Supervisors info:
Κωνσταντίνος Τσιλίδης, Αναπληρωτής Καθηγητής, Ιατρική Σχολή, Πανεπιστήμιο Ιωαννίνων
Ευάγγελος Ευαγγέλου, Αναπληρωτής Καθηγητής, Ιατρική Σχολή, Πανεπιστήμιο Ιωαννίνων
Ιωάννα Τζουλάκη, Επίκουρη Καθηγήτρια, Ιατρική Σχολή, Πανεπιστήμιο Ιωαννίνων
Summary:
Introduction:
Over the past decades, the global patterns of incidence and mortality of the various tumour types have undergone some rapid changes, largely due to the variation in the distribution of inherent to modern lifestyle cancer risk factors. The transition from transmitted to human degenerative diseases, places cancers of the colorectum third in incidence and second in mortality. Amongst the changes taking place in developing societies, shift work that causes the disruption of the natural endogenous circadian rhythm has been nominated as a potential carcinogenic. Nevertheless, the published epidemiological evidence for the association between shift work and colorectal cancer is limited and often contradictory.
The purpose of the current study was to overcome the limitations of observational studies and to investigate whether the genetically determined sleep characteristics have a causal effect on the risk of overall colon cancer, as well as by patient sex and according to cancer anatomical distribution (colon, distant colon, proximal colon, rectal cancer).
Methods:
A two-sample MR study was conducted on summary data and the subsequent sensitivity analysis were implemented to assess potential violation of the instrumental variable assumptions using 341 single nucleotide polymorphisms (SNPs) associated with chronotype, 56 SNPs associated with insomnia symptoms and 90 SNPs associated with sleep duration. Summary data for the association between the selected SNPs and the risk of colorectal cancer in 125,478 participants (58,131 cases και 67,347 controls) were retrieved from GECCO, while summary data for the association between the SNPs and the sleep characteristics were retrieved from published GWAS in 651,295 participants for chronotype (UK Biobank and 23andMe), 254,767 for frequent insomnia symptoms, 470,528 for any insomnia symptoms (UK Biobank, HUNT and Partners Biobank) and 493,298 for sleep duration (CHARGE and UK Biobank).
Results:
Morning preference in men was associated with a 13.6% lower risk of overall colorectal cancer (IVW random effects OR: 1.005, 95% CI: 0.956, 1.057). This finding remained robust after the exclusion of the outlying pleiotropic variants from the analysis and was confirmed by the rest of the sensitivity analysis methods. Respectively, the effect of chronotype on colorectal cancer in women was not statistically significant (IVW random effects OR: 1.005, 95% CI: 0.956, 1.057). From both the analysis of the overall colorectal cancer and the stratified data analysis, there was little evidence for the causal effect of insomnia and sleep duration on the risk of colorectal cancer. Several statistically significant associations emerged between the genetic instruments and a number of secondary traits via Phenoscanner search, including smoking, alcohol consumption, BMI, diabetes, physical inactivity, inflammatory bowel disease and insulin resistance, which may be associated with the risk of colorectal cancer and be responsible for the induction of horizontal pleiotropy.
Conclusions:
There was consistent evidence for a protective effect of morning preference on the risk of colorectal cancer. However, further studies on independent samples are required to validate this finding, as well as the triangulation of evidence stemming from other fields of study (epidemiological observational studies, clinical trials, preclinical studies, genetic engineering studies, candidate gene studies) in order to further investigate the underlying biological mechanisms of the colorectal carcinogenesis causal pathway.
Keywords:
Mendelian randomization, Chronotype, Insomnia, Sleep duration, Colorectal cancer
