All-cause and cause-specific mortality in people with mental disorders: a population-based study on risk evaluation, effect modifiers and excess life-years lost in Hong Kong

Abstract

<p>Mental-disorders (MDs) are associated with premature mortality. Previous studies had limitations including confounding by physical-morbidities and lack of cause-specific mortality evaluation. We aimed to quantify mortality risk and life-expectancy gap across MDs in an Asian population. This retrospective population-based study investigated people with MDs (ascertained by ICD10-codes) in 2006–2021, utilizing comprehensive health-record-database of public-healthcare services in Hong-Kong. Individuals without MDs attending primary-care-clinics during study period served as comparisons. We estimated relative all-cause and cause-specific mortality risk using Cox-proportional hazards-regression models, and calculated excess life-years-lost (LYLs). Effect modification (sex, age, Charlson-comorbidity-index, socioeconomic-status (SES)) on relative risks was evaluated with subgroup-analyses. This analysis included 332,298 people with and 902,927 people without MDs. Any MD was associated with increased all-cause mortality (hazard-ratio (HR): 2.04 [95%Confidence-interval (CI)=2.02–2.07]) and excess LYLs (13.81 years [95%CI=13.60–13.98]). Eating-disorders exhibited the highest all-cause mortality (HR=9.43 [95%CI=6.83–13.02]), followed by developmental-disorders (HR=5.55 [95%CI=4.45–6.92]), personality-disorders (HR=4.50 [95%CI=4.06–4.98]) and substance-use disorders (HR=3.83 [95%CI=3.70–3.96]), with pronounced excess LYLs (14.18–17.41 years). Psychiatric-multimorbidity further increased excess mortality. Suicide was associated with the highest mortality risk (HR=8.69 [95%CI=7.97–9.45]). Natural causes accounted for most deaths (85%; HR range=1.50–2.76), while external causes explained 5% of all deaths (suicide: HR=8.69 [95%CI=7.97–9.45]). Men, younger age and lower SES were associated with increased all-cause and natural-cause mortality, while women and higher SES were linked to elevated external-cause mortality. This study highlighted transdiagnostic nature of premature mortality associated with MDs. Implementation of comprehensive multilevel interventions is warranted to narrow this mortality gap.<br></p>