Personal and Social Influences Regarding Oral Sex Among African American Female Adolescents

Abstract

Objective

To identify personal and social factors associated with performing oral sex among female adolescents.

Methods

Sexually active African American female adolescents (n = 715) recruited from sexually transmitted infection (STI) clinics were assessed for self-esteem, sexual sensation seeking, unprotected vaginal sex (UVS), self-efficacy to communicate about sex and to refuse sex, fear of negotiating condoms, relationship power, peer norms surrounding risky sexual behavior, ever having performed oral sex, and three vaginally acquired STIs.

Results

Prevalence for at least one STI was 29%. More than half reported performing oral sex. Controlling for age, performing oral sex was associated with relatively higher sexual sensation seeking, any UVS in past 60 days, relatively lower self-efficacy to refuse sex, and having peer norms supportive of risky sexual behaviors.

Conclusions

Given the potential for epidemic spread of orally acquired STIs to populations of female adolescents residing in communities with high rates of STI prevalence, this initial research provides guidance for intervention development and expanded research efforts.

Introduction

Among sexually active adolescents, oral sex has emerged as normative. Indeed, recent research has shown that oral sex may be more prevalent than vaginal intercourse.^1–3 Although national estimates have not been published, findings from nonprobability studies suggest that performing unprotected oral sex creates risk for the acquisition of a sexually transmitted infection (STI), such as Chlamydia trachomatis,⁴ pharyngeal gonococcal infection,^5,6 herpes simplex virus (HSV),^7,8 human papillomavirus (HPV),^4,9 and human immunodeficiency virus (HIV).^4,10 Consequences of acquiring an oral STI may be severe, as recent findings have identified oral HPV infection as a central cause of head and neck squamous cell carcinoma,^9,11 in particular, oropharyngeal cancer.¹² Many adolescents are not aware that engaging in oral sex can be a significant risk behavior for the acquisition of bacterial and viral STIs.^4–10 Thus, they tend to view oral sex as less risky than penile-vaginal sex in terms of acquiring an STI and, therefore, are less likely to use barrier protection when engaging in oral sex.^2,3,13–16

Little is known about the factors that influence adolescents' decision to engage in oral sex. A study among adolescents in the Western United States found their reasons for having oral sex included personal, relational, and peer influences, such as “lower self-esteem, pleasure, improving intimate relationships and reducing risks associated with vaginal sex.”¹⁷ Although informative, the study did not differentiate between reasons for receiving oral sex vs. reasons for performing oral sex, and the results were based on adolescents' perceptions of other teens' reasons rather than their own.¹⁷ In another study of tenth grade adolescents, the prevalence of oral sex and social influences of engaging in oral sex were examined. Results showed that 40% of adolescents had engaged in oral sex and were unlikely to report the use of STI protection during oral sex. Also, peer influence, that is, having the perception that their best friend engaged in oral sex, was significantly associated with their own oral sex behavior.¹

Given the risk of STI acquisition, specifically for female adolescents when performing oral sex, more in-depth research is warranted to understand the influences of engaging in oral sex so that STI preventive interventions can be designed. Emerging evidence suggests that certain environmental factors may have an effect on adolescents' sexual behaviors.¹⁸ Although many researchers and experts in the field of adolescent sexual health endorse an ecological approach for understanding sexual risk behaviors,^19–21 much of the empirical work on adolescent sexual health remains devoid of an assessment of broader, environmental factors.²² Thus, to gain a better understanding of what factors influence female adolescents to engage in oral sex, an assessment of their personal characteristics and specific environmental factors that may contribute to their engaging in the behavior is necessary.

The purpose of this study was to identify the personal and social (i.e., environmental) influences of engaging in oral sex among a clinical sample of African American female adolescents, a population in the United States at heightened risk for STIs, including HIV.^23,24 We included self-esteem and sexual sensation seeking, as these personal factors have been associated with sexual risk behavior in previous studies.²⁰ Moreover, the social influences we focused on were relational factors (e.g., power in relationship, communication with sex partner, fear of negotiating condom use) and peer norms, as these factors have been shown to exert a direct influence on sexual risk behaviors among adolescents.²⁰

Materials and Methods

Study sample

Participants were African American female adolescents enrolled in a randomized trial of an HIV prevention program. Data collected at baseline were used for this study. Recruitment sites were an urban, publicly funded STI clinic, a teen clinic based in a large public hospital, and a family planning clinic (all clinics were located in the same urban area of a major city located in the Southeastern United States). The urban area selected for the study is one that experiences high prevalence rates of sexually transmitted diseases (STDs), thereby creating high-risk conditions for the transmission of STIs. From March 2002 through August 2004, project recruiters screened female teens to assess eligibility. Adolescents were eligible to participate if they were African American females, 15–21 years old, who reported vaginal sexual activity in the previous 60 days. Exclusion criteria were being married or pregnant or attempting to become pregnant. Of 1558 screened, 874 adolescents were eligible and were asked to participate in the study. The study achieved an 82% participation rate (n = 715). The Institutional Review Board at Emory University approved the study protocol before implementation.

