Healthy at Any Age
Carmichael, Mary, Newsweek
Byline: Mary Carmichael
In the era of Google, medical advice is more confusing than ever. Here's a guide to what you really need to know, and when.
It's a cliche that we're living in the too much Information Age, and if you Google anything related to health, it's plain to see why. Plug in "breast cancer screening" and you'll get 7.39 million results, including guidelines from the National Cancer Institute, the U.S. Preventive Services Task Force, the American Cancer Society, the Centers for Disease Control, and the Mayo Clinic, all on the first page of links. There's also PubMed, where anyone can access 19 million scientific papers that otherwise might be confined to journals in the musty stacks of academic libraries. A PubMed search for "breast cancer screening" returns 112,737 studies, 23,286 of which are freely available in full. Even a much narrower search--say, for a single genetic mutation that might alter your risk and cause you to need more regular screening--yields reams of information ("185delAG BRCA1": 23,800 Google results, 0.37 seconds). By this point, you can start to feel like you know everything about breast-cancer screening that's worth knowing. But you don't. Neither do doctors.
When it comes to health, we're not living in the Age of Too Much Information so much as the Age of Not Quite Enough. Medical science has generated vast amounts of data, and laypeople have more access to them than ever before, but look closely at that information, and it starts to seem disturbingly -incomplete.
Take, for example, those 19 million studies on PubMed. Some are gold--standard research: large, randomized, controlled clinical trials or meta--analyses that carefully parse millions of data points. But others are poorly designed, without randomized participants, control groups, or statistical methods adequate to make sense of the raw numbers they generate. Even some of those identified as randomized controlled trials may not be, as a recent review of thousands of studies published in Chinese journals found. If the authors of these papers don't know how to properly generate their conclusions, it's hard to see how anyone else can feel comfortable relying on them.
Medical guidelines from task forces and disease societies should be more trustworthy--think of all the brainpower and hours of committee meetings that go into them--but they don't always agree with each other, and they occasionally change. Witness the storm over regular mammograms for all women between 40 and 50. It's been seven months since the U.S. Preventive Services Task Force decreed that they're not necessary, but the debate is far from resolved; the American Cancer Society stands by its old policy of advocating annual screenings starting at 40. For that matter, what about prostate-cancer screening or hormone-replacement therapy, both of which have spent the last few years falling in and out of favor in the medical community? When the authorities can't agree on best practices, nonauthorities are left with even less certainty.
As for the genetic information currently available, no one knows how most of it fits into the larger picture of people's health. Doctors recently analyzed the entire genome of one man, Stephen Quake, and tried to turn the string of nucleotides into something clinically meaningful. They found he had an "increased genetic risk for myocardial infarction, type 2 diabetes, and some cancers," and Quake decided to take a prophylactic statin as a result. Genomics enthusiasts hailed the paper, published in The Lancet, as an important step toward personalized medicine. But the vast majority of us don't have our complete genome sequences available yet (Quake is a bioengineer who sequenced his own DNA). It's also worth noting that the research group turned up "many variants of uncertain importance" in Quake's genome. When scientists connect a gene to a disease, they often don't know what that gene does--the protein it makes or the process it regulates--or what factors turn it on or off, up or down. …