Open questions & controversies

In DWP-001, Powers proposed PFS stems from deficient allopregnanolone: 5-alpha-reductase blockade plus decreased-function AKR1C variants leaves patients unable to synthesize this calming neurosteroid, producing a state resembling severe postpartum depression (which responds to brexanolone, a synthetic allopregnanolone). Melcangi's CSF studies (PFS-004, PFS-005) found decreased tetrahydroprogesterone, isopregnanolone, progesterone, and DHT in PFS patients versus controls — direct biochemical evidence of depleted downstream neurosteroids. The PSSD literature converges here too: Xie et al. (PSSD-003) identify allopregnanolone depletion via 5α-reductase inhibition as a PSSD mechanism. On this view, supplementation (progesterone, allopregnanolone analogues) is the rational intervention.

In DWP-002, Powers revised his own model: the anhedonic/low-libido PFS/PSSD subtype is driven by pathological excess of neurosteroids such as THDOC and androsterone that over-activate GABA-A receptors, causing neural network remodeling analogous to chronic benzodiazepine exposure. He implicated upregulated 3α-HSD and broken glucuronidation/excretion (UGT enzymes, gut beta-glucuronidase), noting >50% of his PFS patients showed near-zero urinary androgens on DUTCH testing. On this view, lowering neurosteroid load (his calcium D-glucarate + indomethacin protocol) is rational — and progesterone supplementation could theoretically worsen the anhedonic subtype. Flag: this is a single clinician's observational model from Reddit, not peer-reviewed; the "excess" molecules have not been directly measured in CSF (Powers himself requested mass-spec collaboration to identify them).

Traish (PFS-001), reviewing 250+ articles, concludes PFS is a real constellation of persistent effects occurring regardless of age, dose, or study length, and explicitly notes inadequate adverse-event reporting in most finasteride trials — i.e., the evidence base was systematically blind to the signal. Irwig's case series (PFS-009) and follow-up (PFS-010) show 96% of men still symptomatic ~14 months after stopping, with duration of use not predicting recovery — a pattern hard to square with simple nocebo. Csoka et al. (PSSD-004) reported persistent dysfunction lasting years in previously healthy individuals. Reisman (PSSD-007) addresses the differential-diagnosis challenge directly, arguing the syndrome is distinguishable from the underlying psychiatric condition.

Skeptics (not extensively represented in this corpus — a known gap) argue: (1) sexual dysfunction is common in the general population and in depression itself, so post-drug attribution may reflect the underlying condition; (2) internet-recruited case series suffer ascertainment bias toward severe cases; (3) nocebo effects are well documented with finasteride specifically. Ben-Sheetrit et al. (PSSD-006) found dose did not predict severity, which susceptibility-model advocates read as supporting genetics — but skeptics could equally read as inconsistent with pharmacologic causality. The corpus's honest limitation: nearly all human evidence is retrospective or case-series; no prospective blinded cohort with pre-drug baselines exists.

Traish (PFS-002) proposed finasteride acts as an endocrine disruptor causing drug-induced epigenetic changes — DNA methylation, histone modification, androgen-receptor upregulation — explaining why symptoms outlast the drug. Csoka & Szyf (PSSD-005) generalized this to a "pharmacoepigenetics" field. Supporting data points: Cauci et al. (PFS-007) linked AR (CAG)n/(GGN)n polymorphisms to symptom patterns in 66 PFS patients; the Baylor gene-expression report (PFS-008, conference-level, peer review unconfirmed) found 1,446 genes over- and 2,318 under-expressed in PFS penile skin with AR overexpression; Diviccaro et al. (PSSD-012) found persistent transcriptomic changes in rat nucleus accumbens after paroxetine withdrawal; Rege's video cites increased DNA methylation of a gene promoter in CSF.

