Characteristics of Canadian Youth Reporting a Very Early Age of First Sexual Intercourse
Boyce, William F., Gallupe, Owen, Fergus, Stevenson, The Canadian Journal of Human Sexuality
Abstract: The present study examined the correlates of having experienced first sexual intercourse (FSI) at a very early age using a large, national classroom sample of Canadian adolescents from the Canadian Youth, Sexual Health and HIV/AIDS Study (Boyce et al., 2003). Most of the available studies on this topic have divided adolescent respondents into categories of "early" and "late" based on the average age of first intercourse. As a result, a portion of the young people identified in these studies as having had "early" first sexual intercourse had actually done so at an age when intercourse would have been becoming normative. The large size of the present sample of adolescents (n=2301; mean age 15.8 years) provided enough males and females who had non-normative very early FSI to compare them with peers who had first intercourse later. Associations were tested on variables in four conceptual categories: family relationships; psychological factors; peers and risk-taking; and partner-related factors. A very early age of FSI (defined as 11 years or less for males and 12 years or less for females), was associated with having experienced pressure to have unwanted sex, having used drugs other than marijuana, and believing that popularity at school is dependent upon rebelling/breaking the rules. While the retrospective nature of our cross-sectional analysis precludes assigning directionality of influences, the possible predictive value of the findings, including the influence of "fitting in" with peers, is considered in relation to future research on this topic.
While a substantial body of research has examined the correlates and risk factors for early first sexual intercourse (FSI) in adolescence, there has been little focus on this subject in a Canadian population context. Most such studies have designated FSI as either "early" or "late" based on the average age of first intercourse and have thus included in their "early" samples youth whose first intercourse occurred at ages approaching normativity. In such cases the early-late dichotomization tells us little about the correlates of FSI at non-normative very early ages. The present study addresses this gap in knowledge by examining correlates of first intercourse at these very early non-normative ages in a large national sample of grade 9 and 11 students who have ever had intercourse drawn from the Canadian Youth, Sexual Health and HIV/AIDS Study (Boyce, Doherty-Poirier, Mackinnon, & Fortin, 2003).
Literature on the non-biological correlates of age of FSI has tended to focus on factors associated with four broad conceptual categories: family relationships; psychological factors, peers and risk-taking; and partner-related factors. The present study also uses these categories and the literature review that follows will thus consider past findings in these areas as a basis for our choice of ages representing very early FSI for male and female students. We note at the outset, and again in the discussion, a principal limitation of cross-sectional studies, such as the present one, in which older students' responses to questions about their current lives are used to identify factors thought to be associated with their having begun intercourse at a very early age a number of years in the past.
The limitation is an inability to determine directionality of influence between possible predictors and past events. This is particularly so in the present study, in which associations are based on ages reflecting very early FSI that are younger than may have been chosen in the "early intercourse" studies reviewed here. In the case of our own findings in particular, we will therefore refer to factors that are "associated with" or "related to" early or very early FSI rather than "predictive" or "preventive" of it. This being said, we do anticipate that a number of the associations found in this study will prove to be contributors to very early FSI and not simply consequences of it.
A number of studies have documented associations between family factors and early FSI. Greater connectedness with parents/family (Resnick et al., 1997) and being part of an expressive family (Rosenthal et al., 2001) have been found to be related to later first intercourse suggesting that positive affective family ties reduce the likelihood of early FSI. Family attitudinal factors have also been found to be influential. For example, parental disapproval of both early sex and adolescent contraceptive use has been found to be associated with a later age of FSI (Resnick et al., 1997). Behavioural control factors, such as greater parental monitoring (Longmore, Manning, & Giordano, 2001; Rosenthal et al., 2001), have been shown to be associated with a later age of FSI.
Family structural factors, using a variety of definitions, have also been identified with age of FSI. Early age of FSI has been associated with dysfunctional family history (Magnusson, 2001), family turbulence (Wu 8,: Thompson, 2001), and not living with both parents (Laflin, Wang, & Barry, 2008; Langille & Curtis, 2002; Wu & Thompson, 2001).
Spencer, Zimet, Aalsma, and Orr (2002) found high scores on a global self-esteem measure to be related to an earlier age of FSI for boys and a later age of FSI for girls. However, Laflin et al. (2008) found higher peer-related self-esteem to be associated with less likelihood of intercourse for both boys and girls. Certain psychological factors associated with school (as opposed to school social or structural factors) have also been related to a youth's age of FSI. Higher academic achievement (Laflin et al., 2008; Resnick et al., 1997; Santelli et al., 2004), greater school connectedness (Resnick et al., 1997), and greater academic participation (Crockett, Bingham, Chopak, & Vicary, 1996) have all been associated with lower likelihood of early age of FSI. Factors associated with religion such as greater religiosity (Resnick et al., 1997; Rostosky, Wilcox, Wright, & Randall, 2004), more frequent church attendance (Crockett et al., 1996; Langille & Curtis, 2002), having taken a virginity pledge (Resnick et al., 1997), and greater prayer frequency (Latin et al., 2008) have been similarly associated with a lower likelihood of early FSI.
