Sleep for all ages: at home or in the laboratory?

Abstract

The first double issue of the Journal of Sleep Research in 2011 contains articles devoted to sleep in children (Lemola et al., 2011; Spruyt et al., 2011), in older adults (Buman et al., 2011; Mesas et al., 2011), in narcolepsy patients (Hoang et al., 2011; Nakamura et al., 2011) and in tinnitus patients (Hebert et al., 2011). As sleep researchers continue to improve the description of sleep in all segments of the population, and sleep is frequently viewed as an important and modifiable lifestyle factor which contributes to health and wellbeing, the question of how and where sleep is assessed becomes ever more relevant. During the past 50 years or so, polysomnographic studies conducted in the laboratory have provided detailed descriptions of sleep physiology in a controlled but not very naturalistic environment. Much has been learned from these un-naturalistic laboratory studies, but the timing and duration of sleep recordings were often dictated by laboratory procedures rather than the research volunteer or patient. Although these studies may have reported total sleep time with resolution of minutes, its meaning is questionable when time in bed is fixed at 8 h. In other words, we cannot conclude much from these studies about sleep duration or sleep need across different ages, or other characteristics such as night-to-night variability in sleep timing and the differences between work and free days. On the other side of the sleep assessment spectrum, there is an extensive literature on sleep and health outcomes based on sleep duration, often assessed by a single question: ‘How long do you sleep?’. Many of these questionnaires have measured sleep duration with only a resolution of hours, rather than minutes and in most cases without separate estimates for work and free days. The validity of this type of assessment and the meaning of the answer can be questioned (Bliwise and Young, 2007). Nevertheless, much has been learned from these epidemiological studies about the association between self-reported sleep duration and health (e.g. Gallicchio and Kalesan, 2009). It is clear however, that there is room for improvement of both the laboratory studies as well as field studies. When sleep is recorded in the laboratory the timing and duration of the recording period could be based on the habitual sleep schedule of the participants and this is how a study in a sample of 22 tinnitus patients was conducted (Hebert et al., 2011). In addition, when assessing sleep in the laboratory, the use of sensitive quantitative electroencephalogram (EEG) analysis methods will aid discovery. Whereas in this study, visual scoring did not detect any differences with controls (n = 22), spectral analysis detected a reduction in slow wave activity which correlated with subjective sleep complaints. After completing the home set-up and laboratory-based polysomnography in 30 patients with suspected obstructive sleep apnoea, Campbell and Neill (2011) provide a confirmatory answer to this question, although concerns about signal loss remain. Bruyneel et al. (2011) compared sleep at home to sleep in the hospital sleep laboratory in a sample of 66 patients suspected of having sleep apnoea. They reported that during home recordings sleep efficiency was better, sleep duration was longer – by as much as 47 min – and rapid eye movement (REM) sleep expressed as a percentage of total sleep time was also higher, while at the same time the sleep apnoea index was not different between the two recordings. Measurement of sleep duration by actigraphy in larger samples of subjects of interest living in their normal environment is another interesting way forward. Lemola et al. (2011) measured sleep duration in 291 8-year-old children by actigraphy. Optimism, self-esteem and social competence were assessed by parents and/or teachers. Although the sample is still relatively small, the study overcomes an important limitation of many epidemiological studies and studies in children in which both the independent variable, i.e. sleep assessments, and outcome variables, i.e. health or psychological measures, were proved by the subjects or the parents. We should not forget however, that the use of actigraphy in the assessment of total sleep time also has limitations in children, as we can learn from the study by Spruyt et al. (2011), who compared actigraphy and polysomnography in a considerable sample of healthy children (n = 149; age 4.1–8.8 years). Furthermore, validation of actigraphy in adults has been mainly conducted only for nocturnal sleep, and validation should also include daytime naps; such a study is now available (Kanady et al., 2011). Thus there is a continuing need to develop and validate new methods which go beyond actigraphy to measure sleep in the home environment, and we are looking forward to a plethora of new sleep assessment devices and their validation (Van de Water et al., 2011).