Sulindac

arylalkanoic (acetic) acid NSAID, prodrug

Also sold as Clinoril.

Analyze another medication Medication directory

Findings for this medication

Research about this medication, graded for how strongly each finding can be attributed to it. Nothing here describes your own levels.

Sulindac has little classical depletion machinery. It is a prodrug reversibly reduced to the active sulfide and irreversibly oxidized to an inactive sulfone, about 50% of a dose appears in urine and 25% in feces, the label describes no CYP induction or inhibition, and it names no vitamin or mineral [1]. The signals that exist run through the drug's own toxicity and renal handling: GI bleeding (pooled upper GI complication RR 2.89, 95% CI 1.90 to 4.42 versus non-use [5]), rare hemolytic and aplastic anemia, edema with blunted diuretic natriuresis [1][13], and hyperkalemia at under 1% on the label [1]. Per Hard Rule 2 these are outcomes, and per Hard Rule 5 the sodium and potassium findings are retention effects, so supplementation would be the hazard, not the remedy. Hepatic injury and pancreatitis are labeled drug toxicities [1] and are not graded as nutrient findings. Red and orange tiers are empty; 9 entries are yellow and 1 is ungraded.

arylalkanoic (acetic) acid NSAID, prodrug · ATC M01AB02 (WHO ATC/DDD Index [2]) · also sold as Clinoril and Imbaral [3]; many generics · reviewed 2026-10-07. This describes research about the medication. It does not say anything about your own levels or health.

How to read the evidence grades

These colours describe how strongly a finding can be attributed to the medication in the research. They are not a severity score and they say nothing about your own health.

Grades describe how strongly a finding can be attributed to Sulindac in the research — not how common it is, how serious it would be, and not anything about you personally.

Red — Severe: Two or more independent Tier A sources, this drug in the cohort, the nutrient directly measured, consistent direction, clinically meaningful magnitude, a named mechanism, no substantial contradicting Tier A evidence.
Orange — Moderate: At least one Tier A source with this drug in the cohort showing a real but bounded deficit: an exposure window, a subgroup, low frequency, or a functional rather than circulating deficit.
Yellow — Low: Mechanistic plausibility only (Tier B), or class extrapolation (Tier C) with no drug-specific human status data; also used for context flags and interactions running the other direction.
Not graded — disclosed evidence gap: Nothing has been measured here, so this finding cannot be graded. That does not establish either safety or harm, and it is not the same as a White result.Source definition: Not gradeable under the rules above because nothing has been measured. Published as a documented gap, never as reassurance that there is no effect.

What the research found

Potassium: retention risk (direction: excess), mixed sulindac-specific data with a large null cohort

Yellow — Low

See study details

Hyperkalemia is a labeled sulindac adverse reaction at under 1% and was reported in a four-patient case series, but a large nested case-control that included sulindac found no significant hyperkalemia association, and a randomized trial found a smaller potassium rise than with indomethacin.

Full detail and limits

What the grade means here: Mechanistic plausibility only (Tier B), or class extrapolation (Tier C) with no drug-specific human status data; also used for context flags and interactions running the other direction. Yellow

Limits on what this shows

Null findings: Lafrance 2012 [10] was null for sulindac in a very large cohort. The 7-day crossover [14] found no sulindac effect on renal function.
At-risk subgroups: People on renin-angiotensin blockers or receiving contrast media (interactions found in [10]) and people with kidney impairment or prostaglandin-dependent renal function (label warning [1]).

Regulatory status

Label adverse reactions list hyperkalemia under Metabolic at an incidence of under 1 in 100. The label also states that because sulindac is excreted in the urine primarily as biologically inactive forms, it may possibly affect renal function to a lesser extent than other NSAIDs, and that NSAID administration may cause a dose-dependent reduction in prostaglandin formation and, secondarily, in renal blood flow [1].

Population evidence

Lafrance 2012 [10] (Tier A): nested case-control in US Veterans Affairs new NSAID users 2000 to 2006, 18,326 hyperkalemia cases (first potassium of 6.0 mEq/L or more) and 355,106 matched controls; the abstract states elevated risks for rofecoxib, celecoxib, diclofenac and indomethacin "but not meloxicam, etodolac, piroxicam, sulindac, ibuprofen, naproxen, and ketorolac"; interactions were found with renin-angiotensin blockers and contrast media. Sulindac was therefore in the cohort. Individual odds ratios were not in the text read. Nesher 1986 [11] (Tier B, case series): four patients developed hyperkalemia of 6.1 to 6.9 mEq/L within 3 to 8 days of starting sulindac, levels normalized 2 to 4 days after stopping, and no other potassium-affecting drugs were reported. Nesher 1988 [12] (Tier A, randomized, 74 patients, sulindac versus indomethacin): PubMed listing text reports a smaller potassium elevation with sulindac (0.5 versus 0.8 mmol/L); only title, design and sample size were read from OpenAlex. Sulindac 200 mg twice daily for 7 days in 10 patients with arthritis and impaired renal function [14] did not affect renal prostaglandin synthesis or renal function in a placebo-controlled crossover (read through a summarizing fetch; naproxen, not sulindac, raised serum potassium slightly).

Mechanism

Present for the class and not specific to sulindac: reduced renal prostaglandin synthesis lowers renal blood flow [1]. The label text read does not state a potassium mechanism. Sulindac-specific renal sparing is reported in small studies [14] but disputed by the case series [11].

