Setting politics aside to collect cross-national data on sexual health of adolescents.

The periodic collection of data to monitor changes in the sexual health behaviors of adolescents is well established in the United States through the Youth Risk Behavior Surveillance System (YRBSS). (1,2) The YRBSS--developed by the U.S. Centers for Disease Control and Prevention (CDC)--includes periodic national, state, and local school surveys plus other non-periodic components (e.g., a survey of alternative school students). (3,4) These surveys all use a similar questionnaire, the Youth Risk Behavior Survey (YRBS). The YRBS measures six categories of behaviors: those that contribute to unintentional injuries and violence; tobacco use; alcohol and other drug use; sexual behaviors that contribute to unintended pregnancy and sexually transmitted disease, including HIV infection; dietary behaviors; and physical activity. Nearly all states have conducted a YRBS, and many do so every other year, typically in the spring of odd-numbered years. A small number of states have occasionally excluded the questions related to sexual health out of concern that their inclusion could jeopardize successful implementation of a state's YRBS.

Recognizing the difficulty that some states have faced in adopting and maintaining a core set of questions related to sexual health as part of the their state YRBS, one might anticipate a similarly high level of difficulty in getting European and other North American countries and regions to adopt a core set of questions permitting cross-national comparisons. In fact, nearly all of the 35 nations or regions (e.g., Wales, Scotland, England) participating in the World Health Organization (WHO)-affiliated Health Behaviors in School-aged Children (HBSC) survey during the 2001-2002 school year adopted four YRBS questions related to sexual health and administered them to national samples of adolescents. This article reports on these data collected as part of the wider 2001-2002 HBSC. The article was adapted from the chapter on sexual health in the HBSC International Report recently published by the WHO Regional Office for Europe. (5) Another adaptation of the sexual health chapter may be found elsewhere. (6)

BACKGROUND

Sexual health is a substantial part of adolescents' general, social, and personal well-being. (7) One of the primary developmental aspects of adolescence is the consolidation of identity in general, and sexual identity in particular. The development of sexuality in adolescence involves physical changes associated with puberty, psychological changes, and interpersonal events. Adolescents need to learn how to be comfortable with themselves, how to deal with their sexual feelings, and how to relate in a healthy way to other people.

Adolescence is both a period of opportunity, when new options and ideas are explored, and a time of vulnerability and risk. Fortunately, most adolescents emerge from these changes with positive sexual and reproductive health outcomes. Nevertheless, many of the normative behaviors associated with adolescence--spontaneity, social immaturity, risk taking, and volatility--can raise issues with respect to sexual health.

The key public health concerns related to teenage sexual health include pregnancy and sexually transmitted infections (STIs), all of which cause significant health, social, and economic problems for young people, and are largely preventable through the coordinated efforts of families, schools, health and education agencies, and community organizations. The development of effective school- and community-based programs depends in part on obtaining information about the nature and extent of sexual behavior among adolescents. Current information on issues related to young people's sexual health is urgently needed to help support development of policies and to target and evaluate programs. We also need a better understanding of the social and cultural determinants of sexual risk taking, as well as corresponding protective factors, so that interventions can both be comprehensive and effectively targeted. Unfortunately, few cross-national data have been available about the sexual health of adolescents.

THE HBSC STUDY

The HBSC study is cross-national research conducted by an international network of research teams in collaboration with the WHO Regional Office for Europe. The HBSC began in the early 1980s as a collaboration involving only a handful of European countries. The aim of the HBSC is to gain new insight into, and to increase understanding of, young people's health, well-being, health behavior, and social context. The HBSC is conducted every four years with national samples of students in their twelfth, fourteenth, and sixteenth years of life (ages 11, 13, and 15) using a standardized international core questionnaire with optional modules. When the United States applied and was accepted as an HBSC member in 1998, the HBSC had 29 country or regional members. By 2002, there were 35 members. The average number of completed questionnaires in each participating country or region is 4,500 (all three age groups combined).

One of the driving forces behind the HBSC is to provide member countries and regions with the opportunity to learn from each others' successes in addressing aspects of adolescent development or reducing adolescent risk behaviors. Specifically, the purpose of reporting cross-national results is to examine differences across countries and regions in key health behaviors, indicators of health status, or factors regarded as determinants of health behavior/status. Evidence that certain countries or regions are faring better than others leads to a series of questions focused on causes of these differences. For example: Do the differences result from deep-seated cultural factors? Do they reflect variations in family upbringing? Do public policy initiatives in countries that are doing better, or that have shown improvement, help to explain the differences? Are there coordinated health-promotion campaigns? Does the public education system make concerted efforts to address the issues that could explain differences in patterns of success? In recent years, for example, cross-national transfers in public policy strategies have occurred in relation to use of seatbelts and bike helmets, increasing awareness of and preventing bullying, and various measures for reduction of tobacco use (e.g., enforcement of sanctions for underage sales, taxes on tobacco products, and health promotion campaigns).

