Abstract
Background
Research on the link between hematological characteristics and cancer risk has gained significant attention. Traditional epidemiological and cell biology studies, have identified correlations between blood traits and cancer risks. These findings are important as they suggest potential risk factors and biological mechanisms. However, these studies often can't confirm causality, pointing to the need for further investigation to understand these relationships better.
Methods
Mendelian randomization (MR), utilizing single-nucleotide polymorphisms as instrumental variables, was employed to investigate hematological trait causal effects on cancer risk. Thirty-six hematological traits were analyzed, and their impact on 28 major cancer outcomes was assessed using data from the FinnGen cohort, with eight major cancer outcomes and 22 cancer subsets. Furthermore, 1,008 MR analyses were conducted, incorporating sensitivity analyses (weighted median, MR-Egger, and MR-PRESSO) to address potential pleiotropy and heterogeneity.
Results
The analysis (data from 173,480 individuals primarily of European descent) revealed significant results. An increase in eosinophil count was associated with a reduced risk of colorectal malignancies (OR 0.7702, 95% CI 0.6852, 0.8658; p = 1.22E-05). Similarly, an increase in total eosinophil and basophil count was linked to a decreased risk of colorectal malignancies (OR 0.7798, 95% CI 0.6904, 0.8808;p = 6.30E-05). Elevated hematocrit (HCT) levels were associated with a reduced risk of ovarian cancer (OR 0.5857, 95% CI 0.4443, 0.7721;p = 1.47E-04). No significant heterogeneity or horizontal pleiotropy was observed.
Conclusions
Our study highlights the complex and context-dependent roles of hematological characteristics in cancers.