Hard Rule 2, 5 and 6 checks

Hyperkalemia is a retention effect of renal handling, not a dietary excess or deficiency. Confounding by indication is presumed (arthritis patients commonly take RAS blockers and diuretics, and the cohort [10] attributes the risk to concurrent exposures).

Counter-evidence held on record

Counter-evidence and confounding The largest sulindac-containing cohort is null [10], the randomized comparison against indomethacin favors sulindac [12], and the positive evidence is four patients [11] without a comparator. Absence of significance in a veteran population (mostly men) is not proof of no effect in an individual with kidney disease. Framing caution Excess and retention risk. Potassium supplements and potassium-based salt substitutes would be the hazard, not the remedy, in people on sulindac with kidney disease or RAS blockers. This is not a potassium depletion claim and no potassium dose is recommended.

Sodium: retention and blunted diuretic natriuresis (direction: retention, not depletion)

Yellow — Low

See study details

The label lists edema as a more than 1% adverse reaction and states sulindac can reduce the natriuretic effect of furosemide and thiazides; a five-volunteer study found sulindac cut sodium excretion by about half over four hours, while a placebo-controlled study in impaired renal function found no overall change in sodium excretion.

Full detail and limits

Limits on what this shows

Null findings: The impaired-renal-function crossover found no overall sodium excretion change [14]. A PubMed listing for sulindac with hyponatremia, hyperkalemia or sodium retention returned 10 records and none reported sulindac-associated hyponatremia [29].
At-risk subgroups: People with heart failure or edema, people on diuretics (label interaction), and people with kidney impairment [1].

Regulatory status

Label: fluid retention and edema have been observed in some NSAID-treated patients; edema is listed under adverse reactions at more than 1%; "Sulindac tablets can reduce the natriuretic effect of furosemide and thiazides" [1]. No hyponatremia statement was found in the label text read.

Population evidence

Dixey 1987 [13] (Tier A, small): five volunteers given single doses of sulindac 200 mg or naproxen 500 mg, alone and with hydrochlorothiazide or piretanide; both NSAIDs reduced sodium excretion by about 50% over four hours, and on top of chronic diuretic therapy both blunted the natriuretic response by roughly 30% (OpenAlex abstract field was null; figures come from a summarizing fetch of the indexed text and were not checked against a full abstract). Sulindac versus naproxen crossover in 10 arthritis patients with stable impaired renal function [14] (7 days, placebo-controlled): neither drug significantly affected sodium or potassium excretion overall, and sulindac lowered mean arterial pressure (summarizing fetch of a bibliographic record).

Mechanism

Present for the class: prostaglandin inhibition reduces renal perfusion and tubular sodium excretion [1]. The two sulindac studies disagree on the size of the effect [13][14].

Hard Rule 2, 5 and 6 checks

Sodium retention and edema are renal handling outcomes, not a dietary sodium deficiency. Confounding by indication is presumed in heart failure populations that already retain sodium.

Counter-evidence held on record

Counter-evidence and confounding The sodium-excretion evidence is five volunteers on single doses [13] against a null in 10 patients [14]; neither measured serum sodium. No sulindac hyponatremia report was found. Framing caution Retention and dilution risk, not a sodium deficiency. This entry must never be read as advice to raise salt intake; any salt or fluid change belongs to the prescriber.

Folate: in vitro inhibition of folate-dependent enzymes by sulindac, no folate status measured in sulindac users

Yellow — Low

See study details

Sulindac was one of several NSAIDs that competitively inhibited folate-dependent enzymes in a 1992 in vitro study, and consumer pages claim sulindac depletes folate, but no study measured folate or homocysteine in people taking sulindac.

Full detail and limits

Limits on what this shows

Null findings: No folate, red-cell folate or homocysteine study in sulindac users was found across the channels listed in the Sources access note. Circulating claims are uncited (below).

Regulatory status

No folate statement in the label text read. The label warns that NSAIDs competitively inhibit methotrexate accumulation and could enhance methotrexate toxicity [1]; that is a drug-drug toxicity interaction (Hard Rule 4), not folate depletion by sulindac.

Population evidence

Baggott 1992 [21] (Tier B, in vitro plus small clinical correlation; sulindac in the tested set): the abstract states that many NSAIDs "including sulindac, indomethacin, naproxen, salicylic acid, ibuprofen, piroxicam and mefenamic acid" were competitive inhibitors of avian liver AICAR transformylase and bovine liver dihydrofolate reductase, that an aromatic ring with a carboxylic acid side chain was needed for transformylase inhibition, that the NSAIDs suppressed folate-dependent serine synthesis in human blood mononuclear cells, and that results in rheumatoid arthritis patients "treated with these drugs" correlated with antifolate potency. Inhibition constants and sulindac-specific potency were not in the text read, and whether the parent drug or the sulfide was tested was not stated. PubMed listing "sulindac folate OR sulindac homocysteine": 15 records, 10 displayed, all chemoprevention reviews and none measured folate or homocysteine in sulindac users [29].

Mechanism

Present in vitro for sulindac [21]. Whether tissue exposure reaches inhibitory levels, and whether any functional folate deficit follows, is not shown. The structure-activity statement in [21] was not evaluated for sulindac here.

Class cohort test

Passed for sulindac: sulindac is named in the tested set of [21], so the source is Tier B for this drug, not class extrapolation. The rheumatoid arthritis correlation in [21] is not sulindac-specific and is a class-level finding.