The 1989-1990 HBSC contained an optional module on sexual health that several countries and regions then participating in the HBSC adopted, at least on a one-time basis. This optional package focused on knowledge about, attitudes toward, and a set of behaviors related to the initial expressions of sexuality. In the 1997-1998 HBSC, fewer than half of the participating countries and regions included any questions about sexual health. The questions that were included varied widely in scope and content; however, a sufficient number of countries had adopted some broadly comparable questions that permitted an exploratory, cross-national analysis, which was included in the 1997-1998 International Report published by the WHO Regional Office for Europe. (8) The lack of comparability across sexual health questions at that time helped establish the justification for coming to agreement on a set of "mandatory" questions on sexual health.

These "mandatory" questions were included, for the first time, on the 2001-2002 HBSC survey. While the questions were "mandatory," a country or region was given the opportunity to "opt out" of them in extremis; i.e., if inclusion of these questions might prove to be a showstopper (that is, the entire survey might fail to go forward if these questions were included). To limit such circumstances, these sexual health questions were recommended for administration only to students in a country or region's population of 15-year-olds. The rationale was two-fold: younger students are unlikely to have experienced sexual intercourse, and schools would be far more likely to refuse to participate in the HBSC if younger students were asked questions about sexual health. Analogously, the YRBS is asked only of high school students who during spring administration are ordinarily 15-18 years old. In the 2001-2002 HBSC survey, of the 35 countries and regions that participated, only four did not ask any of the sexual health questions: Denmark, Ireland, Norway, and the United States. Their collective explanation for exclusion of the "mandatory" sexual health questions was that inclusion likely would have deterred school participation, with a consequent unacceptable reduction in participation rates.

The United States excluded the sexual health questions from the HBSC because they were already on the YRBS; therefore, it was believed that the U.S. Office of Management and Budget (OMB) would likely view their inclusion on the HBSC as redundant. Redundancy with other existing data collections has often been sufficient justification for OMB to deny clearance. In addition, within the United States, a strong infrastructure at state and local levels supports the biennial YRBS. However, the HBSC largely lacks such infrastructure to support state and local clearances. For the purposes of this article, 2001 national YRBS data for 15-year-olds has been added to the cross-national comparisons previously published in the WHO International Report. (9)

Questionnaire Development. The YRBS clearly identifies its purpose as monitoring trends in risk behavior. While many HBSC member countries use HBSC data to monitor risk behavior, the HBSC distinguishes itself from the YRBS and similar surveys by seeking to measure the determinants of adolescent development, including the development of relationships. Focusing on relationships rather than risk behavior would seem to lend itself to a different set of questions. However, a review of commonalities across the sexual health questions found in the 1997-1998 HBSC questionnaires of participating countries helped to identify four constructs that were most critical to incorporate into the HBSC. These four constructs already were measured on the YRBS. The YRBS questions also had undergone extensive cognitive tests, had been widely used for over a decade, and were known to produce reliable data in the United States. (10, 11) Therefore, the four following mandatory sexual health questions were adopted from the YRBS with few changes:

1. Have you ever had sexual intercourse? (Sometimes this is called "making love," "having sex," or "going all the way.") The response options were: Yes; No.

2. How old were you when you had sexual intercourse for the first time? The response options were: I have never had sexual intercourse; 11 years or younger; 12 years old; 13 years old; 14 years old; 15 years old; 16 years or older.

3. The last time you had sexual intercourse, did you or your partner use a condom? The response options were: I have never had sexual intercourse; Yes; No.

4. The last time you had sexual intercourse, what method(s) did you or your partner use to prevent pregnancy? The response options were: I have never had sexual intercourse; No method was used to prevent pregnancy; birth control pills; condoms; spermicidal spray or foam; withdrawal; national choice option; questionnaires could include additional country- or region-specific options if desired; some other method; not sure.

