The trusted source for
healthcare information and
Abstract & Commentary
Synopsis: A high saturated-fat and no-starch diet resulted in modest weight loss without negative effects on serum lipid levels; weight loss can be continued for up to a year with such a diet.
Source: Hays JH, et al. Mayo Clin Proc. 2003;78:1331-1336.
Twenty-three patients with significant coronary artery disease (6 of whom were women), 15 women with polycystic ovarian syndrome (PCOS) and 8 women with "reactive hypoglycemia," were recruited for the study. The patients with cardiovascular disease were all being treated with statins (the doses of which were unchanged during this study) and had well-documented cardiovascular disease. Those with diabetes, documented cerebrovascular disease, or recent acute cardiac events were excluded. The patients with PCOS were referred by gynecologists because of desired pregnancy (8), hirsutism (4), and oligomenorrhea (3). The patients with reactive hypoglycemia were extremely obese (mean BMI 46.8 kg/m2) and were defined by "repeated episodes of postprandial dysphoria relieved by food," without a requirement for documented hypoglycemia. The patients in the PCOS and reactive hypoglycemia groups were not taking statins.
The patients on the high saturated-fat and no-starch diet were advised to consume half of their calories as saturated fat, especially red meat and cheese. Eggs were unlimited, and there was no attempt to restrict cholesterol intake. Fresh fruit and nonstarchy vegetables were restricted, and starch was forbidden. Patients kept dietary logs. Weight, serum lipids, and chemistries were measured at baseline for all subjects and again after 6 weeks for the patients with cardiovascular disease, at 24 weeks for those with PCOS, and at 52 weeks for those with reactive hypoglycemia. Lipid fractions were determined by Nuclear Magnetic Resonance (NMR).
Patients were compliant with the meat consumption requirement, and estimated caloric intake fell. Patients in the cardiovascular group ate 2-4 eggs and 1½ pounds of red meat per day. In this group (with cardiovascular disease), mean body weight fell 10%, or 5.5 kg in weeks, with a resultant mean fall in BMI of 2.2 kg/m2.
Neck circumference decreased 0.4 inches. Insulin sensitivity improved and fasting glucose levels fell. Mean triglycerides fell 58 mg/dL, mean very low-density lipoproteins (VLDL) fell 56 mg/dL, and mean VLDL particle size fell 9 nm. Total cholesterol, HDL cholesterol, and LDL cholesterol levels were unchanged, and HDL and LDL size increased. Mean plasma homocysteine levels increased but C reactive protein levels were unchanged.
Ten patients in this group had been classified with metabolic syndrome prior to the study; after 6 weeks, 8 of these 10 patients were no longer in that category, but 1 patient who had not previously been classified as such developed the metabolic syndrome profile. Five of these patients had fasting ketonuria with elevated serum beta hydroxybutyrate levels.
The patients with PCOS lost a mean of 14% of body weight in 24 weeks, and patients with reactive hypoglycemia lost 19.9% of body weight in 52 weeks. No patient in these 2 groups had changes in lipid levels or positive urine ketones.
Comment by Barbara A. Phillips, MD, MSPH
In the accompanying editorial,1 Gerald Gau, also of the Mayo Clinic, says ". . . I am concerned about the long-term cardiovascular risk shown in the published studies . . .We should continue to examine the risk-benefit profiles of the caloric-restricted more rational diets such as the Mediterranean diet." This is true, of course. But maybe we need to re-examine the notion, evident in the title of his editorial, that one size fits all when it comes to weight loss. The people in this study were experiencing significant medical complications of obesity and had likely experienced failure with a variety of diets, including "rational" ones. I think this type of patient is, unfortunately, all too familiar to most of us.
Patients restrict fat grams, as instructed, but eat huge amounts of calories because they are never satisfied. Fat satisfies; carbohydrates don’t. Ultimately, weight loss depends on reduction in calories. In the current study, patients ate fewer calories and, of course, lost weight. The breakthrough with this study was that they appeared to be able to consume very high-fat foods with actual improvement in serum lipid levels.
Although Hays and colleagues never state it outright, a "high saturated-fat and no-starch diet" is essentially the infamous Atkins diet. This study is one of a handful of papers comparing low- or no-carbohydrate diets with the low-fat diets, such as the NCEP (National Cholesterol Education Program) diet promulgated by the American Heart Association, American College of Cardiology, AMA, and others. Most have been short term and have had very poor compliance rates2,3 but have tended to show that patients lose more weight on a high-fat than a low-fat diet. There is 1 study comparing the Atkins diet with low-fat diets for 1 year.4 In this study, only the patients on the high-fat diet had lost significant weight at the end of the year, but a variety of serum lipid measures had worsened in these patients. On the other hand, those on low- or medium-fat diets did not lose a significant amount of weight but had improvement in some serum lipid measures. This is somewhat like choosing between the devil and the deep blue sea. What we still need to know is which is the greater risk factor: obesity or hyperlipidemia. Personally, I think that hyperlipidemia is overrated as a risk factor, especially when compared with obesity. I am impressed that the patients in this study had significant reductions in neck circumference, since neck size is a robust predictor of the likelihood of obstructive sleep apnea (more than 17 inches in a man or 16 inches in a woman strongly predicts significant sleep-disordered breathing).5
Because of the concern about cardiac risk, many physicians now advocate a variant of a high-fat diet, the "South Beach Diet," which is perceived to be a less extreme version of the Atkins diet. Interestingly, while a Google search for "south beach diet" turned up more than 45,000 matching sites, a PubMed search turned up only 2 articles, one of which was a letter and one of which was irrelevant. In my consequently rather nonscientific reading about the 2 diets, the primary difference seems to be that the Atkins diet stresses avoidance of all carbohydrates, while the South Beach Diet sorts carbohydrates as "good" and "bad," based on their glycemic index.
The take-home message from this paper may be that some patients who are refractory to "rational" diets do finally lose weight by avoiding carbohydrates. This is the first paper to suggest that they may be able to do it without adversely affecting serum lipids.
The sad thing about all this is that Dr. Atkins died just as his work was finally being vindicated!
Dr. Phillips, Professor of Medicine, University of Kentucky; Director, Sleep Disorders Center, Samaritan Hospital, Lexington, KY, is Associate Editor of Internal Medicine Alert.
1. Gau GT. Mayo Clin Proc. 2003;78:1329-1330.
2. Samaha FF, et al. N Engl J Med. 2003;348:2074-2081.
3. Foster GD, et al. N Engl J Med. 2003;348:2082-2090.
4. Fleming RM. Prev Cardio. 2002;5:110-118.
5. Davis RJ, et al. Thorax. 1992;47:101-105.