Introduction
An understanding of ways in which sexual behaviour changes over time is essential to efforts to improve the design of HIV prevention policies. Trend data enable us to identify changes in the optimal focus of preventive interventions, to assess the implications of the changing burden of sexual ill health for the planning of sexual health services and to explore the relative contribution of public health policies and secular trends to changes in sexual behaviour. They are as essential to the design of public health policies as are regional or cross-national comparisons.
Descriptions of changes over time are, however, considerably rarer [1–3] than descriptions of cross-cultural differences [4–11]. Some longitudinal studies have been undertaken among subgroups of the general population, especially men who have sex with men [12,13], but they do not provide information on trends among the general population nor on the social context of sexuality. To a large extent, this is a consequence of the lack of reliable data. In the absence of comparable surveys, data are those from studies using different research tools and samples and it is unclear whether observed changes are real or artefactual [14]. Few countries have the luxury of large cohort studies or repeat cross-sectional surveys investigating sexual behaviour. France appears unique in this respect, in having three large national representative sexuality surveys with high response rates conducted in 1970, 1992 and 2006 [15–17].
The data permit a retrospective analysis of trends in sexual behaviours over almost three quarters of a century. This paper uses data from these three surveys to describe changing patterns of sexual behaviour, the factors influencing them and their implications for HIV/AIDS prevention policies. At a time when prevention orientations based on a restriction of sexual activity such as delaying age at first intercourse or reducing the number of sexual partners are being put forward [18], historical trends provide a unique opportunity to analyse the conditions in which changes in sexual behaviour occur and to bring new results to a major scientific and public health debate.
Methods
Study population
Table 1 presents the main methodological characteristics of the three French national sexual behaviour surveys. All were based on random probability samples. In the first, in 1970, on sexual behaviour and contraception practices, 2625 participants aged 20 and older [15] were interviewed face to face. Questions on sociodemographic characteristics and on family lifestyle and contraceptive practices were included in the face-to-face questionnaire, whereas those on sexual practices were in the self-administered part of the questionnaire. In the second survey on sexual behaviour and HIV prevention undertaken in 1992, 20 055 participants aged 18–69 years old were interviewed by telephone. The sample size was much larger than in the previous survey, data on homosexual practices being needed to study at-risk sexual practices among a smaller population subgroup of men having sex with men [16]. In the third survey on the social context of sexuality carried out in 2006, the focus shifted to the study of the sexual practices of people having at least two sexual partners in the last 12 months and the sample size was reduced accordingly. In all, 12 364 participants aged 18–69 years were interviewed by telephone, younger adults being given higher probability to be included in the sample [17].
;)
Table 1: Methodological characteristics of the three national surveys of sexual behaviour in France.
Measures
We compare data from responses to questions that were common to all three surveys: age at first intercourse, protection at first intercourse, numbers of sexual partners in a lifetime and in 5 years, experience of sexual practices – masturbation, oral sex, anal sex – in a lifetime, frequency of sex and importance of sexual intercourse.
As regards age at first sex and condom use at first sex, which are presented in Figs 1 and 2, the first survey conducted in 1970 (investigating birth cohorts 1920–1950 who had their first sexual intercourse at around 18–20 years of age) allows us to go back further to the end of 1930s till the 1960s; the CSF survey conducted in 2006 among people aged 18–69 years old (birth cohorts 1937–1988 who had their first sexual intercourse at around 15–17 years of age) permit to cover first sexual intercourse that occurred between the 1940s and 2000s).
Fig. 1: Median age at first sexual intercourse, by sex and age cohort.
Fig. 2: Use of contraception (all methods) and condom use at first sexual intercourse by sex and year of first sexual intercourse. Data from the 1970 survey are plotted for first sexual intercourse occurring between 1939 and 1965. Data from the 2006 survey for first sexual intercourse occurring between 1955 and 2005.
We used published data for the 1970 survey [15], as the database no longer exists. Thus, it was not possible to carry out test for trends on these data. We maintained consistency in terms of age group in order to provide comparative data.
Statistical analysis
Differences in sampling design and response rates between the three surveys were taken into account by weighting the data to adjust for the unequal probabilities of selection – that is, the number of eligible persons in the household and age of the respondent. Then, to correct for differences in sex, age, level of education and region between the achieved sample and population estimates, a nonresponse poststratification weight has been applied [19]. The composition of the final samples corresponds well to that of the French population as described in contemporaneous national censuses.
In all analyses, we used weighted observations. The total numbers reported in the tables are unweighted, that is, the number of people who answered the questions. All the percentages are weighted.
