Search This Blog

Monday, 2 May 2016

A brief reading of the report of the new Kevin Hall study

Is the insulin theory of obesity over? Well I'd say it's over when people with diabetes using exogenous insulin to cover high-carb diets no longer have to worry about weight gain, and not before. But this is interesting research, and it proves if nothing else that Gary Taube's NuSci research initiative wasn't just set up to confirm his every thought.Kevin Hall discusses his new study here, with Yoni Freedhoff.

This research contradicts two things - one, the metabolic advantage theory of ketogenic weight loss.
It never made sense to me that wasting energy would make it easy to lose weight. That's "run off that coke" type illogic. The only thing that makes obese people lose weight sustainably is the repair of the appestat. The LCHF diet is great for this because without the carbohydrate foods that stimulate cravings, and with a belly full of fat, it's easy to get eating right again.
The second thing it contradicts is Taube's statement - more a guess or rule of thumb than a hypothesis - in GCBC that all weight loss diets that work, work because they restrict carbs and thus lower insulin. This is how a lot of weight loss diets work, but this study shows that, if calories are held even, the rate of weight loss needn't be proportionate to the lowering of insulin in every diet phase.
It also contradicts the idea that a ketogenic diet causes significant muscle loss. This happens at first, to a small amount, then it's reversed. It's not an ongoing problem that results in people wasting away - and these subjects were in a metabolic ward, so very limited in how much exercise they did.
There are two things the study does not do. It's an isocaloric comparison, so there's no test of which diet would have been more likely to cause free-living people to spontaneously eat and move the right amount to normalise weight. And it's not a study of weight gain, so says little about the metabolic and dietary conditions that made the subjects obese in the first place.
It is likely, but not clear from this report, that the subjects had lower insulin levels in both diet phases than they had while gaining weight or at baseline. If that is true, then the insulin hypothesis of obesity is doing just fine, but is in need of a little adjustment.
What's interesting to me is that what this study does say about LCHF diets confirms two statements in the 1950s and 1960s work of John Yudkin that I've been reading - there is no low carb metabolic advantage, and therefore they can only work as well as they do for weight loss if people spontaneously tend to eat the right amount when eating fat and protein.


PhilT said...

We'll see when it appears - for example blood insulin and ketone data didn't appear in the discussion. My alarms went off when Kevin said there was a 500 calorie (per day ??) discrepancy between actual and expected energy expenditure in the metabolic ward (outside the chamber), which is why there was weight loss when it was supposed to be a eucaloric study.

As he's very wedded to the calorie theory I would be concerned this highlights 500 cals of error somewhere that he's just trying to dismiss as unexpected exercise in the metabolic ward. The trial registration said they were tracking activity.

Although overweight and obese I think the screening took out diabetics. The devil will be in the detail. I would like Kevin to produce a mathematical model of the acute processes of glucose and fat disposal and storage over a timescale up to 10 days. If he can crack that we might learn from it.

Puddleg said...

I's like to know a bit about the diet too - the more the better. How close was it to a real-world LCHF intervention? While this may not have affected readings in the chamber much it might well have effects on both satiety and insulin response if an experimental diet isn't based around real food.

Puddleg said...

Like everyone who commented on this, and like Kevin Hall himself, I spoke too soon.

Jason Fung shows that the metabolic advantage is relative

Michael Eades finds a few problems with the way Hall presented the results - where's the first week's data?

karl said...

As I've said elsewhere - I think of insulin as a feedback loop signal - not useful without knowing the level of insulin sensitivity.

I'm suspicious that PUFAs- driving inappropriate insulin sensitivity are the confounding factor. What if people going on low carb diets are eating less LA? ( Chips - breads - many carb snacks are loaded with LA ).

It always gets back to needing to do synthetic diet research - like Milton Winitz and others did in the 1960s.

What if low-carb diets are only compensating for the effects of PUFA on insulin sensitivity?

Puddleg said...

You mean like this?

I'd say you're right.