Life-Space Mobility in Canadians Adults

Abstract

BACKGROUND Life-space mobility refers to the extent of movement into the environment and community. The Life-Space Assessment (LSA) is a self-report measure that measures life-space mobility. Prior to its use in a Canadian community-dwelling population, the LSA should be validated, and reference values need to be generated. As such, this thesis has the following aims: 1) To establish sex-stratified reference values for the LSA; 2) To determine the construct (convergent and known-groups) validity of the LSA, and 3) To determine the predictive validity of the LSA in predicting health outcomes such as hospitalizations at 3 years among community-dwelling Canadian adults. METHODS The Canadian Longitudinal Study on Aging (CLSA) was used. For reference values, percentile regressions were used to estimate the age-specific percentiles (i.e., 5th to 95th) for LSA scores. For convergent and divergent validity, Spearman correlation coefficients were used to compare the LSA against other measures. For known-groups validity, a change of 5 or more points on the LSA was considered between known-groups (i.e., those with chronic disease and those with a falls history). For predictive validity, Receiver Operating Characteristic curves were used. Analyses were stratified by age and sex. RESULTS References values demonstrated lower LSA scores for females and older age groups. Convergent validity hypotheses (r >0.5) were not met; however divergent validity (r <0.3) and known-groups validity (5 or more points on the LSA) hypotheses were met. Lastly, predictive validity hypotheses (area under the curve >0.7) were not met.