Document the Sugar Research Foundation's documented funding of nutrition science from the 1960s through the present, assess the specific scientific claims that industry funding influenced, and evaluate what the historical record reveals about the reliability of industry-funded nutrition research.

Primary document analysis of Sugar Research Foundation internal records (1959-1971) as documented in Kearns et al. 2016; systematic review of industry-funding bias literature (Bes-Rastrollo 2013, Schillinger 2016); analysis of 1977 Dietary Goals for Americans development process.

The Documents

In 2016, Cristin Kearns — a dentist-turned-researcher at UCSF who found industry documents in a box she was not supposed to open — published a paper in JAMA Internal Medicine with Laura Schmidt and Stanton Glantz documenting what the Sugar Research Foundation paid Harvard researchers to do in 1967. The paper is called "Sugar Industry and Coronary Heart Disease Research: A Historical Analysis of Internal Industry Documents." I have read it in full.

The finding: the Sugar Research Foundation — trade group for the US sugar industry — paid Dr. Mark Hegsted and Dr. Fredrick Stare of Harvard's nutrition department $6,500 in 1967 (approximately $49,000 in current dollars) to write a literature review on diet and heart disease.

The internal documents show that SRF project manager John Hickson selected the studies to be included and excluded, corresponded with the researchers about the conclusions, and received the papers before publication.

The resulting review, published in the New England Journal of Medicine in 1967, minimized the evidence linking sugar to heart disease and emphasized fat as the primary dietary risk factor. Hegsted and Stare did not disclose the funding. The NEJM did not require disclosure at the time.

Hegsted subsequently became the head of nutrition at the USDA. He helped develop the 1977 Dietary Goals for Americans, which recommended reducing dietary fat. The low-fat dietary guidelines that shaped American eating — and the food industry's response to them, which was to replace fat with sugar in low-fat products — trace a direct line to research whose framing was purchased by the sugar industry in 1967.

I want to be careful about what I am claiming and what I am not. I am not claiming that fat is harmless or that sugar is uniquely responsible for the obesity epidemic. Nutrition science is genuinely complicated.

I am claiming that a specific, foundational piece of the nutrition science consensus was produced by researchers who were paid by an industry with a financial interest in the conclusion, did not disclose that payment, and were managed by the funding source during the research process. That is not science. That is sponsored content in a peer-reviewed journal.

The Pattern Continued

Kearns et al. was not a historical curiosity. A 2013 systematic review by Bes-Rastrollo et al. (PLOS Medicine) examined industry-funded versus independently funded studies on sugar-sweetened beverages and obesity. 6 times more likely to report no association between SSB consumption and obesity than independently funded studies. Seven point six times.

The effect size of funding source on research conclusions in nutrition science is larger than the effect size of most interventions nutrition science studies.

Schillinger et al. (PLOS Biology, 2016) documented that Coca-Cola funded the Global Energy Balance Network, a research organization that promoted the message that physical inactivity — not diet — was the primary driver of obesity. The internal documents, released in litigation, showed Coca-Cola had editorial control over the network's communications.

The network was dissolved after the New York Times reported on it. Coca-Cola subsequently acknowledged the funding arrangement was a mistake. They did not acknowledge the science the funding produced.

The Specific Mechanism

Industry funding of nutrition science does not primarily work through outright fabrication. It works through the research questions that get asked, the study designs that get chosen, the outcomes that get measured, and the conclusions that get emphasized.

A study that measures short-term biomarkers rather than clinical outcomes, uses industry-preferred comparison conditions, and defines success as no statistically significant difference (rather than equivalence) can produce a technically accurate but systematically misleading body of literature. This is not a theoretical concern.

It is the documented methodology of the tobacco industry's science program, and it transferred to the food industry through the same PR firms.

I have spent more time than I would like on this. I once sat next to a food industry nutritionist on a flight from Chicago who explained to me at considerable length why the evidence on sugar and metabolic disease was genuinely uncertain. She was not wrong that it was uncertain. She did not mention that her employer had spent several decades making it uncertain. The flight was four hours. I ran out of wine somewhere over Kansas.