Sexual behavior and drinking style among teenagers: a population-based study in Finland

SUMMARY

In this large-scale study, we examined the relationship between an adolescent’s sexual behavior and drinking style. Three aspects of sexual risk-taking were included: early activity, unprotected sexual intercourse and having sex with multiple partners. A distinction was made between different drinking styles, i.e. alcohol drinking and drunkenness-related drinking. Cross-sectional school survey data from the School Health Promotion Study was collected in Finland in 2002–2003. The national sample consisted of adolescents from the eighth and ninth grades (n = 100 790). The mean ages were 14.8 and 15.8 years. Using logistic regression analysis, we investigated the association between sexual behavior and drinking style among teenagers. The likelihood of engaging in sexual intercourse increased with the frequency of alcohol use. In particular, frequent drunkenness-related drinking increased the probability that the teenager had experienced sexual intercourse. The likelihood of engaging in unprotected sex and/or having multiple sexual partners was many-fold for adolescents drinking frequently until they were in a state of drunkenness. Particularly for girls, weekly drunkenness-related drinking was associated with multiple partners. The vast majority of sexually experienced under-aged adolescents drink alcohol, many of them until they are drunk. Thus, it could be effective to combine both alcohol education and sex education, including contraceptive counseling, in early adolescence.

INTRODUCTION

Sexual activity and alcohol use in early and middle adolescence are often regarded as risky behaviors even in liberal and secularized western cultures (Plant and Plant, 1992; Udry and Bearman, 1998; Windle, 2003), and this behavior largely causes concern in a society. On the other hand, it has been emphasized that risk-taking among adolescents is to a certain extent normal (Plant and Plant, 1992). Experimenting both with intimate relationships and alcohol drinking belong to a life-style typical of many teenagers, and initiations of these behaviors often overlap. The timing of these events partly depends on biological maturation: early initiation of sexual intercourse (Capaldi et al., 1996; Zimmerman-Gembeck et al., 2004) and advanced alcohol (Lintonen et al., 2000a; Bratberg et al., 2005; Costello et al., 2007) use are both linked to early maturation age.

There is a generally liberal attitude towards sex in the Nordic countries, and attitudes to adolescent sexuality are quite permissive. Among Finnish adolescents, around one-third of girls and one-quarter of boys have their first sexual intercourse before the age of 16 (Kosunen and Ritamo, 2004). In Western Europe, adolescents experience their first sexual intercourse at the mean age of 17–18 (Bajos et al., 2003). Concurrently, the vast majority of adolescents under the minimum legal drinking age have consumed alcoholic beverages in most European countries (Hibell et al., 2004). National surveys in the USA indicate that under-aged alcohol use is common among American adolescents as well (Johnston et al., 2006).

Alcohol drinking, especially drunkenness-related drinking in adolescence has been generally acknowledged to be a major public health problem. Alcohol use is known to be related to a variety of short- and long-term harms and risk-taking behaviors (Wechsler et al., 1995). According to recent studies, alcohol drinking is particularly linked to sexual risk-taking behavior (Halpern-Felsher et al., 1996; Fergusson and Lynskey, 1996; Zimmerman-Gembeck et al., 2004; Parkers et al., 2007; Lavikainen et al., in press), which commonly consists of early age of first intercourse, unprotected sexual intercourse exposing the adolescent to sexually transmitted infections (STI) and HIV, teenage pregnancies and sexual intercourse with multiple partners (Tapert et al., 2001; Buhi and Goodson, 2007).

In Finland, convergent trends in adolescent sexual health and alcohol use have been identified during the last 10 years. Rates of chlamydia infections and induced abortions began to increase among 15-19-year-old girls in the mid-1990s (Hiltunen-Back et al., 2003; Gissler, 2004). Similarly, adolescent alcohol drinking and drunkenness-related drinking in particular increased right after the early 1980s up to the end of the 1990s. Come the turn of the millennium, the increasing trends in alcohol use among 14-18-year-old adolescents came to a halt. (Lintonen et al., 2000b; Rimpelä et al., 2006) From the year 2002 onward, a decline has been seen in indicators for both induced abortions and substance use (i.e. frequency of drinking and drunkenness, tobacco smoking and social exposure to illegal drugs). This outstanding congruence between the trends of adolescent sexual health and alcohol use is of interest.

