**Background:** Subjects with type 2 diabetes mellitus (T2D) are at increased risk for heart failure (HF), and the prognosis of those with both conditions is poor. The fatty acid binding proteins FABP3 (cardiac-specific) and FABP4 (adipose-tissue-specific) have been linked to metabolic and cardiac disturbances, including HF, and have been associated with mortality in various populations. However, their prognostic value specifically in ambulatory chronic HF (CHF) patients with T2D had not been explored.
**Methods:** This prospective study included 240 ambulatory CHF patients from a structured multidisciplinary HF unit. Baseline serum samples were collected and FABP3 and FABP4 concentrations were measured by sandwich ELISA. Patients were followed for a mean of 5.78 ± 3.30 years. Primary endpoints were all-cause and cardiovascular (CV) death; a composite endpoint of CV death or HF hospitalization was a secondary endpoint. Multivariable Cox regression and Fine-Gray competing risks models were used, adjusting for age, sex, diabetes, ischemic etiology, NYHA functional class, time of evolution, LVEF, NT-proBNP, and obesity.
**Key Results:** T2D was present in 114 patients (47.5%). Subjects with T2D had higher mortality rates (69.30% vs. 50.79%, p = 0.004) and higher serum FABP3 (1829.3 [1104.9–3440.5] pg/mL vs. 1396.05 [820.3–2362.16] pg/mL, p = 0.007) and FABP4 (45.5 [27.6–79.8] ng/mL vs. 34.1 [24.09–55.3] ng/mL, p = 0.006) compared to non-T2D subjects. During follow-up, 143 patients died. In the whole cohort, FABP3 predicted all-cause death and both FABP3 and FABP4 predicted CV mortality. However, in subgroup analyses, the predictive values were only significant in T2D patients: for all-cause mortality, FABP3 (HR 1.25, 95% CI 1.09–1.44, p = 0.002) and FABP4 (HR 2.21, 95% CI 1.12–4.36, p = 0.023); for CV mortality, FABP3 (HR 1.28, 95% CI 1.09–1.50, p = 0.002) and FABP4 (HR 4.19, 95% CI 2.21–7.95, p < 0.001). FABP4 also predicted the composite endpoint (HR 2.07, 95% CI 1.11–3.87, p = 0.022) in T2D patients, but FABP3 did not. Neither FABP predicted outcomes in non-T2D patients. All-cause and CV mortality rates increased across tertiles of both FABP3 and FABP4 in T2D patients.
**Clinical Implications:** This study provides the first evidence that both FABP3 and FABP4 are independent predictors of all-cause and CV mortality specifically in ambulatory CHF patients with T2D, but not in those without T2D. FABP4 additionally predicts the composite endpoint of CV death or HF hospitalization. These findings suggest that FABP3 and FABP4 could serve as clinically useful biomarkers for risk stratification in this high-risk population. The study is limited by its single-center design, relatively small sample size, and retrospective nature, which preclude causal inference. Further research is needed to validate these findings and elucidate the underlying mechanisms, particularly regarding oxidative stress pathways.