Ignorance Is Bliss: The Dark Truth About Chocolate & Diabetes

Does chocolate really prevent diabetes? Kind of like warding off a metabolic dementor? Now, this question has been top of mind since a major publication came out in the BMJ—British Medical Journal—claiming a 21% reduction in type 2 diabetes risk in those who ate at least 5 servings of dark chocolate per week. Chocoholics rejoice! Maybe. But just how legitimate are these findings? We're going to go over the study, and I'll explain to you why the findings unfortunately aren't all they're chalked up to be, or chocolated up to be. But I won't leave you high and dry. I'll also open your eyes to just how complex chocolate science truly is, give my opinion on the overall state of chocolate science, and I'll suggest how to get the most health bang for your chocolate buck with science and practical tips. But first, let's break down what they did in this study, what it says, and what this study doesn't say. This was a large-scale observational study where they looked at associations between chocolate intake and the development of type 2 diabetes over 3 separate cohorts: the Nurses Health Study, the Nurses Health Study II, and the Health Professionals Follow-up Study. There were a total of 111,654 participants included, and they looked at the associations between types of chocolate—dark and milk—versus type 2 diabetes. And the main findings were as follows. There were two: participants who consumed at least 5 servings per week of dark chocolate, but not milk chocolate, showed a 21% lower risk of type 2 diabetes. And 2, intake of milk chocolate, but not dark chocolate, was associated positively with weight gain. Overall, this sounds nice, provided you like dark chocolate and not milk chocolate. And if you want to believe dark chocolate will protect you from diabetes, I recommend you stop this video here and enjoy ignorance as bliss. Oh, you're still watching. Very well. Let's discuss why I'm skeptical that dark chocolate will meaningfully reduce your risk of type 2 diabetes. First and foremost, there was a clear signal of what's called healthy user bias in this study. Healthy user bias means that one group, in this case dark chocolate eaters, tend to have other healthy habits, that can explain the observed effect. Here, lower incidence of type 2 diabetes. And you can see that looking at the data here. I'm showing you data from Table 1, where those with higher dark chocolate intake tended to have more physical activity, more multivitamin use, generally overall diet quality score, so higher diet quality, and lower BMI. And for most of these, it was actually a, an apparent dose-dependent effect, so more dark chocolate consumption coupled with like more physical activity. And conversely, for high milk chocolate users, there was a suggestion of less healthy living, including higher smoking rates and lower diet quality. So is it as simple as those who chose to eat dark chocolate are also just healthier people in general, and those who chose to eat milk chocolate are just less healthy people in general. Well, maybe. In part, it could be that simple. Although a challenge could be, but they, the researchers, adjusted for these factors in their mathematical modeling. That's a counterpoint. However, it's important to realize these statistical adjustments are just best guesses and necessarily incomplete. In fact, the authors themselves write in the paper, we cannot entirely rule out the role of confounding in our observed associations and that residual or unmeasured confounding, or both, may still exist. Furthermore, in a subgroup analysis in this study, they also found no association between dark chocolate consumption and type 2 diabetes risk among individuals with lower diet quality, which is consistent with the idea that confounders—the healthy user bias we mentioned—was carrying the lion's share of the reported effect. Another big red flag was that there was massive study heterogeneity amongst the cohorts, the three we mentioned. In other words, the associations between dark chocolate intake and type 2 diabetes were not at all consistent among the three cohorts. In fact, they were primarily driven by one of the three cohorts, with a supposed 51% reduced risk in type 2 diabetes in the heavy chocolate users in the Health Professionals Follow-up Study, the HPFS, which I think is actually a pretty absurd and unbelievable value, and also one with a giant confidence interval at that, 8 to 74%. And there was no association noted in the Nurses Health Study. The paper actually reads, "In the Nurses Health Study, neither total nor subtypes of chocolate were statistically significantly associated with the risk of type 2 diabetes." Now, if this was actually a biological phenomenon whereby the chocolate was protecting against type 2 diabetes, then I'd expect the data to be more consistent. And instead, what we see is the data appears super noisy, as noisy as an attention-hungry elephant with a bullhorn. And as a last aside before we summarize these points, because I realize I've been throwing a lot of jargon at you, what is dark chocolate anyway? Are we talking about 95% or 100% dark chocolate bars or pure cacao, or 50% dark chocolate truffles? Since this is a self-reported questionnaire that was the basis for this study, these items all get lumped together, although they're obviously not the same. Actually, in the discussion, they do define dark chocolate, which is a somewhat arbitrary term, with a lower bound of 50% cocoa. Now, I don't know about you, but to me that's candy, not true dark chocolate. And that's just my two cents, then again, I'm a bit of a chocolate snob. So, what do we make of all this at a high level? Am I saying that dark chocolate isn't healthy after all? Myth busted? No, not exactly. If we look at the larger body of literature, I think there's something to dark chocolate and the flavanols in dark chocolate— I'm going to put a pin in that word for a moment— being health-promoting. In fact, on balance, I think the literature says more about dark chocolate for being good for cardiovascular health than actually diabetes prevention, and that's a whole other kettle of fish. However, these new data in the BMJ that are promoting all these headlines, they don't inspire confidence in me that dark chocolate, as defined in this study, is actually protective against diabetes. I think the bulk of the apparent protective effect here is more a function of other factors that happen to cluster with dark chocolate consumption rather than the dark chocolate consumption itself. The big point is I really don't put much stock in these data. I think the headlines are just headlines. So before this paper was published, there were