Lifestyles and Sexual Behaviour Among School-Going Adolescents: A Mediational Role of Mental Health

Abstract

Sexual behaviour and mental health among adolescents are major public health issues. This study examines how lifestyles affect sexual behaviour among school-going adolescents in Malaysia, and the potential mediational role of mental health. It is the first to our knowledge to explore the mediating effect of mental health on sexual behaviour with a focus on a fast-growing developing country. Data were obtained from the National Health and Morbidity Survey 2017 (n = 27,497). Structural equation modelling was utilized to examine depression and anxiety as mediators of the relationships between sexual behaviour and smoking, alcohol drinking and illicit drug use, controlling for sociodemographic factors. Results showed that adolescents who smoked, consumed alcohol, used illicit drugs, and had depression and anxiety were more likely to engage in sexual behaviour than others. Depression and anxiety partially mediated the relationships between sexual behaviour and smoking, and illicit drug use. The association between alcohol drinking and sexual behaviour was fully mediated by depression and anxiety. In conclusion, lifestyles may affect sexual behaviour through mediation of mental health. Therefore, policymakers should take mental health factors into consideration when designing adolescent sexual behaviour preventative interventions.

Keywords

adolescent anxiety depression mediation lifestyle sex

Introduction

In today’s rapidly urbanizing and industrializing society, sexual behaviour during adolescence has become an alarming issue across the globe. During puberty, there is a tendency for adolescents to engage in sexual behaviour (Savioja et al., 2017, 2018). It is clearly evident that while adolescent sexual behaviour does not always lead to poor outcomes, it is closely linked to various risk behaviours, such as multiple sex partners, non-use of condom and drug abuse, which are the main factors causing unplanned pregnancies, teen childbearing and sexually transmitted diseases (Miller, 2002; Kotchick et al., 2001; Madkour et al., 2010). These negative consequences can adversely affect adolescent academic performance, mental health, development, as well as marriage in adulthood (Cheah et al., 2016, 2018).

While sexual behaviour is different from risky sexual behaviour, early sexual activity in adolescents is considered risky because it is associated with unprotected sexual intercourse (Jessor, 1992). According to Jessor’s (1992) concept of adolescent risk behaviour, risky sexual behaviour is purposive, functional and has impacts on adolescent psychological outcomes. Jessor (1992) claims that although risky sexual behaviour is harmful to health, it plays an important role in serving personal development and social function in adolescents. For instance, adolescents engage in early sexual activity with the aim of earning peer acceptance and showing maturity. However, risky sexual behaviour can negatively affect the appropriate psychological development in adolescents, such as underachievement of desired social roles and poor development of necessary skills.

Even though adolescent sexual behaviour is illegal in Malaysia, many adolescents have indulged in sex during schooling years. A nationwide study conducted in 2012 illustrated that nearly one-tenth of secondary school students in Malaysia had ever engaged in sexual behaviour (Institute for Public Health, 2012). Almost half of these students had their first sex before 14 years old, 1.4% had multiple sex partners and 56.3% had not used any contraception. Furthermore, a recent study shows that each year, about 14 out of every 1000 Malaysian girls aged between 10 and 19 years become pregnant, accounting for a total of 18,000 pregnancy cases (Nagandla & Kumar, 2020).

There is growing evidence that the decisions of school-going adolescents to engage in sexual behaviour, i.e., participation in genital activities, are influenced by lifestyle factors (Lee et al., 2006; Wong et al., 2009; Peltzer & Pengpid, 2011; Cheah et al., 2016, 2018; Nawi et al., 2017). In particular, adolescents who smoke, consume alcohol and use illicit drugs have been identified to be more likely to indulge in sexual behaviour compared with non-smokers, non-alcohol drinkers and non-illicit drugs users. Sociodemographic factors also play a role (Lee et al., 2006; Wong et al., 2009; Peltzer & Pengpid, 2011; Cheah et al., 2016, 2018; Nawi et al., 2017). For instance, if the adolescents are male, older, having poor academic performance or living with single parents, their likelihood of engaging in sexual behaviour increases. In addition, Jessor (1992) claims that risky sexual behaviour is highly correlated with other risk behaviours because they serve the similar personal and social function in adolescents. For example, adolescents who smoke tobacco and use illicit drug are more likely to indulge in unprotected sexual intercourse compared to non-smokers and non-illicit drug users as they want to prepare themselves for young adulthood.

An interesting but not well-verified hypothesis is that adolescent lifestyles affect sexual behaviour via mediation of mental health. In general, mental health refers to depression, anxiety and other mood, and cognitive disorders. Oginni et al. (2020) examined the mediational role of mental health and found high-risk sexual behaviour to be partially mediated by mental health disparities. Although their study is comprehensive, it did not investigate mental health as a mediator of the association between sexual behaviour and lifestyle among adolescents. The moderate prevalence of mental health disorders among adolescents in Malaysia indicates the possibility that adolescent sexual behaviour is partly or fully explained by mental health. As noted by the Ministry of Health Malaysia, the prevalence of depression and anxiety among school-going adolescents in 2017 was 18.3% and 39.7%, respectively, which was considered high and should not be overlooked (Institute for Public Health, 2017).

Previous studies consistently found that mental health was an important determining factor of sexual behaviour among adolescents. The scenarios in developed and developing countries showed alike. For instance, using data from a longitudinal study conducted in the United States (US), Ghobadzadeh et al. (2019) and Jamieson and Wade (2011) found that depressive symptoms had both direct and indirect effects on sexual behaviour. In Finland, several studies found adolescents of different age groups and genders to be more likely to indulge in sexual intercourse if they had depression (Savioja et al., 2018). Furthermore, Peltzer et al. (2013) conducted their study in Ivory Coast and identified that poor mental health was independently associated with high-risk sexual behaviour. Somewhat similar findings were evidenced by Othieno et al. (2015), who found that students with depressive symptoms were more likely to have multiple sex partners because having depressive feelings could increase the urge to engage in sexual behaviour. Moreover, Pozuelo et al. (2022) performed a systematic review of studies in low-income countries and found likewise that depression was associated with an increased likelihood of participating in sexual behaviour. Therefore, we expect that mental health could be a plausible mediator of the relationships between adolescent sexual behaviour and lifestyles, especially given that lifestyles are connected to mental health. The relationships between mental health and lifestyles have been evidenced in the Malaysian adolescent population (Kaur et al., 2014).

