 When ranked in order of importance, among the interventions available to prevent stroke, the three most important are probably diet, smoking cessation, and blood pressure control. Most of us are doing pretty good on smoking these days. Less than half of us are exercising enough, but according to the American Heart Association, only one in a thousand Americans are eating a healthy diet, even fewer than 1 in 10, or even eating a moderately healthy diet. Why does it matter? Because diet is an important part of stroke prevention. Reducing sodium intake, avoiding egg yolks, limiting the intake of meat, and increasing the intake of whole grains, fruits, vegetables, and lentils. Like the sugar industry, the meat and egg industry spend hundreds of millions of dollars on propaganda. Unfortunately, with great success, I was excited to check out box number one, and was then honored when I did. The strongest evidence for stroke protection is for increasing fruit and vegetable intake with more uncertainty regarding the role of whole grains, animal products, and dietary patterns, such as vegetarian diets. I mean, one would expect they'd do great. Med analyses have found that vegetarian diets lower cholesterol, and blood pressure, and enhanced weight loss, and blood sugar control, and vegan diets may work even better. So all the key biomarkers are going in the right direction, but you may be surprised to learn that there hadn't ever been any studies on the incidence of stroke in vegetarians and vegans until now. And if you think that's surprising, wait until you hear the results. The risks of heart disease in stroke in meat-eaters, fish-eaters, and vegetarians over 18 years of follow-up. Yes, less heart disease among vegetarians, by which they mean vegetarians and vegans combined. No surprise, been there, done that, but more stroke. An understandable knee-jerk reaction might be, wait a second, who did this study? But this is Epic Oxford, world-class researchers whose conflicts of interest may be more likely to read, I was a member of the Vegan Society. What about over-adjustment? I mean if you crunched the numbers over a 10-year period, they found 15 strokes for every 1,000 meat-eaters, compared to only 9 strokes for every 1,000 vegetarians and vegans. Wait, so how can they say there was more strokes in the vegetarians? This was after adjusting for a variety of factors. For example, the vegetarians were less likely to smoke, so you want to cancel that out by adjusting for smoking so that you can effectively compare the stroke risk of non-smoking vegetarians to non-smoking meat-eaters. If you want to know how a vegetarian diet itself affects stroke rates, you want to cancel out these non-diet-related factors. Sometimes, though, you can over-adjust. The sugar industry does it all the time. This is how it works. Imagine you just got a grant from the soda industry to study the effects of soda on the childhood obesity epidemic. What could you possibly do after putting all the studies together to arrive at the conclusion that there was near zero effect of sugary beverage consumption on body weight? Well, since you know that drinking liquid candy can lead to excess calories that can lead to obesity, if you control for calories, if you control for a factor that's in the causal chain, effectively only comparing soda drinkers who take in the same number of calories as non-soda drinkers, then you could undermine the soda to obesity effect. And that's exactly what they did. That introduces over-adjustment bias. Instead of just controlling for some unrelated factor, you control for an intermediate variable on the cause and effect pathway between exposure and outcome. Over-adjustment is how meat and dairy industry-funded researchers have been accused of obscuring the true association between saturated fat and cardiovascular disease. We know that saturated fat increases cholesterol, which increases heart disease risk. Therefore, if you control for cholesterol, effectively only comparing saturated fat eaters with the same cholesterol levels as non-saturated fat eaters, you see how you can undermine the saturated fat to heart disease effect. Now let's get back to this. Since vegetarian eating lowers blood pressure and a lowered blood pressure leads to less stroke, controlling for blood pressure would be in over-adjustment, effectively only comparing vegetarians to meat eaters with the same low blood pressure. That's not fair, right? Since that's one of the benefits of vegetarian eating, not some unrelated factor like smoking, and so it would undermine the afforded protection. So did they do that? No. They only adjusted for unrelated factors, like education and socioeconomic class and smoking and exercise and alcohol, right? And that's what you want, right? You want to tease out the effects of a vegetarian diet on stroke risk, right? You want to try to equalize everything else to tease out the effects of just the dietary choice. And since, for example, meat eaters in the study were on average 10 years older than the vegetarians, you can totally see how, when you adjust for that, vegetarians could come out worse. Since stroke risk can increase exponentially with age, you can see how having nine strokes among 1,000 vegetarians in their 40s could be worse than 15 strokes among 1,000 meat eaters in their 50s. The fact that vegetarians had greater stroke risk, despite their lower blood pressure, suggests there's something about meat-free diets that so increases stroke risk, it's enough to cancel out the blood pressure benefits. But even if that's true, you still would want to eat that way. I mean, stroke is our fifth leading cause of death. Whereas heart disease is number one. So, yes, in this study there were this many more cases of stroke in vegetarians, but there were this many fewer cases of heart disease. But if there is something increasing stroke risk in vegetarians, it may be nice to know what it is, and helps in figuring out how to get the best of both worlds, this is the question we'll turn to next.