Practice Article Summary

A

K

d a s n p a m a y

e m

S 8

1 d

Original article

After the promise: the STD consequences of adolescent virginity pledges

Hannah Brückner, Ph.D.a, and Peter Bearman, Ph.D.b,* aDepartment of Sociology, Center for Research on Inequalities and the Life Course, Yale University, New Haven, Connecticut bDepartment of Sociology, Institute for Social and Economic Research and Policy, Columbia University, New York, New York

Manuscript received September 6, 2004; manuscript accepted January 14, 2005

bstract Purpose: To examine the effectiveness of virginity pledges in reducing STD infection rates among young adults (ages 18–24). Methods: Data are drawn from the National Longitudinal Study of Adolescent Health, a nationally representative study of students enrolled in grades 7–12 in 1995. During a follow-up survey in 2001–2002, respondents provided urine samples, which were tested for Human Papilloma Virus, Chlamydia, Gonorrhea, and Trichomoniasis. We report descriptive results for the relationship of pledge status and sexually transmitted disease (STD) rates as well as health behaviors commonly associated with STD infection. Results: Pledgers are consistently less likely to be exposed to risk factors across a wide range of indicators, but their STD infection rate does not differ from nonpledgers. Possible explanations are that pledgers are less likely than others to use condoms at sexual debut and to be tested and diagnosed with STDs. Conclusions: Adopting virginity pledges as intervention may not be the optimal approach to preventing STD acquisition among young adults. © 2005 Society for Adolescent Medicine. All rights reserved.

Journal of Adolescent Health 36 (2005) 271–278

eywords: Virginity pledge; Adolescent sexual behavior; STD acquisition; STD testing; STD risk factors

( p g t p p i t r t e t r

t c t

Understanding the determinants of sexually transmitted isease (STD) acquisition among adolescents and young dults is critical to assess interventions designed to limit the pread of STDs [1–3]. One set of interventions—adopted by umerous organizations and directly supported by federal olicy—are programs that encourage abstinence by encour- ging adolescents to make pledges to remain virgins until arriage. This article considers the relationship between

dolescent virginity pledges and the sexual behavior of oung adults, focusing on STD acquisition.

In 1993, “True Love Waits” initiated a movement to ncourage adolescents to pledge to abstain from sex until arriage. By 1995, an estimated 2.2 million adolescents

*Address correspondence to: Dr. Hannah Brückner, Department of ociology, Yale University, P.O. Box 208265, New Haven, CT 06520- 265.

tE-mail address: [email protected]

054-139X/05/$ – see front matter © 2005 Society for Adolescent Medicine. All oi:10.1016/j.jadohealth.2005.01.005

12% of all adolescents) in the United States had taken such ledges. Earlier research demonstrated that adolescent vir- inity pledges were associated with a significant delay on he baseline rate for the transition to first sex [4]. Although ledging was associated with delayed sexual initiation, the ledge effect was deeply shaped by social context, most mportantly, the number of other students in the community hat pledged and the social structure of the community with espect to the patterning of friendships. The pledge appears o work where public commitment to abstain from sex is ncoded into shared group activities, thus enhancing iden- ification with the movement and encoding the promise to emain a virgin into the larger social fabric.

If taking a pledge indeed reduces premarital sexual ac- ivity, one might expect that pledgers are less likely to ontract STDs than others because they initiate sexual ac- ivity later, have fewer sexual partners, and are more likely

o have sex in the context of a marital relationship than

rights reserved.

o m m s s y g t a h p t a c p s c S

D

t w l c i e l t

2 D B l i c a e a p r s d t s

t d t d s 0 o s

s y n e p t p p w p p i b w 2 1 b

b a d o f d o m b r u p

M

w p I a g p t p w c e ( p c p p t a A p

272 H. Brückner and P. Bearman / Journal of Adolescent Health 36 (2005) 271–278

thers, all protective factors. On the other hand, pledgers ay be more exposed to infection than others because they ay be less likely to use condoms [4]. We explore the

exual and health behaviors that may mediate the relation- hip between pledging history and later STD status. If oung people take a public virginity pledge to remain vir- ins until marriage, having sex before marriage means that hey break their pledge. Thus, sexually active pledgers have

greater incentive than nonpledgers to hide that they are aving sex. Especially critical are interactions with health rofessionals able to provide services to those who suspect hey may have an STD, pharmacists and others who provide ccess to condoms, contraceptive information, and STD ounseling, and friends and family members who could rovide relevant information about STDs, but may consider uch information unnecessary. Against this background, we onsider the relationship between pledging and self-reported TD-related health care utilization.

ata and methods

The initial results on the impact of virginity pledges on he transition to first sex arose from analyses of the first two aves of data from National Longitudinal Study of Ado-

escent Health (hereafter, Add Health). In this article, we onsider data from the 3rd in-home wave of Add Health nterviews, when respondents were 18–24 years old. This nables us to consider the long-term consequences of ado- escent pledging on the sexual behavior and STD acquisi- ion dynamics of young adults.

