Nobody has identified the happiest people over 70, traced their lives backwards and catalogued what they stopped doing. That study does not exist, which is why every version of this list contains the same things: resentment, comparison, worrying what other people think, holding grudges. They are not habits so much as moods, and moods are close to the hardest thing there is to quit on purpose.
But there is one habit where quitting has been tracked for half a century, where the effect on mood has been measured separately from the effect on the body, and where both halves of the answer are the opposite of what most people assume.
The half about mood
Gemma Taylor, Paul Aveyard and colleagues pooled 26 longitudinal studies that had measured people’s mental health before they stopped smoking and then again afterwards, comparing them against people who carried on. The follow-ups ran from seven weeks to nine years.
The Artful Age
A weekly letter on aging well, family across generations, and the creative life after the kids leave home.
Anxiety fell, by 0.37 of a standard deviation. Depression fell by 0.25. Mixed anxiety and depression fell by 0.31. Stress fell by 0.27. In the other direction, psychological quality of life rose by 0.22 and positive affect rose by 0.40.
These are not enormous numbers in isolation, so the authors supply the comparison themselves, and it is the line that makes the paper memorable: the effect sizes are, in their words, equal or larger than those of antidepressant treatment for mood and anxiety disorders.
They also looked for the group everyone assumes is the exception. There was no evidence that the effect was any smaller among people with physical or psychiatric disorders than in the general population.
Why this is surprising to the people it describes
Ask smokers why they smoke and a large share of the answer is emotional. To take the edge off. To settle down. To get through something. Researchers have documented this repeatedly, in people with and without any psychiatric diagnosis, and the belief is not a rationalization invented after the fact. It corresponds to something real that happens every time a cigarette is lit.
The authors’ explanation for the gap between that experience and their results is the most interesting thing in the paper. Smokers who have not smoked for a while become irritable, anxious and low. Smoking reliably fixes it. So the sensation of relief is genuine and it is repeated many times a day, which is exactly the schedule on which a belief gets built.
What it is relief from, on their reading, is nicotine withdrawal. The cigarettes are curing a condition the cigarettes created, and the relief gets filed as a benefit rather than as a repair. Which would mean that what smokers are most afraid of losing when they stop is a thing they would be getting back.
The objection that has to be answered
The obvious alternative reading is that the arrow points the other way. People whose mood is already improving are probably better placed to quit, and people in a bad patch are not, so quitters would look happier afterwards without quitting having done anything.
The authors take this seriously and have a decent answer. More than half of the studies they included were secondary analyses of randomized cessation trials, in which everybody was attempting to quit. In those studies the decision to try was not contingent on feeling better, and when the researchers split the results by whether they came from trials or from population cohorts, the difference between quitters and continuers held either way.
They also considered whether some third thing, a good stretch in a person’s life, might drive both the quitting and the mood, and argue this is implausible across follow-ups that run out to nine years.
What they do not do is claim proof. They write plainly that observational data can never establish causality, and note that a randomized trial in which people are assigned to keep smoking has problems no ethics committee would accept. They also say they found little evidence of publication bias but cannot rule it out.
The half about the calendar
The other number in the supplied premise is “years ago”, and that one has an answer too, from the longest-running study of its kind.
Richard Doll and colleagues recorded the smoking habits of 34,439 male British doctors in 1951 and followed their deaths for fifty years. Men born between 1900 and 1930 who smoked cigarettes and kept smoking died on average about ten years younger than lifelong non-smokers.
Stopping bought some of that back, and the amount depended on when. Cessation at 60 gained about three years of life expectancy. At 50, about six. At 40, about nine. At 30, about ten, with a survival pattern close to men who had never smoked at all.
And one figure speaks directly to the age in the headline. At the death rates of the 1990s, the probability of surviving from age 70 to age 90 was 7 percent for cigarette smokers and 33 percent for non-smokers.
The three years bought by stopping at 60 is the number worth sitting with, because it is the one that contradicts the fatalism. It is smaller than the ten years available to someone who stops at 30, and it is not nothing.
What this evidence will not carry
The doctors’ study is one sex, one profession, one country, and men born between the 1870s and 1930. The researchers chose doctors deliberately, because they would describe their own habits accurately and could be traced through the medical register, which is excellent for the study and poor for generalising to everyone else. The paper’s own central finding is that the hazard changed sharply between birth cohorts, which is itself a warning against treating any one of these numbers as permanent.
The mental health meta-analysis is built entirely from observational studies, with all the limits its authors name.
And neither paper says what the supplied topic says. Nobody compared the happiness of over-70s who quit against over-70s who never started, or against over-70s who did not quit. The finding is about change within people, measured before and after, which is a different and narrower thing.
What is actually left
None of this is a rebuke, and it is not an instruction. Stopping is genuinely hard, harder after decades, and the people who do it late usually do it with help rather than by deciding to. Anyone still smoking at 70 is not the failed version of the person in the study.
What the evidence offers is narrower and more useful than a demand. It is a correction to one specific belief, held sincerely by millions of people, that the cigarettes are doing something for their mood that they would lose. On the best available measurement, the opposite is what happens.
Which is a strange place for a list of this kind to end up. The genre offers moods to quit and no evidence that quitting them is possible or that it would help. The one entry with fifty years of records behind it is a physical habit, and its most surprising payoff is not the years. It is the mood, which was the thing the whole list was about in the first place.