Raloxifene

Limits on what this shows

• Null/negative findings: The liver study's own oxygen-consumption data show only a modest 14–17% decrease despite a much larger (74–78%) drop in ketone-body output, and effects on long-chain fatty acid oxidation appeared only at the highest concentrations tested, arguing against a blanket mitochondrial toxicity claim.
• At-risk subgroups: None established in humans; the disruptive-direction data used concentrations (25 µM) far above raloxifene's typical free plasma concentration (bioavailability 2%, 95% protein-bound), so clinical relevance at the approved 60 mg/day dose is unverified.
• Counter-evidence: The vascular study and the liver study used the same drug in different tissues with opposite conclusions; neither has been replicated in humans or cross-validated against the other model.

Regulatory status

Not addressed in the FDA label; no mitochondrial toxicity signal listed.

Population evidence

None in humans. Two conflicting animal/ex vivo studies: (1) cultured rat aortic vascular smooth muscle cells and isolated rat aortic rings — raloxifene prevented a phenylephrine-induced rise in mitochondrial membrane potential, reduced mitochondrial reactive oxygen species (via fluorescent probes and JC-1 staining), and preserved uncoupling protein-2 (UCP2) expression, producing vasodilation independent of estrogen receptor or cyclooxygenase pathways. (2) Perfused livers, isolated mitochondria, and peroxisomes from ovariectomized rats — at 25 µM, raloxifene inhibited mitochondrial beta-oxidation of medium-chain fatty acids (ID50 11.24 µM for octanoyl-CoA) and reduced oxygen consumption 14–17%.

Mechanism

Protective direction: mild mitochondrial uncoupling via UCP2, lowering ROS production. Disruptive direction: raloxifene's phenolic structure is proposed to undergo peroxidase-catalyzed bioactivation to a phenoxyl radical, co-oxidizing NADH and perturbing mitochondrial redox balance independent of direct respiratory-chain inhibition.

Counter-evidence held on record

Do not present either direction as established in humans. This is animal/ex vivo mechanistic evidence only, with directly conflicting results depending on tissue and concentration; see the Overlap map for a shared source with the Oxidative stress and ATP production entries below.

Limits on what this shows

• Null/negative findings: The brain study found no change in the antioxidant enzymes SOD or catalase despite the MDA reduction, and no improvement in a concurrent learning-behavior task — the antioxidant signal did not translate to a measured functional outcome in that model.
• At-risk subgroups: None established in humans.
• Counter-evidence: Three additional titles bearing on this category were identified (raloxifene protecting endothelial cells against oxidative stress; an antioxidant effect on heart/liver/brain cortex of ovariectomized rats; raloxifene reported as a better LDL antioxidant than estradiol or tamoxifen in postmenopausal women in vitro) but could not be opened during this review — PubMed and Wiley returned bot-verification pages and the Menopause-journal LDL study sat behind a paywall. They are listed in Sources as inaccessible rather than used as supporting evidence.

Regulatory status

Not addressed in the FDA label.

Population evidence

Ovariectomized female Sprague-Dawley rats (n=24; non-ovariectomized control, ovariectomized vehicle, ovariectomized + raloxifene 1 mg/kg subcutaneously daily for 14 days): raloxifene significantly reduced malondialdehyde (MDA, a lipid-peroxidation marker) across cortex, striatum, hippocampus, and cerebellum (p<0.05), with no significant change in superoxide dismutase or catalase activity. Separately, the same ex vivo perfused-liver study cited under Mitochondrial burden found raloxifene at 0.25–2 µM drove strong NADH co-oxidation via peroxidase-catalyzed reactions with hydrogen peroxide, shifting the beta-hydroxybutyrate/acetoacetate ratio toward a more oxidized cellular redox state.

Mechanism

Antioxidant direction: raloxifene's phenolic hydroxyl groups (structurally related to tamoxifen and vitamin E) are proposed to scavenge lipid radicals directly in brain tissue. Pro-oxidant direction: the same phenolic structure, once oxidized to a phenoxyl radical by peroxidase/H2O2, consumes NADH and can act as a pro-oxidant catalyst rather than a terminal antioxidant.