Of the original Add Health wave 1 respondents (n � 0,745), 15,170 individuals, or 73%, participated in wave 3. ata were collected between August 2001 and April 2002. iomarker data (urine samples) on STD status were col-

ected from 92% of wave 3 respondents. A total of 1183 ndividuals (8%) refused participation in the biospecimen ollection. Urine samples were collected in the field and nalyzed for the presence of three sexually transmitted dis- ases, Chlamydia (CH), Gonorrhea (GC), and Trichomoni- sis (TR). In addition, 7000 female respondents who re- orted ever having had vaginal sex in wave 3 were andomly selected for Human Papilloma Virus (HPV) as- ays. The specific procedure and tests used are described in etail in [5]. Both males and females were tested for the hree bacterial STDs regardless of their sexual activity tatus.

Collection procedures followed a strict protocol; samples hat arrived in a condition not suitable for testing were iscarded. Between 4% (CH), 5% (TR), and 10% (GC) of he samples were not tested for these reasons [5,6]. Pledgers o not differ from others with respect to whether their amples yielded results in the testing (p � 0.32 for CH, p � .17 for TR, p � 0.20 for CG). Pledgers did not differ from thers in the extent to which they refused to provide urine

amples (p � 0.28). Data are weighted to adjust for over- w

ampling of various groups and wave 3 nonresponse. Anal- ses of panel attrition for wave 3 show that generally, onresponse bias has little impact on estimates [7]. If pledg- rs were significantly more or less likely than others to articipate in wave 3, however, weighting may not correct he resulting bias. Among females, respondents who re- orted pledging in wave 1 or 2 were just as likely to articipate in wave 3 as nonpledgers. Among males, those ho pledged in wave 1 or wave 2 were more likely to articipate in wave 3 than others (23% nonresponse com- ared with 29%). However, both pledging and participation n wave 3 is negatively associated with age. The difference etween pledgers and others is significant only for males ho were 17 years and older at the time of wave 1 (31% vs. 2% nonresponse); among males under age 15 and between 5 and 16, nonresponse differs by 3% and 6%, respectively, ut not statistically significantly.

Unless otherwise noted, the analyses reported below are ased on 11,471 respondents with valid data on STD status nd grand sample weights. The majority of the results are erived from cross-tabulating pledge status with various utcome and behavioral measures. To adjust standard errors or the clustered sample design, we used the survey proce- ures provided in STATA [8,9]. For two measures (timing f first sex and first marriage) we used Kaplan-Meier esti- ates of the survivor function; these analyses use weights,

ut do not adjust for clustering. A Wald test based on a obust variance estimator appropriate for weighted data was sed to test for the difference between observed and ex- ected number of failures within each group.

easurement

Pledge status was collected across all three waves. In ave 1 and 2, respondents were asked “Have you taken a ublic or written pledge to remain a virgin until marriage?” n wave 3, respondents were asked “Have you ever signed

pledge to abstain from sex until marriage?” We distin- uish between individuals who reported that they took a ledge in one of the three waves and later said they did not ake a pledge (inconsistent pledgers) and those whose ledge reports are consistent across waves. Respondents ho reported pledging for the first time in the wave 3 were

lassified as consistent pledgers. The third group (nonpledg- rs) comprises respondents who never reported a pledge 80%, n � 9072). Thus, one in five respondents reported ledging in at least one wave, but only 7% (n � 777) gave onsistent answers. The remainder (13%, n � 1622) re- orted a pledge in one wave but said they had not taken a ledge in a later wave. It is unfortunate that the wording of he item was changed in wave 3; however, inconsistencies re frequent even when comparing only wave 1 and 2. bout half of those who reported a pledge in wave 1 and articipated in wave 2 said they had not taken a pledge in

ave 2. Response inconsistency in longitudinal studies is

r t i H i b c p

f ( n i t t l s i i m i s

S fi i f e b a f fi c f t a

b a q a p w t t c w b p i i p r p

o h y a r r t s i fi a c F 3 m i r d O c w

e w p w b

R

S

1 C t i r y t S T b r

t H f b 3 B w t g

273H. Brückner and P. Bearman / Journal of Adolescent Health 36 (2005) 271–278

elatively common. In general, consistency is greatest when he behavior under investigation is more salient. Consider- ng the complete set of “ever” questions for which Add ealth collected data, consistency of pledging appears to be

n the middle of the salience distribution, above smoking, ut below sexual orientation or school expulsion. Here, we onsider all three groups—consistent pledgers, inconsistent ledgers, and nonpledgers.

