While the links between alcohol and any particular disorder are irrelevant in the grand scheme of things where the net effects on mortality and morbidity are what really matter, the link with cancer has been getting a lot of airtime - and a lot of attention from the dedicated team at the University of Otago.
And so this is then interesting. The relationship between alcohol and cancer seems to vary by exercise. While there's a straight linear increase in cancer risk with alcohol consumption among those with little physical activity, those engaging in more than 15 hours per week of MET [some measure of exercise] showed a J-curve relationship - albeit one where the bottom part of the J wasn't statistically different from zero.
The paper also finds the standard J-curve in all-cause mortality risk, but less tightly estimated. The base of the curve is lower among those getting more exercise though, so you should then take moderate alcohol use and exercise to be complements rather than substitutes. Or, at least, exercise seems to knock back the cancer risk of drinking to basically nil except at heavy drinking.
For what it's worth, I get my exercise incidentally: chasing children and Pokemon, or chasing Pokemon with the children. Wanting a beer afterwards is nature's way of telling us that we should have a beer after exercise. Evolution gave us these preferences for a reason....
HT: Evan Roberts