Hard Rule 6 check

Confounding by indication is presumed: patients with rheumatoid disease may be on methotrexate (a label-listed interaction [1]), and inflammation can affect folate and homocysteine independently of the NSAID (reasoning, not a retrieved finding).

Counter-evidence held on record

Counter-evidence and confounding The inhibition is shown in enzyme and cell preparations, not as a lower folate level in people. Circulating claims: a consumer page states "Taking sulindac may deplete vitamin B9" with references that are drug monographs, not folate measurements [26] (Tier D), and a supplement-retailer chart lists "Ibuprofen, Naproxen, Sulindac" under folate with no references [27] (Tier D). Neither claim supplies Tier A or B evidence, so they do not raise the grade. Framing caution Theoretical signal only. Not a folate deficiency claim for sulindac and no folic acid supplementation is recommended here. People also taking methotrexate should take that interaction to their prescriber.

The five body-system areas

Each finding is counted once, in a single area. Where a study is mentioned in a second area it is context only.

Mitochondrial Burden

Each finding below is counted once, in this area only.

Mitochondrial burden: sulindac sulfide uncoupling in isolated rat liver mitochondria and cancer-cell mitochondrial effects, concentration relevance unestablished

Yellow — Low

See study details

The active metabolite sulindac sulfide, but not sulindac or sulindac sulfone, uncoupled isolated rat liver mitochondria and released stored calcium in vitro, and sulindac plus an oxidant collapsed mitochondrial membrane potential in cancer cells but not normal cells; no human mitochondrial measurement exists.

Full detail and limits

Limits on what this shows

Null and reverse findings: Sulindac and the sulfone did not uncouple in [17]. A 2024 retinal cell study (listing text only, not opened) reports sulindac suppressing VDAC1 expression and preventing mitochondrial DNA leakage and ATP loss in a diabetic retinopathy model [29], a reverse direction in a disease model.

Regulatory status

No mitochondrial statement in the label text read. Label hepatic warnings (borderline liver test elevations in up to 15% of patients; rare severe hepatic reactions including fulminant hepatitis, liver necrosis and failure) are drug toxicities and are not graded as mitochondrial findings [1].

Population evidence

None in humans. Leite 2006 [17] (Tier B, in vitro): isolated rat liver mitochondria and HepG2 cells; "sulindac sulfide, but not sulindac sulfone or sulindac itself, caused mitochondrial uncoupling, released preaccumulated calcium from the organelle, and decreased cell viability" in association with ATP depletion from membrane potential dissipation (abstract via OpenAlex summarizing fetch; concentrations not in the text read). Marchetti 2009 [18] (Tier B, in vitro): sulindac 500 micromolar for pretreatment of human lung (A549) and colon (RKO) cancer cells followed by an oxidant produced a large rise in reactive oxygen species and loss of mitochondrial membrane potential (JC-1 dye) with apoptosis, and this was not seen in normal lung or colon fibroblasts; sulindac sulfone was tested at 250 micromolar, sulindac sulfide was not tested. Nadanaciva 2013 [19] (Tier B): sulindac and its sulfide and sulfone metabolites were among 11 NSAIDs tested in isolated mitochondria, rat hepatocytes and zebrafish; the abstract gives no sulindac-specific result and full text was not opened. Listing-level titles (not opened) describe sulindac sulfide and phospho-sulindac (a derivative, excluded) causing mitochondrial dysfunction in cancer cells [29].

Mechanism

Named and demonstrated in vitro: protonophore-like uncoupling by the sulfide with calcium release [17]. The sulfide is the active metabolite [1], so the finding is relevant to what the body forms, but whether free sulfide concentrations in plasma, liver or colon mucosa reach uncoupling levels is not shown; the label reports 93.1, 95.4 and 97.9% protein binding for sulindac and metabolites [1] and a human study found sulfide accumulating in colonic tissue [23].

Counter-evidence held on record

Counter-evidence and confounding Isolated-mitochondria uncoupling is a screening-level finding; the Marchetti concentration (500 micromolar) was not compared with plasma exposure in the text read and the effect required an added oxidant [18]. The hepatic injury signal on the label is a drug toxicity outcome and is not evidence of a mitochondrial nutrient deficit. Framing caution Mechanistic plausibility only. This entry does not support any mitochondrial or antioxidant supplement and is not a deficiency claim.

Oxidative Stress

Each finding below is counted once, in this area only.

Oxidative stress: mixed direction (pro-oxidant effect in cancer cells, antioxidant effect in neurons), no human redox marker study

Yellow — Low

See study details

In vitro, sulindac raised reactive oxygen species and killed cancer cells exposed to an oxidant while sparing normal cells, and in a separate neuron model it lowered valproate-induced oxidative stress; no human antioxidant-capacity or lipid-peroxidation measurement in sulindac users was found.

Full detail and limits

Limits on what this shows

Null and reverse findings: Reverse direction: antioxidant effect in neurons [20]. Normal cells were not sensitized in [18]. No erythrocyte superoxide dismutase, glutathione or malondialdehyde study in sulindac users was found in the listing "sulindac oxidative stress" (10 records, none a human redox marker study) [29].

