Clinic walking speed and daily step cadence both linked to mortality in 4,067 UK adults

A study published on October 23 in JAMA Network Open compared two ways of measuring walking as predictors of death from any cause: a short walking test in a clinic and the step rate recorded in everyday life by a wrist-worn device, according to healthline.com healthline.comhealthline.comhealthline.comOpen source ↗. Both measures were linked to mortality risk and showed similar ability to predict it. The work was part of the Whitehall II accelerometer substudy in the United Kingdom.

Study design and measurements

The analysis included 4,067 participants with an average age of about 69 years; about one-third were female. Data were collected from January 2012 to December 2013, and deaths were identified from national records through August 2024.

For clinical walking speed, participants walked a marked 2.44-meter course at their usual pace three times, and the average speed in meters per second was used. Those who normally used a cane could use it. For free-living walking, participants wore a GENEActiv accelerometer on the wrist for 9 consecutive days. Software detected walking episodes and counted steps, from which the median cadence and the 95th-percentile cadence were calculated.

The risk models were adjusted step by step: first none, then sociodemographic, behavioral and finally health-related factors healthline.comhealthline.comhealthline.comOpen source ↗.

Results

During a median follow-up of nearly 12 years, 739 participants died. Average clinical walking speed was 1.2 meters per second, median daily cadence about 75 steps per minute, and 95th-percentile cadence about 112 steps per minute.

The correlation between clinic speed and daily cadence was weak, yet each was independently associated with lower mortality. After adjustment, each standard increase in clinical walking speed corresponded to a 23% lower risk of death. Higher median and 95th-percentile cadence were associated with 17% and 18% lower risk, respectively.

Each measure alone showed modest predictive ability. Adding either one to a model with known risk factors slightly improved prediction, to a similar degree. Combining them gave a small additional gain that was not significant after accounting for detailed health factors. Predictive power was similar in men and women, although the models performed slightly better for women. Sensitivity analyses, including exclusion of early deaths, gave consistent results healthline.comhealthline.comhealthline.comOpen source ↗.

Neha Sabharwal, a physical therapist who was not involved in the study, said that walking requires the heart, lungs, muscles and brain to work together, and that a slowdown in a person's walking can have causes beyond age.