Metabolic syndrome and 16-year cognitive decline in community-dwelling older adults

Abstract

Purpose: To determine whether metabolic syndrome is associated with accelerated cognitive decline in community-dwelling older adults.

Methods: A longitudinal study of 993 adults (mean 66.8 ± 8.7 years) from the Rancho Bernardo Study. Metabolic syndrome components, defined by 2001 NCEP-ATP III criteria, were measured in 1984-1987. Cognitive function was first assessed in 1988-1992. Cognitive assessments were repeated approximately every 4 years, for a maximum 16-year follow-up. Mixed-effects models examined longitudinal rate of cognitive decline by metabolic syndrome status, controlling for factors plausibly associated with cognitive function (diabetes, inflammation).

Results: Metabolic syndrome was more common in men than women (14% vs. 9%, p = .01). In women, metabolic syndrome was associated with greater executive function and long-term memory decline. These associations did not differ by inflammatory biomarker levels. Diabetes did not alter the association of metabolic syndrome with long-term recall but modified the association with executive function: metabolic syndrome was associated with accelerated executive function decline in diabetic women only. Metabolic syndrome was not related to rate of decline on any cognitive measure in men.

Conclusions: Metabolic syndrome was a risk factor for accelerated cognitive decline, but only in women. Prevention of metabolic syndrome may aid in maintenance of cognitive function with age.

Figure 1

Modeled cognitive function test score trajectories by metabolic syndrome status for men and women. Plots are based on all model coefficients using sex-specific mean values for covariates (age, educational level, depression score, and number of cognitive assessments). Note that the y axis is reversed for the Trails B because increasing Trials B scores indicate poorer performance. Women with metabolic syndrome showed significantly greater longitudinal decline in Trails B performance relative to women without metabolic syndrome (p=.01). Rate of decling on long term recall significantly differed between women with and without metabolic syndrome after additionally controlling for inflammatory biomarkers. Rate of change did not differ significantly as a function of metabolic syndrome status in men. MMSE = Mini-Mental State Exam; Trails B = Trail Making Test Part B; Recall = Buschke-Fuld long-term recall.