Erythromycin

Limits on what this shows

• Null findings and exclusions: A yeast sporulation study naming erythromycin as an inhibitor of mitochondrial protein synthesis (Marmiroli 1981 [34]) reports no respiration or ATP data and is not used. A PubMed query for erythromycin with mitochondria or respiration was uninformative (the OR operator broadened it to 31,418 records, none about erythromycin) and a Crossref title search returned no erythromycin-mitochondria study [35]. No human measurement (muscle, blood cells, lactate) after erythromycin was found.

Direction

Possible mitochondrial translation or function impairment (mechanistic plausibility). Not a nutrient deficiency; hepatotoxicity is a drug toxicity and is not filed here.

Regulatory status

The US label and SmPC do not mention mitochondria; the label mechanism is antibacterial [1][2].

Population evidence (Tier B, in vitro; abstracts only)

Prajapati 2019 [12]: erythromycin and clindamycin "induced caspase activation and cell death in several different human cell lines", compromised the mitochondrial respiratory chain, impaired autophagy flux and reduced lysosome numbers so that damaged mitochondria accumulated; TFEB overexpression restored mitochondrial function and rescued cell death. Cell lines, concentrations, durations and all numbers were not available: the Springer page showed only the abstract and the OpenAlex record had no abstract (abstract read from Springer and a second record, the BVS/MEDLINE page). Adoutte 1972 [13]: Paramecium aurelia exposed to erythromycin for 12 hours to 6 days showed progressive elongation of mitochondria and a considerable decrease in cristae, whereas three resistant mutants kept near-normal mitochondrial structure; concentrations were not stated on the page read, and the page reports structure only, not respiration.

Mechanism

Proposed: the 39S mitochondrial large subunit resembles the bacterial 50S ribosome, so erythromycin binds it and blocks mitochondrial translation; the review statement in Lamb 2015 [14] is "the erythromycins and chloramphenicol selectively bind to the large subunit of the mitochondrial ribosome". Lamb 2015 did not test erythromycin itself (azithromycin was the macrolide tested), so it supports only the hypothesis. Prajapati reports an effect on respiration, with mitophagy impairment as an added mechanism [12].

Counter-evidence held on record

Counter-evidence and confounding Cell-line cytotoxicity data do not establish plasma or tissue exposure in people; the concentrations were not retrieved, so the comparison to the label peak of 1.13 to 1.68 micrograms/mL [1] could not be made. The Paramecium study is a protist. Framing caution Do not state that erythromycin damages mitochondria in people or causes fatigue through energy failure. This is a laboratory hypothesis, graded for mechanistic plausibility only.

Limits on what this shows

• Null findings: No human study measuring malondialdehyde, glutathione or antioxidant capacity in erythromycin-treated people appeared in the top 10 of 213 PubMed records [17].

Direction

Mixed. Reverse-direction (protective) in inflammatory-stress models; pro-oxidant in cytotoxicity models. Not a deficiency.

Regulatory status

Not addressed in the label or SmPC [1][2].

Population evidence (Tier B, in vitro and animal; abstracts and one full text)

Li 2024 [15]: U937 human macrophages, erythromycin 1 microgram/mL pre-incubated 24 hours before cigarette smoke extract; the smoke extract raised ROS and malondialdehyde and lowered SOD and PPAR-gamma, and erythromycin reversed each (P under 0.05; n = 7 for ROS, 8 for MDA and SOD; the text gives directions only, no numbers). A PubMed listing of 213 records [17] shows erythromycin lowering senescence markers via PI3K-mTOR in smoking mice and BEAS-2B cells (PMID 36506578, Front Pharmacol 2022), mixed catalase and TBARS changes in rainbow trout (PMID 26760278), and an erythromycin-induced ROS response in a microalga; none is human. Prajapati 2019 [12]: erythromycin increased ROS as part of respiratory chain impairment in human cell lines. Ranadeera 2025 review [16]: erythromycin suppressed oxidative stress-induced senescence and boosted antioxidant enzymes in mammalian cells but caused oxidative damage in aquatic species; no human data reported in the abstract.

Mechanism

Protective findings: PPAR-gamma and PI3K-mTOR signaling in the cited cell work [15][17]; pro-oxidant findings: respiratory chain impairment [12].

