Abstract
Purpose
Little is known about how adolescent sexual behaviors develop and the influence of personal or perceived social attitudes. We sought to describe how personal, perceived peer and perceived family attitudes towards adolescent sexual activity influences adolescent females’ sexual behaviors over time.
Methods
Between 1999–2006, 358 English-speaking females, aged 14–17 were recruited from three urban adolescent clinics. Participants completed quarterly and annual questionnaires over 4 years. Primary outcomes were engagement in eight sexual behaviors: kissing, having breasts or genitals touched, touching partners’ genitals, and oral (giving or receiving), anal, or vaginal sex. Three attitudinal scales assessed personal importance of abstinence, perceived peer beliefs about when to have sex and perceived family beliefs that adolescent sex is negative.. We used generalized estimating equations to identify predictors of each sexual behavior and compared whether personal, perceived peer or perceived family attitudes predicted sexual behaviors over time.
Results
The odds of reporting each sexual behavior increased with age but were lower among those whose personal or perceived family attitudes were less positive. Participants’ personal attitudes towards adolescent sex were the strongest predictor of engagement in all eight sexual behaviors even after controlling for perceived peer and perceived family attitudes.
Conclusions
Female adolescent’s personal attitudes towards abstinence appear to be the strongest predictor of engagement in a variety of sexual behaviors. Efforts to influence adolescent attitudes towards abstinence may be an important approach to reducing sexual behaviors that increase the risk for pregnancy and sexually transmitted infections.
Keywords: adolescents, adolescent sexual behavior, attitudes, peer group, family
INTRODUCTION
Adolescent sexual behavior is a normal developmental milestone.[1] However, little is known about how adolescent sexual behaviors develop over time.[2–6] Specifically, little is known about how sexual behaviors are acquired, the factors influencing this process or how patterns of engagement in various sexual behaviors change over time.[5, 7–10] There are two major limitations of existing studies. First, the predominance of cross-sectional study designs prevents determination of temporal associations or causation. Second, despite the fact that sexual behaviors other than coitus affect the acquisition of sexually transmitted infections (STI), many focus on penile-vaginal intercourse.[7–11] Studies that explore how a broader range of adolescent sexual behaviors develops are needed as this information has important implications for sexual health educators and developers of adolescent pregnancy and HIV/STI prevention interventions.
Attitudes towards sex, whether an adolescent’s personal attitudes or those of important social referents such as peers and parents, are important behavioral influences.[3, 12–16] However, few studies have examined the influence of adolescent’s personal attitudes towards sex on their sexual behavior relative to the attitudes of key social referents.[12, 17, 18] In this study, we sought to describe changes over time in female adolescents’ reporting of eight sexual behaviors; to describe changes in adolescents’ personal, perceived peer and perceived family attitudes and beliefs towards adolescent sexual activity; to evaluate demographic and attitudinal predictors of reported engagement in eight sexual behaviors; and to assess the relative influence of adolescents’ personal, perceived peer and perceived family attitudes and beliefs towards adolescent sexual activity on adolescents’ reported sexual behavior over time. Because we were interested in behaviors that increase risk for teen pregnancy and STIs, we focused on sexual behaviors with opposite-sex partners.
METHODS
Study Design and Procedures
Data were collected as part of a larger, cohort study of STI risk and protective factors among female adolescents.[7, 19–21] The larger study consisted of up to six annual questionnaires, quarterly interviews and two 84-day diary collection periods each year. The current secondary analysis uses data from the annual questionnaires and quarterly interviews. Enrollment was rolling during the first 4 years of the study; therefore, participants included in this analysis contributed different amounts of follow-up data. Although participants could provide data for up to 6 annual visits, few reported data at years 5 and 6. Thus, we limited analyses to data from the first 4 annual visits. Written informed assent was obtained from adolescents and written consent from parents/legal guardians. This research was approved by the institutional review boards at Indiana University/Purdue University at Indianapolis and the University of Pittsburgh.
Participants
Participants were recruited from three primary health clinics in Indianapolis that serve lower- and middle-income residents and are located in areas with high rates of adolescent pregnancy and STIs. Eligible participants were non-pregnant, English speaking females, aged 14 to 17 years. For the current analyses, participants had to complete at least 2 visits following enrollment. Although 386 participants enrolled in the larger study, 358 are included in the current analysis as 28 had fewer than two visits following enrollment.
Measures
Primary Outcomes
Our primary outcomes were reported engagement in eight sexual behaviors with opposite-sex partners prior to enrollment and each quarterly visit. Sexual behaviors included four non-penetrative (i.e., deep kissing, having breasts or genitals touched, and touching a partner’s genitals) and four penetrative behaviors (i.e., giving or receiving oral sex, vaginal and anal sex).
