It's actually more complicated . . .

I expect most readers have seen news stories about the President's Cancer Panel report. The headline is that the panel claims that environmental exposures -- by implication at least, meaning mostly to human-produced chemicals -- make a far more important contribution to the burden of cancer than has been recognized. This has long been what I must, sadly, call an article of faith among environmentalists, but the report actually does not come to that conclusion, nor does it present any direct evidence for it. This spin must come from a press release, is the only conclusion I can make. In fact, in the executive summary, they say this:

At this time, we do not know how much environmental exposures influence cancer risk and related immune and endocrine dysfunction. Environmental contamination varies greatly by type and magnitude across the nation, and the lifetime effects of exposure to combinations of chemicals and other agents are largely unstudied. Similarly, the cancer impact of exposures during key “windows of vulnerability” such as the prenatal period, early life, and puberty are not well understood.

In other words, we don't know, but maybe it's worse than we think, so we ought to try to find out. That's a less exciting headline, but it has the virtue of being accurate.

Here's some context. Environmental epidemiology is a notoriously difficult endeavor. First, it's quite difficult to measure how much of any given agent people are exposed to. You can find people who live in a place where something or other is known to be in the air, for example, but they will all vary widely in how much time they spend at home, how much time they spend outdoors, how much they open the windows, etc. And living in that place also means that they likely have all sorts of other things in common that might be relevant to their cancer risk. We can try to measure and account for as many of those factors as we can think of, but we can't be sure we haven't missed the one that matters.

What is worse, it might be that exposure to X only matters if also Y. For example, radon exposure was found to increase the lung cancer risk of uranium miners, but it turned out to mostly be a synergistic effect with tobacco smoking. The effect was difficult to detect in non-smokers. Since smoking and environmental tobacco smoke were -- and still are -- by far the most powerful environmental cause of cancer, overwhelming everything else, people in public health (public healthists?) found it hard to get as excited about other kinds of exposures. Also, a lot of much hyped suspected links, for example between environmental endocrine disrupters such as many pesticides and breast cancer, haven't panned out.

Another inconvenient truth is that we do know about other environmental cancer risks, but they don't indict giant soulless corporations but rather your grandmother and your own sorry self. Smoked and grilled meat products and starchy fried foods contain carcinogens. Also, excessive alcohol consumption promotes cancer. And of course, so does getting a sunburn.

Now, we do know that some products of incomplete combustion -- ergo compounds found in motor vehicle exhaust and smokestack emissions -- called polycyclic aromatic hydrocarbons, are carcinogenic. More complicated versions called dioxins are found as contaminants in pesticides as well as smokestack emissions (including, by the way, your fireplace chimney) and paper mill effluent. A lot of reactive hydrocarbon compounds probably can cause genetic damage, mess with hormones, and just cause inflammation and irritation which makes cells divide more than usual, which also may increase cancer risk. The problem is that all of these exposures mix and match and interact, and that it probably matters how old you are and how long it lasts and whether you are overweight or underweight or male or female and what your particular genetic endowment is and who the hell knows what else.

Most chemicals approved for various uses, including as ingredients in food or food packaging, believe it or not, have not been adequately tested for safety. But that is also very hard to do. You can't do experiments by feeding them to people. You can feed stuff to mice, despite PETA's objections, but a) mice aren't people and b) you can't feasibly mimic human exposures. It's just too expensive to give millions of mice a tiny bit of something for a long time. Instead you have to give a small number of mice a whole lot of it. That just ain't the same thing; and mice don't live nearly as long as people anyway so you can't do twenty or thirty year follow up. And sorting out what people are exposed to from their natural history, and trying to relate that to their lifelong risk of cancer, is just really, really hard and can very easily lead you to wrong conclusions.

So yeah, we need to keep studying these questions, and we probably need to do it bigger and better. But so far there are only a few specific things we can tell you. Don't smoke. Don't do shots every night. Don't live near the highway. Maintain an appropriate body weight. Don't work as a hot tar roofer or a paver -- although somebody has to do it.