Prevalence of adolescent mental disorders in Kenya, Indonesia, and Viet Nam measured by the National Adolescent Mental Health Surveys (NAMHS): a multi-national cross-sectional study.

BACKGROUND: Mental disorders are the leading global cause of health burden among adolescents. However, prevalence data for mental disorders among adolescents in low-income and middle-income countries are scarce with often limited generalisability. This study aimed to generate nationally representative prevalence estimates for mental disorders in adolescents in Kenya, Indonesia, and Viet Nam. METHODS: As part of the National Adolescent Mental Health Surveys (NAMHS), a multinational cross-sectional study, nationally representative household surveys were conducted in Kenya, Indonesia, and Viet Nam between March and December, 2021. Adolescents aged 10-17 years and their primary caregiver were interviewed from households selected randomly according to sampling frames specifically designed to elicit nationally representative results. Six mental disorders (social phobia, generalised anxiety disorder, major depressive disorder, post-traumatic stress disorder, conduct disorder, and attention-deficit hyperactivity disorder) were assessed with the Diagnostic Interview Schedule for Children, Version 5. Suicidal behaviours and self-harm in the past 12 months were also assessed. Prevalence in the past 12 months and past 4 weeks was calculated for each mental disorder and collectively for any mental disorder (ie, of the six mental disorders assessed). Prevalence of suicidal behaviours (ie, ideation, planning, and attempt) and self-harm in the past 12 months was calculated, along with adjusted odds ratios (aORs) to show the association with prevalence of any mental disorder in the past 12 months. Inverse probability weighting was applied to generate national estimates with corresponding 95% CIs. FINDINGS: Final samples consisted of 5155 households (ie, adolescent and primary caregiver pairs) from Kenya, 5664 households from Indonesia, and 5996 households from Viet Nam. In Kenya, 2416 (46·9%) adolescents were male and 2739 (53·1%) were female; in Indonesia, 2803 (49·5%) adolescents were male and 2861 (50·5%) were female; and in Viet Nam, 3151 (52·5%) were male and 2845 (47·4%) were female. Prevalence of any mental disorder in the past 12 months was 12·1% (95% CI 10·9-13·5) in Kenya, 5·5% (4·3-6·9) in Indonesia, and 3·3% (2·7-4·1) in Viet Nam. Prevalence in the past 4 weeks was 9·4% (8·3-10·6) in Kenya, 4·4% (3·4-5·6) in Indonesia, and 2·7% (2·2-3·3) in Viet Nam. The prevalence of suicidal behaviours in the past 12 months was low in all three countries, with suicide ideation ranging from 1·4% in Indonesia (1·0-2·0) and Viet Nam (1·0-1·9) to 4·6% (3·9-5·3) in Kenya, suicide planning ranging from 0·4% in Indonesia (0·3-0·8) and Viet Nam (0·2-0·6) to 2·4% (1·9-2·9) in Kenya, and suicide attempts ranging from 0·2% in Indonesia (0·1-0·4) and Viet Nam (0·1-0·3) to 1·0% (0·7-1·4) in Kenya. The prevalence of self-harm in the past 12 months was also low in all three countries, ranging from 0·9% (0·6-1·3) in Indonesia to 1·2% (0·9-1·7) in Kenya. However, the prevalence of suicidal behaviours and self-harm in the past 12 months was significantly higher among those with any mental disorder in the past 12 months than those without (eg, aORs for suicidal ideation ranged from 7·1 [3·1-15·9] in Indonesia to 14·7 [7·5-28·6] in Viet Nam). INTERPRETATION: NAMHS provides the first national adolescent mental disorders prevalence estimates for Kenya, Indonesia, and Viet Nam. These data can inform mental health and broader health policies in low-income and middle-income countries. FUNDING: The University of Queensland in America (TUQIA) through support from Pivotal Ventures, a Melinda French Gates company.

Correlates of mental disorder and harmful substance use in an indigenous Australian urban sample: an analysis of data from the Queensland Urban Indigenous Mental Health Survey.