Although previous studies often examined factors associated with sexual behaviour among adolescents, none had made an attempt to establish the mediating effect of mental health, especially in developing countries. Hence, the objective of the present study is to examine mental health as a mediator of the relationships between lifestyles and sexual behaviour, controlling for the effects of sociodemographic factors in these relationships. This is especially important given that mental disorder among adolescents is an alarming issue in all countries globally, including Malaysia. Malaysia is in transition from a developing country to a developed country, and this is associated with a rapid change in lifestyles (Raaj et al., 2021). As a result, the stress levels increase substantially, leading to serious mental health issue.

The contributions of the present study are numerous. First, to the best of our knowledge, this is one of the few studies that examines how lifestyles affect adolescent sexual behaviour through mediation of mental health, controlling for sociodemographic factors. Second, the country of interest is a fast-growing developing country, Malaysia, where mental disorder is prevalent and study related to adolescent sexual behaviour is lacking. Although not an international sample, our findings are important in the context of dealing with sexual behaviour among the adolescent population in Malaysia and could be generalizable to other developing countries, especially those in Asia that share similar population characteristics. Third, nationally representative data with a large sample size (n = 27,497) are used. Important findings can therefore be generated.

Methods

Data

Data from the National Health and Morbidity Survey (NHMS) 2017 were used for secondary analyses (Institute for Public Health, 2017). The NHMS 2017 was the most comprehensive nationwide adolescent health survey conducted by the Ministry of Health Malaysia. The survey period was from March to May 2017. While the NHMS 2019 is the latest survey, it does not have data on adolescent sexual behaviour and is unavailable for public use at the time the present study was conducted. Hence, it was not used. The objective of the NHMS 2017 was to collect information on school-going adolescents’ health behaviours, sociodemographic characteristics and lifestyle profiles. The inclusion criteria were secondary school students aged between 13 and 17 years across all the states in Malaysia, including federal territories. The single proportion formula for prevalence was used to calculate the sample size. The eligible students were 30,823. However, only 27,497 completed and returned the questionnaires, which was equal to the response rate of 89.2%.

In order to ensure that the data were nationally representative, a two-stage stratified sampling was adopted. In the first stage, 212 secondary schools were randomly chosen from the total of 2738 schools. Probability proportionated to the school size was used. In the second stage, the sampling frame consisted of all classes in each of the selected schools. About 4 to 10 classes were selected from every school using a systematic random sampling. All the students in the selected classes were requested to take part in the survey.

Pretested self-administrated questionnaires were prepared in Malay language, then translated to English. After that, a backward translation was conducted by a professional translator. A pilot study and data collection training were carried out prior to the survey with the aim of minimizing non-sampling errors. A week before the survey, parents of the selected students received the consent forms and were briefed by school teachers. If the students did not obtain consent from their parents or did not want to participate in the survey, they were not allowed to answer the questionnaires. Anonymity of respondents was maintained. The collected data underwent quality check before they were analysed. Ethical approval was obtained from the Medical Research and Ethics Committee of the Ministry of Health Malaysia, as well as the Ministry of Education Malaysia (NMRR-16-698-30,042). Of note, data from the NHMS 2017 are not open access and not all the information can be used for research. Researchers have to seek approval from the Ministry of Health Malaysia prior to using the data.

Outcome Variable

The outcome variable used in the present study, sexual behaviour, was a categorical variable with a binary outcome, indicating whether or not the respondents engaged in sexual behaviour recently. Information on this variable was obtained by asking the respondents ‘Have you ever had sexual intercourse in the past 30 days?’ They responded with either ‘yes’ or ‘no’.

Exposure Variables

Lifestyles were used as the exposures in the present study. They came from three questions: 1) ‘During the past 30 days, did you smoke cigarettes?’ 2) ‘During the past 30 days, did you consume alcohol?’ 3) ‘During the past 30 days, did you use any illicit drugs, such as heroin, glue, amphetamine and ecstasy?’ The respondents were coded as smokers, alcohol drinkers and illicit drugs users if they answered ‘yes’ for these questions, respectively.

Mediators

Mental health variables were used as the mediators of sexual behaviour as well as the outcome variables in unhealthy lifestyles and sociodemographic factors. They consisted of depression and anxiety. The Depression, Anxiety and Stress Scale (DASS-21) developed by Lovibond and Lovibond (1995) was used to identify whether or not the respondents had depression or anxiety (yes vs. no). In the DASS-21, the respondents were requested to answer 21 closes-ended questions related to mental health. The questions were categorised based on the symptoms of anxiety and depression. The possible answers for each question were ‘did not apply to me at all’, ‘applied to me to some of the time’, ‘applied to me to a good part of time’ and ‘applied to me most of the time’. The minimum score of each question was 0, while the maximum score was 3. The total score was then multiplied by two. If the respondents had a total score of >9 (anxiety questions) and >13 (depression questions), they were considered to have anxiety and depression, respectively (Institute for Public Health, 2017). A previous study conducted in Malaysia found that the DASS-21 had high validity and reliability (Nordin et al., 2017). The Cronbach’s alpha values for the subscales of depression, anxiety and stress were 0.863, 0.850 and 0.837, respectively.

