I said I would do this a while back. Most of these were randomly pulled by staff and all names are redacted, so I cannot answer specific questions on any particular one. Some, yes, and you can probably figure out who they are, the outliers.
I think the take home is that numbers do not mean much and do not directly correlate with dosages or even each other. All of them are current and have been with us a while, so whatever symptoms brought them in have been resolved.
Thanks for posting. The hidden variables here are metabolic clearance rate and patient size which will affect their distribution volume. Are these patients all doing the same injection type at the same site?
Clearance can be weakly correlated with age from what Iāve seen and hence thereās usually some correlation between age and shbg. If you tried different doses on the same patient youād get a pretty straight line. Therefore, individual test level is correlated with dose as it has to be based on the scientific principles weāve discovered over time. The challenge is teasing apart the variables I mention above to create a reasonable model for how a patientās level will respond to a given dose beforehand.
In my day job I would pick this data set apart and show you how this would all fit together. Here though, i am slowly getting the message that its fruitless. As I usually find and someone told me long ago, complexity is in the eye of the beholder.
From the data Iāve seen on here, clearance rate seems to have a lot of variability across users ā even when accounting for other potential correlates such as age, weight, SHBG, liver function, etc. I would guess the strength between these variables and clearance rate to be somewhat weak - moderate , and as such, I think it would be difficult to develop such a model using these known variables as a proxy for clearance rate (though, admittedly, I havenāt tested it).
If you use R at all, there are some neat pharmacokinetic packages that predict serum concentrations across time. Iāve played around with them a bit.
Thanks for the feedback. Curious if you clustered MCR based on a few variables using principal component analysis, if the RMSE for the individual confidence limits would be significantly improved vs just a global model, which as you state may be of pretty limited predictive value. Thereās enough data in the literature where it would have been great to include in an analysis, but again this is a hypothesis to test. My point here is thereās more to the story and that dosage does correlate with level, you just have to tease apart the rest of the mess.
Right, I am aware of that, but I figured this would be ok without adding age, bodyweight, diet, CBC, LFTs, other medications and other information.
Some are cypionate, some enanthate. Most glute, maybe a few thigh. All 23g, 1 inch needles. Some are six days post injection, maybe some five. As noted, some less.
Right, if aiming for a number, good luck with that, which was really my point.
Yes, many overthink it. However, I do know that starting them at 150-200mg a week works well 90% of the time.
Itās the same explanation as with all the others. Thatās what works for him and he is happy with the results. Every patient he has referred is on at least double what he is. Itās always tempting to take the protocol that works for you and shove it down everyoneās throat. Itās also a mistake.
My point with this comment in bold (the first few words were left out of the quote) was it is probably fruitless to spend much time on this on here as it is really beyond the scope of this forum (hence it is most likely fruitless to spend this type of rigor here on this forum:
Taking a look at the data you did share (TT is in ng/dL and fT is in pg/mL) and performing a regression of TT, fT and SHBG, thereās a very nice correlation between fT and TT+SHBG. Leverage plots above show that (as has been shown previously), for a given TTā¦ free T is inversely proportional to SHBG. Hence a model similar to this or Vermeulen online calculator will allow one to typically predict fT within 10-20% if they know their TT and SHBG
This simple model predicts fT to within 5 ng/dL on average.
However, back to the original motivation and topic for why you posted these data. I agree with you that with the data you shared it is not possible to build an accurate model to predict TT/fT as a f(dosage and frequency). The hypothesis to test if we added additional effects like age, body mass, body fat, is that we could build a model to predict TT/fT to within say 10% if we used these additional variables to characterize central clearance. I donāt want guys on here to leave with the impression that dose and TT/fT levels are not explictly related, we just donāt have enough data in the table above to test this properly because the picture is incomplete and doing this type of activity is state of the art in the field.
Like @highpull mentioned, whether anyone would want to use such a model is another issue entirely and heās got a system that works for him.
TT is in units of ng/dL and fT in units of pg/mL. To convert fT to ng/dL, take the numbers in pg/mL and divide by 10. I believe these are Quest data looking at the ranges. @highpull could confirm.
Thatās impressive. Iāll have to take a closer look at the stats when I have time. Iāve love to get something published along these lines. if I had the time.
Yes, there is a TT, fT and SHBG correlation. But, no SHBG correlation with dose or injection frequency. At least that I see, anyway. Too many low SHBG guys do fine with once weekly dosing.
Agreed. If you go through the literature lots of times SHBG gets confused and confounded with age as an explanatory effect for dose vs fT serum level. Would be fun to take all the data available and look at clearance which is a function of a bunch of factors that govern the rate limiting step which is absorption:
@dbossa, dead (as in potentially dead) serious questions:
Last week you accused me of doing a cycle (you knew, you assumed?) when I ran into problems earlier this yearā¦
I went to the trouble to (as accurately as I humanly could, see details in linked post below) estimate what my TT and fT levels were during the unfortunate issue (TT between 1700 and 2600 ng/dL with fT between 30-50 ng/dL). I discontinued low dose, therapeutic AAS dosing 4 weeks before prior to the incident:
So now look at the data @highpull shared (you can check out the summary graphs I made above or go to the original post I did and zoom in). Some questions pop into my head:
Which one of these guys are on a cycle?
How do you determine this?
For guys with existing conditions (maybe they know/maybe they donāt), I infer from your comments on my nasty experience that there is potentially an upper limit on how high they should take their TT/fT? Do I understand that correctly? What is that limit? Could you consult your providers or perhaps you have an opinion? Why would guys with heart issues need to stay below a certain TT/fT level?
Based on the answer to question 3, should care be used on how high to initially start or continue a man on TRT/TOT since there may be issues (heart) that are unknown or they may be older? Should guys with these issues not exceed physiologic range?
Thanks for your thoughts on this. I also really appreciate any insight your team has on this. @highpull, Iād value your opinion on this as well.
@dbossa: Maybe thereās a silver lining from what I thought was some real unfairness/hypocrisy/BS in your back and forth with me on my AFIB incident. Perhaps this discussion could help some men who use exogenous T.