The first question includes parenthetical cues to help the young person understand the meaning of the term "sexual intercourse." Validity studies have shown that adolescents can accurately report whether they have engaged in sexual behavior. (12) The question does not specify that sexual intercourse is restricted to vaginal intercourse. However, most young people interpret such cues as indicating vaginal intercourse, and refrain from reporting non-vaginal sex as "sexual intercourse," thereby understating the population at risk of STI. It also should be noted that, by asking only whether young people had ever had sexual intercourse, the question did not identify those who were currently sexually active and, therefore, immediately at risk of pregnancy and STI.

The second question was designed to measure age at first sexual intercourse. This age is generally thought to be significant since those who engage in early first intercourse are thought to be at greater risk for unplanned and/or unprotected sex and, therefore, unintended pregnancy and the contraction of STIs. Moreover, early first intercourse correlates with other modes of risk taking. Alcohol and drug use have clear associations with early first intercourse, which is likely to be unintended and unprotected. (13,14,15)

The third and fourth questions were designed to measure condom and contraception use at last intercourse. Research has shown that adolescents have difficulty in summarizing their use of contraceptives, even for short time periods, because their use is not consistent. Adolescents may use condoms, contraceptive pills, or other methods sporadically, depending on the situation and the sexual partner. In addition, if asked about "typical" behavior, respondents (both adults and young people) are more likely to bias their answers by describing socially desirable behavior. Responses about the last encounter have higher reliability and validity than those on typical behavior.

In the analyses of the data reported in response to the third and fourth questions, young people who responded to either question by saying that they or their partners used a condom during the last intercourse were regarded analytically as having used a condom for dual reasons: that is, to prevent both pregnancy and transmission of STI. In addition, for these analyses, responses to the fourth question on contraceptive methods were combined to provide a summary measure of the proportion of 15-year-olds reporting use of at least one mode of contraception. The pre-coded response for withdrawal was excluded because this method offers little or no protection from pregnancy. Future analyses of the data on the use of condoms and other means of contraception will pay particular attention to national choice responses.

Exclusions of Sexual Health Data. In the 2001-2002 HBSC, 31 out of 35 participating countries and regions included one or more of the four sexual health questions in their questionnaires; as noted above, only Denmark, Ireland, Norway, and the United States did not. Of the 31 that did, data from Malta were excluded from the analyses because Malta's translation of sexual behavior as "sexual experience" rendered its data non-comparable to those of other nations. Two countries asked fewer than all four questions but their data were still included in the analyses. The Czech Republic included only the question about ever having had sexual intercourse. The Russian Federation asked only the questions about ever having had sexual intercourse and the age at first intercourse. Italy asked all four questions, but asked the contraception question differently, so Italy's data on contraception were excluded from the analyses. In two countries, less than a full national sample was asked the sexual health questions. Israel asked all of the sexual health items but, in accord with past practice, not in religious schools. Similarly, Germany used the full set of questions but only in two Lander (areas), Saxony and Berlin.

RESULTS

The four sexual health questions allow the investigation of four research questions.

* What proportion of the population has experienced sexual intercourse?

* What proportion of the sexually active population experienced early intercourse?

* How many in the sexually active population protect themselves and their partners by using condoms?

* How many in the sexually active population protect themselves and their partners against pregnancy by using some type of contraception?

As mentioned, for these analyses, responses to the question on contraceptive methods were combined to provide a summary measure.

Experience of Sexual Intercourse. Counting the Unites States, for which YRBS data have been added, 31 countries and regions included in this analysis asked 15-year-olds whether they had ever had sexual intercourse. The differences in responses are striking (Table 1). The percentages of 15-year-olds who report having had sexual intercourse range from 15% in Poland to 75% in Greenland. In nine countries and regions, mainly in eastern and central Europe, plus Spain, fewer than a fifth of young people report ever having had sexual intercourse. At the upper end of the spectrum, in England, Greenland, Scotland, Ukraine, the United States, and Wales, a third or more have had sexual intercourse.

The gender differences in having ever experienced sexual intercourse are wide. Among boys, positive responses range from 18% in Spain to 71% in Greenland. The Czech Republic, Estonia, Poland, and Spain cluster at the low end with rates of about 20%. At the opposite end of the spectrum, in nine countries and regions, about a third or more of boys have had sexual intercourse. Among girls, positive responses range from 4% in the former Yugoslav Republic of Macedonia to 79% in Greenland. Rates are below 20% in 15 countries and regions but about 33% or more in 6 others.

Interestingly, in the same six countries or regions, more girls than boys declared having had sexual intercourse. The largest differences are found in Germany (23% for boys vs. 34% for girls) and Wales (29% for boys vs. 40% for girls). In eight countries, a more traditional pattern prevails with at least twice as many boys as girls having had sexual intercourse. More than three times as many boys as girls gave positive answers in Greece and Israel, and more than ten times as many in the former Yugoslav Republic of Macedonia (38% for boys vs. 4% for girls).

Age at First Intercourse. Across all countries and regions, the mean age of first intercourse among 15-year-olds who reported ever having sexual intercourse was 14.3 years for girls and 14.0 for boys. The mean age ranges from 13.5 years in Lithuania to 14.6 years old in Ukraine. For girls, the mean age of first intercourse ranges from 13.6 years old in Lithuania to 14.9 years old in Ukraine. For boys, it ranges from 13.5 years old to 14.5 years old in the same two countries. In most countries and regions, it is slightly lower among boys than girls. The greatest gender difference, about one year, is found in Portugal (13.7 years for boys vs. 4.8 years for girls) (Table 2).

Use of condoms. The proportion of sexually active young people who report that a condom was used the last time they had sexual intercourse ranges from 64% in Finland (closely followed by Sweden and the United States) to 89% in Greece and Spain (Table 3). In seven countries or regions, no more than 70% of sexually active young people reported using a condom. In seven others, 80% to nearly 90% report condom use in their last encounter.

In all countries and regions except three (Portugal, Switzerland and England), boys are more likely than girls to report condom use the last time they had sexual intercourse. The gender difference can sometimes be quite large, as in Flemish-speaking Belgium (81% boys vs. 60% girls) and Ukraine (84% boys vs. 60% girls). The proportions reporting condom use ranges from 69% in Portugal to 91% in Greece for boys and from 58% in Sweden to 89% in Spain for girls.

Use of contraception. The proportion of sexually active young people reporting the use of at least one method of contraception (including but not limited to condoms and birth control pills) during their most recent intercourse ranges from 73% in Poland to 95% in the Netherlands (Table 4). Proportions are below 80% in seven countries or regions and at or above 90% in eight others.

Among boys, the proportion reporting use of contraception when they last had sexual intercourse ranges from 73% in Poland to 92% in the Netherlands. Among girls, the proportion ranges from 68% in Ukraine to 97% in the Netherlands. The countries and regions are almost evenly split as to whether boys or girls have a higher rate of contraception use; in many, the rates are nearly identical. Boys are far more likely than girls to use contraceptives in Greece, Hungary, Israel, and Ukraine. Girls are more likely to use contraceptives in England, Germany, Greenland, Portugal, and Switzerland.

DISCUSSION

The responses to the four questions relating to sexual health demonstrate noteworthy differences across the HBSC countries and regions in the proportions of 15-year-olds having had sexual intercourse, the mean age at first intercourse, the use of condoms to prevent STI, and the use of contraceptives during the most recent intercourse. Cross-national differences undoubtedly reflect fundamental cultural, social, religious, and educational differences across countries, as well as differences in public policy. The most important findings demonstrate variations across countries and regions in the use of condoms. While no more than 70% of sexually active young people used a condom the last time they had sexual intercourse in seven countries (including the United States), 80% to 90% of sexually active young people did so in seven others. These findings have important policy implications. In the context of the HBSC, however, further analysis will give an opportunity to explore the determinants of condom use within and across countries in relation to other risk behaviors (especially drug and alcohol use), onset of puberty, school and community bonding, school performance and parental relations.

Examination of the gender differences shows that, in many countries and regions, traditional expectations tied to gender are eroding. For example, while boys are twice as likely as girls to have experienced sexual intercourse in nearly a third of HBSC countries and regions, the genders are almost equal in this experience in many more, and girls are more likely than boys to have experienced intercourse in six (England, Finland, Germany, Greenland, Scotland and Wales). In almost all countries and regions, boys are more likely than girls to report that a condom was used during their last intercourse. The gender difference can sometimes be quite large, as in Flemish-speaking Belgium and Ukraine. These gender discrepancies raise complex questions related to cultural context, family upbringing, public policy, and the content of health education programs.

The HBSC study is not the ideal means of providing a complete picture of age at initiation of sexual activity because even the oldest participants are only in their sixteenth year of life, when the majority of young people have not yet started to be sexually active. Nevertheless, the population identified as sexually active in the study consists largely of early initiators who by definition are seen to be at higher risk of unplanned and/or unprotected intercourse and other risk behaviors associated with impulsiveness. On the other hand, with some noteworthy exceptions, a high percentage of these early initiators in many countries and regions report using condoms. This suggests that young people not only have received the various messages on "safer sex" but also seem largely to have accepted and acted on them.