Covariates

The control variables were the respondents’ sociodemographic characteristics, which consisted of Form, gender, ethnicity, parents’ marital status and school location. In Malaysia, secondary schools are categorised into lower-secondary (Form 1–3) and upper-secondary (Form 4–5). Form 1 and Form 5 are equivalent to 7th and 11th grade in the US education. The age of Form 1 students was 13 years, while the age of Form 5 students was 17 years. Ethnic variable consisted of six categories: Malay (the ethnic majority), Chinese, Indian, Sabah natives, Sarawak natives and other ethnic groups (Others). The respondents’ parents’ marital status was grouped into two categories: married and single (divorced/widowed). No marriage refers to single parents who consider themselves heterosexual and believe that heterosexuality is the normal sexual orientation. Lesbian, gay, bisexual and transgender (LGBT) parents were not considered in the present study. School location was segmented into urban and rural areas.

Hypotheses

Previous studies often found mental health disorder to be a risk factor for adolescent sexual behaviour. For instance, Ghobadzadeh et al. (2019) found that depression had a positive impact on sexual behaviour. Jamieson and Wade (2011) also reported a positive relationship between sexual behaviour and depression. Using Finnish data, Savioja et al. (2018) observed that adolescents with depression had a higher likelihood of indulging in sexual behaviour compared to their peers without depression. Othieno et al. (2015) also noted that having multiple sex partners was associated with depression. Furthermore, findings of Brem et al. (2017) showed both anxiety and depression to have significant effects on sexual behaviour. The positive relationships between poor mental health and sexual behaviour were also evidenced by Pozuelo et al. (2022), who conducted a systematic review, and Xu et al. (2022), who used a sample of British adolescents. In addition, past studies consistently showed that smoking, alcohol drinking and substance use were highly associated with depression and anxiety (Johnson et al., 2000; Poulin et al., 2005; Steuber & Danner, 2006; Wu et al., 2010; Pedrelli et al., 2016; Johannessen et al., 2017; Xu et al., 2022). More specifically, adolescents who smoked, consumed alcohol, and used substance were more likely to develop depression and anxiety than those who did not engage in any unhealthy lifestyles. Therefore, the present study hypothesizes that depression and anxiety mediate the relationships between sexual behaviour and smoking, alcohol drinking, and use of illicit drugs.

Statistical Analyses

Structural equation modelling (SEM) was used to explore the mediating effects of depression and anxiety on the relationships between sexual behaviour and smoking, alcohol drinking, and use of illicit drugs. Sociodemographic variables were controlled for. Bootstrap percentile confidence intervals were calculated for the direct, indirect and total effects of smoking, alcohol drinking and use of illicit drugs on sexual behaviour. One thousand bootstrap replicates were performed. Furthermore, we conducted a logistic regression analysis with sociodemographic and lifestyle variables affecting the mental health variables, so that we were able to identify the risk factors for depression and anxiety. In addition, to examine the combined effects of the exposure variables on sexual behaviour, smoking, alcohol drinking and use of illicit drugs were combined to form an unhealthy lifestyle variable. This variable consisted of four categories: ‘3’ referred to participation in three unhealthy lifestyles; ‘2’ referred to participation in any two unhealthy lifestyles; ‘1’ referred to participation in any one unhealthy lifestyle; and ‘0’ referred to non-participation. The dose-response relationships between this unhealthy lifestyle variable and sexual behaviour were then estimated using a logistic regression as the outcome variable was a categorical variable (with sociodemographic variables controlled for). The odds of participating in sexual behaviour were presented and interpreted. A total of 27,497 respondents were used in the analyses. No observation was deleted because there was no missing data. The 5% level of significance was selected. Stata statistical software was used to perform all the analyses (StataCorp., 2019).

Results

Of all the respondents, 5.22% engaged in sexual behaviour. Overall, the largest proportion of students were in Form 3. The majority of students were females, Malays, staying with married parents and urban dwellers. There were only a small number of students reporting smoking, drinking and illicit drugs use. In terms of mental health, approximately 38.30% and 17.38% of students had anxiety and depression, respectively (Table 1).

Table 1.

Summary statistics of variables (n = 27,497).

Variables	Frequency	Percentage
Outcome
Sexual behaviour
Yes	1435	5.22
No	26,062	94.78
Mediators
Depression
Yes	4778	17.38
No	22,719	82.62
Anxiety
Yes	10,532	38.30
No	16,965	61.70
Exposures
Smoking
Smoker	4140	15.06
Non-smoker	23,357	84.94
Alcohol
Drinker	2473	8.99
Non-drinker	25,024	91.01
Drugs
User	554	2.01
Non-user	26,943	97.99
Covariates
Form
1	5704	20.74
2	5501	20.01
3	5837	21.23
4	5532	20.12
5	4923	17.90
Gender
Male	13,135	47.77
Female	14,362	52.23
Ethnicity
Malay	18,713	68.05
Chinese	4100	14.91
Indian	1428	5.19
Sabah	1781	6.48
Sarawak	921	3.35
Others	554	2.01
Parent marital
Married	23,544	85.62
Single	3953	14.38
Location
Urban	15,899	57.82
Rural	11,598	42.18

Source: NHMS 2017.

Respondents were less likely to have depression and anxiety if they were male [adjusted odds ratio (aOR): 0.702–0.876] or living with married parents [aOR: 0.746–0.903]. Compared to Form 5 students, Form 1 [aOR: 1.266] and 4 [aOR: 1.161] students were more likely to have depression, but Form 2 [aOR: 0.808] and 3 [aOR: 0.831] students were less likely to have anxiety. Chinese [aOR: 1.197], Indians [aOR: 1.828], Sabah natives [aOR: 1.232] and Others [aOR: 1.358] had a higher likelihood of developing depression relative to Malays. The odds of having depression and anxiety were higher among urban dwellers [aOR: 1.114–1.138], smokers [aOR: 1.219–1.538], alcohol drinkers [aOR: 1.600–1.883] and illicit drugs users [aOR: 3.338–3.425] than others (Table 2).

Table 2.

Correlates of sociodemographic and lifestyle variables to mental health (n = 27,497).

