One-week changes on the Independent Living Scales in older adults: Proof of concept for quantifying subtle functional decline.

Abstract

Objective: Identifying subtle change in daily functioning is important for diagnosing late life cognitive disorders. This proof-of-concept study sought to develop standardized regression-based (SRB) equations for the Independent Living Scales (ILS) and examine their application in older adults over one week to quantify subtle functional changes. Method: Two hundred sixteen older adults (86% cognitively unimpaired, 14% cognitively impaired) were administered the ILS, a multi-dimensional standardized measure of daily functioning, during a baseline visit and one week later. In the unimpaired participants, linear regression models predicted one-week ILS scores by baseline ILS scores, demographics, retest intervals, and premorbid intellect. SRB equations were applied to both groups and compared with independent t-tests and chi-square tests. Results: One-week retest scores were significantly (p < 0.001) predicted in the unimpaired individuals, with regression models accounting for 14-51% of the variance. As expected, SRB-derived z-scores differed significantly between groups, with the impaired group demonstrating more negative values for the ILS Full Scale (p = 0.03) and 2 of 5 subscales (p ≤ 0.002). Chi-square analyses revealed that smaller-than-expected retest performance was overrepresented in impaired participants (e.g. ILS Full Scale: impaired = 30%, unimpaired = 5%). Conclusion: Using SRB methodology, change scores on the ILS across one week were developed and showed expected group differences. These change scores, which correct for practice effects and expected variability across one week, have the potential to identify subtle functional decline in those at risk for dementia. Future research could expand this methodology to other functional scales across longer retest intervals.