Date of Award


Document Type

Campus Access Dissertation


Epidemiology and Biostatistics



First Advisor

Burch, James B


Background: Circadian rhythm disruption, sleep abnormalities, and shift work have emerged as cancer risk factors. Genetic variation of clock genes, which aid in circadian rhythm control, may increase cancer susceptibility or lead to disruption of hormone secretion. A Period 3 (Per3) clock gene variable number tandem repeat (VNTR) has been associated with chronotype, sleep and mood disturbances, altered hormone or cytokine secretion, and increased premenopausal breast cancer (BrCA) risk. Methods: Two studies were conducted. One study based in India identified histologically confirmed BrCA cases that were frequency-matched on age (+ 5 years), geographic region, menopausal status, and time of interview (+ 3 months) to controls without a previous cancer diagnosis. Odds ratios (OR) comparing cases and controls were obtained using logistic regression to determine if cases had increased odds of either having a 4/5 or 5/5 Per3 VNTR genotype, an evening chronotype, or abnormal sleep characteristics. In the second study, a random sample of police officers from Buffalo, New York was used to determine if there were any differences in mean waking or diurnal cortisol secretion among those with different Per3 VNTR genotypes using analysis of covariance (ANCOVA) models. Results: The 4/5 or 5/5 Per3 variant was 30-40% more common among BrCA cases, although not statistically significant. Cases were more likely than controls to have morning or evening chronotype relative to those without a circadian preference. Among all women, the only sleep-related statistically significantly finding was that those who went to bed after to 11PM, on average, had reduced risk of BrCA compared to those who went to bed after 11PM. In the study of police officers, those with a 4/5 or 5/5 genotype had a greater waking and diurnal cortisol output compared to those with the 4/4 Per3 variant. Discussion: Circadian disruption and clock gene polymorphisms may have a profound effect on physiological health and disease susceptibility. Although the Per3 VNTR was not associated with BrCA in our study of Indian women, we provide evidence that the Per3 VNTR is associated with both waking and diurnal cortisol output. In addition, this project is the first to identify chronotype as a putative BrCA risk factor.


© 2012, Michael David Wirth