First Heterosexual Intercourse in the United Kingdom: A Review of the Literature
Hawes, Zoe Coline, Wellings, Kaye, Stephenson, Judith, The Journal of Sex Research
Although a wealth of data is available on the sexual behavior of young people in the United Kingdom, there has been no recent review of the literature specifically relating to first sexual intercourse. The event of first sexual intercourse alone is generally not the most reliable indicator of onset of sexual activity. The sexual behavior of young people involves a variety of practices which do not necessarily culminate in intercourse (Henderson et al., 2002). Nevertheless the occasion of first sexual intercourse remains an event of immense social and personal significance (Mitchell & Wellings, 1998). The status of virginity, which is still of great cultural and legal importance, is technically defined in terms of experience of sexual intercourse. The event has major health implications because it marks initiation into the sexual act which, if unprotected, carries the highest risk of such adverse outcomes as unplanned pregnancy and sexually transmitted infection.
Teenage pregnancy rates in England are among the highest in the Western world (Social Exclusion Unit, 1999; Wellings et al., 2001). The promotion of condoms and youth-friendly health services, which has formed the core approach of the Teenage Pregnancy Strategy (TPSE; Social Exclusion Unit, 1999), has had limited success in reducing conceptions of those under 18 years old (The Office of National Statistics, 2009). In addition, many young British people regret their first heterosexual experiences especially when it occurs in the early teenage years (Wellings et al., 2001). There may be some value in considering the wider context within which first sex takes place to ensure a positive experience both during and after the event.
Recent decades have seen declines in the average age at first sex (Wellings et al., 2001), making an examination of the nature of this event particularly timely. The primary objective of this work is to review the literature on the timing, circumstances, and consequences of first (or early) heterosexual vaginal intercourse. A secondary objective is to examine the placing of heterosexual debut in the literature--that is, the way the subject has been defined, contextualized, and examined.
The search strategy for the literature review was developed using an evidence-based and iterative process to ensure that high quality interdisciplinary literature was identified. Key concepts and terms were identified for the review and used to create search strings that were combined to produce the final search strategy. The search strategy was developed on PubMed using both free-text and Medical Subject Headings (MESH) terms. The final search strategy was "Coitus [MESH] AND (debut OR (early NOT (function* OR dysfunction*)))."
The search strategy developed on PubMed was used as a basis for searching a range of electronic databases, which spanned several academic disciplines. The strategy was tailored to the characteristics of the database in question, with the databases' own thesaurus and MeSH terms employed where appropriate. Searches were initially conducted in August 2007 and updated in May 2009. In addition to the electronic searches, literature was also identified by scanning the reference lists of relevant articles identified by electronic search. Finally, literature was identified through contact with experts in the field.
Literature examining the meaning of first heterosexual vaginal intercourse, the timing and circumstances of first sex, the factors associated with the event, and the consequences was included. Academic literature published from 1960 onward in English was examined. Research studies employing quantitative or qualitative methodologies and conducted in the United Kingdom were considered.
Literature was selected according to quality criteria. Only databases that indexed predominantly peer-reviewed journals or that provided an option to limit searches to peer-reviewed literature were searched. In addition, only references including abstracts were included.
Papers were read and the relevant information extracted using pre-designed, loosely structured forms. Pertinent questions were identified in advance and used as a guide to reading. When writing up the findings, emphasis was placed on the most recent literature to provide a contemporary perspective of research in the field.
The Nature of the Evidence
The number of papers informing this review totalled 47, between them describing 37 unique studies.
The majority of the literature examining sexual debut employs a survey methodology, collecting quantitative self-reported data by interview or written questionnaire. These included the surveys conducted as part of the SHARE (Sexual Health and Relationships: Safe, Happy, and Responsible) and RIPPLE (A Randomized Intervention trial of Pupil-led sex Education in England) trials of sex education. Included reviews also drew heavily on self-reported data from surveys, and one employed quantitative meta-analysis techniques (Key, 1995).
In some of the recent literature, the validity of the survey data was increased by triangulating it with other forms of data such as diagnostic test results of sexually transmitted infection, case note analyses, or official government statistics. Case-control studies also involved external validation of clinical, biological, or social status.
