Competing risk analysis of the association between dementia and major depressive disorder: a nationwide population-based study in Taiwan.
Chun-Jen HuangShih-Feng WengJhi-Joung WangHui-Min HsiehPublished in: Aging & mental health (2020)
Objectives: To investigate the association between major depressive disorder (MDD) and the competing risk of Alzheimer's dementia (AD) and non-Alzheimer's dementia (NAD) using a nationwide population-based health insurance database.Methods: From Taiwan's National Health Insurance Research Database, we extracted claims data of 13,067 patients with MDD and 52,268 non-MDD controls matched by index date, age, sex, geographical area, monthly income, and selected comorbidities through propensity score matching. Follow-up durations in person-years were calculated for each person until dementia diagnosis, death, or the end of 2013. Competing incident risks of AD and NAD between patients with MDD and non-MDD.Results: In this study, 335 (2.6%) of the 13,067 MDD patients and 313 (0.6%) of the 52,268 non-MDDs developed AD. During the follow-up period, 73 (0.59%) of the 13,067 MDD patients developed NAD and 80 (0.15%) of the 52,268 non-MDD developed NAD. The patients with MDD had 4.73 and 3.69 times higher risks of AD (adjusted subdistribution hazard ratio [SHR] 4.73; 95% confidence interval [CI] 4.05-5.52) and NAD (adjusted SHR 3.69; 95% CI 2.68-5.08), respectively, than the controls.Conclusions: The patients with MDD had significantly higher incidence rates of AD and NAD than the controls, in particular among aged 65 and above. Additional studies are required to clarify the underlying pathophysiology between the MDD-dementia association and investigate whether prompt intervention in MDD can reduce the risk of dementia.
Keyphrases
- major depressive disorder
- health insurance
- bipolar disorder
- mild cognitive impairment
- cognitive impairment
- end stage renal disease
- cognitive decline
- newly diagnosed
- ejection fraction
- randomized controlled trial
- chronic kidney disease
- healthcare
- type diabetes
- prognostic factors
- human health
- physical activity
- affordable care act
- emergency department
- risk factors
- climate change
- machine learning
- electronic health record
- patient reported outcomes
- deep learning
- drug induced