Delayed sleep phase cases and controls

Abstract

Background: Delayed sleep phase disorder (DSPD) is a condition in which patients have difficulty falling asleep before the early morning hours and commonly have trouble awakening before late morning or even early afternoon. Several studies have suggested that variations in habitual bedtime are 40-50% heritable.

Methods: We recruited a case series of 205 participants, along with 221 controls (DSPD-C) with normal sleep, roughly matched for age, gender, and ancestry. A representative sample of San Diego adults recruited some years before was already available to confirm the control group. Both DSPD and DSPD-C provided blood or saliva samples for DNA and completed extensive questionnaires about sleep habits, sleep history, family history, sleep quality, morningness-eveningness traits, depression, mania, and seasonality of symptoms. The DSPD group wore wrist actigraphs for a median of 13.2 days. The representative sample collected previously had undergone actigraphic recordings, from which 48 hours of data were generally available.

Results: The DSPD and DSPD-C samples showed almost no overlap on morningness-eveningness scores. DSPD cases went to bed and arose about 3 hours later than the DSPD-C and the representative sample. DSPD cases reported more difficulties with sleep, poorer sleep quality, and more depression, but there was no significant difference in a history of mania. DSPD cases reported more family history of late bedtimes, but female DSPD reported that their fathers' bedtimes were later than the fathers of male DSPD.

Conclusion: These results indicate a DSPD phenotype is familial and associated with unipolar depression.

Figure 1

Distribution of morningness-eveningness scores. The percentiles of Horne-Östberg morningness-eveningness scores are plotted for 5 groups: a) absolutely certain DSPD 1, b) fairly certain DSPD 2, b) questionable DSPD 3, d) unlikely or very doubtful DSPD 4–5, and e) DSPD-C matched controls. Low HO scores indicate eveningness and high scores indicate greater morningness. In this study, the best criterion separating DSPD and DSPD-C was a score of 41.

Figure 2

Distribution of actigraphic sleep acrophases. The percentiles with an actigraphic sleep acrophase at or below each time are plotted versus clock hours after midnight. The percentiles are plotted for 5 groups: a) absolutely certain DSPD 1, b) fairly certain DSPD 2, c) questionable DSPD 3, d) unlikely or very doubtful DSPD 4–5, and e) a representative population sample of San Diego adults.

Figure 3

Distribution of bedtimes. The percentiles with a reported bedtime (week prior to completing the questionnaire) at or below each time are plotted versus clock hours after midnight. The percentiles are plotted for 6 groups: a) absolutely certain DSPD 1, b) fairly certain DSPD 2, c) questionable DSPD 3, d) unlikely or very doubtful DSPD 4–5, e) the DSPD-C, and f) a representative sample of San Diego adults.

Figure 4

Distribution of wake-up times. The percentiles with a reported wake-up times (week prior to completing the questionnaire) at or below each time are plotted versus clock hours after midnight. The percentiles are plotted for 6 groups: a) absolutely certain DSPD 1, b) fairly certain DSPD 2, c) questionable DSPD 3, d) unlikely or very doubtful DSPD 4–5, e) the DSPD-C, and f) a representative sample of San Diego adults.

Figure 5

Self reported comparison of bedtimes with other people. Self-reported bedtimes (in clock time after midnight) are reported for participants who stated that on average (after age 21) they went to sleep much earlier, somewhat earlier, about the same time, somewhat later, or much later that other people their age. Thick horizontal lines are medians. Red boxes show the 2^nd and 3^rd quartiles. Bars show the upper and lower 5%.