Journal of Adolescent Health 50 (2012) S26 S36

www.jahonline.org
Original article

Medias Contribution to Sexual Knowledge, Attitudes, and Behaviors for Adolescents and Young Adults in Three Asian Cities
Chaohua Lou, M.D.a,*, Yan Cheng, Ph.D.a, Ersheng Gao, M.D.a, Xiayun Zuo, M.D.a, Mark R. Emerson b, and Laurie S. Zabin, Ph.D.b
a b

Department of Epidemiology and Social Science, Shanghai Institute of Planned Parenthood Research, Shanghai, P.R. China Department of Population, Family, and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland

Article history: Received July 14, 2011; Accepted December 6, 2011 Keywords: Mass media; Unmarried; Sex-related knowledge; Attitude; Sexual behavior; Asian city; Multicenter study

A B S T R A C T

Purpose: Evidence in western countries indicates that the media have associations with adolescents and young peoples sexual behavior that may be as important as family, school, and peers. In this new study of Asian adolescents and young adults in the three cities of Hanoi, Shanghai, and Taipei, the associations between exposure to sexual content in the media and adolescents and young adults sex-related knowledge, attitudes, and behaviors are explored in societies with traditional Confucian culture, but at different stages in the process of modernization. Methods: The data are from a questionnaire-based cross-sectional study conducted from 2006 to 2007, where a sample of 17,016 adolescents and young adults aged 1524 years from Shanghai, Hanoi, and Taipei completed face-to-face interviews coupled with computer-assisted self-interviews for sensitive questions. For the objectives of this article, analysis was restricted to the 16,554 unmarried respondents. Exposure to sexual content in the mass media (including the Internet and traditional media), pornographic videos, and a preference for western/Asian movies/videos were the main media inuence measures. Sex-related knowledge, premarital sexual permissiveness (PSP), and sex-related behaviors were the main outcome measures. The impact of each of four contexts including family, peer, school, and media on sex-related knowledge, attitudes, and behaviors were assessed using multiple linear regression stratied by gender and city, controlling for age, urban/rural residence, education, and economic status. The change in adjusted R2 from the multiple linear regression analysis was adopted to indicate the contribution of family, peer, school, and media variables to respondents sex-related knowledge, attitudes, and behaviors. Results: The contextual factors, including family, peer, school, and media, explained 30%50% of the variance in sex-related knowledge, 8%22% of the variance in PSP, and 32% 41% of the variance in sex-related behaviors. Media variables explained 13%24% of the variance in sexual knowledge, 3%13% in PSP, and 3%13% in sex-related behaviors, which was comparable with that of family, peer, and school variables. These associations differed by city and gender. Conclusions: Access to and use of mass media and the messages they present are inuential factors on sex-related knowledge, attitudes, and behaviors of unmarried Asian adolescents and young adults, and should be considered in future research and intervention programs attempting to improve reproductive health outcomes. 2012 Society for Adolescent Health and Medicine. All rights reserved.

* Address correspondence to: Chaohua Lou, M.D., Department of Epidemiology and Social Science Research on Reproductive Health, Shanghai Institute of Planned Parenthood Research, 2140 Xie Tu Road, Shanghai 200032, P. R. China. E-mail address: chaohual@yahoo.com (C. Lou).

1054-139X/$ - see front matter 2012 Society for Adolescent Health and Medicine. All rights reserved. doi:10.1016/j.jadohealth.2011.12.009

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

S27

The Asian areas of Taiwan, China mainland, and Vietnam share the same Confucian-based culture but have been open to outside inuences socially, culturally, and economically for different periods and in different ways. Research has shown that the levels of adolescents and young adults sexual activities differ in the three areas. In China mainland, with a more open policy in effect since the late 1970s, surveys have shown measurable increases in sexual intercourse among adolescents and young adults in general; for example, in 1998, 1.8% of middle school students in Shanghai reported ever having had sex, and this percentage had risen to 7.0% in 2005 [1,2]. A report in 2004 on an unmarried migrant population aged 1524 years in one district of Shanghai showed even higher rates of premarital sex: 34.8% of males and 12.7% of females were sexually active [3]. Taiwan, one of the most developed regions in Asia, seems to parallel many developed countries in several social aspects. In the Taiwan Young People Survey in 1994, 7% of adolescents aged 1519 years were sexually experienced [4]. More recently, a 2003 study involving girls in Taiwan vocational high schools showed that 25.5% were sexually experienced [5]. Since 1986, with the effects of economic reform, considerable social change has occurred in Vietnam. A survey conducted in six provinces among adolescents and young adults aged 1522 years in 1999 reported that 10% of male and 5% of female adolescents and young adults had had premarital sex [6]. In the Survey Assessment of Vietnamese Youth (widely known as SAVY) conducted in late 2003, it was found that 19.8% of 18 25-year-old boys in urban areas had had sex and 13.6% in rural areas, respectively. For girls, the percentages were still quite low, 2.6% and 2.2%, respectively [7]. Recent ecological models of adolescent health risk behaviors posit that behavior is produced by the interactions between people and their environments, for example, school, peers, and family [8,9]. According to social ecology theory, mass media are another important dimension of young peoples lives that may take on special signicance during adolescence, in particular for sexual risk behaviors [9,10]. With the approach of the information era, no matter whether in the East or the West, access to media gradually becomes prevalent in the daily life of adolescents and young adults. A recent survey in 2008 2009 found that on average, U.S. 8 18-year-olds spent over 7 hours per day using media [11]. In China, a survey on Internet use in 2007 indicated that more than 40% of youths aged 18 24 years visited the Internet [12]. In Taiwan, 75% male college students and 25% female college students were committed to Internet use [13]. In Vietnam, a qualitative study showed that young people used the Internet as a medium for expressing sexual identities and desires [14]. Previous research has shown that adolescents consistently cite mass media as important sources of sexual information [1517]. Moreover, some adolescents may be learning about sex from the media but not be aware of it. However, much of the media that adolescents are exposed to include sexual imagery but rarely portray the consequences of risky sexual encounters or convey healthy sexual messages [9,18,19]. Meghan Smith conducted content analysis on web-based reproductive health information in 2000 and found that 63% of information online can be dened as pornography, and she observed that this may negatively impact youths premarital sexual attitudes and behaviors [20]. Despite data showing that adolescents and young adults are frequent media users and consumers of numerous unhealthy media messages about sexual behaviors, limited research has included media as an inuential factor in models of teens risk