PURPOSE: Limited data exists on the relationship between sociodemographic and cultural variables and the prevalence of specific mental and substance use disorders (MSDs) among Indigenous Australians, using diagnostic prevalence data. This paper utilises data from the Queensland Urban Indigenous Mental Health Survey (QUIMHS), a population-level diagnostic mental health survey, to identify socioeconomic and cultural correlates of psychological distress and specific MSDs in an urban Indigenous Australian sample. METHODS: Using a mixture of household sampling (door-knocking) and snowball sampling (promotion of the survey in the community), 406 participants aged 18 to 89 were recruited across key locations in Southeast Queensland. The study investigated various demographic, socioeconomic, and cultural factors as predictors of psychological distress (measured by the Kessler-5) and MSD diagnoses (utilising the Composite International Diagnostic Interview, CIDI 3.0) using a series of univariate logistic regressions. RESULTS: Individuals in unstable housing (homeless, sleeping rough) and those reporting financial distress were more likely to experience an MSD in the past 12 months and throughout their lifetime. Individuals reporting lower levels of connection and belonging, limited participation in cultural events, and lower empowerment were more likely to have a lifetime mental disorder. CONCLUSION: This data emphasises the importance of addressing systemic and social determinants of health when designing and delivering community mental health services and underscores the need for holistic approaches when working with Indigenous communities.

Using the National Mental Health Service Planning Framework to inform integrated regional planning: a case study in Tasmania, Australia.

BACKGROUND: The aim of this study was to demonstrate the application of a needs-based mental health service planning model in Tasmania, Australia to identify indicative directions for future service development that ensure the equitable provision of mental health services across the State. METHODS: The activity and capacity of Tasmania's 2018-19 mental health services were compared to estimates of required care by: (1) generating estimates of required care using the National Mental Health Service Planning Framework (NMHSPF); (2) collating administrative mental health services data; (3) aligning administrative data to the NMHSPF; and (4) comparing aligned administrative data and NMHSPF estimates to identify priority areas for service development. Findings were contextualised using information about service location, population demographics, and upcoming service development. RESULTS: Bed-based services capacity reached 85% of the NMHSPF estimate. However, access to certain bed types was inequitable across regional areas. Access to jurisdictional clinical ambulatory team-based services was lowest in the South, while overall full-time equivalent staff capacity reached 58% of the NMHSPF estimate. Access to Primary Health Tasmania (PHT) primary care services was highest in the North West; access to Medicare services was highest in the South. Collectively, activity across primary care (PHT, headspace and Medicare) reached 43% of the NMHSPF estimate. Over half of Community Managed Mental Health Support Services were state-wide services. CONCLUSIONS: This study demonstrates the application of a needs-based planning model for mental health services. Findings revealed service priority areas across Tasmania and highlight considerations for needs-based planning.

Analysis of data items and gaps in Australia's national mental health services activity and capacity data collections for integrated regional service planning.

Background: Services data are an important source of information for policymakers and planners. In Australia, significant work has been undertaken to develop and implement collections of mental health services data. Given this level of investment, it is important that collected data are fit for purpose. Objective: This study aimed to: (1) identify existing national mandated and best endeavours collections of mental health services activity (e.g. occasions of service) and capacity (e.g. full-time equivalent staff) data in Australia; and (2) review the content of identified data collections to determine opportunities for data development. Method: A grey literature search was conducted to identify data collections. Where available, metadata and/or data were analysed. Results: Twenty data collections were identified. For services that received funding via multiple funding streams, data were often captured across several collections corresponding with each funder. There was significant variability in the content and format of collections. Unlike other service sectors, there is no national, mandated collection for psychosocial support services. Some collections have limited utility as they do not include key activity data; others do not include descriptive variables like service type. Workforce data are often not collected, and where data are collected, they are often not comprehensive. Conclusion: Findings are an important source of information for policymakers and planners who use services data to inform priorities. Implications: This study provides recommendations for data development, including mandating standardised reporting for psychosocial supports, filling workforce data gaps, streamlining data collections and including key missing data items in some collections.

Mental health systems modelling for evidence-informed service reform in Australia.

Australia's Fifth National Mental Health Plan required governments to report, not only on the progress of changes to mental health service delivery, but to also plan for services that should be provided. Future population demand for treatment and care is challenging to predict and one solution involves modelling the uncertain demands on the system. Modelling can help decision-makers understand likely future changes in mental health service demand and more intelligently choose appropriate responses. It can also support greater scrutiny, accountability and transparency of these processes. Australia has an emerging national capacity for systems modelling in mental health which can enhance the next phase of mental health reform. This paper introduces concepts useful for understanding mental health modelling and identifies where modelling approaches can support health service planners to make evidence-informed decisions regarding planning and investment for the Australian population.

Grading disorder severity and averted burden by access to treatment within the GBD framework: a case study with anxiety disorders.

