[The Framingham function overestimates stroke risk for diabetes and metabolic syndrome among Spanish population].

OBJECTIVE To estimate stroke risk for diabetes, isolated or associated to metabolic syndrome (MS) according WHO and National Cholesterol Education Program (NCEP) criteria. DESIGN Multicentre and prospective cohort study. SETTING Primary health care. PARTICIPANTS Subjects between 55-85 years-old without any evidence of stroke, included from 1998 in a random population sample for MS follow-up during routine practice in Reus (Tarragona, Spain). MAIN MEASUREMENTS Stroke risk was achieved using Framingham function by means of a computerized algorithm using a diagnostic factorial design (diabetes and/or MS). Theoretical stroke risk and cumulated incidence of stroke events (1998-2003) were compared. RESULTS Among 728 subjects (412 women, mean age =66 years old, body mass index =29 kg/m2), 457 (62.8%) did not have diabetes, nor MS, 93 (12.8%) had MS without diabetes, 72 (9.9%) diabetes without MS, and 106 (14.5%) presented both conditions (WHO rules). According NCEP criteria were 60.7%, 14.8%, 7.8%, and 16.7%, respectively. Ten-year estimated stroke risk accounted for (WHO/NCEP) 8.4/9.1%, 10.8/10.5%, 18/17.3%, and 18.8/19.1%. Cumulated incidence for stroke events were: 2.8%, 1.4%, 5.4%, and 3.8% (WHO), and 2.5%, 2.8%, 3.5%, and 5.8%, respectively (NCEP). CONCLUSIONS Stroke risk scores were extremely increased among diabetic subjects irrespective to MS diagnose. The Framingham function probably overestimates stroke risk among Spanish individuals.