Impact statements
Data on changes in adolescent suicidal behaviours are limited. This study is a secondary analysis of repeated cross-sectional data from the Global School-based Student Health Surveys conducted in Indonesia, Myanmar, the Philippines and Thailand in 2007/2008 and 2015/2016. It reveals that suicidal behaviours were prevalent in all countries and increased across all of them in that period, with the largest rises observed in Myanmar and Thailand. Being physically attacked, bullied and experiencing social difficulties were the most significant and consistent risk factors for suicidal behaviours. These findings can draw adequate attention from policymakers, as well as the health, social and education sectors and parents to adolescents’ suicidal behaviours. The study provides evidence informing multicomponent interventions targeting multiple risk factors within society, schools and families to mitigate this public health crisis.
Introduction
Suicide, the act of intentionally ending one’s own life, is the second most common cause of death among adolescents and young adults aged 10–24 years globally (Hawton et al., Reference Hawton, Saunders and O’Connor2012). Suicide is rare before the age of 15 years (estimated at 1.2 deaths per 100,000 boys aged 5–14 years), but it increases rapidly during the rest of adolescence (19.2 deaths per 100,000 male adolescents aged 15–24 years) (Bertolote and Fleischmann, Reference Bertolote and Fleischmann2002; Kolves and De Leo, Reference Kolves and De Leo2014; Naghavi, Reference Naghavi2019). Therefore, early and mid-adolescence are critical periods for the prevention of suicide. Suicidal behaviours are defined as “a range of behaviours that include thinking about suicide (or ideation), planning for suicide, attempting suicide and suicide itself” (WHO, 2014). Understanding the determinants of suicidal behaviours among adolescents is crucial for improving and developing suicide prevention strategies.
Carballo et al. reviewed 44 empirical studies on the risk factors for suicidal behaviours in children and adolescents (Carballo et al., Reference Carballo, Llorente, Kehrmann, Flamarique, Zuddas, Purper-Ouakil, Hoekstra, Coghill, Schulze and Dittmann2020). They identified three main factors that increase the risk of suicidal behaviours: psychological factors (e.g., depression, anxiety, previous suicide attempts, drug and alcohol use and other mental health disorders); stressful life events (e.g., family problems and conflicts with peers); and personality traits (e.g., neuroticism and impulsivity). Hawton et al.’s (Reference Hawton, Saunders and O’Connor2012) review suggests three groups of potential risk factors for suicidal behaviours among adolescents: psychiatric and psychological factors (e.g., mental disorders, drug and alcohol misuse, low self-esteem, poor social problem-solving and hopelessness); sociodemographic and educational factors (e.g., gender, low socioeconomic status and academic pressure); and individual negative life events and family adversity (e.g., parental separation or divorce, adverse childhood experiences and bullying).
Hawton et al.’s (Reference Hawton, Saunders and O’Connor2012) review highlights gender inequality in suicidal behaviours and found that girls are more likely to self-harm than boys, whereas boys are more likely to die by suicide than girls. Gillies et al.’s meta-analysis of 72 studies confirms that the risk of self-harm among girls is 1.72 times higher than that among boys (Gillies et al., Reference Gillies, Christou, Dixon, Featherston, Rapti, Garcia-Anguita, Villasis-Keever, Reebye, Christou and Al Kabir2018). Miranda-Mendizabal et al. conducted a meta-analysis of 67 studies and found that girls had twice the risk of suicide attempts compared to boys, while boys had a 2.5 times higher risk of dying by suicide than their female counterparts (Miranda-Mendizabal et al., Reference Miranda-Mendizabal, Castellví, Parés-Badell, Alayo, Almenara, Alonso, Blasco, Cebria, Gabilondo and Gili2019).
Empirical studies have identified both individual factors and living circumstances that might be associated with suicidal behaviours among adolescents. This suggests that suicidal behaviours are complex, multifactorial and not fully understood, especially among adolescents in resource-constrained settings. There is no clear evidence on whether the prevalence of suicidal behaviours among adolescents has increased or decreased over the last decade. In addition, it remains unclear what the most significant factors contributing to adolescent suicidal behaviours are and if the population attributable fractions of these factors have changed over time.
