Plasticity of the intrinsic period of the human circadian timing system

Abstract

Human expeditions to Mars will require adaptation to the 24.65-h Martian solar day-night cycle (sol), which is outside the range of entrainment of the human circadian pacemaker under lighting intensities to which astronauts are typically exposed. Failure to entrain the circadian time-keeping system to the desired rest-activity cycle disturbs sleep and impairs cognitive function. Furthermore, differences between the intrinsic circadian period and Earth's 24-h light-dark cycle underlie human circadian rhythm sleep disorders, such as advanced sleep phase disorder and non-24-hour sleep-wake disorders. Therefore, first, we tested whether exposure to a model-based lighting regimen would entrain the human circadian pacemaker at a normal phase angle to the 24.65-h Martian sol and to the 23.5-h day length often required of astronauts during short duration space exploration. Second, we tested here whether such prior entrainment to non-24-h light-dark cycles would lead to subsequent modification of the intrinsic period of the human circadian timing system. Here we show that exposure to moderately bright light ( approximately 450 lux; approximately 1.2 W/m(2)) for the second or first half of the scheduled wake episode is effective for entraining individuals to the 24.65-h Martian sol and a 23.5-h day length, respectively. Estimations of the circadian periods of plasma melatonin, plasma cortisol, and core body temperature rhythms collected under forced desynchrony protocols revealed that the intrinsic circadian period of the human circadian pacemaker was significantly longer following entrainment to the Martian sol as compared to following entrainment to the 23.5-h day. The latter finding of after-effects of entrainment reveals for the first time plasticity of the period of the human circadian timing system. Both findings have important implications for the treatment of circadian rhythm sleep disorders and human space exploration.

Figure 1. Raster plot of the study design for a representative individual (2478).

The study protocol is double plotted such that consecutive days are next to and beneath the other. Scheduled sleep episodes in complete darkness (horizontal black lines), exposure to moderately bright light (parallelogram boxes with sun; part of wake episodes Days 6–19, 38–51), and CP procedures (open horizontal boxes; on Days 4–5, 20–21, 36–37, 52–53, 68–69) are indicated. Except for the scheduled wake episodes during the baseline (Day 1–3) and recovery days (Day 70–73; ∼90 lux [∼0.23 W/m²]) and exposure to moderately bright light (∼450 lux [∼1.18 W/m2]), lights were dim throughout the study (∼1.8 lux [∼0.0048 W/m²]). DLMOn_25% (green upward-triangles) and DLMOff_25% (red downward-triangles) are shown during dim light conditions. The intrinsic circadian period of the melatonin, cortisol, and temperature cycles during the FD protocols (Days 22–35, 54–67) were computed by a nonorthogonal spectral analysis technique. The fitted phase of the minimum of core body temperature (gray dotted line) and of the maximum of plasma cortisol (black dashed line) during the FD protocols are indicated. The observed circadian periods during the exposure to the 24.65-h day and to the 23.5-h day and the composite estimate of the intrinsic circadian period as computed by averaging the intrinsic period estimate from plasma melatonin, plasma cortisol, and core body temperature data (τ_composite) for both FD protocols are indicated to the right of the figure.

Figure 2. Lighting regime was effective for entraining circadian timing system to 24.65-h and 23.5-h rest-activity cycles.

(A) The phase angles of entrainment were not different between the CP protocols at baseline and that following the 23.5-h days. The phase angles of entrainment were significantly different between the CP protocols at baseline and that following the 24.65-h days, such that the melatonin DLMOn_25% and DLMOff_25% during the CP protocol following the 24.65-h days relative to the scheduled sleep episode occurred even later than during the baseline CP protocol. Therefore, the lighting regimes were effective in maintaining entrainment to both non-24-h sleep-wake cycles at an appropriate phase angle. During the CP protocol following fourteen 23.5-h days (post 23.5-h), the timing of the scheduled sleep episode occurred 7 clock hours earlier than during the CP protocol preceding these shorter-than-24-h days. During the CP protocol following fourteen 24.65-h days (post 24.65-h), the timing of the scheduled sleep episode occurred 9.1 clock hours later than during the CP protocol preceding these longer-than-24-h days. Open bars, scheduled 8-h sleep episodes during CP protocols; green upward-triangle, DLMOn_25%; red downward-triangle, DLMOff_25%; error bars, SEM; *, P<0.01; **, P<0.005. (B) Light was effective for entraining the circadian timing system, such that the observed circadian period was similar to the period of the imposed non-24-hour sleep-wake cycles. Vertical lines, 23.5-h and 24.65-h day-night cycles (T-cycles); Error bars, 95% confidence interval.

Figure 3. Plasticity of the human circadian period.

Intrinsic circadian period was significantly longer following the 24.65-h day versus than following the 23.5-h day, as estimated during two-week FD protocols in dim light conditions. This was true when using plasma melatonin data (P = 0.002), plasma cortisol data (P = 0.004), core body temperature data (P = 0.02), and the composite (P = 0.002) for period assessment, the later depicted here. There were no significant differences between melatonin, cortisol, and temperature period assessments. There was no significant order effect. Small filled circles; individual changes in intrinsic circadian period following exposure to both non-24-h day-night cycles (post 24.65–post 23.5); large filled circle; group difference; error bars, SEM.