Previous results pertaining to the association between adolescents’ sexual risk-taking, especially non-use of contraceptives, and alcohol use, have been somewhat contradictory (Morrison et al., 2003; Parkers et al., 2007), possibly because drinking is often measured only as the frequency of alcohol use. Thus, only a little is known about the influence of drinking style on adolescent sexual behavior, i.e. alcohol use and drunkenness-related drinking should be regarded as separate elements of drinking style. Drunkenness has been largely connected with the concept of disinhibition as people may become less inhibited under the influence of alcohol (Plant and Plant, 1992). Among adolescents, for instance, being drunk may both increase reckless behavior and the likelihood of failing to use condoms. The main motive for adolescents to drink alcohol might actually be to become ‘disinhibited’ (Abel and Plumridge, 2004).

In this large-scale study, we investigate the associations between sexual behavior, namely engaging in sexual intercourse, engaging in unprotected sexual intercourse and having multiple sexual partners and drinking style among teenagers. Instead of analyzing adolescent alcohol use as such, we have made a distinction between different drinking styles: the difference between frequencies of alcohol drinking and drunkenness-related drinking.

MATERIAL AND METHODS

The School Health Promotion Study (SHPS) is a national cross-sectional survey concerned with health, health behavior and school experiences of adolescents in eighth and ninth grades in secondary school. The SHPS has been carried out from 1996 onwards in two parts: even-numbered years in provinces in Southern Finland, Eastern Finland and Lapland and odd-numbered years in provinces in Western Finland, Oulu and Åland. Participation has been based on voluntariness and the interest of each municipality and the schools within the area (Luopa et al., 2005).

The data utilized in this paper was obtained from SHPS surveys in 2002 and 2003. During that period 374 out of 448 Finnish municipalities fell into the data collection areas. Participation was good: in 2002, 149 municipalities (out of a possible 155) and in 2003, 207 municipalities (out of a possible 219) took part. The total number of adolescents here was 100 790, while one birth cohort in Finland was around 60 000 during those years (i.e. adolescents born in 1987–1989) (Statistics Finland). Of the respondents, 51 698 were eighth-graders (mean age 14.8, SD 0.4) and 49 092 ninth-graders (mean age 15.8, SD 0.4).

Data was gathered by an anonymous classroom inquiry, thus the student questionnaire did not include name or any other information that would identify respondent individually. The structured questionnaire was completed during a school lesson under the supervision of a teacher. To assure confidentiality, all questionnaires per classroom were enclosed in an envelope directed to the SHPS research group. Adolescents absent on the day of survey were not contacted. Based on the information on SHPS data collections, 10–15 percentages of adolescents are absent each day (Kaltiala-Heino et al., 2003).

Sexual behavior

Adolescents were classified as being sexually experienced if they had answered ‘yes’ to the following question: ‘Have you ever experienced sexual intercourse?’ Further, adolescents who had answered in addition or only to either of the questions: ‘How many times have you had sexual intercourse?’ and ‘What kind of contraception did you use in the most recent intercourse?’ were considered as sexually experienced. After re-checking the data manually, nonsensical responses to the questions of sexual experience were excluded.

The number of sexual partners in a lifetime was asked by the question ‘With how many different partners have you had sexual intercourse?’ The original alternatives were ‘one’, ‘two’, ‘three or four’ and ‘five or more’. For the analysis, the number of sexual partners was dichotomized (one to two partners/three or more partners). The concept of multiple partners is regarded as ambiguous, given that the number of partners considered high, in a word multiple, differ from study to study (Fergusson and Lynskey, 1996; Herlitz and Ramstedt, 2005; Stuave and O’Donnell, 2005). In this present study, we consider three or more different partners as multiple.

