Risk and protective factors associated with adolescent depression in Singapore: a systematic review

Abstract

Introduction:

Adolescent depression is prevalent, and teen suicide rates are on the rise locally. A systemic review to understand associated risk and protective factors is important to strengthen measures for the prevention and early detection of adolescent depression and suicide in Singapore. This systematic review aims to identify the factors associated with adolescent depression in Singapore.

Methods:

A systematic search on the following databases was performed on 21 May 2020: PubMed, EMBASE and PsycINFO. Full texts were reviewed for eligibility, and the included studies were appraised for quality using the Newcastle Ottawa Scale. Narrative synthesis of the finalised articles was performed through thematic analysis.

Results:

In total, eight studies were included in this review. The four factors associated with adolescent depression identified were: (1) sociodemographic factors (gender, ethnicity); (2) psychological factors, including childhood maltreatment exposure and psychological constructs (hope, optimism); (3) coexisting chronic medical conditions (asthma); and (4) lifestyle factors (sleep inadequacy, excessive internet use and pathological gaming).

Conclusion:

The identified factors were largely similar to those reported in the global literature, except for sleep inadequacy along with conspicuously absent factors such as academic stress and strict parenting, which should prompt further research in these areas. Further research should focus on current and prospective interventions to improve mental health literacy, targeting sleep duration, internet use and gaming, and mitigating the risk of depression in patients with chronic disease in the primary care and community setting.

INTRODUCTION

Adolescence (10–19 years, World Health Organization [WHO])[1] is a period of rapid physical, cognitive and social transformation that leads to acquisition of the roles and responsibilities of adulthood. While adolescents are mostly physically healthy, this period is marked by experimentation and risk taking,[2] and significant morbidity in terms of mental illness. Half of all mental health problems have an onset before the age of 14, and three-quarters before the age of 25.[3]

Depression is a leading cause of illness and disability globally. With an estimated 1-year prevalence of 4%–5% of depression among youth in the mid- to late-adolescent age group,[4,5] it poses a threat to adolescent well-being. Adolescent depression is associated with health-compromising behaviours, such as smoking and substance abuse, and a higher prevalence of and poorer outcomes for chronic medical conditions such as obesity[6] and insulin-dependent diabetes mellitus.[7] It also heightens the risk of suicide,[8] the third leading cause of death among 15–19-year-olds worldwide,[9] and is an impediment to socio-educational attainment.[10,11] The United Nations Youth Strategy (UN2030) has prioritised the provision of youth-friendly mental health services to address the adverse trajectories of patients and their families that may result from untreated adolescent depression.[10,11]

Locally, the Singapore Mental Health Survey (SMHS 2016)[12] found that the lifetime prevalence of depression in adults aged 18 years and above increased from 5.8% in 2010 to 6.3% in 2016. Mental disorders contribute to 20%–25% of disability-adjusted life years (DALYs), 20%–25% of years of life lost (YLL) and 25%–30% of years lived with disability (YLD). In Singapore, anxiety and depressive disorders are the leading cause of disease burden among 15–34-year-olds and rank fifth in the 0–14 years age category. In 2019, almost 2,000 youth (aged 16–30 years) engaged CHAT (Community Health Assessment Team), a community-based youth mental health outreach, an increase of 36 times from its inception in 2009.[13,14] In Singapore, suicide remains the leading cause of death for persons aged 10–29 years. Among adolescents aged 10–19 years, a high of 19 suicides among teen males were documented in 2018. Among females, ten teen suicides were documented in 2019, a more than three-fold increase from 2018.[15] Hence, at the 2019 Youth Conversations organised by the National Youth Council, mental health was cited as an issue of national concern among youth.

There are many factors associated with depression. These include a strong family history of depression,[16,17,18] sleep deprivation, genetic factors[19] and the presence of chronic medical conditions. Predisposing factors include history of mental disorders, attempted self-harm, disruptive behaviour, learning problems, negative body image, heightened sensitivity to loss and rejection, chronic adversity including maltreatment, family discord, bullying and poverty,[20] as well as frequent clinic attendance, lack of social support and lower socioeconomic status. Stressors such as academic pressure, abuse, personal injury and loss[20,21,22] act as additional risk factors.

Singapore is a Southeast Asian city-state at the crossroads of global trade, imbued with a sociocultural amalgamation of Asian collectivism[23,24,25] and Western individualism,[26] and oriented nationally towards academic and economic competitiveness.[27] This systematic review seeks to understand the risk and protective factors associated with depression among adolescents growing up in multicultural Singapore.

