Figure 1. Study design of the univariable and multivariable Mendelian randomization analyses of the associations between six leukocyte traits and five digestive system cancers in East Asian and European cohorts. CRC: colorectal cancer; ESCA: esophageal cancer; GC: gastric cancer; LC: liver cancer; PC: pancreatic cancer.
Figure 2. Associations between genetically six predicted circulating leukocyte traits and five digestive system cancers in East Asians and Europeans, based on the univariable inverse variance weighted analysis. Odds ratio (OR) and 95% confidence interval (CI) represent the change in OR of cancers per 1 standard deviation (SD) increase in each leukocyte subtype in the blood.
Figure 3. The effect of each leukocyte trait on the risk of gastric cancer and colorectal cancer. (a) The effect estimators of blood basophil count on the risk of gastric cancer in East Asians, based on three univariable MR methods (IVW, MR-Egger, and WM). (b) Leave-one-out sensitivity analysis for basophil count on gastric cancer in Asians. (c) Multivariable MR estimators representing the effect of each leukocyte count on the risk of gastric cancer in East Asians (left), colorectal cancer in East Asians (middle), and colorectal cancer in Europeans (right). (d) The effect estimators of blood eosinophil count on the risk of colorectal cancer in East Asians, based on three univariable MR methods (IVW, MR-Egger, and WM). (e) Leave-one-out sensitivity analysis for eosinophil count in colorectal cancer in Asians. (f) The effect estimators of the blood eosinophil count on the risk of colorectal cancer in Europeans, based on three univariable MR methods (IVW, MR-Egger, and WM). (g) Leave-one-out sensitivity analysis for eosinophil count in colorectal cancer in Europeans. Odds ratio (OR) and 95% confidence interval (CI) represent the change in OR of cancers per 1 standard deviation (SD) increase in each blood leukocyte subtype. IVW: inverse variance weighted; MR: Mendelian randomization; WM: weighted median.