Peixuan Zheng1, Jim Pleuss2, Dusty Turner3, Scott W. Ducharme4, Elroy J. Aguiar1. 1The University of Alabama, Tuscaloosa, AL. 2United States Military Academy-West Point, West Point, NY. 3Center for Army Analysis, Fort Belvoir, VA. 4California State University Long Beach, Long Beach, CA.

Background: The US National Health and Nutrition Examination Survey (NHANES) Physical Activity Monitor (PAM) data for 2011-2014 were recently released. Therein, physical activity (PA) was reported as a novel PAM metric known as Monitor Independent Movement Summary (MIMS) units. Currently, limited information exists regarding the relationship between cognitive function and habitual PA levels as quantified using MIMS metrics. Purpose: To determine the association between habitual PA and cognitive function using a nationally representative dataset of US older adults aged 60 years and above. Methods: The analytical sample comprised n=2516 individuals (mean age: 69.1± 0.2 years; 54.7% females). Cognitive function was assessed using the Digit Symbol Substitution Test (DSST). NHANES PAM data were collected using a tri-axial accelerometer (ActiGraph GT3X+; 80Hz) worn on participants’ non-dominant wrist. Valid wear-time was considered ≥10h/day and ≥4 valid days. PA data were summarized as MIMS/day (the average of daily MIMS across all valid days), peak 30-minute MIMS (Peak-30MIMS; the average of the highest 30 non-consecutive MIMS mins/day, averaged across all valid days). Peak 30MIMS can be considered as a PA index of “best effort”, as it is shaped by both the intensity and persistence of ambulatory behavior within a day, as well as its consistency (regularity) across valid days. Multivariable linear regression (accounting for survey sample weights) was performed to determine the relationship between DSST score and each MIMS metric separately, while adjusting for age, gender, race, education, body mass index and hypertension. Results: The means [SEs] of DSST score, MIMS/day and Peak-30MIMS were 52.2 [0.6], 11308 [100.4] and 36.5 [0.2], respectively. Both MIMS/day and Peak-30MIMS were positively correlated with DSST score (Pearson’s r = 0.3 and 0.4, respectively). After adjusting for covariates, DSST score increased by 0.68 and 0.57 (β-coefficients) for each 1000-unit increase in MIMS/day, or each one-unit increase in Peak-30MIMS, respectively (p-values<0.001). Conclusions: Our findings suggest that higher habitual PA (both daily accumulated and peak effort) is associated with better cognitive test performance in this US older adult population. Such information informs our understanding of the dose-response relationship between PA and cognitive function based on accelerometer-determined PA metrics.

This document is currently not available here.