 The risks of heart disease and stroke in meat eaters versus vegetarians over 18 years of follow-up. Not surprisingly, vegetarian diets were associated with less heart disease, 10 fewer cases per 1,000 people per decade compared to meat eaters, but vegetarian diets were associated with three more cases of stroke. So eating vegetarian appears to lower the risk of cardiovascular disease by 7 overall, but why the extra stroke risk? Did it just be reverse causation? For example, when studies have shown higher mortality among those who quit smoking compared to continuing smokers, we suspect reverse causality. When we see a link between quitting smoking and dying, instead of quitting smoking leading to people dying, more likely dying led people to quit smoking. It's the same reason why non-drinkers can appear to have even more liver cirrhosis because it was their failing liver that led them to stop drinking. This is the so-called sick-quitter effect, and you can see it when people quit meat, too. See how new vegetarians can appear to have more heart disease than non-vegetarians? Why would some older person all of a sudden start eating vegetarian? Maybe it's because they've just been diagnosed with heart disease. So that may be why there appears to be higher rates at first the sick-quitter effect. It's in control for that. You can throw out the first five years of data to make sure the diet has a chance to start working, and indeed, when you do that, the true effect comes clear, significant drop in heart disease risk. So does that likely explain the apparent increased stroke risk? No, because they still found higher stroke risk even after the first five years. Huh. OK, then what's going on? Let's dive deeper into the data to look for clues. This is what you get when you break down the results by type of stroke and type of vegetarian, vegetarian versus vegan. There are two main types of strokes, ischemic strokes and hemorrhagic strokes. Now most are ischemic strokes, or clotting strokes where an artery in the brain gets clogged off, as opposed to hemorrhagic strokes or bleeding strokes where a blood vessel in the brain ruptures. In the United States, it's about 90-10, 9 out of 10 strokes, or clotting 1 out of 10, the bleeding type. And that's what the vegetarians appeared to have significantly more of. Now, the vegans didn't, in fact, statistically have significantly higher risk of any kind of stroke, but that's terrible news for vegans. Vegans have the same stroke risk as meat eaters? What's going on? What's so increasing their stroke risk that it's offsetting all their natural advantages? And the same could be asked of vegetarians. Even though this was the first study of vegetarian stroke incidents, there have been about a half-dozen studies on stroke mortality. And the various meta-analyses have consistently found significant lower heart disease risk, but the lower stroke mortality was not statistically significant. Now with this new study, vegetarians can take comfort in the fact that at least they don't have a higher risk of dying from stroke, but that's terrible news for vegetarians. What, statistically, vegetarians have the same stroke death rate as meat eaters? What's going on, right? What's so increasing their stroke rate that it's offsetting all their natural advantages? Okay, let's run through a couple possibilities. If you look at the vitamin D levels of vegetarians and vegans, they do tend to run consistently lower than meat eaters. And lower vitamin D status is associated with increased risk of stroke. But who has higher levels of the sunshine vitamin? Those who are outside running around and exercising. Maybe that's why their stroke risk is better. What we need are randomized studies. And when you look at people who've been effectively randomized at birth to have lifelong lower vitamin D levels just genetically, you do not see a clear indicator of increased stroke risk. So the link between vitamin D and stroke is probably not cause and effect. And so in terms of an answer to our question, vitamin D doesn't seem to fit, right? We'll explore some other possibilities next.