METHODS

Design and search strategy

The protocol for this review was designed based on the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) statement. We conducted a systematic search on the following databases: PubMed, EMBASE and PsycINFO on 21 May 2020 with adapted search terms for each database [see Appendix]. The following search terms were used: ‘depressi* OR MDD OR dysthymia OR dysthymic disorder’ AND ‘adolescent* OR adolescence OR teen* OR youth*’ AND ‘singapore’ AND ‘factor OR associat* OR relationship* OR Predict OR Prevalence OR Course OR prognos* OR cause OR etiolog* OR aetiolog*’. In order to ensure sufficient data, the search strategy did not limit studies by study design.

Selection criteria

All reports that examined the risk factors for depression in the adolescent population were eligible for inclusion. We took adolescence to refer to individuals aged between 10 and 19 years according to the WHO definition.[1] We considered articles for review if they satisfied the following inclusion criteria: peer-reviewed scientific reports of original research; English language articles; study population with a mean age of 10–19 years; study population resided in Singapore; and studies that reported associations with depression as an outcome measure. The exclusion criteria were informal publications (such as commentaries, letters to the editor, editorials, meeting abstracts), review papers, non-English papers, studies with lack of access to full texts and studies that did not focus on depression as an outcome. The selection of documents for eligibility was evaluated independently, first using the title and abstract and subsequently using the full text, to ensure they met the inclusion and exclusion criteria.

Quality appraisal

All included studies were appraised using the Newcastle Ottawa Scale. The scores for risk of bias are tabulated in Table 1. The quality appraisal process was peer reviewed by the team of authors. The studies were ranked to be of good, fair or poor quality based on the Agency for Healthcare Research and Quality (AHRQ) standards [Table 2].

Table 1.

Risk of bias assessment.

Study	Newcastle Ottawa Quality Assessment Scale
	Selection (max. 5)	Comparability (max. 2)	Outcome (max. 3)
Magiati et al. (2015)[30]	5	0	0
Mythily et al. (2008)[37]	2	0	2
Peh et al. (2017)[32]	2	1	2
Wong and Lim (2009)[33]	4	2	2
Yeo et al. (2019)[35]	5	2	2
Lu et al. (2014)[34]	2	2	2
Lo et al. (2018)[36]	1	2	3
Gentile et al. (2011)[38]	3	2	2

Table 2.

Quality assessment interpretation (AHRQ standards).

Quality assessment	Newcastle Ottawa Quality Assessment Scale
	Selection (max. 5)	Comparability (max. 2)	Outcome (max. 3)
Good quality	3 or 4	1 or 2	2 or 3
Fair quality	2	1 or 2	2 or 3
Poor quality	0 or 1	0 or 1	0 or 1

AHRQ: Agency for Healthcare Research and Quality

Data synthesis

The included papers were qualitatively and thematically analysed to produce a narrative synthesis.[28] A preliminary synthesis was first undertaken by one member of the research team, and the identified categories were subsequently cross-checked and discussed with another member. The researchers met to discuss any disagreements that arose, and any differences in opinion were resolved through consensus. No ethics approval was required for this review.

RESULTS

Search results

The literature search yielded 94, 116 and 18 search results from PubMed, EMBASE and PsycINFO, respectively, totalling 228 studies. After 59 duplicates were removed, 169 titles and abstracts were screened based on the inclusion and exclusion criteria, after which 128 articles were excluded. Of the remaining 41 full-text articles, 33 additional studies were excluded based on the exclusion criteria, leaving eight articles finally. With no restrictions on study design, we have included seven cross-sectional studies and one interventional study. Given the diversity of risk factors, variability in study designs and lack of common data, a meta-analysis could not be performed.

All eight studies were conducted in Singapore, with sample sizes ranging from 108 to 2998 and study settings ranging from hospital (n = 2) to school (n = 6) settings. Regarding the depression measures used, one study did not use a validated depression scale, two studies used similar depression scales and one study used a depression scale validated in Singapore (Asian Adolescent Depression Scale).[29] A summary of the search results is presented in the PRISMA flowchart [Figure 1]. The study characteristics and summary of the results are presented in Table 3.

Figure 1.

PRISMA flowchart shows the selection process for the current systematic review. [Adapted from: Moher D, Liberati A, Tetzlaff J, Altman D G. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement BMJ 2009; 339.]

Table 3.

Summary of study characteristics and results.