Predictor Variables
The dataset contained several measures of adolescents’ personal, perceived peer and perceived family attitudes and beliefs towards adolescent sexual activity. We developed and refined three scale measures that most closely assessed our concepts of interest using confirmatory factor analyses. The content in these scales were not parallel which was a limitation of the data available in secondary dataset. The Adolescent Personal Importance of Abstinence Scale (4 items, alpha 0.72) assessed the importance of being considered a virgin or waiting to have sex until one is older, in love or married. (Sample item: How important is it to you to wait to have sex until marriage?) The Perceived Peer Beliefs about When to Initiate Sex Scale (3 items, alpha=0.61) assessed perceptions of friends’ attitudes towards having sex when in love, with a boy/girlfriend, or waiting until one is older. (Sample item: How does your friend feel about having sex if you’re in love?) Perceived Family Belief that Sex During Adolescence is Negative Scale (4 items, alpha=0.67) assessed family beliefs that sex during adolescence is wrong, dangerous, or reflects immaturity. (Sample item: My parents think that it is morally wrong for teenagers to have sex before marriage.) For simplicity, the three scales will be referred to collectively throughout this manuscript as personal, perceived peer and perceived family attitudes scales. The Personal Importance scale utilized a 3-point likert scale; perceived peer and family belief scales used 4-point likert scales. Higher scores indicated greater acceptance of adolescent sexual activity. To allow direct comparisons of scale scores, each scale’s average score was divided by the total number of likert response options for the descriptive analyses. Unscaled averages were used in regression models since these would not affect calculated odds ratios.
Descriptive Variables
Two socio-demographic variables (age, race) and four sexual history items (age at coitarche, number of lifetime sexual partners, number of sexual partners in past 3 months, and timing of last intercourse before enrollment) were included. Race was self-reported using 6 response options: American Indian/Alaskan Native, Asian, Black/African American, Hispanic/Latina, Native Hawaiian/Pacific Islander, or White. Due to relative sample homogeneity, race was collapsed into a dichotomous variable (Black versus other).
ANALYSIS
Descriptive Analysis
We calculated means/medians for continuous variables and frequencies for categorical variables. Age at first sex, number of lifetime sexual partners and sexual partners within past 3 months were ordinal so medians are reported. We performed bivariate analysis to assess differences in sample characteristics, sexual history, and sexual behaviors as well as personal, perceived peer and perceived family attitudes towards adolescent sexual activity by age and penile-vaginal sexual experience at enrollment (yes/no). Figure 1 shows data only from the 122 participants who completed the questionnaire at enrollment and at all 4 annual visits to illustrate sexual behavior change over time. We did this because it is important to include the same cohort of participants when examining trends over time. To perform statistical testing, either the observations from one time period to the other must be independent or dependent. Had we used all 358 cases at each time period, these analytic criteria would not have been met as there would have been a different number of subjects reporting data at each time point. This would have also reduced our ability to assess significance in changes over time. We looked at the same plots with all 358 observations noting little difference. We also compared demographic characteristics of the 122 and 358 participants and noted no significant differences.
Figure 1.
Figure 1a. Changes in non-penetrative sexual behaviors over time
Figure 1b. Changes in penetrative sexual behaviors over time
Predictor Scales
Before including attitudinal scales in regression analyses, we assessed for correlation between the three scales among the 354 participants who provided complete data for each scale at enrollment. To confirm reliability of each scales’ performance over time, we assessed correlation between the three scales among the 122 providing complete data for each scale at enrollment and at the four annual visits. There was no significant correlation between the three scales in either analysis confirming that each assessed unique constructs and could be placed simultaneously in regression models.
Predictors of Variation in Sexual Behaviors
We used generalized estimating equation (GEE) modeling to assess whether personal, peer or parental attitudes predicted reporting of each sexual behavior. Each participant was the unit of measure and visit number (1–5) was our time variable. We assumed a one degree autoregressive correlation structure such that behavior in adjacent years would be more highly correlated. We first used GEE to determine if age, race and time were significant predictors of sexual behavior and should therefore be considered covariates in subsequent models. Next, we constructed three sets of GEE models for each sexual behavior. Eight unadjusted models were initially created with the dependent variable in each being one of the eight sexual behavior outcomes and the independent variable either the personal, perceived peer or perceived family attitude measure. The second set of eight models (partially adjusted) were identical to the unadjusted models except each was adjusted for age, race and time. To determine whether the relationship between attitude scores and sexual behavior changed over time we assessed for time x attitudinal score interactions. None were significant so interaction terms were dropped from the models. Thus, the partially adjusted models allowed us to determine which attitudinal scale (personal, perceived peer or perceived family) was the strongest predictor of reported sexual behaviors. We controlled for time because the sexual behaviors increased with time. The third set of models (fully adjusted) were adjusted for age, race, time and the other two attitudinal scores. This allowed us to assess the predictive power of personal, perceived peer or perceived family attitudes after controlling for the other two attitude scales simultaneously. We also stratified our analysis by age and by penile-vaginal sexual experience at enrollment. This allowed us to examine whether and how the attitudinal scores varied in their predictive ability when each age-cohort and when sexual experience were considered separately. Estimates were considered statistically significant at p<.05. All analyses were performed using STATA.[22]
RESULTS
Socio-Demographic and Sexual Behavioral Characteristics
Sample characteristics are shown in Table 1. The mean age at enrollment was 15.3±1.1 years. The majority of participants were Black, reported previous penile-vaginal intercourse and sexual intercourse within 3 months before enrollment. More than two-thirds reported deep kissing and having their breasts or genitals touched by a partner.