We examine STD status of the three groups using data rom analyses of urine samples for Human Papillomavirus HPV), Chlamydia (CH), Gonorrhea (GC), and Trichomo- iasis (TR) at the time of the wave 3. We hypothesize that nconsistent pledgers will be less exposed to infection risk han nonpledgers but more than consistent pledgers, and herefore have intermediate STD rates. For the less preva- ent STDs (GC, CH, TR) we aggregate test results into a ingle dichotomous indicator to increase sample size. This s justifiable because all three are transmitted through sexual ntercourse. HPV is analyzed separately because it is much ore prevalent, may be transmitted by noncoital sexual behav-

or, and the testing was based on a different sample (n � 3317; exually active females only).

exual and health behavior. The timing of the transition to rst sex was measured using self-reported age at first vag-

nal intercourse. The month and year of this event was taken rom wave 1 for those who reported being sexually experi- nced at that time; from wave 2 for those who transitioned etween waves 1 and 2. For those who transitioned there- fter, we used self-reported age at first vaginal intercourse rom wave 3. In wave 3, respondents were asked for age at rst intercourse rather than year and month of the event. In ontrast to earlier research, we therefore measure the dif- erences between pledgers and others only in years, rather han months. Marriage age was calculated from the month nd year of the first marriage reported in wave 3.

It was more difficult to determine the timing of pledging ecause neither wave 1 nor wave 2 included a question bout the timing. Although wave 3 contained a retrospective uestion about the timing of pledging, this question was sked only of respondents who reported having ever taken a ledge in wave 3. Thus, those who reported pledging in ave 1 or 2 but not in wave 3 were not asked when they had

aken a pledge. However, in many cases a time-order be- ween pledging and sexual debut could be determined by omparing reported pledge and virginity status across aves. We were thus able to determine the time-ordering etween first sex and pledge for 93% of pledgers. The ledge movement allows nonvirgins to participate in pledg- ng. We refer to adolescents who pledged after having had ntercourse as “secondary virgins.” Of those who reported ledging in any wave, 11% were secondary virgins, 21% eported no vaginal sex in any wave, and 61% had sex after ledging; 6% had missing data or a tie.

Number of sexual partners was measured as the number f

f partners reported in wave 3 with whom the respondents ad vaginal intercourse. We also calculated the number of ears a respondent was exposed as difference between age t first vaginal sex and age at interview. To measure STD isk associated with partners, we used a question that asked espondents for each partner “As far as you know, during he time you and �PARTNER� have had a sexual relation- hip, has �PARTNER� had any other sexual partners?” Sim- lar to our measure of transition to first sex, condom use at rst vaginal sex was measured from self-reported data cross all 3 waves of data collection. In wave 1 and 2, ondom use at first vaginal intercourse was asked directly. or those who reported vaginal sex for the first time in wave , we extracted this information from the relationship infor- ation, because condom use was reported for the first vag-

nal intercourse for each sexual relationship reported by the espondent. Where available, we used the start and ending ates of the sexual relationship to identify the first one. therwise, we used a question that asked the respondent to

hronologically order all relationships to determine which as the first. Add Health asked respondents also about their experi-

nces with STDs. Specifically, all respondents were asked hether they had been diagnosed with various STDs in the ast year; whether they had ever seen a doctor because they ere worried about having a STD; and whether they had een tested for various STDs in the past year.

esults

TD acquisition

Bio-marker rates were as follows: TR (2.3%, 95% CI .8–2.8%), CH (4.2%, 95% CI 3.6–4.9%), GC (0.4%, 95% I 0.3–0.6%), HPV (28.8%, 95% CI 26.3–31.4%). Al-

hough these rates may be somewhat lower than those found n the literature, it bears noting that they are derived from a epresentative random sample of the population aged 18–24 ears. Other studies use clinical samples or special popula- ions and may therefore overestimate STD prevalence [10]. TD acquisition varies significantly by race and ethnicity. able 1 reports current infection with TR, GC, and/or CH y race, ethnicity and gender (Panel A). Results for HPV are eported separately (Panel B).