Study	Themes according to theoretical framework: social, biological, psychological	Study design	Setting	Sample source and source of data	Sample size	Depression measure used	Results
Magiati et al. (2015)[30]	Social: ethnicity Biological: gender	Cross-sectional	School	Self-reported questionnaire in 18 primary schools	1,655	CDI	Chinese had a lower CDI total, negative mood, anhedonia and ineffectiveness than Malays. Males scored higher for ‘interpersonal problems’ than females. Compared to the US population from another study, Singapore children had higher CDI, with females scoring higher in several CDI domains
Mythily et al. (2008)[37]	Psychological: excessive internet use	Cross-sectional	School	Self-reported questionnaire in three secondary schools	2,735	Not reported	Participants who reported excessive internet use, defined by the author as >5 h a day, were more likely to feel sad or depressed
Peh et al. (2017)[32]	Psychological: traumatic childhood	Cross-sectional	Hospital	Self-reported questionnaire in an adolescent psychiatry outpatient clinic	108	PHQ-8	Severity of maltreatment exposure (measured to include physical, sexual and emotional abuse and physical and emotional neglect) was associated with depressive symptoms
Wong and Lim (2009)[33]	Psychological: hope and optimism	Cross-sectional	School	Self-reported questionnaire in a secondary school	340	CES-D	Optimism scores and depression scores were negatively correlated. Hope scores and depression scores were negatively correlated. Hope and optimism contributed significantly to depression even after controlling for optimism and hope, respectively. Hope did not account for depression beyond what was accounted for by optimism. Optimism, pessimism and agency were significantly predictive of depression
Yeo et al. (2019)[35]	Biological: sleep deprivation	Cross-sectional	School	Self-reported questionnaire in six local and international secondary schools	2,313	11-item Kutcher Adolescent Depression Scale	Shorter sleep was associated with a higher global depression score. Having <7 h of sleep compared to an age-appropriate (8−10 h) amount of sleep was associated with depressive symptoms
Lu et al. (2014)[34]	Biological: asthma	Cross-sectional	Hospital	Self-reported questionnaire in one hospital	171	Revised Child Anxiety and Depression Scale	Adolescents with poorly controlled asthma reported significantly more depressive symptoms than subjects with well-controlled asthma and healthy controls. Number of subjects who were disturbed by depressive symptoms was found to be higher among participants with asthma, especially those with poorly controlled asthma. Increased ACT score was associated with lower depression scores
Lo et al. (2018)[36]	Biological: sleep deprivation	Cohort study	School	Self-reported questionnaire in all-girls secondary school	375	11-item Kutcher Adolescent Depression Scale	Increase in TIB on weekdays was associated with decrease in depressive symptoms
Gentile et al. (2011)[38]	Psychological and social: pathological video gaming	Cohort study	School	Self-reported questionnaire in six primary schools and six secondary schools	2,998	Asian Adolescent Depression Scale	Children with more pathological gaming symptoms at baseline had higher levels of depression at 2-year follow-up. Increase in pathological gaming symptoms between baseline and follow-up was associated with further increase in depression at follow-up. Those who became pathological gamers ended up with increased depression, and those who stopped being pathological gamers had lower depression than those who remained pathological gamers

ACT: Asthma Control Test, CDI: Children’s Depression Inventory, CES-D: Centre of Epidemiological Studies-Depressed Mood Scale, PHQ-8: 8-item Patient Health Questionnaire, TIB: time in bed

Risk of bias assessment

The risk of bias assessment scores ranged from 4 to 9 [Table 1]. Half the studies did not have a representative sample (selection score of 2 and below) mostly because they had a target group of a selected group of users rather than a random or non-random sample. Some papers also fared lower in the representativeness of the sample because they did not include the response rate of responders/non-responders or did not justify their sample size. Nonetheless, most of the studies controlled for factors and had good quality of outcomes reported.

Factors identified

The factors identified were categorised into: (1) sociodemographic factors (gender, ethnicity); (2) psychological factors, including childhood maltreatment exposure and psychological constructs (hope, optimism); (3) coexisting chronic medical conditions (asthma); and (4) lifestyle factors (sleep inadequacy, excessive Internet use, pathological gaming).

Sociodemographic factors

Magiati et al.[30] found that a substantial minority (16.9%) of primary school-aged Singapore children (age 8–12 years) reported depressive symptoms based on the Children’s Depression Inventory (CDI), which was higher than that reported by Ramli et al.[31] for Malaysian secondary school-aged adolescents (10.3%).

Although little clinically significant difference in CDI scores was found between Malay, Indian and Chinese children, Chinese children were observed to report lower scores for depressive symptoms (negative mood, interpersonal problems and anhedonia, and total CDI score) compared to their Malay and Indian peers.

While some gender differentiation was noted between Indian girls who reported more emotional symptoms than Indian boys, the study found no clinically significant difference in total depression scores between males and females.