Data collection

As part of the intervention trial, adolescents were asked to provide a self-collected vaginal swab, which was assayed using nucleic acid amplification technology to detect three prevalent sexually transmitted pathogens: Trichomonas vaginalis, C. trachomatis, and Neisseria gonorrhoeae.^25,26 Based on evidence suggesting the possibility of decreased reporting bias, all measures were assessed using audio-computer-assisted self-interviewing (ACASI).²⁷ By providing a voice track that delivered each question to adolescents through headphones, ACASI technology may reduce problems that otherwise would have been posed by low literacy.

Measures

Selection of scales was guided by the underlying theoretical model and by our prior research using valid measures that have shown satisfactory reliability in assessing relevant behaviors and constructs among African American female adolescents.

Performing oral sex

We asked adolescents whether they had “ever performed oral sex on a guy?” Responses were yes/no.

Personal factors

Sexual sensation seeking has been associated with engaging in sexual risk behaviors among various populations and was included as a personal characteristic in the present study.^28,29 Sexual sensation seeking was assessed using an 8-item scale (the Sexual Sensation Seeking Scale for Adolescents [SSSA]), which was developed and validated by DiClemente et al. for use with an African American female adolescent population as part of a National Institute of Mental Health (NIMH)-funded HIV intervention study.³⁰ A sample item is: I enjoy having sex on the spur of the moment. Each item required a response based on a 4-point Likert-type scale: 1(strongly disagree) to 4 (strongly agree). In a previous study with African American female adolescents, the SSSA demonstrated strong internal consistency and showed moderate stability and satisfactory construct validity.³⁰ In the present study, the scale produced a satisfactory interitem reliability coefficient (Cronbach's alpha = 0.72)

Low self-esteem was cited as a reason that adolescents perceived that other adolescents engaged in oral sex and was assessed as a personal characteristic that may influence female adolescents to perform oral sex in the present study.¹⁷ We assessed self-esteem using the Rosenberg Self-Esteem Scale (RSES).³¹ This 10-item scale measures global self-esteem with four response choices ranging from 1 (strongly disagree) to 4 (strongly agree). The RSES has been used widely with diverse populations, including African American female adolescents,³² and has demonstrated validity and reliability.³³ In the present study, reliability was satisfactory (α = 0.85).

In addition, we assessed a behavioral variable, that is, if adolescents engaged in unprotected vaginal sex (UVS) during the past 60 days. Adolescents were asked how many times in the past 60 days they had penile-vaginal intercourse. Then, they were asked: Of those numbers of times you had sex, how many times did you use a male condom? A variable was created that indicated the percentage of condom use, and we dichotomized the variable to represent either 100% condom use (no risk) or <100% condom use (some risk).

Relational factors

Five distinct relational factors were assessed as social or environmental influences of female adolescents performing oral sex. A 6-item scale that was created for young African American women was used to assess adolescents' self-efficacy to communicate with their male sex partners about sex-related issues.³⁴ The partner sexual communication self-efficacy scale included such items as: With a sex partner, how hard is it for you to ask how many sex partners he has had? With a sex partner, how hard is it for you to ask if he would use a condom? Each item required a response based on a 4-point Likert-type scale: 1 (very hard) to 4 (very easy). Interitem reliability of the scale in the present study was adequate (α = 0.83).

Adolescents' perceived power in their relationships with male sex partners was also assessed with the Sexual Relationship Power Scale. The scale has been validated in a previous study specifically designed to establish its psychometric properties with young women.³⁵ The scale comprises 12 items and, in the present study, achieved adequate interitem reliability (α = 0.80). Sample items included: Most of the time we do what my partner wants to do. My partner won't let me wear certain clothes. Each item required a response based on a 4-point Likert-type scale: 1 (strongly disagree) to 4 (strongly agree).