behavior. Increasing evidence in western countries points to the media adolescents use frequently (television, music, movies magazines, and the Internet) as important factors in the initiation of sexual intercourse, showing that there are possible causeand-effect between media use and earlier sexual intercourse [21], and that these effects are comparable with those of family, peers, and school [9]. However, no similar results have been reported in Asian societies thus far. In this article, we use the data from the Three-City Study of Asian Adolescents and Young Adults: Hanoi, Shanghai, and Taipei to explore the role of media in unmarried adolescents and young adults sex-related knowledge, attitudes, and behaviors. First, we hypothesize that exposure to media is an important context for adolescents and young adults sex-related knowledge, attitudes, and behaviors in these Asian cities with their heritage of Confucian-based culture, and that this effect is comparable with other contextual inuences, such as family, peers, and school. Thus, media inuence can explain a part of the variation in respondents sex-related knowledge, attitudes, and behaviors. Secondly, as Taiwan, China mainland, and Vietnam have experienced different social and economic transitions despite sharing similar cultural backgrounds, media development, adolescents and young adults exposure to media, and the relationship between media exposure and adolescents and young adults reproductive health status will be different in the three settings. Thus, we hypothesize that the associations between access to media and adolescents and young adults sex-related knowledge, attitudes, and behaviors will be different in these three Asian cities. Methods Sample and procedures Data for this article come from a 2006 cross-sectional survey of 17,016 male and female, married and unmarried adolescents and young adults, aged 1524 years, conducted in urban Hanoi, Shanghai, and Taipei and rural areas included in their large metropolitan districts by a team of researchers from the Johns Hopkins Bloomberg School of Public Health, the Population and Health Research Center in Taiwans Bureau of Health Promotion, the Shanghai Institute for Planned Parenthood Research and the Hanoi Institute for Family and Gender Studies. In this article, only the 16,554 unmarried respondents were included in the analysis. The sampling methodology has been described in detail in Levels of Change in Adolescents and Young Adults Sexual Behavior in three Asian Cities [22]. Multistage sampling methods were used to ensure representativeness within each city. In Hanoi and Shanghai, both private residences and group living facilities were sampled. In Taipei, students were interviewed in school, with a small nonstudent subsample interviewed at their private residences and group living facilities. The survey was developed by the research team, translated, back-translated, and pilot tested in each site. Interviewers received extensive training. Most of the interview was conducted face-to-face, except that computer-assisted self-interview was used for sensitive questions. All aspects of this study received approval from the Committee on Human Research at the Johns Hopkins University as well as the collaborating local organizations.

S28

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

Measures There are three groups of measures, that is, independent variables or media exposure variables, control variables, and dependent or outcome variables (these were measured by the scores for responses on sex-related knowledge, premarital sexual permissiveness (PSP), and sex-related behaviors). Independent or media exposure variables. The media exposure variables include four indicators, namely, (a) exposure to sexual content in the traditional media, (b) exposure to sexual content on the Internet, (c) exposure to pornographic movies/videos, and (d) origin of preferred movies/videos. Exposure to sexual content in the traditional media. There are four kinds of information/knowledge including AIDS/STDs, sex, pregnancy, and contraception which were asked about in the questionnaire and dened as sexual content. Five types of media are dened as traditional medianamely, radios/televisions/videos, newspapers/magazines, books, bulletins/educational columns, and telephone hotlines. As traditional media in China are unlikely to convey much of sex in its content, we are looking largely at informational traditional media, save for porn. Exposure was assessed by asking respondents whether they had learned of each of the aforementioned sexual information/knowledge subjects from each of aforementioned media sources (yes 1, no 0). The responses were summed and yielded a score ranging from 0 (never heard of any sexual content from any traditional media) to 20 (have heard about all four kinds of sexual information from all ve media types). The score was divided into two groups (zero and above zero) in bivariate analysis and was treated as a continuous variable in the multivariate analysis. Cronbachs alphas for items composing these variables are .85 in Hanoi, .88 in Shanghai, and .89 in Taipei. Exposure to sexual content on the Internet. Similarly, respondents were asked whether they had learned the four kinds of information listed before in the text from the Internet (yes 1, no 0). The responses varied from 0 (never heard of any information from the Internet) to 4 (have heard about all four kinds of sexual information from the Internet). In the analysis, the score was treated in the same way as knowledge from the traditional media. The Cronbachs alphas of items constituting these variables in Hanoi, Shanghai, and Taipei are .88, .90, and .86, respectively. Exposure to pornographic videos. Respondents were asked whether they had watched any pornographic movies/digital versatile discs or seen pornographic pictures or videos. Responses were scored 0 for no and 1 for yes. Origin of preferred movies/videos. Respondents were asked where most of the movies/videos they like to watch come from. The answers were coded as follows: Asian area 1, western countries 2, no preference 3. Control variables. A variety of other potential contextual determinants of adolescents and young adults sex-related knowledge, attitudes, and behaviors including individual, family, peer, and school variables were treated as covariates. Table 1 presents the denition and coding of these variables.

