Associations between population-level BCG and yellow fever vaccination and aortic mortality: A 29-year observational analysis using causal inference methods.

AIMS: Diseases of the aorta are driven by chronic inflammation and matrix metalloproteinase activity. Vaccines such as Yellow Fever (YF) and BCG may be associated with vascular protection through "trained immunity", epigenetic reprogramming of innate immune cells that attenuates systemic inflammatory responses. We investigated the association between population-level vaccination coverage and aortic mortality using contemporary causal inference methods.
METHODS AND RESULTS: This ecological study analyzed 27 Brazilian states over 29 years (1994-2022), encompassing 783 state-year observations and 185,429 deaths. Poisson fixed-effects models with state and year intercepts were employed to isolate within-state effects. States with high YF coverage (>80%) showed a 16% lower age-standardized aortic mortality compared to low-coverage states (IRR 0.84; 95% CI 0.76-0.93; P = 0.001). E-value analysis for this association was 1.67, indicating robustness against moderate unmeasured confounding. For BCG, associations followed a biologically plausible lag structure, peaking at 15 years post-vaccination (IRR 0.91 per 10 pp. increase; 95% CI 0.85-0.98; P = 0.009). Negative control analysis using Diphtheria-Tetanus-Pertussis (DTP) vaccine, which lacks trained immunity effects, showed no association with aortic mortality (IRR 1.02; P = 0.578), arguing against residual confounding from healthcare access. Causal mediation analysis suggested that 65% of the YF vaccine's association operated through pathways independent of infectious disease reduction.
CONCLUSION: Population-level vaccination with YF and BCG is associated with lower aortic mortality in this ecological analysis. These hypothesis-generating findings, supported by negative control validation and temporal consistency across three decades, are compatible with a trained immunity pathway but require confirmation in prospective individual-level studies before causal conclusions can be drawn.