Black males and black females have rates roughly 8 imes that for white males and females, respectively. Asian, ispanic, and others have rates 2 to 5 times that of Whites

or TR, GC, and/or CH. The prevalence for HPV varies etween 16% for Asian females, 25% for white females, 0% for Hispanic females, and 34% for black females. ecause both race/ethnicity and pledging are associated ith STD infection, we consider the effect of the pledge on

he likelihood of having an infection within race/ethnic roups. Owing to small sample sizes, we combine the data

or Hispanics, Asians, and respondents of other race/ethnic-

i s a F w a w f ( r A d s f

g e p b s f i w m b c w r

T S

P

P

T S

P

P

a

274 H. Brückner and P. Bearman / Journal of Adolescent Health 36 (2005) 271–278

ty, and report HPV separately only for Whites. Table 2 hows that there are no significant differences in STD rates cross any of the pledge groups compared with nonpledgers. or most groups, the point estimates are fairly similar as ell. With respect to the summary indicator for TR, GC,

nd CH, the relatively largest differences are found for hite respondents, with point estimates that are 31% lower

or inconsistent pledgers and 6 % for consistent pledgers Panel A). The rates are so low, however, that we cannot eject the null hypothesis of zero difference in the rates. lthough the rates are higher for the other groups, the ifferences are too small to be significant, statistically or ubstantively. Point estimates for HPV are slightly higher or pledgers (Panel B), albeit not significantly so.

able 1 TD prevalence rates by race/ethnicity and gendera

Female

Percent 95%

anel A: GC, TR, and/or CH (%)b

White 3.8 3.1 Black 24.0 20.4 Hispanic 6.4 4.8 Asian 5.9 3.6 Other 15.8 7.8 Totalc 7.5 6.5

anel B: HPV (Female only, %)d

White 25.5 23.0 Black 34.1 29.2 Hispanic 29.5 23.1 Asian 16.3 9.4 Other 23.7 12.1 Total 27.2 26.3

Standard errors adjusted for clustering. a STD status measured at wave 3; percentages weighted using the long b Includes respondents who never reported vaginal sex. c Includes 6 respondents with missing race/ethnicity. d Includes only females who reported vaginal sex in wave 3.

able 2 TD status by pledge status and race/ethnicitya

Nonpledgers (n � 9072)

Inc (n

Percent 95% CI Per

anel A: TR, CH, and/or GCb

White 3.5 2.8 4.3 2. Black 20.3 17.4 23.6 19. Other 8.1 6.7 9.9 6. All 6.9 6.0 8.0 6.

anel B: HPVd

White 25.1 22.3 28.1 25. All 26.5 24.1 29.0 28.

a Pledge status measured across waves 1–3; STD status measured at wave djusted for clustering.

b Includes respondents who never reported vaginal sex. c Includes 6 respondents with missing race/ethnicity.

d Includes only females who reported vaginal sex in wave 3.

If nonresponse were significantly different for the pledge roups, these findings might be owing to selectivity. For xample, nonpledgers at risk for STD might be less likely to articipate in wave 3, which would lead to a downwardly iased estimate for the group. As explained above, no re- ponse bias associated with pledge status was found for emales. There was differential response for males, specif- cally older males, however. By comparing STD rates ithin age groups by pledge status in waves 1 and 2, we can ake an informed guess about the significance of response

ias. Because of the small cell sizes, here we combine onsistent and inconsistent pledgers. If differential response as solely responsible for the findings, we should see lower

ates for pledgers among those who were under age 15 at the

Male

n Percent 95% CI n

3316 2.9 2.2 3.7 2926 1340 16.0 12.9 19.9 1041 923 9.7 7.1 13.2 901 381 5.4 2.6 10.7 427 105 7.0 2.7 16.6 105

6069 5.8 4.9 6.9 5402

1806 809 519 180 53

3367

weight for wave 3.

t pledgers )

Consistent pledgers (n � 777)

p (n)

95% CI Percent 95% CI

1.4 4.1 3.3 1.6 4.8 .355 (6242) 13.5 27.4 18.1 9.6 31.5 .915 (2381) 3.8 11.5 8.0 3.4 17.5 .785 (2842) 4.8 8.5 4.6 3.1 6.8 .150 (11,471)c

18.7 33.2 28.8 19.7 25.4 .770 (1788) 23.2 34.4 26.7 19.4 35.6 .808 (3317)