Confidence intervals and statistical tests were performed according to the complex sampling design, with the survey module of the STATA software version 10, which incorporates the weighting, clustering and stratification of the data (Stata Corporation, College Station, Texas, USA).
Results
Sexual debut and protection of sex
Median age at first sexual intercourse fell for both men and women in the seven decades covered by the surveys (Fig. 1). For women, the years of most marked decline were those during World War II. Median age fell by almost 2 full years during the 1940s, from 22 to just over 20, stabilized in the 1950s and resumed its fall in the 1960s and the 1970s. From 1939 to 1978, median age at first sexual intercourse fell by 3.5 years, whereas from 1979 to 2005, the decline was barely a year. The decline in age at first sexual intercourse over the comparable period was considerably more modest for men, being barely a year. As a consequence, the gap between men and women in terms of age at onset of sexual activity declined from 4 years (22.0 for women vs. 18.1 for men) in 1939 to 4 months (17.6 for women vs. 17.2 for men) in 2005 onwards.
The data show a marked increase in protection at first sexual intercourse (Fig. 2). The proportion of women reporting use of any contraceptive method rose steadily over the period, from 37% in the early 1940s to 92.9% in 2005 (P < 0.001). Among men, the rise was only slightly less marked (P < 0.001). Condom use at first sex was reported by increasing proportions of both women (P < 0.001) and men (P < 0.001) from the early 1940s onward but rose most strikingly during the 1980s and early 1990s (Fig. 2). By the mid-1990s, more than 80% of men and women reported condom use at first intercourse, and the proportion remained at this level during the subsequent decade.
Sexual partnerships
Numbers of partners reported in the lifetime remained stable between all three surveys for men of all ages (11.8 in 1970, 11.0 in 1992 and 11.6 in 2006). For women, mean lifetime number of partners reported increased from 1.8 in 1970 to 3.3 in 1992 and to 4.4 in 2006 (Table 2). The gap between men and women decreased over the same period but remained sizeable and statistically significant (11.6 vs. 4.4, P < 0.001). Trends over time differed between age groups and between men and women. For women aged over 30 at the time of interview, mean number of lifetime sexual partners increased between 1970 and 2006 but remained stable among men of the same age. The mean number of lifetime partners reported by women aged under 30 at the time of interview was higher in 1992 than in 1970, but no increase was observed between 1992 and 2006. For men under 30, a decrease in the mean was seen in the most recent period (10.4 in 1992 and 7.7 in 2006, P < 0.0001).
Table 2: Mean number of partners, during lifetime, by sex and age (95% confidence interval).
Sexual practices
Reported frequency of sexual intercourse changed little between the three surveys; number of occasions of sex varied narrowly around a mean of 10 per month for all age groups and for both men and women (Table 3).
Table 3: Sexual practices and social norms by sex and age: 1970, 1992 and 2006 [percentage (95% confidence interval)].
The proportion of men and women reporting ever-experience of sexual activities other than vaginal intercourse, for example, masturbation, fellatio and cunnilingus, and anal intercourse, increased significantly between the surveys. The increase was more marked for women than men and for respondents aged over 30 than for those aged under 30 (Table 3). The lifetime prevalence of reported masturbation among women aged 20–49 in 2006 was more than three times as high as that in the same age group in 1970 (62% compared with 19%). For men, the corresponding increase was from 71 to 92%. Lifetime prevalence of oral sex and anal sex among women increased from 51 to 91% and from 15 to 41%, respectively, between 1970 and 2006. For men, the increase was from 55 to 94% and from 19 to 51%, respectively.
There were significant changes between 1970 and 2006 in views on the importance of sexual relations to well being. The proportion of respondents who agreed with the statement ‘Sexual intercourse is essential to feeling good about oneself’ increased from 48 to 59.5% among women and from 55 to 68.8% among men.
Discussion
These data, from serially conducted surveys in France, show dramatic changes in sexual behaviour in the last seven decades, especially among women. Compared with those becoming sexually active in the 1930s, young women today begin sexual activity 4 years earlier, more commonly report multiple sexual partnerships and have a more diverse sexual repertoire. Trends are less marked for men, for whom median age at sexual debut has fallen by only 1 year, number of reported sexual partners shows no clear trend and there has been a more modest expansion in the range of sexual practices engaged in. As a result, although sex differences in sexual behaviour remain significant, the gap between men's and women's sexual behaviour has narrowed. During this same period, there was a remarkable consistency across all three surveys in reported frequency of vaginal intercourse in the past 4 weeks. Conversely, the practice of safer sex has increased markedly for both men and women. There has also been a significant increase in the perceived importance of sexual activity to well being.