Dependent or outcome measures. There are three main outcome measures: (a) a sex-related knowledge score, (b) a PSP score, and (c) a sex-related behavior score. Sex-related knowledge score. A total of 32 questions related to reproductive health knowledge (4 on reproduction, 22 on contraception, and 6 on STDs/AIDS) were used to evaluate each respondents knowledge level. A correct answer was credited with a score of 1 and incorrect answer with a score of 0. The sum of the scores for the 32 questions was calculated, the maximum being 32; then the score was converted into a new one with a maximum of 100 (the value of the sum was divided by 32 and multiplied by 100). The higher the score, the greater the respondents sex-related knowledge. The Cronbachs alphas for items composing the knowledge score are .71 in Hanoi, .84 in Shanghai, and .60 in Taipei. Premarital sexual permissiveness score. The score was obtained by creating male and female Guttman scales based on questions concerning the individuals acceptance of various physical acts (kissing, fondling, or coitus) under various conditions of affection/familiarity (with casual acquaintance, with boyfriend/girlfriend, or with anc/ance). These questions were modied from the PSP scale developed by Reiss (1967), which was widely applied in measuring adolescents and young adults PSP in western countries, Taiwan, Japan, and Singapore [2326]. The items for the male and female scales are similar except that the sex referent is female for the female scale and male for the male scale. A one-dimensional measure of PSP was created within each city for each gender using a frequency analysis of the items mentioned before. The rankings were based on the percentage of acceptable responses for each act/familiarity combination. Items with similar percentages of acceptable responses were combined into a single group. The lower the percentage of acceptable responses an item has, the higher its rank, and therefore the more permissive a respondent who nds that behavior acceptable is considered to be. The analysis showed that both the male and female scales in all three cities were similarly ordered from less to more permissive the rankings are (0) disagree with all premarital sexual contact, (1) agreement with kissing with boyfriend/anc (girlfriend/ance) or fondling with anc (ance), (2) fondling with boyfriend (girlfriend), (3) coitus with anc (ance), (4) coitus with boyfriend (girlfriend), (5) kissing with casual acquaintance, (6) fondling with casual acquaintance, and (7) coitus with casual acquaintance. Through the Guttman scaling method, these items were arranged in an order so that an individual who agreed with a particular item was assumed to also agree with items ranked sequentially lower [27]. All coefcients of reproducibility for the Guttman scaling in the three cities are higher than .96. Sex-related behavior score. Respondents sex-related behavior level was evaluated by ve questions on intimacy behaviors (holding hands, hugging, kissing, fondling, and sexual intercourse) in the questionnaire. Respondents were asked whether they had participated in each of the ve behaviors. A Guttmantype scale was used to construct the sex-related behavior score through ranking the items with the proportion of yes within each site. Participants received scores of 15 for holding hands, hugging, kissing, fondling, and sexual intercourse, respectively, reecting the highest level of behavior experienced; adolescents and young adults who reported not having participated in any of

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

S29

Table 1 Contextual and control variables Variable Individual Age Gender Educational level Description of variable/question/statement Range & value

Age of respondent Gender of respondent Student: current educational level Nonstudent: highest educational level

Student Economic status Residence Engagement Family Live with parents

Do you currently go to school? Quintile of the sum of 14 items of assets in current dwelling Urban or rural area Are you currently engaged? Whether you live with biological parent most of the time?

1524 years 0 female, 1 male 1 high school or lower 2 college/vocational school (Hanoi) 3 university or higher 0 no, 1 yes Quintile within each site 0 rural, 1 urban 0 no, 1 yes 1 live with neither 2 live with father 3 live with mother 4 live with both Item 12: 1 not at all, 2 a little, 3 some, 4 a lot Item 310: 1 never, 2 sometimes, 3 often, 4 always Sum range: 10 40 1 disagree, 2 no opinion, 3 agree For each of the four items, 0 no, 1 yes Sum range: 0 4 For each of the four questions, 0 no, 1 yes Sum range: 0 4

Family support score at 13 or 14 years old

Perceived parental upset view of sex Learned sexual content from parents

Would seek help from parents if necessary

Item 12: discussed problems with father/mother; Item 3 4: he/she was interested in you and how you were doing; Item 5 6: he/she believed that you would be successful; Item 7 8: he/she expected you to do your best; Item 9 10: he/she showed you that he loved you (Cronbachs alpha: Hanoi, .80; Shanghai, .86; Taipei, .80) My parents would be very upset if I had sexual intercourse before marriage Learned from parents about AIDS/STDs, sex, pregnancy, and contraception information, respectively (Cronbachs alpha: Hanoi, .74; Shanghai, .78; Taipei, .77) Would you talk to parents if you felt like killing yourself, were being physically abused or threatened, thought you might have a sexually transmitted disease, or needed contraception, (respectively)? (Cronbachs alpha: Hanoi, .71; Shanghai, .67; Taipei, .60) Hours spent in hanging around with friends in a normal week Perceived most of close friends have had sexual intercourse Do you need to have premarital sex to t in with your friends? Item 1 4: Learned from friends about AIDS/STDs, sex, pregnancy, and contraception information, respectively Item 5 8: Learned from boy (girl) friends about AIDS/STDs, sex, pregnancy, and contraception information, respectively Item 9 12: Learned from young relatives about AIDS/STDs, sex, pregnancy, and contraception information, respectively (Cronbachs alpha: Hanoi, .82; Shanghai, .85; Taipei, .86) Item 1 4: Would you talk to friends if you felt like killing yourself, were being physically abused or threatened, thought you might have a sexually transmitted disease, or needed contraception, (respectively)? Item 5 8: Would you talk to boy (girl) friends if you felt like killing yourself, were being physically abused or threatened, thought you might have a sexually transmitted disease, or needed contraception, (respectively)? (Cronbachs alpha: Hanoi, .55; Shanghai, .71; Taipei, .68) Item 1: Students are encouraged to ask questions and speak out in school; Item 2: Teachers treat every student fairly; Item 3: Teachers generally set high expectations of students; Item 4: Teachers provide students with support and encouragement. (Cronbachs alpha: Hanoi, .59; Shanghai, .70; Taipei, 0.51) Grades received as a student Learned from schools about AIDS/STDs, sex, pregnancy, and contraception information, respectively (Cronbachs alpha: Hanoi, .82; Shanghai, .83; Taipei, .78) Would you talk to teachers if you felt like killing yourself, were being physically abused or threatened, thought you might have a sexually transmitted disease, or needed contraception, (respectively)? (Cronbachs alpha: Hanoi, .47; Shanghai, .67; Taipei, .68)

Peers Time spent with peers weekly Perceived peers sexual behavior Perceived peer pressure on sexual behavior Learned sexual contents from peers

0 hour, 2 16 hours, 3 hours 0 no, 1 yes 0 no, 1 yes For each item, 0 no, 1 Sum range: 0 12 yes

Would seek help from peers if necessary

For each item, 0 Sum range: 0 8

no, 1

yes

School School support score

disagree, 2 neither, 3 agree Sum range: 4 12

Grades Learned sexual content from school

1 poor, 2 average, 3 0 no, 1 yes Sum range: 0 4 0 no, 1 yes Sum range: 0 4

good

Would seek help from teachers if necessary

S30

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

Table 2 Respondents exposure to sexual content in the media by city and gender (%) Media variables Hanoi Shanghai Taipei

Male Female Total Male Female Total Male Female Total (n 3,095) (n 3,109) (n 6,204) (n 2,983) (n 3,040) (n 6,023) (n 2,168) (n 2,159) (n 4,3270) Learned about sex from the Internet Learned about sex from traditional media Watched porn Origin of preferred movies/videos Asian Western No preference 57.0* 97.9* 43.3* 67.9* 19.1 13.0 45.3 98.7 6.8 76.9 11.7 11.4 51.4** 98.3** 26.0** 72.2** 15.6 12.2 63.7* 85.7* 56.7* 61.5* 20.8 17.7 48.3 88.2 22.2 70.2 13.6 16.2 56.0 86.9 39.6 65.8 17.3 16.9 84.4* 94.3* 94.1* 20.5 56.6 22.9 80.7 97.2 77.1 21.5 55.8 22.7 82.6 95.7 85.8 21.0 56.2 22.8 .01.

