Segar MW, Khan MS, Patel KV, et al. - Among community-based people with dysglycemia, an accurate prediction of incident heart failure (HF) risk can be achieved by using the WATCH-DM risk score, a clinical risk score. Greater improvement in the HF risk prediction performance was achieved by adding natriuretic peptide (NP) levels, in those with low/intermediate risk vs those with high risk.

• Participants (n= 8,938) were adults with diabetes/pre-diabetes free of HF at baseline from four cohort studies (ARIC, CHS, FHS, and MESA); 5-year risk of incident HF was estimated using machine learning- [WATCH-DM(ml)] and integer-based [WATCH-DM(i)] scores.

• HF occurred in 4.8% within 5 years, and both above-mentioned scores showed high discrimination for predicting HF risk in people with dysglycemia (C-indices = 0.80 and 0.71, respectively), with no evidence of miscalibration (Greenwood–Nam–D'Agostino P ≥0.10).

• Increased NP levels alone had a significantly higher C-index, for predicting incident HF, in those with low/intermediate (<13) vs high (≥13) WATCH-DM(i) scores, among persons with dysglycemia [0.71 vs 0.64].

• On integrating NP levels with the WATCH-DM(i) score, HF risk discrimination improvement and net reclassification improvement (NRI) differed across the spectrum of risk with greater improvement noted at low/intermediate risk [WATCH-DM(i) <13] vs high risk [WATCH-DM(i) ≥13] (C-index = 0.73 vs 0.71; NRI = 0.45 vs 0.17).