One of my pet peeves is how science is reported in the press. Sensationalism sells. Eyeballs are all that matter, so we get disturbing and inflammatory headlines that promise a lot more certainty than is warranted by the underlying science. Usually this is not the fault of the scientists but of the reporters and editors who bury the careful qualifications and uncertainties because they don’t fit the story they are trying to tell. But sometimes the scientists are less than cautious about how their work is perceived.
So like the proverbial boy who cried wolf too many times, when we really need to pay attention to science people are skeptical because the conclusions of science had been oversold in the past.
That seems to be the case in the latest news about the dangers of meat consumption that attracted a lot of attention last week. The headlines were disturbing:
- “All red meat is bad for you, new study says”
- “Red meat is blamed for one in 10 early deaths”
- “Scientists warn ‘red meat can be lethal’”
Wow! It makes you want to toss the burger in the trash mid-meal.
There are a lot of reasons not to eat red meat. It is not particularly healthy, and the production of beef cattle is an inefficient use of land, water, and energy supplies. But the evidence for its lethality, especially in modest quantities, is exceedingly thin.
The observational study on which these headlines depend followed over 120,000 women and men from the Nurses’ Health Study and the Health Professional’s Follow-up Study for 28 and 22 years respectively. The research summary report claims that a single daily serving of unprocessed red meat was associated with a 13% increased risk of death, while a single serving of processed red meat was associated with a 20% increased risk. Furthermore, according to the report, of the people in the sample who died during the study, 1 in 10 deaths could have been prevented if they had consumed less than 1/2 serving of red meat per day.
Sounds scary. But the study simply was not fine-grained enough to draw these conclusions. Nutrition advocate Denise Mungerdoes a great job of breaking down the limitations of this study. Her main points are:
- Its a observational study, not one that actually attempts to control confounding variables, which is especially important since people who eat meat may have other characteristics that increase their health risks.
- It depends on self-reports of what people consume and is thus inherently unreliable, as numerous studies have shown.
- But most importantly, the scary-sounding statistics are not so scary when you put them in context. As Munger writes:
If your risk of dying from a particular disease is 5% to start with, a “20% increased risk” only bumps you up to 6% in the grand scheme of things. That’s a lot less scary. Especially when delectable foods are involved.
Munger’s take-down of this study is worth reading in its entirety. The take- away-point is not that scientists are incompetent or dishonest. It is that science is very hard to do when studying complex systems such as the human body. The right answer seldom emerges from a handful of studies that establish only correlations, not causation, and the initial assessment of an hypothesis is often wrong. The fact that scientists often like the publicity that comes with sensationalism just means they’re human.
No doubt we should limit our consumption of red meat. But the impression that any meat consumption is inherently lethal is misleading.
Exploring the Philosophy of Food and Wine 