Regulatory status

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

Population evidence

None in humans. Marchetti 2009 [18] (Tier B, in vitro; PLoS ONE): sulindac pretreatment at 500 micromolar plus an oxidant (tert-butyl hydroperoxide or hydrogen peroxide) gave a roughly 400% increase in intracellular ROS fluorescence in cancer cells versus individual treatments of 53 to 58%, and the abstract states the effect "does not involve cyclooxygenase (COX) inhibition" and was not seen in normal cells; the paper also reports a small uncontrolled topical sulindac with hydrogen peroxide gel proof-of-concept in ten actinic keratoses (not a systemic redox measurement). Zhang 2015 [20] (Tier B, in vitro and rat; Int J Mol Med): sulindac reduced valproic acid-induced oxidative stress and improved viability in cultured cortical neurons, and treated rat offspring had fewer repetitive movements (abstract via PubMed listing summarizing fetch; markers and concentrations not read). Listing-level titles (not opened) include a 2010 rat study in which oxidative stress contributed to liver injury when sulindac was combined with lipopolysaccharide, and a rat liver gene-expression signature that flagged sulindac as an oxidative-stress or reactive-metabolite compound [29].

Mechanism

Both directions are chemically plausible and reported in vitro. Which dominates in vivo at therapeutic exposure is unresolved. A COX-independent ROS mechanism is stated for the cancer-cell effect [18].

Why not orange

Orange needs a bounded real deficit in a measured nutrient or marker attributable to the drug in a Tier A source. All retrieved evidence is in vitro or animal and runs in both directions.

Counter-evidence held on record

Counter-evidence and confounding Cell-line concentrations are high, models are disease-specific, and the cancer-cell sensitization used an added oxidant [18]. Framing caution Do not describe sulindac as depleting antioxidants and do not describe it as an antioxidant therapy. No antioxidant supplement is implied.

Copper and Iron Balance

Copper and iron are assessed separately. Nothing here is counted twice.

Iron: GI blood loss pathway (drug toxicity, direction: loss), no iron status measured in sulindac users

Yellow — Low

See study details

Sulindac carries the NSAID class warning for GI bleeding and anemia and has a pooled upper GI complication risk of about 2.9 times non-use, but a controlled fecal blood loss trial found no difference from placebo and no sulindac cohort measured ferritin or iron saturation.

Full detail and limits

Limits on what this shows

Null findings: Cohen 1976 [4] was null for fecal blood loss, but it is a 2-week study in healthy young men and does not test older arthritis patients on long-term therapy. Listing-level titles (not opened) show an RCT of sulindac sulfoxide and sulfide on gastric mucosa described as not distinguishable from placebo, and a 1987 retrospective cohort in which sulindac users had the highest upper GI bleeding rate, so the older literature points in both directions [29]. Moore 2008 [6], the systematic review of 51-chromium blood loss trials (47 trials, 1,162 people), did not mention sulindac in the text read, so it is Tier C context only and is not used to support or weaken the sulindac estimate.
At-risk subgroups: The label text read does not enumerate GI risk groups. Listing-level evidence points to pre-existing GI disorders (a cohort reports a much higher odds ratio for adverse reactions in that group; not opened) [29]. Aspirin co-use is a label-listed interaction (aspirin lowers sulfide levels and the combination is not recommended), and the label states studies showed no clinically significant interaction with oral anticoagulants [1].

Regulatory status

Label: NSAIDs increase the risk of serious GI events including bleeding, ulceration and perforation; upper GI ulcers, gross bleeding or perforation occur in approximately 1% of patients treated for 3 to 6 months. Under Hematological Effects the label states anemia is sometimes seen with NSAIDs and that hemoglobin or hematocrit should be checked if signs or symptoms of anemia appear; NSAIDs also inhibit platelet aggregation and can prolong bleeding time. Anemia is listed under adverse reactions at under 1%, with no percentage. Iron is not mentioned [1].

Population evidence

Castellsague 2012 [5] (Tier A, outcome): meta-analysis of 28 observational studies; sulindac pooled RR 2.89 (95% CI 1.90 to 4.42) for upper GI complications versus non-use, in the moderate-risk range of the ranking (abstract level; the outcome is a complication, not iron status). Cohen 1976 [4] (Tier A, randomized, direct blood loss measurement): 40 healthy men, red cells labeled with 51-chromium, sulindac 240 mg or 400 mg daily versus aspirin 4.8 g versus placebo for 2 weeks; aspirin blood loss exceeded both sulindac groups and there was no significant difference between either sulindac group and placebo. Randomized FAP and polyp trials that gave sulindac for months to years reported the following at abstract level: Giardiello 1993 [7] (22 patients, 150 mg twice daily for 9 months) stated no adverse effects from sulindac were noted; Giardiello 2002 [8] (41 young subjects, 75 or 150 mg twice daily for 48 months) evaluated side effects but the abstract gives no hemoglobin or GI bleeding figures; Ladenheim 1995 [9] (sporadic colonic polyps) listed one anemia dropout and two heartburn dropouts among four dropouts from the sulindac group (read through a summarizing fetch of OpenAlex text; abstract field null).

Mechanism

Present. The active sulfide metabolite inhibits COX and prostaglandin synthesis [1], and NSAID mucosal injury of stomach and small bowel can bleed. The label does not state an iron mechanism. Iron loss itself (ferritin, transferrin saturation, serum iron) was not measured in any retrieved sulindac study; a PubMed listing for "sulindac iron deficiency OR sulindac anemia" returned 10 records, none an iron-indices study [29].