Chi-square subdivision method was used for multiple comparison on proportions between Hanoi and Shanghai, Shanghai and Taipei, Hanoi and Taipei, and all p * p .01, Chi-square test between genders. ** p .01, Cochran-Mantel-Haenszel test among sites.

these behaviors were scored 0. The coefcients of reproducibility for the scale in three cities are higher than .97. Statistical analysis The sample was weighted before the analysis, and the weights were calculated according to the probability of each respondent being selected from each sample site. Chi-square tests were used to detect differences in the distribution of demographic characteristics (age-group, gender, education level, etc.) and exposure to media sexual content between genders within each city. The Mantel-Haenszel test was used to detect differences in exposure to media sexual content among cities after controlling for gender. The Wilcoxon-Mann-Whitney test was adopted to compare respondents sex-related knowledge score, PSP score, and their sex-related behavior score between genders. The Kruskal-Wallis test was used to compare respondents sexrelated knowledge, attitudes, and behavior scores among the cities after controlling for gender. Finally, multiple linear regressions were used to compare the impact of each of the four contexts (family, peer, school, and media) on the outcome measures, controlling for individual variables. In the rst set of analyses, each set of contextual factors was entered into a model with each outcome variable. Then, all factors of individual, family, peer, and school were simultaneously entered into one model, and nally, media variables were included in the model. The change in adjusted R2 between the two models (with and without media variables) was used to indicate the contribution of media factors after controlling for other contextual factors. A Cronbachs alpha coefcient higher than .60 indicates an acceptable internal consistency of items making up the composite variables. A coefcient of reproducibility higher than .90 for a Guttman-type scale is considered an acceptable reproducibility for Guttman scaling. In general, the Cronbachs alpha coefcients for composite variables and the reproducibility coefcients for Guttman Scaling within each site in the article meet the minimum level of acceptance. Data were analyzed with SAS 9.1 (SAS Institute Inc., Cary, NC). Cases with missing values were excluded from the analysis. Results Demographic characteristics of respondents The sample for this study includes 16,554 adolescents and young adults (6,204 from Hanoi, 6,023 from Shanghai, and 4,327

from Taipei). The proportion of Shanghai respondents aged 2124 years (31%) was less than that of Taipei and Hanoi (40%). Signicant differences were also observed between the three cities in terms of the background variables of gender, educational level, current student status, economic status, urban/rural residence, and whether the respondent is engaged (p .05). These factors were then controlled for in the multivariate analysis. All the following analyses were stratied by city and gender. Status of respondents exposure to sexual content in the media More than half of respondents in three cities reported having learned about sex from the Internet (45% 84%); the highest percentage was in Taipei, the lowest in Hanoi, and the middle in Shanghai. In all three cities, more respondents reported having learned about sex from the traditional media (86%99%) than from the Internet. The percentage was higher in Hanoi than that in Taipei and Shanghai. A relatively high proportion of Taipei respondents reported having watched pornographic movies or videos (86%), compared with respondents in Shanghai and Hanoi. Most respondents in Hanoi (72%) and Shanghai (66%) preferred Asian movies/videos, whereas more than half of Taipei respondents expressed a preference for western movies/videos. The differences of exposure to sexual content in the media within each of the sources mentioned before were statistically signicant between cities (p .01). Gender differences were also observed within each site, that is, more males than females learned about sex from the Internet, more females than males learned about sex from the traditional media, and considerably more males than females watched pornographic videos (p .01). In Hanoi and Shanghai, signicantly more males than females preferred western movies/videos (p .01) (Table 2). Respondents sex-related knowledge, attitudes, and behaviors Table 3 presents the median scores of respondents sexrelated knowledge, attitudes, and behaviors by city and gender. The sex-related knowledge and the sex-related behavior scores were the highest in Taipei. Shanghai respondents had the lowest knowledge score. The PSP score was the highest for Shanghai respondents, followed by those in Taipei and was the lowest for respondents in Hanoi. All the differences between cities were statistically signicant (p .01). There was also evidence of gender differences in each city. In Hanoi, males had higher sexrelated knowledge, more permissive attitudes to premarital intimacy behaviors, and higher sex-related behavior scores than

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

S31

Table 3 Median score of respondents sex-related knowledge, attitudes, and behaviors by city and gender Reproductive health status Hanoi Male (n 3,095) Sex-related knowledge score (0100) Premarital sexual permissiveness score (07) Sex-related behavior score (05) 56.3* 4.0* 2.0* Female (n 3,109) 53.1 3.0 1.0 Total (n 6,204) 56.3a,** 3.0a,** 1.0a,** Shanghai Male (n 2,983) 43.8 5.0* 2.0* Female (n 3,040) 43.8 4.0 1.0 Total (n 6,023) 43.8b 5.0b 1.0b Taipei Male (n 2,168) 59.4* 5.0* 3.0 Female (n 2,159) 65.6 4.0 3.0 Total (n 4,327) 62.5c 4.0b 3.0c

Student-Newman-Keul test for multiple comparison across sites, values with the same alphabet indicators are not signicantly different. * p .01, Wilcoxon-Mann-Whitney test between genders. ** p .01, Kruskal-Wallis test among sites.

did females (p .01). In Shanghai, males had more open attitudes to premarital sex and were more sexually active than were females (p .01), but they had a similar knowledge level. In Taipei, females had a higher level of sex-related knowledge, and males once again had more liberal attitudes toward sexually intimate behaviors before marriage (p .01), but they had similar sexrelated behavior levels. Comparison of the effects of media, family, peer, and school on outcome variables Media inuences showed a consistent and signicant association with respondents sex-related knowledge, permissive attitudes toward premarital sexual intimacy, and sex-related behaviors in each site (Table 4). In Hanoi, the proportions of variance that media variables explained in male and female respondents sex-related knowl-

edge and permissive attitudes to premarital sexual behaviors are higher than those that family, peer, or school variables explained. The variance accounted for by media variables in male and female respondents sex-related behavior is lower than the variance accounted for by peer variables, but higher than that by family or by school variables. After controlling for other contextual factors, media still added 2%9% of variance to the prediction of male and female respondents sex-related knowledge, PSP, and sex-related behaviors. All contextual factors, including media, explained about 40% of the variance in sexual knowledge, 8% of the variance in sexual permissive attitudes, and 32% of the variance in sex-related behaviors of female and male respondents (Table 4). In Shanghai, as in Hanoi, media variables explained more variance in male and female respondents sex-related knowledge and permissive attitudes toward premarital sexual intimacy than did family, peer, or school variables, and more variance in

