Study 1: No differences once weight controlled
Overweight women do not have any significant differences from normal weight women in terms of REE (resting energy expenditure), once weight has been normalised, according to new American research.
Subjects: 49 women who were previously overweight but were now normal weight were compared with 49 women who had never been overweight. All subjects were pre-menopausal, none had abnormal glucose and all lived sedentary lifestyles.
Method: Longitudinal observational study. Researchers measured energy expenditure and fuel use (using chamber calorimetry) and body composition (using DXA) at baseline and again at 1 and 2 year follow-up.
Results: After adjustment for both lean and fat mass, there were no significant differences between either the sleeping or resting energy expenditure or the 24 hour fuel use of the two groups (p > 0.25). Nor was there any significant statistical association between these variables and patterns of weight gain at follow-up.
Reference: Obes Res. 2003 Aug;11(8):937-44.
Study 2: Differences in obese children
There are differences in REE and metabolism of individual macronutrients between obese and non-obese children, according to recent Spanish research.
Subjects: 39 obese children and 32 controls, aged from 4 to 13 years.
Method: Observational study of REE (open circuit indirect calorimetry), body composition and substrate metabolism (calculated from oxygen consumption, carbon dioxide production and nitrogen excretion).
Results: REE was higher in obese children than normal children after adjustment for lean body mass, as was percentage of fat oxidation, whereas percentage carbohydrate (CHO) oxidation was lower.
Reference: An Pediatr (Barc). 2003 Apr;58(4):316-21.
Study 3: Change in REE after weight loss surgery predictable
New findings from the US show that the change in REE seen in patients who have had massive weight loss after barosurgery is predictable based on their loss of fat mass.
Subjects: 20 extremely obese adult women and men (mean BMI= 50.1).
Method: The subjects were tested before and after stabilisation following weight loss surgery, which occurred after an average of 14 months, and following weight loss of an average of 53 kg.
Results: There were significant decreases in REE after surgery (-2.4 MJ/d; p < 0.001), but this was predicted by changes in lean and fat mass and measured REE was not significantly different from predicted REE.
Reference: Am J Clin Nutr. 2003 Jul;78(1):22-30.
Perhaps by identifying patients with such abnormalities it would even be possible to tailor treatments more effectively using medications that might safely reverse changes in REE. For example, there is some evidence that leptin can do this (ref.4), although a recent clinical trial failed to confirm this (ref.5).
Unfortunately, it has proved very difficult to establish whether alterations in REE are or are not significant in the overall clinical picture of obesity.
In part this is because REE is notoriously hard to measure accurately outside of a highly artificial laboratory environment (hence the use of various predictive equations for it). REE also depends to a significant degree on body composition, particularly fat and lean mass, which of course are the very things that change in obesity and its management.
The three new studies summarised above illustrate some aspects of the state of the science in this area. They highlight the wide variety of situations in which REE measurements might be taken - adults vs children, after moderate vs drastic weight loss etc. They also use a variety of methodologies and ways of adjusting the REE data for body composition. They come to different conclusions.
A meta-analysis published in 1999 on studies of REE in obesity after weight loss included data from 124 formerly obese and 121 control subjects (ref.6). After adjusting for body composition, the formerly obese subjects had a 4.1% lower REE than control subjects (95% CI: 0.2-8.1%, p = 0.04). This difference was due to a sub-set of 13% of the formerly obese who had low REE, compared with only 3% of controls (p<0.003).
The authors of this meta-analysis were careful to point out that the cause of this was far from certain. Were the differences part of the underlying mechanism of obesity, or a consequence of subjects having successfully lost weight (which usually includes some loss of lean mass)? A third possibility is that the lower REE was due to the fact that some of these subjects were still in negative energy balance, despite having stable weight at the time of study.
Although this is a very complex subject, we do not intend to undertake an extensive literature review here, other than to agree with the thrust of an editorial which accompanied the meta-analysis we referred to above. That editorial concluded that there was very little convincing evidence so far to show that defects in energy expenditure explain either the development of obesity or the problems overweight people have in maintaining long term weight loss.
This does not mean the question is settled by any means. But we need to be very wary of promoting the idea that people who find it difficult to lose weight are `slow energy burners'.