Tuesday, January 3, 2017
The J Curve explained
In order to make sense of the seemingly conflicting reports about wine and health there’s one essential thing to understand: the J-shaped curve. It’s a simple concept, universal, in plain sight, and often ignored. It goes like this: Take “nondrinking” as the baseline and plot increased or decreased relative risk of a health issue with increasing levels of daily consumption. Nondrinkers have a certain risk of, say heart attacks, moderate drinkers a lower risk, heavy drinkers a relatively higher risk. Not too complicated. The tricky parts are separating wine drinkers from drinkers in general, and daily moderate drinkers from occasional drinkers.
Wait - water? Obviously not drinking enough water is unhealthy, and questioning the benefits of hydration seems a fool’s errand. But it is possible to take it too far; in 2007 a woman participating in a water drinking contest called “Hold Your Wee for a Wii” was found dead of water intoxication. Superhydration throws electrolyte balances out of whack, with toxic and even fatal levels of water intake surprisingly easy to achieve. A J-shaped curve.
Even lifetime happiness reportedly follows the curve. Young people generally enjoy a sense of well-being and optimism, career and family stress creates a dip through the 20’s and 30’s, then later in life happiness rises above the baseline, at least for most.
Why is this simple model so often overlooked? One reason is that good data points are hard to come by, when it is drinking and eating habits that are being tabulated. People are unreliable self-reporters. Or researchers may have hidden agendas based on the need to publish, so that they focus on only the findings that support their hypothesis. Research on breast cancer and alcohol is particularly fraught with this problem; heavy drinking is unquestionably bad, but difficult in parsing out the subset of women who drink red wine (for example) with regularity and in moderation leads to extrapolation errors. If you simply draw a line from the heavy drinking/high risk corner of the graph down to the no drinking corner, you miss the bottom of the J. And you don’t want to miss the bottom of the J curve.