Peers and risk-taking
Better peer relations (Crockett et al., 1996) and greater levels of peer pressure to have sex (Laflin et al., 2008) have both been found to be associated with an earlier age of FSI as has substance use, an activity normally done in the presence of peers (Crockett et al., 1996; Kinsman, Romer, Furstenberg, & Schwarz, 1998; Santelli et al., 2004; Tapert, Aarons, Sedlar, & Brown, 2001). Crockett et al. (1996) also found minor delinquency, which similarly includes acts commonly done with peers, to be associated with early FSI. Associations have been found between early FSI and perceptions that friends are sexually active (Kinsman et al., 1998), greater peer support for sex (Matin, Coyle, Gomez, Carvajal, & Kirby, 2000; Santelli et al., 2004), attitudinal factors such as believing that engaging in sexual activity leads to gains in social status (Kinsman et al., 1998), and "attitudinal tolerance of deviance" (that is, lower belief in the "wrongness" of acts such as theft, aggression, and lying) (Costa, Jessor, Donovan, & Fortenberry, 1995).
Laflin et al. (2008) found an association between greater abstinence self-efficacy (measured using partner-related items) and lower likelihood of early FSI. However, the dynamics between intimate partners may work to override intentions to delay sexual involvement. Matin et al. (2000) found that having an intimate partner two or more years older was related to an earlier age of FSI as was the experience of unwanted sexual advances (pre-sexual contact). This Marin et al. study was also one of the only studies to address the influence of pressure/ coercion on early FSI.
Identifying an age that can be considered "early" for first sexual intercourse
There is a debate in the literature regarding the age at which FSI should be considered "early". The common dichotomization of early and late FSI categories by placing youth on either side of a sample's midpoint age of FSI (Crockett et al., 1996; Magnusson, 2001; Rosenthal et al., 2001) can misclassify many youth as having early FSI when they were actually at an age that was becoming normative. This dichotomization limits our knowledge about youth whose FSI was at a nonnormative, very early age by grouping them with others who do not really fit this category. Nonnormative very early age of FSI may be related to coercive or forced sex, but it may also be associated with other personality-based and/or social variables. The age at which very early FSI becomes normative is therefore up for debate and clearly depends on a variety of cultural, biological, and environmental factors in any location (Tonkin, Murphy, & Poon, in press).
The focus of the present study
Certain work has explicitly excluded youth who had FSI at a very early age (e.g., age 11 years and under) because these youth are said to be a distinct subpopulation (Kinsman et al., 1998; Resnick et al., 1997). In the present study of grade 9 and 11 students who have ever had sexual intercourse, our focus is on those students who reported FSI at these very early ages. It is thought that youth with early FSI may be most at risk of negative consequences. For example, Coker et al. (1994), in their school-based sample of American adolescents in grades 9-12, found FSI before age 13 to be associated in future years with greater numbers of sexual partners, reduced likelihood of using condoms regularly, and an increased risk of having been pregnant/caused a pregnancy, and having had an STI (for females). Furthermore, they found early FSI before age 13 to be related to fighting, taking a weapon to school, and early tobacco and alcohol initiation. Kinsman et al. (1998) found that, among a high-risk sample of impoverished urban American students in grade 6 (mean age of 11 years), having sex at this age was associated with an expressed intention to have sex, a belief that peers are sexually active, and a belief that sexual intercourse confers social status. The present study used a younger age (11 or less for males, 12 or less for females) to identify students who had very early FSI with a view to determining whether the associations found in prior studies that used a higher age for "early" FSI would be stronger, weaker or no different. The rationale for our choice of cut off ages of 11 or less for males and 12 or less for females is presented below.
Data base and sampling
The data for this study comes from the Canadian Youth, Sexual Health and HIV/AIDS Study (Boyce et al., 2003), a classroom based survey designed to measure sexual health behaviours, attitudes, determinants, and knowledge of key sexual health issues in a nationwide sample of adolescents. Conducted in 2002/2003 (n=11,125), Grade 7, 9, and 11 students (approximately 12, 14, and 16 years) from all provinces and territories were included, with the exception of Nunavut. Active consent was sought and measures were taken to ensure anonymity of the subjects (no names were collected; subjects placed the surveys in sealed envelopes before returning them). Teachers followed a standardised protocol when administering the survey (Boyce et al., 2003).
