The Hormone No One Checks: How Antipsychotics Steal Testosterone While Treating Minds

Antipsychotic medications keep millions of people with schizophrenia, bipolar disorder, and psychosis stable. They are essential, often lifesaving drugs. But they carry a hidden endocrine cost that psychiatry has acknowledged in passing and medicine has failed to measure: systematic testosterone suppression in men.

Not subtle. Not rare. A prevalence of 55% low testosterone in men on antipsychotics. Hyperprolactinemia in 18–72% of males across studies, reaching 70–100% with risperidone. And the screening rate for testosterone in psychiatric populations? Zero audits exist. Not low — literally unmeasured.

This is the story of a drug class that created a massive hormonal problem, built drugs that could fix it, and then never checked whether fixing it worked.

Three Roads to Low Testosterone

The textbook explanation — antipsychotics raise prolactin, prolactin suppresses GnRH, testosterone falls — is incomplete. The full damage operates through three parallel mechanisms, any one of which can cause symptomatic hypogonadism on its own.

Pathway 1: Prolactin-Mediated HPG Suppression

This is the dominant mechanism. D2 receptor blockade in the tuberoinfundibular pathway removes dopamine's tonic inhibition of prolactin secretion. Prolactin rises. The downstream cascade is now causally established: elevated prolactin suppresses kisspeptin neurons in the arcuate nucleus — proven by eLife 2024–2025 knockout studies — which reduces GnRH pulsatility, dropping LH and testosterone.

The dose-response data is clear. A comprehensive dose-response meta-analysis of 165 studies (CNS Drugs, 2025) mapped prolactin elevation across 21 antipsychotics. The pattern: most drugs raise prolactin in a dose-dependent manner that plateaus at moderate doses. Risperidone and paliperidone are the worst offenders. Haloperidol, olanzapine, and lurasidone follow. The effect is greater in females, but the testosterone consequence is uniquely male.

Pathway 2: Direct eNOS Impairment

Even when prolactin-mediated testosterone suppression is corrected, erectile dysfunction often persists. Cheng et al. (Toxicology and Applied Pharmacology, 2025) showed that hyperprolactinemia directly impairs the PI3K-Akt-eNOS signaling cascade in erectile tissue — a mechanism independent of testosterone levels. Bromocriptine reversed this impairment independently of testosterone restoration. This means prolactin attacks erectile function through two routes: one hormonal (low T), one vascular (eNOS).

Pathway 3: Dopaminergic Disruption

One study found no association between sexual dysfunction scores and either prolactin or gonadal hormone levels in antipsychotic-treated men with schizophrenia. The implication: some sexual dysfunction is purely neurological. D2 blockade in mesolimbic reward circuits directly suppresses sexual desire and arousal. Additional receptor effects — alpha-adrenergic antagonism causing erectile failure, serotonin 5-HT2 effects disrupting orgasm, anticholinergic blockade reducing vasodilation — compound the damage through mechanisms that neither prolactin correction nor testosterone replacement can fully address.

The Prolactin-Sparing Spectrum

Not all antipsychotics are equal. The CNS Drugs 2025 dose-response meta-analysis provides, for the first time, a clear hierarchy.

The aripiprazole finding is particularly striking. A 2025 study (J Psychiatric Research, n=128) showed an inverse dose-dependent effect: higher aripiprazole doses produced lower prolactin and less sexual dysfunction. Its partial D2 agonism at the tuberoinfundibular pathway actively suppresses prolactin rather than merely not elevating it. But this mechanism cuts both ways — case reports of hypersexuality and impulse control disorders (gambling, compulsive shopping, binge eating) mean the dose must be carefully titrated.

The head-to-head comparison patients actually need — aripiprazole vs. cariprazine vs. brexpiprazole on male testosterone and sexual function — does not exist. This is a confirmed evidence void, not a search failure. No study has compared the three partial agonists on the endpoint that matters most to men.

The Monitoring Failure

If the medical system recognized antipsychotic-induced hypogonadism as the crisis it is, we would see routine screening. Instead, the audit data reveals systematic neglect at every level.

Read that last number again. Zero. Not one published study has audited whether psychiatrists check testosterone in men starting antipsychotics. The North Kirklees audit identified the primary barrier: a communication gap between psychiatrists and GPs — "no clear instructions mentioned from Psychiatrist." Psychiatrists prescribe the drugs. GPs manage the metabolic consequences. Neither checks the hormones.

This is particularly damning given what we know about illness-related baseline dysregulation. A 2025 IJNP study of 189 drug-naive first-episode schizophrenia patients found that 32.3% of males already had abnormal prolactin before any medication — compared to 8.6% of females. Schizophrenia itself disrupts the HPG axis. Antipsychotics compound an existing vulnerability. Without baseline measurements, we cannot distinguish drug effect from disease effect, and we cannot track deterioration.

The Bone Crisis

Hypogonadism doesn't just cause sexual dysfunction. It destroys bone.

The BJPsych study found bone loss consistently correlated with medication dose, and in men specifically, inversely correlated with testosterone values. A 2024 comprehensive review (Diagnostics) confirmed the pattern across 15 studies covering 3,245 patients with mean antipsychotic treatment duration of 15.8 years.