Adolescents' level of fear about condom negotiation was assessed with the Worry About Aversive Sexual Outcomes Scale. This scale is intended to describe the frequency with which one feels fear or worry associated with negotiating the use of condoms with sex partners. This scale comprised 7 items and was validated with a sample of African American adolescents in a previous study.³⁶ Participants were asked how frequently they feared that if they talked about using condoms with a sex partner, the sex partner would “ignore my request,” “threaten to hit me,” “threaten to leave me,” “swear at me or call me ugly names,” “hit, push, or kick me,” “leave me,” and “go out with other girls.” Each item required a response based on a 5-point Likert-type scale: 1 (never) to 5 (always). Reliability for the scale in the present study was sufficient (α = 0.81).

Adolescents' self-efficacy to refuse sex was assessed using a 7-item subscale developed by Cecil and Pinkerton that assessed perceptions of individuals' ability to refuse sexual intercourse.³⁷ Sample items included: How sure are you that you would be able to say NO to having sex with someone you want to date again? How sure are you that you would be able to say NO to having sex with someone who you want to fall in love with you? How sure are you that you would be able to say NO to having sex with someone who is pressuring you to have sex? Each item required a response based on a 4-point Likert-type scale: 1 (I definitely can't say no) to 4 (I definitely can say no). This scale achieved adequate reliability in the present study (α = 0.87).

Finally, we also assessed whether adolescents believed their partner might have had vaginal sex with another girl concurrently. We asked them specifically: During your relationship with your boyfriend, has he had vaginal sex with another girl? The response was either yes or no. We anticipated that those having this perception might be more likely to perform oral sex, rather than engage in penile-vaginal sex, as an STI risk-reduction strategy.

Peer factors

Normative beliefs or peer norms surrounding sexual risk behaviors were assessed with a scale developed by DiClemente and Wingood for the current study. Adolescents were asked to indicate “how many of your friends think that (1) it's okay to have vaginal or anal sex without a condom, (2) it's okay to have sex with someone you just met, (3) cheating on your partner is okay, (4) it's safe to have sex when you are high on drugs or alcohol, and (5) you don't have to use a condom with someone you know well.” Each item required a response based on a 5-point Likert-type scale: 1 (none) to 5 (all). The scale achieved adequate reliability in the present study (α = 0.77).

Data analysis

Bivariate associations

All continuous scale variables were assessed for normality. We conducted a visual inspection of data plots and examined estimates of skew and kurtosis. We also conducted Kolmogorov-Smirnov tests, which provide inferential statistics on normality. None of the six assessed scales met tests of normality; all were significantly skewed. Several transformations (square root, log) were performed; however, despite these efforts, transformations were not effective in correcting skewness to acceptable levels for any of the scale measures (i.e., skewness <2.0 standard errors [SE] of skewness). Because nonnormally distributed variables can distort relationships and significance tests, we opted to dichotomize these variables by performing a median split. According to MacCallum et al.,³⁸ dichotomization of variables is justified when the distribution is highly skewed. Associations between dichotomous predictor variables and the outcome measure were assessed by the use of prevalence ratios, their 95% confidence intervals (CI), and their respective p values.

Multivariate associations

Correlates achieving a screening level of significance (p < 0.10) at the bivariate level were entered into a hierarchical (two blocks) multiple logistic regression model using a forward stepwise method in the second block. The first block included only adolescents' age (a control variable). The logistic regression model was used to calculate adjusted odds ratios (AOR), the 95% confidence intervals (CI), and corresponding p value.

Results

Characteristics of the sample

Average age of the adolescents was 17.8 years (standard deviation [SD] 1.72). The majority (65.3%) were students currently attending school. Nearly one third (29.4%) of the sample reported having a paying job. The median level of education fell between grades 10 and 11. Overall, 206 adolescents (28.8%) tested positive for at least one of the three vaginally acquired STIs (T. vaginalis, C. trachomatis, and N. gonorrhoeae). More than half (57.6%) reported they had performed oral sex on a male partner. Of those who had performed oral sex, 31.6% (n = 130) indicated they derived no pleasure, 27.7% (n = 114) stated they got a little pleasure, 29.7% (n = 122) indicated they got some pleasure; only 10.9% (n = 45) said they got a lot of pleasure from performing oral sex.

Bivariate associations

Table 1 displays bivariate findings pertaining to performing oral sex. Table 1 also displays all relevant descriptive information. Adolescents were more likely to perform oral sex if they reported (1) higher sexual sensation seeking, (2) lower self-esteem, (3) engaging in unprotected vaginal sex in the past 60 days, (4) lower sexual communication self-efficacy, (5) greater fears related to consequences of negotiating condoms, (6) lower self-efficacy to refuse sex, and (7) perceived peer norms supportive of unsafe sex behaviors.

