Background
Risky health behaviors such as cigarette/alcohol use, as well as the onset of sexual activity, are common from late adolescence to young adulthood in East Asian countries such as Taiwan. These behaviors increase an individual’s subsequent mental and physical health risks to some degree [
1‐
4]. Studies indicate nearly one-quarter of male youth in Taiwan self-report cigarette use at least once by the end of high-school, and the smoking rate more than doubles to 48% by the end of college [
5]. Although female cigarette use was somewhat lower among Taiwanese high school students (4-5%), smoking prevalence, particularly for youth, has continued to rise since Taiwan opened its tobacco market to foreign companies in 1987 [
6]. A recent population-based survey in Taiwan reports alcohol use by age 18 at about 38% [
1]; not surprisingly, that percentage is predicted to increase as adolescents move into early adulthood [
7]. Similarly, the same study [
7] reports premarital sex by age 18 among female adolescents in Taiwan is 22% among 15–19 year-old cohorts, but only 17% among those 20–24 year-old cohorts. Social norms regarding premarital sex are apparently becoming more liberal in recent times.
Problem behaviors (e.g., smoking, drinking, drug use, premarital sex) tend to cluster in youth as distinct social behaviors [
2,
8‐
11]. If the relationship between cigarette/alcohol use and subsequent sexual activity can be understood, interventions can be designed to efficiently and effectively decrease and/or limit the likelihood of adolescents’ misusing these substances and engaging in premarital sex. Thus far, there is relatively little empirical evidence regarding this link in Asian countries like Taiwan.
Mason and colleagues [
12] analyzed data gathered from a sample of 808 American youth, followed from age 10 to age 24. Their findings suggest early alcohol use correlates with a higher likelihood of risky sex. However, that study is based on a Western population, and its analyses did not include factors related to social context and peer influence; therefore, additional risk behaviors often associated with adolescent development were not assessed.
Using the Theory of Planned Behavior [
13‐
15] and Social Learning Theory [
16,
17], this study explores the association between adolescent cigarette/alcohol use and premarital sex from a temporal perspective. The Theory of Planned Behavior asserts an adolescent’s intention to engage in premarital sex flows from a youth’s attitude toward premarital sex, his/her perception of the subjective (cultural) norms associated with such behavior, and his/her beliefs regarding his/her ability to engage in sexual activity. How these factors are influenced by relationships with peers and stressful life events are also central to the dynamics of when, how, and to what degree youth engage in premarital sex. In addition, Social Learning Theory posits peer group interaction portrays a significant role in cigarette/alcohol use and subsequent premarital sex among college students. In a sample of Taiwanese youth who successfully completed a competitive college entrance exam, adolescents report peer pressure as a powerful influence in their decision to initiate and experiment with a variety of risky health behaviors [
16,
18]. Previous research in the Philippines and Taiwan identifies perceived sexual activity by best friends as an influential factor in the initiation of sexual activity in an adolescent’s own life [
8,
19]. In the present cohort study, we extend our analyses to include the temporal relationship between peer pressure and premarital sex during college.
Using longitudinal data from the Taiwan Youth Project, we specifically address this research question: What is the relationship between adolescent cigarette/alcohol use and the likelihood of premarital sex during college, after controlling for premarital sex attitudes, peer influence, life and behavioral events, and family background? The analyses were conducted separately for male and female college youth, due to gender differences suggested by a wide range of existing literature [
7,
19].
Taiwan provides an interesting cultural context to investigate this question. Traditional Taiwanese culture values conservative social norms such as responsibility, restraint, interdependency and collectivism, suggesting premarital sex is not endorsed, especially among single adults [
20]. Taiwanese college students living in this cultural context are exposed to a high degree of normative and institutional influence, which leads them to be “well-behaved” and not engage in smoking, drinking, or premarital sex in late adolescence. Over three-fourths of Taiwanese high school graduates enroll in college [
21,
22]. Studies have shown the school environment serves as a protective factor leading to less deviance, less psychological maladaptation and less premarital sex in Taiwan [
8,
23]. Even within this context, however, risky behaviors such as cigarette/alcohol use and premarital sex occur, creating a salient public health issue deserving further investigation.
