Use of self-report to predict ability to walk 400 meters in mobility-limited older adults.

OBJECTIVES: To determine whether the ability to walk 400 m could be predicted from self-reported walking habits and abilities in older adults and to develop an accurate self-report measure appropriate for observational trials of mobility when functional measures are impractical to collect.
DESIGN: Cross-sectional.
SETTING: University-based human physiology laboratory. PARTICIPANTS: One hundred fifty community-dwelling older men and women (mean age+/-standard error= 79.8+/-0.3). MEASUREMENTS: An 18-item questionnaire assessing walking habits and ability was administered to each participant, followed by a 400-m walk test. Ninety-eight (65%) volunteers were able to complete the 400-m walk; 52 (35%) were unable. Logistic regression was performed using response items from a questionnaire as predictors and 400-m walk as the outcome.
RESULTS: Three questions (Do you think you could walk one-quarter of a mile now without sitting down to rest. Because of a health or physical problem, do you have difficulty walking 1 mile? Could you walk up and down every aisle of a grocery store without sitting down to rest or leaning on a cart?) were predictive of 400-m walking ability and were included in the model. If participants answered all three questions compatible with the inability to walk 400 m, there was a 91% probability that they were unable to walk 400 m, with a sensitivity of 46% and a specificity of 97%.
CONCLUSION: A three-item self-report developed in the study was able to accurately predict mobility disability. The utility of this instrument may be in evaluating self-reported mobility in large observational trials on mobility when functional mobility tasks are impractical to collect.