A stratified sampling technique using a single stage cluster design was performed to select classrooms from particular schools. Stratification variables used to select schools were: a) public versus Roman Catholic school, b) English versus French, c) urbanicity/size of city, d) geographic location, and e) school size. One class per grade per selected school was randomly chosen to participate with the exception of the samples from Prince Edward Island, the Yukon, and the North-West Territories. Here, the whole student populations in Grades 7, 9, and 11 were included due to their small numbers. Only students in Grades 9 and 11 are included in this analysis as detailed sexual behaviour questions were not asked at the Grade 7 level.
Factors representing conceptual categories shown to be associated with early FSI in prior studies were used here. The categories included: family relationships (both affective and structural factors); psychological factors (self-esteem, school, religious, and attitudinal factors); peers and risk-taking; and partner-related factors (sexual coercion). The Canadian Youth, Sexual Health and HIV/AIDS Study addressed each of these thematic categories. All variables were coded into two, three, or four level categorical variables for the purposes of odds ratio analysis.
Dependent variable: Very early FSI
This is a binary outcome variable coded from an item asking "How old were you when you had sexual intercourse for the first time?" In order to avoid the previously described problems of "early-late" dichotomization of FSI but without discarding entirely the concept of an average age of FSI, we combined a probabilistic with a theoretical approach. It is thought that the choice of ages that are two standard deviations below the mean age of FSI (among all those who have ever had sexual intercourse) is sufficient to encompass those who can be considered non-normative (in a probabilistic sense) in relation to their age of FSI. With a view to encouraging future research to adopt our methodology, we chose to use the whole number in years closest to two standard deviations below the mean age of FSI as the critical value determining a very early age of FSI. For boys, those who had FSI at age 11 or younger were thereby placed in the (nonnormative) very early FSI category; for girls the age cutoff was 12 years old or younger. For males, the reference group in the present study was male students with FSI at 12 to 18 years; for females, the reference group was female students with FSI at 13 to 18 years. This approach is consistent with some other research that used the 11 to 12 age range as the upper end of "early" FSI (Marin et al., 2000).
All independent variables were drawn from Boyce et al. (2003) and were measured as described below.
Family relationship variables
Parent relationship scale: This scale measures various aspects of youth relationships with parents and was created through a summation of 11 items. All items were measured using a 5-point scale (strongly agree to strongly disagree) and reverse coded when necessary. The first ten items involving mother or father were: "My mother (or father) understands me."; "I have a lot of arguments with my mother (or father)."; "My mother (or father) trusts me."; "What my mother (or father) thinks of me is important."; and "My mother (father) expects too much of me". The eleventh item was "I have a happy home life." Reliability for this scale was high (Cronbach's alpha = 0.851). Summing these items resulted in a scale ranging from 11 to 55 which was categorised into approximate tertiles: 1 = good parent relations (scores 42 through 55); 2 = moderate parent relations (scores 36 through 41); 3 = poor parent relations (scores 11 through 35).
Family structure: The range of family living situations was dichotomized into living with both parents (coded 0) versus some "other" situation (coded 1).
Self-esteem: Five self esteem-related items that were addressed in Boyce et al. (2003) were entered into a maximum likelihood factor analysis with an oblique rotation. These five items ("I like myself," "I would change how I look if I could," "I have confidence in myself," "I often have a hard time saying "no", "I am often sorry for the things I do") were measured on five-point scales (strongly agree to strongly disagree) that were coded so that low scores indicated good self-esteem. Results indicated a two-factor solution (initial eigen values of 2.07 and 1.08). Three items ("I like myself," "I would change how I look if I could," and "I have confidence in myself") loaded substantially (0.3) on the first factor. Only one item ("I am often sorry for the things I do") loaded greater than 0.3 on the second factor. The self-esteem construct was therefore created through a summation of the three items with factor loadings greater than 0.3 on the first factor (Cronbach's alpha = 0.714). This 3 to 15 scale was divided into tertiles: 1 = good self-esteem (scores 3 to 5); 2 = moderate self-esteem (scores 6 and 7); 3 = poor self-esteem (scores 8 through 15).
Religious attendance in the past year: This four category variable ("never" to "usually every week") was recoded into 1 = regular attendance (sometimes, usually every week); 2 = on special occasions; 3 = no religious attendance in the past year.
School achievement: This variable measured the student's grade average last term or semester and was coded: 1 = 80% or higher; 2 = 60% to 79%; 3 = less than 60%.
Desire to be popular at school: This item asked if a person must rebel or break the rules to be popular at school and was coded (0 = no, 1 = yes).
Peers and risk-taking variables
Smoking frequency: This variable measures current smoking habits and was coded: 1 = do not smoke; 2 = less than once a week; 3 = at least once a week but not every day; 4 = every day.
Close friends smoke: This item was coded 1 = less than half of my friends smoke; 2 = about half smoke; 3 = more than half smoke.
