Abstract
Objective: There is an ever growing body of literature which endorses the view that sleep evolves with normal aging and also modified by being either a man or woman. Changes in sleep patterns have been consistently described from childhood, through adolescence, adulthood and into old age, with sleep continuity measures seen to decline with aging and to a greater extent in men compared to women. However, when sleep is evaluated using subjective means, elderly women are more likely to report poor quality sleep compared to men. We evaluated the subjective (self-reported) and objective (polysomnographic and EEG power spectra) of 181 healthy subjects under normal and experimental challenge by traffic noise with three defined aims; firstly, to describe the nature of aging and sex-related effects upon sleep. Secondly, to evaluate the inter-relationship between objective (PSG & spectral) and subjective (self-reported) measures of sleep. Finally, to assess the influence of these age and sex upon sleep propensity when sleep is experimentally challenged (ESF). Methods: Subjective and objective sleep data from two sister studies were subjected to an exploratory analysis. Polysomnographic measures were derived using standard evaluation methods (R&K guidelines), with EEG power spectra measures up to 32. 0 Hz computed from artefact-free manually scored REM and NREM sleep epochs. Effects by aging, sex, EEG topography, model (ESF) or by interaction were evaluated using two-way unbalanced ANOVA (chapter 3) or with a correlated-errors method (chapters 4 & 5). The Cohen's d statistic (effect size) contrasted the magnitude of the differences in EEG power density values of young subjects versus older subjects or between sexes. Finally, estimates of the age-related changes to polysomnographic measures across the adult life course were modelled using a logistic regression method (chapter 6). Results: Objective measures were more sensitive to the effects of both aging and sex than subjective measures, with age-related trends indicating that the sleep of men was subject to greater modification over the adult life course, compared to women. Declines in delta and theta EEG power were evident, with concomitant increases in alpha, sigma and beta frequencies with aging, across all topographic regions in men, a response which was absent in women. Interestingly, the effects of aging upon quantitative EEG measures in NREM sleep was "topographically global" in men, but "regional" in women. In REM sleep, both men and women exhibited more regional loci for effects, although in women, effects were prominent in anterior (frontal) derivations, with men being more posterior (occipital) derivations. ESF did elicit an age, sex and model interaction for total sleep time (TST) and self-reported wake after sleep onset (sWASO). These results suggested that ESF reduced TST in young women, with less modification to total sleep seen in older women, with older men subjectively reporting more wake episodes following sleep onset than women. Likewise, ESF effects upon REM sleep elicited a markedly variable electrophysiological response in women relative to men. Conclusions: Objective measures of sleep were more sensitive than self-reported measures to the effects of aging and sex, but also illustrated that the sleep measures vary predominantly along aging lines with sex-specific effects being more subtle. Furthermore, the sex-specific effects upon REM sleep may contribute to the predisposition of insomnia symptoms reported by women. With known contributing factors excluded from our clinical population (e. g. pathology, societal pressures and comorbid conditions), our results indicate that the normal evolution of sleep elicits changes in perceptual and electrophysiological parameters, consistent with insomnia. As such, education of the "perceptual nature" of the evolution of sleep and the application of psychological therapies for insomnia (e. g. CBT-I) may provide a better long term management of this disorder.