The challenges of carrying out comparative analyses of sexual behaviour, well documented in cross-national comparisons [8], apply equally in comparisons through time. Reporting of sexual behaviour is particularly susceptible to contextual influences [6,7,14]. Observed changes over time may result from changes in the values attached to particular sexual activities, in the meanings of terms used and in the strength of social sanctions governing their disclosure. A strength of the French surveys, however, has been that questions were asked in the same way in all three surveys. Response rates varied from 74.6 to 81.4% and the distortion in the structure of the samples that may result from refusals has been taken into account by poststratifying the three samples with the same methodology [19]. We can also underline that results on age at first sex and condom use and numbers of lifetime partners are similar to those in other national surveys on AIDS conducted in 1992 and 2004 [20] and sexual behaviour of adolescents conducted in 1995 [21]. Moreover, as shown in Fig. 2, the answers on condom use at first sex of those aged 69 years old in 2006 are remarkably similar to those provided by men and women aged 20 years old who were interviewed in 1970s survey. Furthermore, although most studies carried out elsewhere in Europe in the 1970s were either not repeated [22] or were carried out among specific subgroups of the population, such as married women [23], teenagers [24], students [1] or health service users [25], all three French surveys were based on general population samples. Only Finland has national data over a comparable period (1971, 1992 and 1999), the last one being a postal survey with a low response rate [3].
A considerable challenge lies in the interpretation of these data. Norms relating to sexual behaviour respond to both secular and public health influences, yet the two are difficult to distinguish. The trends observed in this analysis can be seen to have occurred in a rapidly changing social and public health landscape. The second half of the twentieth century in France, as in other European countries, saw increasing sexual autonomy and spending power among young people; liberalising legislation surrounding sexual issues such as contraception and abortion; increasing availability of medical contraception; postponement of childbearing and reduction in the number of marriages; a trend toward sex equality; and the advent of the HIV/AIDS epidemic.
Given the multiplicity and synchronicity of these influences on sexual behaviour, inferences in terms of attribution to either public health interventions or features of the broader social context are difficult to draw from observed trend data. Nevertheless, the chronological sequence of some events permits hypotheses to be ventured. The historical coincidence of the upturn in condom use and the advent of the HIV/AIDS epidemic in the late 1980s and early 1990s, for example, is striking and needs to be seen in the context not only of increased public concern about HIV but also that of the focussed HIV/AIDS public education campaigns, with a focus on routinization of condom use together with vigorous efforts to rid the condom of its strong associations with illicit sex in France [26]. It is nevertheless notable that the momentum of increased condom use continued after the advent of HAART (1996 in France), despite the decreasing numbers of prevention campaigns [27]. A possible explanation being that condom use at first intercourse has indeed been normalized in France, which is an important public health result because condom use at first sex has been shown to be a reliable indicator of preventive practices in adult sexual life [8,17].
In France, no intervention aiming at delaying the age at first intercourse or to reduce the number of sexual partners has ever been implemented. The decrease in the age of women at first sexual intercourse in the post–World War II period went together with a trend toward earlier marriage. It meant little change in premarital sexual behaviour. By contrast, the decrease in the 1960s and the 1970s, which occurred in a context of stability of age at first union, and prior to the widespread availability of medical contraception in France (from the mid-1970s), meant a steep rise in premarital sexual activity, independently of the dissemination of medical contraception [28]. It is worth underlining that this decrease had started before the advent of May 1968, which is generally considered in French social history to be a starting point in terms of family change and relaxation of sexual mores [29]. Furthermore, there has been a steep decline in religious practice over time and the strong relation between religiosity and a restricted sexual repertoire that used to prevail in the 1970s is no longer observed in the 2000s [17].
Similarly, the increasing number of sexual partners reported by women and the observed sex convergence in sexual behaviour need to be seen against a backcloth of secular rather than public health trends – the increased social autonomy of women in other spheres of life, their increasing participation in the labour force and the postponement of childbearing (average age at first birth was 28 for women in France in 2007) [30].
