Insulin-deficient diabetes subtype is linked to higher rates of kidney disease and all-cause mortality

Abstract Background Understanding diabetes heterogeneity across diverse populations is critical for advancing precision medicine. However, population-representative data on diabetes subphenotypes in South America are scarce. We examined Ahlqvist-derived diabetes clusters in nationally representative cohorts of Chilean adults and their associations with complications and mortality. Methods This complex-sample analysis included 308 and 536 adults with diabetes from the Chilean National Health Survey (2009–2010 and 2016–2017). Participants were classified as Mild Age-Related Diabetes (MARD), Mild Obesity-Related Diabetes (MOD), Severe Insulin-Resistant Diabetes (SIRD), or Severe Insulin-Deficient Diabetes (SIDD) using the approach of Bello-Chavolla et al. Self-reported complications and all-cause mortality were assessed, with up to 13 years of follow-up. Results Subtype distributions were consistent across survey waves, with increasing MOD prevalence over time. SIDD was characterized by longer diabetes duration and poor glycaemic control. It exhibited up to threefold higher albuminuria, also among recently diagnosed individuals. In unadjusted analyses, SIDD showed higher mortality risk (HR 4.87, 95% CI 1.87–12.70), whereas MOD demonstrated mortality comparable to normoglycaemic individuals. After multivariable adjustment, the association between SIDD and mortality was attenuated and no longer significant, while HbA1c emerged as a key determinant of mortality risk. Conclusions In contrast to findings in European populations, SIDD was associated with early kidney involvement and higher unadjusted mortality risk in Chile, whereas MOD represents a relatively lower-risk subtype. However, the mortality association was not independent of conventional risk factors, particularly HbA1c. These findings support the clinical utility of diabetes subphenotyping for risk stratification in public health settings, while underscoring the importance of population-specific validation.