Bullied others: This five category item asks how often in the past two months the student has been involved in bullying other students at school (never to several times a week). Based on the cut points suggested for this item by Solberg and Olweus (2003), those who have bullied others "once or twice" or less in the past 2 months were considered non-bullies (coded 0) whereas those who have bullied others "2 or 3 times a month" or more often were considered bullies (coded 1).
Hashish/marijuana use frequency: This seven category variable asks how often the student uses hashish/marijuana (never to every day). This item was coded into: 1 = never; 2 = once a month or less; 3 = twice a month or more.
Use of drugs other than marijuana: Given the relative rarity of use of these substances, student use of glue/ solvents, LSD, ecstasy, cocaine/crack, or magic mushrooms was treated as a distinct phenomenon and these substances were combined into a single variable. Response categories on use of each item ranged from 1 = never to 7 = everyday use. This variable was collapsed into 0 = have not used other drugs; 1 = have used at least one other type of drug at least once.
Close friends use drugs to get stoned: This is a three category item coded as: 1 = less than half of my friends use drugs; 2 = about half of my friends use drugs; 3 = more than half of my friends use drugs.
Close friends have had sex: This item asks how many close friends have had sex. The response categories are coded: 1 = less than half; 2 = about half; and 3 = more than half of my friends have had sex.
Easy to talk to friends of the same sex: This item asks how easy it is to talk to friends of the same sex. Response categories are 1 = very easy; 2 = easy; 3 = not easy.
Easy to talk to best friend: This variable was coded 1 = very easy; 2 = easy; 3 = not easy.
Been pressured to have sex when did not want to: This item asks if the student has ever been pressured to have sex when they did not want to and is a 0 = no, 1 = yes item.
Have had sex when did not want to: This item asks if the student has ever had sex when they did not want to. Coding is 0 = no, 1 = yes.
The analysis was conducted in three stages. First, a bivariate analysis using the chi square statistic was conducted to determine which predictor variables differed significantly by age of FSI. The variables were then entered into separate binary logistic multivariate regression models for each conceptual group. Finally, the variables that were significant in the conceptual group logistic regressions were entered together into a multivariate logistic regression model to examine the simultaneous influence of multiple significant variables. Some prior studies that excluded youth who had FSI at the youngest ages (under 12) explicitly stated that early FSI analyses should be stratified by gender (Laflin et al., 2008) while others that used cut points for age of FSI that approximated those used in the present study did not indicate that a gender-stratified analysis is necessary (Marin et al., 2000). We present the latter option. However, interaction terms between gender and each of the variables in the final model were examined and none were found to be significant (results not shown).
Among the 2309 students in the sample who had ever had sexual intercourse, 54% were females and 46% males (Table 1). Of the 2301 students who reported their age, most (about 93%) were 14-17; average age was 15.8 years. The percentages of students at each grade level who had ever had intercourse were consistent with other studies and have been reported elsewhere (Boyce et al., 2006). Among the 2222 students reporting a valid age of FSI (1003 males, 1219 females), males and females did not differ in this respect. Overall, 6.2% of the combined sample of males and females who have ever had intercourse reported a very early age of FSI as defined in this study (i.e., age 11 or younger for boys and 12 or younger for girls).
Descriptive statistics: Correlates of very early FSI
Table 2 indicates that the majority of variables tested for their association with very early FSI were significant. Compared with those who had FSI at a later age, those with a very early age of FSI were more likely: to have a poor relationship with their parent(s); to believe that one must break the rules/ rebel to be popular at school; to report substance use (marijuana, drugs other than marijuana, smoke every day), to belong to a risk-taking peer group (majority of friends smoke, use drugs, and have had sex); to have bullied others, and to have difficulty talking to peers (best friend, friends of the same sex). Those with a very early age of FSI were also more likely to have been pressured to have sex when they did not want to and to have actually had sex when they did not want to. Family structure, self-esteem, religious attendance, and school achievement were not found to be associated with a very early age of FSI at the bivariate level.
Conceptual category logistic regression models: Very early FSI on bivariate correlates
In the separate regressions for each conceptual category, only one variable from each of the family relationships, psychological factors, peers and risk-taking, and partner-related categories were significantly associated with very early FSI (Table 3). From the family relationship category, those with a poor relationship with their parents were significantly more likely to have had a very early age of FSI than those with a good relationship with their parents (OR = 1.78; 95% CI = 1.15 - 2.75).
From the psychological factors category, believing that one must rebel or break the rules to be popular was associated with very early FSI (OR = 2.18; 95% CI = 1.51 - 3.15) compared to those who did not hold this belief.
In the peers and risk-taking category, use of drugs other than marijuana was significantly associated with very early FSI. Adolescents who had used such other drugs were almost twice as likely to have had a very early age of FSI compared to those who had not used those drugs (OR = 1.85; 95% CI = 1.11 - 3.09).
