Determinants of physical function in community dwelling old people

Abstract

Background. Poor physical function is associated with disability and mortality in old people. Objectives. The aim was to find determinants of physical function in old people. Design. Secondary, cross-sectional analysis. Setting. Community in the Reykjavik, Iceland. Participants. 236 old people (73.7 ± 5.7 years, 58.2% female). Measurements. Timed-up-and-go (TUG), six-minute-walk-for-distance (6MWD), anthropometrics, quadriceps strength, dietary intake, mini-mental-state-examination (MMSE), leisure-time physical activity (LTPA) and blood variables were assessed. Descriptive, bivariate and multivariate statistical analyses were used. Results. There were differences between men and women in energy intake, body composition and muscular strength, but physical function did not differ between men and women. In bivariate analysis, most of the assessed variables correlated with 6MWD and TUG. Stepwise lin-ear models showed that age, body composition, strength, medication, LTPA and MMSE were predictors of physical function but not hematological variables. The association between MMSE and function disappeared when corrected for strength/body weight. Results were similar for both 6MWD and TUG and the strongest predictors in the final models were age and quadriceps strength/body weight. Conclusions. In community dwelling old people, physical function decreases with age. However, it is of clinical relevance that there are modifiable determinants of physical function, in particular strength for a given body weight, LTPA and number of medications, which represent potential targets to maintain physical function in this age group. Our results also indicate that neither cognitive function, nor dietary intake nor blood chemical variables were independently associated with physical function.