Table 4 Percent variance in sex-related knowledge, attitudes, and behaviors explained by each context, after controlling for individual variables (%) Sites Independent variables (R2) Individual Sex-related knowledge score Hanoi Male Female Shanghai Male Female Taipei Male Female Premarital sexual permissiveness score Hanoi Male Female Shanghai Male Female Taipei Male Female Sex-related behavior score Hanoi Male Female Shanghai Male Female Taipei Male Female Family Peer School Media Variables without media All variables Change in R2

18.4 19.6 18.2 20.9 16.9 11.7

1.1 1.1 5.4 1.9 2.7 .8

9.7 6.7 17.2 14.5 3.9 6.8

5.1 5.2 2.0 3.9 4.9 4.0

16.2 12.9 23.9 17.4 13.8 12.9

31.2 30.3 38.6 38.8 26.3 23.2

39.9 36.7 49.5 46.9 34.1 30.3

8.7 6.4 10.9 8.1 7.8 7.1

1.7 2.7 5.4 4.9 3.9 2.0

3.3 1.2 3.3 .9 6.1 3.5

2.5 2.1 7.4 8.4 4.0 6.8

1.1 .5 .5 1.9 1.1 .6

4.4 2.8 13.2 8.7 3.1 5.3

7.2 6.1 14.9 14.6 13.1 11.6

9.2 7.8 21.8 18.5 14.9 13.9

2.0 1.7 6.9 3.9 1.8 2.3

20.3 23.1 18.2 16.9 16.6 12.2

1.5 .5 3.5 1.3 2.8 3.7

8.6 7.2 15.3 17.9 17.1 23.8

.5 .2 .8 1.0 .9 1.8

4.6 4.4 12.8 11.3 3.3 4.8

30.0 30.7 34.6 35.6 36.8 39.2

31.6 32.7 39.5 39.5 39.1 40.6

1.6 2.0 4.9 4.0 2.3 1.4

S32

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

Table 5 Variables regressed on knowledge score by city and gender: standardized Variable Hanoi Male (n 2,619) Individual Age Educational level (ref: HS or lower) Student or not (ref: nonstudent) Economic status (ref: bottom quintile) Urban/rural type (ref: rural) Engaged or not (ref: no) Media exposure Learned about sextraditional media Learned about sexthe Internet Watched porn (ref: no) Preferred movies/videos (ref: Asian) Western No preference Family Whether lived with parents (ref: live with neither) Live with father Live with mother Live with both Perceived family support Perceived parental upset view of sex (ref: disagree) No opinion Agree Learned about sex from parents Would seek parental help when necessary Peer Time spent with peers weekly (ref: 0 hour) 16 hours More than 6 hours Perceived peers sexual behavior (ref: no) Perceive peers pressure on sexual behavior (ref: no) Learned about sex from peers Would seek peers help when necessary School Perceived school support Received good grades as a student Learned about sex from teachers Would seek teachers help when necessary Female (n 2,556) Shanghai Male (n 2,916) Female (n 2,987) Taipei Male (n 2,073) Female (n 2,040)

.12* .15* .06* .01 .02 .01 .29* .01 .12* .05* .01

.21* .13* .01 .03 .02 .01 .25* .01 .11* .05* .03*

.10* .11* .08* .02 .01 .01 .24* .08* .17* .05* .03*

.15* .07* .06* .05* .02 .05* .20* .06* .11* .09* .05*

.12* .09* .11* .04* .09* .00 .21* .08* .11* .10* .07*

.00 .20* .05* .05* .04* .03 .22* .03 .11* .10* .04

.01 .00 .01 .03* .04 .00 .03* .02

.02 .06* .05 .02 .05 .08* .02 .01

.04* .04* .01 .12* .07* .06* .01 .06*

.00 .00 .03 .10* .02 .00 .02 .07*

.05* .02 .02 .06* .11* .07* .01 .06*

.02 .03 .01 .08* .03 .00 .05* .09*

.04 .06* .07* .02 .12* .11* .02 .00 .10* .02

.01 .01 .06* .05* .10* .04* .02 .04* .14* .01 reference group.

.02 .03 .12* .02 .14* .09* .03* .04* .05* .01

.01 .03 .08* .01 .19* .10* .07* .06* .12* .01

.04 .05 .08* .02 .01 .03 .05* .03 .16* .02

.05 .04 .09* .02 .06* .04* .02 .05* .20* .02

Variables without indicating reference are treated as continuous variables. ref * p .05.

male and female respondents sex-related behaviors than family and school variables after controlling for individual variables. When all other context factors were included in the model, media inuences still made a signicant contribution, adding 4%11% of variance to the prediction of male and female respondents sex-related knowledge, PSP, and sex-related behaviors. All context variables, including media, jointly explained nearly 50% variance in sex-related knowledge, 20% variance in PSP attitudes, and 40% variance in sex-related behaviors of female and male respondents (Table 4). In Taipei, after controlling for individual variables, media variables explained higher variance than family, peer, or school variables in male and female respondents sex-related knowledge; explained higher variance than school variables in males and females PSP and higher variance than family variables in females PSP. In male and female respondents sex-related behaviors, although considerably lower than the variance explained by peer variables, the effect is nevertheless higher than those of family and school variables. After controlling for other context factors, media variables contribute 1.4% 8% variance in male and female respondents sex-related knowledge, PSP, and

sex-related behaviors. All the variables, including individual, family, peer, school, and media variables, jointly explained over 30% of the variance in sex-related knowledge, nearly 15% variance in PSP, and 40% in sex-related behaviors of female and male respondents in Taipei. Tables 57 present the associations between media exposure variables and sex-related knowledge, attitudes, and behaviors after adjusting for the effects from other context variables by city and gender. The results show that media variables are signicant contributors across the dependent variables. Irrespective of whether they are male or female, adolescents and young adults in each city who learned something about sex from the traditional media have higher knowledge levels and lower sexrelated behavior levels. Regarding learning about sex from the Internet, Shanghai respondents and Taipei male respondents who learned about sex from the Internet are also more knowledgeable about sex-related subjects; Shanghai and Hanoi respondents who learned about sex from the Internet had higher levels of sex-related behaviors. Shanghai respondents as well as Hanoi and Taipei males who learned about sex from the Internet reported more permissive attitudes to premarital sexual inti-