BACKGROUND: The Global Burden of Disease Study (GBD) estimates burden by cause with major relevance for resource allocators globally. Non-fatal burden estimates are influenced by disorder severity. However, for many disorders, global severity is sourced from a single high-income country survey. We aimed to estimate severity distributions that vary by Healthcare Access Quality Index (HAQI) using anxiety disorders as a case study and present the usefulness of this method in simulating averted and avoidable burden globally. METHODS: In this case study, we estimated treatment use among respondents with anxiety disorder in the 1997 Australian National Survey of Mental Health and Wellbeing (NSMHWB), the source used to estimate severity of anxiety disorders in GBD. Treatment effects were sourced from the Cochrane Database of Systematic Reviews and pooled via network meta-analysis. Severity distribution was established via a meta-regression of their disability weights, derived from 12-item short form survey scores. We simulated the shift in severity across scenarios without access to treatment and with full access to optimal treatment (cognitive behavioural therapy and antidepressants). We interpolated this shift linearly along the HAQI, extrapolated country-specific severity from HAQI scores, and calculated averted and avoidable burden. FINDINGS: The database review sourced 56 reviews, of which eight were eligible for inclusion. These eight reviews reported on 156 randomised controlled trials, with 194 treatment effects. Respondents to the 1997 NSMHWB consisted of 5936 women (55·8%) and 4705 (44·2%) men aged 18 years or older (mean age and ethnicity data not available). The survey-weighted treatment effect size was -0·28 (95% uncertainty interval -0·45 to -0·12). The pooled treatment effect for full coverage optimal treatment was -1·07 (-1·47 to -0·64). The sequela-weighted disability weight among people with anxiety disorder in the NSMHWB was 0·141 (0·042 to 0·275). The estimated disability weight was 0·188 (0·070 to 0·341) after removing the benefits of treatment and 0·056 (0·013 to 0·140) after providing all people with anxiety disorder access to optimal treatment. Globally, 12·5% (4·6 to 21·5) of anxiety disorder burden was averted because of available treatment. However, 71·1% (46·2 to 87·6) of global anxiety disorder burden could be averted if all people with anxiety disorders had access to optimal treatment. INTERPRETATION: Because it is based on guidance from a single survey done in one high-income country, the burden of anxiety disorders in low-income and middle-income countries is probably underestimated by GBD. Despite the availability of effective treatments, low use of these treatments means that most burden is still avoidable. Most of the burden could be averted if all people with anxiety disorders had access to optimal treatment, highlighting the importance of public promotion and referral pathways of treatment for anxiety disorders. Location-specific severity distributions in GBD would greatly increase precision in burden estimates and highlight avertable burden to clinicians, public health practitioners, and policy makers. FUNDING: Queensland Health and Bill & Melinda Gates Foundation.

Educational level and the risk of mental disorders, substance use disorders and self-harm in different age-groups: A cohort study covering 1,6 million subjects in the Stockholm region.

OBJECTIVE: To investigate the associations between low education and risk of mental disorders, substance use disorders and self-harm in different age-groups. METHODS: All subjects in Stockholm born between 1931 and 1990 were linked to their own or their parent's highest education in 2000 and followed-up for these disorders in health care registers 2001-2016. Subjects were stratified into four age-groups: 10-18, 19-27, 28-50, and 51-70 years. Hazard Ratios with 95% Confidence Intervals (CIs) were estimated with Cox proportional hazard models. RESULTS: Low education increased the risk of substance use disorders and self-harm in all age-groups. Males aged 10-18 with low education had increased risks of ADHD and conduct disorders, and females a decreased risk of anorexia, bulimia and autism. Those aged 19-27 years had increased risks of anxiety and depression, and those aged 28-50 had increased risks of all mental disorders except anorexia and bulimia in males with Hazard Ratios ranging from 1.2 (95% CIs 1.0-1.3) for bipolar disorder to 5.4 (95% CIs 5.1-5.7) for drug use disorder. Females aged 51-70 years had increased risks of schizophrenia and autism. CONCLUSION: Low education is associated with risk of most mental disorders, substance use disorders and self-harm in all age-groups, but especially among those aged 28-50 years.

Diagnostic accuracy of the Composite International Diagnostic Interview (CIDI 3.0) in an urban Indigenous Australian sample.