Three studies employed qualitative approaches, drawing their data from in-depth interviews with young people (Holland, Ramazanoglu, Sharpe, & Thomson, 2000; Ingham, Woodcock, & Stenner, 1991; Mitchell & Wellings, 1998). Other studies employed multimethod approaches, incorporating both quantitative and qualitative methods. Theoretical papers adopted a commentary approach or applied a behavioral ecology framework to the study of childhood experience and interpersonal development.
The majority of the literature related specifically to the United Kingdom or Great Britain. Most studies focused on England or Scotland, and one exclusively examined northern Ireland (Schubotz, Rolston, & Simpson, 2004). Other studies were comparative, examining debut across several countries; one of these included data from Wales (Currie et al., 2004). The theoretical papers did not refer to a specific setting.
The terminology used to describe first sexual intercourse varies. Many studies simply refer to the event of first (sexual) intercourse itself; other, more recent, studies refer to sexual debut, whereas the term coitarche is common in the medical literature. Others use terminology such as first sexual experience or the onset of sexual activity, of sexual relations, or of sexual experience, contextualizing first sexual intercourse within the broader range of sexual acts. Studies vary in the extent to which sexual experience is equated with first sexual intercourse. West, Wight, and Macintyre (1993), for example, defined sexual experience as "the experience of having had heterosexual sexual intercourse" (p. 374), whereas Wellings and Field (1996) distinguished between the event of first intercourse itself and a young person's first sexual experience.
Operational definitions of sexual intercourse also vary. NATSAL (The National Survey of Sexual Attitudes and Lifestyles) 1990 and 2000, together with the SHARE and RIPPLE trials, defined heterosexual intercourse as "a boy/man putting his penis into a girl/woman's vagina," or "going the whole way," whereas Ingham et al. (1991) defined it as "vaginal penetration with or without ejaculation" (p. 121).
Assumptions regarding sexual orientation are often apparent in the language used. Whereas some studies explicitly specify heterosexual intercourse or experience, many do not. In some cases, it can be implied by the context--for example, when the research is framed in relation to teenage pregnancy. Two included studies reporting combined data on homosexual and heterosexual first intercourse (Coleman & Testa, 2007, 2008; Schubotz et al., 2004), one defining intercourse as vaginal or anal and as occurring with the same or the opposite sex (Coleman & Testa, 2007, 2008).
First sexual intercourse is contextualized in terms of the risk of adverse outcomes in the majority of studies. Other studies discuss debut in terms of the social, economic, or political environment within which young people live (e.g., with reference to the nature of interpersonal relationships; Holland et al., 2000; Ingham et al., 1991; Mercer et al., 2006; Wight et al., 2008). A few studies discuss first intercourse within the context of sexual behavior more generally or from a developmental or evolutionary perspective.
The findings presented in this review are described in four main sections. First, the data on the timing of first sexual intercourse are presented, alongside an examination of the factors associated with timing. Second, the data on the context or circumstances within which first sexual intercourse occurs are presented; this includes consideration of the nature of the sexual partnership and the motivation for the first sexual encounter. Again, the factors associated with these contextual characteristics are presented. Third, risk reduction practices at first sex are examined, together with the distal factors associated with contraception. Finally, the data on the consequences of first sexual intercourse are presented.
Timing of First Sexual Intercourse
A dominant focus of the literature examining first sexual intercourse is on the timing of the event. In general, timing is seen in relation to chronological age, and the presentation of the data often reflects normative assumptions. Terms such as early sexual intercourse or early sexual experience, for example, may suggest an appropriate time for an individual to begin sexual activity. Timing is most commonly conceived in relation to chronological age and so the data on the timing of first sexual intercourse are often presented in terms of the proportion reporting sex before a certain age. The most commonly selected age is 16, the age of sexual consent in Britain (Wellings & Parker, 2006). NATSAL 2000 (n = 11,161) found the proportion of 16- to 19-year-olds reporting first sex before the age of 16 to be 30% and 26% for men and women, respectively (Wellings et al., 2001). In northern Ireland, where the age of sexual consent is 17, the data were additionally presented before age 17 (Schubotz et al., 2004). In a large multi-method study (n = 1,013), 27% had had sexual intercourse before the age of 16, and almost 70% had had sex by 17 years (Schubotz et al., 2004). More recent surveys carried out in England and Scotland do not contradict these findings (Blenkinsop, Wade, Benton, Gnaldi, & Schagen, 2004; Currie et al., 2004; French et al., 2007; Henderson et al., 2002; Lenciauskiene & Zaborskis, 2008; Stephenson et al., 2004; Wallace et al., 2007; Wight et al., 2008; Wight et al., 2002), although some report rates closer to 40% (Currie et al., 2004; Lenciauskiene & Zaborskis, 2008; Stephenson et al., 2004; Wight et al., 2008; Wight et al., 2002). Median age is also presented in the studies, and is universally reported as 16 years (Schubotz et al., 2004; Wellings et al., 2001).
