Background
Observational studies have identified associations between haematological traits and type-2 diabetes mellitus (T2D). However, it is difficult to infer causal effects due to the potential of confounding. Our study utilizes the Mendelian randomization (MR) approach to address the above limitation and investigate the causal effect of haematological traits (HT) such as white blood cell (WBC) count, platelet count, haemoglobin (Hb) levels, and red blood cell (RBC) count, on T2D in individuals of African ancestry.
Methods
The participating cohorts included participants of African ancestry in the Blood Cell consortium (n = 15,171) and the Million Veteran Program dataset (N=53,445: Ncases =23,305; Ncontrol = 30,140). We applied a univariable, bidirectional, and multivariable Two-sample MR to estimate the causal relationship between haematological traits and T2D using GWAS summary statistics. Genetic instruments for haematological traits and T2D included only the variants that attained a genome-wide significant level (P ≤ 5 × 10−8) of association with the respective exposure. The inverse-variance weighted MR approach was used in the main analysis, complemented by sensitivity analyses that are more robust to the inclusion of pleiotropic variants, including MR-Egger and MR-PRESSO. We further evaluated the genetic correlation between the hematological traits and T2D using linkage disequilibrium score regression (LDSC).
Results
In the main inverse-variance weighted (IVW) MR estimates, genetically predicted mean corpuscular hemoglobin concentration (MCHC), mean corpuscular hemoglobin (MCH) and mean corpuscular volume (MCV) levels were associated with decreased risk of T2D. Genetically predicted higher RBC was significantly associated with an increased risk of T2D. For WBC traits, there was evidence of protective effects of total WBC count and neutrophil count (NEU) on T2D. In the reverse analysis, genetically predicted T2D liability was associated with lower levels of RBC distribution width (RDW) and elevated levels of lymphocyte count (LYM). The multivariable analysis supported direct effects of genetically predicted MCH, MCHC, RDW, MCV, and lower levels of WBC with decreased risk of T2D. Additionally, LDSC showed negative genetic correlation of T2D with haematocrit (HCT) and haemoglobin concentration (HGB), and positive with eosinophil (EOS) and platelet count (PLT), suggesting that there is a genetic basis for the association between T2D and these blood traits.
Conclusion
These findings indicate that haematological traits may differentially play a role in the development of T2D and be affected by T2D. However, further research is needed to validate and explore the biological pathways and mechanisms involved in these associations.
