Here's my theory:
surface level analysis of data says retirement is bad for health, but it doesn't address situations like the following.
Take Boomer Bob. Boomer Bob began working full-time at 20 years old as a manual laborer. Objectively, he had very high purchasing power for his skillset compared to people his age today. However, work is still work, and he comes home exhausted and demoralized everyday. As a result, over the decades, he developed increasing alcoholic tendencies to cope with the psychological distress from work in addition to working even heavier hours to provide for a family relying on his sole income.
Due to Bob becoming an absent father and husband, he experiences a divorce and a distanced relationship with his kids.
Eventually, his pension finally kicked in, he gets the rolex watch, and he never has to work another day in his life—but he kept one thing from work that he couldn't let go—his alcoholism.
As Bob goes into retirement, he spends more and more time with alcohol as he never developed any other hobbies in his life while also essentially losing his family due to work draining all of his attention, energy, and the best hours of the day just to get by.
Bob eventually dies—alone in his home—16 months into retirement from alcohol poisoning.
What does 'science' conclude? 'Retirement is bad', but clearly the situation is much more nuanced than this.
Sure, work technically helped create distance between alcohol and himself, but it's also the cause of him developing alcoholism in the first place.
So, did retirement kill him or was it work that killed him?