What is the “prog study”?

• A clinical trial conducted by a team at Amsterdam UMC where they randomized six groups of 15 post-op transfem people to receive either standard hormone therapy (estrogen alone), estrogen + progesterone, or similar combinations with double the dose of estrogen or progesterone.
• A protocol for the study explaining what was planned was published in 2023 (Dijkman et al., 2023).

Where can I read the full study?

• The full study has not been published.
• Some early results were presented at the conference of the European Professional Association for Trans Health (EPATH) in September 2025.
• The conference abstract from EPATH (Dreijerink et al., 2025), reproduced below¹, is the only published material on the results.
• However, a member of the community was at the presentation and took a photo of the results slide (see below).

What did the study find?

• The control group receiving the standard regimen had an average breast volume increase of ~10% over the study period.
• The groups receiving double the typical dose of estrogen plus either the the typical progesterone dose (group EEP, receiving 200 mg daily) or double the typical progesterone dose (group EEPP, receiving 400 mg daily) displayed larger average increases.
• These increases were reported as statistically significantly larger than the control group (but see caveats below).

Results slide

Picture of results slide taken at EPATH 2025

(*) indicates a statistically significant effect (p < 0.05)

Important caveats

Methodological

• The “standard estrogen dose” at Amsterdam UMC is whatever gets the person into the Dutch target range of 200-400 pmol/L (54-109 pg/ml) (Conemans et al., 2022). This is significantly below the WPATH standard of 367-734 pmol/l (100-200 pg/ml) (Coleman et al., 2022). Thus, the “double estrogen dose” groups were actually receiving a dose in line with international recommendations.
• While all progesterone was meant to be taken orally, people on Reddit who claimed to be involved in the study² have said that some participants opted to take it rectally without telling the study authors. In any event, actual progesterone levels for any group in the study have not been reported.
• The reported percent increases compared participants in one group to participants in the control group, rather than comparing participants’ breast sizes before and after treatment. In other words, the double E + 200 mg P group didn’t have a “37% increase in breast volume”– rather, their increase was 37% greater than the increase the control group had (confusing isn’t it).
• The authors did not measure breast size after discontinuing progesterone. progesterone is known to cause fluid retention which can impact breast size (Dillard et al., 2025), but reverses if progesterone is discontinued. Whether the effect shown in the study is permanent growth or simply fluid retention is therefore unclear.
• 15 people per group (total 90 people) is a small number of participants, and the confidence intervals and error bars for each group are super wide, making it hard to extrapolate. Hopefully further studies will include larger participant groups.

Statistical

• We do not know whether they controlled for confounding variables such as BMI.
• (allegedly, needs confirmation) A member of the community who attended the presentation reports that during Q&A the presenter admitted observing an inverse effect after controlling for BMI. Specifically, there was a trend towards reduced breast volume in the EP and EPP groups. Note that the use of the term “trend” is common parlance in applied statistics for an “apparent but statistically non-significant, or borderline significant” effect. We of course do not know if this was the intended use.
• If so, this would represent an instance of Simpson’s paradox, a classical yet all-to-common statistical phenomena encountered in clinical studies when important confounding variables are not properly accounted for.
• We do not know if the statistical analysis includes an appropriate p-value correction, necessary in the presence of multiple hypothesis testing.
• The original protocol (Dijkman et al., 2023) called for a statistical analysis that would treat predictors, stratification, interaction terms, and confounding variables appropriately using a mixed model approach, which would address the previous concerns. At the conference, however, the results were presented as five simple percent differences between each individual treatment group and the control. These differences appear to have been analyzed separately, rather than as fixed-effect coefficients within a single model. This reporting style suggests that the full mixed model analysis has yet to be performed or could not be completed. Possible reasons include missing data or poor data quality arising for example from participants’ failure to adhere to the protocol.

Bottom line: does progesterone promote breast growth?

• Unfortunately this is still unclear.
• Better designed studies (including serum progesterone measurements, more participants, a discontinuation phase, and better statistics) will be needed to convincingly answer this question.
• Alternatively, if the 2023 protocol was in fact carried out in full as originally designed, then we simply need to wait patiently for the final and complete analysis of the data to pass peer-review and be published.
• It’s common for preliminary results like this to be used to justify funding for larger, better trials. Having a ton of hype around said preliminary results probably helps, even if it’s not justified.
• In any event, hopefully this study will prompt Amsterdam UMC and other Dutch clinics to increase their standard estradiol target range and offer progesterone to anyone who wants to try it.
• See Aly (2020, update #3 2025) over at TFS for an alternative and slightly more optimistic take on this study.

Bibliography