Does Parental Involvement Predict New Sexually Transmitted Diseases in Female Adolescents?

Background  African American female adolescents living in low-income urban areas are at increased risk for sexually transmitted diseases.

Objective  To determine if high levels of perceived parental supervision and communication were associated with reduced gonorrhea (GC) and chlamydia (CT) incidence in low-income, African American, sexually experienced female adolescents, aged 14 to 19 years, attending urban health clinics.

Design  A prospective cohort study was used to determine the predictive value for high levels of parental supervision and communication on GC and CT infection in 158 adolescent females. Multiple logistic regression analysis explored the association between incident infection and perceived parental supervision and perceived parental communication while controlling for relevant demographic and behavioral factors (age, religious involvement, school enrollment, a 2-parent household, having a main sex partner, and having concurrent sex partners).

Results  When adjusted for age and baseline GC and CT infection, high levels of perceived parental supervision were associated with reduced GC and CT incidence (adjusted odds ratio, 0.06; 95% confidence interval, 0.01-0.31). High levels of perceived parental communication were not associated with reduced GC and CT incidence (adjusted odds ratio, 0.55; 95% confidence interval, 0.21-1.42).

Conclusions  The link between parental supervision and disease acquisition is particularly valuable because it provides evidence that parental supervision can result in lower sexually transmitted disease rates in urban high-prevalence populations. This is important for interventions designed to increase parental involvement as a strategy for promoting protective sexual behaviors in female adolescents because it indicates that increased parental involvement can also influence subsequent disease acquisition.

Sexually transmitted diseases (STDs) continue to be an important health problem for adolescent females. The highest age-specific rates in women for gonorrhea (GC) and chlamydia (CT) are seen in female adolescents aged 15 to 19 years, with disproportionately high rates in African Americans.^1,2 Understanding the effect of family context on STD acquisition in adolescent African American females is important because modification of family factors may influence STD acquisition. Studies in adolescents have shown good parental communication and supervision were influential in delaying sexual initiation,^3-5 and suggest an association between reduced parental supervision and an increase in STDs, an increase in number of sex partners, and a reduction in condom use.^6-8 Among poor, urban, African American female adolescents, high levels of perceived parental supervision have been associated with fewer pregnancies and less risky sexual behavior.^9,10 Few studies have examined the effect of perceived parental supervision and communication on adolescent STD acquisition using longitudinal data.

We were interested in determining if high levels of perceived parental supervision and communication predicted reduced GC and CT infections over time in adolescent, low-income, African American, sexually experienced females attending urban health clinics. We hypothesized that high levels of parental supervision and communication would be protective against infection. By identifying family factors that prevent GC and CT infection in sexually active African American female adolescents, interventions can be developed to include adolescents' parents to assist in reducing the burden of STDs in this high-risk population.

A 2-year (August 1, 2000-September 30, 2002) prospective cohort study was used to determine the predictive value for high levels of parental supervision and communication on GC and CT infection in adolescent females. Of the 279 female participants enrolled in the study at baseline, 158 (56.6%) completed the 6-month laboratory screening. There were no differences in age, risk perceptions, or risk behavior between those completing the study and those unavailable for follow-up.

Participants were recruited from 2 urban health clinics. One was a public STD clinic and the other was a hospital-based adolescent medicine clinic. Both clinics served similar populations—primarily African American youth, aged 12 to 21 years. We included in our analysis English-speaking female participants between the ages of 14 and 19 years who, at baseline, had vaginal or anal intercourse with an opposite sex partner in the preceding 3 months, resided within the Baltimore metropolitan area, completed an interview on perceived parental supervision and communication, and provided a urine specimen for laboratory testing. At baseline, 97.9% (279/285) of the eligible adolescents agreed to participate. The study received institutional review board approval from The Johns Hopkins University, and all participants provided written informed consent; parental consent was not required because these adolescents were seeking confidential health services.

Participants completed baseline face-to-face interviews on perceived parental supervision, perceived parent-adolescent communication, sexual practices, and human immunodeficiency virus and STD risk behaviors. Participants were screened for GC and CT at baseline and at 6 months using a urine-based ligase chain reaction test (LCx; Abbott Laboratories, Whippany, NJ).

