents, and may underreport behaviors due to social desirability bias. Finally, this sample was derived from 12 Florida counties, and therefore may not be generalizable beyond these areas. However, Florida has a large, racially and ethnically diverse population that provides an opportunity to analyze health outcomes among diverse groups.

While the literature indicates differences in sexual health outcomes by locale among youth, this study found limited evidence of disparities. Yet, this study did find that rural youth were more likely to intend to use a condom in the future. Understanding the specific disparities can inform contraception and sexual health interventions among rural youth. This is especially critical as access to contraception may become more difficult with the changing political climate.