Mefenamic acid

Limits on what this shows

• Null findings and direction: Dominguini 2022 [26] (abstract and article page read): in rats with cecal ligation and puncture sepsis, mefenamic acid given intraperitoneally at 10, 30 or 50 mg/kg immediately after induction increased the activity of respiratory complexes I, II, II-III and IV and lowered ROS and 8-oxoG at 30 and 50 mg/kg. No sham animals received mefenamic acid, so no healthy-animal mitochondrial effect was read. The authors note mefenamic acid is not a specific NLRP3 inhibitor and that a single dose was used.
• At-risk subgroups: None established. Renal and hepatic impairment are label warnings for other outcomes, not shown here to modify mitochondrial effects.

Regulatory status

No mitochondrial statement in the label text read. The label lists hepatotoxicity and renal toxicity as outcomes (sec 5.3, 5.6) [1].

Population evidence

All Tier B (mefenamic acid in every cohort), none human. Uyemura 1997 [21]: the title names diclofenac sodium and mefenamic acid as potent inducers of the membrane permeability transition in renal cortex mitochondria; per a PubMed listing summary, mefenamic acid induced it at 2 micromolar in isolated rat renal cortex mitochondria, about 50-fold more potent than salicylate (abstract not read; OpenAlex held none). Mingatto 1996 [22]: rat kidney mitochondria, respiration and ATP synthesis; per the listing summary, mefenamic acid both uncoupled and inhibited oxidative phosphorylation (title opened through OpenAlex; abstract not read). Tatematsu 2016 [23] (abstract read through OpenAlex): in rat liver mitochondria, mefenamic acid at 200 micromolar induced phosphate-dependent swelling suppressed by cyclosporin A at 2.5 micromolar; swelling was also seen with flufenamic, meclofenamic and tolfenamic acid and less with diclofenac; the NSAIDs tested dose-dependently obstructed respiratory chain electron transport. Jarrar 2017 [24] (abstract via OpenAlex): female rats given free or liposomal mefenamic acid orally every day for 28 days showed sinusoidal dilation, Kupffer cell hyperplasia and mitochondrial damage on liver electron microscopy (the PubMed listing summary gave 80 mg/kg; not confirmed in the abstract). Jarrar 2020 [25] (abstract via OpenAlex): 28-day rat kidney electron microscopy showed mitochondrial injury, nuclear and lysosomal alterations, tubular cell steatosis and apoptosis, worse with free drug than with one liposomal form. Terada 1974 [34] (title only, not opened) concerns uncoupling structure-activity relations of fenamic acids.

Mechanism

Present at the isolated-organelle level: cyclosporin A-sensitive permeability transition, uncoupling and electron-transport inhibition [21][22][23]. Whether any of this occurs at human tissue exposure is not shown.

Concentration context

The label gives mean peak plasma levels of 10 to 20 mcg/mL after a single 1 g dose and more than 90% albumin binding [1]. By simple arithmetic with a molecular weight near 241 (not retrieved from a source in this review), that is roughly 41 to 83 micromolar total drug, so 200 micromolar [23] is above the total peak and the free fraction is lower still, while the 2 micromolar effect [21] is within the total range. Isolated-mitochondria buffers differ from plasma, so these comparisons are rough.

Hard Rule 2 and 6 checks

Mitochondrial injury in rats is a drug toxicity finding, not a nutrient deficiency, and no mitochondrial cofactor (coenzyme Q10, carnitine, B vitamins) was measured. The sepsis result is disease-model confounded.

Counter-evidence held on record

Counter-evidence and confounding Direction is mixed: injury markers in healthy rats and isolated organelles [21][22][23][24][25] against preserved respiratory chain activity in septic rats [26]. Several effective concentrations are at or above clinical total plasma levels [23]. No human biopsy, muscle or blood-cell respiration data exist in the retrieved sources. Framing caution These are animal and isolated-mitochondria findings, not evidence of mitochondrial damage in people taking mefenamic acid and not a reason to take mitochondrial supplements. No dose is recommended here.