OBJECTIVE: The Composite International Diagnostic Interview 3.0 is a standardised diagnostic interview commonly used in population-based mental health surveys, but has not been used in community-residing Indigenous Australians. This paper seeks to determine whether the Composite International Diagnostic Interview 3.0 can produce valid diagnostic information when compared with a diagnostic interview in an urban Indigenous Australian sample. METHOD: This research was conducted over 10 weeks with adult Indigenous clients of two participating Aboriginal Medical Services in South-East Queensland. Using a cross-sectional, repeated-measures design, participants were administered the Composite International Diagnostic Interview 3.0 by an Indigenous interviewer and within 2 weeks attended a second appointment with an Indigenous clinical psychologist, who produced a diagnostic summary. The Composite International Diagnostic Interview 3.0 diagnoses were compared with the diagnostic summaries and clinical concordance between the two measures was calculated. RESULTS: The diagnostic accuracy of the Composite International Diagnostic Interview 3.0 differed by module. The Post-traumatic Stress Disorder and Major Depression modules had good utility in diagnosing post-traumatic stress disorder and major depressive episodes, respectively; however, the Mania module that provides diagnoses of bipolar disorder was found to be unsuitable for this population. Although there were no identified contraindications for the use of the Generalised Anxiety and Alcohol Use Disorder modules, further research on the diagnostic accuracy of these modules is warranted. CONCLUSIONS: The Composite International Diagnostic Interview 3.0 can accurately diagnose some common mental disorders in an Indigenous Australian population, but was found to be unsuitable for others. Given these findings, care should be taken when using the Composite International Diagnostic Interview 3.0 in epidemiological prevalence studies with Indigenous Australian populations.

The association between mental disorders and subsequent years of working life: a Danish population-based cohort study.

BACKGROUND: Mental disorders can affect workforce participation via a range of mechanisms. In this study, we aimed to estimate the association between different types of mental disorders and working years lost, defined as the number of years not actively working or enrolled in an educational programme. METHODS: In this population-based cohort study, we included all people aged 18-65 years (mean 38·0 [SD 13·9]) in the Danish Civil Registration System from Jan 1, 1995 to Dec 31, 2016. Information on mental disorders was obtained from the Danish Psychiatric Central Research Register and information on labour market characteristics was obtained from administrative registers. Follow-up started at age 18 years, immigration to Denmark, or on Jan 1, 1995, whichever came later; and it ended at age 65 years, death, emigration from Denmark, disability pension, voluntary early retirement, or Dec 31, 2016 (whichever came earlier). As the main outcome, we estimated working years lost for those diagnosed with any mental disorder and 24 types of mental disorders, as well as for the general population of same age and sex. We decomposed total working years lost into periods of unemployment or sick leave, disability pension, voluntary early retirement, or death. Data on ethnicity were not available through administrative registers. FINDINGS: A total of 5 163 321 individuals, 2 642 383 men and 2 520 938 women, were followed up for 65·4 million person-years. Overall, 488 775 (9·47%) individuals were diagnosed with a mental disorder. On average, individuals with mental disorders lost an additional 10·52 (95% CI 10·48-10·57) years of working life compared with the general Danish population. Receiving a disability pension (7·54 [7·49-7·59] years) and longer periods of unemployment (2·24 [2·21-2·27] years) accounted for most of this difference. INTERPRETATION: Our findings foreground the substantial impact of mental disorders on workforce participation. There is a need to invest in programmes that reduce the burden of working years lost and assist people with mental disorders in returning to the workforce. FUNDING: Lundbeck Foundation and Danish National Research Foundation.

Measuring the Prevalence of Mental Disorders in Adolescents in Kenya, Indonesia, and Vietnam: Study Protocol for the National Adolescent Mental Health Surveys.

PURPOSE: In low- and middle-income countries, there are limited data on mental disorders among adolescents. To address this gap, the National Adolescent Mental Health Surveys (NAMHS) will provide nationally representative prevalence data of mental disorders among adolescents in Kenya, Indonesia, and Vietnam. This paper details the NAMHS study protocol. METHODS: In each country, a multistage stratified cluster sampling design will be used. Participants will be eligible pairs of adolescents aged 10-17 years and their primary caregiver. Adolescents will be assessed for social phobia, generalized anxiety disorder, major depressive disorder, attention-deficit/hyperactivity disorder, conduct disorder, and post-traumatic stress disorder using the Diagnostic Interview Schedule for Children, version 5. Demographics, risk and protective factors, and service use information will also be collected. In the parallel clinical calibration study, diagnoses of major depressive disorder, social phobia, and generalized anxiety disorder made using the Diagnostic Interview Schedule for Children, version 5 will be calibrated against a diagnostic assessment by in-country clinicians in a separate sample. RESULTS: Data collection for the national survey and clinical calibration study will commence in 2021, with dissemination of findings and methodology due to occur in 2022. CONCLUSIONS: Accurately quantifying the prevalence of mental disorders in adolescents is essential for service planning. NAMHS will address this lack of prevalence data, both within the NAMHS countries and within their respective regions, while establishing a gold-standard methodology for data collection on adolescent mental health in low- and middle-income countries. More broadly, NAMHS will encourage capacity building within each country by establishing linkages between researcher, clinician, government, and other networks.

