Role of Sleep on Motor Learning in Parkinson's Disease and Healthy Older Adults (TARGET-SLEEP)

Towards Retention of Motor Learning in Parkinson's Disease: Understanding Sleep-related Effects of Consolidation

People with Parkinson's disease (pwPD) often present difficulty consolidating newly learned skills into long-term memory. Sleep facilitates motor memory consolidation in healthy adults, especially in combination with targeted memory reactivation (TMR). TMR works by adding associated sounds during learning that are replayed during sleep and thus reinforce the recently formed neural connections. Importantly, recent work suggested that consolidation during sleep may be preserved in pwPD, but robust findings are lacking and have not involved TMR. The objective of the present study is to address this imperative question by investigating the effect of napping on motor memory consolidation by experimentally manipulating exposure to sleep and TMR for the first time. Concretely, the investigators will first compare the effect of a 2-hour nap to that of a wake control period in pwPD and healthy age-matched controls. A validated motor sequence learning task will be used to test for behavioral markers of motor learning and polysomnography with electroencephalography (EEG) will be conducted to study the neural correlates of sleep-related motor learning effects. In a second experiment, the investigators will then test the effects of adding TMR during post-learning sleep, by comparing performance on two motor sequences of which only one is reactivated during post-learning napping using auditory TMR.

PwPD often fail to retain training effects via the process of motor memory consolidation, by which newly acquired skills transform intro robust and long-lasting motor memories without further practice. Compromised consolidation leads to an inevitable deterioration of daily functioning while hindering the prolonged effects of rehabilitation even in the early stages of the disease (Nieuwboer et al. 2008). Intriguingly, post-training sleep facilitates consolidation in healthy adults (King et al. 2017a) and this effect may be preserved in pwPD (Terpening, 2013). Targeted Memory Reactivation (TMR) is a technique tested in young adults, whereby auditory stimuli are added during motor learning. The learning-related sounds are then replayed during post-training non-rapid eye movement (NREM) sleep to reinforce the recently formed neural connections (Diekelmann et al. 2012). The overarching hypothesis of this project is that bouts of sleep and TMR will improve the consolidation of motor memories and markers of neuroplasticity in pwPD and older adults. To test this, the investigators will employ a 'napping' protocol that accounts for circadian effects while allowing performance after diurnal sleep to be directly compared to that of a wake control group (King et al. 2017a). Consolidation will be defined as the change in Motor Sequence Learning (MSL) of finger tapping after a post-training period of either napping or wakefulness compared to the end of initial training. To further indicate robust consolidation, changes in performance will be assessed after a 24h retention period without further practice as well as during a dual-task as a measure of motor automaticity. A parallel group design will allow within group comparison (nap/wake) as well as between pwPD and controls. In a second study, the effects of TMR on consolidation will be compared across groups using a serial reaction time task (SRT). The first objective (Experiment 1) is to determine whether a 2-hour nap improves the immediate consolidation, 24h retention and dual task interference of an MSL task as compared to a similar period of diurnal wakefulness in people with pwPD and healthy age-matched controls and whether the degree of performance change is different between these groups. Hypothesis 1: The investigators expect to find improved consolidation, 24h retention and reduced dual-task interference of MSL performance following a post-training nap compared to wakefulness in both groups. Possibly, improvements are less apparent in pwPD compared to controls due to their cortico-striatal impairments. The second objective (Experiment 2) is to determine whether TMR improves immediate consolidation, 24h retention and dual task interference in pwPD and healthy older adults by comparing performance on two learned motor sequences before and after a 2-hour nap period, during which one of the two sequences is replayed using auditory TMR. Hypothesis 3: TMR during napping will improve immediate consolidation, 24h retention and dual task interference of the SRT in both healthy elderly and PD. Participants first undergo screening, during which demographics, cognitive capacity and disease severity indexes (including dexterity tests) will be obtained prior to undergoing a diagnostic screening night with polysomnography (PSG) to assess for sleep disorder features. Participants will also complete a test battery on sleep quality scales and mood and wear an Actigraphy watch at home for at least five days and nights prior to the first experiment. During experiment 1, participants learn the MSL by self-initiating a 5-element finger sequence that is presented on screen. After learning, participants will be equipped with PSG, which includes EEG. Based on blinded randomization, they will nap for 2 hours or lie on the bed but remain awake for a similar duration. The wake PSG will ensure that no participant in the wake group falls asleep. Participants will then enjoy a 30-45min break to counter sleep inertia effects, prior to being re-tested on the MSL (Retest 1). The next day, participants will be re-assessed on the MSL for 24h retention testing (Retest 2). During experiment 2, similar procedures will be followed as described above except that participants will learn two new finger sequences that are auditory cued, by means of a serial reaction time task (SRT). For the SRT, participants view a row of empty squares presented in the middle of the screen and each time a square is highlighted the participant is instructed to tap the finger that is spatially associated to that square as quickly and accurately as possible, i.e. a serial reaction time task. The difference between the MSL task of experiment 1 and the SRT task of experiment 2 is therefore that during experiment 1 participants self-initiate a sequence that is explicitly shown to them, whereas in experiment 2 the sequence is cued. The order of sequence blocks during learning and retest as well as the sequence selected for TMR will be randomized across participants. Performance on both sequences will be re-assessed after the break, and again at 24h retention without auditory cues. The MSL and SRT tests in both experiments will be preceded by a psychomotor vigilance test as an objective measure of the participants' vigilance on the day and include a single- and dual-task condition. Power calculation: Based on the findings by Terpening et al. (2013) and Dan et al. (2015), a minimum of 16 subjects per group (NAP, WAKE) will be required according to our power analysis based on the MSL-outcomes using β=0.20 and α=0.05 to detect a significant group difference. To account for potential dropouts, the recruitment target is set 20% higher to ensure adequate power in our final analysis. As such, a total of 40 PD patients and 40 healthy elderly controls will be recruited for experiment 1 (i.e. 20 in each NAP/WAKE group). The best sample estimation at this time for experiment 2 is based on previous TMR studies in younger adults also recruiting 16 subjects per nap/wake group (Antony et al. 2012). Therefore, we will target to recruit a total of 20 PD and 20 healthy elderly controls for Experiment 2, again accounting for 20% potential dropout.

