I'm surprised this has been so highly cited, although it is an interesting paper. I was only a co-author on this paper, which was nurtured by my colleague Dr. J.R. Despres. We showed that waist circumference was as good if not better an estimator of internal abdominal fat than sagittal diameter, but the difference between the two is really small. Abdominal circumference seems to perform a bit better and is simpler to obtain, but for clinical reasons we said we better use waist circumference. It's like body mass index in that it’s only a surrogate for the amount of adiposity but it works pretty well in a clinical setting.

  Considering your surprise, what would you consider your most influential work?

In obesity, that would probably be a paper in the New England Journal of Medicine, an overfeeding study with identical twins, published in 1990. ("The response to long-term overfeeding in identical twins," NEJM 322[21]: 1477-82; 24 May 1990.)

  What was the purpose of that research and what were the results?

What I was looking for was an experimental demonstration that genes were involved in the response of the human body to an affluence of calories. We had heritability estimates based on population studies. I was looking for an experimental way to establish whether genes are involved or not. Here we were fortunate enough to have the collaboration of 12 pairs of identical twins, all young males in their early 20s. All were lean, with no family history of obesity, so they were not considered at risk for obesity. We took two weeks to assess what they were normally eating, while remaining weight-stable, and then we measured them for many phenotypes and many metabolic markers. Then we began an overfeeding experiment. We isolated them in a section of a dormitory that had been reserved for us on the campus at Laval University, in Quebec City, and we fed them 1,000 calories per day more than what they would eat normally. We did this six days a week. On the seventh day, they ate the baseline diet. We did that for 100 days. So overall there were 84 days of overfeeding, and each twin had to handle an extra 84,000 calories. We kept them sedentary, as well. It was a very complex study. It turned out that they gained weight, of course. All of them. The average weight gain was 7 kilograms.

The remarkable thing was that there was a huge range, a three-fold range, between high gainers and low gainers. Some twins gained significantly more weight than others under the same challenge. But we had three and half times more variation between pairs of twins than we had within members of the same pair. That suggested that being genetically similar meant a more similar response to caloric overload. And that was a pretty strong demonstration, the first in humans, that genes make a difference.

  Did you follow up on the experiment to see whether they retained the weight and whether that, too, had a genetic component?

We did, but that was not in the New England Journal of Medicine article. It was published later in Metabolism. We re-measured these twins, four months after the completion of the study and then five years later. Four months after the end of the study, they were all within one kilogram of their baseline. Even the high gainers had reverted back. I think this is largely because they went back to their natural environment, to their friends, their old habits, and fell into the same groove and their body weight was at that level within that groove. Five years later, they were still at their normal weight. We had the same resemblance within pairs of twins. They had gained a little bit as you expect when they age. But they were perfectly normal otherwise. We also did the opposite experiment subsequently: losing weight for overweight or obese identical twins and we recovered exactly the same phenomenon. That paper was published in Obesity Research.

  Why did you choose the New England Journal of Medicine for the first paper, Metabolism for the second, and Obesity Research for the third?

The first one was of very general interest, because it touched upon the predisposition to gain weight in the population. That's why I submitted it to the New England Journal of Medicine. The second dealt more with metabolism. We were asking questions not only about weight but also about blood pressure, glucose intolerance, etc. That's why I elected to send that to Metabolism. As for the third one, Obesity Research, I'll be frank with you: I was involved in the founding of this journal. I decided to provide support to the journal and I submitted it there. It has also been quite highly cited.

  What has been the biggest challenge in your obesity research?

It's associated with the fact that we are trying to identify the genes and mutations responsible for bestowing a predisposition to obesity. When I began this research 20 years ago, I thought that only a few genes would be involved. If you could identify which they were, you could predict who would become obese or at least who was at risk. Now it turns out that this predisposition seems to be determined by a much larger number of genes and each seems to have only a small effect. It becomes considerably more challenging to undertake the genetic dissection of this predisposition for obesity. If you have 30 genes and each contributes one, two, or three percent of the risk, then your experiments are usually not powerful enough to detect them. And you miss them or find them and think their effect is so small that it’s probably not important. In reality, it may be important and these effects may be amplified if you have several of these risk genes together in the same individual. But it is very hard to draw an accurate picture of these genes at the moment. It's been very frustrating. My lab has been working on this for quite a while. We've probably looked at 50 different genes and we're still not able to draw an integrated picture of which genes are making a difference.