Table 1.
Sample Characteristics at Baseline Stratified by Age at Enrollment (N=358)
| Sample n (%) | Age at Enrollment (years) | p-value | ||||
|---|---|---|---|---|---|---|
| Characteristics | 14 (N=103) | 15 (N=106) | 16 (N=88) | 17 (N=61) | ||
| Socio-Demographics | ||||||
| Mean age, years (±SD) | 15.3 ±1.1 | --- | --- | -- | --- | --- |
| Race | ||||||
| Black | 319 (89) | 92 (89) | 95 (90) | 78 (89) | 54 (89) | 0.7 |
| White | 34 (10) | 9 (9) | 11 (10) | 9 (10) | 5 (8) | |
| Hispanic | 4 (1) | 1 (1) | 0 (0) | 1 (1) | 2 (3) | |
| Hawaiian/Pacific Islander | 1 (0.3) | 1 (1) | 0 (0) | 0 (0) | 0 (0) | |
| Sexual History* | ||||||
| Median age at coitarche, years | 14 | 13 | 14 | 14 | 14 | ---* |
| Median number of lifetime sex partners | 3 | 2 | 3 | 3 | 4 | ---* |
| Median number of sex partners in last 3 months | 1 | 1 | 1 | 1 | 1 | ---* |
| Last sexual encounter before enrollment | ||||||
| < 1 month | 137 (56) | 24 (48) | 29 (44) | 41 (56) | 43 (78) | <0.01 |
| 1–3 months | 44 (18) | 11 (22) | 13 (20) | 14 (19) | 6 (11) | |
| 3–6 months | 31 (13) | 7 (14) | 9 (14) | 14 (19) | 1 (2) | |
| 6–12 months | 20 (8) | 5 (10) | 12 (18) | 1 (1) | 2 (4) | |
| > 12 months | 12 (5) | 3 (6) | 3 (5) | 2 (4) | 3 (5) | |
| Sexual Behaviors | ||||||
| Deep kissing | 300 (84) | 81 (79) | 87 (82) | 77 (88) | 55 (90) | 0.18 |
| Had breasts touched by a partner | 293 (82) | 70 (68) | 86 (81) | 81 (92) | 56 (92) | <0.01 |
| Had genitals touched | 240 (67) | 47 (46) | 70 (66) | 72 (82) | 51 (84) | <0.01 |
| Touched a partner’s genitals | 201 (56) | 41 (40) | 52 (49) | 64 (73) | 44 (72) | <0.01 |
| Gave oral sex | 47 (13) | 8 (8) | 8 (8) | 10 (11) | 21 (34) | <0.01 |
| Received oral sex | 110 (31) | 20 (19) | 23 (22) | 32 (36) | 35 (57) | <0.01 |
| Had vaginal sex | 268 (75) | 57 (55) | 76 (72) | 79 (90) | 56 (92) | <0.01 |
| Had anal sex | 20 (6) | 2 (2) | 4 (4) | 8 (9) | 6 (10) | 0.11 |
| Attitudes towards Sex† | ||||||
| Personal attitudes | 0.66±0.18 | 0.65±0.19 | 0.64±0.19 | 0.69±0.17 | 0.67±0.18 | 0.19 |
| Perceived peer attitudes | 0.71±0.16 | 0.66±0.16 | 0.74±0.16 | 0.74±0.16 | 0.73±0.14 | <0.01 |
| Perceived family attitudes | 0.60±0.16 | 0.56±0.15 | 0.60±0.16 | 0.61±0.16 | 0.63±0.15 | 0.03 |
Response options for these questions included age groups so no statistical test could be perform.
Each average scale score is scaled to allow direct comparisons (i.e., divided by the total number of items in the scale).
When stratified by age at enrollment (Table 1), there were no differences in racial composition or median age at first sex. Although the number of lifetime sexual partners increased with age, the median number of partners in the past three months did not. The likelihood of reporting sex during the month prior to enrollment increased with age suggesting older participants were more likely to have intercourse regularly. Participants’ likelihood of reporting each behavior increased significantly with age, except kissing and anal intercourse. The former was very common and the latter uncommon at enrollment across all age groups.
When stratified by penile-vaginal sexual experience at enrollment (data not shown), participants reporting penile-vaginal sexual inexperience were significantly younger (14.7 ± 0.87 vs 15.5 ± 1.1, p<0.01) and less likely to report engaging in all the sexual behaviors (p<0.01 for all) compared to sexually experienced participants.
