Click a topic below for an index of articles:

New-Material

A new addition to the HEART is our Forum-check it out

Cognitive and behavioral predictors of sexually transmitted disease risk behavior among sexually active adolescents
By Renee Sieving; Michael D. Resnick; Linda Bearinger; Gary Remafedi; Barbara A. Taylor; Brian Harmon.

http://www.yapmn.com/articles/C.html

ABSTRACT:
Risky sexual behaviors during adolescence may not change once patterns are set. A total of 410 sexually active youths aged 14 to 21 reported on health beliefs and sexual behaviors at baseline and one year later. Responses to the first survey and knowing how to use condoms predicted responses to the second survey in both sexes. Among girls, oral contraceptive use, ability to discuss sexually transmitted disease prevention with partners, and alcohol use explained variability in risky behaviors. Among boys, perceived risk of sexually transmitted disease, negative opinion of condom use, and barriers to obtaining condoms explained variability in risky behaviors.

Copyright © 1997 American Medical Association

Arch Pediatr Adolesc Med. 1997;151:243-251

FULL TEXT:

In a 1980 report, the National Institute of Allergy and infectious Diseases described sexually transmitted diseases (StDs) as "the most pervasive, destructive, and costly communicable disease problem confronting adolescents today in the United States."[1] Since the issue of this report, the prevalence of sexual intercourse has increased among adolescents as have rates of STDs, including human immunodeficiency virus (HIV) infection.[2] In the United States, it is estimated that 3 million adolescents -- 1 of every 6 -- are infected with an STD every year.' Female adolescents aged 15 to 19 years have the highest rates of gonorrhea of any age group in the United States.[4] Although national data on age-specific chlamydia rates have not been published, unpublished reports indicate that approximately 90% of all chlamydia cases occur among persons younger than 25 years.[4] While adolescents represent less than 1% of all cases of acquired immunodeficiency syndrome in the United States, 18% of all male cases and 23% of all female cases are found among 20- to 29-year-olds.[5] With a mean latency period of 8 to 10 years between HIV infection and the onset of symptoms,[4] many young adults with the acquired immunodeficiency syndrome probably became infected during adolescence.

Among sexually active adolescents, certain known factors increase the potential for exposure to infectious agents and the risk for STD. Among the most proximal of these factors are sexual behaviors including inconsistent use of barrier contraception, having sexual intercourse with multiple partners, and having intercourse on a relatively frequent basis.[6,7] These risk behaviors may be somewhat common among sexually active adolescents. For example, among sexually active secondary school students in the United States, 47.2% report that they or their partner did not use a condom during last sexual intercourse. Female students (46.0%) are less likely than male students (59.2%) to report condom use during last intercourse. Condom use with most recent intercourse is less common among older than among younger students.[8] in a study of 14- to 19-year-olds attending teen health clinics, 40.3% of female and 69.4% of male teenagers reported having more than 1 sexual partner during the previous year.[9] Among 9th-to 12th-graders nationwide, 22% of males and 15% of females report having,@ or more sexual partners during their lifetime.[8] While less is known about patterns of intercourse frequency among adolescents, some sexually experienced adolescents report being abstinent for varying lengths of time. Although 44.9% of 15- to 19-year-old white females involved in a recent study reported ever having had sexual intercourse, only 28.0% had had sex within the past month, and only 14.9% had been sexually active on a consistent basis during the past 12 months.[10]

If sexual risk behaviors represent a common pathway to STD acquisition, it is essential to understand the antecedents of these behaviors as a basis for STD prevention efforts. While sociocultural factors such as gender, age, ethnicity, cultural norms, and socioeconomic status may directly or indirectly influence sexual behavior, these factors either cannot be changed or can be changed only through massive social transformation.[7] Another set of factors, psychosocial or behavioral in nature, may be more amenable to change through school-and community-based interventions.

Several empirical studies have examined psychosocial and behavioral correlates of one type of sexual behavior that increases the risk for STD among adolescents, namely, inconsistency of condom use. Among the psychosocial factors that have been associated with inconsistent condom use among adolescents are low levels of perceived susceptibility to STD and HIV infection,(11,12) negative expectations associated with condom use,[11,13,14] and low levels of self-efficacy or confidence in performing the skills necessary to use condoms on a consistent basis.[15,16] Behaviors that have been associated with inconsistent condom use include inadequate communication with sexual partners about sexual issues,[13,14,17] involvement in relatively long-term relationships,[18] having sexual intercourse on a frequent basis,[19] and use of alcohol and/or other substances in connection with sexual activity.[6,20]

Although numerous studies have examined correlates of condom use among young people, predictors of the more complex cluster of sexual risk behaviors have not been examined extensively. Many of the existing empirical studies on the factors that may have an influence on condom use among teenagers have been cross-sectional, limiting longitudinal inferences. Small sample sizes have often limited analyses to relationships between 1 or 2 independent variables and a designated behavioral outcome.

