Rifaximin

Limits on what this shows

• Null findings: The leak respiration result was read through a summarizing fetch tool: the abstract text was reported as "increased" leak respiration while the extracted figures (145.3 +/- 8.5 versus 175.1 +/- 9.8) were lower than control. Direction of leak respiration is unresolved and is not used. Omar 2022 [17] points in the opposite (biogenesis) direction.
• Hepatic disease confounding: Mitochondrial pathway disturbance is already described in decompensated cirrhosis and ACLF [23], so any future human biomarker study would be confounded by disease.

Regulatory status

No mitochondrial statement in the label [1].

Population evidence

None in humans. Baráth 2022 [16] (Tier B, rifaximin-specific): male Sprague-Dawley rats, n=72, rifaximin 8 mg/kg by gavage twice daily for 3 days. Ileal OXPHOS capacity 170.8 +/- 6.1 versus 209.8 +/- 10.8 pmol/s/mL in controls (p<0.05), about 19% lower; complex II-supported respiration 156.1 +/- 7.4 versus 197.0 +/- 7.9; ileal mucosal injury, median histology grade 2. In vitro at 0.5, 5 and 50 ng/mL, hydrogen peroxide production at 5 ng/mL was 2.9 +/- 0.2 versus 1.9 +/- 0.4 pmol/s/mL. Omar 2022 [17]: rat malathion model with rifaximin co-treatment: PINK1 mitophagy activation at 2 weeks, and at 4 weeks higher PGC-1alpha and TFAM (biogenesis markers) with higher short-chain fatty acids. Pose 2021 [23]: simvastatin plus rifaximin (a combination, so not attributable to rifaximin) shifted metabolites in pathways linked to mitochondrial dysfunction in decompensated cirrhosis.

Mechanism

Proposed. A direct effect on mitochondria was seen in vitro at a concentration in the range of the label's healthy-subject Cmax of 2.4 to 4 ng/mL [1]. The authors note absorption below 1% and attribute raised blood hydrogen peroxide to extensive mucosal damage rather than systemic uptake [16]. Gut tissue concentrations were not reported in the text retrieved.

Cohort test

Rifaximin was the test agent in [16] and [17]. Both are animal or in vitro studies, so Tier B. No rifampin or other rifamycin mitochondrial literature was used.

Counter-evidence held on record

Counter-evidence Opposite biogenesis direction in [17]; animal doses and tissue exposures are not mapped to human dosing; the main study states it did not test a clinical setting [16]. Framing caution Animal and in vitro finding only. Not evidence of mitochondrial toxicity in people, and not a reason to take mitochondrial supplements.

Limits on what this shows

• Null findings: No human oxidative-stress marker measured in rifaximin users was located.

Regulatory status

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

Population evidence

None in humans. Pro-oxidant direction: hydrogen peroxide 2.9 versus 1.9 pmol/s/mL at 5 ng/mL in vitro [16]. Protective direction: malathion testicular toxicity in rats, rifaximin co-treatment improved oxidative stress [17]; chronic-binge ethanol in obese KK-Ay mice, rifaximin 0.1 g/L in the liquid diet prevented ethanol-induced steatosis and oxidative stress and suppressed hepatic NADPH oxidase 2, CD14, TLR4 and TLR2 expression [18]. A 2024 paper on rifaximin, iron overload ferroptosis and liver fibrosis lists oxidative stress, lipid peroxidation and glutathione among its keywords, but its abstract was not available and is not used [20].

Mechanism

In the protective models the effect is attributed to microbiota and innate immune changes [18] or short-chain fatty acid signaling [17], not to antioxidant chemistry of the molecule. Direct redox chemistry of rifaximin was not established in the sources retrieved.

Cohort test

Rifaximin was the test agent in [16], [17], [18]. All are animal or in vitro (Tier B). Disease models (malathion, ethanol) limit transfer to healthy use.

Counter-evidence held on record

Counter-evidence The directions conflict between models, and each protective result occurs in a toxicant-injured tissue, so it may reflect reduced injury rather than a change in baseline redox state. Framing caution Not an antioxidant benefit claim and not a deficiency claim. Do not describe rifaximin as depleting antioxidants or as an antioxidant.

Limits on what this shows

• Null findings: No serum iron, ferritin or transferrin saturation measured in rifaximin users was located (PubMed listing query for rifaximin with iron, ferritin or copper returned no such study).
• Hepatic disease confounding: The cell-model paper itself states liver disease can cause peripheral iron metabolism dysfunction and raised brain iron [19], and anemia is frequent in cirrhosis [22]. Any human iron signal in HE cohorts would be confounded.

Regulatory status

No iron statement in the label. The label's hepatic encephalopathy adverse-reaction table lists anemia; that is an outcome, handled in the Oxygen delivery entry [1].

Population evidence

None in humans. Zhang 2022 [19] (in vitro): ferric ammonium citrate and lipopolysaccharide caused iron deposition and reactive-oxygen-species injury in SH-SY5Y neuronal cells; rifaximin corrected the iron deposition and protected the cells, via STAT3/NF-kB signaling. Yuan 2024 [20]: title reports rifaximin ameliorating iron overload-induced ferroptosis and liver fibrosis in vivo and in vitro; only metadata was available, no abstract or numbers were read.

