NEW STUDY Proves it's BMI not Butter that increases Cholesterol on Low-Carb Diets

I am very excited to announce the publication of a paper that may finally lay to rest the misperception that saturated fat is the culprit increasing LDL cholesterol to crazy high levels on low carbohydrate and ketogenic diets. The paper, published in the American Journal of Clinical Nutrition, is entitled Increased LDL cholesterol on a low carbohydrate diet in adults with normal, but not high, body weight a meta-analysis first authored by my good friend and colleague Adrian Sotomayor, MD, PhD, and a rockstar team of collaborators, if I do say so myself. A really special team. Anyway, let's dig in. Interest in low carbohydrate diets has recently grown not only for the treatment of obesity and type 2 diabetes, but also for other conditions not directly related to obesity, such as inflammatory bowel disease, mental health disorders, chronic kidney disease, and neurological diseases. However, adoption of a low carbohydrate diet has been limited clinically in part due to concerns for elevations in LDL cholesterol. LDL is kind of the boogeyman of low carb, a boogeyman of sorts. LDL cholesterol change in a low carbohydrate diet is highly variable. Most people actually do not see increases in LDL and some see decreases. But among those who do see increases, these can sometimes be astronomical, as in lean mass hyperresponders. The topic of prior publications. So a major question that must be answered to address this boogeyman of low carbohydrate diets is what causes increased LDL cholesterol on low carbohydrate diets? In other words, what is a major source of heterogeneity of LDL cholesterol change when people go low carb? Is it saturated fat? Is it genetics? Or Is it something else? Previously, we showed in a cohort of 548 people consuming a low-carb diet that body mass index was inversely associated with LDL cholesterol, i.e., leaner people had larger increases in LDL cholesterol in a dose-response manner. However, this prior study was limited by the self-reported nature of the dietary assessment and some potential selection bias. So in this new study, A systematic review and meta-analysis of low carbohydrate diet randomized controlled trials, with low carbohydrate diet defined as less than 130 grams of carbs per day, we sought to evaluate the relationship among LDL cholesterol change on low carbohydrate diets and BMI or obesity status. The results were striking. We included a total of 41 trials with 1,379 participants, and the mean intervention duration was about 5 months. The groups of trials were then, um, broken down as such. We grouped them by BMI classes. So we had a normal BMI group with BMI less than 25 kilograms per meter squared, an overweight group with BMI between 25 and 30, an obesity class 1 group with BMI between 30 and 35, and then an obesity class 2 group with with BMI between 35 and 40. What we found was that LDL cholesterol increased substantially in trials with a BMI in the normal range. The mean increase was 41.4 milligrams per deciliter increase, which is very high in those trials including normal weight or lean participants, lean relative to the, you know, public standard. For trials with BMI in the intermediate range, Um, so overweight or class 1 obesity, LDL cholesterol did not change. And in trials with class 2 obesity patients, patients with class 2 obesity, LDL tended to actually decrease. Thus, a key finding of this set of meta-analyses was to confirm that across randomized controlled trials of low carbohydrate diets, those on leaner subjects tended to show LDL increases, whereas those including individuals with overweight or obesity showed no change or decreases in LDL, as in those with more severe obesity. So this was a dose-response manner, i.e., as you go up the BMI classes, LDL changes go to zero and then actually turn negative, which is consistent with the lipid energy model, which I'll get to in a minute. But we didn't stop there. We were able to obtain individual participant-level data for 300, allowing us to perform regressions that confirmed on a low carbohydrate diet an inverse relationship between BMI and LDL cholesterol. Again, leaner people had larger increases in LDL, people who were overweight had no change, and those that had class 2 obesity actually had a decrease in LDL. And we went on to present a prediction-based comparison of the influence of BMI and saturated fat intake using these individual participant level data. BMI had a substantially larger influence than saturated fat on LDL cholesterol change. In fact, the effect of a normal BMI, so being lean, having a BMI under 25, on LDL was over 5 times as large as an effect size as being in the top quartile of saturated fat intake. The effect of just being lean had an impact of about 48 mg/dL, whereas being in the top quartile for saturated fat intake only had an impact of about 9. Mg/dL. So point being, BMI dominates over saturated fat. So in summary, these results reveal an inverse association between BMI and LDL cholesterol change with consumption of low carbohydrate diet. Whereas lean individuals experienced marked elevations in LDL cholesterol, those with high BMI typically experienced no change or a reduction in LDL cholesterol. Furthermore, the importance of BMI in explaining response heterogeneity to low carbohydrate diets appeared much, much greater than that of saturated fat. But why? How? The lipid energy model, which has been, uh, overviewed in prior videos but which I'll briefly overview here, provides a mechanistic explanation for this fascinating inverse relationship between body mass index and LDL change. Briefly, according to this model, depletion of liver glycogen stores with carbohydrate restriction increases fat cell lipolysis. Fat cells break down stored fat, releasing free fatty acids, which get resynthesized into triglycerides in the liver, which are then exported on very low density lipoprotein particles, VLDL. This increased VLDL secretion from the liver, together with faster peripheral turnover of these VLDL, um, at fat and lean tissue, yields a lipid profile characterized by low triglycerides, high HDL, and very high LDL. This can be the lean mass hyper-responder triad, and I encourage you to follow the links below for more detail. In the lipid-energy model and the associated lean mass hyper-responder phenotype, they require further study, but from this paper, this meta-analysis, or these sets of meta-analysis, it's now quite clear to me that the culprit of high LDL cholesterol on a low carbohydrate diet is less so the butter and more so the six-pack. Isn't that interesting?