Counter-evidence held on record

Direction depends on tissue, dose, and assay; do not present raloxifene as a validated antioxidant or pro-oxidant supplement rationale for humans. No human raloxifene-specific oxidative-stress marker data were confirmed during this review.

Limits on what this shows

• Why no grade: Both conditions for ungraded are met: no Tier B or C evidence exists connecting raloxifene to copper status (searched via multiple phrasings across web search and the Crossref works API), and no circulating consumer claim naming raloxifene and copper was found to rebut as white.

Tooling note

Crossref, the channel used for this specific query, returned results but none relevant to raloxifene and copper; this is treated as a genuine gap rather than a tooling failure, since Crossref itself was reachable and responsive. Re-checked in this revision (2026-09-26) with a fresh Crossref bibliographic query ("raloxifene copper ceruloplasmin"): still no relevant results, only unrelated copper/ceruloplasmin biochemistry papers. WebSearch was unavailable (session budget exhausted) and Europe PMC/OpenAlex/Semantic Scholar were rate-limited or proxy-blocked for this query in this session; the gap is confirmed via Crossref alone this round and should be re-checked against those other channels when available.

Limits on what this shows

• Why no grade: Both conditions for ungraded are met: searches across web search and Crossref (multiple phrasings, including "raloxifene iron ferritin" and "raloxifene hemoglobin hematocrit erythropoiesis") returned no raloxifene-specific results, and no circulating consumer claim linking raloxifene to iron status was found.

Tooling note

Crossref was reachable and returned unrelated results (general iron-status epidemiology, ferritin biochemistry) rather than errors, supporting a genuine gap rather than a blocked search. Re-checked in this revision (2026-09-26) with a fresh Crossref bibliographic query ("raloxifene iron ferritin hemoglobin"): still no raloxifene-specific results. WebSearch was unavailable (session budget exhausted) and Europe PMC/OpenAlex/Semantic Scholar were rate-limited or proxy-blocked for this query in this session; the gap is confirmed via Crossref alone this round and should be re-checked against those other channels when available.

Limits on what this shows

• Why no grade: A 2025 paper (Uoti et al., Basic & Clinical Pharmacology & Toxicology) studies efflux/uptake transport and "gut microbial reactivation of raloxifene glucuronides" — i.e., enteric bacteria deconjugating raloxifene's glucuronide metabolites to enable enterohepatic recirculation. This is the microbiome affecting the drug's disposition, the reverse of the direction this category grades (the drug's effect on microbiome composition/diversity), so per the same direction-check logic as Hard Rule 4 it cannot support a graded finding here. No study measuring raloxifene's effect on microbial composition or diversity, and no circulating consumer claim about raloxifene and gut flora, was located.

Tooling note

The full text of the Uoti et al. paper itself was blocked by a bot-verification page on PMC across repeated attempts; only its title and abstract-level description (via search result snippets) could be used, and are disclosed as such rather than cited as verified supporting evidence. Re-checked in this revision (2026-09-26) with a fresh Crossref bibliographic query ("raloxifene gut microbiome microbiota"): still no study of raloxifene's effect on microbiome composition or diversity, only general microbiome-biology results unrelated to raloxifene. WebSearch was unavailable (session budget exhausted) and Europe PMC/OpenAlex/Semantic Scholar were rate-limited or proxy-blocked for this query in this session.

Limits on what this shows

• Null/negative findings: Long-chain fatty acid oxidation was inhibited only at the highest concentrations tested, and the oxygen-consumption decrease (14–17%) was proportionally much smaller than the ketogenesis decrease (74–78%), suggesting a shift in fuel handling rather than a wholesale collapse of oxidative capacity.
• At-risk subgroups: None established in humans; 25 µM is far above raloxifene's free plasma concentration at the approved 60 mg/day dose given 2% absolute bioavailability and 95% protein binding.
• Counter-evidence: No human study of raloxifene and ATP production, mitochondrial energy charge, or exercise/fatigue outcomes was located in this review; the well-known "low energy"/fatigue framing used in some consumer material has no raloxifene-specific human source behind it.