Hard Rule 2 and 6 checks

GI bleeding and anemia are clinical outcomes, not iron deficiency. Confounding by indication is presumed: rheumatoid disease and other chronic inflammation lower hemoglobin independent of the drug, and in FAP or polyp populations the colonic lesions themselves can bleed (reasoning, not a retrieved finding). No before-and-after iron data exist to test the drug as cause.

Counter-evidence held on record

Counter-evidence and confounding Controlled blood loss was null [4], the FAP randomized trials reported no major adverse event signal at abstract level [7][8], and the case-level hematologic reports for sulindac are immune hemolysis and marrow aplasia (see Oxygen delivery/utilization), not iron-deficiency anemia. Circulating claim: a consumer page states "Taking nonsteroidal anti-inflammatory drugs may deplete iron" with no sulindac-specific citation [26] (Tier D). Framing caution This is a toxicity pathway (blood loss), not an intake or absorption deficiency. New iron deficiency or dark stools on sulindac calls for evaluating the bleeding source and the prescription, not an iron supplement by default. No iron dose is recommended here.

Copper (zinc noted): no copper status data for sulindac in either direction

Not graded — disclosed evidence gap

See study details

No study measured copper, ceruloplasmin or zinc status in sulindac users, and the only retrieved record is a 1989 test-tube enzyme study that uses copper and zinc ions as reagents with the sulfide metabolite; it was not opened beyond its title.

Full detail, limits and 1 references

What the grade means here: Nothing has been measured here, so this finding cannot be graded. That does not establish either safety or harm, and it is not the same as a White result. Ungraded

Limits on what this shows

Why no grade: Both conditions of the Step 7 dual test were applied. (1) No Tier B or C evidence supports even a yellow grade for a drug effect on copper: the only retrieved record, Cook and Ternai 1989 [25], is titled "Effect of Cu2+ and Zn2+ on the inhibition of human leucocyte elastase by 6-alkyl-3-(omega-carboxyalkyl)-2-pyrones, oleic acid and sulindac sulfide". Its abstract field was null in OpenAlex and a PubMed single-record fetch was rate-limited, so only the title and metadata were seen. From the title alone the metal ions are reagents in an enzyme inhibition assay, which would be a metal-to-drug interaction in a test tube and not a drug effect on copper status; the direction and size of its result were not read and are not used. (2) No circulating consumer claim for copper or zinc and sulindac was found: the consumer page [26], the retailer chart [27] and the Therapeutic Research chart [28] do not name copper or zinc for sulindac (the chart [28] has no sulindac row at all).

Channels tried

PubMed listing "sulindac copper OR sulindac zinc" (10 records, opened; none a copper or zinc status study: zinc-finger protein, copper-catalysed synthesis, gene-expression and assay-reagent hits only); WebSearch for sulindac copper or zinc complexes (NSAID coordination-chemistry hits) and for sulindac with serum copper or zinc in rheumatoid arthritis (generic rheumatoid arthritis copper and zinc studies, none sulindac-specific), then two follow-up queries on sulindac copper(II) crystal structures and on sulindac with ceruloplasmin or copper absorption were refused because the shared WebSearch budget was used up; OpenAlex single-record lookup for PMID 2713094 (abstract null); OpenAlex and Crossref search-list queries for sulindac and copper (HTTP 429, not retried on the same page); a 2022 review of NSAID coordination compounds (Santos et al., Int J Mol Sci, DOI 10.3390/ijms23052855; Crossref record read, PMC full text blocked by a bot check) whose retrieved record named only a cobalt(II)-sulindac complex reference.

Genuine gap or tooling-limited

Both. No human or animal copper measurement with sulindac surfaced in the listing that opened (a genuine-gap signal), but full text of the coordination-chemistry literature and two search channels were blocked, so a sulindac copper or zinc complex study or a trace-element study could exist unseen. Re-check when search access returns.

Counter-evidence held on record

Hard Rule 7 note Copper is kept as its own entry and is not merged with Iron. This ungraded status is not a finding of safety or of no effect.

Cook L, Ternai B. Effect of Cu2+ and Zn2+ on the inhibition of human leucocyte elastase by 6-alkyl-3-(omega-carboxyalkyl)-2-pyrones, oleic acid and sulindac sulfide. Biol Chem Hoppe Seyler 1989. — In vitro enzyme study. PMID 2713094. DOI 10.1515/bchm3.1989.370.1.11. Seen as a PubMed listing row and an OpenAlex record whose abstract field was null.Abstract and full text not opened; only the title and metadata were seen. Not used as evidence for any grade.

Gut Microbiome Risk

Each finding below is counted once, in this area only.

Gut microbiome risk: mouse data on sulindac and fecal microbes (Tier B) plus an unconfirmed NSAID class signal, with the human direction running from microbiota to the drug

Yellow — Low

See study details

In obese azoxymethane-treated mice, sulindac lowered three fecal taxa linked to tumor burden, but no human study of sulindac effects on gut microbial composition was found; the human data show gut bacteria acting on sulindac (forming the active sulfide in the colon), not sulindac acting on bacteria.

Full detail and limits

Limits on what this shows

Null findings: Mouse studies of sulindac in Apc and Mlh1 mutant models [30] and in a carcinogen colon model [31] reported tumor and inflammation outcomes without a microbiota readout (abstract pages opened), so they do not test this category. No human sulindac stool-microbiome study was found: PubMed listing "sulindac microbiota" returned 8 records, none a human sulindac composition study [29].