Performance on the MSL task (Experiment 1) will be compared between a group that undergoes a 2-hour post-learning NAP and a group that will undergo a 2-hour post-learning WAKE period. Participants will be randomized (1:1) to either the NAP or WAKE group. Performance on the SRT task (Experiment 2) will be compared between the sequence that was replayed during the post-learning nap using auditory TMR (replay) and the sequence that was not replayed (no-replay). The order of sequence blocks during learning and retest as well as the sequence selected for TMR will be randomized across participants. Randomization for both experiments will be done by an independent researcher who is not involved in the measurements of any of the studies using a computerized random number generation technique.

For experiment 2, participants will be told that sounds may be played during the nap or wake period, without further knowledge on the anticipated effects of these sounds.

For experiment 1, the NAP group will undergo a post-learning 2-hour diurnal sleep opportunity (i.e. 'nap') without cues. For experiment 2 the NAP+TMR group will undergo a post-learning 2-hour diurnal sleep opportunity (i.e. 'nap') with auditory TMR. The learning related sounds will be presented to participants at 140% of their minimal auditory detection threshold during stage 2 and stage 3 of NREM sleep.

For experiment 1, the WAKE group will undergo a post-learning 2-hour period of quiescent wakefulness without cues.

Participants perform a self-initiated MSL task by tapping a five-element finger sequence presented on screen as rapidly and accurately as possible with their non-dominant hand for 18 blocks during learning and again at each retest assessment. Each block consists of 50 key presses (ideally 10 sequences) and is followed by a rest block of 15-20 seconds without finger tapping. A two-minute rest period will be implemented after 14 blocks to further minimize the effects of fatigue on the last 4 blocks that are used to calculate the primary outcome. Performance on the MSL will be assessed using the 'Performance Index (PI)' [PI=exp^-(seqDur) * exp^-(Errors/12) * 100], taking both speed and accuracy into account (King et al. 2017b). After learning the MSL, participants are randomly allocated to undergo a post-learning 2-hour diurnal sleep opportunity (NAP) or 2-hour period of quiescent wakefulness (WAKE) before being reassessed on the MSL.

Change in PI between the first 4 blocks immediately after the 2-hour NAP or WAKE intervention (Retest 1) and the last 4 blocks of learning immediately prior to the intervention.

The same MSL task as described above in Primary outcome 1 is again repeated 24-hours after Retest 1 in order to assess whether the sleep-related effects on motor memory consolidation are retained in the long-term (Retest 2).

Change in PI between the first 4 blocks after the 24-hour retention period (Retest 2) and the last 4 blocks of Retest 1 immediately after the 2-hour NAP or WAKE intervention.

Experiment 2 is similar to experiment 1, except that participants will learn two motor sequences that are visually and auditory cued by means of a serial reaction time task (SRT). After learning both sequences, participants will nap for 2-hours, but this time while one of the two auditory sequences will be replayed during NREM sleep. Performance on both sequences will be re-assessed immediately after the intervention (Retest 1), and again at 24h retention (Retest 2). The PI will be used to assess performance on the task and compared between the sequence that was replayed and the sequence that is not replayed.

Change in PI between the first 4 blocks immediately after the nap+TMR intervention (Retest 1) and the last 4 blocks of learning immediately prior to the intervention.

The same SRT task as described above in Primary outcome 3 is again repeated 24-hours after Retest 1 in order to assess whether the sleep- and TMR-related effects on motor memory consolidation are retained in the long-term.

Change in PI between the first 4 blocks after the 24-hour retention period (Retest 2) and the last 4 blocks of Retest 1 immediately after the 2-hour NAP+TMR intervention.

The same MSL task as described above in Primary outcome 1 will be performed for an additional 4 blocks at learning and 4 blocks immediately after the intervention (Retest 1), but this time while participants consecutively perform a shape-counting dual task. During the single-task MSL condition, participants view a fixation cross in the middle of the screen, with the sequence presented above the fixation cross. During the MSL dual-task condition, the fixation cross pseudo-randomly changes shape (e.g. "X" or "O") and participants are instructed to count how often in each block a shape change occurred, while they continue to perform the MSL. Participant's responses on the dual task condition will be recorded after each block.