Table 1.

Bivariate Associations with Performing Oral Sex Among African American Female Adolescents

	% Oral Sex	PR	95% CI	p
Personal factors
Sexual sensation seeking
Low (421)	44.7
High (292)	76.4	1.71	1.51-1.94	0.0001
Self-esteem
High (387)	51.9
Low (326)	64.4	1.24	1.09-1.41	0.001
Engaged in unprotected vaginal sex (past 60 days)
No (287)	47.0
Yes (423)	64.8	1.38	1.21-1.58	0.0001
Relational factors
Sex communication self-efficacy
High (380)	54.2
Low (333)	61.6	1.14	1.001-1.29	0.047
Perceived power in relationships
High (368)	54.9
Low (344)	60.8	1.11	0.98-1.26	0.113
Fear of condom use negotiation
Low (384)	52.1
High (329)	64.1	1.23	1.09-1.40	0.001
Sex refusal self-efficacy
High (342)	52.0
Low (371)	62.8	1.21	1.06-1.37	0.004
Male sex partner has other female sex partners
No (435)	54.9
Yes (161)	60.2	1.09	0.94-1.28	0.25
Peer factor
Peer norms supportive of unsafe behavior
No (376)	49.7
Yes (337)	66.5	1.34	1.18-1.52	0.0001

CI, confidence interval; PR, prevalence ratio.

Multivariate associations

Entering age as a covariate (age was significantly related to performing oral sex, t (709) = 8.78, p < 0.0001) and the seven factors significant in the bivariate analysis into a logistic regression equation, we found that in addition to age, four factors were retained in the final model: sexual sensation seeking, unprotected vaginal sex, sex refusal self-efficacy, and peer norms. The model was significant (χ² = 157.28, df = 5, p = 0.0001) and achieved an excellent fit with the data (goodness of fit χ² = 8.18, df = 8, p = 0.42) (Table 2) and showed that for each additional year of age, adolescents were about 1.5 times more likely to report having performed oral sex (AOR 1.46, 95% CI 1.31-1.62, p < 0.0001). Adolescents with higher sexual sensation seeking scores were >3 times as likely to have performed oral sex (AOR 3.3, 95% CI 2.31- 4.71, p < 0.0001). Adolescents reporting UVS in the past 60 days were nearly twice as likely, relative to than those not having UVS, to report performing oral sex (AOR 1.87, 95% CI 1.33-2.63, p < 0.0001). Adolescents with lower self-efficacy scores to refuse sex were 1.4 times more likely to report performing oral sex (AOR 1.42, 95% CI 1.01-1.99, p = 0.046). Adolescents whose perceived peer norms were supportive of risky sexual behaviors were 1.5 times more likely to have performed oral sex (AOR 1.53, 95% CI 1.09-2.15, p = 0.014).

Table 2.

Adjusted Odds Ratios for Personal and Social Influences of Performing Oral Sex

Correlate	AOR	95% CI	p value
Age	1.5	1.31-1.62	0.001
Higher sexual sensation seeking	3.3	2.31-4.71	0.0001
Having unprotected vaginal sex in past 60 days	1.9	1.33-2.63	0.0001
Lower self-efficacy to refuse sex	1.4	1.01-1.99	0.046
Having perceived peer norms supportive of risky sex	1.5	1.09-2.15	0.014

AOR, adjusted odds ratio.

Discussion

The present study examined the prevalence of performing oral sex and the personal and social influences among a population at heightened risk for STIs, including HIV. More than one half of the adolescents (58%) reported having ever performed oral sex. The lifetime prevalence of performing oral sex observed in the present study is slightly higher than in most studies of adolescents.^2,3,39,40 However, this may be attributable to the fact that our sample of adolescents were all sexually active. This lifetime prevalence rate indicates that although these female adolescents are engaging in vaginal sex, they still engage in oral sex and are at risk of acquiring an oral STI. We acknowledge that in the present study, we assessed performing oral sex vs. performing unprotected oral sex, the latter of which is associated with increased health risk. However, previous research with adolescents suggests that few use condoms during oral sex.¹³ Thus, future research should explore condom use during oral sex and associated risk factors with engaging in unprotected oral sex.

