Abstract

BACKGROUND: Limited evidence suggests elevated risks of cardiovascular disease (CVD) among people diagnosed with tuberculosis (TB) disease, though studies have not adjusted for preexisting CVD risk. We carried out a cohort study using 2 separate datasets, estimating CVD incidence in people with TB versus those without. METHODS: Using data from the United States (Veterans Health Administration) and the United Kingdom (Clinical Practice Research Datalink) for 2000-2020, we matched adults with incident TB disease and no CVD history 2 years before TB diagnosis (US, n = 2121; UK, n = 15 820) with up to 10 people without TB on the basis of age, sex, race/ethnicity and healthcare practice. Participants were followed beginning 2 years before TB diagnosis and for 2 years subsequently. The acute period was defined as 3 months before/after TB diagnosis. TB, CVD, and covariates were identified from electronic routinely collected data (primary and secondary care; mortality). Poisson models estimated incident rate ratios for CVD events in people with TB compared to those without. RESULTS: CVD incidence was consistently higher in people with TB, including during the baseline period (pre-TB) and particularly in the acute period: incident rate ratios were US, 3.5 (95% confidence interval, 2.7-4.4), and UK, 2.7 (2.2-3.3). Rate ratios remained high after adjusting for differences in preexisting CVD risk: US, 3.2 (2.2-4.4); UK, 1.6 (1.2-2.1). CONCLUSIONS: Increased CVD incidence was observed in people with TB versus those without, especially within months of TB diagnosis, persistent after adjustment for differences in preexisting risk. Enhancing CVD screening and risk management may improve long-term outcomes in people with TB.