conflicting data, on whether dark chocolate might protect against diabetes. And now, well, I don't think the story has changed one iota. But we are not done, because now we need to ask the super important question: why? Why are there data conflicting? Why can't we get a straight answer on what seems like a simple question? Well, setting aside the fact that it's next to impossible to do a proper randomized controlled trial isolating for one food in terms of incidence of chronic disease, here diabetes, there's huge variability, as we mentioned, in what we're even calling dark chocolate or chocolate. I think we can agree that dark chocolate, as in a dark chocolate truffle with sugar and additives that's highly processed, is very different than 100% dark chocolate bar or pure cacao. Those are very different things, but in the literature they get clustered. And beyond that, and this will blow your mind, consider that there are tons, at least 4 variables, and probably many, many more, that result in variation in pure cocoa that goes to the manufacturer. And this is even before the manufacturer starts to process it and mix in the cocoa with other ingredients to make the chocolate. So chocolate comes from the Theobroma cacao plant. It's a bean, and there are many different varieties of Theobroma cacao plant, at least 10, and they are grown in different regions with different soil conditions, and they are fermented by different microorganisms. Yes, chocolate is a fermented food. And then they're fermented for different durations, and all this happens well before the chocolate goes to the manufacturer. And then there's roasting, something called conching, and all these steps which can all change not only the chocolate flavor profile but also the nutrient profiles of the chocolate. And again, that's all separate from the added sugar and additives and percent on the chocolate. Kind of crazy, right? It's more complicated than we realize. So it's easy to see how chocolate science can get super confusing. But to boil this down and make things as simple as possible for you, before we get on to our practical tips, on balance, I just want to say my opinion is I think dark chocolate is a perfectly healthy food. It can be high in flavonols, antioxidants, fiber if that's a target for you, healthy fats including healthy saturated fats, and dark chocolate without much added sugar doesn't tend to spike blood sugar given its low glycemic impact. It has a glycemic index of around 20 on a roughly 0 to 100 scale, with 0 being carb-free foods like eggs and meat and 100 being pure glucose sugar. So chocolate's a pretty good option for a dessert that won't spike your blood sugar if you're actually having real, real dark chocolate. And honestly, I think the best thing you can do to reduce your risk of type 2 diabetes is eat in a way that reduces sugar intake and keeps your blood sugar stable, and chocolate can be a part of that method. And it pairs incredibly nicely both metabolically and in terms of flavor profiles with healthy fatty foods like a good macadamia butter or a tahini. So, on balance, I say chocolate is great. You should go for it. But now, how to get the most bang out of your chocolate buck? Well, I think the heuristic "darker is better" still applies. Shoot for something that is at least 80% dark chocolate. The closer you can get to 100, the better. And if you can find it, some chocolate makers will even put the flavanol content on the packaging. Flavanols are a family of antioxidant compounds thought to give dark chocolate its health halo, where it arises in the literature. And the main flavanol in chocolate is called epicatechin, which has also been shown to have health benefits on blood flow, blood pressure, and possibly even mitochondrial function. It's also found in foods like green tea, which, as you probably know, also have a health halo. So if you can find chocolate with the flavanol levels on the package, that's a really good sign. But if you can't, don't worry. Generally, higher levels of percentage, so darker chocolate and less processed chocolates, will boast the higher flavanol levels. That means you want to look for something with less roasting, less alkalinization, less dutching, These are all processes that can reduce flavonols in chocolates. And don't worry, no need to take notes if you just check the caption on the video. I put all the notes there for you. Now, my personal favorite— if I'm going to plug a brand, I have no affiliation with this brand— but is stone ground 95% Teza dark chocolate. The stone ground process preserves a lot of the flavonols, and it's super delicious. It's a little bit grainy, but I actually really like that. And it's also, as an aside, been tested and is low in heavy metals like lead and cadmium, which has been a concern among common chocolate brands. So when I eat chocolate, this is probably the brand I use the most. Again, no affiliation, I just like the brand. And another option is to use pure cacao or cacao powder, which you can mix in with things like macadamia butter or tahini, as I mentioned. And if very dark chocolate like the 90, 95, 100% is too bitter for you, I recommend just making your own dark chocolate at home, which isn't too hard. This way you can add your own sweet source to make sure you're not adding sugar or potentially harmful sugar alcohols like erythritol, which I covered in a recent video, or artificial sweeteners. And if you want a great recipe from my friend Martina, Again, I'm linking it below, and you can use just pure cacao powder and allulose and other simple ingredients, maybe add a dash of vanilla or cinnamon. You can customize it. It's actually really nice, especially if you like cooking. So bottom line, there are lots of healthy ways you can use chocolate, and this video isn't meant to undermine that it can be a health food. But there's a deeper lesson here about what we want to believe versus what the data actually say. Based on a pre-existing narrative of the healthfulness of chocolates, and possibly personal preferences, for those chocolate lovers among us, including myself, we would want to believe these headlines. We want to believe that chocolate reduces type 2 diabetes risk by 21%. But the fact of the matter is, the science was not strong. And it's critical we break this down and dig into the methods and limitation, because science is about looking for the truth, because it's the truth, not when it's just the truth we want to hear. So I'm sorry, life is not like a box of chocolates. It's like a box of confusing as heck headlines. But together we can break them down, elevate our minds, and stay curious. I hope you found value in this one. I'm really interested to see how you respond anyway.