Mechanism

No absorption, chelation or loss mechanism is described. The cell data show a signaling-mediated effect on iron handling in injured cells [19].

Cohort test

Rifaximin was the test agent in [19] and [20]; both are cell or rodent studies, so Tier B. No rifampin or other rifamycin iron data were used.

Counter-evidence held on record

Counter-evidence The findings run opposite to a depletion claim, but they come from supraphysiologic iron loading in cells and are not clinical evidence of benefit. Framing caution Reverse-direction, preclinical. Iron deficiency is not anemia and anemia is not iron deficiency (Hard Rule 2). No iron supplementation or restriction is implied by this entry.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or C source was opened to support yellow, and no circulating consumer claim about rifaximin and copper was found to rebut. This is a disclosed gap, not a finding of safety. Kept separate from Iron (Hard Rule 7). Re-check when full search access returns.

What was searched

PubMed search-listing query "rifaximin iron OR ferritin OR copper" (returned 13 records, none measuring copper in people); the Xifaxan label, which mentions no copper [1]; the Drugs.com Xifaxan page, which mentions no nutrient [24]. One record seen by title only, a 2024 Heliyon paper on synthesizing rifaximin with copper nanoparticles, is materials chemistry, was not opened, and is not evidence about copper status. The Antidote Research rifaximin pages returned 404.

Channel outcomes

PubMed listing pages: worked for some queries and returned 429 for others. OpenAlex single-record lookups: worked. DailyMed: worked. Web search: budget exhausted. Crossref, Europe PMC, Semantic Scholar, Unpaywall and direct PubMed API calls: not attempted (org network policy). Mixed result: a clean zero on the channels that worked, with the consumer-claim audit tooling-limited. Likely a genuine thin literature, but not proven.

Limits on what this shows

• Null findings: Kaji 2017 [4]: 20 patients with advanced cirrhosis, 400 mg three times daily for 4 weeks: ammonia, endotoxin and cognition improved with no change in microbiome composition or diversity. Mouse post-infectious IBS model [12]: minimal change in composition and diversity and negligible effect on Lactobacillus and Bifidobacterium. A 2018 randomized trial in cirrhosis is titled "minor effects on bacterial composition" [11] (title and opening line only seen).
• At-risk subgroups: Patients with cirrhosis or hepatic encephalopathy, where systemic rifaximin exposure is 10- to 21-fold higher [1] and C. difficile infection risk is high [14]; people given repeated courses, in whom resistance selection is the concern.
• Hepatic disease confounding: Cirrhosis and alcoholic hepatitis carry their own dysbiosis: severe alcoholic hepatitis patients had reduced diversity versus controls before treatment [8], and IBS patients differed from healthy controls at baseline [7]. Several cirrhosis cohorts are uncontrolled before and after studies [4][5][6], and co-medications were not reported in the abstracts opened, so shifts cannot be fully separated from disease course.

Regulatory status

The label warns of C. difficile-associated diarrhea and of drug-resistant bacteria with unnecessary use, and describes rpoB-mediated resistance. Its microbiology section contains no statement about effects on intestinal flora composition [1].

Population evidence

Fodor 2018 [3]: TARGET 3 IBS-D trial framework (open-label rifaximin 550 mg three times daily for 2 weeks, then randomized repeat courses or placebo); 16S sequencing of 103 stool samples. After 7 days of treatment seven taxa, including Peptostreptococcaceae, Verrucomicrobiaceae and Enterobacteriaceae, were lower at a 10% false discovery rate; changes were largely transient with minimal sustained alteration up to 46 weeks. Kaji 2020 [5]: 30 patients with advanced cirrhosis, 1200 mg daily for 4 weeks, 16S: diversity and richness preserved, Streptococcus decreased, endotoxin activity and permeability markers fell. Bajaj 2013 [6]: 20 cirrhosis patients with minimal hepatic encephalopathy, 550 mg twice daily, 8 weeks: modest decreases in some bacterial families, altered network connectivity, serum fatty acids higher, endotoxemia lower. Zeber-Lubecka 2016 [7]: 72 IBS patients and 30 controls, 16S plus metabolomics: patients differed from controls at baseline; treatment had limited prolonged effects. Kim 2020 [8]: 24 severe alcoholic hepatitis patients; after rifaximin, 17 of 23 taxa shifted back toward the healthy pattern. Huang 2024 [9]: 80 patients randomized, 400 mg twice daily for 8 weeks, 16S on gastric lavage: composition differed between arms.

Mechanism

Present and intended: antibacterial inhibition of bacterial RNA polymerase, with resistance via rpoB mutation [1]. In HE patients, sensitivity of Streptococcus salivarius to rifaximin tracked conjugated secondary bile acid levels, and Ruminococcus gnavus was rifaximin-nonresponsive [10]. A mouse study proposes that bile-acid complexing raises rifaximin's activity in the duodenum and jejunum, which stool studies would miss [13].

