Aside from the previously mentioned bias of people with health problems retiring early and skewing the results, what about this:
People who retire at any age without a clue as to what to do with their lives die faster, because they are relatively stressed and miserable. (Often true also with people whose spouse dies -- they die in disproportionate numbers in the first year or two because of the stress and change to their lives.)
Early retirees and those planning to ER like those on this board spend a lot of time thinking about what to do in ER, why they would want to do it, and make a gradual transition into it. We sidestep the 'die because you have nothing else to do' problem. However, the average Shell Oil early retiree is ERing because s/he has a heckuva good pension plan and maybe even gets a company incentive to retire early to clear out the dead wood. He/she hasn't invested time in doing the necessary transitional and planning stages well (on average), and so represents an at-risk group.
Bottom line: If you're going to ER, you need to get here to early-retirement.org and start reading and posting. This life transition takes time to get you and a spouse on board with.
Final thought: if you had marginal finances (most DIY ERs) and were bored of ER, you'd just go back to work. If you have a rock solid Shell Company ER plan and get bored of ER, you may be less likely to go back to work, (you don't feel any economic benefit from it) and that just makes matters worse..
Maybe this is all a stretch, but I do think these studies need to try to figure out a link between who has planned for ER vs who just had it thrust upon them. Then see if there is a mortaliity difference. I believe there have been similar studies done for traditional retirees along these lines, showing that comrehensive retirement planning, undertaken over many years, pays big dividends, maybe even longer life expectancy.
