A Systematic Review on Adolescents’ Premarital Sexual Behaviour In Malaysia

Background: Premarital sexual behaviours among adolescents and the associated negative consequences of teenage pregnancy is an increasing public health concerns in Malaysia. Despite the extensive research related to sexual and reproductive health of adolescents, the problems remain reecting the possible poor understanding of the underlying causes contribute to the problem. This systematic review is aimed at critically evaluating the available evidence on the contributing factors related to premarital sexual behaviours among adolescents in Malaysia. Methods: Five databases; Scopus, MEDLINE/ PubMed, CINAHL, Science Direct and Google Scholar were utilized to systematically searched related articles up to July 2018. Searching involved local quantitative, qualitative and mixed methods studies with premarital sexual behaviours among adolescents in Malaysia being the primary outcome. Two reviewers independently extracted data and assessed study risks of bias. Results: We identied 1762 articles, of which only nine articles were included, which was dominated by six cross sectional studies. A total of 41 939 adolescents aged between 12 to 19 years old were included in this review. The prevalence of premarital sexual behaviours among adolescents was ranging between 2.9% to 8.3%, reported by history of ever had sex. Religiosity, attitude towards sexual urge (pornography and masturbation), high-risk behaviours (bullying, smoking and substance intake – alcohol and drugs) and parents-peers inuence were main contributing factors identied to associate with premarital sexual behaviours. Conclusion: Negative attitude towards sexual urge and involvement in high risk behaviours put adolescents at higher risk of premarital sexual behaviour, while religious beliefs and reciprocal relationship between parents and peers inuence may act as protective factors. However, more qualitative, and mixed methods studies needed to get in-depth understanding on the relationship between these factors and premarital sexual behaviours.

reported to have experienced a sexual event, compared to only 1% in 1995 [7]. Signi cant numbers of quantitative studies have been published locally related to risky sexual behaviours among adolescents and its negative sequelae with very limited qualitative or mixed methods studies exploring the same problem. Despite the intensi ed efforts to educate the community as a whole and adolescents particularly, the public health concern related to Malaysian adolescents' sexual behaviours remained. By reviewing quantitative and qualitative studies, this article aims to identify the contributing factors of premarital sexual behaviours among the Malaysian adolescents.

Methods
This review was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [37].

Eligibility criteria
We included observational, qualitative, and mixed methods studies that identi ed, predict, explored and explained factors associated with premarital sexual behaviours among adolescents in Malaysia. We excluded experimental studies, non-English language studies, editorial, proceedings, descriptive studies, protocol, and studies among non-Malaysians living in Malaysia and articles published before the year of 2008.

Data sources and search strategy
We systematically searched for relevant articles published in Scopus, MEDLINE/ PubMed, CINAHL, Science Direct and Google Scholar up to July 2018. The search was limited to 10 years, from 2008 until 2018. Combination of keywords were used: Malaysia AND adolescent OR youth OR teen* OR teenager* AND "premarital sex" OR "premarital sexual activity" OR "premarital sexual intercourse" OR "sex before marriage" AND behaviour OR behaviour in the abstract, title or keywords elds.

Study selection
A pair of authors independently assessed the titles and abstracts of a de ned set of articles based on the keywords used. Each study was recorded as include, exclude or unclear. The full articles were retrieved to further assess the include and unclear categorized articles. Eligible studies were identi ed based on the pre-determined inclusion criteria. Discussion between authors were conducted to identify discrepancies in the assessment in order to achieve consensus on the nal articles should be included in the review.

Data extraction, data analysis and quality assessment
The data from the selected studies were extracted according to prevalence of adolescents' premarital sexual activities, and the contributing factors associated with premarital sexual behaviours. Data extraction from all potential studies was documented in Table 1 (See Table 1). The table included information related to the study design, study characteristics (year of study, study objective and sample size), characteristics of study population (age and setting) and the key ndings of individual articles. The Crowe Critical Appraisal Tool (CCAT) was utilized in the quality assessment as the tool covered methodological assessment of both quantitative and qualitative studies [8]. The CCAT examined the studies based on eight criteria which included preliminaries, introduction, design, sampling, data collection, ethical matters, result, discussion, and conclusion. Total score was then converted into percentage whereby the following categories were assigned to allow for comparison; poor quality (≤ 50%), acceptable quality (51-74%), high quality (≥ 75%) [9].

