Note to our followers: Due to a lack of sufficient funding, will cease daily publication of new content at the end of 2018. Publisher Gary Schwitzer and other contributors may post new articles periodically. If you wish to donate, your gift might help keep the site available to the public for a few more years, by defraying costs of web hosting and maintenance. All of our 6,000+ published articles contain lessons to help people improve their critical thinking about health care. Read more about our change in status. And here's how to make a donation.

‘Reverse’ diabetes with intermittent fasting? What you need to know

Michael Joyce is a writer-producer with and tweets as @mlmjoyce

reverse diabetes with fastingA health care message can get polluted, diluted, and distorted by many people as it moves from researchers, through journals and press releases, and ultimately into the news coverage.

Here’s an example of what happens when it starts at the source with grandiose language used by the study authors, is perpetuated in newswire/releases and the publishing journal, and is ultimately parroted and amplified by news organizations.

The study

A case report of 3 men (ages 40, 52, 67) who received dietary education and underwent intermittent fasting over 10 months; all three were able to go off their insulin, lose weight, and experience very mild reductions in their hemoglobin A1C (a rough average of blood sugar levels over the previous 2-3 months). In this study “intermittent fasting” was roughly three 24-hour periods each week in which the subjects were limited to water, coffee, tea, or bone broth.

The published conclusion: “Medically supervised, therapeutic fasting regimens can help reverse type 2 diabetes” (and minimize the use of prescriptions “and possibly surgical interventions” as well).

The limitations

“The reversal language is unjustified and the evidence here is extremely anecdotal,” said Kevin Klatt, PhD, a researcher in dietetics and molecular nutrition at Cornell University.

“What qualifies as a ‘reversal’ with diabetes is unclear. But what is clear is that following a single test like hemoglobin A1C, or simply going off insulin, does not qualify as reversal. Much more robust data is needed and no sort of clinical recommendations can be made from the results of just three subjects.”

The author’s language

The lead author (or authors) is/are quoted as saying:

  • “This present case series showed that 24-hour fasting regimens can significantly reverse or eliminate the need for diabetic medication” (authors quoted in BMJ news release)
  • “[this study] has the potential to completely reverse type 2 diabetes … it changes everything about how we should treat the disease … it is the ultimate empowerment and the ultimate in savings for public health” (Author, Dr. Jason Fung, quoted in Newsweek)

The coverage

Much of the news coverage was contradictory and, therefore, confusing if not misleading. For example, HealthDay — which used the reversal language in its headline and lede — rightly pointed out that with “just three patients it’s impossible to draw firm conclusions about the use of fasting to treat type 2 diabetes.” Yet, in the very next sentence wrote:

Still, the results are noteworthy, given that one in 10 Americans and Canadians have type 2 diabetes, the investigators said in a journal news release.

So which is it? Impossible to draw conclusions? Or truly noteworthy or even newsworthy?

Equally confusing was Newsweek frequently (and uncritically) using the reversal language but then providing reasons why the study “does not speak to the entirety of diabetes care and management in a way that is truly generalizable” … such as the need for much larger randomized studies, as well as the potential harms of precipitously dropping blood sugar levels with fasting.

Ultimately, it’s irresponsible for researchers who know the science of the disease to imply they can “reverse” it based on the responses of 3 men. Especially when such dramatic claims often lead to reporters carelessly adopting this erroneous and misleading language. It’s a disturbing example of a message that got polluted at the source and trickled down to readers.

Also disturbing is that this case report has been promoted under BMJ’s “Myth Exploded” series, the inclusion criteria of which aren’t carefully defined by the editors. This raises two important questions. First, does any case report have the statistical power to explode a myth? One would think only multiple, randomized controlled trials are robust enough to do that.

And couldn’t anecdotal case reports like this one achieve the opposite? That is, perpetuate a myth?


You might also like

Comments (1)

We Welcome Comments. But please note: We will delete comments left by anyone who doesn’t leave an actual first and last name and an actual email address.

We will delete comments that include personal attacks, unfounded allegations, unverified facts, product pitches, or profanity. We will also end any thread of repetitive comments. Comments should primarily discuss the quality (or lack thereof) in journalism or other media messages about health and medicine. This is not intended to be a forum for definitive discussions about medicine or science. Nor is it a forum to share your personal story about a disease or treatment -- your comment must relate to media messages about health care. If your comment doesn't adhere to these policies, we won't post it. Questions? Please see more on our comments policy.

Andrew DePristo

October 12, 2018 at 5:25 pm

I also want to point out that it is not at all clear whether fasting or weight loss is the cause, as substantial weight loss has already been shown to reduce need for insulin in T2D. “The three patients … also lost substantial amounts of body weight and had reduced waist circumferences.” In addition, the problem is NOT to reduce T2D dependence on insulin by weight loss under supervision, but getting any obese person to keep the weight off long term.

BTW, the data has some peculiarities that warrant investigation:
Patient 1: 84 kg to 74 kg and waist 100 cm to 87 cm;
Patient 2: 61 kg to 50 kg and waist 123 cm to 110 cm;
Patient 3: 97 kg to 88 kg and waist 123 cm to 110 cm.
Each patient lost either 10 or 11 kg AND had exactly the same waist decrease of 13 cm. Given the large differences in initial bodyweight, 97 kg vs. 61 kg, the data warrants further investigation.