659,501 men in Catalonia: ED prevalence was 12.6% in T2DM vs 8.3% in non-diabetic men — a 1.5-fold difference peaking at ages 55–64.
Journal: Primary Care Diabetes | Published: 2026-04-17 | Type: Cross-sectional study | PMID: 42000296 Authors: Martínez-Mores S et al. (Hospital de la Santa Creu i Sant Pau, Barcelona; DAP-Cat group, Fundació Institut Universitari) Funding/COI: Funding not listed. No competing financial interests declared.
Using routinely collected primary care records from the SIDIAP database — covering roughly 80% of the Catalan population — this study documents a 1.5-fold higher ED prevalence in men with type 2 diabetes versus men without. At the population scale, the signal is unambiguous: cardiometabolic burden tracks directly with ED diagnosis and prescription rates. The authors argue ED is systematically under-screened in primary care despite its value as a cardiometabolic risk marker.
This is a large, real-world cross-sectional study drawing on administrative and clinical data from 2010–2023 — a 13.5-year window across roughly 4.8 million Catalan residents. The SIDIAP database is one of the most comprehensive primary care records systems in Europe, which gives this study genuine population-level representativeness. ED was defined via ICD-10 codes and/or ED-drug prescriptions, a dual-ascertainment approach that partially compensates for under-coding of a stigmatized condition.
The multivariable logistic regression model is appropriate for the design, and the authors extracted a wide array of covariates — sociodemographic, lifestyle, clinical, laboratory, comorbidities, and medications — allowing reasonable confounder adjustment. However, cross-sectional design cannot establish temporality: whether diabetes drives ED or shared cardiometabolic conditions drive both simultaneously cannot be determined here.
This is a well-executed population-level cross-sectional study that does exactly what it sets out to do: quantify ED burden in a diabetic primary care population at scale. The 1.5-fold prevalence elevation in T2DM is not surprising clinically, but 659,501 men with 13 years of records gives those numbers weight that small clinic studies cannot. The real finding worth attention is the systematic under-recognition argument — ED defined by diagnosis codes and prescriptions is almost certainly a floor estimate, and if ED functions as a cardiometabolic sentinel symptom, primary care routinely misses the window. The cross-sectional design limits this to association mapping, not causal inference. Read it as a prevalence benchmark with a healthcare-systems critique attached, not as mechanistic evidence.