Magnus Stenhagen, Henrik Ekström, Eva Nordell, Sölve Elmståhl Department of Health Sciences, Division of Geriatric Medicine, Lund University, Skåne University Hospital, Malmö, Sweden Objectives: To determine the relationship between long-term change in activities of daily living (ADL) and falls in the elderly and to identify characteristics of groups at risk for falls.Methods: This was a 6-year, prospective cohort study using data from the Good Aging in Skåne study in southern Sweden, involving

more »

... 1,540 elderly subjects, including the oldest-old (age, 60–93 years). The subjects were recruited from the general population. ADL was measured at a baseline and follow-up assessment, using Sonn and Åsberg's revised scale and the ADL staircase. Falls were recorded in a period of 6 months before the follow-up assessment. The association between falls and change in ADL was calculated using adjusted, multiple logistic regression analysis and presented in odds ratios (ORs).Results: Thirteen percent of the study population reported one or several falls in the measured period. Over the course of 6 years, one in four participants changed their ADL status, and parts of this category had an increased risk for falls compared with those who stayed independent in ADL or who had no change in the ADL staircase. Groups with different characteristics had a prominent risk for falls: those with a reduction of two to eight steps in the ADL staircase (OR, 4.05; 95% confidence interval [CI], 1.62–10.11) and those becoming independent from dependency in instrumental ADL (OR, 4.13; 95% CI, 1.89–9.00). The former group had advanced age with a greater burden of cognitive impairment, gait disability, arrhythmia, and fall risk medications. The latter group had a higher prevalence of ischemic heart disease and low walking speed. Conclusion: Both deterioration and improvement in ADL over the course of 6 years increased the risk for falls in a general elderly population. Interventional efforts may require different strategies, as groups with different [...]