This study aimed to examine changes in the prevalence of suicidal behaviours among 13- to 15-year-olds and the factors associated with these over time in four lower- and upper-middle-income countries in the Association of Southeast Asian Nations (Southeast Asia). We hypothesised that the burden of suicidal behaviours among 13- to-15-year-olds increased during the study period, and that experiences of violence and abuse were the most significant risk factors contributing to this burden in all four countries.
Methods
This study is a secondary analysis of data from the Global School-based Student Health Surveys (GSHS) conducted in Indonesia, Myanmar, the Philippines and Thailand – the Southeast Asian countries with data available from two time points.
GSHS surveys
The GSHS is a collaborative surveillance project between the World Health Organization, the United Nations Children’s Fund and the United States Centers for Disease Control and Prevention. The survey has been implemented in 105 countries and territories (as of 2024) to assess behavioural risk factors and protective factors related to the leading causes of morbidity and mortality among adolescents aged 13–17 years.
The GSHS is designed as a relatively low-cost, school-based survey that collects data through a self-administered questionnaire, which can be completed during a regular class period. Each country develops its questionnaire based on standardised core questions and additional country-specific questions. More details on the questionnaire are available on the GSHS website (World Health Organization). The questionnaires were translated into the local languages of instruction for students using a standardised protocol set by the GSHS international team and were pilot-tested for comprehension in each participating country (World Health Organization). A test–retest reliability study of the core questionnaire among Fijian girls found that 69 of 71 items were reliable, with an average test–retest agreement of 77% and an average Cohen’s κ of 0.47 (Becker et al., Reference Becker, Roberts, Perloe, Bainivualiku, Richards, Gilman and Striegel-Moore2010).
Each survey is administered to a nationally representative sample drawn using a standardised two-stage sampling process. In the first stage, schools are selected with a probability proportional to enrolment size. In the second stage, classes are randomly selected from each school, and all students in the selected classes are eligible and invited to participate.
Individual data from the GSHS are publicly available via the GSHS website (World Health Organisation).
Participants
This study included data from adolescents aged 13–15 years from the GSHS conducted in Indonesia (2007 and 2015), Myanmar (2007 and 2016), the Philippines (2007 and 2015) and Thailand (2008 and 2015). Adolescents aged 16 years and above in the 2015/2016 surveys were excluded to be compatible with the 2007/2008 surveys, which did not include that age group.
Data sources
Suicidal behaviours
Suicidal behaviours were assessed using three questions: (1) During the past 12 months, did you ever seriously consider attempting suicide? (2) During the past 12 months, did you make a plan about how you would attempt suicide? and (3) During the past 12 months, how many times did you actually attempt suicide? Question 3 was not asked in the surveys in 2007/2008. The questions on suicidal behaviours were translated into the local language and pilot tested by the research team in each country.
Socio-demographic characteristics
Adolescents’ age and sex were collected using study-specific questions. Household economic status was assessed using the proxy question “During the past 30 days, how often did you go hungry because there was not enough food in your home?” Response options are: never, rarely, sometimes, most of the time and always.
Health risk behaviours
Alcohol drinking was assessed using two questions: (1) During the past 30 days, on how many days did you have at least one drink containing alcohol? and (2) During your life, how many times have you gotten into trouble with your family or friends, missed school or gotten into fights as a result of drinking alcohol? Drug use was assessed using the question: “During your life, how many times have you used drugs?” Social difficulty was assessed using two questions: (1) How many close friends do you have? and (2) “During the past 30 days, how often were most of the students in your school kind and helpful?”
Bullying and physical attack
Being bullied and physically attacked were assessed using two questions: “During the past 12 months, how many times were you physically attacked?” and “During the past 30 days, on how many days were you bullied?”
Parental supervision
Parental supervision was assessed using two questions: “During the past 30 days, how often did your parents or guardians check to see if your homework was done?” and “During the past 30 days, how often did your parents or guardians really know what you were doing with your free time?”
Data analysis
The analyses were conducted in three steps. First, the prevalence rates (with 95% confidence intervals, CIs) of suicidal behaviours and key risk factors were estimated for each survey. The changes in the proportions of suicidal behaviours and key risk factors between the two surveys in each country were calculated and tested using the Z-test.
