23 And Me, Worth It Or Waste?

This is a frequent question amongst those interested in knowing and doing as much as one can for healthspan (living for as long as possible in optimal health). Most who purchase gene testing kits from companies such as 23 and Me, Ancestry.com and MyHeritage.com are interested because these and other companies claim to be able to determine your ancestry. Others are interested because they believe that the testing will provide actionable health optimization.

These companies are not actually providing a “map” of one’s entire genome, but instead are looking at much relatively smaller portions of it called SNP’s (for “single nucleotide polymorphisms”). These are places on the genome where we have identified variations in one or more nucleotides in the DNA that is part of or near a particular gene. These variations are biological markers correlated with various diseases and other expressions that may be considered favorable or unfavorable. Eg, one SNP may be associated with a greater chance of having higher levels of LDL cholesterol. Another may be associated with an ability to thrive with daily sleep of only 6.25 hours per night.

There are several issues with these types of gene testing that should be recognized in determining their usefulness:

Not Statistically Precise

These relatively lower cost tests for SNP’s are not very accurate because they are not statistically precise. The machines that perform the procedures to identify the nucleotides and their position are not perfect, but their accuracy improves statically with each run – meaning with each attempt by the machine to identify the nucleotide and its position.

Just like the more times you flip a coin, the more likely you are to count the statistically expected equal number of heads versus tail, instead of say a set of 10 flips in which you see 7 heads and 3 tails (this is a very loose analogy, but technically, we are really addressing what is referred to as “Base Call Accuracy” and the probability of incorrect base call). Run times on these big Illumina (the leading manufacturer of the gene sequencing equipment) machines costs money, so to keep the costs down, some of these of companies offering “gene testing” will only check certain areas on the genome, SNPs for which we currently have information such as disease correlation or ability to sleep less as mentioned above, BUT, not necessarily ALL the SNPs correlated with each. And, these companies will only pay for a certain number (typically fewer) of runs.

The result is less accurate information which means you may indeed be as Germanic as your last name, Mueller, implies, but because of the relatively few number of runs (coin flips, or, more technically a low “base call accuracy”) and/or the limited number of appropriate SNPs being tested, the results show you inaccurately as more Italian.

This begs the question as to what information you are really getting versus in what you’re really interested. Ie, there are differences between ancestry versus race, heritage or culture, and in which you are truly interested, and I’ll leave for the reader to determine.

Already Known Information

With few exceptions, MOST of what can be determined via SNP or other (complete) gene testing as correlative to YOU and YOUR health has already manifested. Ie, you either already know about the expression of the gene variant (SNP for elevated LDL cholesterol, eg) or there is nothing you can do about it anyway (the SNPs for eye color, eg).

But let’s say that in the case of your test, the results are fairly accurate and you identify a SNP in your genome that is correlated with higher levels of LDL cholesterol. How useful is this information really? If you are a typical person growing up in a Western Industrial society, you have by now undergone a blood draw for assays of your lipids (cholesterol). You are treating it or not, or perhaps just watching it or ignoring it. So, what value does the SNP testing actually provide you?

But what about Alzheimer’s disease? Let’s say I’m only 40 years old so won’t my identifying the SNP correlated with Alzheimer’s disease (something that typically doesn’t begin showing signs and symptoms until much later) be useful to me? Perhaps, but…

Low Bearing Results

Usually, more than one gene or SNP is not only correlated with a disease or expression, but also may correlate positively OR negatively. So, you might have one or more SNPs that increase your chances of developing a disease and one or more that decrease your chances of developing a disease, AND, for each SNP, a separate “chance” or probability correlated with developing the disease.

So, to make a very simple example, let’s say you have 3 SNPs that are positively correlated with developing a condition of elevated LDL cholesterol and 3 SNPs that are negatively correlated with developing elevated LDL. The 3 positively correlated SNPs are each 25% correlated and the 3 negatively correlated SNPs are each 20% correlated. What does that mean to the individual with the SNPs? Net 5% increased RISK? Not much of an increase, and, if it does manifest, does that mean it (my LDL cholesterol) will be dramatically or modestly increased? We don’t really know. Risk is a “window” of sorts, and just like the risk of, eg, dreaded Alzheimer’s Disease, even with the APOE-e4 gene, only approximately 40% of those diagnosed with Alzheimer’s Disease have this variant. And, not only does the APOE-e4 variant simply represent an increased risk, it is one of several (at least three other SNP variants) risk factors associated with development of Alzheimer’s Disease.

So, how much bearing does the APOE-e4 variant (or ANY of the SNPs we are looking at) really have upon us? It would certainly motivate me to be more proactive about preventing cognitive decline with age if I knew I had a significantly (even mild) increased risk. But what about the increased LDL cholesterol variant or the like that have already manifested or not? Presumably, if concerned about elevated LDL cholesterol, I would have already addressed this – if elevated LDL cholesterol is/was present -without the need for a SNP test to tell me I have a predisposition for it.

So, what might be a reason for SNP or full genome testing?

First, if one is concerned about the possibility of passing on certain genes that are latent (meaning not expressed and affecting you, but could affect your offspring) in you and your mate, then gene testing can be considered proactive.

Second, for a few diseases/morbidities like Alzheimer’s Disease whose onset is later in life, knowing about a predisposition may help you avoid or delay onset.

Third, while it could be undertaken on a SNP by SNP basis, a full, one-time genome sequencing or “mapping” can provide one with information that is not only more accurate (full genome mapping by Illumina is more accurate because, as described previously, many more runs (“coin flips”) are performed to ensure statistical accuracy), but more cost efficient and useful over one’s life as more and more correlations are made between the genes and their variants and diseases and expressions. Remember that your genome is your genome, and, unless artificially manipulated by technology such as CRISPR, doesn’t change.

Recently, the effectiveness of certain drugs, from statins to hypertensive medications to cancer chemotherapy, has been correlated with certain SNPs. Having these variants already identified would speed and ease the process of more effective treatment for a variety of illnesses if necessary.

In addition, so many new correlations with SNPs are being found daily (last estimate I received from Illumina is at least 50 per month), that to have one’s full genome sequenced makes it a fairly simple process to check, eg, if you’re only sleeping 6.25 hours per night is harmful (typically) or if you are one of the few individuals with the genetic variant (“DEC2”) which allows you to sleep only 6.25 hours per night with the same efficacy as another without this variant who sleeps just over 8 hours per night. (Man, I wish!).

With free online software called Promethease which interfaces with SNPedia, a library of SNPs, one can check one’s genome against the most up to date information regarding SNPs and their correlation to disease and expression. And, with costs for whole genome sequencing continually decreasing (Illumina currently offers it for $1,000 vowing to get that to $100 in the near future), it seems reasonable to obtain given the potential beneficial information it can provide.