How Hormones May Shape Fear Learning—and What the Pill Changes

Women develop post-traumatic stress disorder at roughly twice the rate of men. Differences in trauma exposure, social support, and help-seeking explain part of that gap, but they do not close the case. Research on fear learning suggests that reproductive hormones may also affect how the brain updates a once-dangerous cue or context as safe.

The interesting question is not whether estrogen is universally protective. It is whether hormone state—and especially exposure to synthetic hormones—changes the brain systems that distinguish safety from threat. Human evidence points to a real possibility, but the results are uneven, and the clinical advice has not caught up with the science.

The safety circuit is hormone-sensitive

Extinction learning does not erase a fear memory. It teaches the brain that a cue or setting that once predicted danger is now safe. The ventromedial prefrontal cortex (vmPFC), hippocampus, and amygdala do much of that work: the hippocampus supplies context, the vmPFC helps regulate threat responses, and the amygdala detects emotional significance.

A 2025 randomized crossover fMRI trial by Stevens et al. tested whether estradiol can alter this circuit. The study enrolled 110 premenopausal Black/African American women from one urban community, including 34 women with current PTSD. Participants received a transdermal estradiol patch or placebo at two menstrual-cycle phases, and the researchers measured brain responses to social-threat cues.

The patch increased mean serum estradiol to 176.06 ± 89.28 pg/mL, versus 88.66 ± 63.58 pg/mL with placebo (t=6.52, p<0.001). Estradiol also changed BOLD responses in the basolateral amygdala, central amygdala, and vmPFC. The study found no phase-by-treatment interaction (p=0.51), so its results do not show that cycle phase itself determined the response. In this experiment, absolute hormone exposure mattered more than whether testing occurred in one of the two cycle phases.

The proposed biology is plausible but incomplete. Estradiol may facilitate safety learning through estrogen receptor beta, which is concentrated in the vmPFC and hippocampus. At higher exposure, estrogen receptor alpha activity in the centromedial amygdala could increase arousal or weaken inhibition. That is an inverted-U model: enough estradiol may help some functions, while more is not necessarily better. The trial also raised estrone, however, so it cannot isolate receptor-specific effects in people.

Estradiol is not a settled treatment story

This was the first positive result in this research line, not the final word. Stevens and colleagues report two earlier studies in healthy women tested in the early follicular phase in which exogenous estradiol did not change amygdala reactivity to negative stimuli. The 2025 sample included trauma-exposed women and a PTSD subgroup, and it used social-threat cues; those differences could matter. The PTSD subgroup contained only 34 participants, which limits confidence in subgroup conclusions.

For now, the trial supports a narrower claim: acute estradiol can modulate threat-related brain activity under particular experimental conditions. It does not establish that estradiol prevents PTSD, treats PTSD, or reliably improves fear extinction in healthy women. Brain activation is also a biomarker, not a patient outcome.

Progesterone may complicate extinction

Progesterone is often cast as estradiol's opposing force in fear learning, but the human causal evidence remains thin. A 2024 randomized trial by Kaczmarczyk et al. enrolled 116 premenopausal women, tested them during the follicular phase, and assigned them to placebo, estradiol alone, progesterone alone, or estradiol plus progesterone before extinction training.

The design matters because observational menstrual-cycle studies cannot separate progesterone's rise from estradiol's simultaneous changes. Progesterone alone showed no independent extinction-facilitating effect. Earlier neuroimaging work reviewed by the authors found that a single oral progesterone dose increased amygdala reactivity in healthy follicular-phase women without changing mood, which is consistent with active sensitization.

Yet the central combined-hormone result was not fully available for interpretation. Animal work reviewed in the paper found better extinction recall with estradiol plus progesterone than with estradiol alone in ovariectomized rats. That does not answer the human question. The trial was single-blind, used separate groups rather than a crossover design, and could not fully control baseline differences in fear reactivity between groups.

Oral contraceptives may leave a different signal

The most provocative finding concerns oral contraceptive history. In a 2025 behavioral and fMRI study by Davignon et al., 147 participants completed a two-day contextual fear-conditioning and extinction protocol. The key endpoint was return of fear in a context previously learned as safe: a failure of safety-context retrieval rather than a diagnosis of PTSD.

Current oral-contraceptive users had greater fear return in the safe context than comparison participants (p=.015; d=0.64), and past users also showed greater fear return (p=.014; d=0.83). Among current and former users, higher ethinyl estradiol doses were associated with worse outcomes: the high-dose group differed significantly (p=.021; d=0.81), while the low-dose comparison was not significant. Greater safe-context fear return correlated with lower hippocampal and vmPFC BOLD activation across participants.

Most notably, current endogenous estradiol levels and biological sex did not significantly predict this endpoint. Oral-contraceptive history did. That result argues against a simple explanation in which contraceptive effects are merely what low endogenous estradiol looks like. The authors hypothesize persistent changes in hippocampal or vmPFC function, but the study did not demonstrate structural remodeling, epigenetic change, or any other lasting mechanism.

Association is not proof of causation. Participants were not randomized to start contraception, contraceptive history was self-reported, and the study did not define a minimum duration of use. It also ran sequential comparisons involving estradiol status, contraceptive history, and ethinyl estradiol dose in the same sample, which raises the risk of false-positive findings. There was no progestin-only group, so the contribution of progestins remains unknown.

Why the contraceptive evidence still needs caution

A 2022 review by Bartholomew et al. describes several studies in which hormonal-contraceptive users showed weaker extinction recall than naturally cycling women and men. It also discusses a 2015 study by Hwang et al. that found no significant hormonal-contraceptive effect on fear extinction. These studies are observational, so preexisting psychiatric history, health-care use, and other differences between users and non-users could explain part of the association.

Formulation is another blind spot. Combined pills differ in ethinyl estradiol dose and progestin type, including androgenic and anti-androgenic activity. Most of the available evidence does not separate levonorgestrel, drospirenone, norethindrone, and other formulations. A dose association with ethinyl estradiol is a useful lead, not a formulation-specific clinical rule.

What this means for trauma care

The broader epidemiology remains clear: women have a higher PTSD burden, as reviewed by Lehner et al.. Hormones are one plausible contributor among many, alongside trauma type, social conditions, stress biology, and neuropeptides. They do not reduce PTSD to a menstrual-cycle problem.

No published human trial has randomized women with PTSD to receive exposure therapy at different cycle phases. No trial shows that changing a contraceptive prevents PTSD or restores safety learning. People with PTSD should not stop or change contraception on the basis of these studies alone; contraceptive decisions involve pregnancy prevention, bleeding control, migraine risk, clot risk, and individual treatment priorities.

The evidence most strongly supports a research priority, not a self-experiment: prospective studies that measure fear learning and brain function before contraceptive initiation, during use, and after discontinuation, while separating pill formulations.

For clinicians, it is reasonable to ask about contraceptive history when evaluating trauma symptoms and treatment response. For researchers, the next useful test is cycle-timed exposure therapy and formulation-specific contraceptive research. For everyone else, the honest conclusion is narrower: estradiol, progesterone, and synthetic contraceptive hormones may shape safety learning, but the field has not yet shown how to use that knowledge to prevent or treat PTSD.