The purpose of these analyses was to identify cognitive and behavioral factors that predict adolescents' level of STD risk behavior and that are amenable to change through clinic-, school-, and community-based interventions. The outcome measure, STD risk behavior, represents a constellation of behaviors that put adolescents at risk for acquisition of STDs. Data from a 1-year longitudinal study of more than 500 adolescents allow examination of predictive relationships. We addressed the following specific research questions: (1) What is the relationship between adolescents, baseline knowledge, attitudes, and behaviors -- including STD-related knowledge, condom use self-efficacy, outcome expectations, partner communication practices, substance use behavior, use of oral contraceptive pills (OCP), and STD risk behavior -- to STD risk behavior 1 year later? (2) What is the relative strength of these cognitive and behavioral factors, both alone and in combination, in explaining STD risk behavior among sexually active teens? and (3) How much variability in STD risk behavior can be explained by antecedent cognitions and behavioral patterns?

RESULTS

STD RISK BEHAVIORS

Figure 3 provides a descriptive picture of T2 sexual risk behavior among members of the study cohort. Among girls, 46.6% reported having vaginal intercourse with 1 partner during the 1-year interval between T1 and T2 surveys; 14.9% reported having 4 or more partners during this same interval. Nearly 22.0% of girls reported using condoms more than half the time with their most recent partner or partners and 36.4% reported never using condoms with their most recent partner(s). Approximately 45% of girls reported having sexual intercourse more than 20 times with their most recent partner, or, among those who had more than 1 partner during the follow-up interval, more than 20 times with each of 2 most recent partners.

Approximately 50% of boys reported having intercourse with 1 partner during the 1 year between surveys; 14.7% reported having 4 or more partners during that interval. About 41% of boys reported using condoms more than half the time with their most recent sexual partner(s); 30.1% reported never using condoms with their most recent partner(s). About 42.5% of boys reported having sexual intercourse more than 20 times with their most recent partner, or more than 20 times with each of 2 most recent partners.

Examining changes in STD risk behavior over the 1-year study interval, T1 risk scores ranged from 1 to 45, with a mean score of 15.69 for girls. At T2, the mean STD risk score among girls was 15.12, representing a nonsignificant change in level of STD risk behavior. For boys, T1 scores ranged from 1 to 44, with a mean score of 14.76. As with girls, the mean level of STD risk behavior among boys at T2 (mean score, 13.56) was not significantly different from the T1 mean.

BIVARIATE ANALYSES

Bivariate analyses (Table 1) found that girls with the lowest T2 STD risk behavior reported the lowest levels of STD risk behavior at T1 (r=0.390, P=.00). Girls with lowest STD risk behavior at T2 also reported highest levels of condom use self-efficacy at T1 (r=-0.225, P=.003). Compared with other groups at T1, these girls were more likely to refuse intercourse if a partner did not want to use condoms (r=0.143, P=.01); they were also more likely to carry condoms when anticipating intercourse (r=0.174, P=.002). in terms of behaviors with their most recent sexual partners at T1, this group reported having more appropriately timed communication related to STD risk and prevention (r=0.224, P=.00), and dating their partners for longer periods before having vaginal intercourse (r=-0.251, P=.00). These girls were also less likely to be using alcohol in connection with sexual activity at T1; 13.0% of those in the low-risk group vs 30.7% of girls in the high-risk group reported alcohol use in connection with sexual activity with most recent partners (P=.002).