entages weighted using the longitudinal weight for wave 3. Standard errors

CI

4.6 28.1 8.5 9.5

29.5 8.7

28.2 39.2 36.8 23.7 41.3 31.4

itudinal

onsisten � 1622

cent

4 6 7 4

3 5

3; perc

t o c d n p t c c s n w p c n o o m t

S

s T e n t t t 7 y 7 t f T a y c o m p p o

l c m o 4 T s c c n

n w l 1 p d

e p I 8 c p ( c s p b

T

i o p

T T

P

P

a s

275H. Brückner and P. Bearman / Journal of Adolescent Health 36 (2005) 271–278

ime of wave 1 because in this group, both pledgers and thers were equally likely to participate in wave 3. In ontrast, among older males, we should see smaller or no ifferences, because the hypothesis is that in these groups, onpledgers who are at risk for STD are less likely to articipate. Differences are small and not significant in all hree age groups, but they are larger among older adoles- ents (3.4% infected pledgers for the TR/CH/GC indicator, ompared with 5.9 % for nonpledgers). The differences are maller in the youngest group (4.3% vs. 5.1%) and the ext-youngest group (5.6% vs. 7.0%). This pattern is not hat we would expect if the lack of a difference between ledgers and others was solely owing to nonresponse asso- iated with pledge status. There are several possible expla- ations for this finding. Differences between pledgers and thers in sexual risk-taking might be smaller than expected, r other risk factors, such as differences in condom use, ight be more important. We explore these possibilities in

he following.

exual debut and marriage

Table 3 summarizes Kaplan-Meier estimates for the tran- ition to first vaginal sex and first marriage. Cell entries in able 3 denote the age at which the percentile indicated in ach row experienced the event in question. Thus, 25% of onpledgers had experienced sexual debut by the time they urned 15, 50% when they turned 17, and 75% when they urned 18 (Panel A). Consistent pledgers, in contrast, reach he 25th percentile by age 17, the median by age 19, and the 5th percentile at age 24. The inter-quartile difference is 3 ears for nonpledgers, 4 years for inconsistent pledgers, and years for consistent pledgers. The Wald test for equality of

he survivor functions for the three groups yields a �2 of 189 or females and 140 for males (p � .000 for both tests). hus, pledgers experience first sex later than others across dolescence, and a significant minority holds out far into oung adulthood. By age 25, we estimate that 25% of onsistent male pledgers are still virgins, compared with 7% f nonpledgers and 15% of inconsistent pledgers. For fe- ales, the corresponding numbers are 21% for consistent

ledgers, 6% for nonpledgers, and 10% for inconsistent ledgers. Note that these results are for vaginal intercourse nly.

Table 3, panel B shows that female pledgers marry ear- ier than nonpledgers. The summary measures in Table 3 onceal, however, substantial differences in the timing of arriage. We estimate that by age 25 more than half (52%)

f female consistent pledgers are married, compared with 0% for inconsistent pledgers and 34% for nonpledgers. he differences in the survivor functions for females are ignificant with a �2 of 14.6 (p � .001). For males, the orresponding numbers are 45% for consistent pledgers, ompared with 26% for inconsistent pledgers and 25% for

onpledgers. The differences in the survivor functions are p

ot statistically significant for males. Respondents who ere married at the time of wave 3 were significantly less

ikely to test positive for TR/CH/GC (4.1% vs. 7.2%, n � 1,741, p � .0003) and HPV (16.3% vs. 29.6%, n � 3317, � .0000). Among married respondents, STD rates did not

iffer by pledge status (data available on request). Although pledgers tend to have sex later than nonpledg-

rs and tend to get married earlier than nonpledgers, most ledgers do not wait to get married before having first sex. f we consider just those respondents who have had sex, 8% of the pledgers have sex (here again, vaginal inter- ourse) before they get married, whereas 99% of the non- ledgers have sex before marriage. Inconsistent pledgers 94%) are in between, as expected. Of all respondents, not onditioning on having had sex, the pattern is roughly imilar. Specifically, 61% of all pledgers, 90% of all non- ledgers, and 79% of all inconsistent pledgers have sex efore marriage or interview date.

ime-order between sexual debut and pledging

One potentially confounding issue with respect to the nterpretation of the pledge effect is the problem of “sec- ndary virgins.” The pledge movement makes room for eople who have had sex to subsequently take a virginity

able 3 ransition to first sex and marriage, by pledge statusa

Nonpledgers Inconsistent pledgers

Consistent pledgers

anel A: Transition to First Sex

All 25 %tile 15 16 17 Median 17 18 19 75 %tile 18 20 24

Female 25 %tile 15 16 16 Median 17 18 18 75 %tile 18 19 23

Male 25 %tile 15 16 17 Median 17 18 20 75 %tile 18 21 25

anel B: Age at First Marriage

All 25 %tile 24 23 23 Median - - -

Female 25 %tile 23 22 22 Median - - 24

Male 25 %tile 24 24 24 Median - - -

a Kaplan-Meier estimates for survivor function; cells entries denote age t which the percentile indicated in each row experienced first vaginal ex/first marriage. Data are weighted with longitudinal weight for wave 3.

ledge. Secondary virgins may shape the differences in

t i m o a

i T s a s t a p t v I c n p d l s

C

f p a p i i

i m

o o l H v a o b f i n 1 n b s b t c p r u i p l v p a s m

O

t n

T H

S S T T T T E E U

276 H. Brückner and P. Bearman / Journal of Adolescent Health 36 (2005) 271–278

ransition dynamics for those who are consistent versus nconsistent pledgers. This is not the case. Although slightly ore consistent pledgers (14% rather than 10%) are “sec-

ndary virgins,” this difference is not substantial enough to ffect the results reported above.