Psychological factors

Peh et al.[32] who recruited adolescents (age 14–19 years) from a psychiatric hospital reported increased depressive symptoms among patients with more severe childhood maltreatment exposure. The psychometric instruments in this study collected data on physical, sexual and emotional abuse, physical and emotional neglect, self-harm behaviours, emotional dysregulation and depressive symptoms. In this study, 75.9% reported at least one episode of self-harm, while 50.5% reported ten or more episodes in the past 12 months. The study found that emotional dysregulation mediated the association between severity of maltreatment exposure and self-harm frequency.

Wong and Lim[33] explored the discriminant validity of ‘hope’ and ‘optimism’ on depression and life satisfaction among Singapore secondary school students. Both constructs were oriented towards positive outcome expectancies, correlating positively with life satisfaction and negatively with depression. While ‘optimism’ was related more to general positive outcomes, ‘hope’ had the element of personal self-efficacy in goal attainment. ‘Hope’ includes the sub-elements of agency thinking (the motivation to pursue one’s goals) and pathway thinking (the planning of a pathway towards goals attainment). Further correlational analysis found that agency thinking correlated negatively with depression, while pathway thinking did not. Interestingly, the study noted gender differences in that males tended to score higher in the ‘hope’ domain, which was protective for depression symptoms.

Chronic medical conditions

In a cross-sectional study, Lu et al.[34] recruited adolescent patients (age 12–19 years) with asthma from a hospital and then compared the rates of depression with age- and gender- matched non-asthma control individuals from the same neighbourhood. They found that adolescents with poorly controlled asthma had higher scores of depression compared to healthy controls and patients whose asthma was well controlled.

Lifestyle factors

Two studies reported the effect of sleep deprivation on depressive symptoms. Not achieving age-appropriate sleep duration (8–10 h) on school nights was associated with significantly increased odds for depressive symptoms (sadness, irritability, worthlessness, low motivation, thoughts of self-harm) as well as poorer self-rated health and increased odds of being overweight.[35,36] In an interventional study investigating the effects of delaying school start time by 45 min, a sustained increase of time in bed (TIB) and total sleep time (TST) was documented at 9 months. This was associated with improved alertness, reduction in depressive symptoms and improvement in depression scores (from baseline to 9 months change: −1.22 ± 0.48, P < 0.01).[35,36]

Two studies reported associations between excessive use of the internet and related media and depressive symptoms in adolescents. One cross-sectional study reported that 17.1% of adolescents used the internet excessively (defined as more than 5 h of internet use a day), and this was significantly associated with lack of home rules regarding internet use, a lower likelihood of having a confidant, feelings of sadness and depression and perceived poorer grades in school.[37] Gentile et al.[38] conducted a prospective cohort study in a population of secondary school children over 2 years to research the association of pathological gaming with depression in adolescents. The study found the prevalence of pathological gaming to be around 9%. Predictor variables of pathological gaming included impulsiveness, lower social competence and poorer emotional regulation skills. In turn, severity of pathological gaming was associated with depression, anxiety, social phobia and poorer grades. These improved when the individual stopped being a pathological gamer.

DISCUSSION

This review synthesised the scientific literature for factors associated with depression among adolescents in Singapore. Despite a modest yield of eight papers, the findings from this review are reflective of the range of factors associated with adolescent depression locally. The summaries of recommendations are provided in Tables 4 and 5.

Table 4.

Summary of recommendations based on intervention.

Factor	Detection	Recommendation
Mental health literacy	Improved mental literacy in the community (parents, teachers, adolescents) will reduce stigma and allow for early detection of mental illness	Strengthen mental health literacy in the community, including parents, teachers, guardians, students, adolescents. A community-wide effort to strengthen mental health literacy may reduce mental health stigma and aid in the early detection and treatment of depression in the community
		Train adolescents in emotional resilience/regulation. Programmes within schools include the SEL[118] items in schools. A wide range of programmes available to the public include the tMHFA and YMHFA,[125] mindfulness-based programmes and youth-empowerment programmes
		Skills in emotional regulation, coping with adversity and identifying maltreatment,[133,134] and identification and modification of maladaptive cognitions/beliefs and behaviours can be taught to adolescents[135,136,137,138,139]
Primary care practitioner	Primary care physicians should be trained to develop a health index of suspicion and detect depression risk among patients in the community	Healthcare providers should be equipped for early detection and treatment of adolescent patients with mental health problems. This may include asking specific questions in the history, using the HEADSSS framework for health risk behaviour and mental health screening and using screening tools in the waiting area
		Further mental health training opportunities include the GDMH and counselling skills
Sleep	Screening for sleep deprivation or poor sleep hygiene in adolescents	Adolescents should be asked about their sleep
		Recommend a sleep duration of at least 6 h per night
		Sleep enhancement: Sleep duration can be improved with adjustment of school start time, reduction of school workload,[122] setting appropriate bedtime[123] and reinforcing sleep hygiene and parental training
Internet use/gaming	Screening for excessive internet use and gaming addiction	Limiting internet exposure: Parental limitation of internet usage and gaming and rescheduling sleep-disrupting media use at night.[140] Family therapy[141] forms as a great support system to prevent internet addiction in the future
		Adolescents should be asked about internet use and use of social media. The HEADSSSS tool (with additional S for social media) for screening can be used[87,88,89]
Chronic disease	Screening for concomitant depression in adolescents with underlying chronic disease	Controlling chronic comorbidity: Adequate medical treatment for underlying medical comorbidity can reduce discomfort and the risk of depression
		Explore the mental health of patients with poorly controlled chronic disease
		Known factors (not limited to) insulin-dependent diabetes mellitus,[61,62] inflammatory bowel disease[63,64] and asthma[34]

