Skip to main content

The Struggles of a $40 Million Nutrition Science Crusade

The Nutrition Science Initiative promised to study obesity and diabetes the right way. Now it’s nearly broke and all but gone.

Emily Waite

Editor’s note: Since this story's publication WIRED has made several clarifications and corrections, which are described at the bottom of this piece.

On Monday night Gary Taubes will begin his second transatlantic trip in a week—from Zurich to Aspen—then eventually head back to Oakland, which he calls home. The crusading science journalist best known for his beef with Big Sugar is heading back from four days at a nutrition conference, where he spent time advocating for a new study into the role of diet in obesity and diabetes. It’s the same kind of work he has taken up with the Nutrition Science Initiative, his nonprofit dedicated to improving the quality of nutrition research. But while the research is pressing forward, the future of his unconventional organization is less certain.

NuSI (pronounced new-see) launched in September 2012 to much fanfare, including in the pages of WIRED. It quickly raised more than $40 million from big-name donors to facilitate expensive, high-risk studies intended to illuminate the root causes of obesity. Taubes and his cofounder, physician-researcher Peter Attia, contended that nutritional science was so inconsistent partly because it was so expensive to do right. With a goal of raising an additional $190 million, they wanted to fund science that would help cut the prevalence of obesity in the US by more than half—and diabetes by 75 percent—by 2025.

Rehabilitating the entire field of nutrition research was always a long shot. But six years in, NuSI is nowhere near achieving its lofty ambitions. In fact, the once-flush organization is nearly broke and all but gone. It’s been three years since it last tweeted, two years since it’s had a real office; today NuSI consists of two part-time employees and an unpaid volunteer hanging around.

But while the organization is almost out of money, Taubes is not yet out of ideas.

When Taubes and Attia first hatched their “Manhattan Project for nutrition,” they planned to work on it nights and weekends, crowdsourcing funds from the low-carb corners of the internet. Between a 2002 New York Times Magazine cover story titled “What If It's All Been a Big Fat Lie?” and his best-selling book Good Calories, Bad Calories, Taubes had become the country’s anti-sugar agitator-in-chief. Then, in 2011, Taubes received an email from a former natural gas trader named John Arnold, who wanted to help.

In May 2012, just weeks after announcing his and his wife’s new charity aimed at reforming iffy areas of science, the Laura and John Arnold Foundation gave NuSI a $4.7 million seed grant to do nutrition research right. In 2013 they followed that up with an additional $35.5 million commitment over five years, making them NuSI’s lead funder.

At the heart of their mission was the decades-old question of whether all calories are, in fact, created equal. The mainstream view is that it’s simply an excess of calories that makes people fat—no matter whether those calories come from a bagel or a steak or a bowl of broccoli. Taubes and Attia subscribe to a growing minority stance, dubbed the carbohydrate/insulin or C/I hypothesis, that contends obesity is caused by an excess of insulin driving energy into fat stores. In other words, carbs make people fat.

Taubes and Attia thought those questions needed a more streamlined research approach to get real answers. So they formed NuSI to funnel money into a rigorous new set of studies while leaving scientists with the experimental independence that would shield their results from bias.

With the Arnold money in hand, Taubes and Attia started recruiting top researchers in 2012 to conduct four initial studies. They brought on people who disagreed with them, like Kevin Hall, a senior investigator at the NIH’s National Institute of Diabetes and Digestive and Kidney Diseases, whose mathematical models predicted that a low-carb, low-insulin diet would have only a tiny impact on calorie-burning. He would head up one of NuSI’s first studies, dubbed the Energy Balance Consortium.

The EBC’s pilot project would put 17 overweight men inside metabolic wards for two months, feeding them precisely formulated meals and pricking and prodding to see what happened to their bodies on a low-carb diet. If it made them burn calories faster, a potential follow-up study would do the same tests on a bigger group of people. If the effect was minimal, researchers proposed that they could then test the effect of low-carb diets on hunger.

Hall was skeptical they would find anything to support the carbohydrate/insulin hypothesis. But he was assured by the terms of the contract; NuSI would have no control over the pilot study’s published conclusions.

At first, things went according to plan. The EBC researchers met with NuSI quarterly to finalize the study’s design and clinical procedures. NuSI signed a consulting agreement with Dr. Jeff Volek, author of the book The Art and Science of Low Carbohydrate Living, to create the diets and menus.