Variables	Depression	Anxiety
Form
1	1.266	0.992
1	(1.142, 1.403)	(0.917, 1.074)
2	1.092	0.808
2	(0.982, 1.214)	(0.745, 0.876)
3	0.973	0.831
3	(0.876, 1.082)	(0.768, 0.900)
4	1.161	1.007
4	(1.046, 1.288)	(0.931, 1.090)
5	Ref.	Ref.
Gender
Male	0.876	0.702
Male	(0.819, 0.937)	(0.667, 0.739)
Female	Ref.	Ref.
Ethnicity
Malay	Ref.	Ref.
Chinese	1.197	0.741
Chinese	(1.088, 1.317)	(0.685, 0.801)
Indian	1.828	1.108
Indian	(1.608, 2.077)	(0.990, 1.239)
Sabah	1.232	1.200
Sabah	(1.085, 1.398)	(1.086, 1.326)
Sarawak	0.870	0.849
Sarawak	(0.723, 1.048)	(0.737, 0.978)
Others	1.358	1.058
Others	(1.100, 1.677)	(0.888, 1.259)
Parent marital status
Married	0.746	0.903
Married	(0.685, 0.813)	(0.842, 0.969)
Single	Ref.	Ref.
Location
Urban	1.138	1.114
Urban	(1.063, 1.218)	(1.058, 1.173)
Rural	Ref.	Ref.
Smoking
Smoker	1.538	1.219
Smoker	(1.404, 1.685)	(1.131, 1.315)
Non-smoker	Ref.	Ref.
Alcohol
Drinker	1.883	1.600
Drinker	(1.688, 2.100)	(1.452, 1.764)
Non-drinker	Ref.	Ref.
Drugs
User	3.425	3.338
User	(2.807, 4.179)	(2.707, 4.116)
Non-user	Ref.	Ref.
Likelihood ratio	1032.650	733.460
p-value	<0.001	<0.001

Note: Adjusted odds ratios are reported. 95% confidence intervals are shown in parentheses. Source: NHMS 2017.

The estimate for engaging in sexual behaviour comparing between smokers and non-smokers if the respondents did not have depression was 0.0522, while the estimate for engaging in sexual behaviour comparing those with and without depression if the respondents were smokers was 0.0026. Alcohol drinkers with depression had a 0.0047 higher chance of engaging in sexual behaviour than those without depression. However, alcohol drinking did not have a direct effect on sexual behaviour. Illicit drug users without depression had a 0.1598 higher probability of indulging in sexual behaviour than non-illicit drug users. The indirect effect of illicit drug use on sexual behaviour was 0.0104 (Figure 1).

Figure 1.

Relations between sexual behaviour and depression, smoking, alcohol drinking, and illicit drug use. Note: *p < 0.05. Sociodemographic variables (Form, gender, ethnicity, parents’ marital status, school location) are controlled for. Source: NHMS 2017.

Among respondents without anxiety, smokers had a 0.0532 higher probability of engaging in sexual behaviour than non-smokers. For smokers, the chance of engaging in sexual behaviour among those with anxiety was 0.0008 higher than those without anxiety. Having anxiety increased the probability of engaging in sexual behaviour by 0.0016 among alcohol drinkers. However, the direct effect of alcohol consumption on sexual behaviour was insignificant. The estimate for engaging in sexual behaviour comparing between illicit drug users and non-users if the respondents did not have anxiety was 0.1638. The indirect effect of illicit drug use on sexual behaviour was 0.0041 (Table 3) (Figure 2).

Table 3.

The direct, indirect, and total effects of smoking, alcohol drinking and illicit drug use on sexual behaviour (n = 27,497).

Variables	Direct	Indirect	Total
Depression
Smoking	0.0522	0.0026	0.0548
Smoking	(0.0424, 0.0619)	(0.0017, 0.0036)	(0.0447, 0.0646)
Alcohol	0.0097	0.0047	0.0522
Alcohol	(-0.0015, 0.0211)	(0.0030, 0.0063)	(0.0028, 0.0258)
Drugs	0.1598	0.0104	0.1702
Drugs	(0.1219, 0.1978)	(0.0068, 0.0141)	(0.1324, 0.2081)
Anxiety
Smoking	0.0532	0.0008	0.0540
Smoking	(0.0433, 0.0641)	(0.0004, 0.0013)	(0.0441, 0.0648)
Alcohol	0.0111	0.0016	0.0127
Alcohol	(-0.0008, 0.0234)	(0.0008, 0.0024)	(0.0010, 0.0250)
Drugs	0.1638	0.0041	0.1678
Drugs	(0.1264, 0.2058)	(0.0021, 0.0061)	(0.1306, 0.2092)

Note: Bootstrap percentile confidence intervals are shown in parentheses. Depression and anxiety are mediating variables. Sociodemographic variables (Form, gender, ethnicity, parents’ marital status, school location) are controlled for.

Source: NHMS 2017.

Figure 2.

Relations between sexual behaviour and anxiety, smoking, alcohol drinking, and illicit drug use. Note: *p < 0.05. Sociodemographic variables (Form, gender, ethnicity, parents’ marital status, school location) are controlled for. Source: NHMS 2017.

In terms of dose-response relationships, respondents who participated in one, two and three unhealthy lifestyles (smoking, alcohol drinking and use of illicit drugs) had 82%, 218.9% and 653.8% higher odds of engaging in sexual behaviour, respectively, compared to respondents who did not indulge in any unhealthy lifestyle. This finding indicated that the likelihood of engaging in sexual behaviour increased with the number of unhealthy lifestyles. In other words, the more unhealthy lifestyles one had, the bigger was the impact on sexual behaviour (Table 4).

Table 4.

Dose-response relationships between sexual behaviour and unhealthy lifestyles (n = 27,497).