There is considerable interest in trends in the timing of first sexual intercourse and, because the event occurs only once, age cohort analyses of data from cross-sectional studies can be used to show changes over time. In recent decades, the age at which young people become sexually active has fallen (Schubotz et al., 2004; Wellings & Field, 1996; Wellings et al., 2001). Median age at first intercourse was 17 years among the 40- to 44-year-old age group in NATSAL 2000 for both men and women (Wellings et al., 2001). The proportion of women reporting first intercourse before 16 years has increased over recent decades, although not since the mid 1990s (Wellings et al., 2001).
Studies also report on the gender differential in timing of first sexual intercourse, and several studies (Blenkinsop et al., 2004; Currie et al., 2004; Currie & Todd, 1993; Lenciauskiene & Zaborskis, 2008; Wight et al., 2008), but not all (Wellings et al., 2001), have found a higher proportion of girls having had sex by age 16, compared with boys. One English survey (n = 3,820) revealed that girls were, on average, six months older at first sexual intercourse (Wallace et al., 2007).
Data from the SHARE trial (n = 7,630) revealed that the occurrence of a first sexual encounter was not necessarily indicative of regular sexual activity: Of the 14-year-old respondents who reported sex, one-third had done so only once (Henderson et al., 2002).
Considering the appropriateness of first sexual intercourse with reference to readiness contrasts with reference to chronological age, which characterizes much of the research. The concept of readiness to have sex first appeared in the literature in the 1980s, although the concept was first operationalized only in 2000. The literature highlights the disjuncture that exists between readiness considered in physical terms (or sexual adulthood) and readiness understood in terms of social acceptability (or social adulthood; Barron, 1986; Shaughnessy & Shakesby, 1992; Wellings & Field, 1996; Wellings et al., 2001).
The evidence suggests that age may be a simplistic indicator for the appropriateness of sexual activity. NATSAL 2000 operationalized the concept of sexual competence or readiness using self-report of four variables: contraceptive protection, autonomy of decision (not influenced by alcohol or peer pressure), consensuality (both partners equally willing), and absence of regret (right time for me). Using these criteria, early age at first sex was significantly associated with a lack of sexual competence. However, age did not account for all of the lack of competence: 91% of women and 67% of men who reported first intercourse at 13 or 14 years were considered not sexually competent, but so too were 30% of those who had sex later than age 18 (Wellings et al., 2001). Discussing first sexual intercourse within a broader framework that takes contextual variables into account may be more informative in future research.
Factors Associated with the Timing of First Sexual Intercourse
The factors influencing the timing of first sexual intercourse are described as physiological, social, and psychological in origin.
Physiological and Psychological Factors
At the biological level, both early age at menarche and early age at spermarche have been associated with a younger age at first intercourse (Kim & Smith, 1999; Wellings et al., 2001). Twin studies in other settings have shown a small genetic influence (Mustanski, Viken, Kaprio, Winter, & Rose, 2007). Biological and environmental influences are not discrete, and interactions between them have been demonstrated. Belsky, Steinberg, and Draper (1991), for example, suggested that stressful experiences within the family during childhood might impact upon later sexual behavior by initiating earlier maturation and puberty. In support of this theory, a small survey (n = 49) conducted in Sheffield found that family conflict and maternal rejection, as well as anxiety and depression, during childhood were predictors of early menarche (Kim & Smith, 1998).
