353 Discussion Methodological considerations Our analyses presented in both Chapters 9 and 10 would have benefited from inclusion of control samples. In the clustering analyses, groups of people with insufficient sleep syndrome, chronic fatigue syndrome and/or obstructive sleep apnoea could for instance be used to test specificity of our clusters of people with central disorders of hypersomnolence. We aimed to include variables covering most important aspects of central disorders of hypersomnolence in our clustering efforts. It would, however, be interesting to repeat the analyses in an independent sample including additional disease-specific measures derived from polysomnography, MSLT and possibly also MRI. In the discussion of Chapter 10, we had to compare prevalence of opioid use to an external Dutch sample [372], as we lacked inclusion of a similarly aged sample of healthy individuals. The retrospective nature of our questionnaire and interviews could have also introduced possible recall bias. By including only opioid use in the past three years with opioid use verification through correspondence with the prescribing physician or pharmacist, we aimed to minimise these effects. Future perspectives Box 3 – Research agenda Section C • Objective quantification of sleep drunkenness • Are there sex-specific subtypes of narcolepsy type 1? • Are hypocretin neurons destructed or is hypocretin deficiency reversible? • The role of hypocretin in opioid addiction? Objective quantification of sleep drunkenness Our findings highlight the potential of sleep drunkenness and difficulty awakening to differentiate those without cataplexy. These measures represent daytime complaints associated with increased sleep need [331, 457]. For our clustering algorithm these measures were only subjectively assessed and objective quantification could improve reproducibility of such markers. Increased lapses and longer response times during the psychomotor vigilance task (PVT) recently significantly correlated with presence and severity of sleep drunkenness after nocturnal sleep in people with central disorders of hypersomnolence [458]. Much remains unclear on this topic and should be considered in future research, as this study did not evaluate intrasubject reproducibility across multiple measurement days, task performance directly 11
RkJQdWJsaXNoZXIy MjY0ODMw