A recent study suggesting a link between coffee drinking and longer lives has prompted a flurry of coverage—some snarky, some cautious, but mostly celebratory. (We see you there, reaching for another cup of coffee.)
The study published at the prestigious New England Journal of Medicine is about as good as observational epidemiology studies go, but it’s limited by virtue of being observational. Last month on our Crux blog, Gary Taubes wrote a hard-hitting piece about the problems with observational studies. A major limitation of surveying people about their lifestyle habits is that correlation does not imply causation. It can’t prove coffee drinking actually led to living longer. There are always confounding variables.
In this coffee study, for example, they initially found that coffee drinkers died younger, but coffee drinkers are also more likely to be smokers. When they controlled for smoking as a confounding variable though, the result flipped: coffee drinkers lived longer. The researchers recognized there are other confounding variables too, and this is the entire list the researchers controlled for, taken directly from the paper:
The multivariate model was adjusted for the following factors at baseline: age; body-mass index (BMI; the weight in kilograms divided by the square of the height in meters); race or ethnic group; level of education; alcohol consumption; the number of cigarettes smoked per day, use or nonuse of pipes or cigars, and time of smoking cessation (<1 year, 1 to <5 years, 5 to <10 years, or ≥10 years before baseline); health status; diabetes (yes vs. no); marital status; physical activity; total energy intake; consumption of fruits, vegetables, red meat, white meat, and saturated fat; use or nonuse of vitamin supplements; and, in women, use or nonuse of postmenopausal hormone therapy.
But are there more confounding factors lurking unaccounted in the numbers? The Boston Globe interviewed Dr. Jeffrey Drazen, the editor-in-chief of NEJM, who admitted the studies like this are problematic and the decision to publish was controversial within the journal. He points out that the study did not control for health insurance, high blood pressure, or high cholesterol levels.
At first the list of confounding factors from the paper looked pretty exhaustive (and long) to us, but there’s always more factors to consider. So let’s play a game of find the confounding factor. What else do you think should the researchers have controlled for?
