A substantial amount of research has been conducted in developed countries on the negative outcomes that may accompany early sexual activity. Unprotected sexual intercourse can result in unwanted pregnancies and abortions and places young people at an increased risk of infection with HIV and other sexually transmitted diseases.¹ Moreover, pregnant adolescents may go on to become adolescent mothers, who typically have low educational attainment and face economic disadvantage.² Similar outcomes may occur in less-developed countries such as the Philippines.

In the 1995 nationwide Young Adult Fertility Survey (YAFS) of 11,000 Filipino males and females aged 15-24, the majority (82%) reported never having had sexual intercourse (74% of males and 89% of females).³ These results are similar to those reported among young Filipino women in the first YAFS conducted in 1982.⁴ The majority of young Filipino adults choose to abstain from sex until they are married,⁵ perhaps as a consequence of the traditional values fostered through the church, the family and like-minded peers.

Given the risks of unprotected sexual intercourse during the adolescent years, it is important to obtain a comprehensive understanding of those factors associated with protective behaviors like sexual abstinence, as well as to address the needs of young people who may be at risk of engaging in unprotected sexual intercourse. Many diverse factors influence adolescents' level of sexual risk-taking behaviors. Poverty, race, ethnicity, religiosity, age at puberty, peer relations, school performance, involvement in other risk-taking behaviors and family composition and relationships have all been identified as determinants of adolescent sexual behavior among youths in the United States.⁶

Similar factors have been recognized as antecedents of premarital sex among young adult Filipinos.⁷ Research indicates that Filipino youths as a whole remain traditional in their values and actions regarding premarital sex.⁸ However, continued research and updated assessments are needed to accurately examine both the protective and high-risk behaviors and attitudes of Filipino youths. The aim of this study was to determine the correlates of sexual abstinence (defined as not having had sex during one's lifetime) by examining those factors most likely to be associated with adolescent premarital sexual intercourse. Understanding these correlates will be useful in developing interventions to maintain protective behaviors.

Methods

During summer 1995, a pretested, self-administered questionnaire was given to students in all sections of an Introduction to Sociology class from two large urban universities located in metropolitan Manila. The class, required for all students regardless of their college major at both universities, was chosen in order to maximize student participation and to eliminate repeating responders. This also ensured a representative sample of all disciplines at the two universities.

One week prior to the survey date, consent forms were given to students describing the purpose and content of the questionnaire. Any student who agreed to participate in the study was asked to sign, date and return the form. No students waived participation. Thus, all potential subjects participated in the survey.

Trained data collectors administered the surveys during class time. The survey was completed in approximately 45-50 minutes. Responses to each survey were coded by the respondent on an answer sheet for electronic data entry. Students were not asked to identify themselves by name. The completed questionnaires were sealed in envelopes and delivered to the Social Development Research Center in Manila. The lead researcher secured and transported them to the United States for data entry and analysis.

The survey used, the 1995 Adolescent Curriculum Evaluation, consists of 68 items and measures knowledge and attitudes about sex, HIV and AIDS and sexual behaviors.⁹ It has been administered to high school students in the United States, as well as to students in Sierra Leone. The written survey includes yes-no questions, Likert-type scales for attitudinal responses and closed-ended behavioral responses.

Knowledge items were transformed into categorical variables, indicating whether or not the respondent had adequate knowledge of a particular subject: Adequate condom knowledge was defined as correctly answering three out of four condom knowledge questions; adequate contraception knowledge required correct responses to four out of eight contraception questions; adequate pregnancy-risk knowledge required two out of three correct responses to pregnancy-risk knowledge questions; and adequate AIDS knowledge was defined as two correct answers out of three AIDS knowledge questions.

A positive attitude toward premarital sex was defined as agreement with the statement "It's okay to have sex before getting married" and disagreement with the statement "Sexual intercourse is never okay before marriage." A positive attitude toward recreational sex was defined as agreement with the statement "Sexual intercourse is okay even if a couple is not in love." Finally, adequate communication with parents or guardians about sex was defined as communication about three out of four of the following issues: pregnancy, birth control, abstinence, or HIV and AIDS.

The study variables included gender, age, age at first intercourse, monthly family income, self-reported academic performance, family composition during childhood, current living situation, religion, frequency of church attendance, adequacy of condom knowledge, adequacy of contraceptive knowledge, adequacy of pregnancy risk knowledge, adequacy of AIDS knowledge, attitude toward premarital sexual intercourse, attitude toward recreational sex, communication with parents or guardians about sex and sister younger than age 19 ever pregnant.

Chi-square tests were performed to detect differences by gender and by sexual activity status for all dichotomous study variables and to evaluate gender differences for abstinent students' most important reason for not having sex. Dichotomous multivariate logistic regression was used to identify correlates of sexual abstinence by gender. This approach allowed for the simultaneous estimation of the odds ratio of each variable while controlling for the effects of the other variables in the model. The odds ratios comparing sexually abstinent respondents with sexually active ones were used to estimate the association with potential correlates. A backwards elimination procedure was used to allow only statistically significant variables at the 0.1 level to remain in the model.