Change in PI between the 4 blocks of dual tasking immediately after the 2-hour NAP or WAKE intervention (Retest 1) and the 4 blocks of dual tasking during learning prior to the intervention.

The same MSL with dual task as described above in Secondary outcome measure 1 will again be repeated 24-hours after Retest 1.

Change in PI between the 4 blocks of dual tasking after the 24-hour retention period (Retest 2) and the 4 blocks of dual tasking at Retest 1 immediately after the 2-hour NAP or WAKE intervention.

The same SRT task as described above in Primary outcome 3 will be performed for an additional 4 blocks immediately after the intervention (Retest 1), but this time while participants consecutively perform a shape-counting dual task. During the single-task SRT condition, participants view empty squares in the middle of the screen, which are filled (i.e. highlighted) one by one in the order of the sequence that is being learned. During the SRT dual task condition, the same squares pseudo-randomly fill with a different shape (e.g. "X" or "O") and participants are instructed to count how often in each block a shape change occurred, while they continue to perform the SRT. Participant's responses on the dual task condition will be recorded after each block.

Difference in PI between sequences A and B assessed across the 4 blocks of dual tasking immediately after the 2-hour NAP+TMR intervention (Retest 1).

The same SRT with dual task as described above in Secondary outcome measure 7 will again be repeated 24-hours after Retest 1. All comparisons using the PI as the main dependent variable of interest, as well as the tertiary outcomes are listed in the attached statistical analysis plan.

Difference in PI between sequences A and B assessed across the 4 blocks of dual tasking after the 24-hour retention period (Retest 2).

Inclusion Criteria: Right handed Can read and understand Dutch Age equal or greater than 40 years PwPD will have a clinical diagnosis of idiopathic Parkinson's disease made by a Neurologist Completed written informed consent approved by the assigned medical ethical committee Exclusion Criteria: Receiving deep brain stimulation Enrollment in an interventional trial for Parkinson's disease therapy Severe sleep apnea determined as an Apnea/Hypopnea index (AHI) > 30 during the screening polysomnography (PSG) Cognitive impairment that could question the participant's ability to provide voluntary informed consent as determined by an Mini Mental State Examination score <24 Co-morbidities that would hamper interpretation of MSL or SRT learning, such as musculoskeletal abnormalities, as determined by a Neurologist or Physical Therapist.

Pseudonymized data files may be shared upon publication of the results, but on no occasion will any identifiable information or contact details of the participants be presented or made visible. Information on data management is provided to participants on the written informed consent form. The polysomnography data will be made available for open access sharing if approval is obtained from the participant on the participant consent form. The informed consent form will include modules explaining why open access sharing is requested. The subjects will be informed about the data that is intended for open access sharing and will have the opportunity to opt out without any consequences to their current or future participation or care at the University of Leuven (KU Leuven) or University hospitals Leuven (UZ Leuven), via a tick-box on the informed consent form. The subject's privacy will be protected.

Pseudonymised data will only be shared with research projects that have obtained written approval from a Medical Ethical Committee to use the data for a specific research purpose. Requests for accessing the data should then be made to the PI of the study, Prof Nieuwboer.

Nieuwboer A, Rochester L, Muncks L, Swinnen SP. Motor learning in Parkinson's disease: limitations and potential for rehabilitation. Parkinsonism Relat Disord. 2009 Dec;15 Suppl 3:S53-8. doi: 10.1016/S1353-8020(09)70781-3.

King BR, Hoedlmoser K, Hirschauer F, Dolfen N, Albouy G. Sleeping on the motor engram: The multifaceted nature of sleep-related motor memory consolidation. Neurosci Biobehav Rev. 2017 Sep;80:1-22. doi: 10.1016/j.neubiorev.2017.04.026. Epub 2017 Apr 29.

King BR, Saucier P, Albouy G, Fogel SM, Rumpf JJ, Klann J, Buccino G, Binkofski F, Classen J, Karni A, Doyon J. Cerebral Activation During Initial Motor Learning Forecasts Subsequent Sleep-Facilitated Memory Consolidation in Older Adults. Cereb Cortex. 2017 Feb 1;27(2):1588-1601. doi: 10.1093/cercor/bhv347.

Terpening Z, Naismith S, Melehan K, Gittins C, Bolitho S, Lewis SJ. The contribution of nocturnal sleep to the consolidation of motor skill learning in healthy ageing and Parkinson's disease. J Sleep Res. 2013 Aug;22(4):398-405. doi: 10.1111/jsr.12028. Epub 2013 Feb 11.

Diekelmann S, Biggel S, Rasch B, Born J. Offline consolidation of memory varies with time in slow wave sleep and can be accelerated by cuing memory reactivations. Neurobiol Learn Mem. 2012 Sep;98(2):103-11. doi: 10.1016/j.nlm.2012.07.002. Epub 2012 Jul 10.