Finally, from the partner-related category, having been pressured to have sex when one did not want to do so was significantly associated with very early FSI. Those who had experienced such pressure were over twice as likely to have had a very early age of FSI when compared with those who had not experienced pressure to have sex (OR = 2.20; 95% CI = 1.39 - 3.49).
Logistic regression of very early FSI on significant associations from conceptual category regressions
When the four variables that were significant in the group logistic regressions were entered into the combined model, one variable each from the psychological factors, peers and risk-taking, and partner-related categories remained significant (Table 4). The family relationships variable (parent relationship scale) became non-significant at this stage.
The most influential association overall was the partner-related variable "having been pressured to have sex when did not want to" (OR = 2.53; 95% CI = 1.69 - 3.79). Believing that popularity at school is dependent upon rebelling or breaking the rules (OR = 1.87; 95% CI = 1.27 - 2.77) and having used drugs other than marijuana (OR = 1.89; 95% CI = 1.27 - 2.82) were also associated with a greater likelihood of very early FSI in the combined model by nearly a factor of 2.
This study demonstrates the strong association of experiencing sexual pressure with very early age of FSI. The experience of having been pressured to have sex when one does not want it was significantly and strongly related to very early FSI in both conceptually separate and combined analyses. While we do not know whether such pressure was related to the FSI itself (the item asking whether the student had ever actually had sexual intercourse when they did not want to was not significantly related to age of FSI), this finding does suggest that very early FSI may be accompanied by social pressure and may not be entirely voluntary. It highlights the necessity of considering involuntary influences as well as passivity in response to sexual pressure, in addition to rational decision-making approaches to the study of very early FSI.
As we noted at the outset, it is necessary to qualify the foregoing observations, and those that follow, by acknowledging that a principal weakness of cross-sectional analyses of the type conducted here is the inability to determine directionality of influence between predictors and outcomes. This is exacerbated when outcomes (i.e., very early age of FSI) actually precede the collection of data, and when there is a varying period of time between the event and the reporting. In this study, two-thirds of respondents were aged 16 or over so as much as 4 to 5 years or more may have been the norm between experience of very early FSI and reporting. During this time, many of the conceptual category variables may have changed and the very early FSI event itself may have influenced a number of associated factors, especially family, peer and personal factors. Prospective studies of very early FSI would be needed to clarify this issue. That being said, the findings reported here can inform such studies and it seems likely, based on the analysis that follows, that some of the associations reported here will prove to be contributors to very early FSI and not only consequences of it.
For example, our findings show that very early FSI is related to extreme risk taking that is often influenced by peers. Both boys and girls who reported using drugs other than alcohol or marijuana were more likely to have a very early age of FSI. While drug use is not exclusively a social activity, it usually occurs in a group context and has been found in previous work to be related to early sexual involvement (Crockett et al., 1996; Kinsman et al., 1998; Tapert et al., 2001). Our findings are not, however, supportive of Kinsman et al. (1998) who found an influence of friends' sexual behaviour on FSI. This type of peer influence may only apply to older youth who are at a more normative age for FSI. The belief that to be popular at school one must break the rules or rebel was also associated with very early FSI. This is not a direct measure of peer influence although it is a possible measure of antinormative attitudes associated with impressing one's peers.
Our findings offer only partial support for Resnick et al. (1997) who found that the predictors of early FSI were not associated with peers but instead were related to family and school factors, religiosity, and looking older than one's peers. Our study does, however, agree with previous work that has shown that familial factors, such as relationships with one's parents, play an important role in relation to age of FSI (Crockett et al., 1996; Rosenthal et al., 2001). Our present analysis shows that poor relationships with one's parents are related to very early FSI in the conceptual category analysis, but not in the final combined model.
One potential mechanism that emerges from these patterns suggests a pathway for very early FSI: "fitting in" with peers. It is possible that strained relations between the student and parents lead to subsequent efforts by the youth to become embedded in peer groups that have anti-normative tendencies, which this study has found to be related to a very early age of FSI. In this way, "fitting in" through activities such as drug use, exposure to situations in which sexual pressure occurs, and rebellious attitudes, might mediate the association between poor relationships with parents and very early FSI. This is an area that future longitudinal research could clarify.
Methodological strengths and limitations of the present study
This study utilized a large sample from a population data set of a particularly appropriate and important age group of students in Canada. As such, it has certain statistical strengths and the ability to reduce bias in estimates and to increase confidence in the associations identified. The study also was able to ask sensitive questions about FSI and various associated variables that are not often possible in school settings.
