Predictors of hand function in older persons : a two-year longitudinal analysis

OBJECTIVE: To validate a hierarchical model of hand function in older persons, using longitudinal data. DESIGN: Longitudinal observational study (2-year data from an ongoing longitudinal study). SUBJECTS: 689 persons older than age 60, including Continuing Care Retirement Community (n = 230), homebound (n = 204), and ambulatory (n = 255) respondents. Mean age at baseline 76.6 (SD = 8.8). MEASUREMENT: Independent variables included sociodemographics, physician measures of upper joint impairment, self-reported comorbidity, arthritis pain, depression, and anxiety. The dependent variables included grip strength and a timed manual performance test. MAIN RESULTS: Using generalized estimated equations (GEE) to test our hierarchial model, we found that gender and upper extremity joint impairment were the strongest predictors of a longitudinal measure of grip strength. Grip strength, in turn, along with demographics, comorbidity, and a measure of psychological status, was significantly related to timed manual performance. CONCLUSIONS: The longitudinal analyses confirmed a previous cross-sectional finding that upper extremity joint impairment contributes significantly to reduced grip strength, which, in turn, contributes to reduced hand performance on a timed test.