The timing of the circadian clock, circadian period and chronotype varies among individuals. To date, not much is known about how these parameters vary over time in an individual. We performed an analysis of the following five common circadian clock and chronotype measures: 1) the dim light melatonin onset (DLMO, a measure of circadian phase), 2) phase angle of entrainment (the phase the circadian clock assumes within the 24-h day, measured here as the interval between DLMO and bedtime/dark onset), 3) free-running circadian period (tau) from an ultradian forced desynchrony protocol (tau influences circadian phase and phase angle of entrainment), 4) mid-sleep on work-free days (MSF from the Munich ChronoType Questionnaire; MCTQ) and 5) the score from the Morningness-Eveningness Questionnaire (MEQ). The first three are objective physiological measures, and the last two are measures of chronotype obtained from questionnaires. These data were collected from 18 individuals (10 men, eight women, ages 21-44 years) who participated in two studies with identical protocols for the first 10 days. We show how much these circadian rhythm and chronotype measures changed from the first to the second study. The time between the two studies ranged from 9 months to almost 3 years, depending on the individual. Since the full experiment required living in the laboratory for 14 days, participants were unemployed, had part-time jobs or were freelance workers with flexible hours. Thus, they did not have many constraints on their sleep schedules before the studies. The DLMO was measured on the first night in the lab, after free-sleeping at home and also after sleeping in the lab on fixed 8-h sleep schedules (loosely tailored to their sleep times before entering the laboratory) for four nights. Graphs with lines of unity (when the value from the first study is identical to the value from the second study) showed how much each variable changed from the first to the second study. The DLMO did not change more than 2 h from the first to the second study, except for two participants whose sleep schedules changed the most between studies, a change in sleep times of 3 h. Phase angle did not change by more than 2 h regardless of changes in the sleep schedule. Circadian period did not change more than 0.2 h, except for one participant. MSF did not change more than 1 h, except for two participants. MEQ did not change more than 10 points and the categories (e.g. M-type) did not change. Pearson's correlations for the DLMO between the first and second studies increased after participants slept in the lab on their individually timed fixed 8-h sleep schedules for four nights. A longer time between the two studies did not increase the difference between any of the variables from the first to the second study. This analysis shows that the circadian clock and chronotype measures were fairly reproducible, even after many months between the two studies.
