There are few data on the prevalence of type 1 diabetes mellitus^1,2 and no estimates for the entire US population. The National Health and Nutrition Examination Survey (NHANES) is a representative cross-sectional survey of the civilian, noninstitutionalized US population. Although NHANES does not explicitly collect information on type 1 diabetes mellitus, we estimated the prevalence based on age of diabetes diagnosis, the age of insulin initiation, and current use of insulin.³

The protocol for the 1999–2010 NHANES was approved by the National Center for Health Statistics of the Centers for Disease Control and Prevention Ethics Review Board. All participants gave written informed consent. After excluding 2955 infants for whom diabetes data were not collected and 75 children and adults who were missing diabetes data, 59,130 participants remained in the analytic sample.⁴ We considered participants to have type 1 diabetes mellitus if they started insulin within 1 year of diabetes diagnosis, were currently using insulin and were diagnosed with diabetes under age 30 (definition 1) or under age 40 (definition 2).

Data were analyzed using SUDAAN, version 10.0 (Research Triangle Institute, Research Triangle Park, NC), accounting for the complex sampling design used in NHANES. The prevalence of diabetes (by each definition) was calculated overall and by age, race/ethnicity, and sex. The sum of sample weights was used to estimate the total number of the US population with type 1 diabetes. We calculated the percentage of diagnosed diabetes and all diabetes (diagnosed diabetes, or hemoglobin A1c ≥6.5%, or fasting glucose ≥126 mg/dL) that was type 1.

The overall fractions (weighted number) with type 1 diabetes mellitus based on definitions 1 and 2 were 2.6/1000 (740,000 people) and 3.4/1000 (970,000 people), respectively (Table). The proportions of diagnosed diabetes that were type 1 were 4.6% and 6.0%, respectively. The proportions of all diabetes that were type 1 were 3.6% and 4.8% for definitions 1 and 2, respectively. Prevalence of type 1 diabetes mellitus increased an average of 0.08% per 4 years for both definitions during the 12-year period 1999–2010 (test for trend, P = 0.02 for definition 1 and 0.06 for definition 2).

In previous estimates, the SEARCH for Diabetes in Youth Study² had identified a prevalence of 1.5/1000 (95% confidence interval [CI] = 1.5 to 1.6/1000) among youth under the age of 20—lower than our estimate of 2.4/1000 for this age group. The National Survey of Children’s Health found 3.2/1000 (95% CI = 2.6 to 3.7/1000) of children under the age of 18 had diabetes, although diagnosis was based on self-report and included both type 1 and type 2 diabetes.¹

Although type 1 diabetes mellitus may develop at any age,⁵ estimates of the prevalence or incidence in older adults are limited.^6,7 By definition, our estimates missed cases starting after the age of 40, while including some younger people with type 2 diabetes who required insulin within a year of diagnosis. Using a higher age cutpoint for diabetes diagnosis in our study would likely result in an even greater misclassification of type 2 diabetes as type 1. When age of diagnosis was not included in the definition (ie, when we include everyone with diabetes who started taking insulin within 1 year of diagnosis), the prevalence was 5.7/1000, representing an upper bound of prevalence. Ideally, diabetes status should be determined based on data collected at or before diagnosis, particularly data on diabetes autoantibodies.⁸

Our prevalence estimates of type 1 diabetes mellitus in the entire civilian noninstitutionalized US population (2.6/1000 for definition 1 and 3.4/1000 for definition 2) are reasonable given current available data. However, the limitations of these data highlight the need for more comprehensive national data collection, both to fully understand the burden of diabetes and to plan healthcare response.