Accelerometer-Derived Relative Amplitude and Midlife Cognition: A Cross-Sectional Study

Importance

Lower relative amplitude of the rest-activity pattern has been associated with cognitive decline and incident dementia in older adults, but its relationship with domain-specific cognitive performance during midlife is less well characterized.

Objective

To examine cross-sectional associations between accelerometer-derived relative amplitude and fluid and crystallized cognitive performance among midlife adults.

Design, Setting, and Participants

Cross-sectional analysis of 851 adults aged 35-64 years enrolled in the MidCog Study and recruited from 13 primary care clinics, including academic practices and federally qualified health centers. Participants completed a 14-day wrist accelerometer recording and the NIH Toolbox Cognition Battery.

Exposure

Accelerometer-derived relative amplitude, calculated from the contrast between activity during the most active consecutive 10-hour period and the least active consecutive 5-hour period.

Main Outcomes and Measures

Age-and education-adjusted NIH Toolbox T-scores (normative mean = 50, SD = 10) for the Fluid Cognition Composite; attention, executive function, episodic memory, working memory, and processing speed; the Crystallized Cognition Composite; receptive language and expressive language. Multivariable linear regression models sequentially adjusted for sociodemographic characteristics, other rest-activity rhythm metrics, established dementia risk factors, average activity, time in bed, and sleep efficiency.

Results

In the final adjusted model, each 1-SD lower relative amplitude was associated with a 2.27-point lower Fluid Cognition Composite T-score (95% CI, 1.15–3.40). Lower relative amplitude was also associated with lower attention by 2.74 points (95% CI, 1.49–4.00), executive function by 1.49 points (95% CI, 0.51–2.47), episodic memory by 2.03 points (95% CI, 0.81–3.25), and processing speed by 1.56 points (95% CI, 0.49–2.64). Relative amplitude was not associated with working memory or crystallized cognitive outcomes.

Conclusions and Relevance

Among midlife adults recruited from primary care practices, lower relative amplitude was associated with modestly poorer fluid cognitive performance after adjustment for established dementia risk factors and sleep characteristics. Relative amplitude may represent a scalable behavioral correlate of cognitive variability within the normative range. Longitudinal studies beginning in midlife are needed to establish temporality, predictive value, or modifiability.

Key Points