I have consistently found that my HbA1C levels are disproportionately high compared to my actual glucose readings. Specifically, the estimated average glucose (eAG) is significantly higher than my measured average.
I have already pointed to the limitations of the rather (very) bad eAG study but I still think my HbA1C is higher than expected from the glucose.
That’s why I am now examining glycation independently from glucose regulation. I previously analyzed glucose regulation in detail here: A comprehensive and very deep dive into glucose dysregulation - Pathophysiology, Diagnosis, Medications, Genetics, etc
Because I found the insights from that glucose regulation genetic analysis so interesting and actionable, I have decided to skip a detailed description of various glycation pathways and move directly into analyzing the genetics of glycation.
So here it is:
Glycation_Pathways_Genetics.pdf (85.0 KB)
1 Like
Following that I analyzed my own specific genome and got that N of 1 report:
Clinical Implication: The patient’s HbA1c likely overestimates true glycemic exposure. CGM
data is a more accurate reflection of actual glucose control than HbA1c suggests. This should
be considered when making treatment decisions based on HbA1c targets.
Glycation_Pathways_Genetic_Report.pdf (202.6 KB)
2 Likes
Nice unpacking. I don’t follow all of the Tx but I get the directionality.
I may have missed something here… but tell me again what exactly you are doing here (or have done)?
Did you get your full genome sequenced? (from who)? Or just SNPs via 23andme?
Then you ran your results through Claude with a focus on understanding the glycation pathways?
Yes first by nebula.org then by sequencing.com. Both 30x level. The sequencing.com is of lower quality and less than 30x it seems. Then I downloaded the raw sequencing data files (fq), combined the 2 and re-sequenced that to get a 60x level vcf file.
Any sequencing service is OK as long as they let you download your DNA data.
I would advise to go to 100x now that it’s available as 30x will have some missing or low quality SNP.
It’s not possible because the files are 100’s of MB (compressed!) and also this would be a gigantic task as the pathways are all interconnected.
So what I do instead is ask Claude to look at all the pathways and the associated SNPs for some focused biological process like HbA1C and glycation here and make a first general document with the description of those pathways and SNPs.
Then I look up at those SNPs in my genome and feed that back to Claude along with my medication and supplement stack.
That where the magic appears! It has been able to pinpoint genetic issues in my glucose regulation and glycation with an astonishing precision.
A few examples that it correctly predicted just from my SNPs:
- I’m insulin sensitive
- My glucose homeostasis set point is too high
- My insulin release trigger mechanism is impaired
- My hepatic glucose is too high
- Metformin didn’t work for me but Imeglimin will
- I’m not very sensitive to GLP1 which is why I’ve not lost any weight at 6mg/week
I’ve know that through N of 1 experiments but Claude found that right away and added other things I could not know like that I have a beta-cell disfunction and issue with proinsulin processing and tRNA modification.
For the HbA1C/glycation pathways it has found out that I have an extended red blood cell lifespan, Increased reactive glycating agents and an NADPH/glutathione competition.
All that will inflate my HbA1C relative to my glucose level.
The third step is that Claude analyzes my medications and supplements stack to validate what I use, suggest dose adjustment and other potential medications/supplements to add.
In the end I get something incredibly precise and actionable for issues that have puzzled me for years.
As a caveat we generally don’t know the effect size of these SNP so trying to address them with medications and supplements may or may not have a significant effect.
2 Likes