Results
Search results.
We identi ed 1762 articles using the previously stated databases. After excluding 948 duplicated studies, a total of 814 articles were screened based on titles and abstracts, leading to further 765 articles excluded due to study location other than Malaysia, irrelevant topics and non-English written articles. Forty-nine full text articles were screened, of which 40 articles were further excluded for the following reasons: irrelevant outcomes (22), experimental studies (7), study population aged more than 19 years old (10) and articles with uncertain statistical analysis (1). A total of nine studies were included in this review [10][11][12][13][14][15][16][17][18]. Three studies were included in the qualitative analysis compared to seven studies in the quantitative analysis, with quantitative and qualitative ndings of the mixed methods articles were separated accordingly for the purpose of this review. Figure 1 shows the PRISMA owchart.
Quality assessment.
Quality assessment for each study was conducted using Crowe Critical Analysis Tool; an established and validated tool used in assessing quality of observational studies [8]. A total of seven studies were rated as of high-quality with percentage scores ranged between 75 to 85% and the remaining two were rated as of acceptable quality (total score ranged between 51-74%) [9] as tabulated in Table 2. Prevalence of premarital sexual behaviours among the Malaysia adolescents as reported by history of ever had sex, ranged from 2.9-8.3% among secondary school adolescents [10,12,14,15,16] and as high as 62.3% among those that were institutionalised [18]. Two studies were conducted in Selangor which recorded prevalence of 4.6% in 2014 [11] and 2.9% in 2017 [15]. Other studies did not speci cally mention their study location but were being generalized to either peninsular or east Malaysia. On average, the mean age of was 14 years old [10,12,18].

Contributing factors of premarital sexual behaviours among Malaysian adolescents.
All quantitative studies included in this review used questionnaires as their study aimed to measure the association between contributing factors and adolescents' premarital sexual behaviours, presented as odds ratio (OR) or adjusted odds ratio (AOR). We have speci cally categorized the factors identi ed in this review according to the constructs of the Social Cognitive Theory, personal, behaviour and environmental factors.

a. Personal factors
All the studies included reported mixed ndings related to the role of personal factors predicting premarital sexual behaviours among adolescents. Religiosity was mentioned in two of the cross-sectional studies and both qualitative studies [11,13,16,17], with low religiosity was reported to associated with higher risk of premarital sexual behaviours among adolescents. Abdul Manaf et al. [11] reported the highest odds of 11.7 likelihood towards sexual behavior among adolescent due to religiosity (AOR 11.7, 95% CI: 1.3-107.1).

b. Behaviour factors
The review found that most of the articles emphasized the role of behavioural factors predicting premarital sexual behaviours, particularly attitude and involvement in high-risk behaviours. Attitude towards sexual urge, which was also reported by the act of watching and reading pornography and masturbation signi cantly predict sexual behavior in three of the cross-sectional studies and also in the mixed methods study [10,11,13,18]. Meanwhile, high-risk behaviours such as bullying, smoking and substance intake (alcohol and drugs) were also signi cantly predicted sexual behaviours among Malaysian adolescents in ve of the cross-sectional studies [11,12,13,14,15].
History of ever used drug predicted the highest likelihood of 7.71 odds towards sexual behaviours among adolescents [12], compared to 6.1 odds by bullying [11].

c. Environmental factors
Meanwhile, family and peers related factors were identi ed under environmental factors that were associated with sexual behaviours among adolescents. The in uence of family, particularly parental role on involvement of adolescents in sexual behaviours were reported in three of the cross-sectional studies [12,14,15], one qualitative study [16] and the qualitative ndings of the mixed methods study [18]. Having both parents who are still married served as the strongest protective factor (OR 0.41, 95% CI; 0.215-0.789) among Malaysian adolescents, with 59% less likelihood to be involved in sexual behaviours [14]. Meanwhile, the role of peers in in uencing sexual behaviours among adolescents were more prominent in the qualitative ndings in both qualitative studies [16,17] and the mixed methods study [18]. Only two of the cross-sectional studies reported the role of peer in in uencing sexual behaviours among Malaysian adolescents [10,12].

