Bear in mind that the study looked at the combined impact of multiple factors: working out, eating healthy food, and not being overweight were specifically stud

Bear in mind that the study looked at the combined impact of multiple factors: working out, eating healthy food, and not being overweight were specifically studied. So by definition, this "10 more years" population were the people for whom drinking didn't snowball. That said, the individual-factor breakdown really did find that at low doses, rising alcohol consumption correlated with improved health. Undoubtedly some healthy-user bias between no and light drinking, but it's a pretty smooth curve through moderate consumption.