Eating red meat is correlated with increased mortality, to the tune of one hour of life per serving. That got me wondering if even a little bit of red meat is bad for you. Maybe there’s some lower threshold where we don’t have to worry about it.
It turns out that earlier this year a relevant study of Seventh-day Adventists came out. Most Adventists are vegetarians. Those that aren’t still eat very little meat. Multiple groups of scientists have conducted a series of prospective studies of the eating habits and health outcomes of Adventists. The new study analyzed the data for the effects of red and processed meat.
Like many studies, this one broke people into five groups based on their consumption of red or processed meat. Due to the high concentration of vegetarians in their study, the lowest group is entirely vegetarian. A zero-consumption reference group lets us see if even low levels of red meat consumption are worse than none. In short, they are. Here are my plots of their main results.
The lowest meat consumption group ate just 4 grams per day of unprocessed red meat and 0.7 g/day of processed meat. That’s tiny. It’s about one quarter-pounder and two slices of bacon per month. The risk of all-cause mortality in this lowest non-zero consumption group is slightly elevated for unprocessed red meat but actually slightly decreased for processed meat. Both values are within the error bar of no change. It’s tempting to take the result for unprocessed red meat at face value and ignore the error bar. A tiny bit of red meat raises mortality 6%! But then we’d have to do the same with the processed meat result. A tiny bit of bacon decreases mortality 5%!
Let’s call the results for the lowest groups inconclusive. In the top group, they eat about 42 g/day of unprocessed red meat and 9 g/day of processed meat. That’s about a slice of bacon every day and a quarter pounder every couple of days. The 90th percentile for Adventists is closer to the median for Americans as a whole. For this top group, both categories of meat are correlated with higher mortality.
The result for unprocessed meat is straightforward: with every step up in meat consumption, there’s a step up in mortality. For processed meat the results aren’t as clear. The very highest group has much higher mortality, but even the second highest group, at 3.3 g/day, doesn’t show much evidence of harm.
As a rough take away, maybe we can say two slices of bacon per week (the second highest group for processed meat) and one or two hamburgers per month (the first and second lowest groups for unprocessed red meat) aren’t so bad. But it could be that the study just isn’t large enough to detect the harms of such small doses.
Hanging over all this talk is that the study was only observational, not experimental. I’m not familiar with Seventh-day Adventists, but I am familiar with Mormons. If you studied Mormons who do and don’t drink alcohol and assumed that was the only difference between them, you’d be very wrong. The groups are systematically different. The same could be true for vegetarian and non-vegetarian Adventists. In fact, the vegetarian Adventists are more likely to be white, to be married, to have graduate degrees, and to exercise. They’re also less likely to drink, to smoke, to have diabetes, and to eat dairy or eggs. The researchers used models that took all these features into account and re-ran the analysis excluding anyone who ever smoked (with similar results). Still, these models are no match for a large, long-term experiment. Since that’s very difficult and expensive, we may never get one.
How to 115 