Aim: To examine the role of disease-related factors and age-related ph
ysiological changes in affecting gait speed and stride length in the e
lderly. Subjects: 925 men and 890 women aged 70 years and above who we
re ambulant, recruited by random sampling stratified according to age
and sex, from all recipients of Old Age and Disability Allowance in Ho
ng Kong. Design: Gait was assessed by measuring the time taken and the
number of steps required to complete a 16-foot walk. Information on h
ealth and functional status, cognitive function, and depressive sympto
ms was collected, and anthropometric indices obtained. Factors affecti
ng walking speed and stride were examined in the overall population an
d also after excluding those with physical disability or diseases. Res
ults: Results were analyzed separately for men and women since mean wa
lking speed was slower in women, who also took a larger number of step
s. Age, coexisting disease, leg or back pain, poor vision, low level o
f physical activity, functional and cognitive impairment, high depress
ive symptom score, and anthropometric indices were all negatively asso
ciated with walking speed. Fallers also had slower speed. After exclud
ing those with diseases or physical impairment, multivariate analysis
showed that the only factors affecting speed were age in men, and age,
height, and level of physical activity in women. Age and height were
factors associated with stride length in men, and height only for wome
n. Conclusion: Both disease-related factors as well as age-related phy
siological changes contribute to the decline in walking speed and stri
de length.