Referring to the question of used contraception in the most recent intercourse, the original alternatives were ‘none’, ‘condom’, ‘oral contraceptives’, ‘condom and oral contraceptives’ and ‘other method’. The open-ended option, ‘other method’, included answers primarily relating to methods of natural family planning (withdrawal and calendar method) which were classified as non-use of contraceptives. For analyzing purposes, the use of contraceptives was dichotomized (protected intercourse/unprotected intercourse). Options involving condom use, oral contraceptives or both (i.e. double contraception) formed the category ‘protected intercourse’, while options ‘none’ and ‘other method’ were grouped together to form the category ‘unprotected intercourse’. In this present study, we consider non-use of contraceptives as sexual risk-taking behavior.

Maturation

Information concerning adolescents’ maturation was assessed by asking ‘How old were you when you had your first menstruation or ejaculation?’ Alternative responses were ‘10 years or younger’, ‘11 years’, ‘12 years’, ‘13 years’, ‘14 years’, ‘15 years or older’ and ‘I have not yet experienced menstruation/ejaculation’. For the analysis, the two lowest categories were combined into ‘11 years or younger’ and the three highest categories were combined into ‘14 years or older or no menstruation/ejaculation yet’. Categories ‘12 years’ and ‘13 years’ were kept as such.

Drinking style

Adolescents’ drinking style was based on two questions. The first question covered the frequency of alcohol drinking: ‘How often do you use alcohol altogether, e.g. half a bottle of beer or more’. Original alternatives were: ‘I don’t drink alcohol’, ‘rarely’, ‘about once a month’, ‘a couple of times a month’ and ‘once a week or more often’. Subsequently, the categories ‘I don’t drink alcohol’ and ‘rarely’ were combined for the analysis. The second question covered the frequency of drunkenness: ‘How often do you use alcohol until you are really drunk’, with four alternatives: ‘never’, ‘rarely’, ‘one to two times a month’ and ‘once a week or more often’. The concept ‘being really drunk’ is well understood among Finnish adolescents, referring to the consumption of around 100 g of pure ethanol (e.g. six bottles of medium strength beer) (Lintonen and Rimpelä, 2001).

Response activity concerning the questions of adolescents’ sexual experience, sexual partners, used contraception, maturation and drinking styles was substantially high: only 2–3% of the adolescents in the eighth and ninth grades had left the question open.

Statistical analysis

The prevalence of adolescents’ sexual experience was calculated separately for boys and girls within different grade levels. Cross-tabular analyses were performed between gender/grade level and sexual behavior (i.e. number of sexual experiences, number of sexual partners and the use of contraceptives), maturation (i.e. onset of menstruation and ejaculation) and drinking style (i.e. frequency of drinking and drunkenness). Pearson’s chi-square test was applied to examine the differences between sexually experienced and non-experienced adolescents.

The study has three main outcomes: for the logistic regression models all outcomes ‘Experienced at least one sexual intercourse’, ‘Engaged in unprotected sexual intercourse’ and ‘Had three or more sexual partners’ were dichotomized (yes/no). Given that pubertal maturation may vary with gender, analyses were run separately for boys and girls. Regarding the first outcome, all adolescents were included. Subsequently, adolescents with two or more sexual experiences were included for the further logistic regression analysis, i.e. outcomes two and three. The rationale behind this decision lies in following assumptions (i) first sexual intercourse is often unplanned and thus, more often unprotected than later encounters, (ii) multiple partners is possible only if one has experienced at least two sexual intercourses.

Given that grade level and maturation were adjusted for, logistic models for all outcomes included two explanatory variables, namely frequency of alcohol drinking and drunkenness-related drinking. In the univariate model (Model 0), one explanatory variable at a time was entered into the model. In the second model (Model 1), explanatory variables were adjusted for grade and maturation age. The third model (Model 2) included all variables, which were entered in one block. Odds ratios (OR) and their 95% confidence intervals (95% CI) were calculated for each logistic model. All the analyses were carried out by using SPSS version 15.0 for Windows.

Frequency of alcohol drinking and drunkenness-related drinking were included in all logistic regression models as separate variables for a reason: consequently, sole use of alcohol (i.e. ‘alcohol drinking’) may be interpreted as an indicator of frequency of use and ‘drunkenness’ as an indicator of drinking style (i.e. drunkenness-related drinking) (Lavikainen et al., in press).

