Two risky health behaviors of parents with adolescent children--smoking and not using seat belts--appear to be linked to their children's sexual and contraceptive behaviors. According to an analysis of 1994-1995 data from the National Longitudinal Study of Adolescent Health (Add Health),1 an adolescent whose parent smokes has independently elevated odds of ever having sex and of having sex before age 15 (odds ratios, 1.4-1.5), as well as of using a contraceptive method at first sex (1.3 for males only). Further, adolescent males whose parent rarely uses seat belts have significantly increased odds of ever having sex (1.3). There is no independent association, however, between parental heavy drinking and the adolescent sexual and contraceptive behaviors examined.

The data come from the adolescent and parent questionnaires administered in the first wave of the Add Health survey. The data for analysis, which refer to more than 9,000 adolescents of each gender in grades 7-12, were weighted to be nationally representative. Binary logit analyses were performed to assess the influence of a parent's risk behaviors (in 95% of cases, the mother's) on the odds of sexual initiation, early first sex (before age 15) and contraceptive use at first sex (i.e., use of any method or a condom or medical method in particular).

Separate models by gender controlled for socioeconomic variables and the high-risk behaviors of one parent--smoking, drinking heavily (five or more drinks at a time) in the past month, and never or rarely wearing seat belts. The analyses also included controls for the adolescent's own risk behaviors--smoking regularly, binge drinking, having friends who use alcohol or drugs, and engaging in delinquent behaviors (i.e., deliberately damaging others' property, stealing, using or threatening to use a weapon, and selling marijuana or other drugs). The investigators also controlled for parental supervision with measures of how often each parent was present when the child left for school, returned home from school and went to bed. All analyses controlled for family income and welfare receipt and parents' educational attainment, as well as adolescents' age, race, religious affiliation and religiosity, living arrangement and area of residence. In the analyses of predictors of contraceptive use, the investigators also controlled for adolescents' age at first sex and the year in which first coitus occurred.

At the time of the survey, the adolescents in the sample were 15.5 years old, on average. Males were significantly more likely than females to have ever had sex (39% vs. 37%) and to have done so before their 15th birthday (20% vs. 17% of those at least 15 years of age). There was no significant difference by gender, however, in the proportions who had used a method of contraception at first coitus (64-65% of sexually experienced adolescents). Thirty-one percent of the adolescents had a parent who smoked, 13-14% had a parent who drank heavily in the past month and 19-20% had one who never or only sporadically used seat belts. Mean supervision scores (summed over the three items) reported by the adolescents were uniformly higher for their mothers (8.6-8.7 on a scale of 0-12) than for their fathers (4.9-5.3).

Net of all other variables, having a parent who smoked was significantly associated with elevated odds of sexual initiation among both females (odds ratio, 1.5) and males (1.4). A parent's nonuse of seat belts was independently related to increased odds of sexual debut, but only among males (1.3); parental drinking, however, had no effects on adolescents' sexual debut. Levels of each parent's supervision had gender-specific effects only: The increased presence of a mother throughout her child's day was independently associated with lowered odds of sexual initiation among daughters (0.96), and a father's higher level of supervision was similarly associated with reduced odds of sexual debut among sons (0.96).

Having a parent who smoked was independently associated with elevated odds of having had sex before age 15 among both males (1.5) and females (1.4). Whereas neither a parent's drinking nor seat belt nonuse had an impact on an adolescent's early sexual debut, increasing maternal supervision was associated with significantly reduced odds of early first sex among adolescent females (0.9). A father's presence at home, however, had no independent effect on early initiation of sexual activity among adolescents of either gender.

Whereas none of the parental risk behaviors affected the odds of a daughter's contraceptive use at first sex, smoking was independently associated with elevated odds of use at first coitus among sons (1.3). The investigators suggest that smokers' potential openness toward parent-child discussions about contraception may partially explain this unexpected association; when the authors controlled for such discussions, the association was no longer significant. They maintain that parental risky behaviors may also indirectly affect adolescent contracepive behavior through their demonstrated influence on age at first sex, a known predictor of method use among adolescents. Finally, although the three parental risky health behaviors had no independent effect on the type of method used by adolescent children, sons whose fathers scored higher on the supervision scale had significantly reduced odds of using a medical method the first time they had sex.

The authors cite several potential limitations of their study, including the lack of information about the timing of the behaviors studied (which made it impossible to disentangle cause and effect), the unavailability of data from both parents and the wording of the items defining the risk behaviors. Nonetheless, they affirm that their findings "support the notion that risk is 'reproduced' across the generations, perhaps because parents often serve as role models for their children." They conclude that "public health campaigns that urge parents to act responsibly by engaging in health-conscious behaviors are likely to help reduce precocious and unsafe sexual activity among teens."--L. Remez


1. Wilder EI and Watt TT, Risky parental behavior and adolescent sexual activity at first coitus, Milbank Quarterly, 2002, 80(3):481-524.