JMIR Publications

ABSTRACT

Background:

Knowledge about how older adults walk is crucial for effective prevention and treatment of various mobility issues, as well as treatment evaluation, but is to date largely limited to lab-based measurements. Although this provides relevant information about what older adults can do under standardized conditions, it does not give insight into how they actually walk in their daily life, a gap that needs to be addressed urgently.

Objective:

The objective of the current study is to describe how older adults walk in daily life in relation to age, sex, and level of physical function, using wearable sensor data from a large sample of older adults with a wide range of age and function from the HUNT4 Trondheim 70+ study.

Methods:

The current study is based on one-week accelerometer data (Axivity AX3) from 1289 older adults (mean age 77.41 (SD 6.06) years, range 70-105 years; 54.7% women). Physical function was assessed using the Short Physical Performance Battery (SPPB). To investigate the effect of age and SPPB score on gait metrics (daily number of steps, 95th percentile speed, mode speed, 95th percentile cadence, mode cadence, and maximum WB distance) for women and men, univariate gamma regressions with log link were performed on each outcome measure, with age and SPPB score in separate models. Sex differences were investigated through Mann-Whitney U tests.

Results:

Older adults showed large variation in how and how much they walked in daily life across age, sex, and physical function, particularly younger participants and those with better physical function. Most gait metrics decreased at an increasing rate with higher age, with men maintaining their levels up to higher ages than women. Poorer physical function led to an exponential or close-to-linear decrease in all gait metrics apart from habitual cadence, which remained stable up to high age. Women had a lower daily number of steps, gait speed, and maximum distance, yet higher cadence, than men (all Ps<.001). On average, 63% of all walking bouts (WBs) lasted <10 seconds, accumulating to a median of 99 minutes (IQR 66-128). For WBs lasting 10-30 seconds, the accumulated time was 105 minutes (IQR 65-154), WBs lasting 30-60 seconds accumulated 31 minutes (IQR 18-47), while WBs lasting >60 seconds accumulated 113 minutes (IQR 37-219).

Conclusions:

Daily life walking performance is affected more by functional ability than by age itself, except for the oldest ages, and differs significantly between sexes. Although most walking bouts are very short, total accumulated walking time is longer in bouts shorter than 30 seconds than in longer bouts. Future research can build upon our findings, considering both the impact of short WBs and relevant group and sex differences when implementing daily life mobility assessment in both clinical studies and patient follow-up.