Results

Thirty-four class sections from the two selected universities were surveyed. The student response rate was 100%. A total of 1,355 students participated; however, we omitted 60 respondents from the study because they gave inconsistent responses to survey questions regarding demographic information or sexual activity status. The data reported here are based on the responses of the remaining 1,295 students.

Slightly more female (56%) than male (44%) students were surveyed (Table 1, page 170). The majority of respondents were 18 years old or younger (72%). Nearly two-thirds reported a monthly family income of 20,000 pesos or more. (The average per capita monthly income in the Philippines is approximately 2,000 pesos.¹⁰) Sixty-four percent of students lived at home with their parents, and 86% had been raised by both of their biological parents. Some 84% of respondents were Roman Catholic, and 74% attended a religious service more than twice a month. More than 90% of participating students indicated that their performance in school was at least average, although these self-reports were not verified.

Overall, 83% of the study respondents reported being sexually abstinent, while 17% reported having had sexual intercourse at least once in their lifetime. Of those who were sexually active, 52% had begun having sexual intercourse by age 16. Only 3% of all respondents had a sister younger than 19 who had become pregnant. Eighty-eight percent of respondents demonstrated adequate AIDS knowledge, and 74% had adequate knowledge about pregnancy risk. Some 52% of students had adequate knowledge about contraception, and 20% were adequately informed about condoms. Poor communication with parents about sex-related issues was reported by 75% of the respondents. Most respondents held nonaccepting attitudes toward both premarital and recreational sex (88% and 91%, respectively).

Equal proportions of males and females reported being Roman Catholic, although females were significantly more likely than males to attend religious services regularly (80% vs. 66%, p=.001). Males were significantly more likely than females to have ever had sexual intercourse (30% vs. 7%, p=.001), although both males and females who had had sex were equally likely to have done so before age 17 (53% and 49%, respectively).

Both male and female students had poor knowledge of condoms, although young women were significantly more likely than young men to have inadequate knowledge (90% vs. 67%, p=.001). Young women were also less well-informed than young men about contraception in general; 55% of females had inadequate knowledge in this area, compared with 40% of males (p=.001). Both males and females reported high levels of AIDS knowledge, although females were more likely than males to be adequately informed in this area (91% vs. 86%, p=.05).

Female students were significantly more likely (p=.001) than male students to hold nonaccepting attitudes toward premarital sex (93% vs. 81%) and toward recreational sex (96% vs. 84%). Females were significantly more likely than males to have good communication with their parents about sex (29% vs. 20%, p=.001).

The second panel of Table 1 indicates respondent characteristics by their sexual activity status. Three out of four sexually active students were males. Sexually active respondents were more likely than abstinent youths to have a monthly family income of 20,000 pesos or more (71% vs. 63%, p=.027) and to report below-average school performance (9% vs. 5%, p=.018). A greater proportion of abstinent students than sexually active students attended church more than twice a month (76% vs. 64%, p=.001). Abstinent respondents were more likely than sexually active youths to be younger than 18 (74% vs. 61%, p=.001).

Sexually abstinent respondents were significantly less likely (p=.001) than those who were sexually active to have adequate knowledge about condoms (17% vs. 38%) and about contraception in general (49% vs. 64%) and were more likely to view premarital sex as unacceptable (92% vs. 67%) and to have a nonaccepting attitude toward recreational sex (92% vs. 82%). Abstinent respondents were less likely than sexually active respondents to have a sister who had become pregnant before age 19 (2% vs. 6%, p<.01).

Among both male and female respondents, the three most common reasons mentioned for remaining abstinent were wanting to wait until one is older or married, religious values and concern about the risk of pregnancy or disease (Table 2). Females were almost twice as likely as males to cite wanting to wait as their main reason for abstaining from sex, while males were twice as likely as females to mention avoiding the risk of pregnancy or disease. Similar proportions of females and males indicated religious values, parent's values and not being asked as their most important reason for not having sex.

A multivariate logistic regression analysis was used to calculate odds ratios and corresponding 95% confidence intervals for the correlates of sexual abstinence by gender (Table 3). Males 18 years of age or younger were 1.6 times more likely to be sexually abstinent than were males older than 18, and females 18 years old or younger were 2.6 times more likely to be sexually abstinent than were those older than 18. Females who reported an affiliation with a religion other than the Roman Catholic Church were 3.3 times more likely to be sexually abstinent than those who were Roman Catholic.

Among males, family income was associated with sexual abstinence. Those with a monthly income of less than 20,000 pesos were 1.5 times as likely as those with a family income of 20,000 pesos or more to be sexually abstinent. Among young women, those who attended church more frequently were slightly more likely to be sexually abstinent, although the difference was not statistically significant (p<.10).