CLASS EXTRAPOLATION (Tier C, not counted)

Larsen 2017 [18]: two randomized trials in healthy adults; one week of clarithromycin raised urinary 8-oxodG by 22.0% (95% CI 3.6 to 40.4) and 8-oxoGuo by 14.9% (3.7 to 26.1), with no change in plasma malondialdehyde. Cohort test: the macrolide was clarithromycin; erythromycin was not in the cohort. Used only as context; it does not carry onto erythromycin.

Counter-evidence held on record

Counter-evidence and confounding Cell models use unrealistic exposure contexts; the protective macrophage result depends on a cigarette-smoke insult, and the direction differs from the respiratory-chain study. Numeric values for [15] were in figures not available. Framing caution Do not claim that erythromycin causes oxidative stress or depletes antioxidants in people. Part of the laboratory data point to an antioxidant effect.

Limits on what this shows

• Null findings: A PubMed listing for erythromycin and iron (173 records, first 10 read) found no human or animal study of erythromycin affecting iron absorption, ferritin or iron status; other records were guidelines, synthesis, bacterial or P450 heme-iron chemistry, and a pharmacist review that names erythromycin only as a CYP3A4 substrate (its abstract was not retrieved) [25]. A hemolytic anemia case series is filed under Oxygen delivery/utilization, not here.

Direction

Possible binding of iron by the drug in vitro (mechanistic plausibility). Not a depletion; not merged with Copper (Hard Rule 7).

Regulatory status

The label and SmPC name no iron interaction [1][2].

Population evidence (Tier B, in vitro chemistry; PubMed abstract summary)

Muranaka 1997 [19], Biochem Biophys Res Commun 232:183-187: erythromycin plus iron chloride at equal molar amounts formed a complex, stable in neutral buffer, that scavenged superoxide (electron spin resonance, luminol chemiluminescence, cytochrome c reduction) to a degree equal to L-ascorbic acid and reduced superoxide release from phorbol ester-stimulated human neutrophils and alveolar macrophages; erythromycin alone had few such effects. Concentrations and numerical values were not stated in the abstract summary.

Mechanism

In vitro complexation is shown. Whether it occurs in the gut lumen, bile or blood at relevant concentrations, or changes iron absorption or status, was not studied in anything retrieved.

Counter-evidence held on record

Counter-evidence and confounding One in vitro abstract from 1997; a stable complex in buffer does not establish iron loss in a person. Framing caution Do not state that erythromycin lowers iron or causes iron deficiency. Do not suggest iron supplementation.

Limits on what this shows

• Null findings: A PubMed listing for erythromycin and copper (85 records, first 10 read) contained environmental co-resistance studies (copper and erythromycin resistance in Enterococcus), a bacterial resistance association, catalytic degradation of erythromycin by copper peptides, analytical and synthetic papers, and the one complexation paper [24]. No human or animal copper measurement after erythromycin.

Direction

Possible binding of copper by the drug in vitro (mechanistic plausibility). Not a depletion and not merged with Iron (Hard Rule 7).

Regulatory status

The label and SmPC name no copper interaction [1][2].

Population evidence (Tier B, in vitro chemistry; OpenAlex reconstructed abstract)

Egu 2024 [23], Heliyon 10:e35536: erythromycin extracted from commercial tablets was reacted with silver, nickel, cobalt and copper salts; elemental analysis, thermogravimetry, FTIR and UV-visible spectroscopy confirmed complexes; the copper complex had a minimum inhibitory concentration of 1.56 mg/mL and a minimum fungicidal concentration of 6.25 mg/mL against the test organisms. Stoichiometry was not stated in the content read.

Mechanism

Complexation chemistry only, aimed at making antimicrobials. Whether it happens in the gut or blood and changes copper status was not studied.

Counter-evidence held on record

Counter-evidence and confounding The complexes were made at synthetic scale for an anti-infective project; this says nothing about copper handling in an organism. Framing caution Do not state that erythromycin lowers copper or causes copper deficiency. Do not suggest copper supplementation.

