Factors associated with falls in older patients with diffuse polyneuropathy

OBJECTIVES: To identify clinical factors associated with falls by older persons with polyneuropathy (PN). DESIGN: A cross-sectional study of 82 subjects aged 50 to 85 with clinical and electrodiagnostic evidence of PN. SETTING: Electrodiagnostic and biomechanical research laboratories. PARTICIPANTS: Patients referred to the electrodiagnostic laboratory. MEASUREMENTS: History and physical examination, including semiquantitative methods of peripheral nerve function, and clinical balance testing. Falls were defined by retrospective self-report over a 2-year period. RESULTS: Forty (48.8%), 28 (34.1%), and 18 (22.0%) subjects reported a history of at least one fall, multiple falls, and injurious falls, respectively. Factors associated with single and multiple falls were similar, so only results for multiple and injurious falls are reported. Bivariate analysis showed that an increased body mass index (BMI) and more severe PN (as determined by the Michigan Diabetes Neuropathy Score) were associated with both fall categories. Men reporting falls also demonstrated a decreased unipedal stance time. Age, sex, nerve conduction study parameters, Romberg testing, medications, and comorbidities were not consistently associated with either fall category. Logistic regression demonstrated that multiple and injurious falls were associated with an increased BMI and more severe PN, controlling for age, sex, medications, and comorbidities (pseudo R-2 = 0.458 and 0.484, respectively). CONCLUSIONS: Although previous work has demonstrated that all older persons with PN are at increased risk for falls, patients with increased BMI and more severe PN are at particularly high risk and should be targeted for intervention.