Subjects who tested positive by urine ligase chain reaction for GC or CT infection were considered to have a current STD infection. Participants testing positive for GC or CT were treated at no cost, according to standard clinic procedures.

Scales were designed to measure perceived parental supervision and perceived parental communication.

The perceived parental supervision scale consisted of 3 questions adapted from the Silverberg and Small Parental Monitoring Scale.^3,11 The instrument assesses the respondent's perception of parental awareness of her activities and whereabouts with the following questions: (1) "How much do your parents or guardians try to know about where you go at night?" (2) "How much do your parents or guardians try to know about what you do with your free time?" and (3) "How much do your parents or guardians try to know about where you are most afternoons?"

Responses for all 3 questions ranged from "not at all" to "most they can" on a 5-point Likert scale, with upper values reflecting high levels of perceived parental supervision. The responses were z scored and then summed.

The 6-question perceived parental communication scale was adapted from the 10-item Open Family Communication Scale.¹² The following statements measure participants' perceptions of communication with their parents: (1) "I find it easy to discuss problems with at least one of my parents or guardians." (2) "I am satisfied with how at least one of my parents or guardians and I talk together." (3) "At least one of my parents or guardians is a good listener." (4) "I can discuss my beliefs with at least one of my parents or guardians without feeling restrained or embarrassed." (5) "At least one of my parents or guardians tries to understand my point of view." (6) "It is easy for me to express my true feelings to at least one of my parents or guardians."

The responses for the communication scale ranged from "strongly agree" to "strongly disagree" along a 4-point Likert scale, with lower values representing better parent-adolescent communication. For ease of understanding, the communication scale was reverse coded to match the direction of responses for the supervision scale. The responses were z scored and then summed.

To determine if high levels of supervision and communication were protective, we dichotomized both scales at the upper quartile (75%). Our results did not differ whether the scales were continuous or categorical. The internal consistency of each scale as measured by the Cronbach α¹³ was .85 for perceived parental supervision and .91 for perceived parental communication.

The Pearson product moment correlation coefficient (r) was used to examine correlations between scale scores and age. Overall, small to moderate correlations appeared between the perceived parental supervision and communication scales (r = 0.29, P<.001) and between age and supervision (r = −0.12, P = .04), with perceived supervision decreasing with increasing age. No significant correlation was found between perceived parental communication and age (r = −0.08, P = .19).

Logistic regression analysis was used to explore the associations between GC and CT incident infection and perceived parental supervision, perceived parental communication, and other demographic and behavioral factors, such as age, religious involvement, school enrollment, a 2-parent household, having a main sex partner, and having concurrent sex partners. Religious involvement, school enrollment, age, concurrent sex partners, and partner type have all been associated with adolescent sexual risk behavior.^14-17 If these factors were associated in our study, we were interested in controlling for their effects to examine the additional effects of parental supervision and communication. P = .10 was used as the cutoff for inclusion of nonstudy variables in the multiple regression models.

Multiple logistic regression analysis was used to examine perceived parental supervision and communication while controlling for relevant demographic and behavioral characteristics. In the adjusted logistic regression models, P = .05 was used as the cutoff for significance and stepwise selection was used to determine the final adjusted models. Parental supervision and communication were separately analyzed in adjusted models to account for the moderate correlation between the 2 variables. The SAS statistical software package, version 8.0 (SAS Institute Inc, Cary, NC), was used for all analyses.

The characteristics of the participants are shown in Table 1. Participants from the 2 sites were similar with regard to demographic and behavioral characteristics, and differences were not statistically significant (P>.31). Participants' parental characteristics are shown in Table 2. Fewer than 20% of the participants' parents were married or living together. The prevalence of GC and CT in this population at baseline was 30.5% (85/279). The incidence at 6 months was 20.9% (33/158), or 41.76 per 100 person-years or 3.48 per 100 person-months.

Predictors of incident infection are presented in Table 3. The univariate analysis indicated that high levels of parental supervision were predictive of fewer GC and CT infections. High levels of parental communication were not predictive of fewer infections. Other characteristics, such as partner type, condom use at last sex, history of condom use, concurrent sex partners, clinic site, religious involvement, living in a 2-parent household, and school enrollment, were not predictive of infection (results not shown).