Table 1. Baseline (T1) Correlates of Follow-up Sexually Transmitted Disease Risk Behavior(*)

Girls (n=335)                              

T1

Knowledge, attitudes, and beliefs

 Barriers to STD checkups

 (11-item scale)[dagger]                          

NS

Would not carry condoms if 

anticipating sex [double dagger]      

   0.174 (.002)

Would not refuse sex with partner 

not using condom [double

 dagger] 

0.143 (.01)

Reasons for past nonuse 

condoms[double dagger]

Negative outcome expectations 

NS

Low perceived susceptibility to 

STD and pregnancy 

NS

Insufficient access 

     NS

Condom use self-efficacy (7-item

 scale)[dagger] 

-0.225 ([is less than].001)

T1 behaviors

STD risk behavior (45-point scale)

[dagger]

0.390 ([is less than].001)

Alcohol use in connection with sexual

 activity [double dagger]     

0.160 (.004)

Talked with most recent partner(s) 

about STD risks before 

having intercourse [double dagger]           

0.224 ([is less than].001)

Length of relationship with partner 

before having intercourse 

[double dagger] (7 categories) 

-0.251 ([is less than].001)

Boys

                                           (n=75)

T1

Knowledge, attitudes, and beliefs Barriers to STD 

checkups (11-item scale)[dagger] 

0.290 (.02)

Would not carry condoms if anticipating sex [double dagger] 

NS

Would not refuse sex with partner not using condom 

[double dagger] 

0.317 (.05)

Reasons for past nonuse of 

condoms [double dagger]

Negative outcome expectations 

0.290 (.01)

Low perceived susceptibility to STD and pregnancy 

0.391 (.001)

Insufficient access Condom use self-efficacy (7-item 

scale)[dagger] 

     -0.388 (.003)

T1

Behaviors STD risk behavior (45-point scale)[dagger]         

0.266 (.04)

Alcohol use in connection with sexual activity[double dagger]        

NS

Talked with most recent partner(s) about STD risks before 

having intercourse [double dagger] 

NS

Length of relationship with partner before having intercourse 

[double dagger] (7 categories) 

NS

(*) NS measures are those that failed to reach significance at the P=.005 level. STD indicates sexually transmitted disease. [dagger] Zero-order Pearson correlation coefficient; associated P values are in parentheses. [double dagger] Spearman rank-order correlation coefficient; associated P values are in parentheses.

Similar to girls, boys engaging in the lowest levels of T2 STD risk behavior at T2 reported low levels of STD risk behavior at T1 (r=0.266, P=.04). Among boys, lower STD risk behavior at T2 also was associated with higher condom use self-efficacy at T1 (r=-0.388, P=.02). In terms of STD-related health beliefs, boys reporting low levels of STD risk behavior at T2 were more likely to refuse to have sex if a partner was not willing to use condoms (r=0.317, P=.007), had fewer negative outcome expectations related to condom use (r=0.235, P=.05), and had fewer perceived barriers to obtaining condoms (r=0.391, P=.001) and to having a STD checkup (r=0.290, P=.02) at T1. They had higher levels of perceived susceptibility to STD and pregnancy at T1 (r=-0.290, P=.01). For boys, no T1 behavior other than STD risk behavior was predictive of T2 STD risk behavior.

Age was not independently related to level of STD risk behavior among boys or girls. Ethnicity was significantly related to level of T2 STD risk behavior among boys. About 42.0% of white boys, 28.6% of African-American boys, and 12.5% of boys from other ethnic groups reported the lowest levels of STD risk behavior at T2. However, after controlling for age differences between ethnic groups, only boys from ethnic backgrounds other than African American were significantly more likely than white boys to be engaged in high levels of STD risk behavior (relative risk= 5.75; [x.sup.2]=4.32, df=1, P=.004)0.

MULTIVARIATE ANALYSIS

All T1 bivariate measures approaching significance at a P=.10 level were included in a multivariate regression analysis for girls. A final multivariate model (Table 2) containing T1 STD risk behavior, condom use self-efficacy, use of OCP, communication with sexual partners about STD prevention, and use of alcohol in connection with sexual activity with recent partner(s) explained 21.1% of the variance in girls, T2 STD risk behavior.

Table 2. Multivariate Predictors of Sexually Transmitted Disease Risk Behavior at Follow-up Among Girls

OCP use as reason for not using 

condoms with 1 most recent partner 

Did not talk with 1 of most recent partners 

about STD risk before having sexual intercourse 

OCP use as reason for not using condoms 

with both most recent partners 

Used alcohol over half the time in connection 

with sexual activity                     

Did not talk with either of most recent partners 

about STD risk before having sexual intercourse [sections]        

Predictor(*)                         

Sig F[double dagger]