In age-adjusted, multivariate analyses, several factors were significantly related to performing oral sex: greater sexual sensation seeking, having unprotected vaginal sex in the past 60 days, lower self-efficacy to refuse sex, and having the perception that peers endorse unsafe sexual behaviors. The finding pertaining to the personality trait of sexual sensation seeking is particularly intriguing because it transcends traditional demographic and behavioral factors used to predict adolescents' sexual risk-taking practices. Previous research with other populations (e.g., men who have sex with men, college students) has examined this personality trait and found sexual sensation seeking to be associated with sexual risk behaviors.^28,29,41 Because little is known about this personality trait as a predictor of risky sexual behavior among adolescents, our findings contribute to the literature and suggest that sexual sensation seeking may be an important correlate of performing oral sex among African American female adolescents. An interesting point, however, is that we also found that female adolescents who engaged in oral sex were also more likely to engage in another sexual risk behavior, that is, unprotected vaginal sex. Previous research has suggested that adolescents who engage in oral sex may be trying to avoid the risk of acquiring an STI or of getting pregnant.¹⁷ As our findings demonstrate, however, rather than identifying risk compensation (i.e., females who report oral sex would be less likely to report unprotected vaginal sex), we identified a strong positive association between performing oral sex and unprotected vaginal sex. Thus, although many adolescents perceive oral sex as a protective or safe behavior, this multivariate finding supports the possibility that they are not using this behavior to replace an ostensibly more risky behavior (unprotected vaginal sex). This finding suggests that perhaps there is a subset of adolescents who engage in myriad sexual risk behaviors.^22,42 More research is needed to identify the correlates associated with this particularly high-risk group of adolescents.

Only one of the five assessed relational factors was retained in the final model. Adolescents who had lower self-efficacy to refuse sex were more likely to perform oral sex. It may be that for these female adolescents, because they lacked confidence to refuse sex, performing oral sex may be viewed as a strategy to avoid having vaginal sex. This possibility, of course, lies in contrast to the previous speculation that oral sex is not a compensatory behavior for penile-vaginal sex. Indeed, neither possibility can be ruled out until more in-depth research is conducted. In addition to holding the perception that oral sex is less risky, many adolescents and some adults for that matter, tend to equate sex with penile vaginal intercourse.⁴³ Adolescents may be adhering to certain cultural norms that prescribe oral sex as not sex per se and, therefore, feel good about their decision to engage in oral sex. Previous research has supported this normative belief where adolescents viewed oral sex as more acceptable and less of a threat to their values and beliefs than vaginal sex.² On the other hand, in our sample of African American female adolescents, we also found that oral sex was associated with UVS. Thus, lower self-efficacy to refuse sex may be a function of a generalized lack of confidence to refuse all forms of sex whether it is oral, vaginal, or anal sex. More research is needed to explore these relations among this population to determine if giving oral sex is truly a risk-avoidance strategy, a compromise for vaginal or anal sex, or a behavior engaged in because they lack the confidence to say no to their sex partners for all types of sex.

Adolescents who perceived that their friends endorsed unsafe sexual behaviors in general were more likely to perform oral sex. This finding is not surprising, given that peer norms surrounding sexual behaviors have been shown to be robust influences on both risky and protective sexual behavior.²⁰ When adolescents perceive that friends and similar-aged teens engage in risky sexual behavior, even if their perception is skewed, they are more likely to adopt those same behaviors. This finding is consistent with those of Halpern-Felsher et al.,² who found that the ninth grade adolescents in their study believed that more of their peers would have oral sex than vaginal sex in the near future. Prinstein et al.¹ similarly found that having the perception of a best friend engaging in oral sex was related to adolescents' own oral sex behavior. Although we did not ask adolescents directly to report how many of their friends think it is okay to perform oral sex specifically, these results suggest peer norms exert a similarly strong influence on African American female adolescents performing oral sex.

These findings have implications for the prevention of STIs among African American female adolescents. Healthcare providers and clinicians who screen and treat female adolescents for vaginally acquired STIs should also begin to screen for orally acquired STIs. Clinicians and health educators could also provide female adolescents with education about the STI risks of engaging in oral sex (e.g., pharyngeal gonorrhea, oral cancer). Indeed, the severity of STI risk associated with performing oral sex may be greatly underestimated or even dismissed entirely by some adolescents. ^4–10 Sexual risk reduction interventions designed for this adolescent population may benefit from incorporating information about the medical sequelae of orally acquired Chlamydia, gonorrhea, and especially HPV. Further, it is quite possible that our findings may not be unique to this clinic-based sample of African American female adolescents, thereby suggesting the same implications apply to populations of white and Hispanic adolescents as well as other populations of African American female adolescents.