Social and Educational Factors
It is, however, to the social and environmental factors associated with first sexual intercourse that most empirical attention has been paid, and a prime focus in this context has been the family. The literature has examined the impact of family dynamics and of family stability on the timing of first sex. The SHARE study in Scotland found that young people whose mothers were aged under 40 were more likely to experience intercourse before the age of 15 (Henderson et al., 2002; Wight et al., 2008). A case control study sampling 100 predominantly White British young women from disadvantaged inner city areas in Britain found that those who had experienced serious family difficulties were three times more likely to have had sex (Pawlby & Mills, 1997). Evidence from the NATSAL 1990 and 2000 surveys, that young people who lived with only one, or neither, natural parent to age 16 begin heterosexual relationships earlier (Kiernan & Hobcraft, 1997; Wellings et al., 2001), is confirmed in smaller surveys carried out in Scotland, England, and northern Ireland (Blenkinsop et al., 2004; Henderson et al., 2002; Lenciauskiene & Zaborskis, 2008; Wight et al., 2002).
There is also evidence to suggest that young people living in foster or residential care are more likely to have sex before the age of 16 (Crocker & Carlin, 2002; Henderson et al., 2002). In a small-scale study of over 300 female attendees of a Genito-Urinary Medicine clinic in Nottingham, the proportion reporting intercourse before age 16 was more than twice as high among those who had lived away from their parents' home during their childhood compared with those who had not (Crocker & Carlin, 2002).
Other studies examining the impact of family dynamics on timing have found low levels of parental monitoring and high amounts of pocket money (a proxy measure of child autonomy; Wight, Williamson, & Henderson, 2006) predictive of early sexual activity in both sexes (Blenkinsop et al., 2004; Curtis, Lawrence, & Tripp, 1988; Henderson et al., 2002; Lenciauskiene & Zaborskis, 2008; Wight et al., 2002; Wight et al., 2006). The Health Behaviour in School-aged Children (HBSC) study carried out in 2001 and 2002 (n = ~3,000) found a low level of maternal monitoring to be more significant for young men and a low level of paternal monitoring to be more significant for young women in predicting early sexual initiation (Lenciauskiene & Zaborskis, 2008).
There are mixed results in the literature regarding the impact of communication with parents on the initiation of sexual activity. The HBSC study did not find ease of communication with parents to be related to the timing of sexual debut (Lenciauskiene & Zaborskis, 2008). An evaluation study of the APAUSE (Added Power And Understanding in Sex Education) sex education program in England (n= 14,127) discovered that students who talked openly with their parents were more likely to be sexually active (Blenkinsop et al., 2004). On the other hand, NATSAL 2000 and a smaller survey conducted in northern Ireland (n = 1,013) found that young women who discussed sexual matters with their parents were more likely to delay first sex (Schubotz et al., 2004; Wellings et al., 2001).
The literature examining the impact of socioeconomic status on the timing of first sexual intercourse reveals a relationship between social disadvantage and early sex (Blenkinsop et al., 2004; Wellings & Field, 1996; Wellings et al., 2001; Wight et al., 2008). The most recent data were gathered as part of the SHARE and RIPPLE trials (Wight et al., 2008). The proportion of 13- to 16-year-olds with high levels of deprivation (based on housing type, educational level, and employment) who had had intercourse was significantly higher compared with their counterparts with low levels of deprivation (Wight et al., 2008). However, these observations may be misleading, as socioeconomic status is related to educational level. NATSAL 2000 found that, after taking educational level into account, the effect of socioeconomic group was no longer significant (Wellings et al., 2001). A review of the literature by Singh, Darroch, and Frost (2001) and a survey of 18-year-olds conducted in Glasgow (n = 908) also revealed little or no influence of socioeconomic background (West et al., 1993).
Factors affecting debut also include educational level. Several studies have found that young people who continue their education for longer have sex later: For example, in northern Ireland, 35.6% of young men who achieve General Certificate of Secondary Education or equivalent qualifications have sex before their 16th birthday, compared with 21.1% of those who attain A levels or similar and 15.8% of those who achieve higher qualifications (Schubotz et al., 2004; Singh et al., 2001; Wellings & Field, 1996; Wellings et al., 2001; West et al., 1993; Wight et al., 2002). The school environment may also be important: A survey carried out in southeast Scotland among 1,206 fourteen- to fifteen-year-olds found that pupils from schools ranked lower than the national average for academic attainment were more likely to have had sex before the age of 16 (Graham, Green, & Glasier, 1996).
