"That's Terrible!" | What Statin Trials Really Show

What if the test your doctor trusts the most is lying to you? Well, hold that provocative question in your mind while I tell you a cute little story. When I was a kid, some Monday mornings I really didn't want to go to school. So I'd make a mug of hot chocolate. Like, really hot chocolate. Then I'd take a sip, time it perfectly, and ask my mom to take my temperature. "Oh no, I have a fever!" A few hours later, of course the fever was gone because I was perfectly healthy, but I'd already received the prescription I wanted: no school. Now, adorable anecdotes aside, and please don't tell Harvard, here's the point: sometimes measurements mislead us. And today's suspect, it's not hot chocolate. It's LDL cholesterol, the bad cholesterol. Medicine loves LDL. It's easy to track. Easy to change. The mantra goes, "The lower the better." But what if LDL cholesterol, specifically the drop in LDL cholesterol people experience when they take medications, isn't what we think it is? By the end of this video, I promise you'll know why LDL might not be the thing to focus on, what actually predicts heart outcomes, and what to ask your doctor at the next visit. Let's get into it. LDL cholesterol drops in statin trials were horrific, terrible. Can you guess the predictive power of LDL cholesterol reductions on cardiovascular outcomes? Ready for the answer? This is terrible. Am I saying statins are bad or entirely useless? It's complicated. So what? What is the takeaway? What do I need to know today? So here it is. Now, to make this video more digestible, I've broken it up into a few chapters. Chapter 1 covers 20 randomized control trials and about 200,000 patients. It's about a new meta-analysis in the European Heart Journal that basically suggests the association between LDL cholesterol reduction magnitude and actual cardiovascular outcomes is incredibly weak. So let's unpack that. But first, a really quick plug: if you learn something from this video, if you enjoy it, don't forget to check out the associated Stay Curious Metabolism newsletter linked below. That's where we unpack even more nuances, more contradictory studies, the complexities, to find out why we've rapidly become number 3 overall bestseller in science globally on the entire Substack platform. All thanks to you. Anyway, let's get into this study. This was published in the European Heart Journal. It was an umbrella review of meta-analyses of human randomized control trials comprising 20 human RCTs and 194,686 participants with a median follow-up period of almost 5 years. And the researchers set out to answer a simple but really important question: Is the reduction people experience in LDL cholesterol a valid surrogate when it comes to predicting clinical outcomes? So let's define surrogate. That means a substitute marker that can stand in for a clinical endpoint. So the assumption is intuitive. Statins and drugs like them work because they lower LDL, right? Therefore, bigger reductions in LDL should correspond to improvements in cardiovascular events, right? But what did the study find? Simply put, LDL cholesterol drops in statin trials were horrific, terrible, abysmal, and all the synonyms predictors of clinical outcomes across these 20 randomized controlled trials. This can be quantified to add a layer of nuance by something called an R-squared value. This tells you how well one thing predicts another.— it ranges from 0, no predictive power, to 1, perfect prediction. So can you guess the predictive power, the ARD squared, from 0 to 1 of LDL cholesterol reductions on cardiovascular outcomes? Can you guess? Drop it in the comments. Ready for the answer? 0 to 0.1. And the results for a related marker, non-HDL cholesterol, were similar. This is terrible. It's like trying to predict who will win a marathon based on who tied their shoes the tightest, or choosing a CEO based on how expensive his hat is. The authors themselves conclude, now quoting, "These associations were too weak to fulfill the conditions required for a surrogate to be considered valid." So, simply put, this calls into serious question whether LDL cholesterol can be used as a surrogate for clinical outcomes in statin trials. Apparently not. But that brings us to Chapter 2: Replication—replication is important—and Very Little Association. I just want to point out, these new 2025 European Heart Journal data aren't an outlier. In a prior 2022 study published in JAMA Internal Medicine, researchers reviewed 21 human randomized controlled trials and tested, again, whether reductions in LDL cholesterol caused by statins were meaningfully associated with improvements in cardiovascular events and all-cause mortality. And here's what they found: And I'll quote directly, "These data indicate that there is very little, if any, association between the magnitude of LDL cholesterol reduction and the size of statin treatment effect." So pause and think to yourself, what is Nick really saying here? What am I really saying here? Am I saying statins are bad or entirely useless? No, I'm not. This video is not actually anti-statin per se. In fact, admittedly, in the trials reviewed, statins did, at a population level, reduce major adverse cardiovascular events