Tracking supplement/med use
I'm trying to figure out how best to track supplement and medication use. The main criterion for the optimal system, for me, is that it be systematic enough that it will be easy to convert it to whatever format might be needed later; for example, the format needed by some soon-to-be-available machine learning system that finds correlations between behaviors and health outcomes.
I dug around, and found many apps that claim to track supplements, but 1) these seem mostly to be geared towards reminding the user to take something (I don't need that) and 2) it's not clear that the raw data will be available to the user.
Has anyone faced a similar question and come up with a good solution?
I maintain a spreadsheet with three tabs:
1) "Currently taking" and "currently considering" drugs and supplements and a short rationale, dose, where I got it from.
2) Blood tests where each row is a particular test and columns are dates. This one also computes Levine phenotypic age for each column automatically.
3) A log where 1 st column is date, 2nd column is supplement/drug/treatment name, 3rd column mentions start/stop, rationale, how I feel, if I experienced any sides etc.
Is this easy to maintain? Yes. Does it have all the information I need? Yes, pretty much. Does it record foods for every meal? No. Can I do ML with it? Probably. Though I never tried.
I keep a daily text note of what I am doing. I have a weekly routine for supplements with different groups of supplements. I also record weight, blood pressure and all the fitbit sense recordings. I also track Light Therapy, O2 varioxia and Molecular Hydrogen and anything else that seems worth recording that is material (odd health issues, some food, alcohol, test results). That goes into a text document. At the moment I am also tracking blood glucose every 5 minutes with a Dexcom G6 CGM.
Once a week I extract from this daily figures for some of the information.
A key thing I do, however, is to isolate out from this where possible one element. Some supplements such as Vitamin D have multi day effects, but by taking a bolus dose and then stopping it is possible to track that sort of thing. Things like menaquinone-7 actually have an effect on the mitochondria that can be seen in the daily figures.
It is a bit chaotic, but I am managing to get quite a bit of useful information. I look particularly to find outcomes that I can reliably link to inputs.
For example I have just switched from NMN to NR. I don't expect to see any variation in the outcome, but am likely to run with this for a few weeks.
Thanks for the feedback.
I'm already doing something similar to what you both describe.
But I want much more precision in the stored data: timestamps for all lifestyle choices (exercise, intake of food and supplements, etc.); symptom and body state details, etc.
If CRON-O-Meter weren't so irritating I might use the paid version, which allows time stamps, then create custom entries for all supplements and meds. Then I could download a copy of the data (a CSV file) at the end of each day, then add symptoms, reports on well-being, etc. in separate columns.
Overall, I'm not thinking about what patterns in behaviors and outcome I might see on my own, but, rather, what a ML system could make of the data (now or in the future).
Here's an amazing lecture by Michael Snyder about what's possible now by way of finding useful correlations in health data: