Alertness, mood and performance rhythm disturbances associated with circadian sleep disorders in the blind

Abstract

Blind people report disturbances in alertness, mood and performance. In laboratory studies, these waking functions can only be maintained when the wake-dependent deterioration is opposed by appropriately-timed endogenous circadian rhythms. We aimed to quantify whether variations in waking function experienced by blind people living in society were dependent on the phase relationship between the sleep-wake cycle and the circadian pacemaker. The time course of alertness, mood and performance was assessed in 52 blind subjects with and without circadian rhythm disorders every 2 h for 2 days per week for 4 weeks. Sleep-wake timing and circadian phase were assessed from diaries and weekly measurements of urinary 6-sulphatoxymelatonin rhythms, respectively. In those subjects who woke at either a normal circadian phase (n = 26) or abnormally early (n = 5), alertness, mood and performance deteriorated significantly with increased time awake (P < 0.05). In 17 non-entrained ('free-running') subjects, waking function varied significantly with circadian phase such that subjects rated themselves most sleepy (P = 0.03) and most miserable (P = 0.02) when they were awake during the time of peak melatonin production. The internal phase relationship between sleep-wake behaviour and the circadian melatonin rhythm in entrained subjects contributed to predictable differences in the daily profile of alertness, mood and performance. Disruption of this phase relationship in non-entrained blind individuals with circadian rhythm sleep disorders resulted in impaired waking function during the day equivalent to that usually only experienced when awake during the night. Treatment for circadian rhythm disorders should be targeted in normalizing these phase relationships.

Figure 1

Double raster plots of subjective sleep and nap episodes (black bars) for four blind subjects with differing circadian rhythm types: normally entrained (top panel), entrained but with an advanced phase (second panel), entrained with a delayed phase (third panel) and ‘free-running’ (lower panel) with a non-24-hour period (see text for definitions). Sequential study days are plotted vertically and horizontally with clock time on the horizontal axis. Weekly assessments of cosinor-derived aMT6s acrophase times are represented by asterisks with the associated lines of best fit (regression) used to determine circadian period shown for each subject. The average (± SD) night-time sleep onset (23.93 ± 0.92 h) and offset times (7.15 ± 0.81 h) for the blind subjects with light perception and normally-phased aMT6s acrophase times (4.57 ± 0.98 h; n = 22) are denoted to highlight the differences in sleep-wake behaviour. aMT6s, 6-sulphatoxymelatonin.

Figure 2

Average aMT6s profiles for each circadian rhythm type (filled profile) plotted concurrently with the frequency of waking from all sleep episodes (white columns) as a function of aMT6s phase for different circadian rhythm types (see Fig. 1 and text for definitions). aMT6s, 6-sulphatoxymelatonin.

Figure 3

Distribution of subjective alertness ratings during wakefulness in relation to clock time (left panel) and circadian phase (right panel) for normally entrained (NE), abnormally entrained (AE) advanced, AE delayed, and non-entrained (FR, ‘free-running’) subjects. When plotted (inappropriately) relative to clock time, there appears to be little difference between the data distributions of the four groups. When plotted appropriately against circadian time, however, clear differences are observed in the data distribution in the AE and FR subjects with respect to time awake and circadian phase. As most of these subjects attempted to sleep on a normal 24-h day, we were able to make observations of waking behaviour at particular interactions of time awake (Process S) and circadian phase (Process C) that are never observed in NE subjects (NE, top right panel). The group data from non-entrained subjects are distributed near-equally across all possible interactions of time awake and circadian phase because of the natural ‘forced desynchrony’ caused by desynchronization between the near-24-hour sleep-wake pattern and the non-24-hour circadian pacemaker (bottom right panel). The performance data also show similar distributions in relation to time wake and circadian phase for each circadian rhythm type (data not shown).

Figure 4

Average alertness, mood and performance (least squares mean z-score ± SEM) plotted in relation to time elapsed since waking from night-time sleep episodes for normally and abnormally entrained subjects (see Fig. 1 and text for definitions). Parameters with significant changes with respect to time since waking are shown by the filled symbols.

Figure 5

Average alertness, mood and performance (least squares mean z-score ± SEM) double-plotted in relation to circadian phase (left panel) and time elapsed since waking from all sleep episodes (right panel) for free-running subjects. Parameters with significant changes with respect to time since waking and circadian phase are shown by the filled symbols.

Figure 6

Average alertness (least squares mean z-score ± SEM) double-plotted in relation to time elapsed since waking from night-time sleep episodes for AE subjects (●, panel a) and AD subjects (●, panel b). Alertness data from free-running subjects for the corresponding circadian phase and time elapsed since waking are also shown (◯) for the two circadian phase types and demonstrate similar time courses.