I’m kicking this post off with a disclosure: what you read today is heavy on speculation and light on evidence because…well…the evidence in this space is thin, at best. The fact that many of you read this blog for its scientific debunking of internet myths is not lost on me. But ever since receiving a reader’s email about effects of the ketogenic diet (KD) on the thyroid and basal metabolic rate (BMR), I just can’t get this issue out of my head. As a result, I’ve been spending wayyyyy too much time researching this issue, diving down all manner of rabbit holes. I’m hoping that laying out my thoughts on the blog will be cathartic, so I can get back to the work of writing about stuff that actually has clearer answers.
Below is a cut-up version of several emails I received from Clarice*, a reader with initial success on KD, who now is having problems with weight regain, hair loss, and “low” T3 levels. I’ll present her case study in chunks, so that we can address and digest the different points she raises:
I’ve eaten keto for nearly 2.5 years. After 40 lbs weight loss and landing at a beautiful weight just with the diet change, I stayed there and maintained effortlessly for 6 months. Then, with absolutely no change in diet after 1.5yrs, I crept up…pound, 2 pounds, 5 pounds, etc. The weight gain has been very stubborn and I have never been able to take a pound back off, and around this time also began general hair loss. The keto ‘guru’ I now follow has a non-standard approach within the keto community and preaches that things like hair loss and weight regain happen because we low-carbers often undereat and ‘damage metabolism and hurt thyroid function’ by eating too few calories…then regain as per recent media discussions of Biggest Loser contestants regaining weight after low-calorie dieting. I was probably consistently eating quite low-calorie (as keto is very appetite-suppressing and, hey, one only wants to eat so much steak and butter).
Weight Loss, Keto, and Your Metabolism
Let’s unpack a couple of things here. First, is it possible to damage one’s metabolism by eating too few calories? As with many things in Medicine, the answer is, “It’s complicated.” There are plenty of studies in the literature looking at the effects of starvation and hypocaloric diets on BMR. Overall, the evidence is fairly clear that BMR will drop, which is a whole lot of obvious. If you are not ingesting much of a calorie load, you don’t need to expend much energy to handle that load. Not only that, but you’ll be losing weight, so there will be less metabolically active tissue to support, resulting in a lower BMR. Plus, the body really hates to lose weight, so it will often ratchet your BMR down out of proportion to the reduction in the amount of lean tissue it is required to support. So far, I’m sure none of this is news to you.
But would you really call this “damaging” the metabolism, given that a lower BMR is a normal physiologic response to dieting and weight loss? I suppose if you’re gaining weight while trying to lose weight, you might consider a lower BMR to be damage, so let’s just stipulate that point as an issue of semantics and move on. The real question is: is the lower BMR permanent, i.e. are you screwed for the rest of your life because you consumed too few calories in your attempt to lose weight and get healthier?
The answer to that question appears to be: maybe. As there are plenty of bloggers who have written in detail about this issue, I will summarize. Most data I came across suggest that a lower BMR can return to its expected level over time – whether one is battling obesity, recovering from an eating disorder, or post-gastric bypass. However, “The Biggest Loser” study showed persistence of a lower-than-expected BMR in the contestants 6 years after wrapping the television show, despite significant weight regain.
Unfortunately, we don’t know why this suppression of BMR persisted. Some have speculated it was the rapidity of weight loss that shocked the participants’ bodies into permanently low-BMR territory; but, I read two randomized human studies that argue against this hypothesis – here and here. These trials showed that the rate of weight loss did not impact the likelihood of weight regain. Others think that there must have been hormonal or other metabolite abnormalities that account for the lower BMR, but none of the measured variables in the study bears that out (though the study wasn’t powered to detect changes of this nature). Bottom line: we just don’t know.
Getting back to Clarice, does she need to worry that her BMR is too low, resulting in weight regain? Perhaps. She has lost a fair amount of weight, which would result in a lowering of her BMR (again, that process of metabolic adaptation is normal). Let’s also assume that she adheres to KD strictly, meaning she’s in nutritional ketosis. If we buy into the theory that ketones (and dietary fats) are an excellent appetite suppressant, then she may not be ingesting enough calories to signal her body that it needs to raise BMR.
My coach has recently had the most success with ‘reverse dieting’ and keto, where he puts his clients on hypercaloric diets (high fat, low carb, just enough protein) and then drops them to their ‘TDEE’ calorie level and sees weight loss. He advocates this as a better way to lose weight without sacrificing metabolism than calorie-cutting, with more lasting weight loss.
On his plan, I have upped calories (very hard to be consistent!), and have not been able to lose an ounce, but have also not gained. Hair still coming out.
Reverse Dieting and Your Metabolism
I’ll admit, when I first read that her coach recommends eating more calories to lose weight, it raised my eyebrows. I don’t spend much time in the internet weight management space, so “reverse dieting” was a new term for me. In my world, where most people are not strictly following any healthy eating plan, eating more calories is a surefire path to weight gain.
When I looked at the medical literature, though, I discovered data showing that BMR can increase with an increased calorie load. This, of course, makes sense. What I haven’t been able to find is any non-anecdotal evidence that short-term increased caloric intake, followed by reversion to baseline intake, results in persistence of the higher BMR and weight loss.
