White Blood Cells and Blood Pressure: A Mendelian Randomization Study

Abstract

Background: High blood pressure (BP) is a risk factor for cardiovascular morbidity and mortality. While BP is regulated by the function of kidney, vasculature, and sympathetic nervous system, recent experimental data suggest that immune cells may play a role in hypertension.

Methods: We studied the relationship between major white blood cell types and blood pressure in the UK Biobank population and used Mendelian randomization (MR) analyses using the ≈750 000 UK-Biobank/International Consortium of Blood Pressure-Genome-Wide Association Studies to examine which leukocyte populations may be causally linked to BP.

Results: A positive association between quintiles of lymphocyte, monocyte, and neutrophil counts, and increased systolic BP, diastolic BP, and pulse pressure was observed (eg, adjusted systolic BP mean±SE for 1st versus 5th quintile respectively: 140.13±0.08 versus 141.62±0.07 mm Hg for lymphocyte, 139.51±0.08 versus 141.84±0.07 mm Hg for monocyte, and 137.96±0.08 versus 142.71±0.07 mm Hg for neutrophil counts; all P<10^-50). Using 121 single nucleotide polymorphisms in MR, implemented through the inverse-variance weighted approach, we identified a potential causal relationship of lymphocyte count with systolic BP and diastolic BP (causal estimates: 0.69 [95% CI, 0.19-1.20] and 0.56 [95% CI, 0.23-0.90] of mm Hg per 1 SD genetically elevated lymphocyte count, respectively), which was directionally concordant to the observational findings. These inverse-variance weighted estimates were consistent with other robust MR methods. The exclusion of rs3184504 SNP in the SH2B3 locus attenuated the magnitude of the signal in some of the MR analyses. MR in the reverse direction found evidence of positive effects of BP indices on counts of monocytes, neutrophils, and eosinophils but not lymphocytes or basophils. Subsequent MR testing of lymphocyte count in the context of genetic correlation with renal function or resting and postexercise heart rate demonstrated a positive association of lymphocyte count with urine albumin-to-creatinine ratio.

Conclusions: Observational and genetic analyses demonstrate a concordant, positive and potentially causal relationship of lymphocyte count with systolic BP and diastolic BP.

Keywords: Mendelian randomization analysis; blood pressure; hypertension; white blood cells.

Figure 1.

Levels of 5 white blood cell types are associated with systolic blood pressure (SBP), diastolic blood pressure (DBP), and pulse pressure (PP) in the UK Biobank. Estimated marginal means of blood pressure indices, from general linear model analysis adjusted for sex, age, age squared, body mass index, smoking status, and alcohol intake frequency, are presented according to quintiles of counts of white blood cell subpopulations. All ANOVA tests, assessing global between-quintile differences in blood pressure indices were significant at P<10^–11. Post hoc tests revealed that all comparisons between the 1st and the 5th quintile of any cell type count with respect to any blood pressure index were significant at Bonferroni-corrected P<0.05, given 150 tests (5 types of blood cell counts×3 blood pressure indices×10 between-quintile differences) performed.

Figure 2.

Mendelian randomization (MR) analyses testing the effects of 5 white blood cell subpopulation counts on systolic blood pressure (SBP), diastolic blood pressure (DBP), and pulse pressure (PP). Results obtained using 4 MR methods (inverse-variance weighted [IVW], Mendelian randomization-Egger [MR-Egger], weighted median, and MR-PRESSO [Mendelian Randomization Pleiotropy Residual Sum and Outlier]) are presented as a heat map representing causal estimates (1 SD of BP index per 1 SD of cell count). BP indicates blood pressure. *False discovery rate P<0.05 for a particular MR approach.

Figure 3.

Mendelian randomization (MR) analyses testing effect of lymphocyte or eosinophil counts on systolic blood pressure (SBP) and diastolic blood pressure (DBP) levels, before and after exclusion of a single variant from the SH2B3/ATXN2 (SH2B adaptor protein 3/ataxin 2) locus. Results of 4 MR methods (inverse-variance weighted [IVW], MR-Egger, weighted median, and MR-PRESSO [Mendelian Randomization Pleiotropy Residual Sum and Outlier]) are presented as causal estimates with 95% CIs.

Figure 4.

Reverse Mendelian randomization (MR) analyses testing the effects of systolic blood pressure (SBP), diastolic blood pressure (DBP), and pulse pressure (PP) on cell counts of white blood cell subpopulations. Results obtained using 4 MR methods (inverse-variance weighted [IVW], MR-Egger, weighted median, and MR-PRESSO [Mendelian Randomization Pleiotropy Residual Sum and Outlier]) are presented as a heat map representing causal estimates (1 SD cell count per 1 SD blood pressure [BP] index).*False discovery rate P<0.05 for a particular MR approach.