It is also possible that “secondary virgins” are confound- ng the substantive interpretation of the results reported in able 2. For example, some respondents could have had ex, thought they might have contracted an STD, felt bad bout that—or felt bad about having sex in general—and ubsequently pledged to remain a virgin until marriage. If his were the case, the fact that pledgers are as likely to have n STD as nonpledgers could be an artifact of the timing of ledging and sex. This does not seem to be the case—of hose who had sex before they pledged (the secondary irgins) the prevalence rate for TR, GC, and/or CH is 5.5%. n contrast, 7.3% of those who pledged before sex have a urrent STD infection, in contrast to 6.9% of those who ever pledged. If the secondary virgins are playing a role in ledger STD infection rates, it is likely in the opposite irection. Those who have sex and then pledge are less ikely to have an STD than those who pledge and then have ex.

ondom use, oral and anal sex

Condom use at first intercourse is a powerful predictor or subsequent consistent use [11]. As found earlier [4], ledgers are significantly less likely to have used a condom t first intercourse than nonpledgers (Table 4). However, ledgers are not less likely to use condoms at most recent ntercourse reported in wave 3, or in the 12 months preced- ng the wave 3 interviews.

Because virginity is often culturally linked only to vag- nal sex, to preserve virginity, adolescents and young adults

able 4 ealth behavior and virginity pledgea

Nonpledgers I

% 95% CI %

elf-reported TR/CH/GC 3.6 3.0 4.2 elf-reported HPVb 2.7 2.1 3.4 ested (TR, CH, GC)c 30.3 28.2 32.5 2 ested (TR, CH, GC)d 10.0 8.8 11.3 ested (HPV)b 14.9 13.6 16.3 1 ested (HPV)c 16.6 15.1 18.2 1 ver seen doctor for STD 22.9 21.2 24.7 1 ver seen doctor for STDe 20.8 19.3 22.3 1 sed condom at first sex 59.7 58.0 61.4 5

Standard errors adjusted for clustering. a STD status measured at wave 3; Percentages weighted using the long b Females. c Females who report any sexual activity in wave 3. d Males who report any sexual activity in wave 3. e Respondents who report any sexual activity in wave 3.

ay engage in other sexual behaviors that involve exchange n

f fluid and are thus salient for STD acquisition. Overall, ral sex and anal sex are prevalent behaviors in this popu- ation, most commonly in conjunction with vaginal sex. ere we consider those who have oral or anal sex without aginal sex. Amongst those who have only oral sex and/or nal sex, pledgers are over-represented. Overall, about 3% f respondents reported oral sex with one or more partners ut no vaginal sex. Although just over 2% of nonpledgers all into this group, 13% of consistent pledgers and 5% of nconsistent pledgers do (p � .000). Similarly, 0.7% of onpledgers report anal but no vaginal sex, compared with .2% for pledgers. Although too few females report anal but o vaginal sex, for males we find a significant difference etween pledgers and nonpledgers (p � .021). Specifically, lightly more than 1% of male nonpledgers report anal sex ut no vaginal sex, compared with almost 3% for inconsis- ent pledgers and 4% for consistent pledgers. For oral sex, ondom use is almost completely absent—respondents re- orted condom use for first oral sex for only 4% of the elationships that involved oral sex. For anal sex, condom se is also lower than for vaginal sex. Condoms were used n about 30% of relationships involving anal sex when artners had anal sex for the first time. The combination of ow condom use and over-representation of pledgers pro- ides some support for the hypothesis that this behavioral attern is associated with greater than expected STD cquisition among pledgers, although the numbers are mall and provide an insufficient basis from which to ake inference.

ther STD risk factors

Here we consider additional risk factors that are known o be associated with acquisition of an STD. The first is umber of sex partners. Pledgers have fewer partners than

tent pledgers Consistent pledgers p �

95% CI % 95% CI

2.2 4.3 2.8 1.7 4.5 .475 0.8 2.6 1.1 0.4 2.6 .018

19.9 28.2 18.5 13.9 24.1 .000 4.7 11.6 5.6 2.6 11.7 .145 7.9 13.2 7.8 5.8 10.3 .000 8.3 14.1 10.9 8.1 14.4 .002

