Does Personality Predict How Long You Live?
By PsyCove Editorial · 11 min read · Updated 2026-10-27 · Checked against the sources listed below
- The effect is real and modest. A meta-analysis of 20 independent samples put the correlation between conscientiousness and longevity at .11 (Kern & Friedman, 2008).
- One standard deviation is worth years, in a large cohort. In 2,359 adults followed for decades, a standard deviation higher on conscientiousness was associated with a 27 percent lower risk of death (Terracciano et al., 2008).
- It is not just behavior you can see coming. The association held largely independently of smoking and obesity, which are the two explanations people reach for first.
- It cannot be applied to a person. These are group averages from decades of follow-up. Nothing in them supports a statement about how long you, or anyone you know, will live.
What Does the Research Actually Measure?
Every study in this area measures the same shape of thing: whether people who scored higher on a trait at one point were more likely to be alive at a later point.
That design produces an association between a score and a risk of death. It does not produce a cause, a mechanism or a forecast. The distinction is not pedantry: it decides whether a finding can say anything at all about one person. The association would be an artifact if the cohort studies had not adjusted for smoking and socioeconomic position; they adjust, and it survives.
The word the literature uses is mortality risk, and risk is a statement about a group. If a trait is associated with a 27 percent lower risk in a sample, that number describes how the fates of many people were distributed. It does not resolve into a quantity of years owed to any one of them.
How Large Is the Conscientiousness Effect?
Larger than the dismissive summary and much smaller than the popular one, which is the usual position for a finding that survives scrutiny.
The most-cited meta-analysis combined 20 independent samples and found a correlation of .11 between conscientiousness-related traits and longevity, with an interval from .05 to .17. The association was strongest for two facets: achievement, meaning persistent and industrious, and order, meaning organized and disciplined (Kern & Friedman, 2008).
The record goes back further than the meta-analyses. Children first assessed in the 1920s were followed for six decades, and ratings on traits grouped under conscientiousness in childhood predicted who was still alive in old age (Friedman et al., 1993). Whatever the pathway is, it appears to start early and run long.
A single large cohort gives the numbers people find easier to hold. In the Baltimore Longitudinal Study of Aging, 2,359 adults were assessed from 1958 onward and 943 of them died during follow-up. Each standard deviation higher on conscientiousness was associated with about 27 percent lower risk of death. Emotional stability was worth about 15 percent and general activity about 13 percent (Terracciano et al., 2008).
Among the deceased, those a standard deviation above the mean on any of those three traits had lived two to three years longer on average. That is a substantial difference in years, and it is the origin of most of the headlines.
- Conscientiousness: 27 percent lower risk of death per standard deviation (Terracciano et al., 2008).
- Emotional stability: about 15 percent lower risk.
- General activity: about 13 percent lower risk.
- In years: those a standard deviation above the mean on any of the three had lived two to three years longer.
What Does a Correlation of .11 Mean in People?
One way to translate a correlation into something visible is to ask how often a higher scorer would outlive a lower scorer, chosen at random.
That translation is mechanical: a correlation of r corresponds to roughly 50 plus 50r percent of pairs going one way. At .11, the figure is about 55 percent. Pick one high scorer and one low scorer at random and the high scorer is more likely to be the longer-lived of the two, with the odds sitting near 55 to 45 (Kern & Friedman, 2008).
| Expression | Value | What it means |
|---|---|---|
| Correlation with longevity | .11 | A small association, comparable to many well-established effects in social science |
| Chance a higher scorer outlives a lower scorer | ≈55% | Better than a coin, and nowhere near a rule |
| Risk difference per standard deviation | 13–27% | Large enough to matter at population scale and useless as an individual forecast |
Both statements are true at once. That is what a small effect over a long period looks like: nearly invisible in a pair of people, substantial across a city.
Why Would a Trait Affect Lifespan at All?
Because traits shape behavior across decades, and behavior is what accumulates into health.
The review literature groups the pathways into a few families, and none of them requires the trait to act directly on the body (Chapman et al., 2011).
- Health behavior. Conscientious people are more likely to sleep, exercise, wear seatbelts and less likely to smoke heavily. The effect held largely independently of smoking and obesity in one cohort, so this is not the whole story (Terracciano et al., 2008).
- Adherence. Following a treatment plan for years is a trait-shaped behavior. A regimen abandoned in month three does not care how good it was.
