Abstract
The initial rationale of this thesis was to conduct an extensive investigation into the circadian rhythms of blind subjects. An epidemiological study of registered blind subjects (N = 388) established that 58% had a sleep disorder as assessed by the Pittsburgh Sleep Quality Index (PSQI > 5). The incidence and severity of sleep disorder was greater in subjects with no conscious light perception (NPL) (78%, PSQI mean±sem = 8.1±0.5, n = 58) compared to subjects with some degree of light perception (LP) (54%. PSQI mean = 5.8±0.2, n = 330). In order to investigate whether the high incidence of sleep disorder was due to circadian rhythm abnormality, a detailed evaluation of circadian rhythms in the blind was conducted. A total of 49 blind subjects (30 NPL and 19 LP) with varying degrees of severity of visual loss were studied for at least four weeks. Subjects were required to keep daily sleep and naps diaries and wear wrist-worn actigraphs continually. They were asked to complete four analogue mood scales and an auditory performance test every two hours whilst awake for 48 h each week. During the same 48 h each week, urine was collected (4-hourly and 8-hourly overnight) for analysis of 6-sulphatoxymelatonin (aMT6s), the major metabolite of melatonin and considered to be a reliable marker of circadian phase. The majority of LP subjects (14/19, 74%) had normally entrained aMT6s rhythms whereas the majority of NPL subjects had abnormal aMT6s rhythms (23/30, 76%). Free-running aMT6s rhythms were only found in NPL subjects (17/30, 57%) the incidence of which was greater in bilaterally enucleated subjects (11/12) compared to those with one (5/7) or both eyes (2/12) present. All of the NPL individuals with abnormally entrained rhythms (n = 5) had both eyes present. The effect of the circadian system on sleep was investigated using both the subjective and actigraphically-derived sleep data. In group analysis of subjects with entrained aMT6s rhythms, there was a strong positive correlation between the timing of sleep and activity with the timing of the aMT6s acrophase. Regression analysis of sleep and activity showed that some individuals with a free-running aMT6s rhythm also had a free-running activity, sleep or nap rhythm although at a much reduced tau compared to the aMT6s rhythm. The effect of age on the timing of behavioural rhythms was also measured in entrained subjects and found that increased age was associated with an advance in the timing of mean sleep onset. The study also showed that there were significantly more naps of a longer duration in NPL subjects (mean±sem = 0.6±0.1 naps/day, 0.4±0.1 h/day) compared to LP subjects (0.3±0.1 naps/day, 0.2±0 h/day). When analysed with respect to circadian rhythm type, the results showed that normally entrained subjects had significantly fewer naps per day compared to all other circadian rhythm types. The timing of the naps was not random; there were significantly more naps (65%) within a 5 h range before and after the aMT6s acrophase. Further analysis of sleep, naps and activity with respect to circadian phase in free-running subjects showed that the number and duration of naps decreased when the aMT6s acrophase was in a normal phase position (24.00 - 06.00 h) and increased when the aMT6s acrophase was maximally out of phase. There were also significant effects of circadian phase on the timing and duration of night-time sleep with maximal night sleep duration occurring when the aMT6s acrophase was normal and minimal when the aMT6s acrophase was out of phase. The timing of sleep onset and offset also changed with circadian phase, being advanced when the aMT6s was advanced relative to normal (18.00 - 24.00 h) and delayed when the aMT6s acrophase was relatively delayed (06.00 - 12.00 h). It was observed that sleep onset and offset did not vary in parallel with circadian phase and this may provide the basis of a model to investigate the effects of circadian influences on sleep. Preliminary analysis of mood and performance did not reveal a robust rhythm as assessed by fitting cosine curves to the data. There was a significant effect of circadian phase on daily mean mood (calmness) and performance (% gaps) in group analysis of six free-running subjects although more detailed analysis is required before any conclusions can be drawn. Comparisons were performed on the simultaneous subjective and actigraphic measurement of sleep for the whole study in all subjects. There was a high rate of disagreement between the two methods in the measurements of absolute sleep parameters , especially when measuring the number and duration of day- or night-time sleeps. However, both methods gave comparable results when measuring changes in sleep over time and the effects of circadian phase on sleep. These results of this thesis show that individuals with no light perception (NPL) are likely to have abnormal circadian rhythms. This is associated with increased incidence and duration of naps, the timing of which is associated with the timing of the aMT6s acrophase. In some free-running individuals, there are also distinct cyclical variations in the timing of sleep in relation to circadian phase.