How much testosterone is converted to estradiol?

[Nelson Vergel]

A man with 1500 ng/dL of testosterone (many men on ExcelMale.com ) may have

1500 x 0.004 = 6.0 ng/dl = 60 pg/ml of estradiol

Some doctors may choose to treat 45 pg/ml but I disagree and always have.

This would create a T/E ratio of 1500/60 = 25

We have some fertility data on T/E ratios:

Is Testosterone to Estradiol Ratio Important in Men?

 

[Dr Justin Saya MD]

A man with 1500 ng/dL of testosterone (many men on ExcelMale.com ) may have

1500 x 0.003 = 4.5 ng/dl = 45 pg/ml of estradiol

Some doctors may choose to treat 45 pg/ml but I disagree and always have.

This would create a T/E ratio of 1500/4.5 = 15

We have some fertility data on T/E ratios:

Is Testosterone to Estradiol Ratio Important in Men?

I hope not too many have a consistent T level of 1500ng/dL (not my patients anyway...unless very high SHBG and consequently need higher total T to increase free T).

I've been trying to glean some useful clinical info from free T to E2 relationships.

Thus, lets say for a patient with a more appropriate T level of 1000ng/dL.

The 0.3% (0.003) conversion factor would equate to an estradiol of 3ng/dL = 30pg/mL. This also lies within the classic range we see for FREE T levels with total T levels around 1000ng/dL (typically free T will be in the neighborhood of 25-35pg/mL...sometimes significantly higher or lower depending on concurrent SHBG levels).

Specifically, I'm looking into patterns to see if anything "jumps out" with regards to symptoms, quality of life, libido, etc for the different scenarios: free T < estradiol , free T = estradiol, free T > estradiol. I'm starting to lean towards patients generally feeling "better" when: free T equals or greater than estradiol, as opposed to less than estradiol. However, there are so many moving parts/variables that it is a complex analysis.

So for the hypothetical patient with total T 1000ng/dL , estradiol 30pg/ml, would *want* free T equal to or greater than 30pg/mL. Again, this is preliminary and a tough analysis as everyone is different and cases of very high or very low SHBG can really skew things and alter even the most reliable patterns. I'm going to keep watching these parameters for clues to better help my patients.

 

[johndoesmith]

I hope not too many have a consistent T level of 1500ng/dL (not my patients anyway...unless very high SHBG and consequently need higher total T to increase free T).

I've been trying to glean some useful clinical info from free T to E2 relationships.

Thus, lets say for a patient with a more appropriate T level of 1000ng/dL.

The 0.3% (0.003) conversion factor would equate to an estradiol of 3ng/dL = 30pg/mL. This also lies within the classic range we see for FREE T levels with total T levels around 1000ng/dL (typically free T will be in the neighborhood of 25-35pg/mL...sometimes significantly higher or lower depending on concurrent SHBG levels).

Specifically, I'm looking into patterns to see if anything "jumps out" with regards to symptoms, quality of life, libido, etc for the different scenarios: free T < estradiol , free T = estradiol, free T > estradiol. I'm starting to lean towards patients generally feeling "better" when: free T equals or greater than estradiol, as opposed to less than estradiol. However, there are so many moving parts/variables that it is a complex analysis.

So for the hypothetical patient with total T 1000ng/dL , estradiol 30pg/ml, would *want* free T equal to or greater than 30pg/mL. Again, this is preliminary and a tough analysis as everyone is different and cases of very high or very low SHBG can really skew things and alter even the most reliable patterns. I'm going to keep watching these parameters for clues to better help my patients.

Here's a good example, my free test and E2 follow pretty closely, usually free test is 3 or so pg/ml lower than my E2.

I don't have any experience where this was different, so I can't exactly compare with the other scenarios you listed.

In terms of improvement before TRT to now there's a huge difference for me in almost everything except maybe libido, but that I wasn't too concerned with, and may just be my personality or due to my life circumstances at the moment.

I feel as if I'm "leveling out" so to speak now, I assume due to the honeymoon period ending and this becoming my new normal, hopefully not losing benefits! I do wonder if eventually I will forget how bad I felt before TRT, as time goes on and there is less of a contrast.

If there was a way to rate different symptoms, like metrics I would. I guess a 0-10 scale could work?

I'd be interested in volunteering my experience and information for a survey or case study.

