Associations between within-day step accumulation pattern and clinical measures of physical function: a change-for-change analysis of longitudinal data in community-dwelling older adults.

Background: While daily step count and stepping pace are linked to various health benefits in older adults, less is known about how the pattern of step accumulation affects physical function. For example, the same step count could be accumulated through clusters of frequent, short bouts (e.g., during house cleaning) or fewer, longer bouts (e.g., walking to and from work). This study aimed to explore whether stepping patterns, and trends in these patterns, were associated with physical function in older adults.

Methods: We analysed accelerometer data from wrist-worn GENEActiv devices, from four time points over 24 months in n = 597 older adults (age ≥65 years, 68% female) participating in the REtirement in ACTion intervention. A step counting algorithm was used to create bouts of stepping (at least 10 steps > 20 steps/minute) before counting the steps in each bout and the average cadence. Total daily steps (20-175 steps/minute), slower-paced steps (20-62 steps/minute; below the median cadence), and faster-paced walking steps (63-175 steps/minute; above the median cadence) were then calculated. We used the frequency of stepping bouts, the time between them (mean and standard deviation) and their burstiness (short bursts of stepping bouts clustered together), to examine the daily patterns of step accumulation. Linear mixed-effects models were used to assess trends in stepping variables and their association with changes in objectively measured physical function (short physical performance battery: SPPB) over the two-year period.

Results: Total, slower-paced and faster-paced daily steps declined, along with the average number of stepping bouts. The time between stepping bouts increased. All components of burstiness, but not burstiness itself, were associated with changes in physical function, even when faster and slower steps (total steps) were in the same model (fewer stepping bouts = lower SPPB, greater SD = lower SPPB). Mean time between bouts was the strongest independent predictor, whereby a 10-minute increase in time between bouts was associated with a clinically important 0.46 decline in SPPB score (p < 0.001).

Conclusions: Preventing increases in the time between stepping bouts could help preserve physical function in older adults. Future intervention trials targeting how bouts of stepping are spread throughout the day, rather than just total steps, may provide a more effective approach to promoting healthy physical functioning in older age.