Minors who start having sex at earlier ages, compared to those who start later, have higher odds of several harmful outcomes such as engaging in unprotected sex (Kirby, 2002; Ma et al., 2009; Finer and Philbin, 2013; Crosby et al., 2015; Shrestha et al., 2016), having an increased number of sexual partners or of casual sexual partners (O’Donnell et al., 2001; Sandfort et al., 2008; Ma et al., 2009; Crosby et al., 2015; Heywood et al., 2015; Shrestha et al., 2016), perpetrating or being a victim of forced sex (O’Donnell et al., 2001; Shrestha et al., 2016), engaging in paid sex (Shrestha et al., 2016), having sex while drunk or drugged (O’Donnell et al., 2001; Sandfort et al., 2008), pregnancy (O’Donnell et al., 2001; Kirby, 2002; Ma et al., 2009; Crosby et al., 2015; Shrestha et al., 2016), abortion (Ma et al., 2009), STIs (Ma et al., 2009; Wand and Ramjee, 2012; Crosby et al., 2015; Lee et al., 2015; Shrestha et al., 2016), depression (Spriggs and Halpern, 2008), behavior problems (Udell et al., 2010), not attending tertiary education (Parkes et al., 2010) or regretting the age of sexual initiation (Cotton et al., 2004). However, some studies describe adolescent sexual initiation as a rite toward maturity or a transitional step taken by those who are ready (Ott et al., 2006; Masters et al., 2008).

Due to these different views on adolescent sexual behavior a debate has been established regarding which messages should be conveyed to youth. Some prioritize the invitation to children and adolescents to delay their sexual initiation in order to avoid risks (Halperin et al., 2004, 2011). Others maintain that this is unrealistic, and suggest to focus on risk reduction strategies such as condom use or partner reduction (Jemmott et al., 1998; Underhill et al., 2007; Jemmott et al., 2010). And there are those that simply explain both strategies without specifically focusing on any (Centers for Disease Control and Prevention, 2016).

Some of the risks associated with early sexual debut can be partly mediated by other variables such as number of lifetime sexual partners or condom use (DiClemente et al., 2005; de Sanjose et al., 2008; Wand and Ramjee, 2012; Oliveira-Campos et al., 2013). However, other risks might be intrinsic to the fact of initiating sex at a younger age (Stöckl et al., 2013). The lack of maturity could increase the odds of making less autonomous and sound decisions, for example related to sexuality (Patton and Viner, 2007; Casey et al., 2008).

Moreover, it seems that adolescents (or at least adolescent women) are biologically more susceptible to STIs. It has been recently explained that adolescent girls may have a higher risk of HIV infection as a result of their immature cervix. The cervix has a greater amount of genital mucosa with a high susceptibility to the virus (Dellar et al., 2015). Young people are also less likely to be at antiretroviral treatment and therefore, virally suppressed, which means they would have a higher transmission risk (World Health Organization, 2010, p. 57; Wringe et al., 2012).

In order to add information to this debate, we studied the association between age of sexual initiation during adolescence and a selection of well-being outcomes regarding that first relationship. One outcome was more related to sexual health (condom use) and the remaining ones were more related to feelings (5 variables measuring satisfaction) and circumstances related to sexual initiation (12 reasons for having sex). Specifically, we worked with the following research questions:

The study is based on Project YOURLIFE, an international cross-sectional and longitudinal study on youth, love and sexuality. This project has been described in more detail elsewhere (de Irala et al., 2009; Osorio et al., 2012, 2015; Carlos et al., 2016).

The questionnaire was responded by 6,085 adolescents aged 13–18 (2,686 from El Salvador and 3,399 from Peru). Approximately 19% (n = 1,179; 591 Salvadorians and 588 Peruvians) had ever had sexual relationships. After dropping those who didn’t report their age at sexual initiation and those who reported an age below 13 or an age of 18, we obtained a final sample of 996 adolescents (526 Salvadorians and 470 Peruvians) for this paper.

Table 1 shows the main socio-demographic data of these adolescents. Most sexually experienced participants were males (68.4%), and they had a mean age of 15.8 years. They were mostly Catholic, studied in public schools and lived in families with a middle socioeconomic level. We do not report the mean age of sexual initiation because the use of the mean age has been found to be misleading (de Irala et al., 2011, 2014). Conversely, the table contains the percent of participants who initiated at each age.

Some well-being outcomes varied across the different ages of sexual initiation. Figure 1 shows the trends in condom use at first sex as well as in feelings related to the satisfaction concerning that first sexual experience. Participants initiating sex at 17-years-old, compared to those initiating at 13, reported a significantly higher frequency of: condom use at first (62% vs. 38%), good memories of their first sex (86% vs. 68%) and regret over having already had sex (34% vs. 19%). Figure 2 shows the trends over different ages of several reasons for having had their first sex. Love was reported as a reason among participants initiating at 17 significantly more often than among those initiating at 13 (66% vs. 62%). On the contrary, other reasons were more frequent among those initiating at earlier ages.

Multiple logistic regressions (Table 2) show that some of these trends of condom use, satisfaction and circumstances concerning first sex are statistically significant even after adjusting for several confounders. Specifically, these regressions show that initiating sex at an older age was associated with more condom use, better memories of that experience, a higher frequency of referring love as a reason for having had first sex, and a lower frequency of several less positive reasons for having had that first sex, such as having first sex for fun, as a result of curiosity, because most of friends had had sex, because of being afraid of losing the partner, because the partner had said they would leave the participant, because of partner insistence, because of the influence of alcohol or drugs and as a consequence of seeing sexual images.

Separated analyses for each country and for each sex were performed. Results were similar, though we found fewer statistically significant results because of the smaller sample sizes. Sensitivity analyses with different approaches to analyze missing values (see Materials and Methods) showed similar results too (data not shown).

