Most antidepressants do not lower testosterone in any direct, predictable way; the strongest evidence concerns a modest, reversible rise in prolactin with some SSRIs and a separate, better-documented effect of antidepressants on sexual function. Whether your testosterone actually falls depends on the drug, the dose, duration, baseline health, and age.
Search snippets can answer "yes" or "no." The real answer lives in the space between, and that space is where thin explainers collapse. This article separates three things that get blurred together: measured hormone levels, perceived libido, and downstream fertility. It explains how each antidepressant class touches the HPG axis, where the signals are strong versus weak, and what you can actually do with the information.
The HPG axis and why testosterone is regulated
Testosterone production is a feedback loop, not a faucet. The hypothalamus releases gonadotropin-releasing hormone (GnRH) in pulses. That drives the pituitary to secrete luteinizing hormone (LH) and follicle-stimulating hormone (FSH). LH tells Leydig cells in the testes to make testosterone; FSH supports sperm production.
The loop is closed by negative feedback. Rising testosterone and its metabolite estradiol suppress hypothalamic and pituitary signaling, pulling GnRH and LH back down. Anything that raises prolactin strongly, disrupts dopamine tone, or stresses the pituitary can interrupt this loop at the top. Anything that damages Leydig cells or the testes interrupts it at the bottom.
This matters because antidepressants are not one pharmacological family. SSRIs, SNRIs, bupropion, mirtazapine, tricyclics, and monoamine oxidase inhibitors act on different receptors and transporters. Their effects on prolactin, dopamine, and gonadal function therefore diverge. A class-wide statement about testosterone is unsupported.
SSRI mechanisms: prolactin and dopamine tone
Selective serotonin reuptake inhibitors are the most studied class for endocrine side effects. Two mechanisms are plausible, and they point in opposite directions on libido but generally the same direction on prolactin.
First, serotonin stimulates prolactin secretion. Serotonergic drugs reliably raise serum prolactin modestly in many patients. Prolactin suppresses GnRH at the hypothalamus, which can lower LH pulsatility and, over time, testosterone. This is the classic mechanism by which hyperprolactinemia causes hypogonadism.
Second, SSRIs can reduce dopaminergic tone downstream, and dopamine is a brake on prolactin. Less dopamine means less restraint on prolactin output. This reinforces the first mechanism.
The catch is magnitude. For most people on standard SSRI doses, prolactin rises sit within or near the normal reference range rather than producing overt hyperprolactinemia. Overt prolactin-mediated hypogonadism is uncommon but real, and it is the exception that deserves attention. In plain terms: the pathway exists, but most users do not travel far down it.
Quantified context: how large are the effects?
Precise hormone shifts vary widely across studies, and the literature is heterogeneous enough that quoting a single percentage would be false precision. What can be stated qualitatively and defensibly:
Prolactin elevations with SSRIs are typically modest, often within the normal range, with a minority of patients crossing into abnormal territory. Testosterone changes, where reported, are usually small and not consistently statistically significant across cohorts. Sexual side effects are far more common and more reliably measured than hormone changes.
This gap is the counterintuitive core of the topic. Study after study finds sexual dysfunction prevalence far exceeding the prevalence of abnormal testosterone. That tells you the symptoms patients report are not primarily driven by measurable testosterone suppression. They are driven by local and central serotonin effects on arousal, orgasm, and desire.
Another professional caveat: many testosterone studies in this area are small, short, and cross-sectional. They cannot establish whether the drug caused a hormone change or whether depression itself, sleep disruption, weight change, or alcohol use did. Depression independently associates with lower testosterone. Separating drug effect from disease effect is genuinely difficult, and honest reviews say so.
A comparison of antidepressant classes
The table below summarizes qualitative mechanisms and the direction of plausible effects. It is not a dosing reference and does not substitute for clinical guidance.
| Class or agent | Prolactin effect | Plausible testosterone effect | Libido/sexual signal |
|---|---|---|---|
| SSRIs (e.g., sertraline, fluoxetine) | Modest rise common | Small, inconsistent lowering via prolactin | Sexual side effects well documented |
| SNRIs (e.g., venlafaxine, duloxetine) | Modest rise possible | Small, inconsistent | Sexual side effects reported, often less than SSRIs |
| Bupropion | Minimal effect | No consistent lowering; sometimes raised libido | Often used to counter SSRI sexual effects |
| Mirtazapine | Generally little prolactin rise | No consistent lowering | Sexual effects less prominent |
| Tricyclics | Variable, some raise prolactin | Inconsistent | Anticholinergic and sedative effects dominate |
| MAOIs | Variable | Inconsistent | Sexual effects reported |
Two rows deserve emphasis. Bupropion works through norepinephrine and dopamine reuptake inhibition and does not carry the serotonergic prolactin signal, which is why clinicians sometimes add or switch to it when sexual function is the concern. Mirtazapine's noradrenergic and specific serotonergic profile behaves differently from SSRIs on prolactin. Class matters.
