Postural sway and falls in Parkinson's disease: a regression approach

A population-based study was designed to evaluate the clinical associates of postural sway and to identify the risk factors for falls in Parkinson's disease (PD). From a total population of 205,000 inhabitants, 215 PD patients were identified of which 120 home-dwelling cases were finally included in the study. Medical data were collected and patients were clinically examined and tested for static balance using an inclinometric device. Recent falls occurred in 40 (33%) of the subjects and 27 (23%) subjects were recurrent fallers. The fallers had a significantly larger sway area (P = 0.021) and a larger maximum deflection in anterior-posterior (P = 0.016) and lateral directions (P = 0.006) than the nonfallers. A significant correlation was found between the sway measures and the UPDRS total score, motor subcore and UPDRS "bradykinesia" item. A higher UPDRS total score (OR: 1.04, 95% CI: 1.01-1.07) and an increased sway area (OR: 1.25, 95% CI: 1.02-1.54) were independent risk factors for recent falling in PD. In addition, the duration and severity of PD, antiparkinsonian medication, recent falling and the use of a walking aid were associated with increased sway measures. The results can be used to identify PD patients who are at a risk of falling. Both antiparkinsonian medication and nonmedical treatment should be optimized to reduce falls in PD.