Variables	Adjusted OR	Standard error	95% CI
Unhealthy lifestyles
0	Ref.	Ref.	Ref.
1	1.820	0.127	1.588, 2.086
2	3.189	0.387	2.513, 4.046
3	7.538	0.945	5.896, 9.638

Note: OR refers adjusted odds ratio. CI refers to confidence interval. Sociodemographic variables (Form, gender, ethnicity, parents’ marital status, school location) are controlled for. Unhealthy lifestyles refer to smoking, alcohol drinking, and illicit drug use. ‘3’ refers to participation in three unhealthy lifestyles. ‘2’ refers to participation in any two unhealthy lifestyles. ‘1’ refers to participation in any one unhealthy lifestyle. ‘0’ refers to non-participation. Source: NHMS 2017.

Discussion

Using a large dataset and a rigorous statistical approach, the present study offered several important findings. First, depression and anxiety partially mediated the relationships between sexual behaviour and smoking, and illicit drug use. Second, the relationship between alcohol drinking and sexual behaviour was fully mediated by depression and anxiety. It can, therefore, be concluded that adolescents’ lifestyles may affect their decisions to indulge in sexual behaviour through mediation of mental health, regardless of their sociodemographic background. This has to be recognized in a policy development to reduce adolescent sexual behaviour.

Our findings were consistent with the evidence of previous studies that having mental disorders increased participation in sexual behaviour. Based on an African American sample, Rubin et al. (2009) found that having serious depressive symptoms was associated with an increased probability of engaging in sexual behaviour. Findings derived from the South African data showed likewise that adolescents with depressive symptoms were more likely to engage in sexual behaviour than their counterparts without any depressive symptoms (Nduna et al., 2010). In Canada, Wilson et al. (2010) also found a positive relationship between sexual behaviour and risk of depression. Similarly, Agardh et al. (2012) using Uganda data identified that poor mental health was closely linked to sexual behaviour. A study conducted in Finnish high schools also suggested risky sexual behaviour, such as multiple sex partners and avoidance of contraception use to be correlated with depression among students (Kosunen et al., 2003). Three reasons may explain why mental disorders could cause sexual behaviour (Agardh et al., 2012; Savioja et al., 2018; Ghobadzadeh et al., 2019). First, adolescents with mental disorders may engage in sexual behaviour with the aim of reducing their stress. Second, sexual behaviour may be used by depressed adolescents to express anger. Third, having mental disorders may discourage adolescents from taking care of themselves.

Alcohol drinking was identified to affect sexual behaviour indirectly through mental health. In other words, increased sexual behaviour among adolescents who consumed alcohol was fully explained by anxiety and depression. The mechanism was that students who consumed alcohol had a higher tendency to develop anxiety and depression compared with those who were non-alcohol drinkers and consequently were more likely to indulge in sexual behaviour. Therefore, if students could avoid alcohol drinking or developing a mental disorder, their tendency to devote themselves to sexual activity could be reduced. This finding is important in the sense that it can support underage sex-related policy decision-making.

Students who smoked and used illicit drugs were more likely to indulge in sexual behaviour than their counterparts who were non-smokers and non-illicit drug users. These differentials were partly explained by depression and anxiety, indicating that mental health plays a role in determining sexual behaviour among smokers and illicit drug users. Previous studies shared somewhat similar findings that use of tobacco and illicit drugs was positively associated with sexual behaviour (Wilson et al., 2010; Othieno et al., 2015). In addition, the present study found a positive relationship between the number of unhealthy lifestyles (smoking, alcohol drinking and illicit drug use) and the probability of engaging in sexual behaviour. A plausible but unverified explanation for this outcome is that students who adopt an unhealthy lifestyle are less concerned about their health. Hence, the tendency for them to participate in sexual behaviour is generally higher compared with those who live a healthy lifestyle. As such, an intervention measure directed towards reducing sexual behaviour among students could be designed more carefully, taking not only lifestyles into account, but also the mental health condition. Policymakers should be aware of the connections between unhealthy lifestyles, especially smoking, alcohol drinking and illicit drug use, and sexual behaviour.

Findings of the present study supported the theory of adolescent risk behaviour developed by Jessor (1992) that early sexual activity, smoking, alcohol consumption and illicit drug use were highly interrelated. There were two explanations for these findings (Jessor, 1992). The first explanation was related to adolescent social ecology. Particularly, the ecology provided adolescents with opportunities to learn various risk behaviours at the same time. The second explanation was that similar function could be served by different risk behaviours. For instance, adolescents engaged in early sexual activity and smoked cigarettes in order to strive independence from their parents. These findings about interrelated risk behaviours are important to policymakers. To lower the prevalence of early sexual behaviour, policymakers should devote their attention to multiple adolescent risk behaviours as a whole instead of individual risk behaviours. As suggested by Jessor (1992), simultaneous and comprehensive policies focusing on all types of adolescent risk behaviours together are necessary.

While the relationship between sociodemographic characteristics and sexual behaviour was not the main focus of the present study, it was noteworthy. The present study found that students living with married parents were less likely to suffer from anxiety and depression than students living with single parents, and they also had a lower likelihood of indulging in sexual behaviour. Somewhat similar findings were evidenced in previous studies (Kaur et al., 2014; Nalugya-Sserunjogi et al., 2016). The authors found that having married parents was a protective factor of depression because adolescents were closely supervised by their parents and this may indirectly lower the tendency to engage in sex. Based on our findings, we conclude that the role of parents in mental health and sexual behaviour prevention is important and should be given attention by policymakers.

Sociodemographic variables that affected mental health but not sexual behaviour were gender, ethnicity and school location. These variables were well-worth discussing and should be taken into account in the formulation of adolescent mental health preventative interventions. Firstly, males were less likely to develop depression and anxiety than females, which was consistent with the findings of previous studies (Adewuya & Ologun, 2006; Dooley et al., 2015; Anjum et al., 2019). Perhaps this is because female adolescents experience more biological changes related to puberty than males. Secondly, Indians and Sabah natives had a higher risk of depression compared with Malays. This finding was in agreement with those of Kaur et al. (2014), concluding that culture and religion may play a vital role in influencing mental health. Thirdly, the likelihood of developing depression and anxiety was higher among urban students compared to rural students. Finding of Anjum et al. (2019) suggested likewise that residing in urban areas increased the odds of suffering from depression due to hectic lifestyles.

