Innate and adaptive immune cell subsets as risk factors for coronary heart disease in two population-based cohorts

Abstract

Background and aims: Cell-mediated immunity is implicated in atherosclerosis. We evaluated whether innate and adaptive immune cell subsets in peripheral blood are risk factors for coronary heart disease.

Methods: A nested case-cohort study (n = 2155) was performed within the Multi-Ethnic Study of Atherosclerosis (MESA) and the Cardiovascular Health Study (CHS). Cases of incident myocardial infarction (MI) and incident angina (n = 880 total cases) were compared with a cohort random sample (n = 1275). Immune cell phenotypes (n = 34, including CD14⁺ monocytes, natural killer cells, γδ T cells, CD4⁺, CD8⁺ and CD19⁺ lymphocyte subsets) were measured from cryopreserved cells by flow cytometry. Cox proportional hazards models with adjustment for cardiovascular disease risk factors were used to evaluate associations of cell phenotypes with incident MI and a composite phenotype of incident MI or incident angina (MI-angina) over a median 9.3 years of follow-up. Th1, Th2, Th17, T regulatory (CD4⁺CD25⁺CD127^-), naive (CD4⁺CD45RA⁺), memory (CD4⁺CD45RO⁺), and CD4⁺CD28^- cells were specified as primary hypotheses. In secondary analyses, 27 additional cell phenotypes were investigated.

Results: After correction for multiple testing, there were no statistically significant associations of CD4⁺ naive, memory, CD28^-, or T helper cell subsets with MI or MI-angina in MESA, CHS, or combined-cohort meta analyses. Null associations were also observed for monocyte subsets, natural killer cells, γδ T cells, CD19⁺ B cell and differentiated CD4⁺ and CD8⁺ cell subsets.

Conclusions: The proportions of peripheral blood monocyte and lymphocyte subsets are not strongly related to the future occurrence of MI or angina in adults free of autoimmune disease.

Keywords: Biomarkers; Cardiovascular disease; Immunity; Inflammation; Risk factors; T cells.

Figure 1.. Associations of CD4⁺ T helper cell subsets specified as primary hypotheses with incident myocardial infarction.

Cell subsets were analyzed per 1-SD higher values using Cox proportional hazards models with sampling weights. Confidence intervals (CIs) used robust (sandwich) standard error estimates and reflect the Bonferroni-adjusted significance level of p<0.0071. Th1, Th2, and Th17 cells were not meta-analyzed due to differences in phenotyping (as described in the Methods). Models were adjusted for age, sex, race/ethnicity, education, clinical site, systolic blood pressure, use of antihypertensive medications, low-density lipoprotein cholesterol, use of statins, smoking, and diabetes.

Figure 2.. Associations of CD4⁺ T helper cell subsets specified as primary hypotheses with incident myocardial infarction or incident angina.

Cell subsets were analyzed per 1-SD higher values using Cox proportional hazards models with sampling weights. Confidence intervals (CIs) used robust (sandwich) standard error estimates and reflect the Bonferroni-adjusted significance level of p<0.0071. Th1, Th2, and Th17 cells were not meta-analyzed due to differences in phenotyping (as described in the Methods). Models were adjusted for age, sex, race/ethnicity, education, clinical site, systolic blood pressure, use of antihypertensive medications, low-density lipoprotein cholesterol, use of statins, smoking, and diabetes.

Figure 3.. Associations of immune cell subsets included as secondary hypotheses with incident myocardial infarction or incident angina

Cell subsets were analyzed per 1-SD higher values using Cox proportional hazards models with sampling weights and robust (sandwich) standard error estimates. Confidence intervals reflect the Bonferroni-adjusted significance level of p<0.0015. Models were adjusted for age, sex, race/ethnicity, education, clinical site, systolic blood pressure, use of statins, low-density lipoprotein cholesterol, use of antihypertensive medications, smoking, and diabetes.