While girls are generally more honest than boys when reporting age of FSI (Alexander, Somerfield, Ensminger, Jonson, & Kim, 1993; Rosenthal, Burklow, Biro, Pace, & DeVellis, 1996; Upchurch, Lillard, Aneshensel, & Fang Li, 2002), evidence exists to question the validity of self-reported age of FSI (Brener, Billy, & Grady, 2003; Lauritsen & Swicegood, 1997; Upchurch et al., 2002). However, Alexander et al. (1993) in their test-retest reliability study indicated that reported inconsistencies of an individual's reported age of FSI were usually within a year of the initially reported age. Our decision to collapse the normative and late groups as a contrast group was intended to allow a focus on very young students and may have increased the observed differences in predictor variables between our two groups. In addition, Alexander et al. (1993) state that categorising individuals into early, normative, and late provides a valid method for analysing age of FSI and minimises problems associated with erroneous reporting.
Another methodological issue is the possibility that some youth may have used varying criteria for interpreting the meaning of "having sexual intercourse" although this was differentiated in the item we used. Brener et al. (2003), in their review of adolescent health-risk behaviour literature, suggest that the vast majority (93%) of adolescents indicate that they consider vaginal penetration to be sexual intercourse. However 63% and 22% respectively believed anal and oral sex are also intercourse.
Our study excluded youth in private schools and is therefore biased towards more mainstream adolescents. As such, the findings can only be generalised to the broader Canadian adolescent school population and not to those in private or special schools or to those out of school. Some adolescent groups who were missed by this sampling methodology (e.g., street youth and adolescents in custody) are likely to be at greater risk for sexual risk behaviours, anti-normative peer associations, and coercion. This means that the associations between the variables related to engaging in very early FSI are likely to be conservative estimates and could be exacerbated among other non school-based adolescent populations.
The findings reported here highlight the serious challenges presented to programs that seek to minimise the risks incurred by youth engaging in sexual behaviour at a young age. Any program attempting to promote the sexual health of adolescents must address the fact that a person's age of FSI is often at least partly a product of external social influences (peer behaviours, sexual pressure) as well as psychological anti-normative attitudes. There are inevitably those who will engage in sexual behaviours at a very early age and it appears reasonable to suggest that youth isolation from family leads to attempts to fit into peer groups through sexual involvement, even at a very young age.
Alexander, C.S., Somerfield, M.R., Ensminger, M.E., Jonson, K.E., & Kim, Y.J. (1993). Consistency of adolescent's self-report of sexual behavior in a longitudinal study. Journal of Youth and Adolescence, 22, 455-71.
Brener, N.D., Billy, J.O.G., & Grady, W.R. (2003). Assessment of factors affecting the validity of self-reported health-risk behavior among adolescents: Evidence from the scientific literature. Journal of Adolescent Health, 33, 436-57.
Boyce, W., Doherty-Poirier, H., Mackinnon, D., Fortin, C., Saab, H., King, M., & Gallupe, O. (2006). Sexual health of Canadian youth: Findings from CYSHHAS. The Canadian Journal of Human Sexuality, 15, 59-68.
Boyce, W., Doherty-Poirier, H., Mackinnon, D., & Fortin, C. (2003). Canadian Youth, Sexual Health and HIV/ AIDS Study, Factors influencing knowledge, attitudes and behaviours. Toronto, Ontario: Council of Ministers of Education, Canada.
Coker, A.L., Richter, D.L., Valois, R.F., McKeown, R.E., Garrison, C.Z., & Vincent, M.L. (1994). Correlates and consequences of early initiation of sexual intercourse. Journal of School Health, 64, 372-377.
Costa, F.M., Jessor, R., Donovan, J.E., & Fortenberry, J.D. (1995). Early initiation of sexual intercourse: The influence of psychosocial unconventionality. Journal of Research on Adolescence, 5, 93-121.
Crockett, L.J., Bingham, C.R., Chopak, J.S., & Vicary, J.R. (1996). Timing of first sexual intercourse: The role of social control, social learning, and problem behaviour. Journal of Youth and Adolescence, 25, 89-111.
Kinsman, S.B., Romer, D., Furstenberg, F.F., & Schwarz, D.F. (1998). Early sexual initiation: The role of peer norms. Pediatrics, 102, 1185-92.
Laflin, M.T., Wang, J., & Barry, M. (2008). A longitudinal study of adolescent transition from virgin to nonvirgin status. Journal of Adolescent Health, 42, 228-36.
Langille, D.B., & Curtis, L. (2002). Factors associated with sexual intercourse before age 15 among female adolescents in Nova Scotia. The Canadian Journal of Human Sexuality, 11, 91-99.
Lauritsen, J.L., & Swicegood, C.G. (1997). The consistency of self-reported initiation of sexual activity, Family Planning Perspectives, 29, 215-221.
Longmore, M.A., Manning, W.D., & Giordano, P.C. (2001). Preadolescent parenting strategies and teens' dating and sexual initiation: A longitudinal analysis. Journal of Marriage and Family, 63, 322-35.