Regulatory status

No microbiome statement in the label text read. The label describes extensive enterohepatic circulation of sulindac and its sulfone [1].

Population evidence

Bowers 2022 [22] (Tier B, mouse; Cancer Prev Res): azoxymethane-treated male FVB/N mice on a diet with sulindac at 140 ppm; the abstract states "the fecal microbes Coprobacillus, Prevotella, and Akkermansia muciniphila were positively correlated with tumor multiplicity and reduced by sulindac in obese mice" (abstract via a summarizing fetch of a university record; diversity metrics not read). Lemmens 2021 [23] (Tier A for pharmacokinetics, direction microbiota to drug): healthy volunteers given sulindac 200 mg tablets with colonoscopies; sulindac reached the colonic lumen through incomplete absorption and biliary excretion, and "microbiota catalyzed sulindac sulfide production," which accumulated in colonic tissue. CLASS EXTRAPOLATION, Tier C: Le Bastard 2018 [24] (systematic review of non-antibiotic drugs; abstract via OpenAlex summarizing fetch) reported no significant alpha diversity change for NSAIDs; it does not name sulindac, so sulindac in the underlying cohorts is unconfirmed. The retrieval gave a beta-diversity statement that is not relied on.

Mechanism

Plausible in principle through local colonic exposure to the sulfide [23] and NSAID mucosal injury [1]; not demonstrated for sulindac in humans.

Class cohort test

Fails for direct support: sulindac is not named in [24], so that source is Tier C, capped at yellow, and is context only. Bowers [22] is sulindac-specific but a mouse obesity and carcinogen model.

Counter-evidence held on record

Counter-evidence and confounding The mouse effect is in a tumor model where taxa track tumor burden, so drug effect and disease effect are not separated, and the study measured correlation with tumor multiplicity [22]. Human sulfide formation by bacteria is a drug-metabolism finding, not a dysbiosis finding [23]. Framing caution Not a dysbiosis claim for sulindac. No probiotic or other supplement is implied.

Oxygen and ATP Output

Oxygen delivery and ATP production are assessed separately. Nothing here is counted twice.

Oxygen delivery/utilization: rare immune hemolytic anemia and marrow aplasia as drug-toxicity outcomes, plus in vitro red-cell death

Yellow — Low

See study details

Hemolytic anemia and aplastic anemia are labeled rare sulindac reactions with published case reports (an immune, drug-dependent-antibody mechanism in one), and sulindac sulfide triggered red-cell death in vitro; none of this is a nutrient deficiency or a measure of oxygen-carrying function.

Full detail and limits

Limits on what this shows

Null findings: A PubMed listing for sulindac with methemoglobin or hemolysis returned 7 records and none was a methemoglobin study [29]. A listing-level photohemolysis assay reported no sulindac photoactivity (title and summary only) [29].

Regulatory status

Label adverse reactions under 1 in 100: "Thrombocytopenia; ecchymosis; purpura; leukopenia; agranulocytosis; neutropenia; bone marrow depression, including aplastic anemia; hemolytic anemia" [1]. Precaution text on checking hemoglobin or hematocrit if anemia signs appear is as in the Iron entry [1].

Population evidence

Angeles 1994 [15] (Tier B, case report): immune hemolytic anemia with kidney and liver failure after several days of sulindac; direct antiglobulin test positive for IgG, and a sulindac-dependent IgG and IgM antibody was demonstrated, compatible with an immune-complex mechanism. PubMed listing titles (not opened) report further sulindac hemolytic anemia and marrow aplasia case reports from 1980 to 1986 [29]. Zbidah 2012 [16] (Tier B, in vitro): 48-hour exposure of red cells to sulindac sulfide at up to 20 micromolar raised intracellular calcium, ceramide, cell shrinkage and phosphatidylserine exposure with minor hemolysis (abstract via OpenAlex summarizing fetch; species and a comparison with plasma concentrations not stated in the text read). No hemoglobin, hematocrit, red-cell mass, methemoglobin or arterial oxygenation study in sulindac users with a control group was retrieved.

Mechanism

For the immune cases, a drug-dependent antibody [15]. For the in vitro finding, calcium entry and ceramide-mediated eryptosis [16]. Neither describes a nutrient pathway.

Hard Rule 2 check

Anemia and hemolysis are outcomes, not iron, copper or B-vitamin deficiencies, and a drug-induced hematologic reaction is not an oxygen-delivery nutrient finding. Any blood-loss anemia is handled in the Iron entry and is not counted twice.

Counter-evidence held on record

Counter-evidence and confounding The reactions are rare (under 1%) and idiosyncratic, the case reports are single patients, with multi-drug exposure in at least one listing-level report [29], and the in vitro concentration is not mapped to the label-reported 93 to 98% protein binding of sulindac and its metabolites [1]. Framing caution A monitoring prompt (complete blood count if new pallor, jaundice, dark urine or fatigue), not a deficiency claim. No supplement is implied and none would treat an immune hemolytic reaction.

ATP production: ATP depletion with sulindac sulfide in HepG2 cells, same source as Mitochondrial burden

Yellow — Low

See study details

The only energy-output evidence is the same in vitro work that places the sulfide metabolite among mitochondrial uncouplers, including ATP depletion in HepG2 cells; no study measured ATP or energy status in people taking sulindac.