If any of you employ this reverse-dieting technique regularly, I’m genuinely interested in your feedback. It seems to me that this would be a difficult needle to thread: you would need to increase calories just enough – for just long enough – to get the BMR up by enough to actually make a significant difference in overall energy expenditure. Then, you would have to cut calories back down to a level that results in enough of an energy deficit to lose weight. Then, as further metabolic adaptation occurs over time (BMR lowers again), you need to repeat the process – even though the targets you achieved last time may not work as well this time.
As a brief aside, those of you who have thrown the “calories-in, calories out” (CICO) model of obesity in the trash and replaced it with the insulin-carbohydrate model probably object to my focus on caloric intake at the end of the preceding paragraph. While I agree that the macronutrient composition of one’s diet is important, and I even buy into the idea that KD is one of the more effective extreme** eating strategies (sorry, vegans!), I don’t believe that “calories don’t matter.” If you consume an overabundance of calories compared to your output, even if they’re mostly from fat and protein, you’ll gain weight – simple. I honestly don’t understand how anyone can claim that calories don’t matter.
If my coach’s opinions about fasting are accurate, the LCHF (low-carb high-fat) community needs voices to explain the opposition opinion. Fasting is so trendy in the keto community now, and many, many women I know are chronically undereating all the time–one meal a day, and/or multiday fasts every week. I wonder if they’ll eventually start regaining as I have.
Intermittent Fasting Should be Intermittent
Intermittent Fasting (IF) is, indeed, gaining in popularity. I’ve even done a bit of Time Restricted Feeding (TRF) myself. Unlike Clarice’s keto guru, I think IF can be a useful tool in the weight management toolkit. Although clinical trials so far have not yielded impressive results compared to hypocaloric dieting, there is a suggestion that the weight lost with IF is mostly fat, whereas traditional diets lead to loss of both fat and muscle. Regardless of whether IF is a good thing to do, I suspect that Clarice’s coach and I would agree: if one chooses to do IF, it should be intermittent.
Why am I italicizing intermittent? I have noticed what seems to be a trend among IF enthusiasts – they eat progressively fewer meals. They may begin their IF journey with an eating window of 14-16 hours per day. Then, they move on to fasting until dinner a few days per week. Then, they’re fasting from dinner-dinner every day. Finally, they toss in a multi-day fast every month or so on top of that. This is purely speculation, but I think that too much fasting may be detrimental to BMR. It is my contention that this will ultimately cause many fasting enthusiasts to gain weight, particularly those who gradually move back towards a regular eating schedule, consuming more calories than they can burn.
Let me emphatically state that this is my opinion, and it is 100% possible that I am wrong. I am not the Yoda of obesity. And I’ve got news for you – neither is anyone else. That’s why there are scores of theories and hundreds if not thousands of books out there about The Way to lose weight – nobody has it all figured out.
So, my wonders about thyroid and the effect of the diet. TSH and T4 for me are normal, but free T3 is very low, rT3 is high (I went down the alt. med. rabbit hole far enough that I asked for these tests). Endocrinology says I’m fine, alt. med. says I’m severely hypo ;} With no real fatigue, not feeling cold, etc. it doesn’t feel right that I am truly hypothyroid, but the hair loss and weight changes are VERY frustrating and stubborn. So, I’ve been trying to figure it out. My guru/coach does believe in the T3/rT3 stuff, and says he’s seen the ratio improve in his clients when they increase calories (even without increasing carbs). After going to an Endo who dismissed T3/rT3 practically with an eye roll, I’ve been suspicious, which is why I found your articles so valuable.
Keto and Your Thyroid
So now we’re getting down to the good stuff. Does keto really wreck your thyroid? If you read the internet, it’s littered with personal stories of low T3 and high reverse T3 (rT3) on KD. You know my position on the general utility of rT3 levels, but has anyone actually studied this phenomenon? It turns out the answer is, yes.
The role of dietary fat in peripheral thyroid hormone metabolism, a human study published in Metabolism in 1980, is a good representation of what I found about this topic. Instead of going through the study in excruciating detail here, I’ll simply highlight the salient results. First, subjects eating a 100% fat, 1500 calorie diet exhibited a big drop in T3 and a big increase in rT3. In fact, these changes were about equivalent to what we see during total starvation. Wow, right? Does this mean that the thyroid is literally starving – that it can’t use fat to function?
Before tackling that question…in a different study, an 800 calorie diet of 100% carbs didn’t show any change in T3 or rT3, whereas the isocaloric, no-carb subjects showed a significant drop in T3 (with no change in rT3). Isn’t that fascinating? Eating very few calories (800!), as long as those calories were from carbs, didn’t do a whit of damage to thyroid levels. But eating 800 calories on a no-carb diet? T3 tanked.
So far, it’s not looking too good for KD. Think about it: the 100% fat diet from the first study – containing twice the number of calories as the 100% carb diet from the second study – resulted in low T3, while that measly 800 calories of carbs allowed the thyroid to happily chug along. I realize we’re comparing subjects from different studies, which isn’t kosher, but I can tell you that I found several studies all showing fairly similar results.