Limits on what this shows

• Null findings: No human serum or tissue malondialdehyde, glutathione or antioxidant enzyme measurement with mefenamic acid was retrieved. The healthy-animal oxidative status after mefenamic acid alone was not read.
• At-risk subgroups: None established.

Regulatory status

No oxidative-stress statement in the label text read [1][2].

Population evidence

Tier B, mefenamic acid in every cohort, none human. Lee 2023 [27] (MDPI article page read through a summarizing fetch; abstract also via OpenAlex): HepG2 cells, mefenamic acid at 5 to 200 micromolar (main experiments 200 micromolar for 24 hours); mefenamic acid significantly induced ROS, upregulated HMOX1, GCLM, NQO1 and SOD1 through the Nrf2 and SQSTM1 pathway, and protected cells from 1 mM tert-butyl hydroperoxide; the protection disappeared when SQSTM1 was knocked down; no significant cytotoxicity from mefenamic acid alone at the concentrations tested. Dominguini 2022 [26]: in septic rats, TBARS, protein carbonyl, DCF-DA signal and 8-oxoG were raised by sepsis and reduced by mefenamic acid at 30 and 50 mg/kg; no mefenamic acid-only healthy group. Jarrar 2017 and 2020 [24][25] reported ultrastructural damage without redox markers. Dimiza 2011 [29] reported antioxidant and radical scavenging activity for copper(II) complexes of mefenamic acid, not for the free drug.

Mechanism

Present at cell level: Nrf2 pathway activation and SQSTM1 dependence [27]; overlap with mitochondrial effects (ROS from mitochondria was not isolated as a source).

Hard Rule 2 and 6 checks

ROS and Nrf2 changes in cells are not a deficiency of an antioxidant nutrient, and no antioxidant vitamin or mineral level was measured. The sepsis model confounds any redox direction by disease.

Counter-evidence held on record

Counter-evidence and confounding The 200 micromolar concentration is above clinical total plasma peak (see Mitochondrial burden). The protective direction [26][27] and the injury direction [24][25] cannot be reconciled from these data. Framing caution Findings run in both directions in laboratory models and say nothing about human antioxidant status. They are not a reason to take antioxidant supplements. No dose is recommended here.

Limits on what this shows

• Null findings: No 51Cr fecal blood loss study of mefenamic acid and no ferritin or transferrin saturation data in a GI-bleeding context were retrieved. No iron absorption study was found.
• At-risk subgroups: Label class statements: aspirin co-users, warfarin users, prior ulcer history, older adults [1]. Women with heavy menses are a separate group whose baseline iron status is driven by the indication.

Regulatory status

Label sec 5.2: upper GI ulcers, gross bleeding or perforation occur in approximately 1% of NSAID-treated patients at 3 to 6 months and in about 2 to 4% at one year; GI events can occur without warning symptoms. Aspirin at analgesic doses is not generally recommended with mefenamic acid and warfarin has a synergistic effect on bleeding. Sec 5.12 anemia text as in the Oxygen entry. Iron is not mentioned [1][2]. Use for pain is limited to one week or less [1].

Population evidence

Fraser 1983 [13] (study of 36 women, mefenamic acid given during menstrual periods for over one year): menstrual blood loss fell from a placebo-baseline 65.6 mL to 45.3 mL, then 49.2 mL at 6 to 9 months and 42.8 mL at 12 to 15 months; serum ferritin increased significantly between admission and the end of follow-up (abstract reconstructed through OpenAlex; full text not opened; iron co-supplementation is not stated in the abstract). CLASS EXTRAPOLATION: Bofill Rodriguez 2019 [14] (Cochrane review, 19 randomized trials, 759 participants) found NSAIDs reduced menstrual blood loss versus placebo; the abstract names no individual NSAID, so whether mefenamic acid was in the cohort is unconfirmed here, and the source is capped at yellow and used for context only. Fecal blood loss: Moore 2008 [15] (47-trial 51Cr systematic review) failed the cohort test: mefenamic acid or fenamates do not appear in its abstract, results, Table 2, Figure 2 or text per the summarizing fetch. Cooke 1976 [16], a narrative review known here through a PubMed listing summary only, states that mefenamic acid and other aspirin substitutes appear to cause less fecal blood loss than aspirin but that long-term effects were not evaluated.