Discussion
As theories and prior research suggest, our results found cigarette smoking, alcohol use, and premarital sex tend to cluster in youth [
2,
8‐
11]. Moreover, our findings demonstrated the influence of adolescent cigarette/alcohol use on later premarital sex, and showed gender differences related to self-reported premarital sex during college. Overall, adolescent cigarette/alcohol use was higher among Taiwanese youth who were sexually active by age 20 when compared to their sexually abstinent counterparts. In addition, for those abstinent youth, the
predictive effect of adolescent cigarette and alcohol use differed for male and female. For males, adolescent “heavier drinking” was significantly related to later premarital sex (by age 22); however, for females, adolescent “light drinking” was significantly associated with later premarital sex. These results are consistent with previous findings of a positive association between alcohol use and premarital sexual activity [
12,
29,
30]. Our analyses further suggest gender differences in a dose–response effect in which the likelihood of later premarital sex is associated with heavier drinkers in males vs. light drinkers in females. And, these results imply that females may suffer more than males from being alcoholic in social norms. The finding of no significant predictive effect of cigarette smoking on self-reported premarital sex for females is consistent with previous reports [
31,
32], but the significant predictive effect of cigarette smoking for males differs from the existing literature. This might be because the adolescent smoking rate is much lower among females in the sample (2%) than for males (8%). It is recommended future research studies examine how/if adolescent cigarette use links to female sexual activities.
Our findings are consistent with the Theory of Planned Behavior and its emphasis on cultural norms and attitudes as related to attitudes toward premarital sex. Previous studies found adolescent premarital sexual activity to be shaped by or associated with contextual factors [
8,
33‐
35]. These contextual factors (e.g., school attendance and community participation) impose conventional behavioral expectations with adolescents expected to abstain from sex until marriage. Prior research also showed a significant association between liberal attitudes toward premarital sex and premarital sexual activity in Nepal college males [
36]. In our college sample, those adolescents with permissive premarital sexual attitudes were more likely to engage in premarital sex during college, regardless of gender. Premarital sex appears to be increasingly common in the romantic relationships of the Taiwanese young populace [
27]. Our results suggest a cultural shift regarding premarital sexual activity among Taiwanese college-age youth. The increased acceptance of more liberal and permissive sexual attitudes is associated with the increased likelihood of sexual activity among Taiwanese youth. It is recommended that future cigarette, alcohol and sexual risk materials developed for Taiwanese college students take these shifting social norms into consideration and pay special attention to gender specific measures.
In addition, there are potentially serious health and social implications for those engaging in premarital sex without protection such as use of condoms or other contraceptives to prevent unwanted pregnancy or sexually transmitted diseases [
37]. In fact, contraceptive use is not common among Taiwanese college students, although the proportion using contraceptives does increase by age [
7]. In separate analyses, we found that less than half (43%) of college students who had premarital sex by age 20 reported consistent contraceptive use. Approximately half (52%) of college students who had premarital sex by age 22 self-reported consistent contraceptive use. Because premarital sex among college youth will likely become more prevalent, encouraging consistent contraceptive use should become a major point of emphasis for health professionals dealing with Taiwanese college students.
As suggested by Social Learning Theory, peer group influence plays a significant role in cigarette smoking, alcohol use, and premarital sex among college students. Our results partly supported that respondents’ perceptions of their close friends’ risky behaviors (cigarette smoking, alcohol use and sexual activity) were higher among youth who were sexually active by age 20 than those who were sexually abstinent. Even when controlling for adolescent risky health behaviors, multivariate analyses in Model 2 showed later premarital sex (by age 22) was positively associated with perceived close friends’ sexual behavior [
8,
19,
36]. In contrast, the non-significant predictive effect of perceived cigarette smoking of best friends for males and non-significant predictive effect of perceived alcohol use of best friends for both sexes may be due to sample selection. Our college sample remained in the protective school environment throughout the time period of this study, a factor that may contribute to a delay in engaging in premarital sex. This finding might be interpreted via the
age-integrated hypothesis [
38], in which an increasing strength of social bond is particularly influential to college students who remained sexually abstinent by age 20. This bond to a social institution thus serves to mitigate peer influence as well as reduce problem behaviors.