Does Dietary Fat Lower T3, Even in the Presence of Carbs?
Now, here’s the really interesting part of the puzzle: going back to that first study, subjects eating 1500 calories comprised of 50% fat and 50% carbs showed a significant rise in rT3 with a non-statistically significant 24% drop in T3. Though the drop in T3 failed to reach significance, it seemed to be close, and keep in mind that the study was small.
Given that the second study showed no effect on T3 or rT3 with an 800 calorie all-carb diet, you might have expected the subjects eating 750 calories of carbs and 750 calories of fat to show the same thing. After all, don’t you just need a minimum number of carbs to keep the thyroid humming? Maybe not, because the presence of 50% fat in the diet appeared to cause a trend toward lower T3 levels.
So, does KD cause hypothyroidism? A 2017 study in children treated with KD for refractory epilepsy reported that 20/120 patients developed subclinical hypothyroidism (slightly elevated TSH, normal FT4, no symptoms). Unfortunately, this study had so many weaknesses that I’m not sure how much weight to give the findings. And it’s very difficult to argue that we can apply data from kids on antiepileptic drugs to healthy adults on KD. Nonetheless, it’s provocative that such a high percentage of their subjects developed thyroid dysfunction.
HD, we’re begging you…imploring you…does KD cause hypothyroidism or not?
The Ketogenic Diet does not Cause Hypothyroidism
You really want to know my opinion? I’m willing to share it with you, as long as you understand that this is my opinion, and I could be wrong. I agree with Dr. Stephen Phinney of Virta Health, who wrote a great, evidence-based post in 2017, postulating that low T3 in KD does not reflect hypothyroidism. You should read that, as I’m not going to rehash his entire argument here. What I will do, however, is propose an alternate explanation for the low T3 from the one Dr. Phinney proposed.
Dr. Phinney wrote that KD likely increases tissue sensitivity to T3, much like it increases insulin sensitivity. Therefore, it takes a lot less T3 to get the job done. This is certainly plausible – even probable – but I’d like to suggest an additional mechanism that may also play a role. Could it be that consuming a predominance of fat in the diet is simply more metabolically efficient than consuming a predominance of carbs and protein? After all, 1 gram of fat has twice the potential energy (9 cal) of 1 gram of carb or 1 gram of protein (4 cal).
I realize that there are a ridiculous number of inputs when it comes to how nutrients are metabolized, which is why we generally measure BMR and not individual cellular metabolic processes. But is it possible that, when it comes to thyroid hormone’s role in the metabolism of nutrients, it simply takes less T3 to liberate energy from fat than from carbs/protein? In addition, there may be other metabolic pathways that are more prominently involved in the breakdown of fats, while T3 could be a more important regulator of carb and protein metabolism (again, this is speculative on my part and I could be wrong).
Remember the study I mentioned earlier – the one showing a substantial (but not statistically significant) drop in T3 in the subjects on a 1500 calorie diet comprised of 50% fat and 50% carbs? The study duration was very short and the carbs were too high for the subjects to be in ketosis. Therefore, Dr. Phinney’s opinion about ketosis leading to enhanced T3 sensitivity would not hold up in this particular situation. But my theory – as simplistic as it is – would better explain the lab “abnormalities.”
It is my contention – presumably in agreement with Dr. Phinney – that the normal reference range for T3 was never designed to apply to people on KD. How could it, since most people don’t restrict carbs to that degree? The lower levels seen in these people are likely normal for people on KD, especially in the context of normal TSH, normal T4, and a lack of signs/symptoms of overt hypothyroidism. I really can’t stress that enough, by the way – if everyone on KD had hypothyroidism, we’d be seeing a lot more overtly hypothyroid people with numerous symptoms, and their TSHs would be high.
So, KD enthusiasts, enjoy your steak, your avocado, your cheese, and your full-fat yogurt. Let nary a crumb of baguette, a grain of rice, nor a slippery noodle pass your lips in the pursuit of higher T3 levels – you don’t need them.
*Clarice is a pseudonym.
**My good friend Dr. Scher, the Low Carb Cardiologist, hates when I use the “E” word (extreme) to describe the ketogenic diet. But I love to needle him, and I can say whatever I want on my blog, so…tough noogies, Bret.
I welcome your comments below, but note that I will moderate the Comments section more aggressively than usual. I’m not interested in litigating the validity of KD as a dietary strategy; we already did that in the Comments section of Ketogenic Diet – Diabetes Cure? Similarly, if you’re a proud, card-carrying vegan, please refrain from trying to persuade us that your way is the best way. I agree with you that vegetables, fruits, nuts, seeds, beans, and all the other good stuff that comes out of the earth is awesome – but it’s not quite enough to make a nutritionally complete diet. If it was, you wouldn’t have to take supplements like B12, omega-3’s, iron, etc. You may disagree with me, and that’s fine, but send me an email instead of making your points here. I want us to focus on how keto and fasting affect metabolism and thyroid function. Thanks to everyone for reading and commenting.
By reading this site and interacting with me in the Comments below, you agree to abide by my Disclaimer.