I do appreciate the in depth response, explanation, and knowledge Dr saya. Sincerely.

 

[Nelson Vergel]

How to Predict Estradiol and DHT at Different Testosterone Doses

I was able to come up with a table using the predictive model equation derived from data in this study: The Effects of Injected Testosterone Dose and Age on the Conversion of Testosterone to Estradiol and Dihydrotestosterone in Young and Older Men Summary This video discusses a research paper...

www.excelmale.com

 

[S1W]

Interesting comments from Dr. Saya. I’m very curious to hear from the hive as to how many actually have Free T > E2 without using an AI.

With my own labs and with numerous labs from others I’ve seen here, when Free T is in the upper ranges (25-30) E2 seems like it is usually higher.

I’m a lean, fit guy with SHBG around 23, and when my Free T is in the mid to upper 20s, my E2 is usually in the mid to upper 40s.

 

[Nelson Vergel]

Free T is always higher than E2. Why? 2 percent or more of total T is usually free T. Only 0.4% of total T aromatizes to estradiol.

The units used by LabCorp for free T are wrong (Quest uses the right units). For example from the results posted above by @johndoesmith , total T of 914 ng/dl multiplied by 0.02= 18.34 nd/dL, or 183.4 pg/mL. His free T with LabCorp is showing 20.6 pg/mL when it should be 206 pg/mL. His sensitive estradiol is 29.1 pg/mL.

I have explained this before. Dr. Saya is aware of this issue. LabCorp has been notified but never cared to change this.

AI's have no influence on free T. T dose is related to higher or lower SHBG. The higher the T dose, the lower the SHBG and the higher the free T. But, obviously, higher T doses also increase hematocrit and decrease good cholesterol(HDL) more. So, it is a balancing act.

 

[Nelson Vergel]

We have very limited data on testosterone to estradiol ratios (ng/dl divided by pg/ml) of around 14 or more needed for proper sperm count in men not on TRT (data was derived using the regular estradiol test which overestimates estradiol). No data on specific ratios for libido and ED although there are some small studies that mention it without giving specifics.

Is Testosterone to Estradiol Ratio Important in Men?

Couple months ago Nelson had a take on E management and understanding E follows T makes perfect sense and treating symptoms vs numbers on bloodwork. There was alot of conversation about a ratio of keeping E2 (14-20) using the math of T÷E2. I recently moved to Nashville from So Cal and starting...

www.excelmale.com

I have never seen data on free T/ sensitive estradiol ratio but it would be interesting to have a study look at this.

 

[S1W]

Free T is always higher than E2. Why? 2 percent or more of total T is usually free T. Only 0.4% of total T aromatizes to estradiol.

The units used by LabCorp for free T are wrong (Quest uses the right units). For example from the results posted above by @johndoesmith , total T of 914 ng/dl multiplied by 0.02= 18.34 nd/dL, or 183.4 pg/mL. His free T with LabCorp is showing 20.6 pg/mL when it should be 206 pg/mL. His sensitive estradiol is 29.1 pg/mL.

I have explained this before. Dr. Saya is aware of this issue. LabCorp has been notified but never cared to change this.

AI's have no influence on free T. T dose is related to higher or lower SHBG. The higher the T dose, the lower the SHBG and the higher the free T. But, obviously, higher T doses also increase hematocrit and decrease good cholesterol(HDL) more. So, it is a balancing act.

Thanks for the reply. I guess I’m still struggling with the math and rationale here.

For example, using that approach a guy could have a TT of 250 ng/dl and a sensitive E2 of 45 pg/ml.

250x0.02 = 5 ng/dl = 50 pg/ml

Using that math his FT of 50 would be greater than his E2 of 45, though I can’t imagine that anyone would feel ok with a TT of 250 and E2 of 45.

Also struggling to think of a case where anyone’s FT wouldn’t be greater than E2 using that approach.

Where am I going wrong here? I’ll likely never be in a position to debate the likes of @Nelson Vergel and @Dr Justin Saya MD and that’s not what I’m doing here.

I just love learning from you guys and feel like I’m missing something here.