Magnusson, C. (2001). Adolescent girls' sexual attitudes and opposite-sex relations in 1970 and in 1996. Journal of Adolescent Health, 28, 242-52.
Marin, B.V., Coyle, K.K., Gomez, C.A., Carvajal, S.C., & Kirby, D.B. (2000). Older boyfriends and girlfriends increase risk of sexual initiation in young adolescents. Journal of Adolescent Health, 27, 409-418.
Resnick, M.D., Bearman, P.S., Blum, R.W., Bauman, K.E., Harris, K.M., Jones, J., et al. (1997). Protecting adolescents from harm: Findings from the National Longitudinal Study on Adolescent Health. Journal of the American Medical Association, 278, 823-32.
Rosenthal, S.L., Burklow, K.A., Biro, EM., Pace, L.C., & DeVellis, R.F. (1996). The reliability of high-risk adolescent girls' report of their sexual history. Journal of Pediatric Health Care, 10, 217-20.
Rosenthal, S.L., Von Ranson, K.M., Cotton, S., Biro, F.M., Mills, L., & Succop, P.A. (2001). Sexual initiation: Predictors and developmental trends. Sexually Transmitted Diseases, 28, 527-532.
Rostosky, S.S., Wilcox, B.L., Wright, M.L.C., & Randall, B.A. (2004). The impact of religiosity on adolescent sexual behavior: A review of the evidence. Journal of Adolescent Research, 19, 677-697.
Santelli, J.S., Kaiser, J., Hirsch, L., Radosh, A., Simkin, L., & Middlestadt, S. (2004). Initiation of sexual intercourse among middle school adolescents: The influence of psychosocial factors. Journal of Adolescent Health, 34, 200-208.
Solberg, M.E., & Olweus, D. (2003). Prevalence estimation of school bullying with the Olweus bully/victim questionnaire. Aggressive Behavior, 29, 239-268.
Spencer, J.M., Zimet, G.D., Aalsma, M.C., & Orr, D.P. (2002). Self-esteem as a predictor of initiation of coitus in early adolescents. Pediatrics, 109, 581-584.
Tapert, S.F., Aarons, G.A., Sedlar, G.R., & Brown, S.A. (2001). Adolescent substance use and sexual risk-taking behavior. Journal of Adolescent Health, 28, 181-189.
Tonkin, R.S., Murphy, A., & Poon, C.S. (in press). Sexuality and reproductive health in adolescence: Policy implications of early age of sexual debut. In W. Boyce (Ed.), Adolescent Health in Canada: Science, Policy and Human Rights. Kingston, ON: Social Program Evaluation Group, Queen's University.
Upchurch, D.M., Lillard, L.A., Aneshensel, C.S., & Fang Li, N. (2002). Inconsistencies in reporting the occurrence and timing of first intercourse among adolescents. Journal of Sex Research, 39, 197-206.
Wu, L.L., & Thomson, E. (2001). Race differences in family experience and early sexual initiation: Dynamic models of family structure and family change. Journal of Marriage and Family, 63, 682-696.
William F. Boyce (1), Owen Gallupe (1), and Stevenson Fergus (2)
(1) Social Program Evaluation Group, Queen's University, Kingston, ON
(2) School of Kinesiology and Health Studies, Queen's University, Kingston, ON
Correspondence concerning this paper should be addressed to William Boyce, Social Program Evaluation Group, McArthur Hall, Queen's University, Kingston, ON K7M 5R7. E-mail: firstname.lastname@example.org. Note: Owen Gallupe is now working with the Centre for Addiction and Mental Health, Toronto, ON.