Second, the PAF of each potential risk factor for suicidal behaviours was calculated. PAF is defined as the fraction of all cases of a particular outcome in a population that is attributable to a specific risk factor (Mansournia and Altman, Reference Mansournia and Altman2018). In other words, PAF estimates the proportion of cases that would not have occurred if there had been no exposure to a particular risk factor. If the 95% CI of a PAF does not cover the value of zero, it is interpreted as statistically significant.
Finally, we conducted multivariate decomposition analyses (Powers et al., Reference Powers, Yoshioka and Yun2011) to examine the contributions of different factors to changes in the proportions of suicidal behaviours between two time points in each country. This technique partitions the overall change into two components for each factor: (1) endowment effect, which reflects how changes in the distribution of a factor contribute to changes in the outcome, and (2) coefficient effect, which captures how changes in the strength or impact of a factor on the outcome changes contribute to the overall change. For example, an increase in suicidal behaviours among adolescents could be due to a higher proportion of adolescents using illicit drugs (endowment effect) and/or because the effect of drug use on suicidal behaviours became stronger (coefficient effect). The results of these multivariate decomposition analyses can also help to explain the change in PAF of a factor between the two time points.
Because data on suicide attempts were not collected in the 2007/2008 surveys, suicidal behaviours were defined as having suicidal thoughts and/or plans in the past 12 months for all surveys. In addition, PAF analyses for suicide attempts were conducted using the 2015/2016 survey data.
Analyses were performed using STATA Version 16 (StataCorp, 2019). PAFs were calculated using the Stata model PUNAF (Newson, Reference Newson2015) after performing multiple logistic regression analyses. The decomposition analyses were conducted using the “mvdcmp” module (Powers et al., Reference Powers, Yoshioka and Yun2011). All analyses accounted for survey design (cluster effects and sampling weights). Missing data were handled using the pairwise deletion method.
Results
Samples
The number of participants included in the analyses ranged from 1,940 in Myanmar in 2016 to 5,624 in Indonesia in 2015 (see Supplementary Table 1). The sex ratios were not balanced, and the proportions of age groups varied across surveys. These differences were adjusted for in the analyses.
Health risks
Most health risks (Table 1) were prevalent across all four countries at both time points. Among these risks, being physically attacked or bullied – ranging from 37.8% in Myanmar (2007) to 70.3% in the Philippines (2007) – and a lack of parental supervision – ranging from 21.9% in Myanmar (2007) to 63.6% in the Philippines (2015) – were the most common across all countries. The Philippines reported the highest proportions for nearly every risk factor. Alcohol consumption (10.9% in 2008 and 14.6% in 2015) and drug use (6.0% in 2008 and 6.4% in 2015) were most prevalent in Thailand, similar to the Philippines (alcohol consumption: 8.0% and 14.1%; drug use: 4.5% and 6.7%, respectively).
Table 1. Weighted proportions (95% CI) of health risks
Data cell: Weighted proportion, % (95% CI).
Comparing the two time points (Supplementary Table 2), alcohol consumption and drug use increased significantly (p < 0.05) in every country, except for drug use in the Philippines and Thailand. Notably, the largest changes occurred in being physically attacked or bullied and lack of parental supervision, with decreases observed in Indonesia (−22.1%, p < 0.001, 95% CI [−27.4, −16.8]; and − 10%, p < 0.001, 95% CI [−14.5, −5.8], respectively) and increases in Myanmar (24.9%, p < 0.001, 95% CI [17.4, 32.4]; and 15.9%, p < 0.001, 95% CI [10.1, 21.7], respectively).
Suicidal behaviours
Suicidal behaviours (Table 2), including seriously considering suicide and making a suicide plan, were most common in Thailand (13.0% in 2008 and 20.9% in 2015) and the Philippines (17.3% and 17.4%, respectively). The proportions of young people experiencing suicidal thoughts increased in all countries over that period (Supplementary Table 3), except the Philippines (p = 0.938). The largest increases were in Myanmar (10%, p < 0.001, 95% CI [7.3, 12.7]) and Thailand (7.9, p < 0.001, 95% CI [3.7, 12.1]).
Table 2. Weighted proportions (95% CI) of suicidal behaviours in the past 12 months
Data cell: Weighted proportion, % (95% CI).