By August 2014, the EBC researchers had preliminary results on their 17 volunteers: Their published conclusions noted a relatively small difference in energy expenditure. That didn’t mean it was a failure; to the researchers, they had succeeded in verifying the methodology before using it in an even bigger, longer study. “We had to work out these rather complex logistics of getting common food sources distributed among many institutions,” says Rudolph Leibel, one of the consortium scientists working on the pilot at Columbia. “It looked like something the Allies would have organized for all the landings on D-Day.”

Advertisement

But when Hall presented the pilot’s results in person to representatives from NuSI at a meeting in Bethesda in September, they were not so rosy-eyed. When NuSI saw the data, it began providing extensive critiques.

Taubes had issues with both the researchers’ conclusions and many of the study’s designs, which fed participants a “standard American diet” for four weeks before switching them to an extremely low-carb, or ketogenic, regimen with the same amount of calories. It was supposed to get them to a stable weight, or energy balance, to establish a baseline before going keto. But the subjects all lost weight even before they cut out carbs. Taubes contended that was because the standard diet didn’t have enough refined sugary beverages to depict average American consumption.

“From my perspective, the pilot was a failure for several reasons,” Taubes says. “First, it failed to get people in energy balance in the run-in period, and that was a necessary condition to interpret the findings.” In addition, he points out, the design didn’t include a group of nondieters, and nonrandomized trials do not allow for firm conclusions about causality, conditions that everyone in the group knew going in. In his eyes, all the pilot told them was that their method was flawed. "If this was an animal study, they’d have thrown them out,” he says. “Euthanized them and started over.”

But NuSI had already spent $5 million of the Arnolds’ money, and the researchers were eager to get to the second phase of the study. As they worked out the details through 2015, the relationship between EBC and NuSI continued to fray. “There was not a real team,” says Eric Ravussin, EBC’s co-principal investigator and director of Pennington’s Nutrition Obesity Research Center. “As scientists we were in agreement over the pilot results and the new protocols, but NuSI had some concerns. It eventually just became us versus them.”

According to Hall and Ravussin, NuSI began to push back, in a way they felt jeopardized their ability to do good science. In April, the EBC researchers sent NuSI an email requesting to reestablish their academic freedom.

"We lived up to the terms of the contract, as we did with all the research NuSI funded," Taubes says. "Their academic freedom was never even remotely threatened.”

Emily Waite

The relationship between the EBC researchers, NuSI, and the Arnold Foundation deteriorated even further. At a meeting in front of John Arnold, NuSI directors Taubes and Mark Friedman openly quarreled with Hall and his colleagues about what was really necessary to run a good study. And at the end of December, Attia quietly resigned from the organization. Sources close to him say he was unhappy being a full-time fund-raiser; he wanted to get back to research.

NuSI scrambled to fill Attia’s position as president, first with Christopher Ochner, a psychiatrist from the Icahn School of Medicine at Mount Sinai, and a few months later with Julie Eckstrand, NuSI’s then-director of clinical operations, who has since left. At the beginning of 2016, NuSI’s yearly contract with the Arnold Foundation was replaced by a series of three-month bridge contracts. The team of 15 full-time employees and major contractors shrunk to a skeleton crew that could handle the three remaining studies. NuSI shuttered its San Diego headquarters and became a virtual organization.

By January 2016, Hall had had enough. At the end of a planning meeting, he stepped down from his role with the EBC, citing changing expectations about the structure and practice of the NuSI collaboration.

The remaining researchers continued to clash with NuSI about the second phase. In July, the pilot results were finally published in The American Journal of Clinical Nutrition. They received a lot of media attention, in no small part because Hall said the pilot, along with another study he’d conducted previously, “basically falsify” the carbohydrate/insulin hypothesis.

Advertisement

By the end of the summer, the Arnold Foundation had decided not to fund the second phase of the study. NuSI stopped getting checks from the Arnolds, but the foundation didn’t stop funding research into the carbohydrate/insulin question. That fall they opened their search to the wider world, putting out a call for proposals for “rigorous research projects that will assess the role that sugar and/or macronutrients play in metabolic responses and fat accumulation.”

