Exercise therapy for prevention of falls in people with Parkinson's disease: A protocol for a randomised controlled trial and economic evaluation

The number of persons with Parkinson's disease over age 50 in Western Europe's five most populous nations was up to 4.6 million in 2005, and this will double to up to 9.3 million by 2030 [1]. Over 100,000 Australians are currently living with Parkinson's disease [2] and in line with estimates from Western Europe this number is also expected to double by 2030 [1]. Since the most common age of onset is 50–60 years and life expectancy is near normal, people with Parkinson's disease suffer the debilitating consequences of the disease over decades.

The mainstay of medical treatment for Parkinson's disease is pharmacological therapy to boost depleted dopamine levels. Despite optimal medication, people with Parkinson's disease living in the community experience frequent and recurrent falls with devastating consequences. Up to 68% of people with Parkinson's disease will fall and up to 46% of people with Parkinson's disease will experience recurrent falls each year [3‐6]. These rates are around twice of those in the general older population [6]. In addition, a 12-month prospective study (n = 113) found that 27% of people with Parkinson's disease fell at least once each month and 15% fell at least once a week [3].

Among people with Parkinson disease, as many as 65% of fallers will experience an injury secondary to their falls, 33% will suffer a fracture and 75% of falls will lead to use of a health care service [7]. Falls and related fractures are the most common secondary reason that people with Parkinson's disease are admitted to hospital [8]. These falls have devastating consequences and are accompanied by pain, reduced mobility and unacceptably high levels of caregiver stress. Fear of falling is also greater in community-dwelling people with Parkinson's disease than healthy controls [9]. This results in restriction of activities, compromising quality of life and predisposing to secondary reductions in muscle strength and cardiovascular fitness.

Recently Latt (2006) tested a battery of physiological and clinical variables considered to be potential risk factors for falls in people with Parkinson's disease [3]. These included Parkinson's specific impairments, such as slowness of movement, poor balance, freezing of gait and cognitive impairment as well as age-related impairments, such as reduced lower limb muscle strength. After adjusting for past falls, it was found that freezing of gait, poor balance and lower limb muscle weakness were independent predictors of falls. Therefore, the logical targets for an exercise program designed to reduce falls would be freezing of gait, balance and lower limb muscle strength.

Systematic reviews in the general older population, have found that exercise programs which specifically target balance and lower limb muscle strength are effective in preventing falls [10, 11]. In people with Parkinson's disease, lower limb muscle strength and regular exercise are significantly correlated with physical abilities [12‐15], therefore highlighting the role of exercise as an appropriate intervention in this population. Exercise has been shown to improve balance [16‐18] and strength [18, 19] and cueing training [20] has been shown to improve freezing of gait [17] in people with Parkinson's disease. However, no previous randomised controlled trials have investigated an exercise intervention aimed at reducing falls in people with Parkinson's disease by simultaneously targeting all three of these predictors of falls. The Weight-bearing Exercise for Better Balance (WEBB) exercise program has been developed to specifically target poor balance and lower limb muscle weakness for people at risk of falls. The PD-WEBB program used in this study includes progressive weight-bearing balance and strength exercises along with evidence-based cueing strategies to address freezing of gait. (The PD-WEBB program is available from the authors on request).

The primary aim of this randomised controlled trial is to determine the effectiveness of the PD-WEBB exercise program in reducing the rate of falls in people with Parkinson's disease. In addition, we will establish the cost effectiveness of the program from the health provider's perspective and determine the effects of the program on i) risk factors for falls, ii) physical abilities, iii) fear of falling and iv) quality of life.

There is an urgent need to identify cost-effective evidence-based interventions for reducing falls and related injuries in people with Parkinson's disease. To date, no adequately-powered studies have investigated exercise interventions aimed at reducing falls in people with Parkinson's disease. Only three randomised controlled trials have reported falls as outcomes. One study found a 38% (rate ratio = 0.62) reduction in rate of falling from a fully-supervised 8-week program of treadmill walking but was clearly under-powered (n = 18, 95% CI of RR 0.26 to 1.48, calculated by CS from published data) [35]. Another reported improved balance, walking and confidence in walking, and a non-significant 40% (odds ratio = 1.40) increase in risk of falling following a 3-week home-based program of rhythmical cueing of walking [17] (95% CI 0.63 to 3.1). The only study which targeted fallers [16] investigated a 6-week home-based program of multi-facetted exercise and showed promising results on balance and a non-significant, but 26% reduced risk of falls (risk ratio for repeat falls at 8 weeks = 0.74, 95% CI 0.47 to 1.15, calculated by CS from published data). A statistically and clinically beneficial reduction in falls risk might have become apparent if the interventions had been of longer duration.

The PD-WEBB program has been designed for falls prevention and so includes high challenge, yet safe, progressive balance exercises targeting standing balance and walking as well as progressive moderate to high intensity leg strengthening exercises with resistance applied using weighted vests. As different individuals have different preferences for exercise, participants who are randomly allocated to the exercise group will have the option to perform the 6-month program of exercises entirely at home (home-based mode) or in association with their Parkinson's support group (support group-based mode) with an additional home exercise program. Both modes of delivery incorporate semi-supervision by a physiotherapist and are designed to be individually tailored. The randomized controlled trial [17] which has addressed freezing (the RESCUE trial) using cued walking practice found no overall effect on freezing, but a significant effect in a subgroup of freezers. Therefore our exercise intervention will incorporate cueing strategies for overcoming freezing [17, 20, 24] as well as balance and strength exercises.

We will investigate the effect of the PD-WEBB program in people with Parkinson's disease using a randomized controlled trial incorporating the features known to reduce bias (i. e. concealed random allocation to groups, blinded outcome assessment and intention to treat analysis) and large enough to detect any effect of the intervention on fall rates. This project will determine the effectiveness of the exercise intervention in reducing falls, but will also determine the cost effectiveness of the intervention. This program, if found to be effective, has the potential to be implemented at a reasonable and sustainable cost within existing community services. Any reduction in falls for people with Parkinson's disease will reduce the personal and financial costs to individuals with Parkinson's disease, their families, health care resources and the community.