Similarly, the proportion of actual suicide attempts was highest in the Philippines and Thailand, with 4.7% (95% CI [3.4, 6.4]) of participants in Thailand reporting four or more suicide attempts in the past 12 months. The lowest proportion of young people reporting suicide attempts was in Indonesia, with 1.0% (95% CI [0.7, 1.5]) reporting four or more suicide attempts in the past 12 months.
Risk factors for suicidal behaviours
The statistically significant PAFs for the risk factors of suicidal behaviours (suicidal thoughts and/or plans in the past 12 months) are presented in Table 3. Among the risk factors, being physically attacked or bullied consistently contributed the most to suicidal behaviours across countries and time points – ranging from 32.3% (95% CI [24.6, 39.3]) in Indonesia (2015) to 63.7% (95% CI [1.3, 86.6]) in Myanmar (2007). Social difficulties – ranging from 7.6% (95% CI [1.1, 13.6]) in the Philippines (2007) to 57.2% (95% CI [2.6, 81.2]) in Myanmar (2007) – and lack of parental supervision – ranging from 10.1% (95% CI [1.2, 18.3]) in Thailand (2015) to 26.8% (95% CI [15.2, 36.8]) in Myanmar (2016) – were also consistent and significant risk factors. Being female and experiencing poverty were significant risk factors – ranging from 12.8% (95% CI [6.0, 19.2]) in the Philippines (2015) to 33.6% (95% CI [20.6, 44.5]) in Myanmar (2016) but not entirely consistent across countries (not significant in Myanmar [2007] and Thailand [2015]). The magnitude of the association between alcohol consumption and suicidal behaviours was larger in the Philippines (5.8%, 95% CI [3.3, 8.3] in 2007 and 9.6%, 95% CI [6.5, 12.6] in 2015) and Thailand (6.4%, 95% CI [1.1, 11.3] in 2008 and 11.6%, 95% CI [7.8, 15.3] in 2015) than in Indonesia (not significant in 2007; 4.6%, 95% CI [2.6, 6.7] in 2015) and Myanmar (not significant in 2007; 5.4%, 95% CI [2.1, 8.5] in 2016), while drug use contributed the least among the major risk factors. When controlling for these risk factors, age was not consistently associated with suicidal behaviours. Between the two time points, Indonesia and Myanmar showed more significant risk factors in the later surveys, while the results remained consistent in the Philippines and Thailand.
Table 3. Population attributable fractions (PAF) for associated factors of suicidal thoughts and/or plan in the past 12 months
Only significant PAFs (p < 0.05) were included.
The decomposition analyses (Supplementary Table 4) reveal the contributions (in percentages) of changes in the distribution (endowment effect) and the impact (coefficient effect) of each risk factor to the overall change in the proportion of suicidal behaviours between the two time points in each country. In Indonesia, the proportion of suicidal behaviours increased from 5.0 to 8.6% (Table 2). This overall change was mainly attributed to the constant term, which represents the effects of unmeasured or unexplained factors. However, reductions in being physically attacked/bullied, social difficulty and lack of parental supervision together accounted for a 95% decrease in the change in suicidal behaviours, counteracting the increase from other sources. In Myanmar, suicidal behaviours increased from 0.7 to 10.7%. This substantial rise was primarily attributed to the increases in the proportions of being physically attacked/bullied (from 37.8 to 62.7%) and lack of parental supervision (from 21.9 to 37.8%). In Thailand, the proportion of suicidal behaviours increased from 13 to 20.9%. This overall change was explained by increases in social difficulty, drinking alcohol and lack of parental supervision and the contribution of unmeasured factors (constant term). The change in suicidal behaviours in the Philippines was minimal (from 17.3 to 17.4%) and no individual factor made a substantial change.
The PAFs for the risk factors of suicide attempts in 2015/2016 (Table 4) were mostly consistent with those of suicidal behaviours. Being physically attacked or bullied (ranging from 35.1%, 95% CI [21.6, 46.3] in Indonesia to 49.0%, 95% CI [40.7, 56.2] in the Philippines) and experiencing social difficulties (ranging from 16.2%, 95% CI [4.4, 26.5] in Myanmar to 28.7%, 95% CI [20.4, 36.1] in Indonesia) were the most significant risk factors. However, lack of parental supervision was not significant in Indonesia or Thailand. Being female consistently contributed to the risk of suicide attempts (11.0%, 95% CI [2.1, 19.1] in Thailand to 39.4%, 95% CI [24.3, 51.5] in Myanmar). Alcohol consumption and drug use had similar impacts on the risk of suicide attempts across the countries.