11.9 17.8 14.1 11.0 17.8 .000 12.9 17.6 15.9 12.4 20.2 .000 51.2 58.2 54.6 48.5 60.0 .017

weight for wave 3.

nconsis

3.1 1.4 3.8 7.5 0.3 0.9 4.6 5.1 4.9

itudinal

onpledgers. Whereas the typical nonpledger male has had 2.4

p . n p m f s t a e . c t

t r t t t n m t m b t t

H

r t t d n w a t

D

e p l l t b r s e s t e l p

c l

i a p e c i y a a a

c w f t h f p o o a p m n p p c a e c r t S w u s

o [ e f f s c h

A

p K

277H. Brückner and P. Bearman / Journal of Adolescent Health 36 (2005) 271–278

artners, male pledgers have 1.5 partners on average (p � 000). The same pattern holds for females as well, 2.7 for onpledgers and 1.9 for pledgers (p � .000). Inconsistent ledgers are between these ranges (2.0 for males and fe- ales, respectively). Nor are pledgers exposed to STD risk

or as long as nonpledgers. The average number of years of exual activity, or time of exposure, is shorter for pledgers han for others. Consistent pledgers were sexually active for n average of 4.2 years, compared with inconsistent pledg- rs with 4.4 years and nonpledgers with 5.9 years (p � 000). Thus, with respect to both the number of partners and umulative exposure, pledgers are at lower risk to acquisi- ion of an STD than nonpledgers.

It is possible that although pledgers have fewer partners han nonpledgers, their partners might have more concur- ent partners, thus putting them at enhanced risk. We test his idea indirectly, from self-report data. STD rates are in he expected direction—specifically STD rates are 7% for hose with none, and 10% for those with two or more onmonogamous partners (p � .002). But pledgers are uch less likely (6% vs. 12%) than nonpledgers to report

hat their partners are nonmonogamous. Although pledgers ight be less aware of their partners’ nonmonogamous

ehavior than others, from these data there is no support for he idea that the higher-than-expected pledge STD rate is he result of high-risk partners.

ealth behavior

Although those who take virginity pledges have STD ates comparable to those who do not, they are less aware of heir STD status. Table 4 shows that in the year preceding he interview, pledgers were less frequently tested for and/or iagnosed with an STD than nonpledgers. Most notably, onpledging females are almost twice as likely as females ho pledged to be tested, p � .000). Furthermore, pledgers

re significantly less likely to report seeing a doctor because hey are worried about an STD (p � .000).

iscussion

Contrary to expectations, we found no significant differ- nces in STD infection rates between pledgers and non- ledgers, despite the fact that they transition to first sex ater, have less cumulative exposure, fewer partners, and ower levels of nonmonogamous partners. Examination of he point estimates revealed small or nonexistent differences etween pledgers and others, with the exception of white espondents. Advocates for abstinence-only education as- ert that premarital abstinence and postmarital sex are nec- ssary and sufficient for avoiding negative consequences of exual activity, such as STDs. This assertion collides with he realities of adolescents’ and young adults’ lives in sev- ral ways. First, although pledgers experience sexual debut ater than others, most of them will eventually engage in

remarital sex. Those who do report lower frequency of H

ondom use at first intercourse. Those who do not are more ikely to substitute oral and/or anal sex for vaginal sex.

Second, although marriage is protective against STDs, it s not perfect. Although female pledgers marry earlier, and n estimated 12% did not report any premarital sex, married ledgers test positive at the same rates as married nonpledg- rs. The biomarker data we have analyzed in this article annot tell us whether pledgers had a lower risk of STD nfection as young adolescents. By the end of their teenage ears, at the time of wave 3, however, these advantages, if ny, have vanished. As a social policy, pledging does not ppear effective in stemming STD acquisition among young dults.

Furthermore, because most adolescents eventually be- ome sexually active during their teenage years, is it really ise to ban discussion of contraception and STD protection

rom sex education? Although virtually all adolescents say hey learned about STDs in school [12,13], other studies ave shown that many adolescents underestimate their in- ection risk and that they have mistaken ideas about what rotects them from STDs and what does not [14,15]. The rganizations that promote pledging and other abstinence- nly programs have been hostile to programs that combine bstinence education with information about how to prevent regnancy and STDs for sexually active adolescents. The aterials distributed by “True Love Waits” and other orga-

izations teach adolescents that the only protection from regnancy and STDs is abstinence. The all-or-nothing ap- roach advocated by many abstinence-only programs may reate additional barriers to knowledge and protection for dolescents. For example, the emphasis on virginity may ncourage adolescents to limit their sexual activity to non- oital behaviors, which may nevertheless expose them to isks of infection. In this context, it is important to know hat pledgers are less likely than nonpledgers to be tested for TDs, and to have ever seen a doctor because they are orried about an STD. If STDs were more likely to go ntreated among pledgers, higher STD prevalence may re- ult even in the presence of lower incidence rates.