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

S33

Table 6 Variables regressed on premarital sexual permissiveness score by city and gender: standardized Variable Hanoi Male (n 2,619) Individual Age Educational level (ref: HS or lower) Student or not (ref: nonstudent) Economic status (ref: the bottom quintile) Urban/rural type (ref: rural) Engaged or not (ref: no) Media exposure Learned about sex from traditional media Learned about sex from the Internet Watched porn (ref: no) Preferred movies/videos (ref: Asian) Western No preference Family Whether lived with parents (ref: live with neither) Live with father Live with mother Live with both Perceived family support Perceived parental upset view of sex(ref: disagree) No opinion Agree Learned about sex from parents Would seek parental help when necessary Peer Time spent with peers weekly(ref: 0 hour) 16 hours More than 6 hours Perceived peers sexual behavior(ref: no) Perceived peers pressure on sexual behavior (ref: no) Learned about sex from peers Would seek peers help when necessary School Perceived school support Received grades as a student Learned about sex from teachers Would seek teachers help when necessary Female (n 2,556) Shanghai Male (n 2,916) Female (n 2,987) Taipei Male (n 2,072) Female (n 2,040)

.01 .05 .01 .04 .03 .04 .03 .08* .12* .03 .01

.01 .01 .03 .07* .04* .03 .01 .01 .14* .04* .00

.02 .04 .00 .01 .05* .03 .04 .11* .23* .05* .01

.01 .01 .11* .02 .03 .01 .04 .10* .15* .05* .02

.01 .07* .01 .05* .07* .01 .02 .09* .10* .01 .06*

.01 .01 .01 .01 .05* .01 .05 .00 .15* .02 .03

.02 .08 .07 .03 .07* .15* .01 .08*

.01 .14* .11* .04 .04 .02 .00 .01

.01 .03 .00 .03 .10* .12* .01 .06*

.01 .00 .04 .03 .00 .05 .01 .04

.01 .00 .03 .01 .13* .23* .04 .10*

.01 .04 .05* .02 .14* .25* .09* .04

.04 .01 .07* .02 .07* .03 .05* .03 .02 .01

.05 .04 .04* .05* .04 .04* .02 .00 .06* .04 reference group.

.03 .08* .05* .03 .09* .02 .01 .01 .01 .02

.06* .02 .03 .03 .13* .12* .06* .03 .07* .01

.01 .05 .06* .05* .08* .03 .01 .04 .02 .04

.03 .03 .14* .01 .10* .04 .02 .02 .03 .02

Variables without indicating reference are treated as continuous variables. ref * p .05.

macy; however, learning about sex from traditional media had no signicant effect on PSP. Having watched pornographic videos was signicantly associated with higher sex-related knowledge, greater PSP, and a higher level of sex-related behaviors across cities and genders. As to a preference for movies/videos of a specic origin, adolescents and young adults in the three cities who preferred western videos were more likely to have a higher sex-related knowledge level. Shanghai respondents and Hanoi female respondents who preferred western movies/videos were more likely to have more permissive attitudes toward sexual intimacy before marriage, and Shanghai and Hanoi female respondents and Taipei male respondents who preferred western movies/videos reported a higher level of sex-related behaviors. In summary, there are differences in the associations between access to media and adolescents and young adults reproductive health outcomes between the three cities and between genders. Discussion The ndings of this study show that using media to access sex-related information is prevalent among adolescents and

young adults in these three cities. Nearly all respondents report having learned about sex from traditional media, and a majority of respondents learned about sex from the Internet. The survey results are consistent with other studies conducted among Asian adolescents and young adults, which indicate that the various forms of media have been an important source of information for Asian adolescents and young adults in learning about sex [16,17]. The results of this study show that media exposure is signicantly correlated to adolescents and young adults sex-related knowledge, attitudes, and behaviors. Even after considering inuences from other important socialization sources, such as family, peer, and school, media impacts on the knowledge, attitudes, and behaviors of adolescents and young adults are signicant. These ndings are similar to those of research conducted in western countries [9]. Moreover, our study shows that the effects of media are also comparable with other environmental factors. For sex-related knowledge, the impact of media variables is higher than that of family, peer, and school variables. For sexrelated attitudes, the impact of media variables is similar to that of peer variables and higher than that of school variables. In

S34

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

Table 7 Variables regressed on sex-related behavior score by city and gender: standardized Variable Hanoi Male (n 2,619) Individual Age Educational level (ref: HS or lower) Student or not (ref: nonstudent) Economic status (ref: bottom quintile) Urban/rural type (ref: rural) Engaged or not (ref: no) Media exposure Learned about sex from traditional media Learned about sex from the Internet Watched porn (ref: no) Preferred movies/videos (ref: Asian) Western No preference Family Whether lived with parents (ref: live with neither) Live with father Live with mother Live with both Perceived family support Perceived parental upset view of sex(ref: disagree) No opinion Agree Learned about sex from parents Would seek parental help when necessary Peer Time spent with peers weekly(ref: 0 hour) 16 hours More than 6 hours Perceived peers sexual behavior(ref: no) Perceived peers pressure on sexual behavior (ref: no) Learned about sex from peers Would seek peers help when necessary School Perceived school support Received grades as a student Learned about sex from teachers Would seek teachers help when necessary Female (n 2,556) Shanghai Male (n 2,916) Female (n 2,987) Taipei Male (n 2,073) Female (n 2,040)

.29* .02 .04 .06* .03 .02 .04 .04* .15* .00 .01

.36* .01 .00 .08* .01 .02 .06* .04* .14* .04* .02

.21* .07* .07* .08* .01 .03 .01 .05* .23* .01 .04*

.21* .04* .01 .02 .03 .08* .05* .07* .20* .05* .00

.26* .02 .06* .03 .02 .01 .09* .05 .09* .11* .07*

.29* .11* .02 .04* .01 .02 .10* .02 .11* .02 .00

.03 .01 .02 .05* .06 .07* .03 .06*

.03 .01 .02 .01 .04 .04 .04* .04*

.01 .00 .04* .03 .04* .08* .01 .03

.01 .01 .01 .00 .03 .06* .02 .06*

.03 .02 .04 .04 .03 .06* .05* .04*

.01 .01 .04* .02 .01 .04 .11* .10*

.04 .11* .17* .04* .10* .03 .05* .01 .04* .03

.09* .11* .13* .02 .14* .07* .00 .01 .03 .03 reference group.