Female respondents with inadequate knowledge about condoms were 2.6 times more likely to be sexually abstinent than were those with adequate knowledge, and females who did not have a sister who became pregnant before age 19 were 5.6 times more likely to be sexually abstinent than those females with a sister who had an adolescent pregnancy. Female respondents who disapproved of premarital sex were 7.0 times more likely to be sexually abstinent and males who held this view were 2.9 times more likely to be sexually abstinent than were their peers with more permissive attitudes.

Academic performance, living with parents, having been raised by biological parents, knowledge of contraception, knowledge of pregnancy risk, knowledge of AIDS, attitudes towards recreational sex, and communication with parents about sex were not significantly associated with the likelihood of sexual abstinence.

Discussion

Attitudes play an important role in determining a person's behavior.¹¹ In this study, we found attitudinal differences across genders: Females were four times as likely as males to be sexually abstinent (only 54 of 730 females sampled were sexually active), and abstinent females were more likely to have nonaccepting attitudes toward premarital sex than were abstinent males. This finding supports previous research indicating that Filipino females value virginity more than do Filipino males.¹² It suggests that a gender-specific approach to sexuality education that strongly emphasizes decision-making and values-clarification among males is needed for the prevention of early sexual activity.

While a majority of both sexually abstinent and sexually active students demonstrated adequate knowledge about AIDS and pregnancy risk, the majority of respondents had inadequate knowledge about condoms and contraception in general. Although poor knowledge about condoms and contraception appears to be associated with sexual abstinence among both males and females, the fact that sexually active youths have inadequate knowledge in these areas seems most important to address, since this puts youths at risk for unintended pregnancy and infection with HIV and other sexually transmitted diseases. However, increased knowledge about contraception and condoms alone may not be sufficient to reduce sexual risks.¹³

One-third of the population of the Philippines are aged 10-24; therefore, the prevention of risky sexual behavior that may handicap their eventual contribution to society is a worthwhile investment.¹⁴ The findings of this study support previous research indicating that young Filipino adults remain relatively traditional in their sexual values and behaviors. However, with increasing modernization and urbanization, Filipino youths are being exposed to a growing number of conflicting messages that challenge traditional values. These cultural inconsistencies could result in an upward trend of Filipino youths engaging in high-risk sexual behaviors.

Thus, more qualitative research is essential to develop interventions that minimize the potential threats of increased rates of adolescent fertility and HIV infection. Research that incorporates a population-based sample of Filipino youth is needed to validate the use of sexual risk-reduction strategies within the Filipino social context.

Based on the responses from this study's sample population, peer education is a possible approach to promote protective social norms and also to emphasize decision-making skills and values education. Moreover, for youths who choose to be sexually active, interventions that build skills in condom use are necessary. While these strategies are very important, collaboration from the community is also needed to address these issues. In the battle to prevent unintended pregnancies and HIV infection among Filipino youths, researchers, educators and members of affected communities should collaborate to develop interventions that can enhance knowledge, correct misconceptions and maximize protective Filipino values.

Several limitations of this study should be noted. Survey respondents were primarily young and urban, and came from families with relatively high incomes. The study population is therefore not representative of Filipino youths in terms of economic status. We caution the reader against generalizing the findings of this study beyond populations that share its characteristics, both in terms of the economic level of respondents and the study's urban setting.

However, the study does appear to be representative of Filipino youths in terms of sexual behavior: Most Filipinos wait until marriage to become sexually active, and the mean age of the study population, 17.6, is well below the median age of first marriage for women in Manila, which is 24.¹⁵ (Men in the Philippines tend to marry at a later age than do women.) Moreover, similarities between the sexual knowledge, attitudes and behaviors of this study sample and the responses of 11,000 young Filipino adults on the 1995 YAFS are notable. For example, in that nationwide sample, 18% of the participants had sexual intercourse at least once in their lifetime,¹⁶ similar to the proportion in the present study. While these similarities reduce the probability of selection bias operating in this study, the use of a convenience sample does limit the study's applicability as a basis for large-scale recommendations.

Underreporting of sexual behavior has been documented in past studies in the Philippines,¹⁷ and participants in this study, especially females, may have underreported sexual behavior as well. In order to minimize this possibility, the authors followed proper survey protocol to ensure the confidentiality and anonymity of the participants' responses, and reliability checks were built into the survey to assess the consistency of sexual behavior responses. Furthermore, due to the nature of the survey, the study definitions of variables measuring knowledge and attitudes were restricted by the forced-choice format of the survey.

More in-depth studies on the sexual behavior of Filipino youths, using more comprehensive assessments, are necessary. Recent studies in the Philippines have incorporated research methods such as focus groups to assess the sexual and reproductive health concerns of university students.¹⁸ Future endeavors using such qualitative research methods in both rural and urban communities would allow for an improved understanding of the sexual behaviors and attitudes of Filipino youths. Qualitative assessments at varied sites could provide information that would be helpful in developing interventions that promote sexual risk reduction and take account of the social context in which such behaviors occur.