LAUSR

Ms. X returned for follow-up of her Parkinson’s disease. Last time, we were getting concerned; she had many lingering issues with fatigue, anxiety, trouble concentrating, gait slowing, and so forth. We had decided to increase the levodopa dose. Now, she looked remarkably well — less anxious, mood better, feeling energetic... so, how did the dose increase go? In fact, she never increased it; rather, she had started a new exercise program and was working out vigorously 5 days a week. She felt so well that the dose increase seemed unnecessary. This n = 1 experience is becoming more common in clinicians’ offices as the potential benefits of exercise on Parkinson’s are becoming broadly disseminated. Randomized, controlled trials of exercise generally show some benefits for objective testing, using measures such as the Unified Parkinson Disease Rating Scale (UPDRS). However, except for larger effects on gait, objective motor benefits seem relatively modest, at least compared with these n = 1 subjective reports. Might there be benefits to exercise that we just do not capture on our motor evaluations? In this month’s Movement Disorders, Amara et al have started to answer that question by looking at the effects of exercise on sleep. In this randomized, controlled trial, they recruited patients with mild-moderate PD who were not currently engaged in regular highintensity exercise. Recruitment was stratified to age and sex, although patients were not selected based on having sleep problems. Participants were then randomized to either a nonexercise sleep intervention or an exercise program. The exercise program was broad, focusing on strength training, balance, movement speed, and cardiovascular conditioning. It was conducted via inperson sessions 3 times per week. The nonexercise intervention was a sleep hygiene intervention mainly consisting of a single physician consultation and provision of literature on sleep hygiene training, supplemented by monthly telephone follow-up. So, this is essentially a head-to-head study of simple sleep hygiene training versus exercise for sleep in PD. The main outcome measures were objective (polysomnography) and subjective sleep questionnaires. The primary outcome was sleep efficiency on the polysomnogram; sleep efficiency refers to the percentage of the sleep period that one spends actually asleep. Numerous other objective sleep quality and sleep stage variables were assessed. The polysomnogram was conducted once at baseline, once on the same night as the exercise intervention at study end (termed “acute exercise”), and again postintervention on an “off-exercise” night (termed “chronic exercise,” which was set as the primary comparator). Subjective outcomes included questionnaires related to sleep overall (the Pittsburgh Sleep Quality Index [PSQI], somnolence [the Epworth sleepiness scale], and fatigue [the fatigue severity scale]). The primary outcomes here are reported for the perprotocol analysis rather than intention to treat (ie, reporting those who actually followed the intervention, rather than all randomized to the intervention). These analysis methods have different purposes. Intention-totreat analyses are relatively conservative and are essential in licensing a new treatment (mostly because those who drop out from the intervention could be different from the rest, therefore biasing study results). On the other hand, per-protocol analyses may give a clearer picture of the amplitude of effects actually obtainable by the intervention. The primary outcome assessment found significant improvement in sleep efficiency in the exercise group compared with sleep hygiene. On other polysomnogram measures, there were additional reductions in wake after sleep onset, increased total sleep time, and increases in both slow-wave sleep (“deep” sleep) and rapid eye movement sleep. These findings were much clearer on the “chronic exercise” night and were © 2020 International Parkinson and Movement Disorder Society