Falls in Parkinson's disease

Abstract

Background Parkinson's disease (PD) is a slowly progressive neurodegenerative disorder affecting 1% of the population over 60 years. Although motor abnormalities are the core feature of the disease, PD is today considered a multisystem brain disease. While early falls due to impaired balance is considered a red flag for atypical parkinsonism, the frequency of falls at different stages of PD remains uncertain due to methodological limitations in previous studies.

Objectives The overall aim of this thesis was to describe and achieve a better understanding of the epidemiological aspects of falls in PD across different stages of the disease. To accomplish this, the following objectives were outlined: • To determine the frequency of falls and demographic and clinical features of falls in 2 population-based cohorts with PD at different stages of disease (paper I). • To describe the development of falls in a population-based cohort of patients with established PD during 8 years follow-up, and explore risk factors in previous non-falling patients with PD (paper II). • To describe the development of falls in a population-based incident cohort of patients with newly diagnosed PD vs a matched control group during 7 years of follow-up, and determine concomitants and risk factors of falls in the PD cohort (paper III). • To objectively examine physical activity levels (time ambulatory, standing and sitting/lying) in a subgroup of PD patients with and without a fall history last 6 months, and identify potential mediators of an active lifestyle (paper IV).

Methods All included subjects were part of 2 population-based cohorts: (1) The Stavanger Parkinson project recruited patients with PD between September 1992 and May 1993, initially to determine the prevalence of PD in Rogaland county, Norway. The study was extended withperiodically examinations during 12 years. Data from baseline (n=232), 4-year (n=121) and 8-year (n=64) follow-ups are included in this thesis.

(2) The Norwegian ParkWest study is a multicentre longitudinal cohort study of the incidence, neurobiology and prognosis of PD in Western and Southern Norway. All cases fulfilling strict diagnostic criteria of PD (n=265) were included between November 2004 and September 2006. A cohort of 212 patients with PD was eligible for long-term follow-up. In addition, a total of 201 normal control subjects from the same geographical area were recruited between November 2004 and April 2007. A subgroup of 175 control subjects was matched for age and sex. The participants were monitored closely over 7 years.

Proportions of patients falling at different stages of PD were calculated. Associated features and risk factors of falls were explored, both according to baseline values and longitudinal development. Volume, pattern and variability of physical activity (sedentary behavior, standing and ambulatory activity) were measured by triaxial accelerometers. Selected aspects of physical activity were compared between fallers and non-fallers.

Results Patients with established PD (mean disease duration 9 years) had a 10-fold higher frequency of falling compared with drug-naYve patients with incident PD. Falling in the established PD cohort was associated with more disability in everyday life (higher UPDRS ADL score) and motor complications (higher UPDRS complication of therapy score).

More than 1! of newly diagnosed PD patients reported falling at baseline or within the first year of follow-up. The proportion of patients falling increased during the 7-year follow-up, affecting almost 7j of all patients. Non-falling patients with incident PD had more than a 3-fold increased risk of falling compared with control subjects during the 7-year study period. Higher age at disease onset and early postural instability and gait difficulty (PIGD) phenotype were associated with increased risk of incident falls.

In patients with established PD, 72% reported falling after 8 years of prospective follow-up (mean disease duration 16 years). Symptoms representing non-dopaminergic deficiency (higher motor subscore B), higher levodopa equivalent doses and freezing of gait were associated with falls during the first 4 years of follow-up.

In a subgroup of patients from the Norwegian ParkWest cohort (mean disease duration 9 years), those who sustained a fall were more susceptible to being sedentary. Whereas motor impairment (higher UPDRS motor score) was associated with inactivity in non-falling patients, lack of confidence in being able to get up from floor unaided was associated with inactivity in patients with a recent fall history.

Conclusions Patients with PD fall more often than age- and sex-matched normal controls when newly diagnosed and during longitudinal follow-up. Newly diagnosed patients with PIGD phenotype and higher age may be candidates for specialized assessment and treatment interventions for preventing falls. Patients and caregivers need to be aware of a potential increase in sedentary behavior among patients with PD, especially following a recent fall. Practicing how to get up from floor may be beneficial in patients at risk of falling. The complex heterogeneity of PD and the identification of patients who are moving in and out of a frailty continuum remains a challenge for future research.