The contribution of general medical conditions to the non-fatal burden of mental disorders: register-based cohort study in Denmark.

BACKGROUND: General medical conditions (GMCs) often co-occur with mental and substance use disorders (MSDs). AIMS: To explore the contribution of GMCs to the burden of disease in people with MSDs, and investigate how this varied by age. METHOD: A population-based cohort of 6 988 507 persons living in Denmark during 2000-2015 followed for up to 16 years. Danish health registers were used to identify people with MSDs and GMCs. For each MSD, years lived with disability and health loss proportion (HeLP) were estimated for comorbid MSDs and GMCs, using a multiplicative model for disability weights. RESULTS: Those with any MSD lost the equivalent of 43% of healthy life (HeLP = 0.43, 95% CI 0.40-0.44) after including information on GMCs, which was an increase from 25% before including GMCs (HeLP = 0.25, 95% CI 0.23-0.27). Schizophrenia was associated with the highest burden of disease (HeLP = 0.77, 95% CI 0.68-0.85). However, within each disorder, the relative contribution of MSDs and GMCs varied. For example, in those diagnosed with schizophrenia, MSDs and GMCs accounted for 86% and 14% of the total health loss; in contrast, in those with anxiety disorders, the same proportions were 59% and 41%. In general, HeLP increased with age, and was mainly associated with increasing rates of pulmonary, musculoskeletal and circulatory diseases. CONCLUSIONS: In those with mental disorders, the relative contribution of comorbid GMCs to the non-fatal burden of disease increases with age. GMCs contribute substantially to the non-fatal burden of disease in those with MSDs.

The global gap in treatment coverage for major depressive disorder in 84 countries from 2000-2019: A systematic review and Bayesian meta-regression analysis.

BACKGROUND: The treatment coverage for major depressive disorder (MDD) is low in many parts of the world despite MDD being a major contributor to disability globally. Most existing reviews of MDD treatment coverage do not account for potential sources of study-level heterogeneity that contribute to variation in reported treatment rates. This study aims to provide a comprehensive review of the evidence and analytically quantify sources of heterogeneity to report updated estimates of MDD treatment coverage and gaps by location and treatment type between 2000 and 2019. METHODS AND FINDINGS: A systematic review of the literature was conducted to identify relevant studies that provided data on treatment rates for MDD between January 1, 2000, and November 26, 2021, from 2 online scholarly databases PubMed and Embase. Cohort and cross-sectional studies were included if treatment rates pertaining to the last 12 months or less were reported directly or if sufficient information was available to calculate this along with 95% uncertainty intervals (UIs). Studies were included if they made use of population-based surveys that were representative of communities, countries, or regions under study. Studies were included if they used established diagnostic criteria to diagnose cases of MDD. Sample and methodological characteristics were extracted from selected studies. Treatment rates were modeled using a Bayesian meta-regression approach and adjusted for select covariates that quantified heterogeneity in the data. These covariates included age, sex, treatment type, location, and choice of MDD assessment tool. A total of 149 studies were included for quantitative analysis. Treatment coverage for health service use ranged from 51% [95% UI 20%, 82%] in high-income locations to 20% [95% UI 1%, 53%] in low- and lower middle-income locations. Treatment coverage for mental health service use ranged from 33% [95% UI 8%, 66%] in high-income locations to 8% [95% UI <1%, 36%] in low- and lower middle-income countries. Minimally adequate treatment (MAT) rates ranged from 23% [95% UI 2%, 55%] in high-income countries to 3% [95% UI <1%, 25%]) in low- and lower middle-income countries. A primary methodological limitation was the lack of sufficient data from low- and lower middle-income countries, which precluded our ability to provide more detailed treatment rate estimates. CONCLUSIONS: In this study, we observed that the treatment coverage for MDD continues to be low in many parts of the world and in particular in low- and lower middle-income countries. There is a continued need for routine data collection that will help obtain more accurate estimates of treatment coverage globally.

The Productivity Commission inquiry into mental health.