- Occupational and material pathways. Traits influence what jobs people get and keep, which influences income, which influences health. This is a selection pathway rather than a biological one.
- Social pathways. Relationship stability and social contact both predict health, and both are partly shaped by how agreeable and how stable a person is.
The honest version of the mechanism is therefore unglamorous. It is not that discipline is medicinal. It is that discipline makes a hundred small decisions go one way instead of another, and the hundredth decision is made thirty years after the first.
What Can This Not Tell You?
A correlation between a trait and lifespan cannot tell you how long you will live, and it cannot be used to estimate that, for three separate reasons.
- The unit is the group. Every figure above describes how risk was distributed across hundreds or thousands of people. Individual outcomes include car accidents, infections and diagnoses that no questionnaire sees coming.
- Correlation is not a lever. Knowing that a trait is associated with longevity does not mean raising the trait raises the years. The causal direction is partly the reverse: serious illness changes personality.
- The effect is small at the level of a person. A correlation of .11 is not a signal you can read off one life. Using it on an individual is like using a city's average rainfall to decide whether to take an umbrella tomorrow.
- A model fitted on people with your age, sex and health history, rather than on the whole sample.
- An interval rather than a point, because no individual estimate in this literature comes with one.
- A comparison against the base rate for someone in your position, which is the number that would actually matter.
There is also a measurement limit worth stating. A trait score is a position taken at one moment. People move within their own range across weeks and shift on average across decades, so the trait that predicted an outcome in a 1958 cohort is not a fixed possession of the person who reads this (Fleeson, 2001; Roberts et al., 2006). The ordering of people on a trait is steadier than the absolute positions of individuals, so a cohort finding can survive even as every member of it moves (Roberts & DelVecchio, 2000).
We design the result page to report four positions and no health estimate, because nothing a four-dimension score contains could support one. Our data on this question is borrowed from cohort studies, and their unit of analysis is a group of hundreds or thousands.
Is an Effect This Small Worth Knowing About?
Yes, for a reason that has nothing to do with prediction: it changes what you notice.
A finding that a trait is associated with longevity, through behavioral pathways you could name, points at the pathways rather than at the trait. Sleep, adherence, follow-up appointments and social contact are all modifiable. The trait is not. The research is useful precisely because it directs attention to the modifiable part.
The same logic explains why the finding does not license fatalism in either direction. “I am low on conscientiousness so it makes no difference” and “I am high on conscientiousness so I am safe” are both misreadings of a group average, and the second one is the more dangerous of the two because it feels like good news. Set against other predictors of the same outcomes, these effects are comparable in size to socioeconomic status and cognitive ability (Roberts et al., 2007), which is another way of saying they are one input among several rather than a verdict.
Where Does Type Language Fit Into This?
Nowhere, and the absence is worth naming rather than papering over.
The literature above is built on continuous trait measures, mostly the Big Five. There is no comparable body of research on four-letter codes and mortality, and the reason is structural: the codes are cut from the same trait space, so a study of type would be a study of those traits with most of the resolution removed (Bess & Harvey, 2002).
If someone tells you that a particular type lives longer, ask for the study. The honest answer is likely to be a Big Five result restated as a type claim, with the precision lost on the way through. The cautionary literature on the indicator has been making this point for decades: claims made on its behalf routinely outrun what it measures (Pittenger, 2005).
Is This a Medical Test?
A personality score is not a diagnostic instrument, and it cannot be used to estimate how long you will live. The studies cited report group-level associations between a questionnaire score and mortality records. They do not produce an individual prognosis.
The distinction matters because the vocabulary sits close to medicine, and three things get lost in the translation.
- Association is not risk. A hazard ratio describes how two groups differ on average. Turning it into your risk needs your baseline risk, which a personality score does not contain and cannot estimate.
- No pathway is measured. The plausible routes from a trait to mortality run through health behavior, occupational sorting and adherence to treatment. Those are measured by other instruments. The questionnaire measures none of them.
- Clinical questions need clinical answers. If you are worried about your health, your sleep or your mood, the useful next step is a clinician rather than a questionnaire. This site reports four positions on four dimensions and stops there.
Read the effect sizes as descriptions of populations, which is what they are. That is interesting enough without borrowing the authority of a medical test. If you want a question for your next appointment, "does anything in my health record interact with how I handle stress?" is worth more than a personality score.
What Should You Take From This?
Four things, none of which involves estimating your own lifespan.
