By: Sean Bourke, MD
Note: This article was originally published in the July/August 2014 issue of San Francisco Medicine.
When Dr. Conrad Lai and I started JumpstartMD in 2007, we approached the venture from the standpoint of two Stanford-trained emergency medicine physicians passionate about creating a medically supervised weight-loss and wellness program designed to provide a practical, sustainable, and preemptive response to the chronic and preventable medical conditions that account for 75 percent of our nation’s health care spending.
At the time, our thinking echoed the conventional wisdom that we learned in medical school—namely, that a low-fat diet pays the most dividends for health. It’s advice that’s aligned for the last fifty years with that of the American Diabetes Association (ADA), American Heart Association (AHA), and the U.S. Department of Agriculture’s (USDA) Dietary Guidelines for Americans and Food Guide Pyramid.
But the years of patient interactions; engaging with thought leaders, many from the Bay Area (Gary Taubes and Drs. Robert Lustig, Steve Phinney, Peter Attia, and Ronald Krauss); and my own reading, including Bay Area native Nina Teicholz’s recent book, The Big Fat Surprise, which informed many of the arguments in this article, have brought me to the conclusion that the vast majority of contemporary dietary recommendations that we’ve assumed to be correct are simply wrong, the product of weak data overinterpreted by well-intentioned experts eager to end the rising tide of diseases of modernity.
A “Great Experiment”
This “great experiment” in low-fat living begat an era unprecedented in human history. We now consume, and export, what can arguably be described as the world’s deadliest diet, the nutritional equivalent of cigarettes.
The spark? Amid an ambient fat phobia and obsession—first with total cholesterol and later with low-density lipoproteins (LDL) as primary risk factors for cardiovascular disease—experts made low-fat dietary recommendations that, despite being premature and unfounded, became institutionalized and entrenched. These events triggered a reciprocal—and unanticipated—creep in the consumption of refined carbohydrates, which now make up the bulk of our modern diet in the face of increasing evidence that they are (1) precursors to the most atherogenic small LDL particles (particles that can predict a proclivity to develop diabetes even prior to a rise in insulin), (2) promoters of fat formation through the insulin response, (3) proinflammatory, and (4) associated with advanced glycation end (AGE)-products that may accelerate aging.
I see the consequences of this misguided thinking in my practice regularly. Recently, a patient, expressing frustration at her poor weight loss during a diet recall, was surprised to learn that the 38 grams of carbohydrates in her daily Frappuccino—32 of which came from added sugar, which is 7 grams higher than the daily added sugar limits now recommended for women by AHA—were not “completely innocuous” from a weight-loss and health standpoint.
And why wouldn’t our fellow Americans assume as much when every trip to the grocery store bombards them with messaging that this or that product is “low fat” or “low in cholesterol.” (Dietary cholesterol, meanwhile, was definitively shown as far back as the 1950s to have minimal—that is, a “trivial”—effect on serum cholesterol, yet you’d never know it given our low-fat, low-cholesterol emphasis.)
Shouldn’t we own up to our contribution to the disastrous public health debacle that is our country today? The Food Guide Pyramid has had it all wrong since its formation, and yet it’s been propelled upward by misguided politics posing as science.
The path forward? It starts with education and awareness. And that education and awareness begins with an understanding of the flimsy science behind the dietary recommendations that our friends, families, and patients accepted in good faith long before they’d been definitively proven. Unfortunately, support for that science still exists.
Where It All Started
In the second half of the twentieth century, no man and no study had a greater influence on our arguably misguided direction than Ancel Keys, the scientist who secured a seat in 1961 on the AHA’s Nutrition Committee and who led the famous “Seven Countries Study,” the first multi-country epidemiologic study ever undertaken.
Further review of this pivotally influential study has shown that, aside from the fact that epidemiologic studies can identify associations but not causality, the scientific rigor and methodologies on which Keys based his conclusions were deeply flawed for a host of reasons: selection bias and lack of randomization, incomplete data, postwar dietary aberrations from the norm, flawed sampling such as one-third of the diet recall surveys for the population of Crete occurring during the Lenten period, when saturated fat intake was halved), and more. These flaws were buried while Keys’s recommendations were pushed forward through his great force of will, savvy leveraging of media and other influencers, and marginalization of dissenting viewpoints.
And what of the controlled studies to definitively prove Keys’s diet-heart hypothesis that fat, and saturated fat specifically, led to heart disease? In fact, alternative hypotheses, many implicating sugar and refined carbohydrates as a primary cause of atherogenic dyslipidemia (something Keys described as a “mountain on nonsense”) and a key culprit in our obesity epidemic, are now being borne out by science.