Figure 1 illustrates that the proportion of participants who reported engaging in each non-coital (Figure 1a) and coital (Figure 1b) sexual behavior increased significantly over time. Similar time trend relationships were noted when graphs were stratified by age or penile-vaginal sexual experience at enrollment (data not shown). When stratified by age, the increasing prevalence of sexual behaviors over time was found to be influenced primarily by increasing sexual behaviors of those aged 14 or 15 at enrollment. By age 16, reporting of each sexual behavior was similar regardless of age or penile-vaginal sexual experience at enrollment.
Attitudes towards Adolescent Sexual Activity
Participants perceived friends as having more accepting attitudes towards adolescent sexual activity compared to personal or perceived family attitudes (Table 1). Participants perceived family attitudes were less accepting than their own or their peers. Participants’ attitudes did not vary significantly with age at enrollment. However, perceived peer and perceived family attitudes were more accepting as enrollment age increased. The personal (0.54 ± 0.17 vs 0.70 ± 0.17, p<0.01), perceived peer (0.62 ± 0.16 vs 0.75 ± 0.14, p<0.01) and perceived family attitudes (0.53 ± 0.15 vs 0.62 ± 0.15, p<0.01) of participants who reported being sexually inexperienced at enrollment were significantly less accepting of adolescent sexual activity compared to those who reported being sexually experienced.
Changes in participants’ personal, perceived peer and family attitudes towards adolescent sexual activity over time are shown in Table 2. Compared to enrollment data, participants’ personal attitudes, perceived peer and family attitudes all became more accepting over time with the greatest changes in participants’ personal and perceived family attitudes.
Table 2.
Longitudinal Changes in Personal, Perceived Peer and Perceived Family Attitudes towards Teen Sex*
| Attitude Scale | Time (Summary score† ± standard error) |
||||
|---|---|---|---|---|---|
| Enrollment | Year 1 | Year 2 | Year 3 | Year 4 | |
| Personal | 0.66 ± 0.01 | 0.72 ± 0.01 | 0.75 ± 0.01 | 0.74 ± 0.02 | 0.77 ± 0.02 |
| Perceived peer | 0.71 ± 0.01 | 0.73± 0.01 | 0.73 ± 0.01 | 0.76 ± 0.01 | 0.77 ± 0.01 |
| Perceived family | 0.60 ± 0.01 | 0.64 ± 0.01 | 0.66 ± 0.01 | 0.65 ± 0.01 | 0.69 ± 0.02 |
Higher scores indicate greater acceptance of teen sex.
Each average scale score is scaled to allow direct comparisons (i.e., divided by the total number of items in the scale).
Predictors of Variation in Sexual Behaviors
In our initial exploratory models, participants’ reported sexual behaviors varied by age and race. Increasing age was associated with almost twice the odds of reporting having had one’s breasts (1.66; CI: 1.24, 2.20) or genitals touched (1.85; CI: 1.43, 2.39), touching a partner’s genitals (1.86; CI: 1.44, 2.40), giving (1.61; CI: 1.22, 2.13) or receiving (1.53; CI: 1.22, 1.92) oral sex and three times the odds of engaging in vaginal sex (2.91; CI: 2.12, 4.00). Black race was associated with six times the odds of reporting giving oral sex compared to non-black participants (6.04; CI: 2.50, 14.57).
Table 3 shows the relative predictive power of personal, perceived peer and family attitudes towards adolescent sexual activity at each visit in our unadjusted, partially and fully adjusted analysis. In the fully adjusted analysis, almost every model that included perceived peer or perceived family attitude scores as the predictor variable, the odds of reported a sexual behavior decreased after adjusting for age, race, time and the other two attitudinal scores. When personal attitude scores were the predictor variable, odds ratios also generally decreased after partial or full adjustment. Participant’s personal attitudes towards abstinence appear to be a stronger predictor of reporting sexual behaviors than perceptions of peer or family attitudes as demonstrated by the greater number of significant odds ratios after controlling for covariates and the larger odds ratios in instances where more than one attitude scale predicted a behavior (e.g., vaginal sex). When these models were stratified by age at enrollment (data not shown), personal attitudes were most influential at younger ages but none of the attitude measures were significant by age 17. Perceived peer and perceived family attitudes were not consistent predictors of any of the sexual behaviors and did not appear to become more influential as enrollment age increased. Similarly, personal attitudes appeared to be a stronger predictor of sexual behavior regardless of reported penile-vaginal sexual experience at enrollment relative to perceived peer or perceived family attitudes (data not shown).
Table 3.
