The blood microbiome and its association to cardiovascular disease mortality: case-cohort study

Abstract

Background: Little is known about the association between bacterial DNA in human blood and the risk of cardiovascular disease (CVD) mortality. Methods: A case-cohort study was performed based on a 9 ½ year follow-up of the Oslo II study from 2000. Eligible for this analysis were men born in 1923 and from 1926 to 1932. The cases were men (n = 227) who had died from CVD, and the controls were randomly selected participants from the same cohort (n = 178). Analysis of the bacterial microbiome was performed on stored frozen blood samples for both cases and controls. Association analyses for CVD mortality were performed by Cox proportional hazard regression adapted to the case-cohort design. We used the Bonferroni correction due to the many bacterial genera that were identified. Results: Bacterial DNA was identified in 372 (82%) of the blood samples and included 78 bacterial genera from six phyla. Three genera were significantly associated with CVD mortality. The genera Kocuria (adjusted hazard ratio (HR) 8.50, 95% confidence interval (CI) (4.05, 17.84)) and Enhydrobacter (HR 3.30 (2.01, 5.57)) indicate an association with CVD mortality with increasing levels. The genera Paracoccus (HR 0.29 (0.15, 0.57)) was inversely related. Significant predictors of CVD mortality were: the feeling of bad health; and the consumption of more than three cups of coffee per day. The following registered factors were borderline significant, namely: a history of heart failure; increased systolic blood pressure; and currently taking antihypertensive drugs now, versus previously. Conclusions: The increasing levels of two bacterial genera Kocuria (skin and oral) and Enhydrobacter (skin) and low levels of Paracoccus (soil) were associated with CVD mortality independent of known risk factors for CVD.