In the multivariate analysis, parental supervision and parental communication were run in 2 separate models. Age, baseline GC and CT infection, and parental supervision remained in the final model. When adjusted for age and baseline GC and CT infection, perceived parental supervision remained predictive of fewer infections. High levels of parental communication were not predictive of GC and CT infection in the adjusted model.

Our prospective results showed that high levels of perceived parental supervision led to a reduction in the laboratory-confirmed incidence of GC and CT in African American female adolescents, regardless of their age. These results add to increasing research^3,8,10,18 on the effect of family-level factors on adolescent STD risk. Parental supervision has been associated with other aspects of adolescent sexual behavior in similar populations,^3,6 and our results showed a protective effect against GC and CT infections as well.

Parental involvement as a strategy for promoting protective behaviors among adolescents is increasingly a subject of research,¹⁹ and our results provide further evidence that interventions designed to increase parental involvement may affect not only adolescent behavior but disease acquisition as well. Results from a recent randomized controlled trial²⁰ indicate that increased levels of parental supervision resulted from a program designed to promote parental supervision. Our results indicate that interventions designed to promote parental supervision in high-risk single-parent families also may reduce adolescent STD incidence. Because most adolescents enrolled in our study came from single-parent homes (81.4%), our results are particularly salient for urban high-risk populations because they indicate that the protective effect of parental supervision occurs regardless of family composition.

In our study, high levels of parental communication had little effect on GC and CT incidence. This finding is consistent with other prospective communication studies,⁵ which found perceived parental communication had no effect on sexual activity at 6 months. One explanation for this may be the difficulty of measuring parent-adolescent communication.²¹ Communication has been shown to vary by the sex of the adolescent and parent,^22-24 and to depend on the parental style of communication.²⁵ Our study measured only perceived parental communication as a factor for family connectivity. It did not examine the content, style, or delivery of the message—all of which may influence adolescent sexual behavior.

Other factors, such as partner type, condom use, religious involvement, and school enrollment, were not predictive of infection. While these factors have been associated with sexual risk behavior, we were unable to show they influenced disease acquisition. It is possible that these factors may have less influence on disease acquisition over time. However, a more likely explanation may be that our population was relatively homogeneous in behavior and, therefore, we were unable to detect an effect.

A limitation of our study is the proportion of adolescents who returned for follow-up testing at 6 months (56.6%). There were no differences at baseline in age, risk perceptions, or risk behavior between those completing the study and those unavailable for follow-up; therefore, we do not believe those participants unavailable for follow-up were at a different risk for GC or CT than those who remained in the study.

Our findings suggest future studies should examine the influence of perceived parental supervision and communication on STD infections in an adolescent male population. Research^18,26,27 indicates sex differences exist in perceived supervision and communication; therefore, these parental factors may not have similar effects among male adolescents.

Understanding the link between parental supervision and STD acquisition will allow us to develop effective interventions that include parents. Interventions that include videotaped instruction and role playing have been effective at increasing parental monitoring among single mothers of high-risk African American youth.²⁰ Our results indicate interventions that increase parental supervision may influence STD acquisition and are important to incorporate with adolescent-centered risk reduction activities.

Parental involvement as a strategy for promoting protective behaviors among adolescents is increasingly a subject of research. Studies in adolescents have shown good parental communication and supervision were influential in delaying sexual initiation, and suggest an association between reduced parental supervision and an increase in STDs, an increase in number of sex partners, and a reduction in condom use. Few studies have used longitudinal data to examine the influence of parental supervision and parental communication on disease acquisition. Our study shows that perceived parental supervision was associated with GC and CT incidence, while perceived parental communication was not associated with GC and CT incidence. Our results provide further evidence that interventions designed to increase parental involvement may affect not only adolescent behavior but disease acquisition as well.

Accepted for publication March 8, 2004.

This study was supported by grant R01-AI36986 from the National Institute of Allergy and Infectious Diseases, Bethesda, Md; and grant MCJ000987 from the Maternal and Child Health Bureau, Health Resources and Services Administration, Rockville, Md.

Correspondence: Jonathan M. Ellen, MD, Department of Pediatrics, The Johns Hopkins School of Medicine, Park 307, 600 N Wolfe St, Baltimore, MD 21287 (jellen@jhmi.edu).