Saw a thread today from a 19yo with crashed libido whose bloodwork came back ‘very high’ and he didn’t know what to do with that. I’ve seen this pattern enough times on this sub now that I want to think out loud about it, because the standard answer (it’s psychological, lower your stress, get more sleep) isn’t satisfying when the labs are this divergent from the symptoms. Before I push back on the conventional take, the strongest version of it is real. Libido is genuinely a CNS phenomenon as much as a hormonal one, dopamine and prolactin and sleep architecture matter a lot, and plenty of guys with picture-perfect labs have a dead libido for non-endocrine reasons. I’m not dismissing that. But here’s what I keep coming back to. Total T is a confounded number. It includes: - free T (the bioavailable fraction, usually 1-2%)
- albumin-bound (weakly bound, kind of available)
- SHBG-bound (locked up, not bioavailable) If your SHBG is elevated, your total T can look great while your free T is functionally low. SHBG goes up with thyroid hormone, with certain meds, with low insulin states, with liver stuff, with aging. A 19yo with high SHBG and high total T can absolutely have free T in the bottom quartile. My hypothesis (and I want pushback on this): a non-trivial chunk of the ‘high T but no libido’ cases on this sub are SHBG-bound total T elevations being misread as functional androgen sufficiency. The patient sees the total number, the doc sees the total number, everyone says ‘labs are fine’ and moves on. The actual bioavailable hormone is unremarkable or low. What would falsify this? A few things: 1. If we actually saw free T measured (calculated or direct) in these cases and it tracked with total T, my hypothesis is wrong.
- If SHBG measurements in symptomatic high-T guys came back normal-to-low, my hypothesis is wrong.
- If estradiol was the better predictor of libido state than free T in this cohort, my hypothesis is incomplete at best. The assay piece matters too. LC-MS/MS vs immunoassay can give meaningfully different totals at the high end. A 19yo with a ‘very high’ immunoassay result might be reading closer to upper-normal on mass spec. I’m a biostatistician not a clinician so take the mechanism with appropriate salt, but the pattern is statistical enough that I think it’s worth asking: when total T is high and libido is dead, has anyone actually looked at free T, SHBG, and estradiol in the same panel? genuinely curious what people have seen here, especially anyone who got the full panel and figured out what was actually going on.
The framework here is more rigorous than most of what this question usually gets, and I mean that sincerely. The SHBG hypothesis is mechanistically sound, and laying out the falsification criteria is more intellectual honesty than most threads bother with. The part I’d add to, though, is that calculated free T via the Vermeulen equation assumes a fixed albumin around 4.3 g/dL, and that assumption tends to break down in exactly the patients where SHBG is already confounding the total T picture, so you can end up chasing a calculated free T that’s still not accurately representing what’s actually available. Direct free T by immunoassay is, if anything, less reliable at the individual level. The estradiol framing you called “incomplete at best” is actually where I’d push hardest, bc elevated E2 can suppress libido centrally even when free T looks adequate once you correct for SHBG, and a single cross-sectional panel can’t separate those two explanations without knowing whether the E2 elevation is primary or downstream from aromatization of the elevated total T. The re-challenge design that would actually disentangle them basically never gets done.
The Vermeulen-albumin point is the one I most want to concede, you’re right that assuming 4.3 g/dL is doing real work in cases where the patient already has something atypical going on, and the people most likely to get a calculated free T ordered are exactly the people most likely to violate that assumption. Fair. Where I’d push back, gently, is on framing E2 as the dominant axis here. The “elevated E2 suppresses libido centrally” claim has a surprisingly thin evidence base when you actually try to pin down the effect size, vs the noise from sleep, prolactin, dopaminergic tone, relationship variables, etc. I’ve seen the AI-letrozole and clomid-recovery cohorts where guys with wildly different E2s reported essentially indistinguishable libido outcomes. Mechanism plausible, outcome correlation weaker than the forum consensus treats it as. Your point that a single cross-sectional panel can’t separate aromatization-driven E2 from primary E2 elevation is exactly right, and it’s the same problem the SHBG story has. We’re all reading tea leaves from one snapshot. The re-challenge design would settle it. It will never get funded. So we keep arguing about n=1 panels on the internet, which is, I guess, what this sub is for.
The case for treating E2 as secondary noise is genuine, and I take the prolactin/sleep/dopaminergic confounders seriously. But clomid and letrozole cohorts are almost the worst possible test for the specific E2-libido claim. Clomid is a SERM that blocks central estrogen receptors directly, so serum E2 and receptor-level signalling are decoupled in that population regardless of what the blood result says. Letrozole cohorts have T and E2 moving in opposite directions simultaneously, so you can’t isolate E2 as the variable of interest at all. Comparing libido outcomes across either of those isn’t a test of what E2 does in isolation, it’s a test of what happens when you change several things at once. I’d be more persuaded by the weak-E2-correlation conclusion if it came from a design where E2 was actually the thing varying.
eta: one more thing