The ankle brachial index and change in lower extremity functioning over time: the Women's Health and Aging Study

To define the association between baseline ankle brachial index (ABI) level and subsequent onset of severe disability.

Prospective cohort study.

Baltimore community.

Eight hundred forty-seven disabled women aged 65 and older participating in the Women's Health and Aging Study.

At baseline, participants underwent measurement of ABI and lower extremity functioning. Measures of lower extremity functioning included patient's report of their ability to walk one-quarter of a mile, number of city blocks walked last week, number of stair flights climbed last week, and performance-based measures including walking speed over 4 meters, five repeated chair stands, and a summary performance score. Functioning was remeasured every 6 months for 3 years. Definitions of severe disability were developed a priori, and participants who met these definitions at baseline were excluded from subsequent analyses.

Participants with an ABI of less than 0.60 at baseline had significantly higher cumulative probabilities of developing severe disability than participants with a baseline ABI of 0.90 to 1.50 for walking-specific outcomes (ability to walk a quarter of a mile, number of city blocks walked last week, and walking velocity) but not for the remaining functional outcomes. In age-adjusted Cox proportional hazards analyses, hazard ratios for participants with a baseline ABI of less than 0.60 were 1.63 for becoming unable to walk a quarter of a mile (P = .044), 2.00 for developing severe disability in the number of blocks walked last week (P = .004), and 1.61 for developing severe disability in walking speed (P = .041), compared with participants with a baseline ABI of 0.90 to 1.50. Adjusting for age, race, baseline performance, and comorbidities, an ABI of less than 0.60 remained associated with becoming severely disabled in the number of blocks walked last week (hazard ratio = 1.97, P = .009) and nearly significantly associated with becoming unable to walk a quarter of a mile (hazard ratio = 1.54, P = .09). In fully adjusted random effects models, a baseline ABI of less than 0.60 was associated with significantly greater decline in walking speed per year (P = .019) and nearly significantly greater decline in number of blocks walked last week per year (P = .053) compared with a baseline ABI of 0.90 to 1.50.

In community-dwelling disabled older women, a low ABI is associated with a greater incidence of severe disability in walking-specific but not other lower extremity functional outcomes, compared with persons with a normal ABI over 3 years.