Carbs: low is good; is *lower* better? this study suggests “nope” http://t.co/5MATDl8Msr
— Bill Lagakos (@CaloriesProper) February 27, 2015
This was met with much backlash from the low carb cavalry, because, well, if low is good then lower must be better…
I’m not anti-keto; but I’m not anti-science. FACT.
“…some people are not genetically equipped to thrive in prolonged nutritional ketosis.” –Peter Attia
It’s not a bad study.
Some of the criticism of this study included small sample size; that’s wrong because #context: you don’t expect many more than 20 participants in a study where all food is provided for 6 weeks. $$$
The prescribed diets:
And in typical Atkin’s fashion, carbs were slowly titrated back (5g/wk) in the keto group (which may partially explain why ketones declined as the study progressed).
The participants were obese & sedentary, and according to HOMA-IR (a measure of insulin resistance), they were insulin resistant. Why is this important? Because in general, obese insulin resistant patients respond better to low carb than low fat diets, whereas the reverse is true for insulin sensitive patients… but what about low carb vs. lower carb?
Apparently, it doesn’t matter — the difference between groups was not significant, even though low carbers appeared to lose more fat mass than keto dieters (NS).
Further, even though keto dieters were getting slightly more protein, neither this, nor the ketones, spared muscle. BEHOLD ->
The only statistically significant difference between groups in the above table was blood ketone levels (duh)… and while it wasn’t a massively deep ketosis (0.333 – 0.722 mM), it still counts… anyway, higher blood ketone levels obviously do not guarantee better fat loss.
So, what gives?
1. I don’t think this finding is particularly surprising (or wrong), because: 1a) we already know low carb works well in this population; and 1b) “if low is good then lower must be better” is low-level thinking. Really.
2. Food intake was well-controlled. This is an important point.
Low carb usually wins against low fat. So why is this pill so hard to swallow? 42% carbs in a pretty big energy deficit is modestly ketogenic (~0.2 mM), and still conducive to the loss of body fat without causing hunger. Some might argue that 157 grams of carbs isn’t Low Carb™.
- WHAT ARE YOU EVEN ARGUING? That the low carb diet was really low fat, and therefore low fat beat keto?! O_o …MAYBE if they were insulin sensitive; but they weren’t.
- 42% carb sounds like a lot, but 157 grams isn’t a lot for an obese adult on 1500 kcals, who is probably burning ~3000 kcal/d… they were still mildly ketotic [~0.2 mM]. Energy balance must be considered when deciding whether a certain amount of carb is “low.” Context always matters.
Both diets were insulin-suppressive and reduced hunger. Both diets caused fat loss. Taubes 101.
Another critique of this study was that it is grossly underpowered — a potentially legitimate concern; however, “underpowered” is a specific statistical term which means, roughly, that the study failed to detect a real difference, usually because of small sample size and/or large variability. In this study, the authors were looking for changes in LDL cholesterol, so that’s how they determined sample size and duration. They found no difference (remember: this wasn’t low carb vs low fat; it was low carb vs lower carb). This doesn’t mean the study was underpowered; it could simply mean there was genuinely no difference. And as far as fat mass is concerned, they didn’t find a statistically significant difference there, either… but looking at the figure below, if they had found a significant difference, it would favor low carb over keto:
Another critique is that the study was run by Barry Sears, who has skin in the game. He was essentially testing his diet (an omega-3 rich low[ish] carb diet). But unless you can find inherent flaws in the study, you’re saying he falsified data. Gravitas. That’s a bold accusation.
Another critique was the use of BIA for assessing body comp. OK, for only 20 subjects, maybe they could’ve used a more expensive and time-consuming method. But to say it’s inaccurate… 1) this study wasn’t done in ’80’s. Modern BIA devices are far more accurate than the old ones. And 2) most of the criticism cites plasma volume shifts and changes in hydration status that are known to occur during carb restriction — a potentially valid criticism — but this wasn’t low carb vs. low fat… both diets were carb-restricted.
Protein and calories were controlled… two important variables. Eg, “Body-weight loss and weight-maintenance depends on the high-protein, but not on the ‘low-carb’ component of the diet…” (Soenen et al., 2012) <- that study seems relevant here.
In my opinion, one of the biggest strengths of carb restriction for weight loss is the spontaneously reduction in food intake seen in obese insulin resistant patients. That was removed from the picture in this study by strictly controlling food intake… sure, the participants could’ve cheated, but you can say that about any study, which pretty much invalidates all of the studies.
A keen Redditor pointed out that the sum of saturated, monounsaturated, and polyunsaturated fats was less than the total fat in Table 1. OK, but if the values are computed with the individual fats instead of total fat, not much changes – and the keto diet ends up with fewer calories than low carb…
In real life, under ad libitum conditions, obese insulin resistant patients seeking weight loss will be well-served by a low carb or ketogenic diet (and an education about diet and lifestyle). But that may be due to the other dietary changes: spontaneously lower food intake, [usually] more protein, etc. Not magic. And in real life, not everyone even knows what a carb is, so recommending “very low carb” will probably result in “modestly reduced carb.”
Don’t be a science-denier! In this study, low carb beat keto, and there were no major flaws (that I could find).
A seafood-rich low carb diet is a great start — whether or not it’s actually ketogenic may not matter as much — the more important point in this context, is that insulin resistant patients will benefit from carb restriction.
In a recent Tim Ferriss podcast, Peter Attia further explained why he abandoned keto and opted for low carb instead. Yeah yeah yeah Attia doesn’t have obesity. But still… since he can obviously [easily] adhere to a ketogenic diet long-term, I don’t think he’d make this switch if he thought it was unhealthy. I concur; they can both be healthy diets. (and no, I don’t think Attia’s decision says anything meaningful about LC or keto.)