A 12-year polysomnographic study in Huntingtons: sleep problems predict disease onset and severity

Increasing evidence suggests that the sleep pathology associated with neurodegenerative diseases can in turn exacerbate both the cognitive deficits and underlying pathobiology of these conditions. Treating sleep may therefore bear significant, even disease-modifying, potential for these conditions, but how best and when to do so remains undetermined. Huntingtons Disease (HD), by virtue of being an autosomal-dominant neurodegenerative disease presenting in mid-life, presents a key model condition through which to advance this field. To date, however, there has been no clinical longitudinal study of sleep abnormalities in HD, and no robust interrogation of their association with disease onset, cognitive deficits and markers of disease activity. Here we present the first such study. HD gene carriers (n=28) and age- and sex-matched controls (n=21) were studied at baseline and 10- and 12-year follow up. All HD gene carriers were premanifest at baseline, and were stratified at follow up into prodromal/manifest and premanifest groups. Sleep abnormalities were assessed through two-night inpatient polysomnography (PSG) and two-week domiciliary actigraphy, and their association was explored against i)validated cognitive and affective outcomes (Montreal Cognitive Assessment, Trail A/B task, Symbol Digit Modalities Task [SDMT], Hopkins Verbal Learning Task [HVLT], Montgomery-Asberg Depression Rating Scale [MADRS]) and ii)serum neurofilament-light (NfL) levels. Statistical analysis incorporated cross-sectional ANCOVA, longitudinal repeated measures linear models and regressions adjusted for multiple confounders including disease stage. 15 HD gene carriers phenoconverted to prodromal/early manifest HD by study completion. At follow-up, these gene carriers showed more frequent sleep stage changes (p=<0.001,partial eta squared=0.62) and higher levels of sleep maintenance insomnia (p=0.002,partial eta squared=0.52). The latter finding was corroborated by nocturnal motor activity patterns on follow-up actigraphy (p=0.004,partial eta squared=0.32). Greater sleep maintenance insomnia was associated with greater cognitive deficits (Trail A p=<0.001,R squared=0.78;SDMT p=0.008,R squared=0.63;Trail B p=0.013,R squared=0.60) and higher levels of NfL (p=0.015,R squared=0.39). Longitudinal modelling suggested that sleep stage instability accrues from the early premanifest phase, whereas sleep maintenance insomnia emerges closer to phenoconversion. Baseline sleep stage instability was able to discriminate those who phenoconverted within the study period from those who remained premanifest (area under curve=0.81,p=0.024). These results demonstrate that the key sleep abnormalities of premanifest/early HD are sleep stage instability and sleep maintenance insomnia, and suggest that the former bears value in predicting disease onset, while the latter is associated with greater disease activity and cognitive deficits. Intervention studies to interrogate causation within this association could not only benefit patients with HD, but also help provide fundamental proof-of-concept findings for the wider sleep-neurodegeneration field.