Regulatory status

Not addressed in the FDA label; no fatigue/energy-metabolism warning listed beyond general adverse-event reporting.

Population evidence

Perfused livers, isolated mitochondria, and peroxisomes from ovariectomized rats 3 weeks post-surgery versus metestrus-phase controls. At 25 µM raloxifene: total ketone-body production (acetoacetate + beta-hydroxybutyrate) fell approximately 74–78% from endogenous substrate and 26–43% from exogenous fatty acids; oxygen consumption fell 14–17%; CO2 production rose without a parallel rise in oxygen consumption, indicating metabolic diversion away from oxidative phosphorylation; mitochondrial beta-oxidation of medium-chain fatty acids was dose-dependently inhibited (ID50 11.24 µM for octanoyl-CoA-derived substrate).

Mechanism

Inhibition of mitochondrial and peroxisomal fatty acid beta-oxidation, the pathway that feeds acetyl-CoA into the TCA cycle and oxidative phosphorylation for ATP synthesis, proposed to run through raloxifene's peroxidase-mediated bioactivation and NADH/NAD+ disruption rather than direct respiratory-chain blockade.

Counter-evidence held on record

This is a single ex vivo animal-liver source at a supraphysiologic concentration; it does not establish that raloxifene lowers ATP production or causes fatigue in patients at approved doses. See the Overlap map — this is the same source cited under Mitochondrial burden and Oxidative stress above.

Limits on what this shows

• Null/negative findings: Hematocrit, the parameter most directly tied to oxygen-carrying capacity, showed only small, inconsistent within-normal-range changes, i.e., no meaningful raloxifene effect on red-cell mass was demonstrated. The study's own conclusion was that raloxifene treatment is "not associated with adverse changes in hemorheological factors."
• At-risk subgroups: None established; this is a general postmenopausal osteoporotic population, and the effect (a decrease in viscosity) runs in a favorable rather than a risk-conferring direction.
• Counter-evidence: Single trial, not independently replicated in this review. It also sits in tension with raloxifene's boxed warning for increased venous thromboembolism risk (highest in the first 4 months) and the RUTH trial's stroke-death signal, since lower blood/plasma viscosity would not on its own predict higher clotting or stroke-death risk. The two findings are not necessarily contradictory (viscosity is only one determinant of thrombotic risk, and the boxed warnings rest on separate procoagulant-factor data), but this favorable viscosity signal must not be read as evidence against those warnings.

Regulatory status

Not addressed in the FDA label; no hemorheological, hematocrit, or oxygen-delivery warning is listed for raloxifene beyond general adverse-event reporting.

Population evidence

"The hemorheological effects of raloxifene in postmenopausal women with osteoporosis. Results of a 3-year placebo-controlled clinical trial" (PMID 12122230, Clinical Hemorheology and Microcirculation, 2002), opened via Europe PMC's REST API in this revision. Fifty-seven postmenopausal osteoporotic women were randomized to placebo or raloxifene (60 or 120 mg/day) for three years, with blood viscosity, plasma viscosity, hematocrit, and fibrinogen measured at intervals. Mean values for all hemorheological indices showed small, inconsistent changes that stayed within the normal reference range in both arms. Raloxifene-treated subjects showed a small but statistically significant decrease over time in high-shear-rate blood viscosity and in plasma viscosity relative to placebo. No thromboembolic events occurred in either arm during the study.

Mechanism

Not established by the trial itself; no specific molecular mechanism was proposed. A plausible pathway is raloxifene's estrogen-receptor-mediated effect on plasma fibrinogen and lipid-related viscosity determinants, consistent with estrogen's known hemorheological effects, but this was not tested directly.

Counter-evidence held on record

This is a single Tier A trial showing a modest, favorable rheological effect and no adverse change in hematocrit; it does not establish a clinically meaningful effect on tissue oxygen delivery in either direction, and it must not be read as contradicting raloxifene's boxed thromboembolism and stroke-death warnings, which rest on separate coagulation-factor data.