Figure 4.

The light and melatonin phase response curves re-referenced to habitual sleep timing. The light phase response curve is adapted from a phase response curve generated to a single 1 hour pulse of white broad spectrum bright light [55]. There is some debate as to whether humans are as sensitive to light during the day as this phase response curve suggests. The melatonin phase response curve is adapted from a phase response curve generated to 3 days of a daily dose of 0.5 mg of exogenous melatonin [96]. Accordingly, we have reduced the amplitude of the melatonin phase response curve by a factor of 3 to better estimate the effects of a single dose. This phase response curve may overestimate the phase shifting effects of exogenous melatonin as the phase shifts were measured in the absence of a fixed light-dark cycle. As discussed in the main text, phase response curves are useful general guides, but cannot be used to precisely predict phase shifts in circadian timing in individual patients.