Varying Associations Between Body Mass Index and Physical and Cognitive Function in Three Samples of Older Adults Living in Different Settings.

BACKGROUND

The study investigates variations in the associations between body mass index (BMI) and (a) physical and (b) cognitive function across three samples of older adults living in different settings, and moreover determines if the association between BMI and physical function is confounded by cognitive abilities.

METHODS

One hundred ninety-five patients of a geriatric day hospital, 322 persons receiving home care (HC), and 183 nursing home (NH) residents were examined regarding BMI, cognitive (Mini-Mental State Examination), and physical function (Barthel Index for activities of daily living). Differences in Mini-Mental State Examination and activities of daily living scores between BMI groups (<22, 22-<25, 25-<30, 30-<35, ≥35kg/m(2)) were tested by analysis of covariance considering relevant confounders.

RESULTS

Activities of daily living and Mini-Mental State Examination impairments increased from the geriatric day hospital over the HC to the NH sample, whereas prevalence rates of obesity and severe obesity (35%, 33%, 25%) decreased. In geriatric day hospital patients cognitive and physical function did not differ between BMI groups. In the HC and NH samples, cognitive abilities were highest in obese and severely obese subjects. Unadjusted mean activities of daily living scores differed between BMI groups in HC receivers (51.6±32.2, 61.8±26.1, 67.5±28.3, 72.0±23.4, 66.2±24.2, p = .002) and NH residents (35.6±28.6, 48.1±25.7, 39.9±28.7, 50.8±24.0, 57.1±28.2, p = .029). In both samples significance was lost after adjustment indicating cognitive function as dominant confounder.

CONCLUSIONS

In older adults the associations between BMI and physical and cognitive function were dependent on the health and care status corresponding to the setting. In the HC and the NH samples, cognitive status, as measured by the Mini-Mental State Examination, emerged as an important confounder within the association between BMI and physical function.