Limits on what this shows

• Population evidence: children (Tier A, with a null alpha result): Thavamani 2023 [6]: single-center observational case-control pilot (Cleveland), 11 children on prokinetic erythromycin ethylsuccinate and 22 age-matched controls with feeding intolerance, 8 and 20 analysed after quality control; most were under 3 years; duration 4 to 32 weeks (median 12, IQR 9 to 23); 16S and ITS sequencing; no pre-exposure samples, no randomization. Shannon diversity (P = 0.71 bacteria and fungi) and richness (P = 0.6 bacteria, P = 0.35 fungi) did not differ; bacterial beta diversity at genus and species level did not differ (species P = 0.7129); fungal beta diversity at species level did (P = 0.0297). Candida abundance rose and Galactomyces fell; Prevotella rose and Clostridium fell (no p-values in the text). The authors attribute the species-level difference "likely" to the smaller erythromycin group. Full text via MDPI page read through a summarizing fetch; the abstract was confirmed by a PubMed record (PMID 37998808).
• At-risk subgroups: Long-term low-dose users (4 weeks or more) [5]; infants and children on prolonged prokinetic courses (4 to 32 weeks), for fungal composition [6]; people with C. difficile risk factors (label) [1].

Direction

Disturbance risk (excess-risk framing). This is not the depletion of a nutrient, and no supplement or probiotic is implied.

Regulatory status

The label warns that C. difficile-associated diarrhea "can range from mild diarrhea to fatal colitis" and "must be considered in all patients who present with diarrhea following antibiotic use" [1]. The SmPC reports pseudomembranous colitis as rare and GI effects as the most frequent, dose-related reactions [2]. No label text addresses the microbiome itself.

Population evidence: adults (Tier A)

Choo 2023 [5]: single-blinded randomized parallel 4-week trial in healthy adults, 20 participants, erythromycin ethylsuccinate 400 mg twice daily (n = 10) vs azithromycin 125 mg twice daily (n = 10); stool and fasting serum at baseline and end of treatment; no placebo arm (full text via summarizing fetch of the ASM article; a second record, the OpenAlex works lookup, confirmed title, PMID 37347185 and the conclusion that long-term macrolide use depleted keystone gut bacteria). Taxon count fell with erythromycin (P = 0.006) and diversity fell (P = 0.010); evenness and total bacterial load did not change significantly. Combined PERMANOVA P = 0.0002 for beta diversity, with erythromycin and azithromycin not different from each other (P = 0.826). Both drugs depleted B. longum, B. adolescentis, A. muciniphila and B. wadsworthia and raised Eggerthella lenta, R. gnavus and Coprococcus comes; erythromycin alone lowered R. bicirculans, R. inulinivorans and Actinomyces sp. ICM47 and raised Blautia obeum and Flavonifractor plautii (genus abbreviations as given in the source summary). Predicted functional pathways changed with erythromycin (P = 0.014; carbohydrate metabolism, glycolysis and fermentation reduced). Effect sizes and confidence intervals were not retrieved, only P-values.

Supporting Tier B

Maier 2021 [9]: 144 antibiotics tested against human gut commensal species; macrolides and tetracyclines "inhibited nearly all commensals tested but also killed several species", and erythromycin was the antibiotic in the gnotobiotic mouse experiment in which dicumarol protected Bacteroides (abstract and figure captions only; concentrations and strain counts not retrieved). Choo 2023 [5] mouse arm: erythromycin 20 mg/kg in drinking water for 90 days lowered stool acetate, propionate and butyrate; short-chain fatty acids were not measured in the human arm.

CLASS EXTRAPOLATION (Tier C, capped at yellow, not counted toward this grade)

Korpela 2016 [7]: 142 Finnish children aged 2 to 7 years with sampled stool and purchase records; macrolide use was linked to fewer Actinobacteria (Bifidobacterium, Collinsella fell 4 to 10-fold), about twofold more Bacteroides and Proteobacteria, lower bile-salt hydrolase and higher macrolide resistance, with changes persisting beyond 6 months. The cohort test was run on the full text: the macrolides were "azithromycin and clarithromycin"; erythromycin appears only as the reagent in a resistance culture test. Erythromycin was therefore NOT in the cohort. The McDonnell 2021 meta-analysis [8] pooled four azithromycin trials (Shannon mean difference -0.86, 95% CI -1.59 to -0.13) and, from the content read, named no erythromycin study. Both are class extrapolation and were used only as context.

Excluded

Eckernas 1991 [33] (dirithromycin in 20 volunteers) is a different drug and is excluded. A PubMed listing for erythromycin and microbiome returned 3 records with erythromycin as the exposure, one of them marine invertebrates [36].

