Longevity genetics: what FOXO3, APOE and centenarian studies establish

Studies of people who live to extreme old age have turned up a handful of genuine longevity genes — most reliably the APOE gene, and in some populations FOXO3 [s1][s2]. But those signals explain only a small share of who reaches very old age: once you correct for the fact that people tend to marry others with similar lifespans, the true heritability of human longevity comes out well below 10% [s3]. Long life runs in families much less than the phrase suggests, and far less than a genetic test could ever predict for an individual.

How much of lifespan is inherited

"Long life runs in families" is usually supported by heritability estimates that have sat consistently in the range of about 15% to 30% [s3]. A 2018 study using Ancestry public family trees — pedigree data on hundreds of millions of historical people — reproduced those nominal figures, then showed why they mislead [s3]. Much of the apparent correlation between genetic relatives was also present between non-genetic in-law relatives, a pattern that points to assortative mating: people pair up with partners whose lifespan-influencing traits resemble their own [s3]. After modelling that effect, the authors concluded that the real heritability of longevity for birth cohorts across the 1800s and early 1900s was well below 10%, and had been generally overestimated [s3].

That reframes everything that follows. The genes below are real, but they operate inside a phenotype that is mostly not inherited.

FOXO3 and the insulin/IGF-1 pathway

The most reproducible longevity gene outside APOE is FOXO3, part of the insulin/IGF-1 signalling pathway that regulates ageing across many species. In a 2008 nested case-control study of long-lived American men of Japanese ancestry, variation in FOXO3A was strongly associated with reaching old age: the odds ratio for carrying two copies of the minor allele versus two copies of the major allele was 2.75 (P = 0.00009) [s1]. The long-lived men also showed lower rates of cancer and cardiovascular disease, better self-reported health, and biological markers of greater insulin sensitivity, several of which tracked with the FOXO3A genotype [s1]. FOXO3 associations have since been reported in other populations, which is why it is treated as a credible signal rather than a one-cohort fluke — though the effect is a shift in odds, not a guarantee.

APOE is the signal that holds up genome-wide

When researchers scan the whole genome rather than testing candidate genes, one locus dominates. A 2019 meta-analysis of genome-wide association studies used a strict definition of longevity — 11,262 people surviving to at least the age marking the 90th survival percentile and 3,484 reaching the 99th, against 25,483 controls who died at or below the 60th percentile [s2]. The APOE gene came through clearly: the ε4 variant (rs429358) was associated with lower odds of surviving to both the 90th and 99th percentile ages, while the ε2 variant (rs7412) showed the opposite, protective direction [s2]. A second signal near the gene GPR78 (rs7676745) was associated with lower odds of reaching the 90th percentile [s2].

APOE is already the best-known genetic risk factor for Alzheimer's disease and cardiovascular disease, so its appearance in longevity scans is partly a story about avoiding the diseases that kill people — the same genetic architecture that underlies specific illnesses shows up when you select for the people who escaped them longest [s2].

What this establishes, and what it does not

The honest summary is narrow. A small number of variants — APOE most robustly, FOXO3 in particular populations — genuinely shift the odds of exceptional longevity, and they cluster in pathways (lipid handling, insulin/IGF-1 signalling) that biology already implicated in ageing [s1][s2]. But they account for a fraction of a trait whose heritability, properly measured, is in the single digits [s3]. That is why no consumer genetic test can tell you when you will die, and why the search for centenarian "longevity genes" has not produced a target that translates into a therapy for everyone else.

It also reframes the Blue Zones debate: if inheritance explains under a tenth of the variance, the clustering of long-lived people in particular places is overwhelmingly a story about environment, behaviour and shared circumstances rather than shared DNA. The genes matter to the underlying biology — the same insulin/IGF-1 and other hallmark pathways mapped in ageing research — more than they matter to any individual's odds.

This article describes research findings and is not medical advice.

Sources

1. [s1] FOXO3A genotype is strongly associated with human longevity. Proceedings of the National Academy of Sciences, published online September 16, 2008. https://doi.org/10.1073/pnas.0801030105
2. [s2] A meta-analysis of genome-wide association studies identifies multiple longevity genes. Nature Communications, published online August 14, 2019. https://doi.org/10.1038/s41467-019-11558-2
3. [s3] Estimates of the Heritability of Human Longevity Are Substantially Inflated due to Assortative Mating. Genetics, published online October 31, 2018. https://doi.org/10.1534/genetics.118.301613