Abstract

Aim: (i) to identify patterns of chronic diseases using latent class analyses (LCA) and (ii) to examine their association with all-cause mortality in the Mexico City Prospective Study. Materials and Methods: 154 369 participants (67.2% women) were included in this prospective analysis. From 17 self-reported chronic conditions, four clusters were identified following a LCA approach: (a) Cluster 0 (63.3%), healthy group (reference group for the analyses); (b) Cluster 1 (7.5%), other chronic diseases: high prevalence of asthma, peptic ulcer, and peripheral artery disease; (c) Cluster 2 (21.0%), hypertension and some diabetes; (d) Cluster 3 (8.2%), mainly diabetes as the principal condition. Associations between the clusters and all-cause mortality were investigated using Cox proportional hazard models adjusted for confounding factors. Sensitivity analyses were performed by sex and fulfilling the glycated haemoglobin (HbA1c) target for the cluster with the highest mortality risk. Results: Over a median follow-up of 20 years (IQR: 18.6–21.3), 30 286 participants died. After adjusting, Cluster 3 (the diabetic cluster), where those with the highest mortality risk (hazard ratio [HR]: 2.60 [95% CI: 2.51–2.68]), followed by Cluster 2 (HR: 1.72 [1.67–1.77]) and Cluster 1 (HR: 1.26 [1.20–1.31]). Sensitivity analyses revealed that participants who did not achieve the HbA1c target, as well as men in Cluster 3, had the highest mortality risk. Conclusion: Even if we identified higher all-cause mortality across the three clusters, our findings highlighted the disproportionate burden of diabetes, particularly among men, as well as the urgent need for targeted and culturally sensitive interventions. © 2025 John Wiley & Sons Ltd.