Aspirin

Limits on what this shows

• Key negative/null findings: No modern human study measuring mitochondrial function, mitochondrial DNA content, or a validated mitochondrial-burden biomarker in low-dose cardioprotective aspirin users was located; this is an evidence gap, not a tested-and-negative result.
• At-risk subgroups: The only population with a well-documented clinical correlate of this mechanism in the sources located is severe salicylate overdose/poisoning, a toxicology emergency rather than a nutrient-status question. A historical clinical association between aspirin and Reye syndrome in children with concurrent viral illness is widely referenced in the toxicology literature reviewed here (see Trost & Lemasters, above), but the current U.S. label's specific warning language could not be independently re-retrieved within this review's tooling and is not asserted as label text in this entry.

Regulatory status

Not addressed as a nutrient-depletion finding in the label material reviewed for this dossier. StatPearls' salicylate toxicity reference describes mitochondrial uncoupling as a toxicity mechanism at blood salicylate concentrations of roughly 40 mg/dL and above (mild toxicity), rising through 80–100 mg/dL (moderate) and above 100 mg/dL (severe) — concentrations far above what 81 mg/day cardioprotective dosing produces.

Population evidence

No Tier A human study of low-dose cardioprotective aspirin and mitochondrial function was located. All supporting data is Tier B, in vitro or animal: isolated gastric mucosal mitochondria (Jørgensen et al., Scandinavian Journal of Clinical and Laboratory Investigation, 1976) showed salicylate- and aspirin-induced uncoupling of oxidative phosphorylation; membrane-biophysics modeling (Gutknecht, Journal of Membrane Biology 1990, and Molecular and Cellular Biochemistry 1992) describes salicylate-driven proton transport across lipid bilayers and mitochondrial membranes as the physical basis of this uncoupling; isolated rat-liver mitochondria and cultured rat hepatocytes (Trost & Lemasters, Journal of Pharmacology and Experimental Therapeutics 1996, and Toxicology and Applied Pharmacology 1997) showed salicylate triggers the mitochondrial permeability transition, framed by the authors around Reye syndrome and toxic liver injury; and isolated rat intestinal mitochondria (Somasundaram et al., Gut 1997) showed mitochondrial damage as part of the direct-contact phase of NSAID-induced intestinal injury.

Mechanism

Named: salicylate is a weak lipophilic acid that shuttles protons across the inner mitochondrial membrane, collapsing the proton gradient that drives ATP synthase (chemiosmotic uncoupling) — a mechanism independent of, and distinct from, the COX-1/prostaglandin pathway that drives both the antiplatelet effect and the iron-status finding above.

Counter-evidence held on record

Every study located that demonstrates uncoupling uses isolated mitochondria, cultured cells, or intact animals exposed to salicylate concentrations at or above the human toxic range; none tested the 81–325 mg/day dosing range this dossier is otherwise scoped to. Framing caution This finding must never be presented as "aspirin depletes a mitochondrial nutrient" or paired with a supplement recommendation such as coenzyme Q10 or carnitine. Mitochondrial uncoupling by salicylate at toxic concentrations is aspirin's own established toxicological mechanism, not a nutrient the drug depletes, and must be labeled as such wherever this entry is surfaced.

Limits on what this shows

• Key negative/null findings: None of the three studies located could be confirmed, from material actually opened, to show either a significant increase or a significant decrease in a named oxidative-stress marker attributable to aspirin alone.
• At-risk subgroups: Not established. The populations studied (coronary bypass patients, patients with major depression) carry independent, disease-driven elevations in oxidative stress that were not excluded as competing explanations by the sources located.

Regulatory status

Not addressed in the label material reviewed.

Population evidence

The APRICOT trial (Gąsecka et al., Platelets, 2025) is a randomized controlled trial comparing aspirin dosage regimens on oxidative stress and platelet reactivity in patients undergoing coronary artery bypass grafting — Tier A by design (drug-specific, human, randomized) — but its full text and result magnitudes could not be independently opened for this review; it is named only, and no statistic from it is asserted here (see Sources; flagged unverifiable). Kuliczkowski et al. (Blood Coagulation & Fibrinolysis, 2016) examined the relationship between high on-aspirin platelet reactivity and oxidative stress in coronary artery bypass patients; likewise named only, full text not independently retrieved. Gałecki et al. (Human Psychopharmacology, 2009) measured oxidative-stress parameters including copper-zinc superoxide dismutase and total antioxidant status in depressive patients, but the aspirin arm was combined with fluoxetine throughout, so the drug-specific contribution cannot be isolated from this source.

