Predictors and consequences of injurious falls among older adults : a holistic approach

Sammanfattning: The field of research on falls among older adults is well studied. Despite this, there are some knowledge gaps that need to be addressed: 1) research studying injurious falls, as opposed to any falls; 2) knowledge on sex differences, and specific risk profiles for injurious falls; 3) development of an effective screening tool for community-dwelling older adults, that can detect people at risk of first-time falls, who may be targeted by preventive interventions; and 4) what factors influence the risk of losing independence, in a long-term perspective, after an injurious fall. The purpose of this thesis is to fill these gaps through the following aims: to detect risk profiles of injurious falls among older adults, to enable early detection of those at risk, and to examine the long-term consequences of fall injuries on everyday function. To reach these aims we used data from the ongoing population-based, Swedish National study on Aging and Care in Kungsholmen (SNAC-K). These are our main findings: The results from Study I indicate that risk factors for injurious falls tend to cluster within individuals, forming specific risk profiles, rather than appearing one by one. It is possible to predict elevated fall risk up to 10 years in advance and it also seems possible to distinguish groups of people at different levels of risk. In Study II we concluded that women and men share risk factors in many cases, but the levels of significance vary between the sexes. A few risk factors indeed seem to be sex specific. We also concluded that short-term (0–3 years) and longterm (4–10 years) risk factors differ, distinguishing specific acute and long-term risk profiles. We developed a screening tool for first time injurious falls in Study III, consisting of: age, cohabitation status, IADL dependency, and a balance test. Scores on the screening tool were weighted according to sex-stratified coefficients. The predictive value (measured with Harrell’s C statistics) of the scores were 0.75 and 0.77, for women and men, respectively. To be able to predict first time fallers up to 5 years in advance opens up for the possibility of primary prevention alongside with secondary prevention for recurrent fallers. With Study IV, we showed that sociodemographic and health related factors (living alone, physical inactivity, and self-rated poor health) measured before an injurious fall modified disability trajectories up to 12 years after baseline. These results enable identification of extra vulnerable fallers, who might need extra rehabilitation and attention after an injurious fall, with the goal to maintain independence. In conclusion, the results of this thesis suggest that 1) it might be more appropriate to study fall risk profiles, rather than risk factors in isolation; 2) risk factors for injurious falls may differ by sex and length of follow-up; 3) we created a screening tool for first time injurious falls, that is easy and quick to administer and has the possibility to predict falls up to 5 years in advance; and 4) sociodemographic and health related factors may help to identify extra vulnerable fallers, who might need extra rehabilitation and attention after an injurious fall, to maintain independence.