Friday, February 08, 2008

More Weight Gained Over 6-Years on Low-Fat Diet

Last week I posted about our collective belief, as a nation, that weight loss and maintenance is a problem solved by simple math - balance the "in" with the "out" columns of the calorie ledger and weight will be lost when the "out" exceeds the "in" and will be maintained when the two are equal.

Or so we're told.

I threw out the idea that weight isn't simply about math, but about chemistry.

The latest issue of the American Journal of Clinical Nutrition is out, and within is a very intriguing study where this very issue was central to the research conducted. It's one piece of the "weight is chemistry" equation, so let's take a look.

In the study, A novel interaction between dietary composition and insulin secretion: effects on weight gain in the Quebec Family Study, researchers followed subjects for six years to see what effect, if any, levels of insulin secretion had on weight.

The objective was to "determine whether physiologic differences in insulin secretion explain differences in weight gain among individuals consuming low- and high-fat diets."

This interest to investigate such was born from clinical trials where subjects consumed low-fat diets where results often revealed that there are huge variations between individuals consuming similar dietary ratios of macronutrients and reported calories; these differences are often chalked up to psychological and behavioral factors - motivation, compliance with diet, under-reporting of foods consumed, etc., ignoring the possibility that there is a physiological variation that may be at work.

In this study, the researchers noted the findings of Dansinger et al who compared four dietary approaches (Atkins, Ornish, LEARN and Zone), noting that "Of those assigned to the Ornish very-low-fat diet, mean weight loss was again small, 2 kg, but individual weight change ranged from almost –30 kg to >10 kg. This interindividual variation is commonly attributed to differences in motivation and compliance, but biological factors may also be contributory."

Thus the researchers in this study considered the the following for a test hypothesis: Low-fat diets are inherently high in carbohydrate because, for most people, the third major nutrient, protein, remains within a fairly narrow range. Carbohydrate has the most potent effect on insulin secretion of the major nutrients. Therefore, individuals with high insulin secretion consuming a low-fat diet might be especially susceptible to weight gain.

To investigate this, they "examined the associations between insulin concentration at 30 min (insulin-30) during an oral-glucose-tolerance test (OGTT) and change in body weight or waist circumference in the Quebec Family Study (QFS)" over a period of six years."

The findings were very interesting.

The primary findings of this study are that a proxy measure of insulin secretion strongly predicted weight gain and change in waist circumference over 6 y in adult whites, especially among those consuming lower-fat diets.

...our results and those involving GI suggest the existence of a unique physiologic phenotype that responds poorly to high insulin-stimulating diets, regardless of whether these diets are high in carbohydrate or have a high GI. Moreover, the combination of a high-carbohydrate and high-GI diet—that is, a diet high in glycemic load (GL)—may produce especially great weight gain among individuals with this phenotype.

While the researchers reached the conclusion that these findings suggest there exists a "unique physiological phenotype that responds poorly to high insulin-stimulating diets" - such assumes we're supposed to consume a low-fat, high carbohydrate diet; if we're not responding well to such a diet, it's because some have a unique physiological phenotype rather than consuming a diet that is at odds with human physiology.

I'd counter that it's just as likely it has nothing to do with a "unique phenotype" among us, but rather the expected results of a modern diet at odds with our physiology. Minor quibble.

Nonetheless, the finding is important as it points directly to individual metabolic factors playing a role within the context of the diet - low fat verus high fat - not just calories. In this study, it was insulin secretion. It mattered, especially in those consuming a low-fat diet.


  1. Anonymous12:50 PM

    Regina, is it possible that this study would allow us low-carbers to present ourselves as a "niche market" in the world of weight-loss diets?

    Perhaps we could persuade our doctors to check our insulin concentration at 30 minutes during an oral-glucose-tolerance test as they did in the study. If the findings are correct, we might be able to characterize ourselves as being in a population that responds poorly to high insulin-stimulating diets. That way we wouldn't be threatening an almost religious belief in the idea that a calorie is a calorie, but we could be placed into a special category that requires an unusual type of intervention, i.e., a low-carb type of weight reduction diet.

  2. Anonymous6:24 PM

    Please also add to your comments that the mortality observed in this study was actually LOWER than the general mortality in the general population with similar makeup as the ACCORD study. i would be very careful to make assumptions without looking at the trial data first.

  3. Actually I think the recognition that not everybody responds to a given diet the same way, and the need to individualize prescriptive diets based on metabolic factors is a huge step in the right direction, if it happens.

  4. Anonymous10:04 AM

    If true, this diet-phenotype interaction might explain why some of us have to go close to zero carbs to maintain our weight loss, while others can eat between 60 and 100 grams of carbohydrate a day and still maintain.

    The almost-zero-carb group might secrete a relatively high level of insulin in response to carbohydrate ingestion, while those who can tolerate more carbs secrete a relatively lower level of insulin.

    This would be in line with the observation that animals with their ovaries or testes removed have higher serum insulin. It could explain why our carb tolerance seems to decrease as we get older.

  5. I'm a little late on this, but I might suggest that the "unique phenotype" exists in the minority of people who seem to be able to process a high-carb diet *without* accompanying glucose/insulin problems that the rest of us experience to one degree or another.