Longitudinal associations between physical activity and preferred walking speed in community-dwelling older adults: a 4-year follow-up study

Abstract Background The ability to engage in physical activity reflects preserved physical function, autonomy, and well-being in older adults, making physical activity a key outcome of healthy aging. Walking is a common form of physical activity in later life, so understanding how walking speed is associated with overall activity is particularly important. This study examined 4‑year changes in physical activity energy expenditure (PAEE) in community‑dwelling older adults and assessed preferred walking speed in relation to PAEE. Methods This study used data from AGNES, a 4-year follow-up study of initially 75-, 80-, and 85-year-old adults. PA was measured using thigh-worn accelerometers and ECG monitors, with PAEE estimated using a branched equation model integrating heart rate and acceleration data. Preferred walking speed was measured using a self-paced six-minute walk test. Associations between PAEE, walking speed, and their 4-year changes were examined using linear mixed models. Results Among 299 participants, PAEE declined by 16.9 kJ·kg − 1 ·day − 1 over 4 years. Preferred walking speed showed a positive association with PAEE, with each 1 km·h − 1 higher speed corresponding to 4.7 kJ·kg − 1 ·day − 1 higher PAEE. There was limited evidence for an association between changes in preferred walking speed and changes in PAEE over the 4‑year follow‑up period. Conclusions In community-dwelling older adults, PAEE declined substantially with age, more steeply than in total physical activity previously reported in similar cohorts. Walking speed was strongly associated with PAEE, but its change was not consistently associated with changes in PAEE.