School sex education has been shown to have a marked effect on the timing of first sexual intercourse. NATSAL 1990 and 2000 surveys found that those young people who cited school as their main source of information about sexual matters were less likely to have had sex by the age of 16, compared with those who cited other sources, such as friends or the media (Wellings & Field, 1996; Wellings et al., 2001; Wellings et al., 1995). These results are corroborated by a smaller survey carried out in northern Ireland (n= 1,013; Schubotz et al., 2004) However, specific programs of sex education (e.g., peer-led or experimental programs) have been shown to have little or no effect on sexual debut, compared with routinely provided sex education (Buston, Williamson, & Hart, 2007; Henderson et al., 2002; Stephenson et al., 2003; Wight et al., 2000; Wight et al., 2008; Wight et al., 2002; Wight et al., 2006). The SHARE trial, for example, compared theoretically based sex education, which included a five-day teacher training program and 20 teaching sessions delivered over two years with current practice in a trial of over 8,000 thirteen- to fifteen-year-old pupils. The study found no differences between the two groups in relation to the onset of sexual activity one year into the program (Wight et al., 2002).
Ethnicity also appears to be associated with the timing of sexual debut (Blenkinsop et al., 2004; Bradby & Williams, 1999; Coleman & Testa, 2007; Singh et al., 2001; Wight et al., 2008). A recent survey conducted by the Trust for the Study of Adolescence (TSA) among a group of ethnically diverse 15- to 18-year-olds in London (n=2,602) found Black males (African and Caribbean) and White females (British) to have the greatest experience of sexual intercourse below the age of 16 (p < .001). Young Asians were least likely to have experienced intercourse before 16 (Coleman & Testa, 2007). These results are reflected in the findings of other studies (Blenkinsop et al., 2004; Bradby & Williams, 1999; Singh et al., 2001; Wight et al., 2008).
Both religious affiliation and religiosity (strength of belief) have been shown to be related to sexual initiation (Coleman & Testa, 2008; Henderson et al., 2002; West et al., 1993). Women with no religious affiliation, for example, more commonly report sexual intercourse before the age of 16, contrasting with those of Muslim or Hindu faith (p< .001; Coleman & Testa, 2008). In Scotland, attendance at religious services has been shown to be related to first sexual intercourse (Henderson et al., 2002; West et al., 1993), and frequency seems to be a factor--those attending church weekly were less likely to have had sex than those who reported attending less than weekly or never (West et al., 1993).
Context of First Sexual Intercourse
Studies reporting on contextual factors surrounding first intercourse are fewer than those on the timing of the event. The literature particularly focuses on a young person's relationship with his or her partner and the motivation for initiating sex for the first time.
Nature of the Partnership
A number of studies have examined the nature of the first sexual partner. Several surveys have found that most young people first have sex with a person they regard as a boyfriend or girlfriend (Coleman & Testa, 2008; Henderson et al., 2002; Wight et al., 2008) or with whom they are in a steady relationship (Coleman & Testa, 2008). There are exceptions, however, and in northern Ireland, only a minority of young people reported having been in a steady relationship with the person with whom they first had sexual intercourse (Schubotz et al., 2004). In Scotland, the majority of young people first have sex with a partner they have known for one month or more, but there is significant variation by gender, with young men more likely to engage in first sex with a casual partner (p < .001; Wight et al., 2008).
The majority of young people report having sex for the first time with
a partner of the same age or older, although there are interesting gender differences in regard to age (Henderson et al., 2002; Ingham et al., 1991; Mercer et al., 2006; Wight et al., 2008). In NATSAL 2000, young men, on average, reported being the same age as their partner at first sex, although 30% first had sex with an older female, and 25% first had sex with a younger female. Young women, however, were found to be two years younger, on average, than their first heterosexual partner: 71% of women first having sex with an older male and less than 5% first having sex with a younger male (Mercer et al., 2006). These patterns have been observed elsewhere (Wight et al., 2008). NATSAL 2000 also found that the age difference between partners varied according to age at first sex such that there was more likely to be an age difference among those who first had sex at an older age (Mercer et