DEMENTIA IN ELDERLY: PREDICTING AND PREVENTING RISK
MetadataShow full item record
Objectives: Specific aims of the study are: (1) To develop and validate a late-life dementia risk index using EMR database to predict the risk of developing dementia in elderly population. (2) To examine comparative effectiveness of antidepressants classes on risk of dementia in elderly with depression. Methods: The retrospective cohort study was conducted using General Electric (GE) Centricity Electronic Medical Records (EMR) database (2000-2011) to address research objectives. The Multivariate logistic regression using stepwise procedure was used to develop late-life dementia risk index. The risk of dementia among users of different antidepressant classes was modeled using Cox hazard regression model. Results: The late-life dementia risk model included older age (65-70 years, 71-75 years, and > 75 years), gender (female), race (black, hispanic and other racial group), depression, brain injury, cerebrovascular disorder, hyperlipidemia, hemoglobin level, creatinine level, albumin, and BMI. The final index had a maximum possible score of 25 points and a c-index of 0.725 (0.722-0.728). The results were identical using the 10-fold cross-validation approach (c statistic, 0.740; 0.736-0.744). The parsimonious model with only demographics and medical conditions included age, gender, race, depression, brain injury, cerebrovascular disorder, hypertension, and hyperlipidemia. The c-statistics of the model was 0.726 (0.723-0.730). The multivariate survival analysis after controlling for potential confounders found no significant difference (SNRI (HR 1.12 (95% CI, 0.92-1.35), Modified Cyclics (HR 0.79 (0.57-1.10) and Tetracyclics (HR 1.15 (0.85-1.55) across different antidepressant classes with respect to risk of dementia. Conclusion: The late-life dementia risk index developed using readily available EMR database can be used in routine clinical and research practice. In this retrospective analysis, no significant difference was observed between different antidepressant treatments groups with respect to risk of dementia. Large-scale prospective clinical trials are needed to examine comparative effectiveness of antidepressant treatment for delaying onset of dementia in elderly with depression.