I'm guessing that if you do something to improve your health enough that it adds years to your life, the rest of your health should be improved as well.

Meaning, maybe standing also improves your back and you ache less in later years.

It's hard for me to imagine doing something that improves your life expectancy at the expense of some part of your body.

This is massive speculation on my part.

This isn't a valid general principle. People who quit smoking late in life add extremely low-quality disabled life years.

As a general principal, I would disagree with the position that it's not valid. Certainly some people who quit smoking late in life prolong a painful cancerous demise, however I would imaging the worst of the low-quality disabled years are often a result of the limited, desperate nature of the available potential life-saving treatments: radiation/chemotherapy, etc. A significant percentage of smokers quit late in life and avoid those extremely low-quality disabled life years. Also, it seems perspective typically changes as 'quality of life' decreases, generally resulting in the desire for continued respiration;-)

Short version: you're thinking lung cancer. It's not, it's COPD, and it's near ubiquitous for people who have smoked a while.

Assume the 2 years is evenly distributed across the rest of your life. Meaning each day at a given level of "health" is extended by the fraction of 2 years

Why would we assume that other than that it might make us feel better?

I don't have a great answer for other motives and it seems like a diffucult prediction to draw strong conclusions on. However, it's certainly not binary, so I'd argue that the possibly of making us 'feel better' is reason enough to assume.

Anybody can drive, not everybody can go hike up the mountain and experience nature's full beauty. If you think sitting on your ass is quality of life, you are missing out.

Would 2 years of pleasure be preferable to 4 years of torture? Where would the balance be for you?

Edit: Was updated to be more in-line with Parent's comment, and less direct.

What happens after those 2 years or 20 [1] years? And then I die?

I'd take the pleasure.

[1] The post originally said "20" years then updated to 4 years, but my answer is the same.

To some, it's the same as asking if you'd rather eat tasty meals and live until 62 - or never enjoy what you eat, and constantly torture yourself with exercise, and live until 80.

May you maximize both your years and your prosperity.

I'm 65.

Being dead three years ago does not appeal to me!

I'm willing to bet that's an age-independent invariant :).

Sounds like questions only you can answer.

How can I answer that? I'm not a medical researcher or statistician. And since I'm not a cat, I can't determine it empirically.

I feel like there's a type error in even asking that question, because you're comparing a distribution over people with a concrete person. Like a homeless person with no income hearing that the average household income is $52,000 and wondering which president is on his dollar bills, or whether the number of $1 bills within his portion is prime.