Table 4. Population attributable fractions (PAF) for associated factors of suicide attempt in the past 12 months
Only significant PAFs (p < 0.05) were included.
Discussion
This study estimates that the prevalence of suicidal behaviours among 13- to 15-year-olds was higher in Thailand and the Philippines than in Indonesia and Myanmar. Being physically attacked or bullied and experiencing social difficulties were the most significant and consistent risk factors for suicidal behaviours among young people across countries and time points. The proportions of young people experiencing suicidal behaviours rose in all countries from 2007/2008 to 2015/2016, with the largest increases observed in Myanmar and Thailand. The changes were explained largely by the changes in being physically attacked and bullied, experiencing social difficulties, drinking alcohol and lack of parental supervision.
There are several strengths to this study. First, the inclusion of data from multiple time points in four Southeast Asian countries allows for cross-country and cross-time comparisons. Second, the samples are nationally representative. Third, data were collected using a questionnaire that had been previously tested and demonstrated a moderate level of inter-rater reliability and acceptable internal consistency. Finally, the study examines several of the most common health risk factors for 13- to 15-year-olds. We acknowledge some limitations. Suicidal behaviours were not measured consistently across time points, as data on suicide attempts were unavailable for 2007/2008. In addition, the study did not include certain factors, such as adverse childhood experiences, including traumatic events and harsh and humiliating punishments, physical illness and family characteristics, including parents’ education levels and family composition. The GSHS did not include young people who are not attending school, but included homeless and refugee children, who are at higher risk of suicidal behaviours. This study did not have data on the deaths caused by suicide, which could be important for policymaking. These data fell outside the scope of the GSHS and cannot be collected about these children at this stage. However, the findings of our study point out the importance of adding them to future surveys. The GSHS questionnaire has a moderate level of inter-rater reliability, which could introduce errors into this study. Finally, we did not have data from during or after the coronavirus disease 2019 pandemic. Research evidence suggests that the pandemic had significant short- and long-term effects on child and adolescent health and well-being (Wolf and Schmitz, Reference Wolf and Schmitz2024). Therefore, it may have influenced the prevalence of suicidal behaviours and their associated factors in these countries (Bridge et al., Reference Bridge, Ruch, Sheftall, Hahm, O’Keefe, Fontanella, Brock, Campo and Horowitz2023).
Being physically attacked or bullied has been well-established as a significant risk factor for adolescent suicidal behaviours (Hawton et al., Reference Hawton, Saunders and O’Connor2012;Phuong et al., Reference Phuong, Huong, Tien, Chi and Dunne2013; Cluver et al., Reference Cluver, Orkin, Boyes and Sherr2015; Sharma et al., Reference Sharma, Nam, Kim and Kim2015; Miranda-Mendizabal et al., Reference Miranda-Mendizabal, Castellví, Parés-Badell, Alayo, Almenara, Alonso, Blasco, Cebria, Gabilondo and Gili2019). This study adds further evidence, showing that these are the most consistent and significant risk factors across a range of countries and cultures. Physical violence and bullying may trigger feelings of shame, depression, anxiety, powerlessness and social isolation, all of which reduce a person’s capacity to cope with stress (Brodsky and Stanley, Reference Brodsky and Stanley2008). There are multiple sources of violence against adolescents, including harsh parental punishment, peer aggression, gender-based violence and teacher-perpetrated violence. Bullying can occur in various forms (physical, psychological and social) and settings (online, at school and in public spaces), though little research has compared the mental health impacts of these different forms.