 

[Cataceous]

Rather than saying that LabCorp's units for the direct free T test are wrong, it might be better to say that they are measuring some other parameter that is somewhat correlated with free testosterone. The correlation of direct free T with actual free T is only so-so, and you should not simply multiply by 10 to make the comparison. Equilibrium dialysis tests are said to be the gold standard for measuring free testosterone, but even with these I don't think there is yet harmonization allowing direct comparisons between different labs. The Tru-T free T calculation is supposed to correlate well with equilibrium dialysis. And the parameters it uses—total T, SHBG and albumin—suffer from less variation between different labs. Therefore it is a good candidate for generating a more universally comparable value, which may also be used in these ratios.

It seems likely that among the various possible ratios, the relative amounts of the free hormones should have the most importance. That is, how does free estradiol compare to free testosterone? This comparison accounts for differing SHBG levels. It's already known that for a fixed free estradiol level—which is the primary HPTA regulator—lower SHBG means that free testosterone is relatively lower than free estradiol. This is thought to be one of the reasons why men with low SHBG face greater challenges.

@S1W: It does appear that in 2016 Dr. Saya was relying on the LabCorp direct free T test with normal range 7.2-24 pg/mL. But to require estradiol to be lower than these numbers seems to be pushing it pretty low. The conversion factor from testosterone to estradiol that he's citing, 0.3%, seems at the very low end of normal to me. I would have said 0.4-0.5% is more typical.

 

[Nelson Vergel]

Rather than saying that LabCorp's units for the direct free T test are wrong, it might be better to say that they are measuring some other parameter that is somewhat correlated with free testosterone. The correlation of direct free T with actual free T is only so-so, and you should not simply multiply by 10 to make the comparison.

No. That is not what is the case.

 

[Cataceous]

No. That is not what is the case.

If I'm understanding EIA correctly, they're basically measuring the intensity of a color, which makes the units arbitrary. So from a technical point of view the units are wrong and LabCorp won't change them because of the inertia of years of testing with this scale. Nonetheless, from a practical standpoint I think their direct free T test is still better characterized as I described it: measuring some other parameter that is somewhat correlated with free testosterone. This emphasizes the inaccuracy of the test method; it is poor enough that it should not be used for any important decisions. Changing the units will not improve the test's substandard correlation with actual free testosterone levels, so in fact this conversion should be discouraged.

 

[Nelson Vergel]

Sometimes I don’t know why we have to discuss things that are obvious mathematically. If you call a test free testosterone, you should be reporting free testosterone. That is what the patient or their insurance paid for.

 

[Cataceous]

... If you call a test free testosterone, you should be reporting free testosterone. That is what the patient or their insurance paid for.

If only things were so simple. @madman has extensively documented why these immunoassay-based tests should not be used. A sample:

The older, direct analogue RIA methods have been discredited and are no longer recommended for use [6–8].[R]

An analog-based free testosterone immunoassay reported free testosterone test results that were related to total testosterone concentrations under varied experimental conditions. This alleged free testosterone assay did not detect serum free testosterone (the test results it reported were nonspecific) and should not be used for this purpose.[R]

... this class of assay has been criticized as having poor accuracy, sensitivity, and between-assay comparability and being influenced by the dilution of serum (2)(3)(4)(5)(6). It is of note that the reference interval for free T by analog-based assay is much lower (concentrations about one-fifth as high) than that for the equilibrium dialysis assay (2). This calibration difference between assays is a major problem.[R]

The problems with LabCorp's direct free T test are even borne out in my results. Using results from tests on the same days, see how the Tru-T free testosterone calculation compares to LabCorp's direct test results. Linearity is the expected result:

 

[Nelson Vergel]

No need to make this more complicated than it is. They are using the wrong units and they should know better. Look at the two graphs you just posted. The units are different but the number is the same.

 

[Cataceous]

... Look at the two graphs you just posted. The units are different but the number is the same.

If you "fix" the units by multiplying the direct values by 10 then you get to the same order of magnitude, but the data sets are still not remotely comparable. Even setting aside the lack of linearity, the direct measurements have a slope that is half that of the Tru-T values, implying that the scale factor should be more like 15 to 20, and then a substantial offset correction is needed.

There is a way to simplify all this: Don't use the direct free T test. Instead use equilibrium dialysis or ultrafiltration or Tru-T.

 

[tareload]

If you "fix" the units by multiplying the direct values by 10 then you get to the same order of magnitude, but the data sets are still not remotely comparable. Even setting aside the lack of linearity, the direct measurements have a slope that is half that of the Tru-T values, implying that the scale factor should be more like 15 to 20, and then a substantial offset correction is needed.