.08* .14* .14* .02 .18* .03* .01 .00 .03 .02

.06* .11* .09* .04* .25* .10* .02 .02 .04* .03

.07* .10* .24* .04* .26* .04* .00 .02 .11* .01

.04 .05* .22* .05* .29* .04* .04* .01 .04 .01

Variables without indicating reference are treated as continuous variables. ref * p .05.

terms of sex-related behaviors, the impact of media variables is only lower than that of peer effects, but higher than the impact of family and school. The various types of media present a great deal of sexual information in a compelling and easily accessible format. For adolescents and young adults with a heightened romantic and sexual interest during their sexual development, the various forms of media provide a relatively safe and minimally embarrassing way to learn about sex, especially in the Asian region, where talking about sex with adolescents and young adults is still a source of discomfort to educators and parents. Thus, media may serve as a kind of sexual super peer for adolescents and young adults seeking information about sexuality [15,28,29]. However, much of the sexual content in the media describes attractive sexual behavior but rarely depicts the negative consequences of that behavior, which may make adolescents and young adults become more permissive regarding sex and encourage experimentation and imitation [9]. Our research shows that learning about sex from the Internet, watching pornographic videos/movies, and preferring western movies/videos are associated with more permissive attitudes to premarital sex-

ual intimacy and a higher level of sex-related behaviors, ndings consistent with previous research [30 32]. However, in contrast with previous ndings [21,33], the study demonstrates that learning about sex from the traditional media (including radio/ television/videos, newspaper/magazines, books, bulletin/educational column, and telephone hotlines) was associated with respondents lower level of sex-related behaviors. One possible reason is that traditional forms of media are more easily supervised and monitored compared with the Internet; for example, in China, most of the traditional media types are controlled by the government. Thus, traditional media are more likely to transmit scientic and healthy sexual information to adolescents and young adults. Because this study did not collect detailed information on adolescents and young adults exposure to the specic sexual content of the various forms of media, more research is needed to explore the differences in the sexual content that Asian adolescents and young adults receive from the traditional media as opposed to the Internet. National, regional, and gender differences have been observed in many aspects in this study. In terms of media exposure, more Taipei respondents learned about sex from the Internet and

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

S35

watched pornographic videos than did those in Hanoi and Shanghai. Additionally, more Taipei respondents preferred western movies/videos, whereas those in Hanoi and Shanghai preferred Asian movies/videos. More males than females learned about sex from the Internet and watched pornographic movies/ videos. Respondents in Taipei also had the highest sex-related knowledge and the highest level of sexual activity. Across all three cities, male and female respondents had similar sexrelated knowledge levels, but male respondents had more permissive attitudes to premarital sexual intimacy, and in Hanoi and Shanghai a higher level of sexual behaviors. Concerning the inuence of contextual factors on respondents sex-related attitudes, among Hanoi and Shanghai respondents, media factors explained more variance than did the other contextual factors (peer, school, and family); among Taipei male respondents, family factors explained more variance than the other factors, and for Taipei female respondents, peer factors explained more variance than the other factors. The results of gender differences in this study are consistent with previous research [34]. Taipei is situated in one of the most developed areas of Asia, where rapid media development and broad contact with the outside world provide adolescents and young adults with greater opportunity to be exposed to all kinds of information in the media. This may contribute to the high levels of sexual knowledge and sex-related behaviors found in adolescents and young adults in Taipei compared with their counterparts in the other two cities. The more conservative nature of the societies of Hanoi and Shanghai may result in reduced communication between adolescents and young adults and their parents, peers, or teachers concerning sexual topics, and media consequently becomes a more important source of sexual information. This could explain the high impact of media on PSP attitudes in Hanoi and Shanghai but not on those in Taipei. Regarding the diversity of associations between different contextual factors and outcome variables, more research is needed to identify and explore the underlying causes. The specic sexual content of the media is one area in need of much more investigation. This study is limited in several ways. First, the data presented here are cross-sectional, and thus, the direction of causal inuence is unclear. The process may even be a bidirectional one. For example, sexually active adolescents and young adults may be more likely to seek pornographic videos and western movies to t their preferred lifestyle, or conversely, exposure to pornographic videos and western movies may make young people more predisposed to engage in sex. In terms of the latter, longitudinal studies have found that increased exposure to sexual content on television predicts earlier initiation of sexual intercourse among adolescents and young people [21]. Second, the measurement of media exposure is not precise: for example, the frequency of access to sexual content in the mass media is not included in the questionnaire, which limits the depth of the analysis in this study. Finally, we failed to explore the possible differences in the sexual content of the media between cities that the adolescent respondents received because the questionnaire was not designed for this purpose. Despite these limitations, several policy/research implications can be identied from our ndings. First, media play as important a role in adolescents and young adults sex-related knowledge, attitudes, and behaviors as do other contextual factors. Considering the reach of media is broader than that of other environmental variables, media inuences on adolescents and young adults sexual development should be a concern of re-