Discussion
Prevalence of premarital sexual behaviours among adolescent in Malaysia.
The review found the prevalence of pre-marital sexual behaviour among Malaysian adolescents ranged between 2.9 and 8.3%. The review nding was higher in comparison to nding by national health survey in 1996 which reported prevalence of ever had sex among adolescents at 1.8% but the prevalence range were in keeping with subsequent series of national surveys on adolescents' at 8.3% in 2012 [19] and 7.3% in 2017 [7]. The poor response rate and the lack of representativeness in the selection of study population in the study by Nawi et al. [15] may have contributed to the low prevalence of premarital sexual behaviour among adolescents. Additionally, the use of self-reported and non-anonymous questionnaires among seven studies may have led to under-reporting of the prevalence of pre-marital sex among Malaysian adolescents.
Nevertheless, the reported prevalence of pre-marital sex behaviour among Malaysian adolescents in these studies were lower as compared to the neighbouring countries, including 11.2% in Brunei [20] and 11.0% in Thailand [21], but higher as compared to Cambodia (8%) and Vietnam (3%) [22]. Possible causes for the variation could be due to characteristics of the respondents, tools used to collect the data and locality of study.
Contributing factors associated with adolescents' premarital sexual behaviours.
The role of religion was the only personal factor consistently reported in few of the articles reviewed, particularly in the qualitative results. Having a more conservative religious connection reported to delay sexual debut [23]. A review done in the United States which analysed evidence of adolescents' religiosity causal relationship with their sexual behaviours noted all 10 longitudinal studies conducted from 1980 to 2000 reported that religious adolescents and those who were actively attending religious activities were unlikely to be involved in sexual activity [23], re ecting the important role of religious beliefs in shaping adolescents' sexual behaviour.
Meanwhile, attitude towards sexual urge and involvement in high-risk behaviours were important factors contributed towards sexual behaviours among Malaysian adolescents, identi ed from this review. Masturbation and reading or watching pornography were two manifestations of negative attitude towards sexual urge reported to predict sexual behaviours among adolescents in this review. Masturbation or solitary sex is de ned as an act of exciting one's sexual organs by rubbing, stroking, pressing, folding or other forms of manipulation to reach orgasm [24]. According to Herbenick et al. [25], the physiological changes and emergence of sexual urges among adolescents may be manifested in the form of masturbation or other activities, ranging from hugging and kissing to vaginal intercourse. Association between masturbation and pornography watching has been reported in previous studies as well. In a case study of undergraduate students in Bangladesh found that, the prevalence of masturbation among the students was 33% and was signi cantly higher among male students [26]. The study also Similarly, risky behaviours such as cigarette smoking, intake of drugs and alcohol were also found as main contributors towards adolescents' premarital sexual behaviour. These factors were also consistently reported as signi cant predictors of premarital sexual behaviour in several other Asian countries such as India, China, Indonesia, Thailand and Vietnam [22]. A recent local study also reported that history of ever used drugs (AOR = 10.201) and ever smoked (AOR = 1.628) were signi cantly associated with sexual activity of adolescents [27]. According to Hops et al. [28], adolescents who abuse substances are more likely to engage in health-risking sexual behaviour (HRSB) and are at particularly high risk for HIV/AIDS. Evidence showed that as many as 22.1% of adolescents engaged in substance use during their most recent sexual encounter [29]. Adolescents who regularly abuse substances are more likely to become sexually active at an earlier age [30], have more sexual partners [31], and are more likely to have unprotected sex [32]. A meta-analysis on risky sexual behaviour and substance use among adolescents reported the role of moderators such as sex, ethnicity and age on the relationship between drug use and risky sexual behaviour [33].
On the other hand, parents and peers were found to be important protective factors towards sexual behaviours among Malaysian adolescents. Wang et al., [3] suggested that there are reciprocal relationships between parenting, peer in uence and adolescent sexual risk involvement. Parents and peers are two major potential in uences on risky and protective behaviours among adolescents by creating a social context for behavioural development, act as role models, and provide opportunity and reinforcement for risk and protective behaviours [3]. Interestingly, literatures have also reported the interplay between parental and peer in uences in order to better understand risky behaviours among adolescents. Parental monitoring may buffer negative peer in uence on adolescent risk involvement [34], in uence the type of peers with whom adolescents associate [35] and decrease the likelihood that adolescents will a liate with deviant peer groups [36]. To effectively reduce sexual risks among adolescents, peers and parents should be included in any related intervention [3].