A timeline of when effects appear
Timing helps distinguish transient adjustment from persistent change. The table is qualitative because individual trajectories vary.
| Period | What tends to happen | What it means |
|---|---|---|
| Week 0-2 | Serotonin and prolactin shifts begin; mood not yet improved | Hormone changes can precede symptom benefit |
| Week 2-6 | Sexual side effects often emerge; mood response builds | Most libido complaints declare themselves here |
| Month 2-3 | Prolactin, if elevated, tends to stabilize | Persistent elevation is worth measuring |
| Month 3-6 | Hormone changes, if any, usually plateau | A stable picture emerges |
| On discontinuation | Prolactin and sexual function typically recover | Reversibility is the rule, not the exception |
| Post-discontinuation | Recovery may lag weeks to months | Timeline is variable and individually reported |
The practical takeaway: a hormone panel drawn five days into treatment tells you little. Waiting until the medication has reached steady state gives a more interpretable result. That is typically several weeks, though pharmacokinetics differ by drug.
Red flags and edge cases
Most people on antidepressants do not develop clinically meaningful testosterone suppression. A smaller group does, and the warning signs deserve attention. Seek evaluation rather than self-adjusting medication.
New or worsening loss of morning erections, or persistent low libido beyond the first weeks Enlarged or tender breast tissue, or nipple discharge (suggests hyperprolactinemia) New fatigue, hot flashes, or reduced body hair Infertility concerns or changes in semen volume Symptoms in someone with known pituitary or testicular conditions Any combination of the above with headache or visual field changes (pituitary warning)
The last bullet matters because a prolactin-secreting pituitary adenoma can mimic or amplify drug effects. That is a differential diagnosis, not a drug side effect, and it needs imaging and endocrine input.
Misconceptions that mislead
Several ideas circulate widely and are not well supported.
The first is that "antidepressants tank testosterone." The evidence supports a modest, inconsistent, and often reversible effect in a subset of patients, chiefly through prolactin. Framing it as a uniform crash overstates the data.
The second is that "low libido equals low testosterone." In this population, sexual side effects are largely serotonergic and central, and they occur at rates far higher than abnormal testosterone. Assuming a hormonal cause can lead to unnecessary testosterone therapy and missed dose or drug adjustments.
The third is that "if prolactin is high, stop the drug." Abrupt discontinuation risks relapse and discontinuation syndromes. Dose reduction, switching, or adding a dopamine agonist is a clinical decision, not a self-directed one.
The fourth is that "natural testosterone boosters fix it." Supplements marketed for testosterone have weak, inconsistent evidence and are not a substitute for addressing the actual mechanism.
Practical actions and thresholds
If you and your clinician suspect an endocrine effect, here is a rational sequence.
Do not stop the antidepressant abruptly. Discuss timing and taper with the prescriber. Ask whether a morning total testosterone and a prolactin level are appropriate. Morning sampling matters because testosterone is diurnal. Consider a repeat measurement if the first is borderline, since single values are noisy. If prolactin is elevated, the conversation should turn to dose, drug choice, or endocrine referral rather than self-treatment. Track symptoms and dates, not just how you feel, so the pattern is visible to the clinician. Treat reversible contributors: sleep, alcohol, weight, and thyroid status all influence testosterone.
On thresholds, be careful. Reference ranges differ by laboratory, assay, and population. A "low" result at one lab can be normal at another. This is one reason cross-study percentages conflict. The practical rule is to interpret a value against the lab's own range and against symptoms, not against a number pulled from a headline. There is no universal testosterone cutoff that applies to everyone.
Limitations and what remains unknown
The honest summary is that this is an active research area with real but modest signal and substantial noise. Most available evidence is short-term and observational, and it struggles to disentangle drug effects from depression, lifestyle, and age. Long-term, adequately powered studies with serial hormone measurements are limited.
We can say, with reasonable confidence, that serotonergic antidepressants can raise prolactin modestly and, in a subset of patients, plausibly reduce testosterone through that pathway. We can say the effect is usually small, often reversible on stopping, and far less common than sexual side effects. We cannot say that antidepressants uniformly lower testosterone, quantify a reliable percentage for any given drug, or predict individual response. Anyone who tells you otherwise is outrunning the data.
If you are concerned about your hormones on an antidepressant, the right move is a conversation with the prescriber, a properly timed lab panel, and a plan. Not a supplement, not a cold-turkey stop, and not a statistic from a thin explainer.