One of the strengths of the present study is the exploration of the mediational role of mental health in sexual behaviour among school-going adolescents. As expected, mental health is an important mediator of the relationships between adolescent sexual behaviour and lifestyles. Furthermore, the present study has the novelty of the population given the paucity of study conducted in Malaysia and other developing countries. Moreover, findings estimated from a large dataset and a robust methodological approach have important contributions to literature and policy development. In particular, our findings suggest that policymakers could make an effort to safeguard school-going adolescents who smoke, consume alcohol and use illicit drugs from developing anxiety and depression, so that their tendency to indulge in sexual behaviour can be reduced. Special attention could also be paid to adolescents with single parents. In addition, an intervention strategy directed towards reducing mental disorders among high-risk groups, such as females, non-Malays and urban dwellers could be introduced.

One of the limitations of the present study is that in spite of the high correlations in our statistical models, we are unable to make conclusions about causality because of cross-sectional data. Furthermore, in Malaysia, school-going adolescents are barred from engaging in sexual behaviour. Hence, minor reporting error is unavoidable. There is also a possibility of underreporting sexual behaviour for legal and gender reasons. Moreover, we are uncertain that whether or not adolescents engaged in unprotected sexual behaviour. A future qualitative study is suggested to concentrate on improving the knowledge of the mechanisms explaining the association between mental health and sexual behaviour.

Conclusion

The present study sheds light on the relationships between sexual behaviour and lifestyles. Controlling for sociodemographic factors, smokers, alcohol drinkers and illicit drug users have been found to be more likely to engage in sexual behaviour than non-smokers, non-drinkers and non-illicit drug users because they tend to have anxiety and depression. With the findings of the present study, more effective intervention measures directed toward reducing adolescent sexual behaviour can be formulated.

Footnotes

Acknowledgments

The authors would like to thank the Director General of Health, Malaysia for his permission to use the data from the National Health and Morbidity Survey (NHMS) 2017 and to publish this paper.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Yong Kang Cheah

Author biographies

Yong Kang Cheah is an Associate Professor in School of Economics, Finance and Banking at Universiti Utara Malaysia. He completed his Postdoc at University of Chicago in 2019 and Ph.D. at University of Malaya in 2014. He obtained a Master of Economic Management in 2010 and a Bachelor of Social Sciences (Economics) in 2009 at Universiti Sains Malaysia. His research interests are Health Economics, Public Health, Health Behaviour and Mental Health. His works have been published in numerous journals, such as Journal of General Internal Medicine, Public Health, Journal of Diabetes, Journal of Pediatric Nursing, Psychiatry Research, Nursing and Health Sciences, British Food Journal, Nutrition and Cancer, and Socio-Economic Planning Sciences.

Nabilah Hanis Zainuddin completed and graduated with Bachelor of Medicine and Bachelor of Surgery from Melaka-Manipal Medical College. She has been with Ministry of Health for the past 9 years and is currently working as a medical officer in Sector Biostatics and Data Repository, National Institutes of Health, Malaysia. Her research interests are Biostatistics, Epidemiology and Non-communicable Diseases.

Chee Cheong Kee completed and graduated with Bachelor of Science in Nutrition and Community Health from Universiti Putra Malaysia and Master degree with Clinical Nutrition from National University of Malaysia. He has been with Ministry of Health Malaysia for the past 25 years and is currently working as a researcher in Sector for Biostatistics and Data Repository, National Institutes of Health, Malaysia. He provides staff in Ministry of Health Malaysia with biostatistics and research methodology training. He is also a statistical consultant to National Institutes of Health. His fields of research interest include Nutritional Assessment and Epidemiology, Smoking, and Non-communicable Diseases.

Kuang Hock Lim is currently a researcher at Institute for Medical Research, Ministry of Health Malaysia. He studies tobacco use among youth and adults in Malaysia with a focus on prevalence, second-hand smoke exposure, and anti-tobacco policies. He is also actively involved in validation of the questionnaire. He has published more than 100 scientific papers and has a Scopus H-Index of 20.

Mohd Azahadi Omar obtained his Master in Community Health (Epidemiology & Statistics) from National University of Malaysia in 2003. Currently, he is the Head for Sector for Biostatistics and Data Repository, National Institutes of Health, Malaysia. His main research interests are Research Methodology, Biostatistics, and Burden of Disease. He has been involved in numerous population-based studies and served as a Principal Investigator (PI), co-PI and research team member. He has published more than 90 manuscripts in peer-reviewed journals and presented more than 120 research articles in local and international conferences.

Yoon Ling Cheong is a research officer in Institute for Medical Research, National Institutes of Health, Ministry of Health Malaysia. She obtained her Ph.D. in Geography at University Humboldt Berlin, Germany (2015), Master in Remote Sensing and GIS at Universiti Putra Malaysia (2007), and Bachelor of Information Technology at Universiti Tenaga National (2002). Her research interests are Spatiotemporal Studies using Geographic Information System and Remote Sensing on infectious diseases, Non-communicable Diseases, and Environment-related Diseases. She has published spatiotemporal research on dengue, Zika, and tuberculosis.