There is a way to simplify all this: Don't use the direct free T test. Instead use equilibrium dialysis or ultrafiltration or Tru-T.

Excellent discussion guys. This has given many people fits as you have to understand what the direct free T test is actually measuring and that it is only correlated with free T (add on that the next layer of complexity talking about Vermeulen calc free T vs Tru-T and the non-scientifically training person will have his or her head hurt really bad.

But I can understand a provider using it as a windsock

Benefits of Estrogen for TRT Patients

I thought people slept in hyperbaric chambers to INCREASE their HCT, so they could have more endurance? I mean it is just “oxygen in the blood” right? This patient feels good? He is nearly anemic. Do anemic people feel good? Hey he feels good forget about those labs they don’t men nuffin! ~...

forums.t-nation.com

Benefits of Estrogen for TRT Patients

Well my natural T was 265 (shudder)

forums.t-nation.com

 

[tareload]

If only things were so simple. @madman has extensively documented why these immunoassay-based tests should not be used. A sample:

The older, direct analogue RIA methods have been discredited and are no longer recommended for use [6–8].[R]

An analog-based free testosterone immunoassay reported free testosterone test results that were related to total testosterone concentrations under varied experimental conditions. This alleged free testosterone assay did not detect serum free testosterone (the test results it reported were nonspecific) and should not be used for this purpose.[R]

... this class of assay has been criticized as having poor accuracy, sensitivity, and between-assay comparability and being influenced by the dilution of serum (2)(3)(4)(5)(6). It is of note that the reference interval for free T by analog-based assay is much lower (concentrations about one-fifth as high) than that for the equilibrium dialysis assay (2). This calibration difference between assays is a major problem.[R]

The problems with LabCorp's direct free T test are even borne out in my results. Using results from tests on the same days, see how the Tru-T free testosterone calculation compares to LabCorp's direct test results. Linearity is the expected result:
View attachment 11096View attachment 11097

Combine both plots into one using "parity plot" to show your point better (remove dose variable from the combined plot). After "correction" into same units for abscissa and ordinate, it's clear that the slope is not close to 1 whether you use calc Tru-T, calc Vermeulen or actual eq-dialysis data. Hence, it's not a units issue.

 

[Cataceous]

Combine both plots into one using "parity plot" to show your point better (remove dose variable from the combined plot). After "correction" into same units for abscissa and ordinate, it's clear that the slope is not close to 1 whether you use calc Tru-T, calc Vermeulen or actual eq-dialysis data. Hence, it's not a units issue.

If the dose/linearity illustration isn't needed then I have a lot more data, which I plotted previously:

 

[tareload]

If the dose/linearity illustration isn't needed then I have a lot more data, which I plotted previously:
View attachment 11146

Very nice. Nelson's factor of 10 solution sorta works with the Tru-T numbers (slope would be closer to 1.5 ish if you used the calc Vermeulen numbers (for someone with reasonable SHBG). So direct RIA off by a factor of 14 (in comparison with Tru-T) or ~5-6 (in comparison with Calc Vermeulen free T).

Thanks for putting that graphic together.

This plot looks similar to your plot:

https://www.tandfonline.com/doi/abs/10.3109/13685538.2013.835800?journalCode=itam20

In recent years, the RIA method has been criticized by
some experts as inaccurate due to substantial numerical
discrepancies between RIA and EqD results, and too heavily
influenced by either TT [7] or by SHBG [8]. The numerical
differences lead to confusion in interpreting clinical results
and can complicate efforts to establish biochemical standards
for the diagnosis of TD. Based on these issues, some experts
advocate that the RIA method should not be used in clinical
care and that conclusions of previous research studies
utilizing the RIA method may not be valid [9]. Nonetheless,
RIA remains in widespread use and some authors have argued
it provides clinically meaningful information in men when
interpreted using an assay-specific reference range [10].
Currently, there is a paucity of data directly comparing the
two methods against the gold standard in a clinical population
of men. Our goal in this study is to examine the relationships
between RIA, cFT, EqD, and TT in a population of men
presenting to an outpatient andrology clinic.

 

[tareload]

Sometimes I don’t know why we have to discuss things that are obvious mathematically. If you call a test free testosterone, you should be reporting free testosterone. That is what the patient or their insurance paid for.

In the case of RIA direct method, they should only charge you 16% +/- of the regular price since you've got to push all those units around.