GDMH: graduate diploma in mental health, SEL: socio-emotional learning, tMHFA: teen Mental Health First Aid, YMHA: Youth Mental Health First Aid

Table 5.

Summary of recommendations based on factors.

Factor identified	Subfactor	Reference	Recommendations
Sociodemographic	Gender and ethnicity	Magiati et al. (2015)[30]	Identifying those at risk for earlier detection of mental health issues
			Better mental health literacy: Greater advocacy for mental health awareness will aid in early detection of depression and allow prompt treatment
			Examples (not limited to): Teen and Youth Mental Health First Aid,[125] Silver Ribbon,[127] SAMH[126]
Psychological	Childhood maltreatment exposure	Peh et al. (2017)[32]	Screening for emotional dysregulation or maltreatment history in adolescents to determine possible underlying depression risk
	Psychological constructs (hope, optimism)	Wong and Lim (2009)[33]	Emotional resilience training: Skills in emotional regulation and coping with adversity
			Examples (not limited to): REACH,[120,121] cultivating resilience,[129,130,131,132] identifying maltreatment,[133,134] modification of cognition/beliefs and behaviour[135,136,137,138,139]
Chronic medical condition	Chronic medical conditions	Lu et al. (2014)[34]	Screening for concomitant depression in adolescents with underlying chronic disease
			Controlling chronic comorbidities: Adequate medical treatment for underlying medical comorbidities can reduce discomfort and the risk of depression
			Known factors (not limited to) insulin-dependent diabetes mellitus,[61,62] inflammatory bowel disease[63,64] and asthma[34]
Lifestyle	Sleep duration	Yeo et al. (2019)[35]	Screening for sleep deprivation or poor sleep hygiene in adolescents
		Lo et al. (2018)[36]	Sleep enhancement: Sleep duration can be improved with adjustment of school start time, reduction of school workload,[122] setting appropriate bedtime,[123] reinforcing sleep hygiene and parental training
			Recommended sleep duration: 6 h per night
	Excessive internet use and related media	Gentile et al. (2011)[38]	Screening for excessive internet use and gaming addiction
		Mythily et al. (2008)[37]	Limiting internet exposure: Parental limitation of internet usage and gaming and rescheduling sleep-disrupting media use at night.[140] Family therapy[141] forms as a great support system to prevent internet addiction in the future
			HEADSSSS tool (with additional S for social media) for screening can be used[87,88,89]

REACH: Response, Early interventions and Assessment in Community mental Health, SAMH: Singapore Association for Mental Health

Sociocultural factors

Sociocultural factors play a significant role in terms of mental wellness.[39,40] A study of local adults has shown that Indians have significantly higher rate of depression compared to Chinese and Malays.[41] While Magiati et al.’s[30] paper did not show any significant differences between Chinese, Indian and Malay primary school children in terms of self-reported depression, the tendency towards lower scores for depressive symptoms among Chinese students compared to their Malay and Indian counterparts may be related to the adversity experienced by the ethnic minority groups[42] and possibly an under-reporting of symptoms of Chinese children in order to avoid ‘loss of face’.[43]

In terms of gender, adolescent females have been observed to be at a higher risk of depressive symptoms,[44,45,46] attributable in part to the pubertal changes in the hormone–brain relationship[47,48] and to the emergence of received sociocultural gender roles that may include the need to internalise symptoms[49] among girls more than among boys.[50,51,52] In this review, primary school-aged Indian girls were found to have higher scores for emotional symptoms than Indian boys, which may be reflective of the increasing influence of traditional gender roles during adolescence.[53] Nonetheless, the full effects of sociocultural differentiation on well-being may not be well reflected in Magiati et al.’s[30] paper as the school-aged participants would have just been at the cusp of assuming their sociocultural and gender roles.