Mechanism

GI bleeding: present at class level (NSAID mucosal injury; label sec 5.2). Menstrual blood loss reduction: prostaglandin pathway inhibition is the stated class action [1]. Iron absorption: no mechanism and no study were found; the iron and hemoglobin PubMed listing returned only menorrhagia, hemolysis and review records.

Hard Rule 2 and 6 checks

GI bleeding and anemia are outcomes, not iron deficiency. Iron deficiency needs measured iron indices. Confounding by indication is presumed: in the menorrhagia study the women began with an iron-depleting condition, so a ferritin rise reflects reduced blood loss against that baseline and the ferritin change was not compared with a non-mefenamic arm.

Counter-evidence held on record

Counter-evidence and confounding The label percentages are class-level NSAID rates. The one study that measured ferritin [13] shows a favorable direction, has no ferritin comparison group and does not address the GI bleeding pathway at all, so the two directions cannot be netted against each other. Framing caution The bleeding pathway is a toxicity pathway, not an intake or absorption deficiency. New iron deficiency or black stools while taking mefenamic acid calls for evaluating the bleeding source and the prescription, not an iron supplement by default. The menstrual finding does not mean the drug raises iron. No iron dose is recommended here.

Limits on what this shows

• Null findings: No copper, ceruloplasmin or zinc-to-copper ratio data with mefenamic acid were retrieved.
• At-risk subgroups: None established.

Regulatory status

No copper statement in the label text read [1]; the summarizing reads found no mention of minerals.

Population evidence

Tier B, in vitro chemistry with mefenamic acid as the ligand. Dimiza 2011 [29] (abstract read through OpenAlex): synthesis and characterization of copper(II) complexes of mefenamic acid with heterocyclic ligands, with DNA intercalative binding, serum albumin binding, antioxidant and radical scavenging activity and soybean lipoxygenase inhibition reported. A PubMed listing for mefenamic acid and copper (10 records) held further Cu(II) complex and catalyst papers (titles and listing summaries only, not used), environmental and analytical papers, and menorrhagia or intrauterine device clinical records; none measured copper status.

Mechanism

Chelation chemistry in the test tube. No gastrointestinal chelation or urinary copper loss was shown, so no depletion mechanism is established.

Hard Rule 2 and 7 checks

Complex formation by a synthesized compound is not a copper deficiency. Copper is graded separately from Iron. The intrauterine device records in the listing concern bleeding or pain treatment in people using copper-bearing devices and are not a dietary copper interaction.

Counter-evidence held on record

Counter-evidence and confounding This grade rests on coordination chemistry that shows capability only. The same paper supplies the antioxidant test results cited in the Oxidative stress entry. Framing caution This is laboratory chemistry, not evidence that mefenamic acid lowers copper or that copper should be supplemented. No dose is recommended here.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or Tier C evidence supports even a yellow grade for a microbiome effect of mefenamic acid, and no circulating consumer claim exists to rebut as white. This is a disclosed gap, not a finding of safety. Repair status: re-searched across PubMed (three zero-record listings with different phrasings plus an antibacterial listing), OpenAlex single-record lookup and a Maseda 2019 check; still no mefenamic-specific microbiome study. It is a genuine literature gap as far as PubMed indexing shows, with one residual tooling limitation: the Maier supplementary drug-by-strain table and general web search could not be accessed, so a later revision should re-check whether mefenamic acid is in that screen.

