Progression of mild cognitive impairment to dementia due to AD in clinical settings

Objectives: To describe the positive predictive value of mild cognitive impairment (MCI) and the factors associated with progression in routine practice. Methods: A retrospective cohort study was conducted from the French National Alzheimer Database. Among 446,439 patients cared for in the participating centers between January 2009 and January 2014, 45,386 (10.2%) were classified as having MCI and 23,676 had at least one follow-up visit. Annual progression rate was used to describe the progression of patients with MCI to dementia due to Alzheimer disease. Hazard ratios of dementia due to Alzheimer disease were estimated using Cox regression model. Results: Annual progression rate (95% confidence interval) was 13.7% person-years (py) (13.5%-13.9%) with higher rate for amnestic MCI (aMCI) (18.2% py [17.9%-18.5%]) than for nonamnestic MCI (naMCI) (9.5% py [9.3%-9.6%]). Separate regression models were performed for each MCI subtype. Higher education, older age, female sex, and lower Mini-Mental State Examination score were associated with an increased risk of progression for both subtypes. Use of anxiolytics (adjusted hazard ratio [95% confidence interval]: 0.77 [0.66-0.91]) was a protective factor for aMCI whereas antidepressant drugs (1.16 [1.04-1.29]) were associated with an increased risk. For naMCI, prescriptions of antidepressants (0.85 [0.74-0.98]) and anti-psychotics (0.55 [0.32-0.93]) were protective for progression. Conclusions: Under circumstances emulating routine clinical practice, the positive predictive value of an MCI diagnosis is in line with previous clinical studies and the external validity of the concept is strengthened. Distinguishing between aMCI and naMCI is particularly relevant.