Adverse childhood events

The lifetime prevalence of adverse childhood events (ACE) of Singapore residents aged 18 years and above is 63.9%.[54] ACE and childhood maltreatment are associated with increased risk of adolescent depression and mental disorders across the lifespan.[55,56,57] Postulated mechanisms include the scars of increased self-criticism, negative cognitive styles and the fear of rejection later in life.[32] Peh et al.[32] identified emotional dysregulation as a mediator between childhood maltreatment exposure and the risk of self-harm. Wong and Lim,[33] on the other hand, explored the hope and optimism constructs and their negative correlation with depression.[58,59] This concurs with other studies which showed how lack of goal-directed energy (agency thinking) was more predictive of depression than lack of pathway thinking.[60]

Chronic medical conditions

Our review found that poorly controlled asthma was associated with a greater risk of depressive symptoms. This is consistent with other studies which have found that at least one in four adolescents with asthma have symptoms of depression, possibly attributed, at least in part, to restrictions in daily functioning.[34] Chronic conditions linked to adolescent depression in the literature include insulin-dependent diabetes mellitus[61,62] and inflammatory bowel disease.[63,64] These point to the importance of exploring the mental wellness of the adolescent patient who presents at the clinic with poorly controlled asthma.

Sleep

Sleep is crucial for mental wellness, and the cumulative effects of chronic partial sleep deprivation may adversely affect mood and emotional regulation in adolescents.[65,66,67] A nightly sleep duration of 6 h or less increased the risk of depressive symptoms and subsequent major depression.[68] Sleep inadequacy and depression have a bidirectional relationship: insomnia is both a symptom of depression[69] and a factor that adversely affects the success of depression treatment.[70] Sleep inadequacy may be perpetuated by the local school schedules[71] that start 1 h earlier than the stated recommended start time, according to the American Academy of Paediatrics,[72] American Medical Association[73] and the American Academy of Sleep Medicine.[74] This may be reflective of the East Asian inclination to prioritise academic achievement over sleep, with students often staying up late to finish uncompleted school assignments.[75]

Excessive internet use

Touted as a smart city[76] with near-total computer and internet penetration (89% and 98% of households, respectively),[77] Singapore has an adolescent population that widely uses the internet and is adept at online gaming. However, excessive use of the internet and pathological gaming are associated with depressive symptoms. Both may spur a downward spiral that compounds social isolation[78] and reinforces depression.

In a milieu that often values public self-restraint[79,80] above public self-expression,[43,81,82,83,84,85] the internet world provides a ready means of escape from the restrictions of the real world.[43] As a mediating factor between internet dependence and depression, the tendency to internalise feelings may play a significant role in the onset of depression and may be predictive of poorer mental health and well-being in middle adulthood.[86] So entwined is the cyber world with the contemporary adolescent experience that ‘social media use’ has been added[87] to the HEADSSS tool[88,89] to encourage practitioners to inquire about cyber wellness when assessing the psychosocial risk of adolescents.

Conspicuously absent: academic stress and parental expectations

Academic achievement is a prime preoccupation in Singapore households. The consistent high national ranking in the Programme for International Student Assessment (PISA),[90,91] regular stellar performances in the International Baccalaureate (IB) diploma programme[92] and a thriving private supplementary tutoring mill, which starts from the preschool years for some,[93] attest to this national hyperfocus.[94] Though schoolwork was ranked by youths as the most stressful aspect of their lives in a national survey,[95] it was surprising that academic stress was not explicitly studied as a factor in the papers reviewed.

Closely linked to academic stress and equally conspicuous because of its absence is the matter of parental expectations. Different from Western households that may value social development[96] no less than academic achievement, schooling may often be a major focus of family time, effort and source of contention in East Asian households.[97,98,99] Risk factors that relate to parenting styles and depression include parental overprotection,[100,101,102] psychological control[103] and the teenage fear of being the source of parental disappointment.[104] Compared to their European–American counterparts, Asian parents tend to adopt an authoritarian parenting style[105,106] that demands from children the commitment to hard work, self-discipline and obedience,[107] often measured by academic performance. The adolescent tussle for autonomy in such circumstances is a ground for distress and depression,[108] and practitioners do well to provide guidance to parents on the effects of parenting on adolescents’ well-being. The high reported rate of depression (CDI, 16.9%) in Magiati et al.’s[30] paper, articles on sleep deprivation on school nights[109,110] and their effects on school performance and depression,[66,67,111,112,113] as well as studies on excessive computer use[37,38] and their adverse effects on school performance allude to the influential role that parenting styles and academic stress play in adolescent mental health in Singapore.