By 1993, after the AHA had been recommending a low-fat diet for thirty years and the USDA for fifteen, the Women’s Health Initiative followed 20,000 women in a low-fat diet group to study whether this recommendation actually worked. After a decade of follow-up, the low-fat cohort was no less likely to contract various cancers, suffer from stroke or heart attack, or lose more weight. (JAMA, 2006; 295, no. 6. 629-42, 643-54, 655-66; JAMA, 2006; 295, no. 1:39)
In 2007, Stanford’s Christopher Gardner compared premenopausal women on the Atkins diet with those on the Zone (moderately low carbohydrate), LEARN (moderately low in fat, moderately high in carbohydrate), and Ornish (very low in fat and very high in carbohydrates) diets. The conclusion: The Atkins group lost more weight and had better “overall metabolic effects” than the lower-fat (higher-carb) groups. (JAMA, 2007; 297:969)
In 2008, Iris Shai et al performed a well-controlled two-year study of 322 individuals comparing outcomes on three diets: the AHA low-fat diet, a Mediterranean diet, and the Atkins diet. Atkins dieters not only lost the most weight but also saw the greatest decline in risk factors. The low-fat group fared the worst. In fact, the more fat one ate, the healthier one appeared. (NEJM, 2008; 359: 229)
In 2009, Jeff Volek and others showed a more favorable impact of a carbohydrate-restricted, high-fat diet on metabolic syndrome than a low-fat diet. (Lipids, 2009; 44:297)
In 2010, Dr. Ron Krauss and others published a 350,000-patient meta-analysis stating that “there is no significant evidence for concluding that dietary saturated fat is associated with increased risked for coronary heart disease or cardiovascular disease.” (AJCN, 2010; 91:535) Chowdhury et al recently corroborated the same result with a nearly 650,000-patient meta-analysis that concluded that saturated fat does not cause heart disease and that “current evidence does not support cardiovascular guidelines that encourage . . . low consumption of total saturated fats.” (Annals Int Med, 2014; 150:398)
In April 2014, UCSF’s Laura Saslow et al published a study comparing type II diabetics on a medium-carbohydrate, low-fat, calorie-restricted American Diabetes Association diet with a group on a very low-carbohydrate, high-fat, non-calorie-restricted diet meant to induce nutritional ketosis. Over three months, the nutritional ketosis group lost twice as much weight (5.5 kg versus 2.6 kg) and was able to discontinue diabetes medication and sulfonylurea medications at four and six times the rate, respectively, of the moderate carbohydrate group. (PLOS ONE, 2014; vol. 9, issue 4) One-year data is pending.
The Cholesterol Connection
For decades, part of the rationale against saturated fats stemmed from concerns that saturated fat can affect total and calculated LDL (LDL-C) cholesterol. Yet the march of science has shown that both total cholesterol and LDL-C are not the best predictors of cardiovascular risk. LDL-C isn’t even a biologically active entity but instead an indirect estimate.
For example, thirty years into the Framingham follow-up study, total cholesterol was shown to be an ineffective marker of risk. In fact, 80 percent of the people who develop coronary artery disease have the same total cholesterol as those who do not. (Lloyd-Jones et al, Arch Int Med 2001; 161:949) Of nearly 137,000 patients hospitalized with coronary artery disease, more than 75 percent had LDL-C levels below 130 mg/dl and 23 percent had LDL-C levels below 70 mg/dl. (Sachdeva et al, AHJ; 157:111)
To the contrary, increasing evidence—as presented in the Malmo Trial (Musunuru et al, ATVB; 29:1975, 2009)—shows that carbohydrates, and refined carbohydrates in particular, confer cardiovascular risk independent of saturated-fat intake by acting as precursors to atherogenic smaller LDL particles. (Krauss et al, AJCN; 83:1025) By contrast, saturates tend to drive generation of larger LDL particles, which are generally considered benign.
Seeing the Big Picture
When I think about the studies above and marvel at the tide of illnesses that parallel our rising obesity rates (http://www.cdc.gov/obesity/data/adult.html), I pause in knowing that the major macronutrient changes we have made since the onset of the diabesity epidemic was a recommended reduction in our percentage of calories from fat, and saturates in particular, and a reciprocal unintended 25 percent rise in our consumption of carbohydrates.
That’s not to say that we’ve definitively determined that all fats are fine—relatively new industrialized oils and solid polyunsaturated fats, for example, reveal another untested experiment, a matter of importance given that they now account for 8 percent of our total caloric intake and there is some suggestion they may drive cancer risk or, more generally, a proinflammatory state through an unprecedented excess of omega-6 fatty acids—or that there might not be issues with saturates for some people, or that, more generally, we shouldn’t make blanket nationwide nutrition recommendations when what we need is to personalize optimal dietary recommendations in a way that matches human heterogeneity.
But it is to say that for decades the U.S. had a stable prevalence of obesity of around 12 percent and that in 1961, the prevalence of diabetes in the adult population was 1 percent. Today, those numbers are closer to 35 percent and 11 percent, respectively, with 25 percent of the country being “prediabetic.” The diabesity epidemic is arguably happening to people in large part because of misguided recommendations about dietary fat that were founded on flawed science.
If true, the problem is eminently fixable, if we can let the truth surface and use education and awareness to move us forward. At JumpstartMD, our practice continues to evolve in response to the latest thinking and observes unparalleled outcomes in weight loss and health improvements in the process.
Sean Bourke, MD, is CEO of JumpstartMD, Northern California’s largest medical weight loss and wellness practice. A member of the SFMS and the American Society of Bariatric Physicians, Dr. Bourke is a graduate of Yale College and the University of Southern California School of Medicine. He received his postgraduate training in emergency medicine at Stanford University. JumpstartMD has two offices in San Francisco and nine others encircling the greater Bay Area.