References

Adewuya

A. O.

Ologun

Y. A.

(2006). Factors associated with depressive symptoms in Nigerian adolescents. Journal of Adolescent Health, 39(1), 105–110. https://doi.org/10.1016/j.jadohealth.2005.08.016

Agardh

Cantor-Graae

Ostergren

P. O.

(2012). Youth, sexual risk-taking behavior, and mental health: A study of university students in Uganda. International Journal of Behavioral Medicine, 19(2), 208–216. https://doi.org/10.1007/s12529-011-9159-4

Anjum

Hossain

Sikder

Uddin

M. E.

Rahim

D. A.

(2019). Investigating the prevalence of and factors associated with depressive symptoms among urban and semi-urban school adolescents in Bangladesh: A pilot study. International Health, 14(4), 354–362. https://doi.org/10.1093/inthealth/ihz092

Brem

M. J.

Shorey

R. C.

Anderson

Stuart

G. L.

(2017). Depression, anxiety, and compulsive sexual behaviour among men in residential treatment for substance use disorders: The role of experiential avoidance. Clinical Psychology & Psychotherapy, 24(6), 1246–1253. https://doi.org/10.1002/cpp.2085

Cheah

Y. K.

Lim

H. K.

Kee

C. C.

(2018). Factors associated with underage sex in Malaysia: A secondary data analysis of Malaysia global school-based student health survey 2012. Journal of Population and Social Studies, 26(4), 305–320. https://so03.tci-thaijo.org/index.php/jpss/article/view/114795/111246

Cheah

Y. K.

Lim

H. K.

Kee

C. C.

Mohd Ghazali

(2016). Demographic and lifestyle factors associated with sexual activity among adolescents in Malaysia. Vulnerable Children and Youth Studies, 11(4), 324–331. https://doi.org/10.1080/17450128.2016.1218095

Dooley

Fitzgerald

Mac Giollabhui

(2015). The risk and protective factors associated with depression and anxiety in a national sample of Irish adolescents. Irish Journal of Psychological Medicine, 32(1), 93–105. https://doi.org/10.1017/ipm.2014.83

Ghobadzadeh

McMorris

B. J.

Sieving

R. E.

Porta

C. M.

Brady

S. S.

(2019). Relationship between adolescent stress, depressive symptoms, and sexual risk behavior in young adulthood: A structural equation modeling analysis. Journal of Pediatric Health Care, 33(4), 394–403. https://doi.org/10.1016/j.pedhc.2018.11.006

Institute for Public Health . (2012). Global school-based student health survey 2012. Ministry of Health Malaysia.

10.

Institute for Public Health . (2017). National health and morbidity survey 2017: Adolescent health survey. Ministry of Health Malaysia.

11.

Jamieson

L. K.

Wade

T. J.

(2011). Early age of first sexual intercourse and depressive symptomatology among adolescents. Journal of Sex Research, 48(5), 450–460. https://doi.org/10.1080/00224499.2010.509892

12.

Jessor

(1992). Risk behavior in adolescence: A psychosocial framework for understanding and action. Developmental Review, 12(4), 374–390. https://doi.org/10.1016/0273-2297(92)90014-S

13.

Johannessen

E. L.

Andersson

H. W.

Bjorngaard

J. H.

Pape

(2017). Anxiety and depression symptoms and alcohol use among adolescents – a cross sectional study of Norwegian secondary school students. BMC Public Health, 17(1), 494. https://doi.org/10.1186/s12889-017-4389-2

14.

Johnson

J. G.

Cohen

Pine

D. S.

Klein

D. F.

Kasen

Brook

J. S.

(2000). Association between cigarette smoking and anxiety disorders during adolescence and early adulthood. JAMA, 284(18), 2348–2351. https://doi.org/10.1001/jama.284.18.2348

15.

Kaur

Cheong

S. M.

Naidu

B. M.

Kaur

Manickam

M. A.

Noor

M. M.

Ibrahim

Rosman

(2014). Prevalence and correlates of depression among adolescents in Malaysia. Asia-Pacific Journal of Public Health, 26(5S), 53S–62S. https://doi.org/10.1177/1010539514544356

16.

Kosunen

Kaltiala-Heino

Rimpela

Laippala

(2003). Risk-taking sexual behaviour and self-reported depression in middle adolescence – a school-based survey. Child: Care, Health and Development, 29(5), 337–344. https://doi.org/10.1046/j.1365-2214.2003.00357.x

17.

Kotchick

B. A.

Shaffer

Miller

K. S.

Forehand

(2001). Adolescent sexual risk behavior: A multi-system perspective. Clinical Psychology Review, 21(4), 493–519. https://doi.org/10.1016/S0272-7358(99)00070-7

18.

Lee

L. K.

Chen

P. C. Y.

Lee

K. K.

Kaur

(2006). Premarital sexual intercourse among adolescents in Malaysia: A cross-sectional Malaysian school survey. Singapore Medical Journal, 47(6), 476–481. http://www.smj.org.sg/sites/default/files/4706/4706a2.pdf

19.

Lovibond

S. H.

Lovibond

P. F.

(1995). Manual for the depression anxiety and stress scales (2nd ed.). Psychology Foundation.

20.

Madkour

A. S.

Farhat

Halpern

C. T.

Godeau

Gabhainn

S. N.

(2010). Early adolescent sexual initiation as a problem behavior: A comparative study of five nations. Journal of Adolescent Health, 47(4), 389–398. https://doi.org/10.1016/j.jadohealth.2010.02.008

21.

Miller

B. C.

(2002). Family influences on adolescent sexual and contraceptive behavior. Journal of Sex Research, 39(1), 22–26. https://doi.org/10.1080/00224490209552115

22.

Nagandla

Kumar

(2020). Prevalence of teenage pregnancy in 2015-2016 and its obstetric outcomes compared to non-teenage pregnancy at hospital Tuanku Ja’Afar Seremban (HTJS), Negeri Sembilan, Malaysia: A retrospective case-control study based on the national obstetric registry. Malaysian Family Physician, 15(2), 2–9. https://e-mfp.org/wp-content/uploads/2020/07/v15n2-original-article-1.pdf

23.