Social difficulties and loneliness have also been identified as significant risk factors for adolescent suicidal behaviours in previous studies (Hawton et al., Reference Hawton, Saunders and O’Connor2012). The findings from this study support that this relationship is consistent over time and across settings. These results support the interpersonal-psychological model (Joiner, Reference Joiner2005), which emphasises how feelings of thwarted belongingness can drive the contemplation of suicide. Social difficulties may be related to difficult social circumstances, personality traits, low self-esteem, hopelessness or even symptoms of depression, which directly contribute to suicidal behaviours (Carballo et al., Reference Carballo, Llorente, Kehrmann, Flamarique, Zuddas, Purper-Ouakil, Hoekstra, Coghill, Schulze and Dittmann2020).
Parental supervision plays a significant role in adolescent suicidal behaviours. It can be linked to both positive and negative parenting styles (Burešová et al., Reference Burešová, Bartošová and Čerňák2015). Excessive parental supervision may result from authoritarian parenting, which is characterised by strict rules, firm control over children and anxiety. Conversely, appropriate parental supervision is associated with an authoritative parenting style, which involves problem-solving with the child, setting clear rules and expectations and maintaining open communication. A lack of parental supervision is linked to negative parenting, such as neglectful parenting, which involves a lack of involvement and provides little nurturance or guidance, as well as permissive parenting styles, which are child-driven and involve failing to give or enforce rules. In addition, a lack of parental supervision may be associated with poor family relationships, parents’ mental health problems or substance abuse and household financial difficulties, which in turn influence suicidal behaviours. The impact of insufficient parental supervision on adolescents’ suicide attempts is not consistently found across different countries in this study. This suggests that the relationships between parental supervision and adolescents’ mental health outcomes are complex and non-linear and can be modified by culture and tradition.
Drinking and drug use are two well-established health risk behaviours, particularly among adolescents (Swahn and Bossarte, Reference Swahn and Bossarte2007). There is consistent evidence of an association between alcohol consumption and suicidal behaviours. A study conducted in 17 European countries found this association to be significant (odds ratio [OR] = 1.49, 95% CI [1.34–1.66]) (Kokkevi et al., Reference Kokkevi, Rotsika, Arapaki and Richardson2012). The US National Youth Risk Behaviour Survey confirmed that preteen alcohol use initiation was statistically significantly associated with suicidal ideation (adjusted OR = 1.89, 95% CI [1.46–2.44]) and suicide attempts (adjusted OR = 2.71, 95% CI [1.82–4.02]) (Swahn and Bossarte, Reference Swahn and Bossarte2007). Carballo et al.’s (2020) review also indicated that both cross-sectional and longitudinal studies consistently report alcohol misuse as a risk factor for suicidal behaviours in various settings. Similarly, evidence of the relationship between substance use and suicidal behaviour is well-documented in the literature (Miranda-Mendizabal et al., Reference Miranda-Mendizabal, Castellví, Parés-Badell, Alayo, Almenara, Alonso, Blasco, Cebria, Gabilondo and Gili2019). The study of 17 European countries found a significant association between substance use and self-reported suicide attempts among adolescents (OR = 2.22, 95% CI [2.02–2.43]) (Kokkevi et al., Reference Kokkevi, Rotsika, Arapaki and Richardson2012). Carballo et al.’s (2020) review further confirms that illicit drug use increases the risk of suicidal behaviours, and this risk escalates when adolescents use drugs simultaneously with alcohol. However, the relationship between health risk behaviours and mental health problems can be bidirectional, as young people struggling with mental health issues such as depression or anxiety may turn to alcohol or drug use to self-soothe or numb difficult emotions. The data from this study does not allow for disentangling it.
