| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Letters to the Editor |
Harvard School of Public Health, Department of Nutrition, 665 Huntington Avenue, Boston, MA 02115
Dear Sir:
The arguments of Bray and Popkin (1) overemphasize effects that are minor and quite possibly only artifacts, but in the end we agree on much. Their analysis of Chinese data indicating that a major change in fat intake is associated with a minuscule 0.03-point difference in body mass index (in kg/m2) supports rather than refutes my conclusion that dietary fat is not a major determinant of body fat. The unpublished abstract of a European trial that they cite purporting a 1–2-kg effect of fat reduction in obese patients is similar to other short-term studies, but longer-term studies show no progressive effect and even regression after 6 mo. Even if this difference in weight were maintained, it is not clinically important and probably not even perceptible or measurable in an individual obese person. Admonishing individuals to make major changes in lifestyle when the effects are so small puts our credibility at risk.
Bray and Popkin argue that it was "inappropriate" for me to present data only for the 6 studies lasting 
1 y. However, they combined long-term and short-term studies, even though they acknowledged that they showed different outcomes. When these studies are combined, the results are dominated by the much larger number of short-term studies, which leads to misleading conclusions about the long-term effects of dietary fat, which is the real outcome of interest. I specifically noted the report by Knopp et al (2) because it was large and lasted 1 y; it thus provides an independent test of the prediction from the Bray and Popkin meta-analysis. Their prediction was completely refuted. Bray and Popkin highlighted the reported 6-kg weight loss in the smaller subgroup of men with combined hyperlipidemia eating a 25%-fat diet. However, this is a typographical error (R Knopp, personal communication, 1999), which is implied by its inconsistency with the corresponding figure and the statement in the text that "greater fat restriction was not associated with greater weight reduction" (2). The correct weight loss was 3.2 kg, which is compared with 1.8 kg in the highest-fat group and 2.8 kg in the medium-fat group.
I did not address animal studies or short-term human studies of modified-fat products, both because of space constraints and the doubtful relevance to long-term effects of dietary fat in humans. If we want to know the long-term effects of reducing dietary fat in humans, the most relevant studies are long-term, randomized trials of dietary fat reduction in humans; these consistently show effects that are either null or very small.
Bray and Popkin concluded that "...our main argument is that the effect of energy density of foods affects total energy intake" and that a reduction in total energy intake "can be facilitated by reducing the amount of fat in the diet." Our exchange has moved us forward, because there does seem to be agreement that, within the range of realistic diets, differences in thermic effects between fats and carbohydrates do not seem to be important for weight control. They do not discuss but appear to accept the unimportance of the fat-partitioning hypothesis of weight regulation, which predicts that body fat will be proportional to the percentage of energy from dietary fat because only energy intake from carbohydrate is regulated. This is firmly refuted by the data from long-term trials that indicate that major changes in the percentage of energy from fat have little if any effect on body fat. However, if energy density is the central dietary factor in determining body weight, then the focus on fat alone is too narrow because it ignores carbohydrates, the major source of energy in almost all human diets. Although pure fat is more energy dense than pure carbohydrate, we eat these as complex foods. The food industry has shown that almost any food can be created in a low-fat version that is just as energy dense as the high-fat form. Without an equal emphasis on excess energy from both carbohydrate and fat, we will surely get more of the same.
Finally, Bray and Popkin's conclusion about energy density must be regarded as an interesting but unproven hypothesis. This should not serve as a basis for individual dietary advice nor policy without evidence from long-term human studies. I suspect the hypothesis may prove to be overly simplistic.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