Mechanism

Present. Binding of the bacterial 50S ribosomal subunit blocks protein synthesis [4], commensal bacteria with susceptible ribosomes are exposed, and biliary excretion delivers drug to the gut [1].

Why orange, not red

One prospective erythromycin cohort (10 recipients, no placebo, P-values only) shows a real measured shift with a named mechanism, which clears orange. Red is withheld because (a) there is no second consistent erythromycin cohort: the pediatric pilot found no difference in richness or alpha diversity [6]; (b) the pilot differs in design, age, indication and duration so it neither confirms nor refutes the adult result, but it is contradicting evidence on direction; (c) no recovery or persistence data exist for erythromycin; and (d) the larger macrolide cohorts did not include erythromycin [7][8]. No cohort linked the change to a clinical or nutrient outcome.

Counter-evidence held on record

Counter-evidence and confounding The adult trial has no placebo, so regression and sampling effects cannot be excluded, and it was a 4-week low-dose regimen used in respiratory disease, not a typical 7 to 10 day infection course. The pediatric pilot compared children with feeding intolerance to other children with feeding intolerance, so indication, age and feeding route confound it, and the antibiotic exposure of controls was not described. Microbial change is not shown here to alter vitamin, short-chain fatty acid or other nutrient status in people. Framing caution This is a disturbance risk with long-term low-dose use, not a depleted nutrient and not a reason to take a probiotic or supplement. Do not present it as permanent damage; recovery was not studied. Persistent or severe diarrhea is a clinician question because of C. difficile [1].

Limits on what this shows

• Null findings: No human or animal ATP, lactate or respirometry measurement after erythromycin was found; Lamb 2015 [14] did not test erythromycin. The OR-operator PubMed listing and the Crossref title search were uninformative [35].

Direction

Possible reduction of cellular ATP output (mechanistic plausibility). Not a nutrient deficiency and not independent of the Mitochondrial burden entry.

Regulatory status

Not addressed in the label or SmPC [1][2].

Population evidence (Tier B, weak; abstract only)

Prajapati 2019 [12]: reduced respiratory chain activity "leading to bioenergetic crisis and generation of reactive oxygen species" in several human cell lines. No ATP, oxygen-consumption or membrane-potential numbers, cell lines or concentrations were available in the abstract-level text read.

Mechanism

Proposed: impaired mitochondrial function and mitophagy [12]; the mitoribosome hypothesis in [14] is background only.

Counter-evidence held on record

Counter-evidence and confounding Rests on one abstract; see the Overlap map. Effects in proliferating human cell lines in culture do not transfer to people. Framing caution Do not attribute low energy, fatigue or reduced ATP to erythromycin from this entry.

Limits on what this shows

• At-risk subgroups: None established; both cases were children, and one received a sulfonamide co-drug.

Direction

Rare immune hemolysis reducing red-cell mass (drug toxicity). Not an iron or oxygen-carrying deficiency; anemia is not iron deficiency (Hard Rule 2). Not an intended mechanism.

Regulatory status

Hemolysis is not listed in the sections of the label read; reported reactions include hepatic, allergic, cardiac and GI effects [1].

Population evidence (Tier B, case reports; PubMed records, abstract level)

Wong 1981 [20]: "Erythromycin-induced hemolytic anemia", J Pediatr, a case report in a male infant treated for otitis media with C3, IgM and Coombs testing (no abstract text in the record; age, hemoglobin and outcome not retrieved). Nance 1988 [21]: a 3-year-old girl on erythromycin ethylsuccinate plus sulfisoxazole developed severe hemolytic anemia; red cells coated with erythromycin base lysed in vitro; the authors call it the second reported case of immune hemolytic anemia linked to erythromycin. Of 236 records in a PubMed listing for erythromycin with hemolysis, hemolytic anemia or methemoglobinemia, only these two were identified in the first 10 as caused by erythromycin [22].

Mechanism

Immune-mediated, drug-dependent hemolysis in the second case [21]; not shown for the first. Methemoglobinemia and arterial oxygenation were not separately found.

Counter-evidence held on record

Counter-evidence and confounding Two cases in decades of use; the second involved a combination product containing sulfisoxazole, a separate hemolysis candidate, and the abstracts give no hemoglobin values. No trial reported a hemoglobin signal for erythromycin in the material read. Framing caution Do not state that erythromycin lowers oxygen delivery or causes anemia as a routine effect. This is a rare adverse reaction, not a depletion.