Higher proportions of suicidal behaviours among girls than boys are widely reported (Miranda-Mendizabal et al., Reference Miranda-Mendizabal, Castellví, Parés-Badell, Alayo, Almenara, Alonso, Blasco, Cebria, Gabilondo and Gili2019). In the general population of Western countries, while the rate of suicide deaths is higher among men, women report more suicidal thoughts and attempts. This phenomenon is known as the “gender paradox in suicide” (Canetto and Sakinofsky, Reference Canetto and Sakinofsky1998). Among adolescents, this paradox can be explained by gender differences in emotional and behavioural problems. The higher prevalence of externalising disorders (e.g., conduct disorder, substance use disorder and deviant behaviour) and a preference for highly lethal methods among boys are the main factors contributing to the higher suicide death rates among boys compared to girls (Mergl et al., Reference Mergl, Koburger, Heinrichs, Székely, Tóth, Coyne, Quintão, Arensman, Coffey and Maxwell2015). In contrast, gender differences in the rates of internalising disorders (e.g., anxiety and mood disorders) among girls may account for the higher rates of suicidal thoughts and attempts among girls compared to boys (Allison et al., Reference Allison, Allison, Roeger, Martin and Keeves2001). In addition, higher rates of suicidal behaviours among girls may also be partially explained by gender-based violence. Girls are significantly more likely to be victims of sexual abuse and dating violence, which are major contributing factors to suicidal behaviours (Wunderlich et al., Reference Wunderlich, Bronisch, Wittchen and Carter2001; Miranda-Mendizabal et al., Reference Miranda-Mendizabal, Castellví, Parés-Badell, Alayo, Almenara, Alonso, Blasco, Cebria, Gabilondo and Gili2019).
The data from this study suggest that suicidal behaviours are a significant public health problem among adolescents aged 13–15 years in Southeast Asian countries. However, the prevalence of suicidal behaviours varies widely across these countries. Compared to recent global estimates, the prevalence rates in the Philippines and Thailand are around the global average (Lim et al., Reference Lim, Wong, McIntyre, Wang, Zhang, Tran, Tan, Ho and Ho2019; Biswas et al., Reference Biswas, Scott, Munir, Renzaho, Rawal, Baxter and Mamun2020) and similar to those in other low- and middle-income countries (Uddin et al., Reference Uddin, Burton, Maple, Khan and Khan2019). In contrast, Indonesia and Myanmar have some of the lowest prevalence rates in the world. When compared to the average prevalence in Asian countries (Biswas et al., Reference Biswas, Scott, Munir, Renzaho, Rawal, Baxter and Mamun2020), the rates in Indonesia and Myanmar are similar, whereas the rates in the Philippines and Thailand are substantially higher. In the Southeast Asian region, Peltzer et al. (Reference Peltzer, Yi and Pengpid2017) found that the prevalence of suicide attempts among older adolescents in Thailand was the highest among six Southeast Asian countries, particularly among boys. Pengpid et al. (Reference Pengpid and Peltzer2020) also reported that the prevalence of suicidal behaviours is higher in the Philippines and Thailand than in other countries in the region, including Vietnam, Laos and Indonesia. The higher prevalence of suicidal behaviours in the Philippines and Thailand compared to other countries in the region may be partly explained by a higher prevalence of physical violence and bullying victimisation, health risk behaviours and a lack of parental supervision, all of which significantly contribute to suicidal behaviours (Hawton et al., Reference Hawton, Saunders and O’Connor2012; Carballo et al., Reference Carballo, Llorente, Kehrmann, Flamarique, Zuddas, Purper-Ouakil, Hoekstra, Coghill, Schulze and Dittmann2020).
The data indicate an increasing trend in suicidal behaviours across most of these countries. The substantial increases observed in Myanmar and Thailand can be attributed to the rising levels of some key risk factors, including physical violence, bullying, social difficulty, alcohol use and lack of parental supervision. These findings suggest that policies, programmes and interventions aimed at addressing these factors have the potential to effectively prevent suicidal behaviours among adolescents.
In this study, the proportions of 13- to 15-year-olds reporting both suicidal ideation and attempts among those with suicidal ideation are 36% in Indonesia, 63% in Myanmar, 69% in the Philippines and 59% in Thailand. An analysis of the data from the GSHS across 59 low- and middle-income countries estimated that the pooled prevalence rates of suicidal ideation and suicide attempts among adolescents are similar (Uddin et al., Reference Uddin, Burton, Maple, Khan and Khan2019). Theories and empirical evidence suggest that suicidal ideation leads to suicide attempts (O’Connor and Nock, Reference O’Connor and Nock2014; Castellví et al., Reference Castellví, Lucas-Romero, Miranda-Mendizábal, Parés-Badell, Almenara, Alonso, Blasco, Cebrià, Gabilondo and Gili2017). Nevertheless, individuals who have previously attempted suicide may be more likely to experience recurring suicidal thoughts if the underlying causes persist. In the GSHS, both suicidal ideation and attempts were assessed within the past year. While this does not allow for causal inference, it suggests a strong association between the two.
