Sample composition is shown in . There were significant (p < .001) between-study differences in the proportion of minority ethnic groups and those in social rented housing.
Sample characteristics: socio-demographic information and sexual behavior according to partner type
Of the eligible sample (N = 10,250), 3,766 teenagers reported sexual behavior with either a same-sex or an opposite-sex partner or both, and are included in multivariate analyses. Almost four in 10 teenagers (39.3%, N = 3,565 reported heterosexual intercourse without any report of same-sex behavior, and 2.3% (N = 201) reported same-sex genital contact (). Most teenagers reporting same-sex genital contact had also experienced heterosexual intercourse (last row of table, for combined sexes the bisexual group, 1.6% of the sample, N = 137, comprised 72% of those with same-sex partners, allowing for weighting). A minority of participants reporting a same-sex partner (nine of 201) did not answer questions concerning opposite-sex partners and are treated in these analyses as having same sex partners only. Similarly, 330 participants reported an opposite-sex partner but did not answer questions concerning same-sex partners and are treated as having opposite-sex partners only.
Girls were more likely than boys to report same-sex kissing with tongues and heterosexual intercourse (both p < .001), but there were no other gender differences in reporting of sexual behavior. Although a slightly higher percentage of SHARE teenagers reported heterosexual intercourse than in the RIPPLE study (p < .01), there were no other significant (p < .05) between-study differences in rates of other sexual behaviors with same- or opposite-sex partners.
Among boys, the prevalence of unwanted FS was higher for first homosexual genital contact than for first heterosexual intercourse in the exclusively heterosexual group (). Among girls, there were no differences in rates of unwanted sex according to partner type. In boys and girls, the prevalence of sexual risk-taking was higher for those with partners of both sexes, as compared with teenagers with exclusively opposite-sex partners. Similar effects of partner type were apparent in both the RIPPLE and SHARE studies when examined separately (not shown).
Prevalence of unwanted first sex and sexual risk according to partner type, by gender: univariate comparisons
We now consider attitudinal and behavioral factors reported at baseline (age 13 or 14 years) that may confound differences in sexual outcomes. Univariate analyses revealed some significant (p < .05) or borderline significant (p < .08) differences in attitudinal and behavioral measures according to partner type (). Teenagers with same-sex partners were more religious and more knowledgeable about sexual health, and (boys) were more likely to expect tertiary education than the exclusively heterosexual group. Boys with same-sex partners had lower self-esteem, and girls reported poorer communication with their mother. Most of these differences were also seen when comparing teenagers reporting bisexual behavior with exclusively heterosexual counterparts. Overall, several factors in the bisexual group were protective against sexual risk-taking (greater knowledge, religiosity, and expectations of tertiary education). However, girls with bisexual behavior reported factors associated with greater sexual risk (poor communication with mother, substance use, and expectation of early parenthood).
Attitudinal and behavioral differences according to partner type, by gender: univariate analyses
Results are provided for stage one multivariate analysis using both complete case information and the imputed data set. Coefficients/odds ratios are similar, although for pressure and regret outcomes the imputed data set shows a greater risk associated with same-sex partner for boys. This is consistent with a reduction in bias because of lower disclosure of negative experiences by teenagers with same-sex partners. In this study, we describe results using the imputed data set.
Partner pressure and regret were compared for first same-sex genital contact and opposite-sex intercourse (among teenagers not reporting same-sex genital contact). The latter group were older than the same-sex group (mean ages respectively, 14.4 years, SD: 1.15 and 13.4 years, SD: 2.9, p < .001), and were more likely to have expected sex (55% vs. 25%, p < .001). Age and expectation of sex were strongly associated with the two outcomes, and were included as covariates at stage 1 (). There was a strong gender difference in the effect of partner type. Boys with a same-sex partner were more likely to report partner pressure and regret, although there was no effect of partner type among girls. The only potential confounder for the effect of partner type on unwanted sex among boys arising from univariate analyses in was self-esteem. However, there was only a small effect of adjusting for self-esteem on odds associated with same-sex partner in stage 2, .
Multivariate analysis of partner pressure and regret according to partner type, comparing first genital contact with same-sex partner with first heterosexual vaginal intercourse
Dividing the same-sex partner group and comparing again with boys reporting opposite-sex partners only (not in ), the effects were similar for boys reporting same-sex genital contact only (pressure: OR = 2.11, 95% CI = .75–5.91; regret: OR = 3.73, 95% CI = 1.51–9.25) and boys who reported bisexual behavior (pressure: OR = 2.80, 95% CI = 1.23–6.35; regret: OR = 1.79, 95% CI = 1.00–3.22).
The RIPPLE data set contained a wider range of contextual measures, and indicated that same-sex encounters were more likely to involve alcohol or drugs and no prior partner relationship, although no more likely to involve an older partner. Further exploration (not shown) confirmed boys' greater likelihood of negative feelings after first same-sex genital contact, taking account of additional contextual information.
Sexual risk was compared for teenagers reporting bisexual behavior and those reporting heterosexual intercourse only (, stage 1). Bisexual behavior was significantly associated with greater risk (boys: three measures, girls: four measures). Baseline differences in early parenthood, substance use, and poor communication with mother appeared to be potential confounders of these effects among girls (). For girls, effects of partner type were reduced but remained significant after adding these covariates in stage 2, . For boys, there was less effect of adding baseline covariates.
Multivariate analysis of sexual risk according to partner type, comparing teenagers with both same- and opposite-sex partners and teenagers with opposite-sex partners only
Further adjusting the pregnancy models for characteristics of sexual behavior (age and partner pressure at first heterosexual intercourse, number of partners, not shown in ) attenuated the risk associated with bisexual behavior to nonsignificance among girls (OR = 1.85, 95% CI = .98–3.51), but not among boys (OR = 3.53, 95% CI = 1.86–6.67).