Comparison with our counterparts

Singapore is a multiethnic, multilingual society with coexisting traditionally Asian and modern Westernised values, with risk factors for depression largely similar to those reported in the global literature.[114,115,116] Despite Western influence, local mental health literacy falls short of that other developed countries, with a pervasive tendency to label depression as a ‘weakness’ caused by lack of resilience,[39] rather than as a medical condition. With influence from traditional Asian values, Singaporean children tend to have higher rates of internalising problems compared to externalising problems,[43] which is the opposite of that seen in Western children, further deterring early detection of adolescent depression.

Practical implications

This review underlies the need for an all-of-society effort to address adolescent mental health. In addition to policy-level initiatives, this would include enlisting schools, social service organisations (SSOs), psychological services, ground-up community efforts,[117] as well as the primary and hospital specialist healthcare services as part of an overarching evidence-informed strategy to address youth mental wellness.

Sleep habits

Sleep deprivation is both a harbinger and symptom of distress and depression. Primary care physicians should enquire about the sleep patterns and sleep hygiene of an adolescent patient during consultation, and encourage a healthy sleep routine.

Internet use

Adolescents and their parents who present at the clinic should likewise be asked about the adolescent patient’s use of the internet. Excessive internet use may prompt the need to screen for distress, schoolwork and relationships, as well as depression and other mental health problems. The accessibility of primary care in the community makes it an excellent venue to enquire about internet use and the adolescents’ experience of stress and coping and to provide opportunistic advice on healthy gaming and internet use, particularly for sleep-deprived gamers.

Chronic disease

In our review, poor asthma control was found to be a risk for adolescent depression. Given the bidirectional relationship of chronic disease control and depression, the patient with poor chronic disease control should as much be screened for depressive symptoms[34,61,62,63,64] as a patient with depressive symptoms is asked about the control of his/her chronic disease condition.

Mental health literacy

The strengthening of mental health literacy across all levels of society — policymaking, school, parental/family and personal/adolescent levels — facilitates the possibility of a collective strengthening of emotional awareness and self-regulation in the community and earlier help-seeking for mental health matters among adolescents.

Educational efforts to address adolescent mental wellness in our schools include socio-emotional learning (SEL)[118] in the Character and Citizenship Education (CCE) curricula widely embedded in primary and secondary schools.[119] The system of Response, Early interventions and Assessment in Community mental Health (REACH)[120,121] team working with school counsellors addresses the mental health issues that may arise in the school setting.[117,120,121] In addition, improving sleep through bold interventions, such as the adjustment of school start times, a reduction of schoolwork,[122] a deliberate setting of bedtimes,[123] reinforcement of good sleep hygiene and parental education, plays an important role in adolescent mental wellness.

At the community level, SSOs and ground-up initiatives provide a diverse range of community-based youth services and empowerment programmes for youth (e.g. Impart,[124] mental health first aid (MHFA),[125] Project 180,[124] Singapore Association for Mental Health (SAMH),[126] Silver Ribbon,[127] TOUCH Youth Intervention[128]). They equip adolescents with the skills to cope with adversity, cultivate resilience,[129,130,131,132] identify maltreatment,[133,134] modify maladaptive cognitions[135,136,137,138,139] and address pathological Internet use; also, they provide interventions to reduce sleep disruption[140] and family therapy[141] if and when required. Teen Mental Health First Aid (tMHFA) and Youth Mental Health First Aid (YMHFA) are programmes that have been shown to be effective in improving mental health literacy.[125]

Role of primary care

Noticeably absent in this review was the role of primary care in supporting adolescent mental health. Perhaps the body of work of primary care physicians in managing adolescent mental health has been left largely unpublished. As the first port of call for patients in the community, primary care physicians in the polyclinics and the approximately 2,000 primary care clinics in private practice should be equipped for the early detection and prompt intervention of adolescent depression when necessary. Local training opportunities for primary care physicians include the Graduate Diploma in Mental Health, and other programmes such as positive psychology, cognitive behavioural therapy and other counselling skills. Primary care physicians also need to be connected to local resources, such as family service centres which may provide counselling and therapy services to clients.