Unadjusted and Adjusted Odds of Reporting Sexual Behaviors where Each Attitudinal Scale Was the Primary Predictor
| Sexual Behaviors | Unadjusted Models | Partially Adjusted Models (adjust for age, race, time) | Fully Adjusted Models (Adjust for all variables*) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Personal | Peer | Family | Personal | Peer | Family | Personal | Peer | Family | |
| Deep kissing | 1.68 | 1.61 | 1.44 | 1.54 | 1.40 | NS | NS | NS | NS |
| Breast touching | 3.63 | 1.90 | 1.99 | 3.38 | 1.59 | 1.51 | 2.37 | NS | NS |
| Genital touched | 2.90 | 1.93 | 1.86 | 2.82 | 1.81 | 1.49 | 2.78 | 1.63 | NS |
| Touch partner genital | 2.55 | 1.66 | 1.62 | 2.29 | 1.53 | 1.30 | 2.31 | NS | NS |
| Oral sex received | 2.02 | 1.72 | 1.28 | 1.69 | 1.42 | NS | 2.46 | 1.44 | 0.63 |
| Oral sex given | 1.54 | NS | NS | NS | NS | NS | 1.88 | NS | NS |
| Anal sex | NS | 1.79 | NS | NS | 1.99 | NS | NS | 2.06 | NS |
| Vaginal sex | 3.49 | 2.18 | 1.93 | 3.83 | 2.08 | 1.57 | 5.32 | 1.96 | NS |
NS=Non-significant
Age, race, time as well as personal, perceived peer and perceived family attitudes towards teen sex.
DISCUSSION
Our findings support and extend previous research on adolescent sexual behavior development. Similar to other studies, [2, 5, 8, 23] we found that reporting of sexual behaviors increases with time. Reporting penetrative sex (e.g., oral or vaginal) increased significantly with time, except anal sex which remained uncommon. Reporting non-penetrative acts (e.g., genital touching) varied greatly at enrollment but was uniform by the final study visit. These findings suggest that anal sex is an uncommon part of these adolescents’ sexual repertoires.[5, 7, 24] It also suggests that adolescents’ sexual repertoires evolve to encompass a larger set of behaviors. Although longitudinal studies examining how sexual behaviors cluster during sexual encounters are needed to further clarify this finding, our conclusion is supported by recent studies examining clustering of sexual behaviors.[7]
Also similar to previous studies[5, 25, 26], we noted that sexual behaviors varied by age and race. Older adolescents were more likely to report most of the behaviors examined. Regarding racial differences, Black participants were more likely to report giving oral sex compared to participants of other races. Although our sample was predominantly Black, we were powered to show differences between Black and non-Black participants. Assuming 10% of our subjects engaged in oral sex (this was the minimum over the years of the study) we had 80% power to detect an odds ratio of 4.92 at an alpha of 0.05 with only one year of data. Given that we have multiple years of data, our power was even higher. The observed differences might reflect the overall high prevalence of sexual behaviors in our cohort or real racial differences in sexual behavior preference.
We found that the relationship between attitudes towards adolescent sexual activity and adolescents’ sexual behaviors did not change over time as reflected by the lack of significance in the time x attitude score interaction terms. Female adolescents’ personal, perceived peer and perceived family attitudes towards adolescent sexual activity all became more permissive with time. However, in adjusted analyses personal importance of abstinence was the strongest predictor of adolescents’ sexual behavior reporting. Although perceived peer beliefs about when to have sex and perceived family belief that sex during adolescence is negative were predictive of some sexual behaviors, their influence was inconsistent showing no clear, clinically meaningful associations. This contrasts with previous studies’ which have found that family norms have the greatest influence on early adolescent sexual behaviors with peer norms becoming increasingly important with age.[12–14, 16, 27–29] When interpreting these findings, it is important to acknowledge that each of the social referents whose attitudes we examined are likely influenced by each other in complicated ways that our analysis has not completely disentangled. In addition, it is important to remember that sexual behaviors are not simply a function of personal attitudes and perceived social contextual norms but of multiple personal, social and broader environmental factors. Still, our findings remain interesting given the premise of several health behavior theories commonly used to understand and predict sexual behaviors, such as the Theory of Reasoned Action, that posit that perceived social norms are an important influence on behaviors.[30]
Our findings highlight the important influence of adolescents’ personal attitudes on their sexual behaviors. These findings suggest that, regardless of adolescents’ age or penile-vaginal sexual experience, interventions aimed at influencing adolescents’ personal attitudes towards sex may be most effective at changing their sexual behaviors. More specifically, our data suggest that effective interventions may be those that begin early and increase adolescent females’ perceptions of the importance of waiting to engage in sex until certain milestones are reached (e.g., love, maturity, marriage) or that help them contemplate whether becoming sexually active is consistent with their personal values. It is possible that love, emotional commitment or attachment to a partner – or the perception that these states had been achieved – could have been the predominant motivator for participants’ decisions to initiate or extend their sexual repertoires. This analysis cannot determine whether these factors were the case. What we do know is that for an unfortunate number of female adolescents, their initial sexual episode was tinged with regret.[31–34] This regret (when excluding sexual victimization) may stem from a mismatch between a female adolescent’s perception of the status of the romantic relationship at the time the sexual act took place and her reevaluation of that relationship at a later point in time. Hence, our suggestion that a potential intervention approach may be to help female adolescents explore their personal definitions for love, marriage and other romantic commitments.