STD risk behavior at baseline 

[is less than]0.001

Condom use self-efficacy 

0.02

OCP use as reason for not using condoms with 

1 most recent partner 

0.04

Did not talk with 1 of most recent partners about 

STD risk before having sexual intercourse 

0.08

OCP use as reason for not using condoms with 

both most recent partners 

0.13

Used alcohol over half the time in connection with 

sexual activity 

0.25

Did not talk with either of most recent partners about 

STD risk before having sexual intercourse[sections]                

0.72

(*) Predictor variables are listed in order of importance using standardized [Beta] weights, F values, and partial correlation coefficients as criteria. Mallow Cp was used as the criterion for retaining variables in the final regression model. Final model [R.sup.2]=0.211. STD indicates sexually transmitted disease; OCP. oral contraceptive pills. [dagger][R.sup.2] change values were taken from models containing variable plus all variables entered in earlier steps. [double dagger] The significance of F statistic (Sig F) associated with each variable was taken from a final model including all the above variables [sections] Although not meeting the criteria for inclusion, this item was retained in the model to maintain interpretation of the dummy-coded variable sequence.

T1 STD risk behavior emerged as the most important predictor of girls' T2 STD risk behavior (standardized [Beta] weights, partial correlation coefficients, and F statistic values were used as criteria for ranking independent variables in order of their predictive importance); girls reporting the lowest levels of STD risk behavior scores at T1 were engaging in the lowest levels of STD risk behavior scores at T2 (change in[[Delta]] [R.sup.2]=0.16). Girls with high levels of condom use self-efficacy at T1 were also engaged in lower STD risk behavior at T2 ([Delta][R.sup.2]=00.002). Those who reported OCP use as a reason for not using condoms at T1 were engaged in higher levels of STD risk behavior at T2 ([Delta][R.sup.2]=0.02) as were those with high-risk patterns of partner communication at T1 ([Delta][R.sup.2]=0.006)0. Finally, use of alcohol in connection with sexual activity with most recent partner(s) at T1 was predictive of higher levels of T2 STD risk behavior ([Delta][R.sup.2]=0.004).

COMMENT

Findings from this study highlight important cognitive and behavioral predictors of STD risk behavior. For girls, a model including T1 measures of risky sexual behaviors, condom use self-efficacy, use of OCP, communication patterns with sexual partners about STD prevention, and use of alcohol in connection with sexual activity accounted for 21.1% of the variance in STD risk behavior 1 year later. For boys, the most important predictors of STD risk behavior at T2 included risky sexual behavior at T1, perceived barriers to STD prevention, low levels of perceived susceptibility to STD, negative outcome expectations associated with condom use, and low levels of condom use self-efficacy. Adding to the body of empirical research on adolescent sexual behavior, these findings are of relevance to persons who fund, design, and implement STD prevention programs for adolescents, as well as for those who conduct research in this area.

From a population perspective, absolute levels of STD risk behavior among both boys and girls in this cohort remained about the same during the 1-year interval of this study. Neither a shift toward self-protective behavior nor toward increased risk-taking behavior was seen during the 1-year follow-up. Previous studies examining changes in the levels of sexual risk behavior among adolescents over time have demonstrated mixed findings. While several studies[9,31] have reported developmental progression toward caution among older adolescents and male adolescents, other findings have documented increased risk-taking behavior over time among female adolescents.[9]

Among adolescents in this cohort, the level of sexual risk behavior at T1 was the most important predictor of sexual risk behavior 1 year later; participants engaging in highest levels of STD risk behavior at T1 also were engaging in highest levels of risk behavior 1 year later. The stability of high-risk sexual behavior over time suggests that to be most effective, preventive interventions may need to be initiated early, before high-risk behavior patterns are established and become difficult to modify.

After accounting for T1 behavior, condom use self-efficacy emerged as the strongest predictor of change in sexual risk behavior over this 1-year interval among females; girls with highest levels of condom use self-efficacy at T1 were engaged in the lowest levels of STD risk behavior at T2. While the small sample of boys prohibited multivariate analyses, bivariate findings suggested that T1 condom use self-efficacy may be a critical predictor of STD risk behavior among boys as well. These findings concur with several earlier studies on the importance of self-efficacy in shaping sexual behavior[16,32]; however, they differ from those of Shafer and Boyer,[6] who found a nonsignificant relationship between self-efficacy and condom use among a group of sexually active ninth-graders. Differences in findings may be attributable to differences in populations surveyed and/or differences in measures used to assess various constructs.