Channels tried

PubMed listing "mefenamic acid AND (gut microbiota OR microbiome OR intestinal bacteria OR diarrhea)" (opened, 10 records): overdose, enteropathy, menorrhagia, a fenamate chloride-secretion paper, a urolithin A conjugate paper and similar; no microbiome composition study. PubMed listing "mefenamic acid AND (bacteria OR bacterial OR microbiota)" (opened, 10 records): environmental biodegradation, electrochemical transformation, an efflux pump review that lists mefenamic acid as a substrate of resistance-nodulation-division pumps (listing text only, not opened), a cyclodextrin nanofiber formulation with antimicrobial activity, a bacterial P450 metabolism paper and similar; none examined human or animal gut bacteria. Maier 2018 [28] (Nature, in vitro screen of non-antibiotic drugs against gut bacterial strains; text read through a summarizing fetch): the only fenamate named was meclofenamic acid, described as just below the hit threshold in Extended Data Figure 2b; mefenamic acid was not named in the text read, and the supplementary drug-by-strain table was not accessible. Wikipedia [4], Healthline [33] and People's Pharmacy [32] contain no microbiome claim. WebSearch reported its shared budget exhausted in both the original review and this repair. Repair-pass PubMed listings (opened, each reported by the fetch tool as a zero-record result): "mefenamic acid"[title/abstract] AND "gut"[title/abstract] AND bacteria[title/abstract]; mefenamic acid gut microbiome dysbiosis; mefenamic acid fecal microbiota 16S. A repair-pass listing "mefenamic acid"[title/abstract] AND antibacterial[title/abstract] (opened) returned a biofilm study of mefenamic acid against Pseudomonas aeruginosa and Staphylococcus epidermidis (Abidi 2019, listing summary only, not opened; not gut commensals and not a gut microbiome measure), an MRSA antibiotic-synergy study of NSAIDs (Chan 2017, listing summary only, not opened) and the efflux pump review already noted; none addressed gut microbiota. Maier 2018 [28] was not re-readable at supplementary-table level. An indomethacin-only NSAID microbiome mouse study (Maseda 2019, mBio, abstract read via OpenAlex; DOI 10.1128/mBio.02282-18) names no fenamate and was not used. Europe PMC and Semantic Scholar are barred by robots rules for the fetch tool, and Crossref, OpenAlex search and several PubMed listings were rate-limited (HTTP 429) during repair.

Class extrapolation considered

CLASS EXTRAPOLATION: meclofenamic acid, a fenamate, was reported just below the hit threshold in the in vitro screen [28]. Mefenamic acid was not confirmed to be in that cohort. A sub-threshold result for a sibling does not support a signal, so it was not used to assign yellow.

Context, not evidence for the drug

Diarrhea and enteropathy are labeled or reported outcomes (see the Nutrient absorption entry) but neither shows a change in gut bacteria.

Limits on what this shows

• Null findings: No retrieved source reports hemoglobin, hematocrit, arterial oxygenation or tissue oxygen delivery as an outcome of mefenamic acid in controlled data. The 1968 screen found only one transient positive antiglobulin test among 36 long-term users [6]. Methemoglobin was not found in the label text read or in the PubMed iron and hemoglobin listing.
• At-risk subgroups: Not defined in the sources read. The 1968 cases were long-term users; no dose, age or HLA pattern was extractable.

Regulatory status

Label sec 5.12 (hematological toxicity) says anemia has occurred in NSAID-treated patients, possibly from occult or gross blood loss, fluid retention or an incompletely described effect on erythropoiesis, and says to monitor hemoglobin or hematocrit if signs or symptoms of anemia occur [1][2]. One label read listed hemolytic anemia among less frequent hematopoietic reactions [1]; a second label read returned no hemolytic anemia text [2]. Both reads were through a summarizing fetch tool, so the label status of the term is unresolved.