Nalugya-Sserunjogi

Rukundo

G. Z.

Ovuga

Kiwuwa

S. M.

Musisi

Nakimuli-Mpungu

(2016). Prevalence and factors associated with depression symptoms among school-going adolescents in Central Uganda. Child and Adolescent Psychiatry and Mental Health, 10, 39. https://doi.org/10.1186/s13034-016-0133-4

24.

Nawi

A. M.

Roslan

Idris

I. B.

Hod

(2017). Bullying and truancy: Predictors to sexual practices among school-going adolescents in Malaysia – a cross-sectional study. Medical Journal of Malaysia, 72(5), 298–305. http://www.e-mjm.org/2017/v72n5/adolescents-sexual-practices.pdf

25.

Nduna

Jewkes

R. K.

Dunkle

K. L.

Jama Shai

N. P.

Colman

(2010). Associations between depressive symptoms, sexual behaviour and relationship characteristics: A prospective cohort study of young women and men in the Eastern Cape, South Africa. Journal of the International AIDS Society, 13(1), 44. https://doi.org/10.1186/1758-2652-13-44

26.

Nordin

Kaur

Soni

Por

L. K.

Miranda

(2017). Construct validity and internal consistency reliability of the Malay version of the 21-item depression anxiety stress scale (Malay-DASS-21) among male outpatient clinic attendees in Johor. Medical Journal of Malaysia, 27(5), 264–270. http://www.e-mjm.org/2017/v72n5/Malay-DASS-21.pdf

27.

Oginni

O. A.

Jern

Rijsdijk

F. V.

(2020). Mental health disparities mediating increased risky sexual behavior in sexual minorities: A twin approach. Archives of Sexual Behavior, 49(7), 2497–2510. https://doi.org/10.1007/s10508-020-01696-w.

28.

Othieno

C. J.

Okoth

Peltzer

Pengpid

Malla

L. O.

(2015). Risky HIV sexual behaviour and depression among University of Nairobi students. Annals of General Psychiatry, 14(1), 16. https://doi.org/10.1186/s12991-015-0054-2.

29.

Pedrelli

Shapero

Archibald

Dale

(2016). Alcohol use and depression during adolescence and young adulthood: A summary and interpretation of mixed findings. Current Addiction Reports, 3(1), 91–97. https://doi.org/10.1007/s40429-016-0084-0

30.

Peltzer

Pengpid

(2011). Prevalence and social correlates of sexual intercourse among school-going adolescents in Thailand. Scientific World Journal, 11(1), 1812–1820. https://doi.org/10.1100/2011/532109

31.

Peltzer

Pengpid

Tiembre

(2013). Mental health, childhood abuse and HIV sexual risk behaviour among university students in Ivory coast. Annals of General Psychiatry, 12(1), 18. https://doi.org/10.1186/1744-859X-12-18

32.

Poulin

Hand

Boudreau

Santor

(2005). Gender differences in the association between substance use and elevated depressive symptoms in a general adolescent population. Addiction, 100(4), 525–535. https://doi.org/10.1111/j.1360-0443.2005.01033.x

33.

Pozuelo

J. R.

Desborough

Stein

Cipriani

(2022). Systematic review and meta-analysis: Depressive symptoms and risky behaviors among adolescents in low- and middle-income countries. Journal of the American Academy of Child and Adolescent Psychiatry, 61(2), 255–276. https://doi.org/10.1016/j.jaac.2021.05.005

34.

Raaj

Navanathan

Tharmaselan

Lally

(2021). Mental disorders in Malaysia: An increase in lifetime prevalence. BJPsych International, 18(4), 97–99. https://doi.org/10.1192/bji.2021.4

35.

Rubin

A. G.

Gold

M. A.

Primack

B. A.

(2009). Associations between depressive symptoms and sexual risk behaviour in a diverse sample of female adolescents. Journal of Pediatric and Adolescent Gynecology, 22(5), 306–312. https://doi.org/10.1016/j.jpag.2008.12.011

36.

Savioja

Helminen

Frojd

Marttunen

Kaltiala-Heino

(2017). Delinquency and sexual experiences across adolescence: Does depression play a role? European Journal of Contraception and Reproductive Health Care, 22(4), 298–304. https://doi.org/10.1080/13625187.2017.1374361

37.

Savioja

Helminen

Frojd

Marttunen

Kaltiala-Heino

(2018). Adolescent sexual behavior – family characteristics, parental involvement, and associated mental disorders. International Journal of Sexual Health, 30(3), 295–308. https://doi.org/10.1080/19317611.2018.1494077

38.

StataCorp (2019). Stata statistical software: Release 16.1. Stata Corporation.

39.

Steuber

T. L.

Danner

(2006). Adolescent smoking and depression: Which comes first? Addictive Behaviors, 31(1), 133–136. https://doi.org/10.1016/j.addbeh.2005.04.010

40.

Wilson

Asbridge

Kisely

Langille

(2010). Associations of risk of depression with sexual risk taking among adolescents in Nova Scotia High Schools. Canadian Journal of Psychiatry, 55(9), 577–585. https://doi.org/10.1177/070674371005500906

41.

Wong

M. L.

Chan

R. K. W.

Koh

Tan

H. H.

Lim

F. S.

Emmanuel

Bishop

(2009). Premarital sexual intercourse among adolescents in an Asian country: Multilevel ecological factors. Pediatrics, 124(1), Article e44-e52. https://doi.org/10.1542/peds.2008-2954

42.

Goodwin

R. D.

Fuller

Liu

Comer

J. S.

Cohen

Hoven

C. W.

(2010). The relationship between anxiety disorders and substance use among adolescents in the community: Specificity and gender differences. Journal of Youth and Adolescence, 39(2), 177–188. https://doi.org/10.1007/s10964-008-9385-5

43.

Norton

Rahman

(2022). Adolescent sexual behavior patterns, mental health, and early life adversities in a British birth cohort. Journal of Sex Research, 59(1), 1–12. https://doi.org/10.1080/00224499.2021.1959509