This study addressed a seriously under-explored subject with regard to whether stressful life events are associated with the link between adolescent cigarette/alcohol use and premarital sex [
39]. As prior research suggests, individuals may engage in premarital sex to escape from stressful life events, and such an effect may last as long as 2 years [
40]. Consistent with the previous study using a U.S. sample [
39], our findings demonstrated experiencing a stressful personal event was associated with a higher likelihood of engaging in premarital sex during college for both sexes in Taiwan. In addition, considerable evidence suggests an association between stressful life events and psychological distress in adolescence. Our findings further support this relationship. Additional research is needed to examine the relationships between and among stressful life events, initial sexual encounters, and long-term psychological well-being.
Consistent with prior research [
8,
33,
36], family characteristics were found to be significantly related to later premarital sex among college students in Taiwan, but the results slightly differ for males versus females. The analyses suggest male college students with an educated mother were less likely to have premarital sex between 20 and 22; this relationship was not observed for female college students. In contrast, level of family income was inversely associated with increased odds of premarital sex involvement between 20 and 22 for female college students, but not for male college students. We suspect the differential gender effects on premarital sex may be due to the correlation between family SES and parenting in the college sample. While resourceful family reveals protective function, lower SES background tends to perform less parental control. Additional research is thus needed to compare the family SES context between males and females, and disentangle the various effects including parenting. Given families appear to be particularly influential for Taiwanese college students, understanding the ways in which
family impacts sexual behavior is essential.
There are two major limitations in this study. Firstly, endogeneity may be present in the relationship between substance misuse and sexual activity [
41]. Hence, we adopted a cohort sample to disentangle simultaneous causation between adolescent cigarette/alcohol use and premarital sex during college. After controlling for several other relevant variables in the model, our research was able to document the importance of substance use in predicting premarital sex during college. Secondly, in a Taiwanese context, our results may suffer from under-reporting of sexual behavior by youth due to the social desirability of sexual abstinence [
42]. However, it should be noted extensive efforts were made in two stages to retain samples in the process. At the first stage, original samples (from Wave 1) were sent a notice of a follow-up survey. Each sample was then visited at least 3 times in order to conduct the interview at the field site. For participants who did not have access to or failed to provide consent for an interview, a second stage effort was involved. The district supervisor re-visited these samples’ family to gain access for interview. Since the survey began in 2000, New Year’s holiday greeting cards have been mailed annually to participants to maintain some contact. These efforts to generate long-term cooperative relationship with the participants since early adolescence helped to gain their trust, with less than 3% providing inconsistent answers regarding self-reported sexual experiences in the two surveys [
8].
Conclusions
In sum, our analyses clearly highlight gender differences in the temporal relationship between adolescent cigarette/alcohol use, perceived peer characteristics, stressful life and behavioral events, and the likelihood of engaging in premarital sex among Taiwanese college youth. Whereas group differences show higher proportions of cigarette and alcohol use among the youth who engaged in premarital sex than those who were sexually abstinent for both sexes, the multivariate analysis in the sexually abstinent group further indicates that two years later, adolescent cigarette smoking decreased as a predictor of subsequent premarital sex for females, while the effect of alcohol use remained a significant predictor of premarital sex. Reducing drinking, providing immediate counseling for those who experience a personal stressful event, and promoting positive attitudes toward intimate relationship before marriage during adolescents, may be helpful intervention message content in delaying the age of onset of premarital sex in young adulthood. Specifically, interventions promoting the delay of the onset of drinking and delivering appropriate sexual education during high school may be particularly useful in preventing Taiwanese college students from engaging in risky sexual behavior.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
CC was responsible for development of study hypotheses, data analysis, and drafting of the article. CCY and KK contributed to critical revision of the article. All authors involved in the writing of the paper, and all approved the final submission. All authors read and approved the final manuscript.