These findings have important implications for the design of strategies to prevent HIV/AIDS transmission. Public health interventions have more chance of being effective where their messages are in harmony with existing trends than where they run counter to them. Prevention strategies have tended to polarize into those focusing on risk avoidance on the one hand and risk reduction on the other. In the first category are interventions promoting abstinence and partner reduction [31], in the second are harm limitation strategies such as contraception and prophylaxis. Pleas have been made recently for public health messaging to embrace delaying sexual initiation and reducing the number of sexual partners to prevent further HIV transmission [18,32] although recent data do not support such perspective [33,34]. The evidence from these French data moreover is that, in general, messages to use a condom appear to have had a great purchase among young people and this is consistent with evidence from other affluent countries [9,26].
Similarly, a consequence of the increasing importance placed on sexual activity by French men and women for their well being, pleas to abstain are likely to be less favourably received. Prevention strategies also need take account of shifts in gender relations over time. A consequence of the convergence in the behaviour of men and women in terms of sexual activity is that women may no longer be seen as the ‘guardians of sexual morality’.
There are also trends here that may be currently underexploited in preventive messages. The increase in the diversity of the repertoire of sexual activities, including nonpenetrative practices, could be harnessed more robustly to preventive messages. Messages relating to nonpenetrative sex have rarely featured in public health messaging, and where they have, the reaction has been one of concern rather than support [35]. Yet, the evidence is that benefits of nonpenetrative sexual practices to public health may be as significant as those of condom use [36].
The monitoring of changes in sexual behaviours is important to designing prevention policies in sexual and reproductive health. Historical trends provide important insights into the design of public health policies and interventions. The challenge for sexual health interventions is to create synergy with changes in the broader social context, to harness existing shifts to preventive efforts and to exploit emerging trends to public health advantage.
Acknowledgements
The surveys in 1992 and 2006 were funded by the French National Agency of AIDS Research (ANRS). The Fondation de France, the Institut National de la Prevention et d'Education pour la Sante and the Direction de la Recherche, de l'Evaluation, des Etudes et des Statistiques (DREES) also contributed to the funding.
The two last studies were approved by the French ‘Commission Nationale Informatiques et Libertes’, which has examined ethics and anonymity issues since 1975.
Authors contribution: N.B. and M.B. were the principal investigators of the 2006 survey on sexual behaviour in France; they conceived and wrote the paper. N.Beltzer was the scientific coordinator of the 2006 survey; she commented on the first draft of the paper and approved the final version, as did all the other members of the research team (A.A., M.F., V.G., A.L., C.LeV., H.L., S.L., N.R., L.T., J.W.); C.L. carried out the statistical analysis. K.W. assisted in the drafting of the paper.
References
1. DeBuono B, Zinner S, Daamen M, McCormack W. Sexual behavior of college women in 1975, 1986, and 1989. N Engl J Med 1990; 332:821–825.
2. Turner C, Danella R, Rogers S.
Sexual behaviour in the United States, 1930–1990: trends and methodological problems.
Sex Transmit Dis 1995;
22:173–190.
3. Haavio-Mannila E, Kontula O. Single and double sexual standards in Finland, Estonia and St Petersburgh. J Sex Res 2003; 40:36–49.
4. Bochow M, Chiarotti F, Davies P, Dubois-Arber F, Dür W, Fouchard J,
et al. Sexual behaviour of gay and bisexual men in eight European countries. AIDS Care 1994; 6:533–549.
5. Bajos N, Wadsworth J, Ducot B, Johnson AM, Le Pont F, Wellings K,
et al. Sexual behavior and HIV epidemiology: comparative analysis in France and Britain. AIDS 1995; 9:735–743.
6. Cleland J, Ferry B. Sexual behaviour and AIDS in the developing world. London: Taylor & Francis; 1995.
7. Hubert M, Bajos N, Sandfort T. Sexual behavior and HIV/AIDS in Europe. Comparisons of national surveys. London: UCL Press; 1998.
8. Wellings K, Collumbien M, Slaymaker E, Singh S, Hodges Z, Patel D, Bajos N. Sexual behaviour in context: a global perspective. Lancet 2006; 11:1706–1728.
9. Underhill K, Montgomery P, Operario D. Sexual abstinence only programmes to prevent HIV infection in high income countries: systematic review. BMJ 2007; 335:248–260.
10. Springer A, Kelder S, Orpinas P, Baumler E. A cross-national comparison of youth risk behaviors in Latino secondary school students living in El Salvador and the USA. Ethn Health 2007; 12:69–88.
11. Goodwin R, Kozlova A, Nizharadze G, Polyakova G. High-risk behaviors and beliefs and knowledge about HIV transmission among school and shelter children in Eastern Europe. Sex Transm Dis 2004; 31:670–675.