Full detail and limits

Limits on what this shows

Null and reverse findings: No study that measured intact-tissue ATP in sulindac-treated animals or people was found. A 2024 listing-level report (not opened) describes sulindac preventing ATP loss in a retinal disease model [29].

Regulatory status

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

Population evidence

None in humans. Leite 2006 [17] (Tier B, in vitro): sulindac sulfide decreased HepG2 cell viability in association with ATP depletion resulting from uncoupling-associated membrane potential dissipation; sulindac and sulindac sulfone did not uncouple. Nadanaciva 2013 [19]: sulindac and its metabolites were in the panel; sulindac-specific ATP or respiration results were not read.

Mechanism

Uncoupling collapses the proton gradient that drives ATP synthase, so ATP yield per oxygen consumed falls in the test system [17]. ATP levels in intact human tissue were not shown.

Counter-evidence held on record

Counter-evidence and confounding Cells compensate for modest uncoupling, protein binding lowers free drug [1], and the evidence is not independent of the Mitochondrial burden entry (see Overlap map). Framing caution Do not translate this into a fatigue or low-energy claim for sulindac users. No supplement is implied.

Claims checked and not supported

These circulate online for this medication. They are listed here only so they can be answered, never as findings, and they are not graded.

A claim with no supporting evidence is not necessarily disproven. Where evidence actually argues against a claim, that is stated below.

Sodium depletion by sulindac ("NSAIDs may deplete sodium")
Where it comes from: Lima Memorial Health System consumer library page, Tier D, which states NSAIDs other than diflunisal may deplete sodium and lists no sulindac-specific evidence [26].
Why it fails: Contradicted by the direction of the drug's own data: the label lists edema and reduced natriuretic effect of diuretics [1], and the only sulindac sodium-excretion study found a roughly 50% reduction in sodium excretion over four hours [13]. The signal is retention, not loss, and no sulindac hyponatremia report was found [29]. The retention findings are graded separately in the Sodium entry with a framing caution that salt supplementation would be the wrong response.

How the body handles this medication

Prodrug and metabolism: Sulindac undergoes reversible reduction to the sulfide metabolite and irreversible oxidation to the sulfone metabolite; the label states the activity resides in the sulfide. The label text read gives no cytochrome P450 statement and does not describe enzyme induction or inhibition.[1] DailyMed label, clinical pharmacology (read via summarizing fetch)
Enterohepatic circulation and colon exposure: Sulindac and its sulfone undergo extensive enterohepatic circulation relative to the sulfide. In healthy volunteers given 200 mg tablets, sulindac reached the colonic lumen through incomplete absorption and biliary excretion, and gut microbiota catalyzed sulfide formation that accumulated in colonic tissue.[1] label; [23] Lemmens 2021 (abstract)
Elimination and half-lives: About 50% of the dose is excreted in urine and about 25% in feces. Mean effective half-lives are 7.8 hours (sulindac) and 16.4 hours (active sulfide). Absorption from tablets is similar in extent to solution; no named absorption transporter appeared in the text read.[1] label, clinical pharmacology
Protein binding: Sulindac and its metabolites are 93.1, 95.4 and 97.9% bound to plasma proteins, predominantly albumin, which may lower free drug available for the in vitro effects described in the mitochondrial entries (inference, not a measured value).[1] label, clinical pharmacology
Renal and hepatic handling: Sulfide concentrations were significantly reduced in end-stage renal disease; hepatic metabolism is an important elimination pathway and patients with liver disease may need reduced doses. The label says sulindac is excreted primarily as inactive forms and may possibly affect renal function less than other NSAIDs, while also warning of prostaglandin-dependent renal blood flow reduction, renal papillary necrosis with long-term NSAID use and edema.[1] label, clinical pharmacology, warnings (renal effects)
Interactions with direction: Aspirin depressed plasma sulfide levels (combination not recommended); DMSO is not to be used with sulindac (reported peripheral neuropathy); sulindac can reduce the natriuretic effect of furosemide and thiazides; NSAIDs raise lithium levels, can enhance methotrexate toxicity, and warrant renal monitoring with cyclosporine; sulindac modestly reduced the uricosuric action of probenecid. No iron, calcium, zinc, folate or other mineral or vitamin interaction was listed in the text read. Pancreatitis has been reported; the label says to stop the drug if suspected.[1] label, precautions and warnings

What the research says is worth watching

General information from the medication label and the studies above. It is not personal medical advice, and any change belongs with your prescriber.