Although relational factors have been widely addressed in previous interventions with female adolescents and young women,^44–46 our findings support an emphasis on teaching sexual refusal skills, thereby building self-efficacy for refusal of oral sex as well. It is noteworthy that oral sex and UVS were associated risk behaviors in this study. Thus, building refusal self-efficacy for both oral sex and UVS would appear to be a potentially valuable intervention approach. This emphasis on sexual refusal may be particularly unique to young females who have sex with male partners who are several years older.^47,48 Young women often may feel an obligation to engage in sex with their male partners; thus, addressing this perception as well as providing young women with the requisite skills for refusing sex without harming the relationship itself may be important assets for intervention programs.

The finding pertaining to sexual sensation seeking suggests that comprehensive sex education programs might benefit adolescent females by focusing on the novelty of using condoms especially during oral sex and promoting condom use during oral sex as an erotic sexual behavior.⁴⁹ In essence, rather than trying to change a personality trait (sensation seeking), a more expedient prevention approach may be to use that trait in a protective manner. With an ever expanding number of innovations in condom manufacturing, the prospect of sexually active adolescent females finding an erotic aspect to condom use (and possibly the use of lubricants added to condoms) may indeed promote safer sex behaviors. This seems like a reasonable approach, given that a personality trait, such as sexual sensation seeking, may be recalcitrant to efforts of change. Instead, it may be more effective to introduce these adolescents to flavored condoms, for example. Of course, sexual sensation seeking is not at all unique to young women. Indeed, if young women can achieve condom-protected oral sex, this form of sexual release may meet sexual sensation seeking needs of their partners as well. Male partners may need to use extrathin condoms to more fully experience sexual sensation from oral sex.

The findings also support the influence of peer norms and indicate that sex education programs should continue to address peer norms and perhaps include an emphasis on oral sex behaviors in addition to penile-vaginal sex.^20,44 It is possible that peer norms are optimally permissive for oral sex, given that adolescents may perceive this as a low-risk activity. The perception of safety, however, is most likely predicated on oral sex as a pregnancy prevention strategy and not a reduced risk of STI acquisition. Making adolescent females aware of this possibility is clearly the obligation of sex education programs. Many studies have shown that a significant number of adolescents are engaging in oral sex and do not understand the risks involved because they perceive that their friends and peers are engaging in the behavior and it is normative. Perhaps one strategy to change perceived norms may be incorporating messages in programs that oral sex is not viewed as risk free and that condom use with oral sex is warranted and a common practice.

Finally, it is worth noting that descriptively we observed that only 11% of the young women indicated they derived a lot of pleasure from performing oral sex. This suggests that pleasure derived from performing fellatio may not be common and that it may be distinct from the personality trait of sexual sensation seeking. Future research investigating associations among pleasure, the trait sexual sensation seeking, and the relational variables assessed in this study may be beneficial.

Limitations

These findings are limited by the use of a cross-sectional study design and the use of a convenience sample. Findings are also limited by the validity of the self-reported measures. Although we found one relational factor significant, we cannot rule out the possibility that other relational factors that tap into relationship dynamics and that were not measured in the present study could be associated with performing oral sex. The findings may not be applicable to other adolescent females, thereby suggesting that further research with other populations of female adolescents residing in communities with high rates of STI prevalence is warranted. Further, we cannot dismiss the possibility that young women's reports of having sex were biased as a function of sexual sensation seeking, thereby confounding the observed association between sensation seeking and oral sex.

Conclusions

Four factors were associated with performing oral sex: higher sexual sensation seeking, engaging in penile-vaginal sex without condoms, less self-efficacy to refuse sex, and perceived peer norms supportive of risky sexual practices. These factors are each amenable to change through education-based and skill-building intervention efforts. Given the potential for epidemic spread of orally acquired STIs to populations of high-risk adolescent females and the severity of the associated sequelae of oral STIs, this initial research provides important guidance for intervention development and provides impetus for expanded research efforts.

Acknowledgments

This study was supported by a grant from the Center for Mental Health Research on AIDS, National Institute of Mental Health (R01 MH061210) to R.A.C. and R.J.D. This work was supported in part by the Social and Behavioral Sciences Core of the Emory Center for AIDS Research (P30 AI050409).

Disclosure Statement

The authors have no conflicts of interest to report.