Table 1 Demographic characteristics of grade 9 and 11 students surveyed who had ever had sexual intercourse. Valid Frequency percent Gender (n=2309) Male 1054 45.6 Female 1255 54.4 Current age (n=2301) Mean= 15.8 years 13 8 0.3 14 455 19.8 15 285 12.4 16 1009 43.9 17 396 17.2 18 118 5.1 19 24 1.0 20 4 0.2 21 2 0.1 Age of male FSI * (n=1003) Very early (11 and under) 60 6.0 Later (12 and above) 943 94.0 Age of female FSI * (n=1219) Very early (12 and under) 78 6.4 Later (13 and above) 1141 93.6 * FSI means first sexual intercourse Table 2 Descriptive statistics for final model: Very early FSI predictors. Very early (a) Later (a) Family relationship variables Parent relationship scale Good 27.2 33.0 Moderate 20.8 31.4 Poor 52.0 35.7 Family structure Both parents 58.3 62.7 Other 41.7 37.3 Psychological variables Self-esteem Good 26.2 30.1 Moderate 29.2 31.5 Poor 44.6 38.4 Religious attendance in past year Regular 27.5 25.6 Special occasions 40.6 43.3 Never 31.9 31.1 School achievement 80%+ 26.1 31.2 60% to 79% 56.0 56.3 Less than 60% 17.9 12.6 Desire to be popular at school No 43.1 62.3 Yes 56.9 37.7 Peers and risk-taking variables Smoking frequency Do not smoke 36.0 52.5 Less than once a week 5.1 7.3 At least once a week 4.4 6.4 but not every day Every day 54.4 33.8 Close friends smoke Less than half 32.8 48.0 About half 12.7 17.9 More than half 54.5 34.1 Bullied others No 71.7 85.4 Yes 28.3 14.6 Hashish/marijuana use frequency Never 31.9 43.7 Once a month or less 17.8 23.2 Twice a month or more 50.4 33.1 Use of drugs other than marijuana No 53.5 76.0 Yes 46.5 24.0 Close friends use drugs to get stoned Less than half 32.6 44.8 About half 13.3 18.5 More than half 54.1 36.7 Close friends have had sex Less than half 20.8 26.8 About half 10.8 21.9 More than half 68.5 51.3 Easy to talk to friends of the same sex Very easy 43.6 47.6 Easy 31.6 39.3 Not easy 24.8 13.2 Easy to talk to best friend Very easy 60.2 70.3 Easy 22.6 22.3 Not easy 17.3 7.4 Partner-related variables Been pressured to have sex when did not want to No 61.5 81.8 Yes 38.5 18.2 Had sex when did not want to No 64.2 80.9 Yes 35.8 19.1 Chi (b) Df p Family relationship variables Parent relationship scale Good 14.08 2 0.001 Moderate Poor Family structure Both parents 1.01 1 0.314 Other Psychological variables Self-esteem Good 2.05 2 0.358 Moderate Poor Religious attendance in past year Regular 0.45 2 0.800 Special occasions Never School achievement 80%+ 3.82 2 0.148 60% to 79% Less than 60% Desire to be popular at school No 19.98 1 0.000 Yes Peers and risk-taking variables Smoking frequency Do not smoke 23.81 3 0.000 Less than once a week At least once a week but not every day Every day Close friends smoke Less than half 22.92 2 0.000 About half More than half Bullied others No 18.64 1 0.000 Yes Hashish/marijuana use frequency Never 16.84 2 0.000 Once a month or less Twice a month or more Use of drugs other than marijuana No 32.54 1 0.000 Yes Close friends use drugs to get stoned Less than half 16.27 2 0.000 About half More than half Close friends have had sex Less than half 15.58 2 0.000 About half More than half Easy to talk to friends of the same sex Very easy 14.56 2 0.001 Easy Not easy Easy to talk to best friend Very easy 17.31 2 0.000 Easy Not easy Partner-related variables Been pressured to have sex when did not want to No 33.39 1 0.000 Yes Had sex when did not want to No 22.09 1 Yes 0.000 (a) Percentage. (b) Pearson chi square value (predictor by very early FSI, first sexual intercourse). Table 3 Conceptual category logistic regression models of very early FSI on bivariate predictors. OR 95% CI Model A: Family relationships variables Parent relationship Good 1.00 scale Moderate 0.77 0.45-1.32 Poor 1.78 ** 1.15-2.75 Model B: Psychological variables Desire to be popular No 1.00 at school Yes 2.18 ** 1.51-3.15 Model C: Peers and risk-taking variables Use of drugs other No 1.00 than marijuana Yes 1.85 * 1.11-3.09 Model D: Partner-related variables Been pressured No 1.00 to have sex when Yes 2.20 ** 1.39-3.49 did not want to * p<0.05; ** p<0.01. Note: Only significant variables reported. FSI means first sexual intercourse. Table 4 Logistic regression of very early FSI on significant predictors from conceptual category regressions. OR 95% CI A. Family relationships variables Parent relationship scale Good 1.00 Moderate 0.64 0.36-1.12 Poor 1.29 0.81-2.05 B. Psychological variables Desire to be popular at school No 1.00 Yes 1.87 ** 1.27-2.77 C. Peers and risk-taking variables Use of drugs other than No 1.00 marijuana Yes 1.89 ** 1.27-2.82 D. Partner-related variables Been pressured to have No 1.00 sex when did not want to Yes 2.53 ** 1.69-3.79 * p < 0.05; ** p < 0.01. FSI means first sexual intercourse.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Characteristics of Canadian Youth Reporting a Very Early Age of First Sexual Intercourse. Contributors: Boyce, William F. - Author, Gallupe, Owen - Author, Fergus, Stevenson - Author. Journal title: The Canadian Journal of Human Sexuality. Volume: 17. Issue: 3 Publication date: Fall 2008. Page number: 97+. © 1999 SIECCAN, The Sex Information and Education Council of Canada. COPYRIGHT 2008 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.