RESULTS

One-third of the girls and 27% of the boys in the ninth grade had engaged in sexual intercourse at least once (Table 1). For the eighth-graders the figures were 18 and 16%, respectively. More than half of the sexually experienced girls in the ninth grade reported engaging in sexual intercourse 10 times or more often, whereas in other sub-groups this percentage proportion varied between 34 and 41% (Table 1).

Approximately half of the sexually experienced adolescents reported only one sexual partner (Table 1). Nevertheless, 25% of the girls in the eighth grade and 34% in the ninth grade reported that they had had at least three partners. Among boys, the figures were 29 and 33%. While 80% of the adolescents reported having used contraception during their most recent intercourse, the proportion of unprotected intercourse varied between 16 and 24% in eighth and ninth grade boys/girls. Boys reported non-use of contraceptives somewhat more often than girls did (Table 1).

When comparing sexually experienced adolescents with their non-experienced peers, age at the onset of menstruation/ejaculation was lower among adolescents with sexual experience (Table 2). Similarly, sexually experienced adolescents were different from their non-experienced peers in terms of alcohol drinking: the frequency of drinking and drunkenness-related drinking was higher among adolescents with sexual experience. The majority of their non-experienced peers reported abstinence or minor alcohol use (Table 2).

First, the multiple logistic regression analysis was performed for all adolescents to examine the relationship between sexual experience and drinking style (Table 3). Higher levels of alcohol drinking and drunkenness-related drinking were related to increased likelihood of having sexual experience among both genders. Compared with adolescents abstaining from alcohol and drinking or rarely being drunk, frequent drinkers, especially those with a drunkenness-related drinking style, were more likely to engage in sexual intercourse. This association remained evident after adjusting for grade and maturation in boys and girls (OR = 14.1) (Table 3, Model 2).

Table 4 presents the results of models regarding the relationship between unprotected intercourses and drinking style among adolescents having experienced intercourse at least twice. After adjusting for grade and maturation, weekly drunkenness-related drinking among boys was strongly associated with engaging in unprotected sexual intercourses, whereas among girls both weekly alcohol drinking and frequent drunkenness-related drinking increased that likelihood (Model 2). In general, higher level of drunkenness-related drinking was associated with an increased likelihood of engaging in unprotected intercourse. Maturation prior to the age of 12 increased the likelihood of non-use of contraceptives, notably among boys. (Table 4).

Table 5 presents the results of models regarding the relationship between having had multiple sexual partners and drinking style. Among girls and boys, higher frequency of alcohol use and drunkenness-related drinking were both associated with an increased likelihood of having had three or more sexual partners. Drunkenness-related drinking was, however, associated with multiple partners more evidently than frequent alcohol drinking (Model 2). Particularly for girls, weekly drunkenness-related drinking increased that likelihood (OR = 4.21) (Table 5, Model 2).

DISCUSSION

In this large-scale study, we used school-based data collected in 2002 and 2003. The nationally representative samples comprised a total of 100 790 Finnish adolescents in both eighth and ninth grades. The main finding of this study was that drunkenness-related drinking substantially increased the likelihood of engaging in sexual risk-taking behaviors compared with alcohol drinking. In particular, recurrent drunkenness among both boys and girls was associated with engaging in unprotected sexual intercourse and having multiple sexual partners. Apparently, the association was especially strong in the context of having had multiple sexual partners.

In this study, we investigated separately three aspects of sexual risk-taking: early activity, unprotected sexual intercourse (being exposed to sexually transmitted infections and pregnancies) and having sex with multiple partners (Tapert et al., 2001; Buhi and Goodson, 2007). In the large population-based sample our results confirmed earlier findings of the association between alcohol use and the initiation of sexual intercourse (Capaldi et al., 1996; Fergusson and Lynskey, 1996; Windle, 2003). Our study also showed that this association is straightforward: the more frequently drinking occurred, the more probable it is that teenagers, both boys and girls, had experienced sexual intercourse at an early age. Moreover, this study clearly showed the importance of drunkenness-related drinking style in this context. In both genders, the risk of engaging in sexual intercourse was many-folded even if an adolescent was drunk only every now and then, less than once a month. The risk was very high (14-folded) if this happened every week.

