A Web-based risk test endorsed by the American Medical Association, the American Diabetes Association, and the Centers for Disease Control and Prevention would label 59%-81% of adults as prediabetic if it were applied to the general U.S. population, according to a Research Letter to the Editor published online Oct. 3 in JAMA Internal Medicine.

The test is intended to evaluate the risk for prediabetes of any adult patient, and it advises those it identifies as prediabetic to see their physicians for confirmatory blood glucose testing. But given the results of their study, labeling this many people as prediabetic constitutes overmedicalization and could have the unintended consequence of diluting access to health care for people who actually have type 2 diabetes and other chronic conditions, said Saeid Shahraz, MD, PhD , of the Predictive Analytics and Comparative Effectiveness Center, Tufts Medical Center, Boston, and his associates.

The investigators assessed how many people this test would classify as prediabetic by analyzing data for a nationally representative sample of 10,175 adults who participated in the National Health and Nutrition Examination Survey in 2013-2014. They extracted the same information from participant responses to the NHANES survey that is elicited by the Web-based test: age, sex, history of gestational diabetes, first-degree relatives with diabetes, hypertension, physical activity level, and weight.

Among Americans aged 40 years and older, this test would classify an estimated 73.3 million, or 58.7%, as prediabetic; among those aged 60 years and older, it would classify 80.8% as prediabetic. The test also would advise all those people to see their physicians and undergo blood glucose testing for confirmation, Dr. Shahraz and his associates said (JAMA Intern Med. 2016 Oct 3. doi: 10.1001/jamainternmed.2016.5919).

This would be “premature” for many reasons. Intensive lifestyle modifications are not known to be beneficial for such patients, even those who are found to have impaired glucose tolerance. And there is no direct evidence that preventing type 2 diabetes actually reduces the risk for diabetes-related complications. Moreover, the natural progression from prediabetes to type 2 diabetes is likely to be slow, which calls into question the need for any intervention, particularly for older patients likely to die from competing causes, the researchers noted.

“A valid method to examine for prediabetes should avoid unnecessary medicalization by labeling a disease predecessor as a medical condition and seek to concentrate on people at highest risk to allow for efficient distribution of limited health care resources,” they added.