Among girls, multivariate analyses revealed that a particular pattern of OCP use was another important predictor of increased STD risk behavior. Specifically, girls who reported not using condoms because of OCP use at T1 were engaged in higher levels of STD risk behavior at T2. These girls may represent a group for whom condoms are considered primarily a backup method of birth control. In a previous study, Weisman and colleagues[19] found that consistent condom use was less likely among female adolescents who were consistent OCP users. Both studies suggest that professionals talking with adolescents about OCP use for pregnancy prevention also need to emphasize STD preventive behaviors including limiting number of sexual partners and using condoms on a consistent basis.

Another relationship highlighted by the multivariate analysis is between girls, STD risk behavior and partner communication practices. Girls who were able to talk with their partner(s) about aspects of STD risk such as previous sexual partners, condom use, and history of STD before having intercourse or without having intercourse were significantly less likely to engage in high-risk sexual behavior than were girls who did not talk with partner(s) about these issues or talked only after having had sex. Other studies have found partner communication to be an important correlate of sexual risk behavior among incarcerated youth[17] and among adolescents using family planning clinics.[14] The present results extend earlier findings by documenting the importance of this dynamic in a longitudinal context.

Although bivariate analyses uncovered relationships between several of the health belief measures and STD risk behavior, none of these relationships remained significant in multivariate analyses. This finding generates hypotheses for future study on the possible indirect effects of health beliefs on sexual behavior. It may be that an individual's beliefs directly influence such factors as self-efficacy and partner communication practices, which in turn have a more proximal and direct influence on sexual risk behaviors.

Several study limitations are worthy of discussion. First, the cohort used in these analyses does not represent a general sample of adolescents in the United States. All study participants were adolescents from a large metropolitan area who used health care services; therefore, they may not represent all sexually active metro-area youth, as not all youth use health clinics. Second, our data were collected through self-report. Many of the questions deal with personally sensitive information, introducing the possibility of social acceptability bias. The measure of condom use consistency used in this study does not appear to have a large social acceptability bias, as this self-report was found to be highly correlated with information obtained through clinician interviews and STD examinations.[33] However, the findings of Sandberg and colleagues[34] raise the possibility of the presence of measurement bias with other cognitive and behavioral measures. in future research, multiple methods and sources to assess concepts such as patterns of partner communication may provide a more accurate representation of underlying constructs.

This study has highlighted important individual-level influences on adolescents, sexual risk behavior. To understand the broad range of influences on adolescent sexual behavior, future research should incorporate individual-level factors, such as those used in the present study, with measures that tap social influences, such as perceived peer-group norms, peer behavior, or levels of family communication around sexual norms and contraceptive behavior. By examining intrapersonal and behavioral influences in a broader social context, indirect and direct influences can be elucidated, further clarifying the complex processes underlying adolescent sexual risk behavior.

Based on the findings presented herein, we suggest that school-, clinic-, and community-based efforts aimed at reducing and preventing STD risk behaviors begin early, before the onset of consistent patterns of high-risk sexual behavior. Among sexually active adolescents, interventions should include components that increase condom use self-efficacy, and build skills to communicate with sexual partners about STD prevention issues before having intercourse. Interventions should also address reasons adolescents give for nonuse of condoms, including use of OCP and, particularly for boys, low perceived susceptibility to STD and insufficient access to STD prevention services.

SUBJECTS AND METHODS

SUBJECTS

Data for the present study are from a 1-year longitudinal study of health beliefs, sexual behaviors, and STD acquisition among 549 adolescents, ages 14 to 21 years at baseline (T1), recruited from 7 school- and community-based clinics in the Minneapolis -- St Paul, Minn, metropolitan area. Potential subjects were recruited for the longitudinal study while attending participating clinics for routine health visits. All clinic attendees were eligible for the study unless they had completed antibiotic therapy in the previous 6 months, as such treatment could affect the detection of an STD. An additional exclusion criterion for adolescents using school-based clinics was enrollment in the 12th grade at T1, as 1-year follow-up (T2) could not be assured. T1 data were collected from mid-1989, to mid-1990; T2 data were obtained from mid-1990 to mid-1991, approximately 1 year after individuals, initial participation. All of the subjects were recontacted to request T2 participation by the clinic staff or by the study staff by using follow-up information that was provided by the adolescents at T1.

The cohort selected for this analysis consisted of 335 girls and 75 boys who completed surveys at T1 and T2 and who reported being sexually active at T2 (ie, having had vaginal intercourse with 1 or more partners of the opposite sex in the previous year). Of this sexuall