The report of the Productivity Commission Inquiry into mental health was released in November 2020, estimating the economic cost of mental illness in Australia at over $200 billion a year. The report makes wide-ranging recommendations for improving the mental health of the population, reforming the mental health treatment system, and in the way mental health is managed in other sectors of society.

The cost of mental disorders in Denmark: a register-based study.

The aim of the study was to undertake a detailed analysis of healthcare cost, public transfer payments, and income loss associated with a broad range of mental disorders in Denmark. Based on all persons living in Denmark, we identified those with a hospital diagnosis of one of 18 types of mental disorders and 10 age- and sex-matched controls per case. For each mental disorder, the outcomes were nationwide totals, cost per case, and cost per capita, investigated by sex, age strata, and the number of years after diagnosis. We found a substantial annual income loss of 5 billion Euros and excess healthcare cost of 1 billion Euros for persons with any mental disorder. Each mental disorder was associated with an income loss, excess healthcare cost, and excess public transfer payments compared to matched controls. An interactive data visualisation site with summary data is available at https://nbepi.com/cost .

Register-based metrics of years lived with disability associated with mental and substance use disorders: a register-based cohort study in Denmark.

BACKGROUND: Mental disorders account for a substantial proportion of the years lived with disability (YLDs) globally. These estimates have generally been calculated top down based on summary statistics. The aim for this study was to calculate YLDs and a novel related measure, Health Loss Proportion (HeLP), for 18 mental and substance use disorders, based on person-level register data (bottom up). METHODS: A cohort of 6 989 627 Danish residents (5·9% had a diagnosis of a mental or substance use disorder registered in the Danish Psychiatric Central Research Register) was investigated. YLDs (the duration of disease multiplied by a disability weight) were calculated for the disorder of interest and for comorbid mental and substance use disorders. HeLPs were estimated as YLDs associated with an index disorder and comorbid mental and substance use disorders divided by person-years at risk in persons with the index disorder. All analyses were adjusted for mental and substance use comorbidity using a multiplicative model of disability weights. FINDINGS: Major depressive disorder was the most prevalent disorder, although schizophrenia was the leading cause of YLDs in both sexes combined (YLDs 273·3 [95 % CI 232·3-313·6] per 100 000 person-years). People diagnosed with schizophrenia lost the equivalent of 73% (63-83%) of healthy life per year due to mental and substance use disorders, the largest HeLP of all mental and substance use disorders. Comorbidity of mental and substance use disorders accounted for 69-83% of HeLPs in people with either cannabis use disorders, other drug use disorder and ADHD. By contrast, comorbidity explained 11-23% of the HeLPs in people with autism spectrum disorders, conduct disorder, and schizophrenia. INTERPRETATION: Substantial variation in disability was observed across age, sex, and disorders. The new HeLP metric provides novel details of the contribution of comorbidity to the disability associated with mental and substance use disorders. FUNDING: The Danish National Research Foundation, Queensland Government Department of Health, European Union's Horizon 2020, Lundbeck Foundation, Stanley Medical Research Institute. TRANSLATION: For the Danish translation of the abstract see Supplementary Materials section.

Estimating the risk of suicide associated with mental disorders: A systematic review and meta-regression analysis.

BACKGROUND: Mental disorders (MDs) are known risk factors for suicide. This systematic review updates the evidence base for this association and improves upon analytic approaches by incorporating study-level and methodological variables to account for measurement error in pooled suicide risk estimates. METHODS: A systematic review was conducted to review studies on MDs as risk factors for suicide. Relevant studies were searched using PubMed, Embase, PsychINFO, and existing reviews from 2010 to 19. Studies were eligible if they were longitudinal/case-control studies, representative of the general population, used diagnostic instruments, and quantified suicide risk. The outcome assessed was relative risks (RRs) for suicide due to MDs. A multi-level meta-regression approach was used to obtain pooled RRs adjusted for covariates and between-study effects. FINDINGS: We identified 20 eligible studies yielding 69 RRs. Disorder type, age, sex, use of psychological autopsy, study design, and adjustment for confounders were tested as predictors of pooled suicide risk. Overall, all disorders were significant predictors of suicide with predicted adjusted RRs ranging from 4·11 [2·09, 8·09] for dysthymia to 7·64 [4·3, 13·58] for major depressive disorder. INTERPRETATION: Our results indicate that MDs are important risk factors for suicide. This systematic review provides pooled RRs that have been adjusted for methodological sources of bias. Findings from our paper may inform suicide prevention strategies as part of national health agendas.