But here is the twist: clozapine appears protective. Studies found that patients on clozapine combined with prolactin-raising antipsychotics had significantly higher BMD T-scores and higher testosterone than those on prolactin-raising antipsychotics alone. Clozapine's limbic-specific D2 binding means less tuberoinfundibular prolactin release — hormonal collateral damage is reduced because the drug targets the right circuits.

The Non-Adherence Trap

Sexual dysfunction is the side effect patients care about most and clinicians ask about least.

The numbers tell the story: sexual dysfunction affects 30–80% of men on antipsychotics. Only 3% of psychiatrists routinely evaluate sexual functioning using specific psychometric tests. Drug side effects carry the highest odds ratio for medication non-adherence in schizophrenia — OR 1.96 (95% CI: 1.48–2.59) per a 2023 meta-analysis — and sexual dysfunction is the most bothersome of all side effects.

The cycle is vicious. A man stops taking his medication because it destroyed his sex life. He relapses. He gets hospitalized. He restarts — often at a higher dose or on a depot injection that cannot be easily discontinued. The FAERS pharmacovigilance analysis (Andrology, 2025) revealed a paradox: aripiprazole had the most sexual dysfunction reports (33.7%), likely because its partial D2 agonism can trigger hypersexuality and impulse control disorders in some patients — a different kind of sexual side effect, but one that drives reports.

The Fertility Blindspot

The fertility impact is perhaps the most neglected dimension. Men with schizophrenia are often of reproductive age when treatment begins — 80% of cases onset during the reproductive period. Yet the evidence base for antipsychotic effects on male fertility is almost entirely animal-derived.

What the animal data shows is unambiguous: all antipsychotics impaired sperm parameters in a 2023 comparative study. Haloperidol and risperidone significantly decreased testosterone. All reduced inhibin B. A 2021 olanzapine study in rats showed decreased normal sperm morphology, reduced serum LH, FSH, and testosterone. A 2025 ESHRE abstract found olanzapine combined with a high-sugar diet reduced total sperm motility.

In humans? One of the few direct studies — 20 patients, of whom 10 were on antipsychotics — found significantly reduced sperm concentration and progressive motility. The FAERS analysis (Tan et al., Andrology 2025) identified antipsychotics as high-risk drugs for semen quality reduction. A combined FAERS+EudraVigilance analysis (Gong et al., 2026) flagged 19 high-risk drugs for male infertility, including antipsychotics.

But no large-scale, well-controlled human study of antipsychotic effects on semen quality exists. The field is making treatment decisions about reproductive-age men based on rat models and pharmacovigilance signals.

The Evidence Voids

What makes this story different from a gap-in-the-evidence complaint is the specificity of what's missing. These aren't data I failed to find. They are data that don't exist — confirmed by exhaustive searches across PubMed, ClinicalTrials.gov, FAERS, EudraVigilance, and Cochrane.

Two systematic review protocols — Schneider-Thoma et al. (F1000Research, 2025) and Herder et al. (PLOS ONE, 2025) — are building network meta-analyses to rank all antipsychotics by sexual dysfunction and prolactin. When published, they will provide the definitive comparison the field lacks. But even those protocols focus on sexual dysfunction as the outcome — not testosterone levels, not bone density, not fertility.

The Illness Complication

A crucial nuance: schizophrenia itself disrupts the HPG axis. The 2025 IJNP study of drug-naive patients found that males had higher prevalence of abnormal prolactin than females (32.3% vs 8.6%, p<.001). In the elevated-prolactin group, prolactin was positively correlated with both progesterone and testosterone — the opposite of the expected suppression pattern. In the normal-prolactin group, the correlation patterns were conventional.

This means two things. First, some HPG dysregulation precedes medication. Second, different regulatory mechanisms operate at different prolactin levels — this is not a simple linear suppression story. Without baseline hormonal measurements, clinicians cannot distinguish drug-induced damage from illness-related disturbance, and they cannot track whether treatment changes are helping or harming.

Management: A Hierarchy

When a man on antipsychotics presents with low testosterone, sexual dysfunction, or bone loss, the management follows a clear hierarchy — though the evidence base is thinner than it should be.

A critical caveat from the Sangngarm 2025 prospective cohort: switching or adding aripiprazole produced dramatic prolactin reductions (~68 ng/mL), but sexual dysfunction was not fully resolved by prolactin normalization alone. This confirms the triple-mechanism model — pathways 2 (eNOS) and 3 (dopaminergic) persist even when pathway 1 is corrected.

What Needs to Change

The medical system created this problem. It has the drugs to fix it. But it has never built the infrastructure to detect it, measure it, or track whether interventions work. Here is the minimum that psychiatric care owes men on antipsychotics:

None of this is practiced. None of it is guideline-mandated. Eight of 12 international guidelines recommend baseline prolactin, but they diverge on monitoring and management. Zero recommend testosterone screening. The North Kirklees audit identified the reason: no clear instructions travel from psychiatrist to GP. The monitoring falls into a gap between specialties.

The psychiatric community built prolactin-sparing antipsychotics — aripiprazole, brexpiprazole, cariprazine, lumateperone — that could largely prevent this crisis. The irony is that they exist and are underutilized, while the hormone they were designed to protect goes unmeasured. The 5.3% baseline screening rate in Riyadh is not a local failure. It is a global one.

The strongest version of this story is not that antipsychotics damage testosterone. It is that the medical system knows they damage testosterone, has drugs that don't, and still doesn't check.