Flag: no published study in the corpus demonstrates, in living humans, a specific lasting methylation or histone change caused by finasteride/SSRIs that (a) was absent before drug exposure, (b) persists after discontinuation, and (c) correlates with symptoms versus recovery. The human data are cross-sectional (patients vs controls, no pre-drug baseline), small-n, or conference-level. Animal transcriptomic persistence (PSSD-012) is suggestive but not human epigenetic proof. The AR polymorphism association (PFS-007) shows genetic susceptibility, not drug-induced epigenetic change — these are different claims often blurred together.

Healy et al. (CROSS-001): 300 cases across 37 countries showed genital numbness, pleasureless orgasm, libido loss, and impotence across antidepressants, 5α-reductase inhibitors, and isotretinoin — three pharmacologically unrelated drug classes producing the same syndrome, implying a common downstream pathway. Giatti et al. (PFS-003) argue PFS and PSSD share neuroactive-steroid, neurotransmitter, and gut-microbiota mechanisms, possibly converging on enzymes like PNMT. Xie et al. (PSSD-003) explicitly connect SSRIs to the 5α-reductase/neurosteroid pathway — a direct pharmacological bridge. Healy's 2022 consensus criteria (CROSS-002) formalize the syndromes as parallel entities with shared diagnostic logic.

Flag: phenomenological similarity does not prove mechanistic identity. Finasteride blocks 5α-reductase directly; SSRIs act primarily on serotonin transporters (with 5α-reductase effects shown only preclinically, PSSD-009); isotretinoin is a retinoid with no established link to either pathway. The objective biomarker evidence is asymmetric: CSF neurosteroid disruption is documented in PFS (PFS-004, PFS-005) but, in this corpus, not equivalently in PSSD or post-retinoid patients. Pudendal neuropathy evidence exists for PFS (DISC-013, a minority of patients); in PSSD the same tests were mostly normal (CUR-003; see OQ-010). Grouping may reflect a shared final common pathway (e.g., GABAergic remodeling) while upstream injuries differ — or it may be a useful-but-loose clinical umbrella.

In DWP-002, Powers reports that temporary androgen deprivation ("castration trials") reliably restored androgenic signaling in the androgen-signal-loss PFS subtype (though not the anhedonic subtype) — consistent with his model of intracellular weak-metabolite crowding ("musical chairs") that clears when androgen input pauses. Patient reports in the corpus (community discussions) describe a ~2-week full-remission "window" on testosterone reintroduction followed by crash — phenomenology consistent with a clearance dynamic. As a probe (not treatment), it generates a testable prediction: if signaling reboots after deprivation, the defect is dynamic/clearable rather than structural.

Flag: this is one clinician's uncontrolled observational claim from Reddit. Chemical castration carries known risks (bone loss, mood/cognitive effects, cardiovascular and metabolic consequences) and there are no controlled data, no pre-registered endpoints, no independent replication, and no ethical-oversight framework described. Mainstream endocrinology would require extraordinary justification for inducing hypogonadism as an experiment. Selection and reporting biases in a self-selected patient population could easily produce apparent "responses."

The deficiency model gives a clean rationale: if PFS involves depleted allopregnanolone downstream of 5AR blockade (Melcangi CSF data, PFS-004/PFS-005), replacing the missing neurosteroid should help — and postpartum depression, with a similar symptom profile, responds to brexanolone (synthetic allopregnanolone). Powers (DWP-001) reported two PFS-like patients improving on rectal progesterone, while explicitly flagging N=2 as weak evidence. The 2019 Powers lecture background and the "Rethinking Reductase" presentation both describe progesterone→allopregnanolone as the rate-limiting calming pathway. Rege's video lists neuroactive-steroid therapy among investigational avenues.

No RCTs exist. Under Powers' own revised 2026 model (DWP-002), the anhedonic subtype suffers neurosteroid excess — supplementing progesterone/allopregnanolone in those patients could theoretically worsen the condition, which is why he pivoted to lowering neurosteroid load instead. Rege's video explicitly caveats all five treatment avenues as "none conclusively established." Without subtype stratification, trials could wash out real subgroup effects in both directions.