searchers and health practitioners. Regional and gender differences of these inuences should also be considered. Second, using the mass media to provide sexual and reproductive health education for adolescents and young people should be encouraged because media have proven to be a popular source of sexual information for adolescents and young adults. Third, many adolescents and young adults, irrespective of gender, learn about sex from the Internet and watch pornography, and these have been found in this study to be associated with more permissive sexual attitudes and higher levels of sexual behavior. Therefore, it is necessary to help adolescents and young people learn how to evaluate what they get from the media by the introduction of successful communication programs between teachers and students, and between parents and children. Finally, the preference for western movies/videos is associated with adolescents and young adults greater cognizance of sexual attitudes and behaviors, which implies the necessity of preventing adolescents and young adults from passively absorbing imported sexual information on the one hand, and, on the other, of being encouraged to use constructively the potential benets of introducing popular western movies/videos with accurate sexual information. Acknowledgments The authors would like to acknowledge and thank the Bill and Melinda Gates Institute for Population and Reproductive Health, Johns Hopkins Bloomberg School of Public Health, United States, who provided nancial support for the study. The authors are greatly indebted to all those who have made this article possible. The authors wish to acknowledge Dr. David Bishai and Prof. Robert Blum for their hard work and important help in the study design, sampling, data cleaning, and article editing. The authors would also like to sincerely thank our Vietnam, Taipei, and Shanghai colleagues for their substantial contribution in the collection of these data. References
[1] Gao ES, Tu XW, Zhao SL, et al. Article review on reproductive health of Chinese adolescents and unmarried youth. Beijing, China: China Population Press, 2002. [2] Lou CH, Wang XJ, Cheng Y, et al. Relationship with opposite sex and sexual behavior and inuencing factors of them among vocational school students. Chin J Public Health 2007;23:985 6. [3] Lou CH, Shen Y, Gao ES, et al. Sex-related behaviors among unmarried migrant young people. Reprod Contracept 2004;24:34 8. [4] Kim Choe M, Lin HS. Effect of education on premarital sex and marriage in Taiwan. Honolulu, Hawaii: East-West Center, 2001; East-West Center Working Papers, Population Series, No.108-16. [5] Wang RH, Wang HH, Hsu MT. Factors associated with adolescent pregnancya sample of Taiwanese female adolescents. Public Health Nurs 2003;21:33 41. [6] Mensch BS, Clark WH, Anh DN. Premarital sex in Vietnam: Looking beyond reproductive health. Stud Fam Plan 2003;34:249 63. [7] de Lind van Wijngaarden JW. Exploring factors and processes leading to HIV risk among the most vulnerable children and adolescents in Vietnam (literature review). Hanoi, Vietnam: UNICEF, 2006. [8] Kotchick BA, Shaffer A, Forehand R, Miller KS. Adolescent sexual risk behavior: A multi-system perspective. Clin Psychol Rev 2001;21:493519. [9] LEngle KL, Brown JD, Kenneavy K. The mass media are an important context for adolescents sexual behavior. J Adolesc Health 2006;38:186 92. [10] Brown JD, Cantor J. An agenda for research on youth and the media. J Adolesc Health 2000;27(Suppl 2):27. [11] Rideout V. Generation M2: Media in the lives of 8- to 18-year-olds. Menlo Park, CA: Kaiser Family Foundation, 2010. [12] China Internet Network Information Center. The 19th statistical report on the development of Internet Network in China. Available at: http://www. cnnic.cn/html/Dir/2007/01/22/4395.htm. Accessed Jan 23 2007. [13] Ko CH, Yen JY, Chen CS, et al. Psychiatric comorbidity of internet addiction in college students: An interview study. CNS Spectr 2008;13:14753.

S36

C. Lou et al. / Journal of Adolescent Health 50 (2012) S26 S36

[14] Ngo AD, Ross MW, Ratliff EA. Internet inuences on sexual practices among young people in Hanoi, Vietnam. Cult Health Sex 2008;10(Suppl):20113. [15] Strasburger VC, Wilson BJ, Jordan AB. Children, adolescents, and the media, 2nd edition. Thousand Oaks, CA: Sage, 2009. [16] Nguyen VT. Reproductive health education in school: Situation and challenge. Hanoi, Vietnam: National Committee for Demographic and Family Planning, Association for Development Co-operation, 1998. [17] Qi YL, Tang WH. Analysis of reproductive health education and demands among unmarried youths. Chi Pop Sci 1999;13:59 62. [18] Pardun CJ, LEngle KL, Brown JD. Linking exposure to outcomes: Early adolescents consumption of sexual content in six media. Mass Commun Soc 2005;8:79 91. [19] Rich M. Sex screen: The dilemma of media exposure and sexual behavior. Pediatrics 2005;116:329 31. [20] Smith M. The content and accessibility of sex education information on the internet. Thousand Oaks, CA: Health Educ Behav, 2000. [21] American Academy of Pediatrics. Policy statementsexuality, contraception, and the media. Pediatrics 2010;126:576 82. [22] Zabin LS, Emerson MR, Li N, et al. Levels of change in adolescent sexual behavior in three Asian cities. Stud Fam Plann 2009;40:112. [23] Reiss IL. The social context of premarital sexual permissiveness. New York, NY: Holt, Rinehart, and Winston, 1967. [24] Clayton RR, Bokemeier JL. Premarital sex in the seventies. J Marriage Fam 1980;42:759 75. [25] Chang J. What do education and work mean? Education, non-familial work/ living experiences and premarital sex for women in Taiwan. J Comp Fam Stud 1996;27:1326.

[26] Chia SC. How peers mediate media inuence on adolescents sexual attitudes and sexual behavior. J Commun 2006;56:585 606. [27] The Ahfad University for Women, Omdurman, Sudan. Methods for social researchers in developing countries. Available at: http://srmdc.net/ chapter7/8.htm. [28] Strasburger VC. Sex, drugs, rock n roll: Are the media responsible for adolescent behavior? Adolesc Med 1997;8:40314. [29] Zhang L, Li X, Shah IH, et al. Parent-adolescent sex communication in China. Eur J Contracept Reprod Health Care 2007;12:138 47. [30] Wang RJ, Huang Y, Lin YC. A study of masturbatory knowledge and attitudes and related factors among Taiwan adolescents. J Nurs Res 2007;15:233 42. [31] Hggstrm-Nordin E, Hanson U, Tydn T. Associations between pornography consumption and sexual practices among adolescents in Sweden. Int J STD AIDS 2005;16:1027. [32] Shen Y. KAP study on sexual and reproductive health among 1524 aged unmarried young migrants in Changning district, Shanghai [D]. Shanghai, China: Fudan University, 2003. [33] Sandoval GA. Sexual experiences of the Filipino youth: Demographic patterns and attitudinal correlates. In: Gao ES, Shah IH, Lou CH, Tu XW, eds. Reproductive Health of Adolescents and Unmarried Youth: Status, Perspectives and Strategies. Shanghai, China: The Second Military Medical University Press, 2002. [34] Huong PQ. Adolescent sexual and reproductive health: Vietnamese situation. In: Gao ES, Shah IH, Lou CH, Tu XW, eds. Reproductive Health of Adolescents and Unmarried Youth: Status, Perspectives and Strategies. Shanghai, China: The Second Military Medical University Press, 2002.