Bitterroot Forest Fire

More Heat Than Light: Carbs and Insulin

We have not heard the last of an intense and seemingly bitter debate about the role carbs and insulin play in promoting obesity. At ObesityWeek, David Ludwig and Kevin Hall sparred about a new publication by Ludwig and colleagues. Letters to the editor of the BMJ are keeping that hot debate going.

At the core of the debate is the central question for Ludwig’s publication. Does a diet with less carbs lead people to burn more calories after they’ve lost weight?

Intense Transparency

Ludwig et al – commendably – published all of the data from their study. And so, Hall re-analyzed it to see if the questions he raised about methods changed the results. In a word, it did. He and Juen Guo wrote:

When conducting the analysis originally planned by the authors, we found that the significant increases in TEE with the low carbohydrate diet that were reported by Ebbeling et al. disappeared. Furthermore, the significant TEE effect modification by baseline insulin secretion also disappeared.

Naturally, Ludwig takes issue with this finding:

We further addressed this issue in a prior Rapid Response, showing that Hall’s previously stated concerns about changing body weight were unfounded, and that statistical models including pre-weight loss baseline did not materially change the result.

Stepping Back

Obviously, intense sentiments about this subject are in play. Ludwig and many others believe that an excess of carbohydrates in our diet is in part responsible for growing obesity prevalence. Carbs are the dietary bad actors of the day.

However, Hall makes a good point in saying that scientists have been studying this question for decades. Does the ratio of carbs to fat affect the calories we burn? Most studies have said no.

Ludwig’s paper points in a different direction, but some messy questions about his methods remain up for debate. Though that debate is important, we agree with Ben Goldacre, who says “Everyone animated on either side of the low-carb ‘debate’ is dreary, no exceptions.”

Please. We need more light and less heat on this subject. Convictions are great, but objective science comes first.

Click here and here for further perspective, here for the original study, and here for all the points and counterpoints published in the BMJ.

Bitterroot Forest Fire, photograph by John McColgan / flickr

Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.


December 7, 2018

3 Responses to “More Heat Than Light: Carbs and Insulin”

  1. December 07, 2018 at 8:19 am, Kevin Hall said:

    Thanks Ted,

    Your readers may also be interested in reading the pre-print of a manuscript describing the reanalysis of these data according to the originally pre-specified analysis plan:

    Best wishes,


    • December 07, 2018 at 11:13 am, Ted said:

      Thanks, Kevin, and thanks, Karl. That link was buried in some of the other references so it’s good to have it out there. Looking forward to your paper appearing in final, peer-reviewed form, Kevin.

  2. December 07, 2018 at 10:47 am, Karl J. Kaiyala said:

    It was interesting to read the entire Goldacre twitter thread: apparently JAMA rejected the manuscript, which reminds me that in 2012 when Ebbeling & Ludwig published in JAMA their initial pilot work purporting to find an energy expenditure advantage to low carb, this claim was rendered in terms of a statistical model that made me go “hmmmmmm…..” and as a guy who does not find p-hacking to be at all dreary I got a letter published (JAMA: 308(11), p.1087). Maybe JAMA has upped their game in terms of statistical rigor. Who knows, but of one thing we can be quite sure: beware low carb ideologues in highly prominent journals.