Our study has several important strengths. The cohort design allows us to assess the directionality of observed associations between reported behavioral change and attitudes. We used a multivariate analytic technique that allowed simultaneous examination of three attitudinal factors affecting adolescent sexual behaviors while accounting for the repeated measures design.
There are several key limitations of this analysis. We recruited a convenience sample of urban adolescents from a population at high-risk for early sexual involvement and teen pregnancy. While our findings may not easily apply to other adolescent populations, such as those with lower levels of early sexual involvement, they offer important insights about how sexual behaviors as well as personal and perceived attitudes towards adolescent sex vary over time in a high-risk adolescent female population. Our population had high levels of sexual involvement prior to enrollment meaning we cannot disentangle the effects of prior sexual activities on the behaviors reported during the study period. Ideally, sexual behavior development should be examined over time in a cohort of female adolescents engaging in few to no sexual behaviors at enrollment. Our primary outcome measures assessed heterosexual activities and therefore do not capture same-gender sexual behaviors. Another potential limitation is that we measured perceived, not actual, peer and family attitudes towards teen sexual behavior. However, this was intentional as health behavior theories posit that it is an individual’s perceptions of and internalization of perceived social norms that has proximal effects on health behaviors.[35] Finally, our measures of personal, perceived peer and family attitudes were developed and refined as part of this study. Consequently their validity has not been evaluated in other contexts. While confirmatory factor analysis showed items for each scale loaded into a single domain, additional evaluations using other samples are needed to confirm generalizability. Although our findings are significant, it is important to keep in mind that this analysis does not include measures of social or environmental factors known to influence adolescent sexual behaviors such as partner, sibling or community norms variables.
CONCLUSION
Adolescents’ sexual behaviors as well as their personal and perceived social attitudes towards adolescent sex change during adolescence. However, an adolescent’s personal attitudes towards sex, particularly perceived importance of abstinence, appears to be the strongest predictor of reported sexual behaviors. The message for parents, health providers and intervention developers is that efforts to influence female adolescents’ attitudes towards sex, particularly abstinence, may be an important approach to reducing engagement in sexual behaviors that increase the risk for pregnancy and STIs.
Acknowledgments
Earlier versions of these analyses were presented at the Society for Adolescent Medicine (SAM) annual meeting in Denver, Colorado in March, 2007; at the 7th Annual Women’s Health Research Day sponsored by the Center for Women’s Health Research at the University of North Carolina in Chapel Hill, NC on April 5, 2006; and, at The Robert Wood Johnson Clinical Scholars Program 2005 National Meeting on November 17, 2005.
Funding sources:
This project was funded by the National Institute of Allergy & Infectious Disease to the Midwest Sexually Transmitted Diseases and Topical Microbicide Cooperative Research Center (U19 AI 31494); the National Institute for Child Health and Development, Development of Relationship Dynamics Related to STI/HIV Risk among Adolescent and Young Adult Women (NICHD, R01 HD044387); the Robert Wood Johnson Clinical Scholars Program at the University of North Carolina-Chapel Hill; and by the NIH Roadmap Multidisciplinary Clinical Research Career Development Award Grant (1 KL2 RR024154-01) from the National Institutes of Health. The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH. Information on NCRR is available at http://www.ncrr.nih.gov/. Information on Re-engineering the Clinical Research Enterprise can be obtained from http://nihroadmap.nih.gov/clinicalresearch/overview-translational.asp.
Abbreviations
- STI
sexually transmitted infection
- GEE
generalized estimating equations
Footnotes
Author Contributions
Study Design: Aletha Akers, Dennis Fortenberry, Donald Orr, Melanie Gold, Ada A. Adimora
Data Collection: Dennis Fortenberry, Donald Orr
Data analysis: Aletha Akers,
Data interpretation: Aletha Akers, Dennis Fortenberry, Melanie A. Gold
Manuscript preparation: Aletha Akers, Dennis Fortenberry, Melanie A. Gold, Ada A. Adimora, Donald Orr
Conflicts of Interest: The authors have no conflicts of interest to report.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Contributor Information
Aletha Y. Akers, Email: aakers@mail.magee.edu, Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Pittsburgh, 300 Halket Street, Pittsburgh, Pennsylvania 15213-3180, Office: (412) 641-8756, Fax: (412) 641-1133.
Melanie A. Gold, Clinical Associate Professor of Pediatrics, Department of Pediatrics, Division of Adolescent Medicine, University of Pittsburgh School of Medicine, Staff Physician and Research Coordinator for Student Health Services, Division of Student Affairs, University of Pittsburgh Student Health Service.