Mechanism

Not established in either direction for aspirin specifically at a cited human magnitude. Two theoretically opposed mechanisms are biologically plausible and neither was confirmed with drug-specific human data in this review: COX-1/COX-2 inhibition could plausibly lower inflammation-driven oxidative burden, while any local mitochondrial uncoupling (see entries above, toxic-dose only) would plausibly raise it. Because the dose ranges relevant to each proposed pathway differ, this dossier does not assert a net direction.

Counter-evidence held on record

Confounding by indication: coronary artery bypass and major depression are both independently associated with altered oxidative-stress markers; none of the three studies located isolates aspirin's contribution from the underlying disease or, in the Gałecki study, from the co-administered antidepressant. Framing caution Do not present this as "aspirin increases oxidative stress" or "aspirin is an antioxidant." Both directions are represented among candidate mechanisms and named sources here, and this review could not verify either with drug-specific, unconfounded human data.

Limits on what this shows

• Key negative/null findings: A narrative systematic review (BMC Geriatrics, 2010) of five double-blind, placebo-controlled secondary-prevention RCTs (combined n=5,879; mostly men aged 50–70 with existing cardio/cerebrovascular disease) found no consistent association between low-dose aspirin and change in hemoglobin. The same review's primary-prevention data (two smaller studies, combined n=609, adults ≥70) did show a small but statistically significant Hb fall at one year, consistent with the ASPREE result.
• At-risk subgroups: Age ≥70; long-duration daily use (years, not weeks or months); primary-prevention (rather than established-CVD secondary-prevention) users; women, who start with lower baseline iron stores; anyone on concurrent anticoagulant or additional antiplatelet therapy (a compounding bleeding-risk factor, though this combination was largely excluded from ASPREE itself).

Regulatory status

The U.S. OTC aspirin label does not list a nutrient-depletion warning. It does warn of gastric mucosal irritation, ulceration, and bleeding as labeled adverse effects — a distinct, already-disclosed clinical-outcome warning that must not be re-labeled as this iron-status finding (see framing caution).

Population evidence

ASPREE secondary analysis (McQuilten et al., Annals of Internal Medicine, 2023): randomized, placebo-controlled, n=19,114 community-dwelling older adults, aspirin 100 mg/day. Anemia incidence 51.2 vs 42.9 events per 1,000 person-years (HR 1.20, 95% CI 1.12–1.29). Ferritin <45 µg/L prevalence at year 3: 13% (aspirin) vs 9.8% (placebo); overall ferritin decline 11.5% greater with aspirin (95% CI 9.3–13.7%). Hemoglobin declined an additional 0.6 g/L per 5 years with aspirin (95% CI 0.3–1.0). The authors reported these effects as independent of adjudicated major bleeding events, consistent with a chronic, subclinical (rather than clinically overt) blood-loss mechanism.

Mechanism

Named: COX-1 inhibition suppresses gastroprotective prostaglandins (PGE2/PGI2) in gastric and duodenal mucosa, producing erosions and chronic occult blood loss even at low cardioprotective doses — a direct mucosal-injury mechanism, mechanistically distinct from a pure antiplatelet effect (e.g., clopidogrel, which inhibits platelet aggregation without this direct prostaglandin-suppression/mucosal-injury pathway).

Counter-evidence held on record

Five RCTs in secondary-prevention cardiovascular patients (n=5,879) found no consistent hemoglobin change, a materially different population (younger, established disease, possibly different bleeding-risk and nutritional profiles) — confounding by indication between primary- and secondary-prevention aspirin users is a live, unresolved factor here, not a settled one. Framing caution The consumer-facing surface must never describe this as "aspirin causes GI ulcers" or file the ulcer/erosion itself as the nutrient finding. The labeled adverse effect is GI mucosal injury/bleeding; the nutrient-status finding here is the downstream, separately measured decline in ferritin and hemoglobin that this chronic, often subclinical blood loss produces over years of use. The two are causally linked but must be stated as distinct, sequential facts, and this finding must not be generalized to short-term use, secondary-prevention populations, or higher analgesic/anti-inflammatory dosing without saying so.

