Premium
P4‐365: Targeting Dementia Screening in Primary Care Sites
Author(s) -
Perkins Anthony J.,
Fowler Nicole R.,
Harrawood Amanda M.,
Boustani Malaz A.
Publication year - 2016
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2016.07.110
Subject(s) - medicine , dementia , randomized controlled trial , logistic regression , depression (economics) , anxiety , psychiatry , disease , economics , macroeconomics
Background: Vascular burden, subclinical cardiovascular disease and depression are prevalent in older adults. Mechanisms underlying associations among vascular burden, subclinical cardiovascular disease, depression, and subsequent cognitive decline are unclear. We hypothesized that depressive symptoms will partially mediate the associations of vascular burden and subclinical cardiovascular disease with cognitive decline and that the percent of the total effect of these associations mediated by depressive symptoms will increase as we increase the time lags from two years to four years for each exposure-mediator and mediator-outcome relationship. Methods: We used N1⁄42,927 participants from Cardiovascular Health Study between 65 and 74 years of age without a diagnosis of clinical cardiovascular disease at baseline who were followed longitudinally for up to 18 years. Depressive symptoms, measured bymodified CES-D, were ascertained two years and four years after study baseline and cognitive change, assessed by the Digit Symbol Substitution Test, was ascertained four years and six years after baseline for each available assessment up to 18 years post-baseline. A latent growth curve model was used to model the intercept and slope of cognitive function. Covariates for adjustment were age, sex, race, marital status, income level, and educational level. Causal mediation analyses using structural equations modeling were conducted to examine the indirect effect of the predictors on level and slope of cognitive function through depressive symptoms. The product test was used to determine significant mediation. Results: Overall, 12% (ab1⁄4-0.02,p1⁄40.30) of the total adjusted effect of vascular burden measured at baseline on cognitive change was mediated through depressive symptoms two years after baseline. The percentage fell to 2% (ab1⁄4-0.003,p1⁄40.37) for depressive symptoms assessed four years after baseline. The total adjusted effect of subclinical cardiovascular disease on cognitive change mediated through depressive symptoms two years after baseline was 17% (ab1⁄4-0.01,p1⁄40.41) and after four years was 13% (ab1⁄40.002,p1⁄40.69). Conclusions:Depressive symptoms accounted for a small, clinically insignificant portion of the total association between predictors and cognitive decline. These findings do not support the hypothesis that mood disorders underlie the association between vascular burden and cardiovascular disease and cognitive decline.