Further analyses of contraceptive methods are planned, with the object of understanding of the differences in specific contraceptive practices across HBSC countries and regions, including "national choice" options, and to develop an efficacious, age-appropriate measure of these practices. Further analysis will also explore the differences in risk and protective factors connected with sexual behavior among individuals and across countries. Having a clearer picture of effective and ineffective contraceptive options that adolescents may be using in particular countries will contribute to a growing cross-national understanding of age-appropriate practices.

TABLE 1. PROPORTION REPORTING TO HAVE HAD SEXUAL INTERCOURSE:
15-YEAR-OLDS

                    BOYS  GIRLS

Greenland           70.8  78.8
England             35.7  40.4
Ukraine             47.2  24.0
Wales               28.7  40.1
Scotland            32.9  34.6
USA                 36.5  30.6
Belgium-French      34.4  23.2
Finland             23.0  33.1
Sweden              25.3  30.9
Germany             22.5  33.5
Russian Federation  40.9  16.4
Slovenia            30.8  21.6
Belgium-Flemish     26.3  23.7
Portugal            30.2  20.3
Canada              24.7  24.0
Italy               27.2  20.5
Netherlands         24.2  21.6
Switzerland         25.1  20.6
France              26.1  18.3
Greece              33.6   9.6
Austria             22.1  19.1
Israel              32.4   9.7
Hungary             25.5  16.4
TFYR Macedonia      37.4   3.6
Lithuania           26.4  10.8
Czech Republic      19.4  17.2
Estonia             20.1  15.8
Latvia              21.8  14.1
Spain               18.0  14.8
Croatia             23.2   9.7
Poland              20.9   9.2

TABLE 2. MEAN AGE OF FIRST SEXUAL INTERCOURSE: 15-YEAR-OLDS WHO REPORT
HAVING HAD INTERCOURSE

                    BOYS  GIRLS

Ukraine             14.5  14.9
Greece              14.3  14.6
Israel              14.3  14.7
Italy               14.4  14.5
Russian Federation  14.2  14.7
Spain               14.2  14.5
Czech Republic      14.2  14.4
Latvia              14.2  14.4
Poland              14.1  14.7
Wales               14.3  14.3
Croatia             14.0  14.5
Estonia             14.1  14.3
Greenland           14.3  14.2
Hungary             14.1  14.4
Portugal            13.7  14.8
Slovenia            14.1  14.4
Switzerland         14.0  14.4
Belgium-Flemish     14.0  14.2
England             14.0  14.1
Finland             14.1  14.2
Canada              13.9  14.2
Germany             13.7  14.2
TFYR Macedonia      13.9  14.6
Scotland            14.0  14.0
Netherlands         13.8  14.0
USA                 13.7  14.1
Belgium-French      13.7  14.0
Sweden              13.7  13.8
France              13.5  13.9
Austria             13.5  13.6
Lithuania           13.5  13.6

TABLE 3. CONDOM USE LAST INTERCOURSE: 15-YEAR-OLDS

                 BOYS  GIRLS

Greece           91.2  82.5
Spain            89.1  89.1
Israel           89.3  73.6
TFYR Macedonia   85.1  84.0
France           87.0  77.0
Austria          84.8  78.9
Switzerland      78.3  83.0
Latvia           81.0  77.3
Lithuania        82.2  70.4
Hungary          84.5  71.8
Netherlands      83.3  72.4
Greenland        77.9  74.5
Ukraine          83.7  59.2
Canada           79.8  71.7
Slovenia         80.4  67.6
Croatia          75.2  73.1
Estonia          75.9  70.5
Poland           73.4  72.5
Belgium-French   78.1  64.8
Portugal         68.5  77.8
Belgium-Flemish  81.3  59.7
England          69.6  70.8
Scotland         76.2  63.4
Germany          75.6  64.4
Wales            75.2  63.6
USA              69.0  60.2
Sweden           72.9  57.6
Finland          72.6  58.6

