Differential effects of APOE genotype and vascular risk factors on MCI and dementia

Abstract

Risk factors for incident mild cognitive impairment (MCI) are less well understood than those for dementia (DEM). The Mayo Clinic Study of Aging is a prospective, longitudinal study of elders aged 70-89 yrs. The present analysis concerns 1969 participants (1640 cognitively normal (CN); 329 with MCI) who were diagnosed by a consensus diagnostic process as non-demented at their initial evaluation. Participants were re-evaluated twice at 15 month intervals. The diagnostic process included a physician examination, a neuropsychological battery, and an informant interview. Standard published diagnostic criteria for CN, MCI and DEM were used. At the initial examination, a medical history including a history of stroke was obtained from the subjects. Vascular risk factors were assessed. APOE genotype was determined. Cox regression analyses were used to determine associations between predictors and incident MCI or DEM. Among initially CN subjects over 30 months (88.1% follow-up rate: mean age 79.5; 50.1% women; 43.3% with < 12 yrs education; 6.7% history of stroke; 24.7% APOE e4 carriers), there were 208 incident MCI cases (134 amnestic and 55 non-amnestic MCI). Age>80 yrs, male sex and <12 yrs of education were risk factors for incident MCI. Controlling for those demographic variables, carriage of APOE e4 allele (hazard ratio (HR) =1.47, 95% CI=1.10-1.97), history of stroke (HR =2.17, 95% CI=1.45-3.23) and diabetes (HR=1.46, 95% CI=1.06-2.01) were associated with incident MCI. Hypertension was not a risk factor for incident MCI. Among initially MCI subjects (85.1% follow-up rate), 71 (25.4%) became demented. Neither age, sex, education, APOE e4 carriage, history of stroke, diabetes nor hypertension were significant predictors of incident dementia in persons with MCI. The transition from CN to MCI, but not from MCI to DEM, was influenced by demographic features, APOE e4 genotype, diabetes and history of stroke. Once cognitive impairment is present, the predictors we studied had no impact on subsequent progression. Stroke history, even of events remote in time, is an important risk factor for cognitive impairment. Persistence of associations of incident MCI with diabetes and APOE genotype in this elderly cohort are remarkable.