
OBJECTIVES Late-life difficult-to-treat depression (LL-DTD) is associated with dementia, but the factors associated with comorbid dementia in this population remain unknown. We aimed to identify sociodemographic and clinical factors associated with comorbid dementia in LL-DTD. DESIGN & SETTING Cross-sectional study using electronic mental health records from a south London mental healthcare provider. PARTICIPANTS Patients aged ≥ 60 years with a first recorded depressive episode between 2008 and 2023, meeting criteria for LL-DTD (n = 3886). MEASUREMENTS LL-DTD was defined as ≥ 2 antidepressants prescribed or a natural language processing algorithm indicating treatment resistance. Dementia was ascertained using ICD-10 codes. Sociodemographic characteristics, depressive symptoms, and physical health conditions were ascertained from the electronic health record. Multivariable logistic regression was used. A sensitivity analysis restricted dementia to diagnoses recorded more than three months after the index episode. RESULTS Of 3886 patients with LL-DTD, 890 (22.9%) had dementia. Older age, female sex, and Black and Asian ethnicity were associated with higher odds of dementia (aOR 1.07, 1.24, 1.83 and 1.73, respectively). Somatic multimorbidity, cardiovascular and cerebrovascular conditions were also associated with higher odds. Anxiety (aOR 1.66) and cognitive problems (aOR 3.98) showed positive associations, while severe or psychotic depression and several ICD-10/11 depressive symptoms showed negative associations. When restricted to dementia recorded after the index episode, the association with cognitive problems halved (aOR 1.93) and those with apathy and ADL problems were lost. CONCLUSIONS In our analysis, somatic multimorbidity, cardiovascular conditions and anxiety were associated with comorbid dementia in LL-DTD, supporting comprehensive physical and neuropsychiatric assessment.
BACKGROUND Mental disorders are a leading cause of disability globally, yet data on their burden among older adults is limited. This study aims to quantify the global burden of mental disorders in adults aged ≥ 65 years from 1990 to 2021. METHODS Age-standardized rates (ASRs) of prevalence, incidence, and disability-adjusted life years (DALYs) were calculated per 100 000 population from the Global Burden of Disease Study 2021. Average annual percentage change (AAPC) was used to evaluate temporal trends. Inequality was assessed via slope and concentration indices and projections to 2035 were generated using Bayesian age-period-cohort modeling. Frontier and decomposition analyses were used to identify drivers of burden and global performance gaps. RESULTS In 2021, the ASRs were 14,425.87 (95% uncertainty interval [UI], 12,865.46-16,226.30) for prevalence, 6798.78 (95% UI, 5397.88-8542.31) for incidence, and 1998.27 (95% UI, 1470.34-2580.78) for DALYs. From 1990-2021, the age-standardized DALY rate increased by an average of 0.107% per year (95% confidence interval, 0.070-0.144), while population growth accounted for most of the increase in total DALYs. Anxiety and depressive disorders were the dominant contributors. Inequality analysis showed that DALYs remained disproportionately concentrated in low sociodemographic index countries, with point estimates suggesting widening absolute and relative inequalities. Frontier analysis identified countries such as Uganda and Gambia as having substantial performance gaps. By 2035, the global DALY rate is projected to rise to 2132.10 per 100,000. CONCLUSIONS The increasing burden and persistent inequalities highlight the need to plan mental health services for growing older populations. Projections suggest further increases under the model assumptions.