12. Stolte IG, Dukers NH, Geskus RB, Coutinho RA, de Wit JB. Homosexual men change to risky sex when perceiving less threat of HIV/AIDS since availability of highly active antiretroviral therapy: a longitudinal study. AIDS 2004; 18:303–309.
13. Prestage G, Van de Ven P, Mao L, Grulich A, Kippax S, Kaldor J. Contexts for last occasions of unprotected anal intercourse among HIV-negative gay men in Sydney: the health in men cohort. AIDS Care 2005; 17:23–32.
14. Spira A, Bajos N, Giami A, Michael S. Cross-national comparisons of sexual behavior surveys. Methodology difficulties and lessons for prevention. Am J Public Health 1998; 88:730–731.
15. Simon P, Gondonneau J, Mironer L, Dourlen-Rollier A-M.
Rapport sur le comportement sexuel des Français [Sexual behaviour in France.]. Paris: Pierre Charon et René Julliard; 1972.
16. ACSF Group.
AIDS and sexual behaviour in France.
Nature 1992;
360:407–409.
17. Bajos N, Bozon M.
Enquête sur la sexualité en France. Pratiques, Genre et Santé. [Sexuality in France. Practices, Gender and Health.] Paris: La Découverte; 2008. See also
Sexuality in France. Practices, gender and health. Oxford: Batwell Press; 2010, in press.
18. Coates TJ, Richter L, Caceres C. Behavioural strategies to reduce HIV transmission: how to make them work better. Lancet 2008; 6:36–51.
19. Warszawski J, Messiah A, Lellouch J, Meyer L, Deville JC. Estimating means and percentages in a complex sampling survey: application to a French national survey on sexual behaviour (ACSF). Stat Med 1997; 16:397–423.
20. Beltzer N, Lagarde M, Wu Zhou X, Vongmany N, Gremy I.
Les connaissances, attitudes, croyances et comportements face au VIH/sida: evolutions 1992, 1994, 1998, 2001, 2004. [Knowledge, attitudes, beliefs and behaviours regarding HIV/AIDS. Trends 1992, 1994, 1998, 2001, 2004.]. Paris: Observatoire Regional de Sante d'Ile de France; 2005.
21. Lagrange H, Lhomond B.
L'entree dans la sexualite en France [Sexual debut in France.]. Paris: La Decouverte, Paris; 1997.
22. Zetterberg H.
Sexual life in Sweden. New Brunswick: Transaction Publishers; 2002 [first edition 1967].
23. Chesser E. The sexual, marital and family relationships of the English woman. London: Hutchinson's Medical Publications; 1956.
24. Schofield M. The sexual behaviour of young people. Revised ed. Harmonds Worth, Middlesex: Penguin Books; 1968.
25. McCormack WM, Alpert S, McComb DE, Nichols RL, Semine DZ, Zinner SH. Fifteen month follow-up study of women infected with
Chlamydiae trachomatis. N Engl J Med 1979; 300:123–125.
26. Wellings K, Field B. Stopping AIDS. AIDS/HIV public education and the mass media in Europe. London: Longman; 1996. pp. 38–42.
27. Gremy I, Beltzer N. HIV risk and condom use in the adult heterosexual population in France between 1992 and 2001: return to the starting point? AIDS 2004; 18:805–809.
28. Prioux F. Age at first union in France: a two-stage process of change. Population-E 2003; 58:559–578.
29. Artières P, Zancharini-Fournel M, 68.
Une histoire collective [Mai 68: a collective story.]. Paris: La Decouverte; 2008.
30. Van de Kaa D. Europe's second demographic transition. Populat Bull 1987; 42:1–59.
31. Barnett T, Parkhurst J. HIV/AIDS: sex, abstinence, and behaviour change. Lancet Infect Dis 2005; 5:590–593.
32. Potts M, Halperin D, Kirby D, Swidler A, Marseille E, Klausner J,
et al. Reassessing HIV prevention. Science 2008; 320:749–750.
33. Jemmott JB, Jemmott LS, Fong GT. Abstinence and safer sex HIV risk-reduction interventions for African American adolescents: a randomized controlled trial. JAMA 1998; 19:1529–1536.
34. Santelli J, Ott MA, Lyon M, Rogers J, Summers D, Schleifer R. Abstinence and abstinence-only education. A review of US policies and programs. J Adolesc Health 2006; 38:72–81.
35.
Warning: teenagers may view noncoital sex as a safe option.
Contracept Technol Update 2005;
26:4.
36. Donovan B, Ross MW. Preventing HIV: determinants of sexual behaviour. Lancet 2000; 355:1897–1901.
