The perils of the Atkins diet
Atkins diet is more expensive than conventional diets, yet no more effective
By Midhat Farooqi
July 16, 2003
link to article
The Atkins diet is in fashion once again. Dr. Robert Atkins, the man behind the diet, died in April of an accidental fall, giving new life to his book, "Dr. Atkins' New Diet Revolution," which currently holds the number one spot on USA Today's best-seller list. The book is also No. 1 on the New York Times' best-seller list of paperback advice books. This is disappointing, considering that the Atkins' diet is expensive and no more effective than a conventional diet in the long term.
The diet's revival can be attributed to the publication of recent scientific studies supporting it. The New England Journal of Medicine published one such prominent study in May. Since then, supporters of the diet have used it to claim that Atkins was correct all along: restricting carbohydrates and eating only fat and protein makes a good diet. They claim that this is the way to eat now and for life. However, the study does not completely vindicate the Atkins diet: the results, though promising, were only collected over the short term, and they may be deceptively higher due to incomplete analysis. As it turns out, this is not the way to eat, now or ever.
The study involved 63 obese men and women who were assigned to either a low-carbohydrate, high-protein and high-fat diet -- namely, Atkins -- or a low-calorie, low-fat and high-carbohydrate diet -- the more conventional dieting approach. The researchers then measured weight loss at three, six and 12 months.
The study was a controlled trial -- meaning that various other factors, which could influence the results, were minimized, such as the subjects taking weight-altering drugs. The study assigned the 63 participants to either diet at random, further reducing the likelihood of bias, and professional contact was minimal to imitate the approach used by most dieters.
The study showed that the Atkins diet produced greater weight loss than the conventional diet for the first six months. Those on Atkins lost 7 percent of their body weight, while those on the conventional diet lost 3.2 percent. However, at 12 months, there was no significant difference in the amount of weight loss produced by the Atkins' diet as compared to the conventional diet.
This point is the key, as it suggests that the long-term efficacy of the Atkins' diet is similar to that of a conventional diet. But because the conventional diet is cheaper -- fruits, vegetables and low-fat products are generally less expensive than excessive amounts of meat and low-carbohydrate products -- why follow the Atkins diet?
There are two results of a good diet: one is losing weight initially, and the other is keeping it off. In the study, people on the Atkins diet gained weight they lost in the first six months off the diet, much more than those on the low-fat diet. The Atkins diet may have produced more weight loss but it also resulted in a larger weight rebound.
The authors of the experiment state that this rebound may be due to the small sample size: the smaller the sample, the larger the sampling error. But this applies to all the results: they may not represent the larger population. This problem is further exacerbated by the fact that all of the subjects were obese. Some people have metabolic differences, such as a decreased sensitivity to insulin. Thus, the results might not apply to people who are merely overweight or not fat at all. More importantly, the results might have been different if normal and overweight people were included in the survey.
Not only was the sample size of 63 people too small, only 37 participants finished all 12 months. Forty-one percent of the participants did not complete the study, which further reduces the sample size, hence, increasing the sampling error. Furthermore, the results do not include the weight loss (or gain) of these individuals as produced by the two diets. Since the researchers measured only the compliant subjects, this is a potential source of bias. The results may be misleading.
The outcome would be more convincing if an intent-to-treat analysis, one that includes data from all the participants regardless of whether they dropped out of the study, was performed. A larger sample size comprised of all people, not just those who are obese, would further strengthen the results. But, experiments are needed to determine the long-term safety and effectiveness of the Atkins diet. Until then, it remains a fad diet that works questionably at best and is an expensive eating habit at worst.