Circulating CD38 shows stronger association with incident heart failure in women despite lower levels: sex differences and kidney-function context.

backgroundCD38, an NADase implicated in aging and metabolic stress, promotes cardiac dysfunction via NAD⁺ depletion. Whether circulating CD38 is associated with incident heart failure (HF), whether this association differs by sex, and how it relates to kidney-function context remain unclear.

methodsWe measured plasma CD38 in 42,584 UK Biobank participants (1,659 incident HF events) and evaluated its association with HF using sequential Cox models, attenuation analyses, competing-risk models, and incremental prediction metrics. Complementary mouse studies employed chronic β-adrenergic stress (isoproterenol) to investigate sex-specific responses and test the effects of CD38 inhibition (78c), TGF-β1 inhibition (SB431542), and their combination.

resultsCD38 was associated with incident HF in models adjusting for traditional risk factors (M1: HR per SD 1.25, 95% CI 1.18-1.32), with complete attenuation after eGFR adjustment. Despite higher CD38 levels in men, the association was stronger in women (M1: HR 1.39 vs. 1.17 per SD; sex interaction P = 0.004). After eGFR adjustment, the association remained positive in women (M3: HR 1.17) but became inverse in men (M3: HR 0.90). The CD38-HF association showed substantial attenuation after adjustment for NT-proBNP, ANGPT2, and IL-6, consistent with shared statistical variance with downstream biomarker burden rather than causal mediation. Fine-Gray competing-risk models (3,212 deaths without HF) produced similar results. In mice, isoproterenol induced cardiac remodeling and systolic impairment in both sexes and elicited greater metabolic and renal stress responses in females despite their higher baseline cardiac NAD⁺/ATP levels. CD38 inhibition restored cardiac NAD⁺/ATP and reduced renal stress markers, with numerically greater reductions in females, while TGF-β1 inhibition reduced fibrotic signaling. Combined inhibition produced complementary effects across metabolic and fibrotic readouts.

conclusionsCirculating CD38 is not an independent predictor of HF after accounting for kidney function and downstream biomarker burden, but appears to reflect a broader cardiometabolic-renal stress context. Despite lower circulating levels, CD38 showed a stronger association with HF in women than in men. Proof-of-concept mouse experiments showed that CD38 inhibition preserved cardiac NAD⁺/ATP and reduced renal stress markers, while TGF-β pathway inhibition reduced fibrotic signaling, providing mechanistic support for further sex-aware HF research.