James E. Bost, Associate Professor of Medicine, Associate Professor of Clinical and Translational Science, University of Pittsburgh, Data Center, Center for Research on Health Care.
Ada A. Adimora, Division of Infectious Disease, Department of Internal Medicine, University of North Carolina Chapel Hill.
Donald P. Orr, Professor of Pediatrics, Nursing Research and Nutrition/Dietetics Director, Section of Adolescent Medicine, Department of Pediatrics, Indiana University
J. Dennis Fortenberry, Professor of Pediatrics, Section of Adolescent Medicine, Department of Pediatrics, Indiana University
References
- 1.Committee on Adolescent Health Care Services and Models of Care for Treatment P, and Healthy Development. Adolescent Health Services:Missing Opportunities. Washington, D.C: National Research Council; 2009. [Google Scholar]
- 2.Dariotis JK, Sonenstein FL, Gates GJ, et al. Changes in sexual risk behavior as young men transition to adulthood. Perspect Sex Reprod Health. 2008 Dec;40(4):218–225. doi: 10.1363/4021808. [DOI] [PubMed] [Google Scholar]
- 3.Brooks-Gunn J, Furstenberg FF., Jr Adolescent sexual behavior. Am Psychol. 1989 Feb;44(2):249–257. [PubMed] [Google Scholar]
- 4.Brooks-Gunn J, Paikoff R. Sexuality and developmental transitions during adolescence in Health Ricks and developmental transitions during Adolescence. New York: Cambridge University Press; 1997. [Google Scholar]
- 5.Lindberg LD, Jones R, Santelli JS. Noncoital sexual activities among adolescents. J Adolesc Health. 2008 Sep;43(3):231–238. doi: 10.1016/j.jadohealth.2007.12.010. [DOI] [PubMed] [Google Scholar]
- 6.Halpern CT, Waller MW, Spriggs A, et al. Adolescent predictors of emerging adult sexual patterns. J Adolesc Health. 2006 Dec;39(6):926, e921–910. doi: 10.1016/j.jadohealth.2006.08.005. [DOI] [PubMed] [Google Scholar]
- 7.Hensel DJ, Fortenberry JD, Orr DP. Variations in coital and noncoital sexual repertoire among adolescent women. J Adolesc Health. 2008 Feb;42(2):170–176. doi: 10.1016/j.jadohealth.2007.07.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Porter CP, Ronis DL, Oakley DJ, et al. Early adolescents' sexual behaviors. Issues Compr Pediatr Nurs. 1999 Apr–Sep;22(2–3):129–142. doi: 10.1080/014608699265338. [DOI] [PubMed] [Google Scholar]
- 9.Gates GJ, Sonenstein FL. Heterosexual genital sexual activity among adolescent males: 1988 and 1995. Family planning perspectives. 2000 Nov–Dec;32(6):295–297. 304. [PubMed] [Google Scholar]
- 10.Schwartz IM. Sexual activity prior to coital initiation: a comparison between males and females. Archives of sexual behavior. 1999 Feb;28(1):63–69. doi: 10.1023/a:1018793622284. [DOI] [PubMed] [Google Scholar]
- 11.Smith EA, Udry JR. Coital and non-coital sexual behaviors of white and black adolescents. American journal of public health. 1985 Oct;75(10):1200–1203. doi: 10.2105/ajph.75.10.1200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Whitbeck L, Yoder K, Hoyt D, et al. Early adolescent sexual activity: A developmental study. Journal of Marriage and the Family. 1999;61:934–946. [Google Scholar]
- 13.Crockett LJ, Bingham CR, Chopak JS, et al. Timing of first sexual intercourse: the role of social control, social learning, and problem behavior. J Youth Adolesc. 1996 Feb;25(1):89–111. doi: 10.1007/BF01537382. [DOI] [PubMed] [Google Scholar]
- 14.Meschke LL, Bartholomae S, Zentall SR. Adolescent sexuality and parent-adolescent processes: Promoting healthy teen choices. Fam Relat. 2000 APR;49(2):143–154. doi: 10.1016/s1054-139x(02)00499-8. [DOI] [PubMed] [Google Scholar]
- 15.Henry DB, Schoeny ME, Deptula DP, et al. Peer selection and socialization effects on adolescent intercourse without a condom and attitudes about the costs of sex. Child Dev. 2007 May–Jun;78(3):825–838. doi: 10.1111/j.1467-8624.2007.01035.x. [DOI] [PubMed] [Google Scholar]
- 16.Sieving RE, Eisenberg ME, Pettingell S, et al. Friends' influence on adolescents' first sexual intercourse. Perspect Sex Reprod Health. 2006 Mar;38(1):13–19. doi: 10.1363/psrh.38.013.06. [DOI] [PubMed] [Google Scholar]
- 17.Kinsman SB, Romer D, Furstenberg FF, et al. Early sexual initiation: the role of peer norms. Pediatrics. 1998 Nov;102(5):1185–1192. doi: 10.1542/peds.102.5.1185. [DOI] [PubMed] [Google Scholar]
