Exactly 1 month ago, my first biological age measurement of 2020 was 32.75y (https://michaellustgarten.wordpress.com/2020/02/14/biological-age-32-75y-chronological-age-47y-first-2020-measurement/). When considering that my chronological age is 47y, that’s a 14 year improvement, but I wasn’t (and still aren’t) satisfied. When I sent my blood for analysis, I was battling a mild upper respiratory infection (cough, no fever), which likely raised my WBCs, thereby resulting in a higher biologic age. Also, I was experimenting with a higher intake of meat, eggs, and cheese, to see what affect that it would have on my circulating biomarkers. On that blood test in February, my creatinine levels were higher than my 2015-2020 average value, and if those foods were associated with circulating levels of creatinine, reducing them should also reduce creatinine, and accordingly, further improve my biological age. I also assumed that all other variables on Levine’s Phenotypic Age calculator would be unchanged.
On March 9 2020, I sent my blood for analysis so that I could calculate biological age with Levine’s PhenotypicAge. Almost exactly as expected, my WBCs (4.7 * 10^3 cells/microliter) were closer to my 2015-2020 average value (4.6), rather than the higher value (5.8) in my blood test last month. Similarly, reducing my intake of beef, eggs, and cheese brought creatinine from 1.08 to 0.97 mg/dL, which is closer to its 5-year average (0.94 mg/dL). As a result, I further reduced my biological age by 1.14 years to 31.61y, which is 15+ years younger than my chronological!
Because I track my diet every day, I can investigate the correlation between my meat, eggs, and cheese intake with creatinine. I now have 8 blood tests that correspond to dietary data, and interestingly, there is a moderately strong correlation between my average daily beef+egg+cheese intake with creatinine (r = 0.55). Based on these data, I’m going to continue to minimize consumption of these foods, with the goal of optimizing creatinine.
On a final note, I also expected to further reduce my CRP from 0.3 to something lower, but it slightly increased to 0.37 mg/L. While that is far from a high value, lower is better, and in future blood tests I’ll try to figure out how to further reduce it.
If you’re interested in calculating your biological age, here’s the Excel link:
Michael Lustgarten 
I have been doubting the accuracy of the Levine phenoage. I have seen enough talk out on the web saying people’s results did not even closely match DNA age methylation testing. I decided to do the blood draw for both at the EXACT same time to compare.
I have been running -10 to -13 for the last year which I have done 3 separate tests of the Levine phenoage.
My new test for phenoage is -13 and my MyDNAge results are only -2.
MyDNAge claims a .98 correlation for blood and you claim .94 correlation for phenoage. Which clock is correct?
LikeLike
Horvath’s initial DNAm clock, which is what DNAmAge is measuring didn’t correlate with smoking, so how accurate is it, really? More recent iterations like DNAmGrimAge are better, as smoking is correlated with epigenetic age acceleration, but that DNAm test is not yet commercially available.
In terms of Levine’s clinical measures, they’ve been studied for 50-100y+, so we know how they change with age, and what’s associated with an increased mortality risk. That’s why I place more value on that, when compared with the epigenetic clocks, for now.
LikeLike
They measure ~6 times the number of loci of the original Horvath clock. Below is the reply I received from MyDNAge about what they measure:
“Thank you for reaching out to us. We obtained the exclusive license for the 353 CpG sites of the original Horvath’s clock and we are the only entity that can provide commercial services for Horvath’s Clock age prediction. There are >2,000 CpG loci in our DNAge® panel which includes 90% (i.e. 336 loci) of the CpG sites in the original Horvath Clock.
Dr. Horvath published 2 additional clocks for similar utilities, i.e. PhenoAge and GrimAge after Horvath Clock. He claimed that GrimAge can be use to calculate lifespan and the time the donor have left to live. Unfortunately, the CpG loci for our DNAge® is not 100% overlapped with CpG loci in the PhenoAge. Out of the 513 CpG loci that is used in the PhenoAge, there are only 41 CpG loci overlapped with the original Horvath Clock and there are 42 CpG loci overlapped with our DNAge®. Dr. Horvath claimed that he used 1,030 CpG loci and additional protein markers for the recent GrimAge that published this year. However, the detail CpG loci information for GrimAge is not publicly available.”
Earlier I told you about my 71 year old friend who was a 50+ year smoker. He only recently quit smoking yet measured -13 with phenoage. How accurate is the smoking correlation with the Levine spreadsheet in his case? Phenoage which was used to derive DNAmPhenoage only correlates .10 with smoking.
IMO you might have the risk of death of the average 31 year old but I feel it is misleading to say you have that biological age.
LikeLike
In terms of your smoking friend with good data for PhenoAge, that’s anecdotal. In large population-based studies, having “good” values for its 9 clinical variables would put the individual at having a lower biological age and at a lower risk for all-cause mortality. That’s a population-based average. Is it applicable to all subject data? Obviously not, as your friend is likely an outlier.
What does biological youth look like? If you’ve read my posts, you know, for example, that abumin is high in youth, and decreases during aging. Conversely, creatinine increases during aging, and is used to derive the age-related decrease for eGFR. I’ve studied all of the clinical variables for PhenoAge, and beyond merely getting a “biological age” readout, isn’t part of the goal to have clinical markers that are prevalent in youth, and that are associated with a reduced risk of death? Independent of Levine’s PhenoAge, that’s what I mostly have.
You can argue that I’m being misleading about my “biological age”, but this is not subjective, it’s based on published data. If I’m 80 years old, and I have clinical markers of a 20yr old, am I really biologically 80? And yes, there are exceptions to that rule (i.e. your smoking friend), but most will not be like that.
LikeLike
So if you are 80 with clinical markers of a 20 year old you will likely live 60 more years to 140?
LikeLike
That’s unknown, but that’s the best that one can do until proper rejuvenation therapies are developed.
LikeLike
I just want to say thanks for posting all of your research and experiments. To me, yours is the most sensible approach and the best that can possibly be done to extend healthspan until more is learned.
LikeLike