Taubes continued to work with the Arnold Foundation to review some of those proposals. They accepted one from David Ludwig, a professor of nutrition at the Harvard School of Public Health and co-principal investigator on a different NuSI dietary experiment that launched in July 2013. On that study, which is now complete and currently under peer review, Ludwig says his team’s experiences with NuSI were positive. “They maintained oversight of study progress and milestones; we maintained control over study conduct and had full scientific independence,” he says. Now, with $12 million from the Arnold Foundation, Ludwig is independently completing the study NuSI’s EBC had been established to conduct.

The Arnold Foundation declined to respond to specific questions about how it came to end its relationship with NuSI about $14 million short its commitment. A spokesperson emailed the following statement: “This research was designed to answer scientific questions in the fields of nutrition and obesity. While the foundation no longer directly supports NuSI initiatives, we continue to fund work in the field of nutrition science and remain open to further investments in this area. The NuSI project was a worthwhile effort and remains an important health-related issue for Americans today.”

It’s still too soon to assess what NuSI has added to the nutrition science canon. Results from the two outstanding NuSI-backed studies are due later this year. The fourth and largest one, conducted at Stanford, randomized 600 overweight-to-obese subjects into low-fat versus low-carb diets for a year and looked at whether or not their weight loss could be explained by their metabolism or their DNA. Published this February in JAMA, the study found no differences between the two diets and no meaningful relationship between weight loss and insulin secretion.

Obesity docs like Yoni Freedhoff, a professor of family medicine at the University of Ottawa, aren’t surprised that NuSI hasn’t sparked an epistemological revolution. “From the outset, their approach was simply that knowledge will be enough to drive behavior,” says Freedhoff, who has argued that efforts to prove one diet is better than another do a disservice to patients by implying there’s only one right way to lose weight. He’d love to see research dollars be spent instead on studying how to improve adherence to different eating strategies.

Taubes believes there might still be a need for NuSI in the future. “Our convictions have gotten us this far, and despite some disappointments these questions still seem vitally important to test,” he says. “I say this to my wife all the time: ‘Maybe I’m a quack.’ All quacks are sure they’re right. Isn’t that the defining characteristic of a quack? But the fact is that we funded four studies, and the three randomized trials were highly successful operationally. One of these has been published in a top journal with interesting results, and I remain hopeful that we will soon see if the last two studies will move some needles.”

For now, Taubes thinks NuSI could evolve into something a bit more humble. Between its current coffers and the agreements he’s working on, he thinks NuSI can stay afloat for several years, eventually supporting more outside research, though on a much more modest scale. He’s proposed instituting a scientific oversight committee to make sure everyone agrees on methods and statistical analyses from the outset.

Or maybe NuSI doesn’t need to exist at all. Between the allies Taubes planned to meet in Zurich and the philanthropists he’s lined up to fund his new study, it’s possible he could continue to carry out NuSI’s core mission without his novel nonprofit. The important thing, he says, is just that the research gets done, one way or another.

He’s also got more articles and books he still wants to write, not exclusively about sugar. It’s tricky, though. “I know I clearly have conflicts that other journalists just don’t have, and that’s a tightrope I haven’t figured out how to walk yet,” Taubes says. “This nutrition science crusade—right or wrong—expands easily to fill all the time in my life that can be allotted to work. So I’m going to figure out how to partition time better in the future.”

In between flights and conference dinners, he’s been checking his email for notes on an upcoming article about a new kind of observational study that uses genetic variation to mimic a randomized control trial. While the story isn’t strictly related to nutritional science, Taubes now has the kind of conflicts of interest that make publications wary. He’s writing for another outlet after his old editor at Science wouldn't touch his work. Taubes founded NuSI to support objective science; now it's his own objectivity he has to defend.

Update: Corrections appended, 7/31/2018, 9:53 pm EDT

This story has been updated to clarify the nature of Gary Taubes' trip, the intentions and motivations of NuSI and Taubes, the organization’s current leadership and financial status, and Taubes’ relationship with a news outlet. Language has also been changed to clarify the relationship between NuSI and the researchers who ran its studies, the chronology of NuSI meetings and administrative changes, the precise nature of the carbohydrate/insulin hypothesis, and to more clearly describe study designs, methodologies, and findings. The updated story also includes comment from Taubes on researchers’ allegations that NuSI infringed on their academic freedom and information about other studies funded by the Laura and John Arnold Foundation.

Update: Correction appended, 8/8/2018, 3:20 pm EDT
This story has been amended to correct information about a contract between NuSI and the EBC.

// // //