Recommendations for research

This review points to further areas of research in adolescent mental health. Well-designed interventional studies to improve sleep, address excessive internet use, improve the emotional resilience and mental health literacy of adolescents, and studies that venture to modify parenting styles and/or address academic stress may be of value. Research may also branch into factors associated with depression that were not surfaced in this review, such as adolescent temperament,[142] social isolation and loneliness,[143] peer rejection,[144] parental history of depression,[145,146] poor parent–child relationship,[147] poor family attachment and detachment from family activities.[148,149]

On a backdrop of the 2008 WHO/WONCA joint report recommending the greater integration of mental health in primary care,[150] the potential for primary care involvement in adolescent mental health is notable. As mental illness remains largely undetected even in patients seeking medical help, there is value in reducing the missed opportunities for early detection and management of mental illness for adolescents who show up at primary care clinics in the community, especially since onset during adolescence is more frequent.

A health services research agenda to improve the mental health of adolescents in Singapore should start with a mapping of the extent that primary and secondary health care is involved in adolescent mental health detection and management. Interventions such as practitioner training, the strengthening of partnerships between primary and secondary care, and the use of screening questionnaires and technological innovations[151,152] that have shown promise elsewhere to improve mental illness detection at primary care may, subsequently be planned.

CONCLUSION

The results of the eight studies included in this review serve to prompt healthcare practitioners to further explore the mental wellness of adolescent patients in the clinic. In addition, they underline the importance of an all-of-society effort to strengthen mental health literacy in the population to stem the inordinate number of teen suicides seen in recent years. Health services research in adolescent mental health that enlists the extensive reach and care continuity characteristic of primary care in collaboration with paediatricians and psychiatrists shows promise for developing a home-grown evidence base for the improved mental health outcomes of adolescents in Singapore.

Conflicts of interest

There are no conflicts of interest.

Supplemental digital content

Appendix at http://links.lww.com/SGMJ/A159

APPENDIX

Search strategy and results

A search was conducted on 4th April 2020 on PubMed (94), Embase (116), and PsycINFO (18).

Supplementary Table 1.

PubMed

No.	Searches	Results
#1	(“Depression”[Mesh] OR “Depressive Disorder”[Mesh]) OR (“Depressi*”[title/abstract] OR “MDD”[title/abstract] OR “Dysthymia”[Title/Abstract] OR “Dysthymic disorder”[Title/Abstract])	431,555
#2	(“Adolescent”[Mesh] OR “Child”[Mesh]) OR (“Adolescent*”[title/abstract] OR “adolescence”[title/abstract] OR “teen*”[title/abstract] OR “youth*”[title/abstract])	308,2035
#3	(“Singapore”[Mesh]) OR (“Singapore*”[Title/Abstract])	19,344
#4	“Factor*”[Title/Abstract] OR “associat*”[Title/Abstract] OR “relationship*”[Title/Abstract] OR “predict*”[Title/Abstract] OR “prevalence”[Title/Abstract] OR “course”[Title/Abstract] OR “prognos*”[Title/Abstract] OR “Cause”[Title/Abstract] OR “Etiolog*”[Title/Abstract] OR “Aetiolog*”[Title/Abstract]	960,0393
#5	#1 AND #2 AND #3 AND #4	94

Supplementary Table 2.

Embase

No.	Searches	Results
#1	‘depression’/exp OR ‘depressi*’:ti,ab OR ‘mdd’:ti,ab OR ‘dysthymia’:ti,ab OR ‘dysthymic disorder’:ti,ab	704,475
#2	‘adolescent’/exp OR ‘child’/exp OR ‘adolescent*’:ti,ab OR ‘adolescence’:ti,ab OR ‘teen*’:ti,ab OR ‘youth*’:ti,ab	377,4927
#3	‘singapore’/exp OR ‘singapore*’:ti,ab	29,976
#4	‘factor*’:ti,ab OR ‘associat*’:ti,ab OR ‘relationship*’:ti,ab OR ‘predict*’:ti,ab OR ‘prevalence’:ti,ab OR ‘course’:ti,ab OR ‘prognos*’:ti,ab OR ‘cause’:ti,ab OR ‘etiolog*’:ti,ab OR ‘aetiolog*’:ti,ab	12,460,808
#5	#1 AND #2 AND #3 AND #4	116

Supplementary Table 3.

PsycINFO

No.	Searches	Results
#1	(Depressi* OR MDD OR Dysthymia OR Dysthymic disorder).ti,ab.	284,185
#2	(Adolescent* OR adolescence OR teen* OR youth*).ti,ab.	301,282
#3	Singapore*.ti,ab.	4,269
#4	(Factor* OR associat* OR Relationship* OR Predict* OR Prevalence OR Course OR Prognos* OR Cause OR Etiolog* OR Aetiolog*).ti,ab.	2,123,263
#5	#1 AND #2 AND #3 AND #4	18

Funding Statement

Nil.