Retrospective #253: The “Diabetes Plateau”

“Diabetes rates among adults in the United States are finally leveling off, new data from the Centers for Disease Control and Prevention suggests,” a 2014 Diabetes-in-Control piece leads. The headline of the piece is less speculative. It simply declares, “Diabetes Rates in the US Have Finally Plateaued.” It grabbed my attention, and with the primary audience being clinicians, the Diabetes-in-Control also had a self-congratulatory tone that I think is unwarranted.

The Diabetes-in-Control piece is based on “findings from a National Health Interview Survey (NHIS) on both prevalence and incidence of diabetes from 1980 to 2012,” published online in the Journal of the American Medical Association (JAMA). Prevalence refers to its widespread, existential presence while incidence, in the epidemiological sense, relates to the initial appearance of a new condition (T2DM) in the population. In the NHIS survey, both were self-reported. The findings which led to the headline are a reduction of the incidence of Type 2 diabetes in the general population.

Among the 664,969 adults aged 20 to 79 who responded to the NHIS survey, the prevalence of diabetes was 3.5% in 1990, 7.9% in 2008, and 8.3% in 2012. The incidence per 1000 people was 3.2% in 1990, 8.8 in 2008 and 7.1% in 2012. Percentage change in prevalence and incidence was not significant in the 1980s; however, both jumped significantly from 1990 to 2008. Then, from 2008 to 2012, prevalence plateaued and incidence diminished significantly.

“However,” Diabetes-in-Control highlighted, “both diabetes incidence and prevalence continued to increase at a significantly greater rate for young adults aged 20 to 44 years compared with older adults and for black and Hispanic adults compared with white adults.” In addition, “The rate of increase in prevalence was higher for those with a high school education or less compared with those with more than a high school education.” Hmmm. Another sociodemographic disparity. I wonder if dietary choices had anything to do with that. (Catch the sarcasm, pullese.)

The sub-head in the article asks, “Is It All About Obesity?” Phrasing it as a question, of course, begs the answer and wrongly suggests obesity is causal. The author notes that a statistical adjustment) “reduced the annual percent change in incidence by about a third…” with the author telling Diabetes-in-Control, “This suggests that the leveling off of obesity that occurred over the same period explains [!?!] a large part of the diabetes plateau, but not all of it.” The knock-out punch was, “The BMIs in the NHIS study were self-reported and therefore most likely underestimated.”

Searching for another explanation, the Diabetes-in-Control author speculates, “Beyond obesity, other possible influences on the reduction in diabetes rates include improvements in diet and activity levels and changes in diagnostic criteria.” The author implies that we are getting obese at a lower rate than before; that she adduces this from the finding of this epidemiological study that since the incidence of diabetes diminished significantly from 2008 to 2012, that obesity associated with the incidence and prevalence of Type 2 diabetes did as well.

The evidence she offers for this association – the decline in incidence of Type 2 diabetes with an assumed decline in obesity, contrary to every other study I’ve seen – is “improvements in diet and activity levels, and changes in diagnostic criteria.” Oh, how I wish that were so. That we could realize an actual result by just saying it over and over: Eat a balanced, USDA recommended one-size-fits-all diet and exercise more, and, abracadabra, you will become thin.

Maybe the explanation for “the diabetes plateau” lies instead in the “changes in the diagnostic criteria.” Do you think this could affect the way that 664,969 people self-reported their condition? Could that have “confounded” the study results to the point that all the other “statistical adjustments” were totally insignificant by comparison? Do ya think?

“The 1997 lowering of the fasting plasma glucose cutoff from 140 mg/dL to 126 mg/dL [for a Dx of T2DM] may have increased the diabetes incidence…,” the CDC study author noted. I love it! “May have,” she says. Whereas, “the 2010 shift to use HbA1c for diagnosis may have reduced it, since HbA1c detects fewer cases of hyperglycemia,” she said. Interesting indeed, and contrary to my understanding. The reason the medical establishment has shifted from “fasting” to A1c is for the express purpose of capturing blood glucose postprandial “excursions,” i.e. hyperglycemic spikes due to impaired glucose tolerance (IGT) over a 3-month period. Fasting glucose testing does not do that.