Take a group of obese people (> 250 lbs) and put them on a massive calorie restricted diet. They lose weight and metabolic rate plummets. Weight loss fail? In most cases, yes. But a recent study showed that the decrepit post-weight loss metabolic rate gradually improves in parallel with an increase in dietary fat ingestion to such a degree that even after two long years: totally food intake was almost back to normal, energy expenditure improved, and all this happened despite continual weight loss. In other words, obesity is not permanent.
Decreased energy density and changes in food selection following Roux-en-Y gastric bypass (Laurenius et al., 2013)
Statistically speaking, no diet on Earth comes close to RYGB in terms of weight loss success. Long term. Seemingly permanent. It’s the closest thing to a cure we’ve got.
Body weight is down by 30%, and energy expenditure is rising faster than a speeding bullet. because food intake is increasing while body weight is dropping – they’re probably more active too ** weight loss than more exercise –
But there’s a more mystical aspect to RYGB
Leptin dose was 0.08-0.14 mg/kg fat mass & adjusted as necessary until plasma leptin levels reached pre-weight loss levels. Baseline: 23 ng/mL; weight loss: 16.8 ng/mL; weight loss + leptin: 24.8 ng/mL.
But maybe energy expenditure crashed because of the diet? It was quite possibly absolutely the worst diet anyone could intellectually conceive: 40% corn oil, 45% glucose, and 15% casein (Leibel, Rosenbaum, and Hirsch, 1995). 
OK, now my money is on diet. If you lose weight via corn oil & glucose, then leptin levels plummet while leptin sensitivity becomes exquisite. And so leptin injections work. In the Laurenius RYGB study, we don’t really know what they were eating, but it wasn’t worse than 40% corn oil, 45% glucose, and 15% casein. obviously.
Which brings us here:
Randomized double-blind placebo-controlled study of leptin administration after gastric bypass (Korner et al., 2013)
4.5 years after RYGB surgery in patients who dropped 30.8% of their body weight and were clinically hypoleptinemic – ie, they’re leptin levels were 21.6 ng/mL, whereas a similar BW-matched cohort weighing 94.9 kg was at 33.5 ng/mL. Leptin injected at 0.05 mg/kg BW BID = super duper physiological leptinemia of 234 – 281 ng/mL. It was a crossover study, so here are the data for those who got placebo first (left) and leptin first (right):
but it didn’t work all that well :/
What went wrong?
Lesson 1: they bollixed the crossover. There was NO WASHOUT PERIOD. But this would’ve worked in favor of leptin. Skip ahead if you are fluent in Prelude to a crossover and part deux. In those assigned to placebo first, nothing would’ve happened in the placebo phase, then they would’ve lost weight in the leptin phase. Leptin +1. In those assigned to leptin first, they would’ve lost weight in the leptin phase, they regained it in the placebo phase. Leptin +2, placebo -1.
This actually appears to have happened – see the small rise in body weight in the placebo group? That’s the sign of a bollixed crossover.
Lesson 2: leptin resistance. It took a few weeks for leptin levels to accumulate to such high levels. It’s possible that by week 8 (150 ng/mL) leptin was working, so weeks 8-12 they lost weight. But 150 ng/mL is too high. Leptin resistance started kicking in and by week 12, when leptin levels were pushing 250 ng/mL the result was full blown leptin resistance. Leptin stopped working. The data support this possibility.
Alternatively, maybe they were just eating better than 40% corn oil, 45% glucose, and 15% casein, and weren’t starving. And maybe the RYGB cured them, so leptin injections worked as well in them as it does in regular people. Ie, not at all, unless violent doses of leptin are injected (0.3 mg/kg, which is 6 times higher than the dose which increased plasma leptin levels 10-fold in Korner’s post-RYGB patients). (although this would appear to negate the leptin resistance theory [shhh!]).
From Heymsfield 1999:
Diet alone = paltry weight loss, rarely successful probably because of dire need for leptin injections.
RYGB
Lesson 3: 40% corn oil, 45% glucose, and 15% casein = epic FAIL
