Determinants of fractal motor activity regulation: insights from a large community-dwelling multi-ethnic population

Quantification of alterations in fractal motor activity regulation (FMAR) has the potential to improve early detection of Alzheimer's disease. To improve the clinical relevance of such marker, our analysis aims to investigate potential determinants such as age, sex, physical activity and race and ethnicity in a large community-dwelling multi-ethnic cohort of older Asian, Black, Hispanic and White individuals.

We analysed week-long actigraphy recordings from 1508 older adults (55 to 90 years) of the MESA Sleep dataset. Long-range correlations in activity levels, a marker of fractal regulation, were assessed via a scaling exponent extracted with a Detrended Fluctuation Analysis over two different time ranges. Bayesian linear models were used to compare scaling exponent values between ethnic groups and test their associations with various determinants, such as age, sex and physical activity levels.

We found opposite effects of age on FMAR at both short and long time scales, respectively ( β _Age = 0.014/ σ _Age , 95% C.I. [0.009, 0.019] and β _Age = -0.015/ σ _Age , 95% C.I. [-0.026, -0.005]), as well as differences between women and men at long time scales ( ${\beta }_{\text{Sex}}^{\text{Men}}=+0.028$ , 95% C.I. [0.012, 0.044]). We also confirmed that higher activity levels were linked to less degraded FMAR ( β _MDA = 0.007/ σ _MDA , 95% C.I. [0.003, 0.011] at short time scales and β _MDA = 0.032/ σ _MDA , 95% C.I. [0.024, 0.041] at long time scales). Finally, we observed that Black and Hispanic participants, as well as Asian participants, exhibiting lower values in scaling exponents at short and long time scales, respectively ( β _Black = -0.026, 95% C.I. [-0.035, -0.017], β _Hispanic = -0.018, 95% C.I. [-0.027, -0.009] and β _Asian = -0.065, 95% C.I. [-0.091, -0.039]), compared to White participants.

Using a large multi-ethnic cross-sectional dataset, our analysis provided support and clarified the interaction effect of age and sex on FMAR. It also confirmed, in humans, the influence of activity levels on FMAR observed in rodents and demonstrated for the first time that FMAR alterations are modulated by race and ethnicity.