TABLE 4. USE OF CONTRACEPTIVES LAST SEXUAL INTERCOURSE: 15-YEAR-OLDS

                 1 OR MORE METHODS USED
                 BOYS  GIRLS

Netherlands      92.4  97.0
France           92.1  92.5
Germany          87.8  94.9
Austria          90.2  93.0
Sweden           92.2  90.5
Switzerland      87.3  95.2
Spain            89.8  90.6
Belgium-Flemish  90.5  89.6
Greenland        86.8  91.2
Greece           91.2  82.5
Israel           91.7  77.4
TFYR Macedonia   88.2  84.0
Finland          88.2  86.0
Slovenia         89.2  83.3
Lithuania        88.1  81.6
Canada           86.8  85.5
USA              89.3  83.0
Latvia           86.9  84.0
England          80.4  87.5
Wales            82.4  84.8
Belgium-French   82.2  81.5
Hungary          84.5  72.5
Estonia          79.3  77.3
Portugal         74.8  82.7
Ukraine          85.0  62.5
Scotland         81.2  73.8
Croatia          76.7  74.6
Poland           73.4  72.5

The full report and further information about HBSC is available at www.hbsc.org.

References

1. L. Morris, C. W. Warren, and S. O. Aral, "Measuring adolescent sexual behaviors and related health outcomes," Public Health Reports, Supplement 1, 108 (1993): 31-36.

2. Centers for Disease Control and Prevention, "Trends in sexual risk behaviors among high school students--United States, 1991-2001," Morbidity and Mortality Weekly Report 51 (2002): 856-859.

3. L. J. Kolbe, L. Kann, and J. L. Collins, "Overview of the Youth Risk Behavior Surveillance System," Public Health Reports, Supplement 1, 108 (1993):2-10.

4. N. D. Brener, L. Kann, et. al., "Methodology of the Youth Risk Behavior Surveillance System," Morbidity and Mortality Weekly Report 53, RR-12 (2004): 1-16.

5. J. Ross, E. Godeau, and S. Dias, "Sexual Health," in Young People's Health in Context. Health Behaviour in School-Aged Children: International Report from the 2001/2002 Survey, eds. C. Currie, et al., Health Policy for Children and Adolescents, no. 4 (Copenhagen: WHO Regional Office for Europe, 2004), 153-160.

6. J. Ross, E. Godeau, S. Dias, "Sexual health in young people--Findings from the HBSC study," Entre Nous 58 (2004): 20-23.

7. D. Raphael, "Determinants of health of North-American adolescents: evolving definitions, recent findings, and proposed research agendas," Journal of Adolescent Health 19 (1996): 6-16.

8. J. Ross and W. Wyatt, "Sexual behaviour," in Health and Health Behaviour among Young People, eds. C. Currie, et. al., Health Policy for Children and Adolescents, no. 1 (Copenhagen: WHO Regional Office for Europe, 2000), 115-120, accessed at <http://www.who.dk/document/e67880.pdf>.

9. J. Grunbaum, et al., "Youth Risk Behavior Surveillance--United States, 2001," Morbidity and Mortality Weekly Report 51.SS-4 (2002): 1-64.

10. N. D. Brener, et al., "Reliability of the Youth Risk Behavior Survey Questionnaire," American Journal of Epidemiology 141.6 (1995): 575-80.

11. N. D. Brener, et al., "Reliability of the 1999 Youth Risk Behavior Survey questionnaire," Journal of Adolescent Health 31 (2002): 336-342.

12. D. P. Orr, J. D. Fortenberry, M. Blythe, "Validity of self-reported sexual behaviors in adolescent women using biomarker outcomes," Sexually Transmitted Diseases 24 (1997): 261-266.

13. B. J. Flanigan and M. A. Hitch, "Alcohol use, sexual intercourse and contraception: an exploratory study," Journal of Alcohol and Drug Education 31 (1986): 6-40.

14. J. A. Robertson and M. A. Plant, "Alcohol, sex and risks of HIV infection," Drug and Alcohol Dependence 22.1 (1988): 75-78.

15. B. Traeen and I. Lundin Kvalem. "Sex under the influence of alcohol among Norwegian adolescents," Addiction 9.7 (1996): 995-1006.

James Ross 

[e-mail]

Macro International Inc., Calverton, Maryland 20705

Emmanuelle Godeau

Service medical du Rectorat

Association pour le developpement

d'HBSC and Inserm U558

Toulouse, France

Sonia Dias

Centro de Malaria e Outras Doencas Tropicais, Instituto Higiene e Medicina

Tropical, Universidade Nova de Lisboa

Lisbon, Portugal

Celine Vignes

Service medical du Rectorat and Inserm U558

Toulouse, France

Lori Gross

Macro International Inc.

Calverton, Maryland 20705