Coronary Artery Calcium and Cognitive Decline in the Diabetes Prevention Program Outcomes Study

Abstract

Background Our aim was to investigate the association of coronary artery calcium (CAC) with cognitive function in adults with impaired glucose tolerance or type 2 diabetes. Methods and Results The Diabetes Prevention Program was a randomized controlled trial comparing an intensive lifestyle intervention, metformin, or placebo for prevention of type 2 diabetes among patients with prediabetes. After 3 years, intensive lifestyle intervention and placebo were stopped, the metformin arm was unmasked, and participants continued in the DPPOS (Diabetes Prevention Program Outcomes Study). Approximately 14 years after randomization (Y14), CAC (Agatston score) was assessed with computed tomography, and cognitive performance was assessed with the Spanish English Verbal Learning Test (SEVLT) and Digit Symbol Substitution Test. SEVLT and Digit Symbol Substitution Test were reassessed 5 years later (Y19) along with the Modified Mini-Mental State Exam. We examined cross-sectional and longitudinal associations between CAC and cognition among 1931 participants using linear and logistic regression. In unadjusted analyses, compared with no calcification, CAC score >300 was associated with decreased performance on all cognitive tests at Y14 in both sexes. Additionally, CAC >300 was associated with a greater 5-year decline in SEVLT Immediate Recall in both sexes and SEVLT Delayed Recall in women. After adjustment for demographic, genetic, metabolic, vascular, and behavioral covariates, CAC score >300 remained associated with greater decline in only SEVLT Delayed Recall in women. Conclusions In women with prediabetes or diabetes, CAC >300, compared with no calcification, was independently associated with greater decline in verbal memory. Registration information clinicaltrials.gov. Identifier: NCT00038727.

Keywords: cognitive impairment; coronary artery calcium; diabetes; prediabetes; subclinical atherosclerosis.

Figure 1. Cognitive function and 5‐year decline by CAC severity, stratified by sex.

3MS indicates Modified Mini‐Mental State; CAC, coronary artery calcium; DSST, digit symbol substitution test; SEVLT‐DR, Spanish English Verbal Learning Test, Delayed Recall; and SEVLT‐IR, Spanish English Verbal Learning Test, Immediate Recall. 3MS was administered at Year 19 only. *P<0.05.

Figure 2. Adjusted mean change in cognitive function according to CAC level groups, stratified by sex, age, and diabetes status.

CAC indicates coronary artery calcium; DM, diabetes mellitus; DSST, Digit Symbol Substitution Test; SEVLT‐DR, Spanish English Verbal Learning Test, Delayed Recall; and SEVLT‐IR, Spanish English Verbal Learning Test, Immediate Recall. Shown are least‐square means of cognitive test scores, adjusted for age, race and ethnicity, Diabetes Prevention Program randomization arm, APOE‐ε4 status, education level, depressive symptoms, and interaction between CAC category, and sex, age, and diabetes status. The test for interaction that evaluated whether the association between CAC and cognitive function differed by sex, age, or diabetes status showed significant differences by age and diabetes for women.