Population evidence

Scott 1968 [6] (case series with a screen): three patients developed warm-antibody autoimmune hemolytic anemia while taking mefenamic acid and each recovered after withdrawal; in vitro attempts to show inhibition or enhancement of antibody activity by the drug did not succeed; direct antiglobulin tests on red cells from 36 long-term users were positive, transitorily, in one (abstract reconstructed through OpenAlex; full text not opened). Chan 1991 [8]: a single case of Stevens-Johnson syndrome, cholestatic hepatitis and hemolytic anemia (title and metadata only; no abstract available). Robertson 1971 [9]: a case report titled haemolytic anaemia associated with mefenamic acid (title and metadata only). Shepherd 1986 [12]: a PubMed listing summary describes bullous pemphigoid, Coombs-positive hemolytic anemia and diarrhea related to mefenamic acid, resolving after discontinuation (listing summary only). Reviews: Sanford-Driscoll 1986 [7], whose PubMed listing summary states that mefenamic acid was the most frequently implicated NSAID and that the mechanism is autoimmune, similar to methyldopa (OpenAlex held no abstract, so this is the listing summary only); Homberg 1999 [10] and Worlledge 1973 [11] list mefenamic acid among drugs that cause drug-induced hemolytic anemia (listing summaries only).

Mechanism

Present: an immune mechanism. Per the reviews above it is autoantibody mediated; the mechanism was not tested in the case series, where drug-dependent in vitro effects were not demonstrated [6].

Hard Rule 2 and 6 checks

Hemolytic anemia is a clinical outcome of an immune drug reaction. It is not iron, B12 or folate deficiency and not a measure of oxygen delivery. Attribution rests on temporal association and recovery after withdrawal in case reports; incidence cannot be estimated from them.

Counter-evidence held on record

Counter-evidence and confounding The reaction is uncommon by the one screen available [6], the case reports lack denominators and are subject to publication bias, and the two label reads disagree on whether hemolytic anemia is listed [1][2]. The reviews are known to this review only through listing summaries. Framing caution This is a drug reaction, not a deficiency. New fatigue, pallor, dark urine or yellowing of the skin or eyes while taking mefenamic acid calls for prompt medical evaluation and a review of the prescription. Iron, folate or B12 supplements do not address it. No dose is recommended here.

Limits on what this shows

• Null findings and direction: Dominguini 2022 [26]: in septic rats mefenamic acid increased respiratory complex activities (an opposite direction in a disease model); ATP content was not among the measurements described in the retrieved text. The 28-day rat studies [24][25] reported ultrastructure only, not ATP.
• At-risk subgroups: None established.

Regulatory status

No ATP or energy-metabolism statement in the label text read [1][2].

Population evidence

Tier B, mefenamic acid in every cohort, none human. Mingatto 1996 [22] measured respiration and ATP synthesis in rat kidney mitochondria according to its title; the PubMed listing summary states mefenamic acid uncoupled and inhibited oxidative phosphorylation (the listing summary also gave an ATP-synthesis IC50 value that is not repeated here because the abstract could not be read). Tatematsu 2016 [23]: mefenamic acid and related fenamates dose-dependently obstructed electron transport in rat liver mitochondria (abstract read). Uyemura 1997 [21]: permeability transition in rat renal cortex mitochondria (listing summary).

Mechanism

Present in isolated systems: uncoupling and electron-transport inhibition reduce ATP generated per oxygen consumed; permeability transition dissipates the membrane potential that drives ATP synthase.

Hard Rule 2, 7 and 9 checks

Reduced mitochondrial ATP synthesis in a test tube is not a nutrient deficiency and not fatigue in people. This entry shares its evidence with Mitochondrial burden and Oxidative stress; see the Overlap map. It is graded separately per Hard Rule 7 and is not an independent confirmation.

Counter-evidence held on record

Counter-evidence and confounding The mefenamic ATP-specific result is known only through a listing summary and a title [22]. Effective concentrations are in the micromolar to sub-millimolar range, with some at or above clinical total plasma levels. The sepsis model points the other way [26]. No human data. Framing caution This is a mechanistic signal from isolated organelles, not evidence that mefenamic acid lowers energy levels in people and not a reason to take energy or mitochondrial supplements. No dose is recommended here.