ObjectivesYKL-40 is an acute phase protein elevated in patients with infectious and inflammatory diseases. We tested the hypothesis that baseline elevated YKL-40 is associated with increased risk of future infectious disease in healthy individuals in the general population. MethodsWe prospectively followed 94 665 individuals from the Danish general population for up to 23 years and analysed for plasma YKL-40 levels (n = 21 584) and CHI3L1 rs4950928 genotype (n = 94 184). Endpoints were any infection, bacterial pneumonia, urinary tract infection, skin infection, sepsis, diarrhoeal disease, and other infections. ResultsFor YKL-40 percentile category 91–100% versus 0–33%, the multifactorially and C-reactive protein (CRP) adjusted hazard ratios were 1.71 (95% confidence interval 1.50–1.96; p 4 × 10−14) for any infection, 1.97 (1.64–2.37; p 4 × 10−13) for bacterial pneumonia, 1.62 (1.24–2.11; p 0.002) for urinary tract infection, 1.74 (1.31–2.32; p 2 × 10−4) for skin infection, 1.76 (1.25–2.46; p 0.004) for sepsis, 1.90 (1.29–2.78; p 0.002) for diarrhoeal disease and 2.71 (1.38–5.35; p 0.01) for other infections. In multifactorially and CRP-adjusted models, a twofold increase in YKL-40 was associated with increased risk of all infectious disease endpoints. Mendelian randomization did not support causality, as CHI3L1 rs4950928 was associated with 94% and 190% higher YKL-40 levels (for CG and CC versus GG genotype), but not with increased risk of any infectious disease endpoint. DiscussionBaseline elevated plasma YKL-40 was not a cause but a strong marker of increased risk of future infectious diseases in individuals in the general population.