Background: Simple prediction scores could help identify adults at high risk for diabetes.
Objective: To derive and validate scoring systems by using longitudinal data from a study that repeatedly tested for incident diabetes.
Design: Prospective cohort, divided into derivation and validation samples.
Setting: The ARIC (Atherosclerosis Risk in Communities) study, which followed participants for 14.9 years beginning in 1987 to 1989.
Participants: 12 729 U.S. adults (baseline age, 45 to 64 years; 22.8% black). Follow-up was 96.1% at 5 years and 72.2% at 10 years.
Measurements: Anthropometry, blood pressure, and pulse (basic system) plus a fasting blood specimen assayed for common analytes (enhanced system). Diabetes was identified in 18.9% of participants. Risk score integer points were derived from proportional hazard coefficients associated with baseline categorical variables and quintiles of continuous variables.
Results: The basic scoring system included waist circumference (10 to 35 points); maternal diabetes (13 points); hypertension (11 points); and paternal diabetes, short stature, black race, age 55 years or older, increased weight, rapid pulse, and smoking history (< or =8 points each). The enhanced system included glucose (6 to 28 points); waist circumference (5 to 21 points); maternal diabetes (8 points); and triglycerides, black race, paternal diabetes, low high-density lipoprotein cholesterol concentration, short stature, high uric acid, age 55 years or older, hypertension, rapid pulse, and nonuse of alcohol (< or =7 points each). When applied to the validation sample, ascending quintiles of the basic system were associated with a 10-year incidence of diabetes of 5.3%, 8.7%, 15.5%, 24.5%, and 33.0%, respectively. Quintiles of the enhanced system were associated with a 10-year incidence of 3.5%, 6.4%, 11.5%, 19.3%, and 46.1%.
Limitations: The risk scoring systems had no question regarding previous gestational diabetes, and knowledge of parental diabetes may be uncertain. The analyzed cohort was restricted by age and race; the systems may be less effective in other samples.
Conclusion: Basic information identified adults at high risk for diabetes. Additional data from fasting blood tests better identified those at extreme risk.