Healy & Mangin (PSSD-013) show PSSD cannot be counted with current methods: inconsistent coding scatters cases across health records under vague labels, blocking the large database studies needed for estimates — and there is no dedicated ICD code for PFS or PSSD in either ICD-10 or ICD-11 (per the corpus ICD mappings). Xie et al. (PSSD-003) report persistent symptoms in up to 74% of affected cohorts beyond 6 months — a conditional figure, not population prevalence, but indicating that among those affected, persistence is the rule. Clinical-recognition gaps (PSSD-008) mean cases never enter records at all.

The drugs are taken by tens of millions; if persistent syndromes were common, skeptics argue, the signal would be unmistakable even with poor coding. Irwig's series (PFS-009, 71 men) and Ben-Sheetrit's survey (PSSD-006, 183 possible cases from 532 surveyed) are drawn from self-selected populations, inflating apparent frequency. Flag: both sides are arguing from absence of data — the methods literature (PSSD-013) is the most honest position: prevalence is currently unquantifiable, and any confident number in either direction outruns the evidence.

Healy et al. (CROSS-002) produced multidisciplinary consensus criteria for PSSD, PGAD, PFS, and post-retinoid sexual dysfunction: prior drug exposure, enduring dysfunction ≥3 months post-discontinuation, exclusion of other causes. This gives researchers a shared vocabulary, enables consistent corpus tagging, and provides the differential-diagnosis framework Reisman (PSSD-007) calls for. Operational criteria routinely precede official nosology in medicine's history.

Neither DSM-5-TR nor ICD-10/ICD-11 contains PFS or PSSD as entities (corpus ICD mappings confirm; only proxies exist). Consensus among a self-selected expert panel is not the same as field-tested validity: inter-rater reliability, sensitivity/specificity against mimics (notably depression-related dysfunction), and predictive validity are all unmeasured. PSSD-013's coding analysis shows the criteria have not penetrated routine clinical documentation.

Giatti et al. (PFS-003) list gut microbiota among shared PFS/PSSD mechanisms. Diviccaro et al. (PSSD-011) showed paroxetine altered colonic steroid synthesis and microbiota in rats — a concrete gut-steroid-axis finding. Powers' 2026 model (DWP-002) implicates gut beta-glucuronidase in disrupting androgen glucuronidation/excretion, tying the microbiome directly to his hormonal mechanism. Rege's video lists microbiome modulation among investigational avenues. The gut produces and metabolizes steroids, so a gut-mediated mechanism is biologically plausible.

Flag: all human evidence in the corpus is associative. Microbiome shifts accompany countless conditions (depression, stress, diet change, the drugs themselves) without being causal. The preclinical paroxetine-gut finding (PSSD-011) is in rats, and Powers' beta-glucuronidase link is an observational inference from DUTCH patterns, not a demonstrated causal chain. No human study shows that modifying the microbiome changes PFS/PSSD symptoms.

Genital anesthesia/numbness is a hallmark of both syndromes: reported across all three drug classes in Healy's 300 cases (CROSS-001), in early PSSD case reports (PSSD-004), and in PFS cohorts. Melcangi et al. (DISC-013, J Steroid Biochem Mol Biol 2017) found abnormal pudendal somatosensory evoked potentials in 4 of 16 PFS patients, those with severe erectile dysfunction — the first objective evidence of peripheral nerve involvement. Giatti et al. (PFS-003) frame the syndromes as mechanistically linked. Corneal confocal microscopy in small mixed post-drug series suggests small-fibre involvement across drug classes (LIT-040, SIDE-002).

Flag: pudendal SEPs were abnormal in only 4 of 16 PFS patients (DISC-013), and the same tests were normal in 8 of 9 PSSD patients at Queen Square (CUR-003), whose authors propose a central disturbance of erogenous sensation. Genital anesthesia could arise peripherally (pudendal neuropathy), spinally, or centrally (altered sensory processing from neurosteroid/GABAergic disruption) — Powers' 2026 model emphasizes central GABAergic remodeling, which could produce numbness without any peripheral lesion. Similar symptoms do not entail identical lesions.