Abstract Thinking —
Trigger warning: Frank weight loss science talk and calories in, calories out.
One of the less vitriolic members of FCJ has finally sent along some studies in response to some thing I said in this comment. But since Steve only quotes the abstracts, he left himself wide open for criticism since he doesn’t actually say what’s in the study. But thanks to Kala, who sent me two of the studies (the third is out of print), I can comment.
What I said was “the more you restrict, the greater the results, but the less sustainable — the less you restrict, the less the results, but the more sustainable.” In response, Steve cites three studies. The first study he quotes says, “…the literature shows that initial weight loss is positively, not negatively, related to long-term weight maintenance.”
Fair enough. But if you actually read the first study cited, the evidence is rather flimsy. It supposedly shows that the more weight you lose initially, the better the long-term outcomes. Bear in mind, this is a review, not an actual experiment or a meta-analysis that looks at all of the data. The authors are providing their opinion based on a few select studies.
But all of the studies they cite are based on very low calorie diets (VLCDs). One of the studies compares 43 adults who are divided between an 8-week VLCD of 475 calories (2 MJ) per day to a 17-week “conventional” diet of 1,200 calories (5 MJ). Both groups were also given 20 mg of ephedrine and 200 mg of caffeine thrice daily. All participants were then randomized to a one year maintenance group that either prescribed an ad lib low fat, high carb diet or a fixed-energy diet of 1,850 calories (7.8 MJ) per day or less.
The results? Both groups lost an identical amount of weight after their respective time periods (28 pounds), but a year in these respective groups, a change had taken place:
Regained weight at follow up was greater in fixed energy intake group than in ad lib group [25 pounds] v [12 pounds], group difference [13 pounds].
Not exactly reassuring. And that’s after just one year post-weight loss. The real failure rates start showing up in data during that second year, continuing through five years. And considering that the ad lib (aka eat however much you want) group maintained a greater weight loss backs up exactly what I was saying.
The other study included in this analysis was a 5-year study comparing behavioral therapy (BT) to BT and VLCD combined. Researchers followed 59 morbidly obese patients and had follow-up data for 87% of them. The review that Steve links to mentions the weight loss, which found the following:
At the end of the treatment programme the mean weight loss was [20 pounds] in the behaviour therapy group and [51 pounds] in the combined group. Interestingly, the overall mean weight loss from pre-treatment to a 5 year follow-up was [11 pounds] and [37 pounds], respectively. Similar results have been reported by other investigators using gastric surgery for weight maintenance.
I don’t have access to this study, so I don’t know the exact caloric level, but VLCDs are typically 800 calories or less (aka starvation level). And, like I said, the only proven way to achieve significant weight loss is by restricting calories and increasing exercise drastically (i.e., The Biggest Loser). But as I said, and Steve quoted, “the more you restrict, the greater the results, but the less sustainable — the less you restrict, the less the results, but the more sustainable.”
The review that uses this study leaves on a very important bit of information that shows that what I said was exactly right: the retention rates. In the VLCD group, 56% of the subjects dropped out, while in the BT group, 28% dropped out.
So, twice as many people were unable to stick with a VLCD, meaning that for most people, such a small amount of calories is unsustainable.
Finally, the review’s main author, Arne Astrup, cites his own study of 40 obese women on a 1,000 calories (4.2 MJ) diet as evidence that the more calories you restrict, the more long-term success you’ll have.
Astrup et al. studied predictors of weight loss during a 36-week hypocaloric diet programme in obese women, and found that patients in the upper 50-percentile vs. lower 50-percentile of attained weight loss week 36 (Upper vs. Lower 50-percentile: –[39 pounds] vs. –[22 pounds]) had maintained more weight loss at follow-up 2.5 years later (–[16 pounds] vs. +[6 pounds]).
This is the only study that suggests that the more you lose up front, the less you gain back. But two things to bear in mind with this… this is not a randomized trial, but a review of 40 consecutive dieters. Hardly a rigorously controlled study. Secondly, after the 2.5 year followup, just 14 patients had maintained a weight loss of greater than 11 pounds.
An author can cherry pick his evidence (as opposed to a meta-analysis, which sweeps up all studies within a particular set of criteria), but he can barely make the case that losing a bunch of weight up front on a diet of 1,000 calories or less is the way to go. The study’s authors admit that the success is tenuous at best:
So even when the theoretically predicted weight loss for each patient in a given programme is the same, in real life there will still be substantial variations in weight loss between patients. The cause of these differences in patient responsiveness includes variability in physiological factors such as energy expenditure, physical activity and exercise, fat oxidation rates, and plasma concentrations of thyroid hormones and catecholamines, where high levels are associated with better short- and long-term weight loss outcome. Other hormones such as leptin levels may also be predictive of weight loss outcome. A high number of previous slimming attempts, depression, dietary disinhibition and binge eating is associated with relapse after weight loss and poor long-term outcome. Due to the numerous factors inﬂuencing weight maintenance, any speciﬁc beneﬁcial effect of a greater initial weight loss per se on long-term outcome needs to be assessed by randomized, clinical trials
The second study that Steve quotes says, “According to our review, successful weight maintenance is associated with more initial weight loss…” Sounds pretty clear cut. The only problem is that the study itself contains no hard numbers, no data, no definitions of “successful” to judge what the report’s authors are talking about. We’re just supposed to take it on the author’s authority that losing more weight is better… except, when you go into the actual report under the section titled “Weight loss patterns,” the relationship is more nuanced than the abstract suggests:
Initial weight loss has been identified as a predictor for later weight loss, and also for weight loss maintenance in various treatments… The greater the initial weight loss, the better is the subsequent outcome… According to other results, larger amounts of weight loss during the total intentional weight loss phase have predicted more weight regain.
Once again, it’s a mixed bag. But when you actually scratch the surface of this study, you start to see that it’s not exactly based on an impressive data set:
To enable a review of the literature using very different deﬁnitions of weight maintenance we have used a more inclusive deﬁnition of weight maintenance that implies an initial weight loss that has been subsequently maintained for at least 6 months.
So, this study relies on studies that include a definition of weight maintenance that lasts 6 months. Again, we’re talking actual, long-term weight maintenance of more than 1 year. But the methodological pitfalls get even worse:
Another methodological consideration is that the time of assessment for evaluation of weight maintenance can differ. Pre-treatment factors determining later weight loss maintenance have the greatest informative value for recommendations on treatment assignments. Sometimes the patients are instead assessed at the time of discharge from active treatment and further weight development is predicted from these characteristics.
Loose definitions of weight maintenance and inconsistent study criteria makes for weak tea at best. But what these studies do say, across the board, is that most weight loss attempts fail, as stated bluntly in Steve’s second study:
The long-term weight loss results in weight loss programmes are usually modest. Most patients who lose weight will regain the weight.
There is a big difference between living on starvation calories and restricting calories, and it’s disingenuous to conflate the two. Often what is required isn’t ‘restriction’ in any real sense, but simply bringing calories into an optimal range.
The first study, which included actual quantifiable data, basically showed that the kinds of weight loss that I’m assuming FCJ endorses for us fatties (50 pounds or more) is best achieved through a VLCD of less than 1,000 calories. That is starvation level.
If Aiden is correct, then tell me what the “optimal range” of calories is for losing 50 pounds. And can you point to a single, well-designed, controlled study that has ever demonstrated that your optimal range of calories has ever yielded long-term success for those who follow it.
This is the last post I’m writing on the subject of Fitness CircleJerk, unless someone from your braintrust can provide that information. If I’m wrong, I’ll apologize. So go ahead, FCJ: prove me wrong.