Systematic and rigorous evaluation of the health impact f abstinence-only programs has rarely been undertaken 16]. The results presented in this article show that a careful valuation should accompany the generous federal and state unds that abstinence-only programs have enjoyed. At least or one such program, pledging, our findings suggest that a hort-term evaluation of the behavioral impact of sex edu- ation programs is not sufficient to predict the longer-term ealth impact on STD rates.

cknowledgments

This research uses data from Add Health, a program roject designed by J. Richard Udry, Peter S. Bearman, and athleen Mullan Harris, and funded by a grant P01-

D31921 from the National Institute of Child Health and

H o R d H C 2 h o

R

[

[

[

[

[

[

[

278 H. Brückner and P. Bearman / Journal of Adolescent Health 36 (2005) 271–278

uman Development, with cooperative funding from 17 ther agencies. Special acknowledgment is due Ronald R. indfuss and Barbara Entwisle for assistance in the original esign. Persons interested in obtaining data files from Add ealth should contact Add Health, Carolina Population enter, 123 W. Franklin Street, Chapel Hill, NC 27516- 524. The authors thank two anonymous reviewers for their elpful comments. Authorship is in reverse alphabetical rder. Both authors contributed equally to this article.

eferences

[1] Centers for Disease Control and Prevention. Tracking the hidden epi- demic: trends in STDs in the United States [cited 2004 Jul 14]. Available from: www.cdc.gov/nchstp/od/news/RevBrochure1.pdf.htm.

[2] Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance, 2001. Altlanta, GA: Department of Health and Human Services, CDC, 2002.

[3] Weinstock H, Berman S, Cates W Jr. Sexually transmitted diseases among American youth: incidence and prevalence estimates, 2000. Perspect Sex Reprod Health 2004;36:6–10.

[4] Bearman P, Brückner H. Promising the future: abstinence pledges and the transition to first intercourse. Am J Sociol 2001;106:859 – 912.

[5] Add Health Biomarker Team. Biomarkers in Wave III of the Add Health Study [cited 2004 Jul 14]. Available from: http://www.cpc.

unc.edu/addhealth/files/biomark.pdf.

[6] Carolina Population Center. Wave III data documentation [cited 2004 Jul 14]. Available from: http://www.cpc.unc.edu/projects/addhealth/ codebooks/wave3.

[7] Chantala K, Kalsbeek WD, Andraca E. Non-response in Wave III of the Add Health study [2004 Jun 28]. Available from: http://www.cpc. unc.edu/projects/addhealth/files/W3nonres.pdf.

[8] Chantala, K, Tabo J. Strategies to perform design-based analysis using the Add Health Data [cited 2004 Jul 14]. Available from: www.cpc.unc.edu/projects/addhealth/files/weight1.pdf.

[9] Stata Corp. Stata Statistical Software: Release 8.0. College Station, TX: Stata Corporation, 2003.

10] Miller WC, Ford CA, Hobbs MM, et al. Prevalence of chlamydial and gonococcal infections among young adults. JAMA 2004;292:801–2.

11] Shafii T, Stovel K, Davis R, Holmes KK. Is condom use habit- forming? Condom use at sexual debut and subsequent condom use. Sex Transm Dis 2004;31(6):366–72.

12] Lane T. Despite having received relevant education, youth lack knowledge of STDs. Perspect Sex Reprod Health 2003;35:51.

13] Brückner H, Bearman P. Dating behavior and sexual activity among young adolescents. In: Albert B, Brown S, Flanigan CM, eds. 14 and Younger: The Sexual Behavior of Young Adolescents. Washington, DC: Campaign to Prevent Teen Pregnancy 2003:31–56.

14] Kaiser Family Foundation. Seventeen. A series of national surveys of teens about sex: birth control and protection [cited 2004 Dec 6]. Available from: www.kff.org/entpartnerships/phip070604pkg.cfm.

15] Kaiser Family Foundation. Seventeen. A series of national surveys of teens about sex: sexually transmitted diseases [cited 2004 Dec 6]. Avail- able from: www.kff.org/entpartnerships/seventeen_surveys.cfm.

16] Kirby D. Emerging Answers: Research Findings on Programs to Reduce Teen Pregnancy. Washington, DC: National Campaign to

Prevent Teen Pregnancy, 2001.