Yes, there is a genetic component to your healthspan, lifespan, and longevity. Well, if you have some essential background in genetics, I would assume you already knew this. In this blog post, I want to show you how you can get a partial insight into your health span, lifespan, and longevity status using 23andMe data. I am intentionally saying partial insight as these traits (healthspan, lifespan, and longevity) are rather very complex and controlled by hundreds if not thousands of genetic variants (we will assume, among others, mainly SNPs)
But first, to explain what the terms healthspan, lifespan, and longevity
What are healthspan, lifespan, and longevity?
There are many different definitions for these terms or how these are also called ‘aging phenotypes,’ but I will use the shortest ones:
Healthspan is a period of life lived in good health
Lifespan is a total of years lived. It includes healthspan and non-healthspan
Longevity refers to the ability to live a long life beyond the species-specific average age at death. For example, for humans (even though it varies from country to country), the average age at death for humans is between 72 and 75 years.
As mentioned above, I wanted to explore the genetics behind healthspan, lifespan, and longevity.
I did this by using publicly available data from genome-wide association studies focused on healthspan, lifespan, and longevity and combining it with my personal 23andMe data to check how I am standing with longevity. This blog post basically outlines this small experiment of mine. As I wrote, I used genetic variants (SNPs) that are readily available in my 23andMe data, which can give you a bit more information about your healthspan and longevity. You can learn more about how to download your raw 23andMe data from this page.
Important to underline when we talk about healthspan, lifespan and longevity, we refer to three different measures of aging or those that are often called “aging phenotypes”. As I briefly explained above, there is a clear difference between those three terms. However, it is essential to underline that there are slightly different ways how healthspan as well as other phenotypes, can be measured. When we link this information to genetics (by doing GWAS studies) these differences are also reflected in the genetic variants (genes) associated with given measured phenotypes. But, these are nitty gritty stuff that I don’t want to go into at this stage (I will try to devote a separate post on that to explain it in more detail). In any case, we will be focused on all the genome-wide association studies focused on identified genetic variants (genes) being associated with healthspan, lifespan,e and longevity.
In Short, How Did I Do It?
It is important to note right away that this is far from being a report aiming to provide a definitive report on genetic predisposition for having longer healthspan/lifespan. Still, it will give you more insights into what your genetic component of healthspan and longevity looks like based on this set of SNPs that are highly associated with healthspan and longevity.
So, how did I do it?
STEP 1: I looked into genome-wide association studies that were focused on identifying genetic variants that are highly associated with healthspan, lifespan, and longevity. I reviewed over 5800 genome-wide association studies and looked for SNPs highly associated with those traits, healthspan, lifespan, and longevity. I ended up with 31 SNPs with very low P-values (with the highest value being around 6x10-06 and SNP with the lowest P-value being 1x10-126)
STEP 2: I took those 31 SNPs and checked how many of those are available in my 23andMe data, and out of those 31 SNPs, 14 (one duplicated in two studies) happened to be available in my 23andMe data.
STEP 3: The final step was to extract my genotype information for those 14/13 SNPs and compare it with the effect allele for those same SNPs in corresponding GWAS studies. Here, I did not build any fancy scoring system (like polygenic risk scoring), even though I should have done that sometime in the future. Assuming additive effect, I created a simple visual table with three effect scenarios: orange (genotype with no effect allele), light green (genotype with one effect allele), and green (genotype with both effect alleles).
Genes from 23andMe Data Being Associated With Healthspan, Lifespan and Longevity
Here, I will shortly present those 13 SNPs and corresponding genes strongly associated with healthspan and lifespan and also available in 23andMe data:
SNP rs4970836 is located between the CELSR2 and the PSRC1 genes. This study links rs4970836 to aging phenotypes.
SNP rs1230666 is located in the MAGI3 gene. This study links rs4970836 to aging phenotypes.
SNP rs12203592 is located in the IRF4 gene. This study links rs4970836 to aging phenotypes.
SNP rs34831921 is located in the HLA region, specifically between HLA-DRB1 and HLA-DQA1. This study links rs4970836 to aging phenotypes.
SNP rs10455872 is located in the LPA gene. This study links rs4970836 to aging phenotypes.
SNP rs7859727 is located in the CDKN2B-AS1 gene. This study links rs4970836 to aging phenotypes.
SNP rs1126809 is located in the TYR gene. This study links rs4970836 to aging phenotypes.
SNP rs4268748 is located in the DEF8 gene. This study links rs4970836 to aging phenotypes.
SNP rs6511720 is located in the LDLR. This study links rs4970836 to aging phenotypes.
SNP rs429358 is located in the APOE. This study links rs4970836 to aging phenotypes.
SNP rs7859727 is located in the CDKN2B-AS, and these two studies' Multivariate genomic scan implicates novel loci and haem metabolism in human aging and Identification of 12 genetic loci associated with human healthspan link rs7859727 (CDKN2B-AS gene) to aging and healthspan, respectively.
SNP rs34872471 is located in the TCF7L2 gene. This study links healthspan to rs34872471 (the TCF7L2 gene).
SNP rs3205087 is located in the DFFB gene. This study links increased health span in old age to rs3205087 (the DFFB gene).
SNP rs1123415 is located between the UBASH3A gene and the RSPH1 gene.
What do 23andMe data tell about my healthspan, lifespan, and longevity?
Table 1 illustrates all those 13 SNPs and also my results.
Trait | SNP & Effect Allele | Gene | My 23andMe Genotypes |
Aging traits* | rs4970836-G | CELSR2, PSRC1 | GG |
Aging traits* | rs1230666-G | MAGI3 | AA |
Aging traits* | rs12203592-C | IRF4 | CC |
Aging traits* | rs34831921-A | HLA-DRB1, HLA-DQA1 | AC |
Aging traits* | rs10455872-A | LPA | AA |
Aging traits* | rs7859727-C | CDKN2B-AS1 | CT |
Aging traits* | rs1126809-G | TYR | -- |
Aging traits* | rs4268748-T | DEF8 | CT |
Aging traits* | rs6511720-T | LDLR | GG |
Aging traits* | rs429358-T | APOE | CT |
Aging traits* | rs7859727-C | CDKN2B-AS1 | CT |
Healthspan | rs7859727-T | CDKN2B-AS1 | CT |
Healthspan | rs34872471-C | TCF7L2 | TT |
Increased healthspan in old age | rs3205087-G | DFFB | AG |
Increased healthspan in old age | rs1123415-A | UBASH3A, RSPH1 | AG |
Based on the genotype results from 23andMe data (based on those 13 available SNPs), I have a relatively good genetic predisposition toward longer age (lifespan) and increased health span. To explain again the color coding:
Orange means that genotype with no effect alleles. So, no positive effect, and we don't have any alleles with negative effects. So, we will consider these genotypes as population averages.
Light green means that the genotype has one effect allele, and assuming an additive genetic model it has a positive effect, but not like genotypes with two effect alleles.
Dark green means that the genotype has both effect alleles, and assuming an additive genetic model, it has a double positive effect compared to genotypes with only one positive effect allele.
Also, I have to underline again and say that this is based only on those 13 highly significant SNPs. You can use your 23andMe raw data and do the same thing for yourself. If you want, drop me a message, and I will do it for you for free.