Conclusion and implications
This study highlights that suicidal behaviours among 13- to 15-year-olds is an increasing concern in Southeast Asian countries, particularly in the Philippines and Thailand. Multiple risk factors have been identified in this study, with physical violence, bullying victimisation and social difficulties being the most significant contributors. It should be approached with caution when generalising the findings on the prevalence in this study to other countries, as the prevalence rates are likely to vary widely across different settings. However, the identified risk factors are more generalizable, as they align closely with evidence from previous studies in other contexts.
The findings from this study have important implications for policymakers, health professionals, social workers, educators and parents. Mental and school health policies should be regularly reviewed and evaluated to allow for necessary updates and improvements to address the needs of adolescents. Locally appropriate, multi-component interventions involving the health, social and education sectors are essential to addressing this public health crisis. These interventions need to target multiple risk factors within society, schools and families while also strengthening adolescents’ coping skills and resilience. Among the identified modifiable risk factors, it is clear that physical violence against adolescents, bullying, and loneliness should be prioritised. In addition, health professionals, social workers, teachers and parents need to enhance their knowledge and skills related to adolescent suicidal behaviours to recognise warning signs early and provide effective support.
Open peer review
To view the open peer review materials for this article, please visit http://doi.org/10.1017/gmh.2025.10030.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/gmh.2025.10030.
Data availability statement
The data that support the findings of this study are available at the Global School-based Student Health Surveys website (https://www.who.int/teams/noncommunicable-diseases/surveillance/systems-tools/global-school-based-student-health-survey).
Acknowledgements
The authors are especially grateful to the adolescents who contributed their experiences to this research and the Global School-based Student Health Surveys staff who contributed to the design and conduct of the surveys.
Author contribution
T.T. and J.F. designed this study. H.N. managed the data. T.T. and H.N. conducted the statistical analysis. T.T. wrote the draft of this paper. All authors provided the interpretation of results and critically reviewed the draft. All authors reviewed and agreed on the content of the final submitted version. T.T. and H.N. have accessed and verified the data. T.T. was responsible for the decision to submit the manuscript.
Financial support
J.F. is supported by a Finkel Professorial Fellowship, which is funded by Finkel Foundation. The funders had no role in study design; in collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.
Competing interests
The authors declare none.
Ethics statement
This study is a secondary analysis of data from the Global School-based Student Health Surveys (GSHS). All GSHS surveys obtained approval in their country by a national government agency and an institutional ethics board or committee.
Open access
Comments
22 Oct 2024
Professor Judy Bass
Editor-in-Chief
The Cambridge Prisms: Global Mental Health
Dear Professor Bass,
We are pleased to submit our manuscript, “Suicidality among Adolescents in Four South East Asian Countries – Trends and Contributing Factors,” for consideration for publication in The Cambridge Prisms: Global Mental Health, Special Issue "Self-harm and Suicide: A Global Priority.
While the importance of adolescent suicidality is well-recognized in high-income countries, it has not been adequately addressed in low- and middle-income countries. This study aimed to examine changes in the prevalence and factors associated with adolescent suicidality in Indonesia, Myanmar, the Philippines, and Thailand—four middle-income countries in Southeast Asia—during 2007/08 and 2015/16, to inform policy and intervention efforts.
The data show that suicidality was prevalent in all countries, with the highest rates observed in Thailand and the Philippines. The prevalence of suicidality increased across all countries during the study period, with the largest rises in Myanmar (10%) and Thailand (7.9%). Being physically attacked, bullied, and experiencing social difficulties were the most significant and consistent risk factors for suicidality among adolescents across countries and time points.
These findings strongly suggest that adolescent suicidality requires more attention from policymakers, as well as from the health, social, and education sectors, and parents. Locally appropriate, multicomponent interventions targeting multiple risk factors within society, schools, and families are urgently needed to mitigate this public health crisis.
We are grateful for your consideration of this manuscript, which we believe would be of interest to readers of The Cambridge Prisms: Global Mental Health.
Yours sincerely,
Thach Tran, PhD, BEc, MIRB, MSc
Senior Research Fellow
Global and Women’s Health
School of Public Health and Preventive Medicine
Monash University