It’s thousands of tiny little things, pushing and pulling lifespans up and down.
As the screenshot notes, it’s both diet and access to healthcare.
It’s also other lifestyle factors, like amount of walking or driving, amount of alcohol consumed, tobacco use, etc.
It’s social and economic factors, like income, education levels, employment status, type of job, disability status, marital status, number of close friends.
It’s mental health issues, and related statistics like suicide rates, substance abuse rates, etc.
There are environmental factors, like environmental exposure to certain hazards or pollution, sunlight exposure, altitude, certain illnesses isolated to certain climates, maybe things like localized microbiomes (although those are also correlated with foods eaten and things like that).
There are also genetic factors for individual families or potentially ethnic groups.
And perhaps the one that can’t be ignored entirely is just plain old recordkeeping. Some places have high rates of people living past 100, but don’t seem to have much in the way of a lifestyle or environmental explanation, and may more accurately trace back to unreliable birth records 80+ years ago such that people might be mistakenly reported as living longer than they actually did.
That’s just the unusual prevalence of 100+ year olds, in the so called “blue zones.” Overall country life expectancy statistics aren’t thrown off by that type of fraud as much, because the vast majority of people don’t live anywhere close to 100, and these specific blue zones are a very small overall portion of the larger country.
For the most part, we can observe a correlation between wealth/income and life expectancy, where the blue zones are outliers on that general trend (both long lived and very poor). So there’s no reason to believe that these small communities are poisoning the overall stats in any significant quantity.