- 18.Romo LF, Lefkowitz ES, Sigman M, et al. A longitudinal study of maternal messages about dating and sexuality and their influence on Latino adolescents. J Adolesc Health. 2002 Jul;31(1):59–69. doi: 10.1016/s1054-139x(01)00402-5. [DOI] [PubMed] [Google Scholar]
- 19.Hensel DJ, Fortenberry JD, Orr DP. Situational and relational factors associated with coitus during vaginal bleeding among adolescent women. Journal of sex research. 2007 Aug;44(3):269–277. doi: 10.1080/00224490701443940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Katz BP, Fortenberry JD, Tu W, et al. Sexual behavior among adolescent women at high risk for sexually transmitted infections. Sexually transmitted diseases. 2001 May;28(5):247–251. doi: 10.1097/00007435-200105000-00001. [DOI] [PubMed] [Google Scholar]
- 21.Ott MA, Shew ML, Ofner S, et al. The influence of hormonal contraception on mood and sexual interest among adolescents. Archives of sexual behavior. 2008 Aug;37(4):605–613. doi: 10.1007/s10508-007-9302-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.StataCorp. Stata Statistical Software: Release 10. Version 10. College Station, TX: StataCorp LP; 2007. [Google Scholar]
- 23.Stanton B, Fang X, Li X, et al. Evolution of risk behaviors over 2 years among a cohort of urban African American adolescents. Arch Pediatr Adolesc Med. 1997 Apr;151(4):398–406. doi: 10.1001/archpedi.1997.02170410072010. [DOI] [PubMed] [Google Scholar]
- 24.Schuster MA, Bell RM, Kanouse DE. The sexual practices of adolescent virgins: genital sexual activities of high school students who have never had vaginal intercourse. American journal of public health. 1996 Nov;86(11):1570–1576. doi: 10.2105/ajph.86.11.1570. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Brewster KL, Tillman KH. Who's doing it? Patterns and predictors of youths' oral sexual experiences. J Adolesc Health. 2008 Jan;42(1):73–80. doi: 10.1016/j.jadohealth.2007.08.010. [DOI] [PubMed] [Google Scholar]
- 26.Eaton DK, Kann L, Kinchen S, et al. Youth risk behavior surveillance--United States, 2007. MMWR Surveill Summ. 2008 Jun 6;57(4):1–131. [PubMed] [Google Scholar]
- 27.Buhi ER, Goodson P. Predictors of adolescent sexual behavior and intention: a theory-guided systematic review. J Adolesc Health. 2007 Jan;40(1):4–21. doi: 10.1016/j.jadohealth.2006.09.027. [DOI] [PubMed] [Google Scholar]
- 28.Shoveller JA, Johnson JL, Langille DB, et al. Socio-cultural influences on young people's sexual development. Soc Sci Med. 2004 Aug;59(3):473–487. doi: 10.1016/j.socscimed.2003.11.017. [DOI] [PubMed] [Google Scholar]
- 29.Santelli JS, Kaiser J, Hirsch L, et al. Initiation of sexual intercourse among middle school adolescents: the influence of psychosocial factors. J Adolesc Health. 2004 Mar;34(3):200–208. doi: 10.1016/j.jadohealth.2003.06.004. [DOI] [PubMed] [Google Scholar]
- 30.Health Behavior and Health Education: theory, Research, and Practice. 3. San Francisco, CA: Jossey-Bass; 2002. [Google Scholar]
- 31.Roese NJ, Pennington GL, Coleman J, et al. Sex Differences in Regret: All For Love or Some For Lust? Pers Soc Psychol Bull. 2006;32(6):770–780. doi: 10.1177/0146167206286709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Oswalt SB, Cameron KA, Koob JJ. Sexual regret in college students. Archives of sexual behavior. 2005 Dec;34(6):663–669. doi: 10.1007/s10508-005-7920-y. [DOI] [PubMed] [Google Scholar]
- 33.Martino SC, Collins RL, Elliott MN, et al. It's better on TV: does television set teenagers up for regret following sexual initiation? Perspect Sex Reprod Health. 2009 Jun;41(2):92–100. doi: 10.1363/4109209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Wight D, Parkes A, Strange V, et al. The quality of young people's heterosexual relationships: a longitudinal analysis of characteristics shaping subjective experience. Perspect Sex Reprod Health. 2008 Dec;40(4):226–237. doi: 10.1363/4022608. [DOI] [PubMed] [Google Scholar]
- 35.Montano DE, Kasprzyk D. The theory of reasoned action and the theory of planned bahavior. In: Glanz K, Rimer BK, Lewis FM, editors. Health Behavior and Health Education: Theory, Research and Practice. 3. San Francisco, CA: Jossey-Bass; 2002. [Google Scholar]