tudca / udca · graded evidence · pulled 2026-08-31

TUDCA (tauroursodeoxycholic acid)

The default liver supplement of every oral-steroid cycle has never been studied alongside an oral steroid. Not once. What it does have is a real drug-class pedigree — UDCA is approved medicine for cholestatic liver disease, with a measured dose-response.

The short version
  • Zero studies of TUDCA with any 17α-alkylated steroid — no trial, no case series, no pharmacokinetic interaction study. DATA (an absence, verified by search)
  • UDCA is evidence-based medicine for cholestasis — the most-used supportive agent in drug-induced cholestatic liver injury across 41 studies. DATA
  • The only dose-response RCT (PBC, n=96): 13–15 mg/kg/day beat 10, which beat 5 — recurrence 0% vs 2.9% vs 19.4%. Dose matters, and the effective dose is large. DATA
  • The injury it guards against is real and characterized: stanozolol cholestasis arrives after ~55 days, high bilirubin with normal GGT. DATA

The absence at the center

“TUDCA on cycle” has never been tested DATA
Systematic negative (searched 2026-08-31, Europe PMC, many phrasings): no trial, no cohort, no case series, and no pharmacokinetic study of TUDCA (or UDCA) co-administered with any 17α-alkylated anabolic steroid exists. The most popular pairing in harm-reduction practice has a literature of exactly zero papers. Everyone dosing it on-cycle is running an uncontrolled experiment on themselves.

What UDCA actually has behind it

EvidenceDesignDoseResult
He 2026RCT, primary biliary cholangitis, n=965 / 10 / 13–15 mg/kg/dbiochemical recurrence 19.4% / 2.9% / 0% — clean dose-response
Bessone 2025review, 41 studiesUDCA = most widely used supportive agent in cholestatic DILI
Werner 2026cohort audittarget ≥13 mg/kgonly 61.8% of PBC patients actually reach the effective dose
The hepatology evidence base is about cholestatic disease in general — PBC, DILI, cholestasis of pregnancy — never about steroid-induced injury specifically.

The injury being guarded against

Stanozolol cholestasis: the cleanest steroid-liver registry DATA
Latency registry n=18: cholestatic injury emerged at a median 55 days of use, with high bilirubin and — the diagnostic signature — GGT in the normal range. This is a pure cholestatic (bile-transport) phenotype, not hepatocellular crash, which is precisely the injury pattern UDCA-class drugs address in other diseases.
For oxymetholone the risk is quantified without any protectant DATA
The 16-week HIV-wasting RCT (n=89, no TUDCA anywhere in the trial): ALT >5× baseline in 35% at 150 mg/day, 27% at 100 mg/day, 0% on placebo. That is the measured baseline risk the supplement industry sells against — and the trial that measured it included no liver protectant at all.

Dose — the only anchor that exists

If you extrapolate at all, extrapolate from 13–15 mg/kg MODEL
Premise (data): the one RCT dose-response shows 13–15 mg/kg/day is where UDCA's effect lives in cholestatic disease; 5 mg/kg demonstrably fails. Premise: typical supplement servings are far below that per-day level for most adults. Logic: sub-13 mg/kg/day dosing borrows the drug's reputation without its pharmacology. Prediction: low-dose TUDCA is functionally a placebo with a bile-acid label — untested, and the one dose curve we have points the wrong way for it.

Other claims in circulation

Insulin sensitivity: no human RCT DATA
The “TUDCA fixes insulin resistance” line rests on cell and animal work plus one small open-label ulcerative-colitis pilot (1.75–2 g/day, n=13, preprint) — no randomized human trial with an insulin endpoint exists.

What does not exist

No on-cycle trial · no prophylaxis trial · no PK interaction · no dosing study in healthy adults DATA
Verified by search: no study of TUDCA during AAS use of any kind, none testing it as prophylaxis against 17aa liver injury, none on whether it changes steroid blood levels, and no dose-finding study in healthy adults at all. Its reputation is a reasonable mechanism (choleretic bile acid, cholestatic injury phenotype) wearing the credibility of an unrelated disease's RCTs.