Research: DHEA

TL;DR Evidence has weakened over time. Recent meta-analyses (Conforti 2025, Zhang 2023) and the largest RCT (Wang 2022, n=821) show DHEA gives ~1 extra egg per retrieval, lower gonadotropin dose, shorter stim, and possibly lower miscarriage rate — but no average live-birth benefit. The subgroup that does respond appears to be women with low baseline testosterone who efficiently convert DHEA to T (Gleicher 2013). My low T (0.4 nmol/L) and low-normal DHEA-S (4.5 µmol/L) put me in that subgroup. Standard dose: 25 mg × 3/day = 75 mg/day for 8–12 weeks pre-stim. Reversible androgenic side effects.

Papers

Fertility and Sterility · 2025

Conforti A et al. Therapeutic management in women with a diminished ovarian reserve: a systematic review and meta-analysis of randomized controlled trials

F&S

Finding: 38 RCTs, n=1,336 on DHEA arm. DHEA increased oocytes retrieved by +0.60 (p=0.03). No significant difference in mature eggs, clinical pregnancy, live birth, or miscarriage.

Frontiers in Endocrinology · 2023

Zhang J et al. Efficacy of dehydroepiandrosterone priming in women with poor ovarian response undergoing IVF/ICSI: a meta-analysis

DOI

Finding: 32 studies (14 RCTs). In the RCT-only subgroup, DHEA increased AFC (+1.18), reduced basal FSH (−1.99), reduced gonadotropin dose (−382 IU), shortened stim by ~1 day, and reduced miscarriage rate (RR 0.46, p=0.001). No effect on clinical pregnancy or live birth in the RCT subgroup. Authors recommend ≥3 months for max effect.

J Assist Reprod Genet · 2013

Gleicher N et al. Starting and resulting testosterone levels determine IVF pregnancy rates in women with DOR

PMC

Finding: In DOR women on 75 mg/day DHEA for ≥6 weeks, baseline DHEA-S did NOT predict pregnancy — but cycle-start free testosterone did. Each increase in free testosterone improved pregnancy odds 2.6-fold. Signal: conversion efficiency (DHEA → T), not baseline androgens per se, predicts who benefits. Caveat: CHR/Gleicher group has a known publication bias.

Summary

Outcome	DHEA effect	Source
Oocytes retrieved	+0.60 (significant)	Conforti 2025
AFC	+1.18 (significant)	Zhang 2023
Gonadotropin dose required	−382 IU	Zhang 2023
Stimulation days	−1 day	Zhang 2023
Miscarriage rate	RR 0.46 (significant)	Zhang 2023
Mature (MII) oocytes	No significant difference	Conforti 2025
Clinical pregnancy rate	No significant difference	Both
Live birth rate	No significant difference	Both + Wang 2022 RCT (n=821)

Dose / duration: 25 mg × 3/day = 75 mg/day, micronized pharmaceutical grade, for 6–12 weeks pre-stim (Zhang notes ≥3 months for max effect).

Mechanism: DHEA → testosterone/DHT in the ovary → androgen receptors on granulosa cells → amplified FSH responsiveness, increased follicular recruitment from the primordial pool.

Relation to my history

My Dec 2025 baseline: testosterone 0.4 nmol/L (low; trial average ~1.0) and DHEA-S 4.5 µmol/L (low-normal). Per Gleicher 2013, the women who respond are those who efficiently convert DHEA into testosterone. Whether I'm a converter is unknown until tested — but the headroom for conversion is large. The recent meta-analyses don't promise a live-birth boost on average, but the miscarriage-rate reduction (RR 0.46) is interesting given my prior negative cycle.

Confidence: how well each finding applies to me

Finding	Confidence	Note
DHEA gives ~1 extra egg per retrieval	High	Consistent across recent meta-analyses.
DHEA reduces gonadotropin requirement and shortens stim	High	Practical / logistical signal.
DHEA reduces miscarriage rate	Medium	Zhang 2023 RCT subgroup; Conforti 2025 didn't find this. Intriguing but uncertain.
DHEA does NOT improve live birth on average	High	Wang 2022 (largest RCT) + 2023/2025 meta-analyses agree.
I'm in the subgroup most likely to convert / respond	Medium	Low T + low-normal DHEA-S = theoretical fit. Conversion efficiency not measured yet.
Side effects are reversible at standard dose	High	Acne, oily skin, mild hirsutism; voice deepening rare and dose-dependent.

Questions to raise with CZ clinic (likely Repromeda Dr. Filková on May 25)

Do you routinely use DHEA priming for DOR patients with my profile (low T, low-normal DHEA-S, age 40)?
What dose and duration — is 75 mg/day for 8–12 weeks your standard protocol?
Will you measure baseline and cycle-start total + free testosterone and SHBG to confirm I'm converting DHEA effectively?
What's the stopping criterion if I develop androgenic side effects?

Caveats

The largest RCT (Wang 2022 BJOG, n=821 across 9 Chinese centers) found no LBR benefit on the primary endpoint. This is the single most robust trial and it's negative.
Older Gleicher/CHR cohort studies were more optimistic — selection bias and publication bias likely played a role.
ESHRE 2025 ovarian stimulation guideline treats DHEA as low/very-low certainty evidence; no strong recommendation either way.
Contraindications: hormone-sensitive cancers, liver disease, untreated PCOS.