Limits on what this shows

• Why no grade: Both conditions for UNGRADED are met. (a) No Tier B/C evidence: a targeted search located exactly one study measuring copper in relation to aspirin/salicylates — a rat preeclampsia model (Cholik et al., Biological Trace Element Research, 2025) that measured iron, zinc, and copper together and reported effects on hair iron, hepcidin suppression, and brain/placental zinc, but did not report a significant finding for copper specifically in the material retrieved; a separate search for aspirin plus copper or ceruloplasmin literature returned no on-topic results at all. (b) No circulating claim: the three consumer drug-nutrient depletion charts already reviewed for this dossier (Coastal Pharmacy and Wellness; InVite Health; Riordan Clinic — Sources [11]–[13]) list folic acid, vitamin C, iron, potassium, and zinc for aspirin, and none lists copper.
• What would fill it: A human or animal study directly measuring serum copper or ceruloplasmin in aspirin-exposed subjects with a stated result, or a circulating consumer claim naming aspirin and copper, would move this out of UNGRADED into either a graded tier or the fails-verification table.

Regulatory status

Not addressed in the label material reviewed.

Population evidence

None located that meets Tier A, B, or C for aspirin and copper status specifically, beyond the null copper result noted above.

Mechanism

None of the nine mechanism classes screened for this dossier (absorption interference, accelerated catabolism, renal wasting, cofactor antagonism, mitochondrial interference, dilutional/hormonal disturbance, microbiome disruption, changed intake, increased utilization) points to a copper-specific pathway for aspirin in any source located.

Limits on what this shows

• Key negative/null findings: No study located, human or animal, directly measures aspirin reducing microbial diversity or depleting a specific taxon in the way consumer nutrient-depletion charts frame drug-nutrient relationships. A proposed colorectal-chemoprevention mechanism involving aspirin and Fusobacterium nucleatum is sometimes discussed in the broader oncology literature, but no aspirin-specific human cohort study connecting aspirin exposure to Fusobacterium or overall microbiome composition was located; this is an evidence gap, not a finding either way.
• At-risk subgroups: Not established for aspirin specifically. If a real drug-to-microbiome effect exists, it would plausibly interact with the subgroups already flagged for the GI-mucosal-injury/iron pathway (long-term, older-adult, primary-prevention users), but this is inference, not a finding, and is labeled as such.

Regulatory status

Not addressed in the label material reviewed.

Population evidence

Tsujimoto et al. (European Journal of Clinical Pharmacology, 2021) followed 32 patients already taking low-dose aspirin (100 mg/day) who were additionally started on one of two proton-pump inhibitors, sampling intestinal microbiota at days 0, 30, 90, and 180; the reported finding was that add-on PPI therapy (more strongly with vonoprazan than esomeprazole) increased the proportion of Lactobacillales, correlated with serum gastrin. Every patient in this study was already on aspirin at baseline, so it evaluates a PPI effect layered on aspirin, not aspirin's own effect against a drug-free comparator. Chen et al. (European Journal of Pharmacology, 2023) examined long-term oral aspirin combined with atorvastatin and human gut microbiota, but aspirin was co-administered with a statin throughout, so the drug-specific contribution cannot be isolated (Tier C for aspirin alone).

Mechanism

Two animal mechanisms were located, and both run in the reverse direction from a "drug depletes microbiome" claim. Zhao et al. (Advanced Science, 2026) identified a gut bacterium, Lysinibacillus sphaericus, producing a carboxylesterase (EstB) that hydrolyzes aspirin locally and worsens aspirin-induced intestinal injury in the animal model used — the bacterium acting on the drug and amplifying its known GI-injury mechanism, not the drug suppressing the bacterium. Kim et al. (Pharmaceutics, 2022) found that coffee intake increased aspirin absorption in mice by modifying the gut microbiome, again describing microbiome-to-drug influence rather than drug-to-microbiome depletion.

Counter-evidence held on record

The only human study located had no aspirin-naive comparator arm (confounding by indication here is compounded by an outright missing comparator, not merely an unbalanced one), and the Chen et al. combination-therapy study cannot separate aspirin's contribution from atorvastatin's. Framing caution Do not present this as "aspirin harms your gut bacteria" or "aspirin causes dysbiosis." The strongest documented relationships in the sources located run from the microbiome toward the drug's own local metabolism and injury pattern, the opposite of a nutrient-style depletion claim, and must be labeled as an interaction running the other direction.

Limits on what this shows

• Key negative/null findings: No human study of ATP production, lactate, or bioenergetic markers at low-dose cardioprotective or standard analgesic aspirin dosing was located; only the overdose/toxicology literature describes this outcome.
• At-risk subgroups: Salicylate overdose/poisoning only, by every source located; no subgroup on therapeutic dosing has been shown to carry this risk in the material reviewed.