These results are a contribution to the body of knowledge suggesting that adolescents who have their first sexual relationship at earlier ages are more frequently subject to a selection of negative well-being outcomes.

Regarding condom use, the results from El Salvador and Peru are similar to those found in previous studies from other countries: condoms are more frequently used when sexual initiation occurs at an older age (Ma et al., 2009; Finer and Philbin, 2013; Crosby et al., 2015; Shrestha et al., 2016). This is not surprising, given the fact that younger adolescents are less mature and more impulsive. Actually, it is known that, during adolescence, there is a gap between the relatively slower maturation of the prefrontal cortex, responsible for planning and logical decision making, and the limbic system were emotions and rewards are generated and sought (Patton and Viner, 2007). In fact, in previous publications from this same study in El Salvador (de Irala et al., 2008, p. 59), Peru (Corcuera et al., 2010, p. 111) and Spain (Calatrava, 2010 Unpublished, p. 77), we had shown that the difficulty to obtain condoms or the lack of knowledge were not the most important reasons stated by adolescents when condoms were not used in their first sexual relationships. Conversely they overwhelmingly responded that non-condom use was due to the fact that the sexual encounter was not planned or expected. Similar results were found in another study among Irish youth (Layte et al., 2006, p. 28).

Within the issue of satisfaction with the first sexual encounter, the results are also consistent with previous research. A study in South Africa found that early sexual initiation was associated not only with less condom use and more sexual partners, but also with not feeling that they had been ready and wanted to have sex at the time of sexual initiation (Harrison et al., 2005). In our study, adolescents who initiated sexual relationships at an older age had better memories of their first experience. The remaining 4 associations measuring less satisfaction concerning the first sexual experience were not statistically significant although point estimates were in the expected direction yielding better estimates at older ages.

Our study also finds an association between age of sexual initiation and the circumstances and reasons for having had first sex. Those who initiated at older ages reported love more frequently and reasons related to lack of personal autonomy less frequently: peer pressure, partner pressure, substance use and sex because of having seen sexual images. This is consistent with the British results from the National Survey of Sexual Attitudes and Lifestyles (NATSAL). Some studies within this project have found an association between earlier age of first sex and different main reasons for first intercourse having occurred: peer pressure (Wellings et al., 2001) and being drunk (Schubotz et al., 2004). Some of these studies have built a composite measure (“sexual competence”), one of which components is to report autonomous reasons for the first intercourse. Sexual competence was also found to be associated with older age of first sex (Wellings et al., 2001).

The reasons for first sex do have important sexual health implications. For example, the lack of the aforementioned “sexual competence” was associated with a higher risk of having STIs, low sexual function, unplanned pregnancies and non-volitional sex (Palmer et al., 2016). Similarly, intimate reasons for first sex (for example, loving the partner) were associated with using contraceptives and with discussing contraceptive use before first intercourse (Stone and Ingham, 2002). Finally, in the United States, Else-Quest et al. (2005) defined that the first sexual experience occurred in a negative context if (among other reasons) the participant’s main reason for first sex was peer pressure or the influence of drugs or alcohol. Such negative context was associated with sex dysfunctions, sex guilt, STIs, poorer health, and less life satisfaction.

In summary, our results seem to confirm the association of an earlier age of sexual initiation with a selection of negative outcomes. This seems to pertain to a cross-cultural phenomenon: several important aspects related to sexual initiation in adolescence improve when such initiation takes place at older ages. In this study, what our outcomes have in common is that they are not “long term” negative consequences of sexual initiation rather immediate potential problems resulting from having first sex at that moment.

The main possible limitation of this study is that respondents might have differentially and erroneously recalled events (condom use, satisfaction and reasons for first sex) depending on age of initiation. The worse-case scenario would be that those that had their first sex at older ages might have overestimated their condom use, satisfaction and quality of reasons for first sex compared to those initiating sex at younger ages. We have no reason to believe that this is likely the case in our study. Rather, as the rosy view shows, past events tend to be evaluated better (Mitchell et al., 1997). That would imply that those having initiated sex earlier would recall a rosier experience than they really had, and consequently this could imply that their experience might have been even worse than what their replies show.

As in any cross-sectional study, it is important to correctly interpret the associations because the time sequence between independent and dependent variables has to be well established and/or understood. The interpretation of the association between earlier sexual initiation and less condom use or less satisfaction is straightforward. These outcomes are the result of younger age at sexual initiation. The association between earlier sexual debut and variables such as peer or partner pressure or substance use (reasons for first sex), is more correctly interpretable by affirming that they suggest that earlier sexual initiation takes place more frequently under those cited circumstances.

As for the strengths, the study consists of two large representative samples of school-attending adolescents in El Salvador and Peru. The results have been adjusted for possible confounders and are consistent across countries, and sensitivity analyses show similar results.

Our results seem to confirm that earlier sexual initiation is associated with worse outcomes related to the sexual health and well-being of adolescents. There is room and epidemiological reasons to empower youth (through family, schools, social media and other agents) to delay their sexual debut so as to be better prepared to be more in charge of their reproductive health when sexually active at later ages, as well as to make better decisions concerning the consequences of having sex.

The study was approved by the Ethics Committee of the University of Málaga (Spain). Survey procedures were designed to protect student privacy by ensuring voluntary and anonymous participation. Students were informed that their participation was voluntary and anonymous, and that they could leave the room and/or leave the questionnaire unanswered. They were also told that they could leave any question unanswered if they wished. They were also informed that answering the questionnaire was considered as consent to participate in the study.

All authors participated in the design and implementation of the project. AO and JdI designed this specific study. AO performed the analyses and wrote the initial draft. All authors revised the draft, made substantial contributions and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer KR and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.