compared to placebo. The key finding I'm trying to draw out is that the degree to which LDL is reduced didn't predict improvements. So why is this important, and why the disconnect? That brings us to Chapter 3: One Pill, Many Mechanisms. The answer to resolve the discrepancy is revealed in a critical nuance we need to appreciate. Statins have—brace for jargon—pleiotrophic effects. This means they do more than just lower cholesterol. For example, statins can promote vasodilation via the production of nitric oxide, improve blood flow to vital organs, including the heart, and reduce inflammation, which are core drivers to many chronic diseases. So, in people who are metabolically unhealthy, which, let's face it, includes the majority of the population, about 90% of the population, those effects—vasodilation, anti-inflammation—they're meaningful. In a world that is overweight, inflamed, and insulin resistant, statins can provide certain benefits that are distinct from LDL cholesterol reduction. And those might actually be the real reasons statins improve outcomes, at least in some patients. Yet cholesterol reduction gets basically all the credit. And this distinction matters a lot because at the level of the individual patient, we need to work out a cost-benefit analysis, a risk-benefit analysis. Yes, statins lower LDL and ApoB, and they can improve vascular function and reduce inflammation when those variables are high at baseline in metabolically unhealthy patients. But they can also lower GLP-1, impair insulin sensitivity, increase diabetes risk, and harm mitochondria. The point is it's not black and white, it's not simple, and in any given person, it's not clear how these variables will balance out unless you give it critical thought. It's complicated. Biology is complicated, and that's the point. So here's what we know: statins help in some people to reduce cardiovascular events, not because of LDL cholesterol reduction alone. LDL drop is actually a weak predictor of actual outcomes in these trials. Inflammation and metabolic dysfunction, they really matter. Metabolic health matters more than cholesterol. Now, here is where I could expand this video into a 3 or 4 hour lecture reviewing the contradictory data about the nuances of causality versus importance and the Mendelian randomizations, etc., etc. And frankly, if you want that, I have a lot more content. You just need to see the associated newsletter on the other content I've linked below. Go as deep as you want, I challenge you to. However, I know the majority of viewers are now asking, "So what? What is the takeaway? What do I need to know today?" So, here it is. On balance, statins can reduce the risk of major adverse cardiovascular events, like heart attacks and strokes, in people with pre-existing cardiovascular disease, either a prior heart attack or cardiovascular disease on imaging, and/or those who are metabolically unhealthy—think obesity, high triglycerides. Low HDL, diabetes, etc. But for those healthy individuals with isolated high LDL or ApoB, the data do not support a net benefit of statins, and if you want more on that, here's another video where we hammer home that point. Literally, I do it with a hammer. You'll see. Now, where statins do help some patients, it's likely not just because of LDL cholesterol reduction or ApoB reduction alone. Instead, the benefits probably stem from their pleiotrophic effects, like improving vasodilation and reducing reducing systemic inflammation when those markers are high at baseline, and potentially influencing other important physiological processes. And it's also important to remember, statins aren't the only medication for LDL cholesterol and lipid manipulation. There are many other medications: ezetimibe, bempedoic acid, PCSK9 inhibitors, berberine, and so on. I provide more details on those in the associated newsletter, again check it out below, and other content. Really, this is like a Russian nesting doll approach, consider this the outer doll, and you can go much deeper. But that brings us to Chapter 4: The Conclusion. Surrogates can mislead. Biology isn't simple. So let's wrap up. What's the takeaway? LDL cholesterol is a convenient marker to measure. It's easy to test, easy to change, easy to talk about. But just like that mug of hot chocolate that started us off, that can cause a fever, a drop in LDL cholesterol doesn't necessarily mean your heart is safer. The real actors— inflammation, metabolic health— they're playing behind the scenes. They don't get enough credit. In medicine, we like simple numbers, but human biology is not simple. Surrogates can mislead, especially when they're easy to manipulate. The more precise our understanding of the physiology and a drug's impact on biology, which you've learned a little about here, really it can help improve precision care. So, no, this isn't an anti-statin video. It's an anti-shortcut video. It's a call to action for nuance and hard discussion, because when it comes to your health, and especially your heart, easy answers can be dangerously satisfying. And sometimes, the metrics that we trust the most are just an excuse to skip school. But you didn't today. I celebrate you for that. Stay curious. Tell me what further questions you have.