Functional changes

Serum fatty acids rose in the human MHE cohort [6]. A rat malathion model reported higher acetate- and propionate-producing flora and short-chain fatty acids after 4 weeks [17] (Tier B, rodent). Human fecal bile acid and short-chain fatty acid measurements on rifaximin were not located.

C. difficile and resistance (the drug's own risk)

Reigadas 2017 [14]: retrospective, 388 cirrhotic patients over 6 years, 127 with diarrhea, 46 with confirmed C. difficile infection, 14 of them (30.4%) on rifaximin prophylaxis for hepatic encephalopathy. Rifaximin resistance was found in 34.1% of isolates overall and 84.6% among treated patients; rifamycin therapy and ribotype 001 were risk factors for resistant strains. A 2020 cohort title reports chronic rifaximin use in cirrhosis associated with a lower C. difficile rate [15] (abstract not opened; no numbers used). In rats, C. difficile did not rise after rifaximin [16]. The direction is mixed and this is a safety context, not a nutrient finding.

Direction and magnitude

Hard Rule 2: this is a compositional change in the gut, not a nutrient deficiency and not shown to be harm. Orange rather than red because the effect is modest and partly transient, and one Tier A cohort found no change [4]. Boundary: dosing period and the weeks after it.

Cohort test

Rifaximin was the named drug in every human cohort cited [3][4][5][6][7][8][9]. No rifampin, other rifamycin or other antibiotic data were used. Not a class extrapolation.

Counter-evidence held on record

Counter-evidence and confounding Null composition result in cirrhosis [4], mouse data showing minimal change [12], and the transience seen in IBS-D [3][7]. Most cohorts are small (20 to 80 patients), several lack a control arm, and sequencing was of stool except for gastric lavage [9]. Full text of every cited paper was not opened; abstracts were read through OpenAlex records condensed by a summarizing fetch tool. Framing caution The customer-visible wording must keep the boundary: modest shifts in which bacteria are present, mostly temporary, with harm not demonstrated. This is not a diagnosis of dysbiosis and not a reason to take a probiotic or any supplement. New or worsening diarrhea during or after a course is a clinical question for a prescriber, because of the C. difficile warning.

Limits on what this shows

• Hepatic disease confounding: The trial population has cirrhosis with recurrent hepatic encephalopathy [1], and anemia is common in that population [22]. The label gives no confidence interval or adjustment; a 4-point imbalance in a selected adverse-event table could be chance or disease course.

Regulatory status

Label sec 6.1, HE Trial 1 (550 mg twice daily, N=140 versus placebo N=159): anemia 11 (8%) versus 6 (4%). The table is limited to adverse reactions with at least 5% incidence and a higher incidence on rifaximin than placebo, so it is a selected list [1].

Population evidence

No other rifaximin study measuring hemoglobin, hematocrit or tissue oxygenation was located. Anemia is frequent in cirrhosis and predicts worse outcomes [22], and that paper reports the prognostic value of hemoglobin change depended on rifaximin co-use; the abstract gives no further detail and is not used for a drug effect.

Mechanism

Absent. No marrow, hemolysis, blood loss or nutrient-absorption mechanism for rifaximin was found.

Hard Rule 2

Anemia is a clinical outcome, not an oxygen-delivery deficiency caused by the drug and not iron deficiency. It is carried as a context flag only.

Cohort test

Rifaximin was the study drug in the label trial [1]. Table read through a summarizing fetch tool.

Counter-evidence held on record

Counter-evidence The IBS-D and traveler's diarrhea tables were not checked for anemia. The imbalance is not tested for significance in the extract. Framing caution Do not describe this as reduced oxygen-carrying capacity from rifaximin. It is an unadjusted adverse-event count in a cirrhosis trial.

Limits on what this shows

• Null findings: No human ATP or "low energy" outcome was located. This entry is kept independent of Mitochondrial burden and of Oxygen delivery/utilization (Hard Rule 7); the shared source is disclosed in the Overlap map.

Regulatory status

No energy-metabolism statement in the label [1].

Population evidence

Baráth 2022 [16]: ileal OXPHOS capacity 170.8 versus 209.8 pmol/s/mL (p<0.05); respiration is a surrogate for ATP-linked capacity, not an ATP measurement. Omar 2022 [17]: higher PGC-1alpha and TFAM in a rat toxicity model. Pose 2021 [23]: combination therapy with simvastatin shifted carnitine and beta-oxidation pathway metabolites in decompensated cirrhosis; not attributable to rifaximin alone.

Mechanism

Possible via the reduced respiratory capacity [16], not demonstrated for ATP output.

Cohort test

Rifaximin was the test agent in [16] and [17]; animal data, Tier B.

Counter-evidence held on record

Counter-evidence Opposite biogenesis direction [17]; confounding by cirrhosis metabolic disturbance [23]. Framing caution Not a fatigue or low-energy claim. This is a mechanistic flag from one animal study.