Regulatory status

Not addressed in the label material reviewed as a nutrient- or energy-status finding.

Population evidence

Same Tier B source base as mitochondrial burden above (Jørgensen 1976; Gutknecht 1990, 1992). StatPearls' salicylate toxicity reference (NBK499879) links this bioenergetic failure to the clinical picture of overdose, stating that as salicylate "is metabolized, it causes an uncoupling of oxidative phosphorylation in the mitochondria," with rising lactate and worsening metabolic acidosis reported in that same clinical toxicology picture at the mild-to-severe blood-level thresholds noted above.

Mechanism

Named: uncoupled mitochondria continue consuming oxygen but lose proportional ATP yield, forcing a compensatory shift to anaerobic, lactate-producing metabolism. This is a bioenergetic-efficiency effect, not a substrate or cofactor deficiency, and it is downstream of, not separate from, the same proton-shuttling chemistry described in the mitochondrial burden entry.

Counter-evidence held on record

This entry and "Mitochondrial burden" above describe two facets, structural mitochondrial injury and its bioenergetic output, of one underlying toxic-dose mechanism. They are graded and presented separately because they were commissioned as separate categories, but neither should be read as independent corroborating evidence for the other; see the Overlap map below. Framing caution As with mitochondrial burden, this must never be presented as a therapeutic-dose finding or linked to a supplement recommendation. It is a described mechanism of aspirin's own overdose toxicology, not evidence the drug depletes a nutrient involved in ATP synthesis.

Limits on what this shows

• Key negative/null findings: De Furia et al. (1973), above, found no net alteration in oxygen equilibrium in vivo despite confirmed acetylation, directly contradicting a straightforward "aspirin impairs oxygen delivery" reading of the in vitro affinity data.
• At-risk subgroups: Historic interest in this mechanism centers on sickle cell disease, where increased hemoglobin oxygen affinity was investigated as a potential therapeutic effect (reducing sickling) rather than a harm, at doses different from the low-dose cardioprotective context of this dossier; a human sickle-cell trial (Osamo et al., Acta Haematologica, 1981) was named in the search results but not opened in enough detail for this review to state its dosing or outcome, and is not otherwise relied upon here.

Regulatory status

Not addressed in the label material reviewed.

Population evidence

Bridges et al. (Journal of Clinical Investigation, 1975) demonstrated acetylation of hemoglobin by aspirin both in vitro and in vivo using human blood, reporting that purified acetylated hemoglobin showed slightly increased oxygen affinity and decreased heme-heme interaction. Klotz & Tam (Proceedings of the National Academy of Sciences, 1973) and Walder et al. (PNAS, 1977) reported comparable increases in oxygen affinity for acetylated normal and sickle hemoglobin in vitro, work pursued because increased oxygen affinity was hypothesized to be therapeutically useful in sickle cell disease, not identified as an adverse effect. Xu et al. (Biochemical Pharmacology, 2000) confirmed aspirin acetylation at the beta-Lys-82 residue by NMR in a human hemoglobin variant. Against this, De Furia et al. (PNAS, 1973) measured erythrocytes after in vivo aspirin exposure and reported that, despite acetylation occurring, "the reaction does not result in any alteration in the oxygen equilibrium" — a direct functional counter-finding from within the same literature.

Mechanism

Named and direct: aspirin's acetyl group transfers covalently onto lysine residues of hemoglobin (most specifically documented at beta-Lys-82), a direct chemical modification of the protein itself, mechanistically unrelated to any of the classical depletion mechanisms (chelation, malabsorption, renal wasting, enzyme induction) screened for this dossier. No nutrient cofactor, vitamin, or mineral is depleted, consumed, or displaced by this reaction; it modifies the hemoglobin molecule directly.

Counter-evidence held on record

This is aspirin's own direct chemical action on a body protein, not evidence that aspirin depletes a nutrient involved in oxygen transport (e.g., iron, already covered separately in the Orange tier above via a wholly different, GI-blood-loss mechanism). The two must not be conflated: this entry concerns a direct modification of existing hemoglobin, not the iron supply used to build it. Framing caution Do not present this as "aspirin impairs oxygen delivery" as a headline claim; the only in vivo functional measurement located found no significant net effect on the oxygen equilibrium, and the in vitro affinity-increase literature was pursued as a potential therapeutic mechanism in sickle cell disease, not identified as a harm.