Individual Genetic Contributions to Aging and Longevity are Tiny to the Point of Being Swamped by Statistical and Technical Differences
The genetics of natural variations in human longevity is an interesting subject for study, and there is great enthusiasm for genetics and gene therapy in this day and age, but nonetheless the genetics of longevity has next to no relevance to the future of medicine to treat aging. The results from a great many studies have shown that the contribution of each gene is tiny, and associations between gene variants and aging are only rarely replicated between study groups, suggesting that genetic contributions are (a) highly dependent on one another, and (b) highly dependent on environmental circumstances. The same gene in different human lineages, or the same gene in the same lineage with a different diet or lifestyle, will result in quite different tiny contributions to the pace of aging.
The effects are so small, in fact, that they are probably in many cases statistical or methodological artifacts: change the methodology used to gather or process the data, and different associations show up in the same study population, a point that is well illustrated in the research linked below. Even for the few genes in which variants do show fairly reliable associations, like FOXO3 and APOE, it is still the case that these are tiny effects in the grand scheme of things: perhaps some people have a 10% greater chance of reaching the age of 100 than would otherwise be the case. That would be enough to produce statistically significant enrichment of a gene variant in extremely old individuals. But are the mechanisms involved worth chasing in order to attempt to produce a therapy? How about if a collection of variants doubled the odds of making it to 100? No in either case. Not when there are far greater gains to be achieved via the SENS approach to human rejuvenation or similar strategies based on repair of cell and tissue damage.
In this article we clarify mechanisms of genetic regulation of human aging and longevity traits. The objective of this article is to address the issues in previous research of not reaching a genome-wide level of statistical significance and lack of replication in the studies of independent populations. We performed a genome-wide association study (GWAS) of human life span using different subsets of data from the original Framingham Heart Study (FHS) cohort corresponding to different quality control procedures, and we used one subset of selected genetic variants for further analyses. We used a simulation study to show that this approach to combining data improves the quality of GWAS with FHS longitudinal data to compare average age trajectories of physiological variables in carriers and noncarriers of selected genetic variants.
We used a stochastic process model of human mortality and aging to investigate genetic influence on hidden biomarkers of aging and on dynamic interaction between aging and longevity. We investigated properties of genes related to selected variants and their roles in signaling and metabolic pathways and showed that the use of different quality control procedures results in different sets of genetic variants associated with life span. We selected 24 genetic variants negatively associated with life span and showed that the joint analyses of genetic data at the time of biospecimen collection and follow-up data substantially improved significance of associations of 24 selected single-nucleotide polymorphisms with life span. We also showed that aging-related changes in physiological variables and in hidden biomarkers of aging differ for the groups of carriers and noncarriers of selected variants.
The results of these analyses demonstrated benefits of using biodemographic models and methods in genetic association studies of these traits. Our findings showed that the absence of a large number of genetic variants with deleterious effects may make substantial contribution to exceptional longevity. These effects are dynamically mediated by a number of physiological variables and hidden biomarkers of aging. The results of these research demonstrated benefits of using integrative statistical models of mortality risks in genetic studies of human aging and longevity.
"The effects are so small, in fact, that they are probably in many cases statistical or methodological artifacts: change the methodology used to gather or process the data, and different associations show up in the same study population, a point that is well illustrated in the research linked below."
I don't think it's well illustrated at all. All these longevity GWASes are small under-powered GWASes, and non-replication of genome-wide statistically-significant hits from one small study to another small study is expected even though most of those hits are genuine and have high posterior probability. The twin studies and GCTAs confirm that the additive variants for about a third of variance are there to be found, the lack of genome-wide statistically-significant results implies small individual effect sizes and thus high polygenicity, and so like most complex traits, it's just going to take large sample sizes (as we've already seen with intelligence and schizophrenia, where the naysayers were dispelled by hundreds of hits once sample sizes got to where they needed to be). And Farmingham is just not big enough. I also don't understand their focus on genotyping: genotyping and quality control issues will produce false *negatives*, not false positives as they keep saying at the end, because SNPs will be dropped from analysis and unavailable for the meta-analysis. (Dropping data reduces power which increases false negatives; random measurement error in genotyping introduces additional noise and likewise will reduce power / increase false negatives.) There's going to be way more measurement error in their phenotype measurements like BMI or life history than there will be in the SNP arrays.
If you or Yashin want to argue that aging will differ genetically dramatically over the course of a lifetime and from cohort to cohort, the right way to show this is to compute genetic correlations (https://en.wikipedia.org/wiki/Genetic_correlation) on full polygenic scores: this can estimate longitudinal stability within individuals (in twin studies) and genetic overlap between cohorts/countries/time-periods. Use LD score regression. Or if you can't do that, at least calculate power...
I've never personally understood the infatuation of a certain subgroup of aging researchers with statistical analysis of genetics. I don't foresee this having any therapeutic potential and it's not like they are implying it either.