Regarding the aspect of having had multiple sexual partners, we set the limit at three partners given that our participants were young, the mean age being <16 years. There is no unambiguous definition for what ‘multiple sex partners’ means, and it greatly depends on cultural terms. Also here, our findings are congruent with those of others, indicating that frequent drinking is significantly related to an increased number of sexual partners. Valois et al. (1999) found that alcohol use was significantly and consistently linked to an increase in the number of sexual partners for all race and gender groups. However, we did not find any difference between genders in our secondary school sample. Based on a large US high school study (Youth Risk Behavior Surveillance), boys are more likely to have multiple partners than girls (Valois et al., 1999).

Our study showed that drunkenness was associated with non-use of contraception. This refers to condom use, in the first place, because it is the most important method of contraception in early stages of sexual activity. In line with our findings, regular substance use has been linked with decreased condom use at the age of 16 (Parkers et al., 2007). The study of Parkers and others has also showed that being drunk at the time of sexual intercourse explained the non-use of condoms among adolescent substance users (Parkers et al., 2007). On the contrary, a diary study among American adolescents aged 14–19 suggests that alcohol drinking is not associated with condom non-use (Morrison et al., 2003). Inconsistent findings may result from the used measurements of alcohol use. As our study showed, both frequency of alcohol drinking and drunkenness-related drinking are needed when measuring the link between alcohol use and certain risky behavior among adolescents.

One of the study’s strengths is that it is based on a large and representative sample of the Finnish adolescent population aged 14–16 years. The coverage of secondary schools was remarkably good in both years under study (2002–2003). Our study has showed the relationship between adolescent sexual behavior and drinking style, by using two different measures for the style, i.e. alcohol drinking and drunkenness-related drinking. For the most part, studies concerning adolescent sexual health and behavior have focused on girls; we have, however, included both genders. Given that adolescents’ gender, grade level and maturation might influence both sexual behavior and drinking style, we ran analyses separately for boys and girls.

When interpreting the findings of this study, some limitations should, however, be noted. First, causality cannot be assumed from this cross-sectional survey data. Second, only Finnish data have been used. Both sexual behavior and alcohol use are culture-sensitive phenomena; thus our results illustrate the Finnish adolescent population. For instance, socio-cultural aspects regarding adolescent sexuality may vary from country to country: in the Nordic countries attitudes are more liberal (Edgardh, 2002), whereas in the United States a more restrictive sexual culture is seen (Ponton and Judice, 2005). Conversely, trends in adolescents’ alcohol use have been rather similar in Western Europe and the United States (Hibell et al., 2004; Monitoring the Future, 2006).

In addition, the questionnaire used in the SHPS study required us to employ the experience of at least two intercourses as a cut-off point in logistic regression models (outcomes two and three). While a number of studies have focused on the conditions of an adolescent’s first intercourse (Stone and Ingham, 2002; Manlove et al., 2003), we included adolescents with experience of two or more intercourses to detect particularly those with the possibility to report the used contraception and multiple partners. Finally, information concerning adolescent sexual behavior and alcohol use has been gathered by self-reports. Particularly among adolescents, it is a challenging task to assess these private behaviors. Therefore, it should be noted that surveys may be biased by inaccurate self-reports and by a reluctance to co-operate (Bagnall, 1991). Nevertheless, recent research suggests that self-reports of adolescent alcohol drinking and drunkenness may be regarded as valid (Lintonen et al., 2000b; Lintonen and Rimpelä, 2001).

CONCLUSION

To conclude, we found that the vast majority of sexually experienced teenagers drink alcohol and drinking is drunkenness-oriented for many of them. Our findings suggest that drinking style is strongly associated with adolescent sexual risk-taking behavior. Accordingly, drunkenness-related drinking style is associated not only with the increased likelihood to engage in sexual intercourse but to engage in unprotected intercourse and to have multiple partners as well. These harmful aspects warrant special attention: in particular it would be important to effectively combine both alcohol education and sexual education in school settings to reduce both adolescents’ alcohol use and various harms.

FUNDING

This work was supported by the Finnish Foundation for Alcohol Studies (H.L.), Jenny and Antti Wihuri Foundation (H.L.), and Yrjö Jahnsson Foundation (5454 to H.L.).

REFERENCES