Lab / label
Hemoglobin and hematocrit
The label directs a hemoglobin or hematocrit check if signs or symptoms of anemia appear on long-term NSAID treatment [1]. Add ferritin and transferrin saturation if anemia is found, because anemia alone does not show iron deficiency. Unexplained pallor, jaundice or dark urine merits a full blood count promptly (rare hemolytic and aplastic reactions [1][15]).
Symptoms
GI bleeding signs
Black or tarry stool, blood in the stool, epigastric pain or vomiting blood should be reported promptly [1]. Highest concern with aspirin or anticoagulant co-use or prior GI bleeding [1][29].
Lab / renal
Potassium and creatinine
Worth checking potassium and kidney function when sulindac is combined with an ACE inhibitor, ARB, potassium-sparing diuretic, or in chronic kidney disease or dehydration [1][10][11]. Retention is the risk; potassium supplements and potassium-based salt substitutes are the hazard.
Lab / clinical
Sodium, blood pressure, edema
Track weight gain, swelling and blood pressure, particularly in heart failure, cirrhosis or on diuretics, since sulindac can blunt the effect of furosemide and thiazides [1][13]. This is a water-handling signal and is not a reason to add salt.
Lab / liver
Liver tests and pancreatitis symptoms
The label reports borderline liver test elevations in up to 15% of patients and rare severe hepatic reactions, and warns that pancreatitis has been reported and the drug should be stopped if it is suspected [1]. These are drug toxicities, not nutrient deficiencies, and are listed here only because they alter the monitoring plan.
Timing
Aspirin, DMSO and supplement separation
Aspirin lowered sulfide levels and the label does not recommend the combination [1]; the text read gave no separation interval, so timing is a pharmacist matter. No vitamin or mineral supplement interaction is listed. No supplement dose is recommended or appears anywhere in this dossier, so no upper-limit comparison was needed.
Drug interactions
Methotrexate, lithium, cyclosporine, probenecid
The label flags enhanced methotrexate toxicity, raised lithium levels, renal monitoring with cyclosporine and a modest reduction in probenecid's uricosuric action [1]. These are drug-drug interactions, not nutrient depletions.
Special populations
Pregnancy, renal impairment, older adults, children
Use at about 20 weeks of gestation or later has been associated with fetal renal dysfunction and oligohydramnios [1]. Sulfide levels fall in end-stage renal disease and hepatic impairment may need dose reduction [1]. Older adults were not separately reviewed. Pediatric dosing was not reviewed here, although the 2002 FAP trial enrolled ages 8 to 25 [8].

Overlap map, to prevent double counting

Mitochondrial burden
Shares evidence with: ATP production
Note: Both rest on Leite 2006 [17] (sulfide uncoupling and ATP depletion in HepG2) and the Nadanaciva panel [19]. One body of in vitro evidence viewed as two endpoints, not two independent lines; Hard Rule 7 keeps the entries separate, not independent.
Mitochondrial burden
Shares evidence with: Oxidative stress
Note: Marchetti 2009 [18] is cited in both (membrane potential loss in one, ROS in the other). It is a single in vitro paper; the ROS rise and the membrane potential loss are one experiment.
Iron
Shares evidence with: Oxygen delivery/utilization
Note: Both cite the label hematological effects sentence on anemia [1]. The Iron entry uses it for blood loss anemia and the Oxygen entry uses the label's hemolytic and aplastic anemia list; any blood loss anemia is mediated through the Iron entry and is not counted twice. The Ladenheim anemia dropout [9] is used only in Iron.
Sodium
Shares evidence with: Potassium
Note: Same label renal prostaglandin framing [1] and the same 10-patient impaired-renal-function crossover [14]. The sulindac-containing cohort [10] and the case series and randomized comparison [11][12] bear on potassium only; the diuretic interaction study [13] bears on sodium only.
Gut microbiome risk
Shares evidence with: Mitochondrial burden
Note: No shared source. The colonic sulfide accumulation finding [23] is cited in the Microbiome entry (microbiota acting on the drug) and as one exposure-context sentence in Mitochondrial burden; it is a pharmacokinetic result, not independent mitochondrial confirmation.
Folate
Shares evidence with: No other entry
Note: The label methotrexate interaction [1] is a drug-drug toxicity interaction, not a folate depletion signal, and is recorded in the Folate entry only. The circulating claims [26][27] are shared with Iron and Sodium only as claim origins.
Copper
Shares evidence with: Iron
Note: No shared source. Copper is ungraded and the one record seen [25] is not used as evidence. Graded separately per Hard Rule 7.

All references used above

Sulindac

Findings for this medication

What the research found

Potassium: retention risk (direction: excess), mixed sulindac-specific data with a large null cohort

Sodium: retention and blunted diuretic natriuresis (direction: retention, not depletion)

Folate: in vitro inhibition of folate-dependent enzymes by sulindac, no folate status measured in sulindac users

The five body-system areas

⚡Mitochondrial Burden

Mitochondrial burden: sulindac sulfide uncoupling in isolated rat liver mitochondria and cancer-cell mitochondrial effects, concentration relevance unestablished

💥Oxidative Stress

Oxidative stress: mixed direction (pro-oxidant effect in cancer cells, antioxidant effect in neurons), no human redox marker study

⚖️Copper and Iron Balance

Iron: GI blood loss pathway (drug toxicity, direction: loss), no iron status measured in sulindac users

Copper (zinc noted): no copper status data for sulindac in either direction

🌿Gut Microbiome Risk

Gut microbiome risk: mouse data on sulindac and fecal microbes (Tier B) plus an unconfirmed NSAID class signal, with the human direction running from microbiota to the drug

💨Oxygen and ATP Output

Oxygen delivery/utilization: rare immune hemolytic anemia and marrow aplasia as drug-toxicity outcomes, plus in vitro red-cell death

ATP production: ATP depletion with sulindac sulfide in HepG2 cells, same source as Mitochondrial burden

Claims checked and not supported

Sodium depletion by sulindac ("NSAIDs may deplete sodium")

How the body handles this medication

What the research says is worth watching

Overlap map, to prevent double counting

Mitochondrial Burden

Oxidative Stress

Copper and Iron Balance

Gut Microbiome Risk

Oxygen and ATP Output