- Keep the size of the effect. A correlation of .11 is a real signal and a weak one. Any version of this finding that sounds decisive has dropped the number.
- Look at the pathways, not the trait. If the association runs through sleep, adherence and social contact, those are the things worth attention, and they are open to everyone regardless of score.
- Do not convert a group risk into a personal one. The same study that reports a 27 percent risk reduction reports it for a standard deviation, in a sample, over decades.
- Be suspicious of type-shaped versions. Four letters are not a measure of conscientiousness, and the mortality literature does not contain them.
Frequently asked questions
Does conscientiousness really predict how long you live?
It is associated with longevity across studies, with a meta-analytic correlation of .11 across 20 samples (Kern & Friedman, 2008). That is a small association that is consistent enough to be worth studying and far too small to use as a personal forecast.
How many years does personality add?
In one large cohort, people a standard deviation above the mean on conscientiousness, emotional stability or activity had lived two to three years longer on average (Terracciano et al., 2008). That is a difference between groups of people, not a number of years attached to an individual.
Is it just that conscientious people smoke less?
Partly, but not entirely. The association in the Baltimore cohort was largely independent of smoking and obesity, which suggests the trait also works through adherence, employment and social pathways (Chapman et al., 2011).
Can I change my personality to live longer?
Traits do shift on average across adulthood (Roberts et al., 2006), but the research links traits to outcomes without showing that raising a trait raises lifespan. The actionable part is the behavior the trait is associated with, which can be changed directly.
Does my MBTI type predict lifespan?
No research supports that. The longevity literature uses continuous trait measures, and type codes are cut from the same dimensions with most of the resolution removed (Bess & Harvey, 2002). A type-based longevity claim would be a trait finding restated with less precision.
Further Reading (primary sources & authoritative references)
- Kern, M. L., & Friedman, H. S. (2008). Do conscientious individuals live longer? A quantitative review. Health Psychology, 27(5), 505–512. — DOI
- Terracciano, A., Löckenhoff, C. E., Zonderman, A. B., Ferrucci, L., & Costa, P. T. (2008). Personality predictors of longevity: Activity, emotional stability, and conscientiousness. Psychosomatic Medicine, 70(6), 621–627. — DOI
- Friedman, H. S., Tucker, J. S., Tomlinson-Keasey, C., Schwartz, J. E., Wingard, D. L., & Criqui, M. H. (1993). Does childhood personality predict longevity? Journal of Personality and Social Psychology, 65(1), 176–185. — DOI
- Chapman, B. P., Roberts, B., & Duberstein, P. (2011). Personality and longevity: Knowns, unknowns, and implications for public health and personalized medicine. Journal of Aging Research, 2011, 1–24. — DOI
- Roberts, B. W., Kuncel, N. R., Shiner, R., Caspi, A., & Goldberg, L. R. (2007). The power of personality: The comparative validity of personality traits, socioeconomic status, and cognitive ability for predicting important life outcomes. Perspectives on Psychological Science, 2(4), 313–345. — DOI
- Roberts, B. W., Walton, K. E., & Viechtbauer, W. (2006). Patterns of mean-level change in personality traits across the life course: A meta-analysis of longitudinal studies. Psychological Bulletin, 132(1), 1–25. — DOI
- Fleeson, W. (2001). Toward a structure- and process-integrated view of personality: Traits as density distributions of states. Journal of Personality and Social Psychology, 80(6), 1011–1027. — DOI
- Roberts, B. W., & DelVecchio, W. F. (2000). The rank-order consistency of personality traits from childhood to old age: A quantitative review of longitudinal studies. Psychological Bulletin, 126(1), 3–25. — DOI
- Bess, T. L., & Harvey, R. J. (2002). Bimodal score distributions and the Myers-Briggs Type Indicator: Fact or artifact? Journal of Personality Assessment, 78(1), 176–186. — DOI
- Pittenger, D. J. (2005). Cautionary comments regarding the Myers-Briggs Type Indicator. Consulting Psychology Journal: Practice and Research, 57(3), 210–221. — DOI
Related reading
- Is personality inherited? — how much of a trait is stable enough to be worth linking to anything.
- Which systems have real evidence? — where the trait literature sits relative to the type literature.
- Personality and stress — one of the proposed pathways, examined on its own.
- Where MBTI stops — why four letters cannot carry a mortality claim.
- The four dimensions this site measures are on the test page, and they are not a health measure.