Undiagnosed diabetes may be less common in U.S. than previously estimated

The overall prevalence of diabetes (diagnosed and confirmed undiagnosed) increased from 5.5% of the U.S. population in 1988 to 1994 to 10.8% in 2011 to 2014, according to data from the National Health and Nutrition Examination Survey (NHANES).


Fewer Americans have undiagnosed diabetes than previously thought, according to a recent analysis of National Health and Nutrition Examination Survey (NHANES) data.

The study used results of HbA1c and fasting glucose tests from U.S. adults ages 20 years and older participating in NHANES in 1988 to 1994 or 1999 to 2014. Confirmed undiagnosed diabetes was defined as having both a fasting glucose level of at least 7.0 mmol/L (126 mg/dL) and an HbA1c of at least 6.5% without a self-report of diagnosed diabetes. Results were published online by Annals of Internal Medicine on Oct. 24.

The overall prevalence of diabetes (diagnosed and confirmed undiagnosed) increased from 5.5% of the U.S. population in 1988 to 1994 to 10.8% in 2011 to 2014, the study found. Confirmed undiagnosed diabetes also increased from 0.89% of the population in 1988 to 1994 to 1.2% in 2011 to 2014 but decreased as a proportion of total diabetes cases. In 1988 to 1994, 16.3% of diabetes cases were undiagnosed, compared to 10.9% in 2011 to 2014. Risk factors for undiagnosed diabetes included being overweight or obese, being older, being a racial or ethnic minority, and lacking health insurance or access to health care.

Current CDC estimates say that 30.1 million U.S. adults have diabetes, 7.2 million of them undiagnosed, the study authors noted. These statistics are also derived from NHANES but based only on a single blood test result. This study, which required confirmatory testing, suggests that only 25.5 million U.S. adults had diabetes in 2015, and 2.77 million of them were undiagnosed. “Our analysis demonstrates that we are doing a better job with diabetes screening and diagnosis than might be inferred from previous estimates; overall, only a relatively small portion of the U.S. population who would be identified clinically as having diabetes has undiagnosed disease,” the authors wrote.

The study's results support current targeted screening recommendations including middle-aged and older adults, overweight or obese adults, and high-risk racial/ethnic groups (including Asian-Americans), according to the authors. The results also show the importance of paying attention to any discordance between a patient's HbA1c and fasting blood glucose results. In the study, discordant results were more common as patients aged, “suggesting a high rate of false-positive results in older adults and thus an even greater need to use confirmatory testing in this population,” the authors said. The study had a number of limitations, including the use of a single blood sample for both tests.

An accompanying editorial summarized the study's implications for clinicians. “First, these results underscore the importance of validating one diabetes test with another to make a diabetes diagnosis in patients without symptoms,” the editorialist said. Physicians should also know the risk factors for confirmed undiagnosed diabetes and provide risk reduction and close follow-up to patients with conflicting screening test results and risk factors for diabetes, the editorial said.