Cluster analysis reveals distinct inflammatory phenotypes in cardiometabolic disease

Cardiovascular disease and type 2 diabetes mellitus share inflammatory pathways; however, their combined effect on systemic inflammation remains unclear. Although chronic low-grade inflammation is recognized in both conditions, the heterogeneity of inflammatory responses in patients with comorbidities has not been systematically characterized. This study aimed to identify distinct inflammatory phenotypes in patients with CVD, T2DM, or both and evaluate their clinical and biomarker profiles.

We conducted a cross-sectional study of 240 outpatients from Brazilian cardiology clinics, who were categorized into three groups: CVD + T2DM+ (n = 51), CVD + T2DM- (n = 95), and controls without either condition (n = 94). Serum levels of IL-6, IL-1β, TNF-α, and cardiac biomarkers (CK-MB, NT-proBNP, D-dimer) were measured using chemiluminescence immunoassays. Gaussian Mixture Models identified inflammatory clusters, which were validated using silhouette analysis. Statistical comparisons were performed using Kruskal-Wallis tests for continuous variables and chi-square tests for categorical variables, with ridge regression to assess phenotype predictors.

Analysis revealed three distinct inflammatory phenotypes: Phenotype 1 (n = 74) showed the lowest inflammatory markers (IL-6:2.00 pg/mL; CRP: 3.22 mg/L) and CVD prevalence (32.4%). Phenotype 2 (n = 83) had the highest CVD burden (81.9%) but unexpectedly low TNF-α (6.37 pg/mL) despite elevated D-dimer (0.76 µg/mL). Phenotype 3 (n = 83) exhibited the most severe inflammation (TNF-α: 12.90 pg/mL; hs-CRP: 5.00 mg/L) and the highest comorbidity rate (25.3% CVD + T2DM+). Phenotype 3 also showed significantly higher CK-MB (2.46 vs. 1.00 ng/mL, p < 0.001) and IL-8 (12.6 vs. 10.85 pg/mL, p < 0.001) levels than the other groups. Regression analysis identified age (OR = 2.24, 95% CI: 1.69-3.00) and CVD (OR = 1.47, 95% CI: 1.04-2.04) as significant phenotype predictors.

This study identified three clinically distinct inflammatory phenotypes in cardiometabolic diseases, with T2DM exacerbating systemic inflammation in patients with pre-existing CVD. The discordant phenotype (high CVD burden but low TNF-α levels) challenges current inflammatory models and suggests alternative pathways in some patients. These findings support the need for phenotype-specific approaches for risk stratification and targeted anti-inflammatory therapies in high-risk subgroups.