ORIGINAL ARTICLE Is circadian type associated with sleep duration in twins? Nathaniel F WATSON, 1,2,3 Dedra BUCHWALD, 4 Carolyn NOONAN, 4 Michael V VITIELLO, 5,3 Allan I PACK 8 and Jack GOLDBERG 6,7 1 Department of Neurology, 2 University of Washington Medicine Sleep Center, 3 Center for Research on the Management of Sleep Disturbances, 4 Department of Medicine, 5 Department of Psychiatry and Behavioral Sciences, 6 Department of Epidemiology, University of Washington, 7 Vietnam Era Twin Registry, Veterans Affairs Epidemiologic Research and Information Center, Seattle, Washington, and 8 Department of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA Abstract We used the community-based University of Washington Twin Registry to investigate the genetic association between circadian type and sleep duration. Habitual sleep duration was obtained by self-report and circadian type was measured using a reduced 5-item Horne–Östberg Morningness– Eveningness questionnaire. Univariate and bivariate genetic analyses were fit using structural equation models. We used multinomial logistic regression to examine the overall and within-pair effects of circadian type on sleep duration in total and monozygotic (MZ) restricted twin samples. We surveyed 1620 twins, mostly female (70%), from same-sex pairs (1098 monozygotic, 522 dizygotic) with a mean age of 36 years (SD = 15). Among all twins, 23% were habitual short (<7 h/night), 68% normal (7–8 h/night), and 9% long (9 h/night) sleepers. Overall, 33% were morning-type, 52% neither-type, and 15% evening-type. The heritability of sleep duration was 31% and 40% for circadian type (both P < 0.05). The bivariate analysis did not reveal shared genetic influences on circadian type and sleep duration. When analyzing twins as individuals, eveningness was associated with short and long sleep duration in the total (P < 0.001) and MZ restricted (P < 0.05) twin samples. These findings were attenuated in the within-pair analysis in both samples suggesting that familial factors, namely genetics and shared environment, are in the pathway of interest or confound the association between circadian type and sleep duration. Because both short and long sleep is associated with adverse health outcomes, our findings suggest circadian evening-type may represent an endophe- notype for poor health. Key words: circadian, dizygotic, genetics, monozygotic, sleep duration, twins. INTRODUCTION The circadian clock is an interrelated network of tran- scriptional, translational autoregulatory loops that oscillate through a 24-h period. 1 These molecular oscil- lations define circadian rhythms and influence behav- iors such as alertness, cognitive performance, and sleep ability. The nature of these behaviors defines an indi- vidual’s circadian type or preference for morning or evening activity. Fidelity of circadian rhythms is crucial to maintaining physiological homeostasis and disrup- tion of this rhythm, through shift work or other behav- iors, can result in gastrointestinal, cardiovascular, and carcinogenic pathology. 2–4 Numerous genes, including Dec2, BMAL1, NPAS2/CLOCK, Per, Cry, and CK1s, Correspondence: Dr Nathaniel F Watson, University of Washington Medicine Sleep Center, Box 359803, 325 Ninth Avenue, Seattle, WA 98104-2499, USA. Email: nwatson@uw.edu Accepted 17 September 2011. Sleep and Biological Rhythms 2012; 10: 61–68 doi:10.1111/j.1479-8425.2011.00526.x 61 © 2011 The Authors Sleep and Biological Rhythms © 2011 Japanese Society of Sleep Research