Methadone

Limits on what this shows

• Null findings: Low-dose methadone (10 to 60 mg/day) left testosterone in the normal range in the 1975 study [30]. Women showed no significant suppression in [27]. A 2025 bone study found that the association of methadone with lower lumbar spine T-scores became non-significant after adjustment for estradiol and free testosterone, which supports a hormone-mediated pathway rather than a vitamin D one [23].
• At-risk subgroups: Men on higher methadone doses (the dose-response in [27] and [30]); men with other causes of low testosterone (obesity, age, comorbidity). Women were not suppressed in [27].

Direction

Hormone suppression (endocrine effect). Hard Rule 2: low testosterone is the drug's own pharmacologic effect, not a nutrient deficiency, and no intake change is implied.

Regulatory status

Label section 6 (read via a summarizing tool on a DailyMed mirror): "Cases of androgen deficiency have occurred with use of opioids for an extended period of time" [2]. Label section 12.2 (medlibrary mirror, summarized): opioid effects on ACTH, cortisol, LH, prolactin and GH secretion [3]. The SmPC lists amenorrhea and impaired potency as uncommon [4].

Population evidence

Bawor 2014 [27]: multicentre cross-sectional study in Ontario clinics, 231 methadone patients and 783 non-opioid controls; men on methadone had testosterone of 100.10 versus 414.74 ng/dL in controls, inversely related to methadone dose (about 0.97 ng/dL lower per 10 mg); no significant suppression in women (read through a Springer-page summarizer). Hypogonadism study [28]: cross-sectional, 103 men, 84 on methadone and 19 on buprenorphine; low total testosterone (below 12.0 nmol/L) in 65% on methadone (39% below 8.0 nmol/L) versus 28% (11%) on buprenorphine; methadone-treated men had lower total testosterone than buprenorphine-treated men and reference groups (institutional repository record, summarized). Rubinstein 2016 [29]: retrospective cohort of 1,159 men on a stable single opioid for non-cancer pain, hydrocodone as referent; methadone OR 7.33 (95% CI 3.29 to 16.33) for androgen deficiency (abstract via OpenAlex reconstruction; full text not opened). Mendelson 1975 [30][31]: heroin use was consistently associated with low plasma testosterone; high-dose methadone (80 to 150 mg/day) depressed testosterone similarly while low-dose methadone (10 to 60 mg/day) stayed in the normal range, and testosterone recovered as methadone was reduced (PubMed page, summarized); in a 10-day inpatient heroin study, testosterone fell with heroin 45 to 65 mg/day and stayed low through a 7-day methadone withdrawal (institutional record, summarized).

Mechanism

Named in the label: opioids affect LH (and ACTH, cortisol, prolactin and GH) secretion (label 12.2 [3]). The specific upstream pathway for methadone was not verified in an opened source.

Why orange, not red

The evidence is cross-sectional or retrospective. No retrieved study measured testosterone before methadone and again on stable maintenance, and heroin use itself lowers testosterone [30][31], so confounding by indication cannot be excluded for the maintenance-versus-control contrasts (Hard Rule 6). The Rubinstein comparison is against hydrocodone, another opioid, not against non-use [29]. The hormone is directly measured and the mechanism is plausible, but this is not a nutrient, so the entry is bounded by population and dose.

Counter-evidence held on record

Counter-evidence and confounding Confounding by indication is presumed: opioid dependence, prior heroin use, hepatitis, HIV, age, smoking and obesity all affect testosterone, and the control groups were not matched on opioid history [27][28]. Sample sizes and adjustment sets for [28] were not captured. [30] and [31] are 1975 reports read only as summaries. Framing caution This is a hormone-axis effect of long-term use in men, not a nutrient deficiency. Do not suggest zinc, vitamin D or herbal testosterone products from this entry. Evaluation and any treatment belong to a prescriber.

Limits on what this shows

• Null findings: Dutch prospective study [7], 130 patients on stable methadone for pain: no correlation of QTc with serum potassium or magnesium (abstract via OpenAlex). Veterans cohort [9], 49 men on maintenance: no significant association of dichotomized QTc with potassium or magnesium, while calcium was associated (via a Brieflands page summarizer). Malaysian prospective cohort [8], 46 patients: calcium, potassium and magnesium measured as confounders; QTc was inconsistently related to dose and to methadone level over 8 weeks. A 2005 case report of a pain patient with QTc swinging between 407 and 569 ms on 60 to 240 mg/day recorded normal serum electrolytes, magnesium and calcium [53]. A Tenerife case-control [13] found potassium lower in maintenance patients than controls (4.87 versus 5.36 mmol/L) but serum magnesium higher (28.7 versus 25.3 mg/L), so the two electrolytes do not move together.
• At-risk subgroups: People on diuretics or laxatives, with cardiac hypertrophy or liver impairment, on CYP3A4-inhibiting drugs, or at daily doses above about 100 mg (SmPC [4]).

Direction

Risk context: the electrolyte modifies the drug's cardiac effect (Hard Rule 4). The depletion direction (drug lowers K or Mg) is label-listed without frequency or comparator and is unverified. Hard Rule 2: QT prolongation and torsades de pointes are cardiac toxicity, not nutrient deficiencies.

Regulatory status

US label section 5.4 (via summarizing tools [1][2]): QT prolongation and torsades de pointes occurred with methadone; monitor patients with risk factors "e.g., cardiac hypertrophy, concomitant diuretic use, hypokalemia, hypomagnesemia"; methadone inhibits cardiac potassium channels. Section 6, Metabolic: "hypokalemia, hypomagnesemia, weight gain" listed among adverse reactions from clinical studies or postmarketing reports, with no frequency [1][2]. Section 7 lists "drugs capable of inducing electrolyte disturbances: diuretics, laxatives, and, in rare cases, mineralocorticoid hormones" [2]. SmPC 4.4 and 4.8: QT cases especially above 100 mg/day, electrolyte disorders as risk factors, hypokalemia and hypomagnesemia with frequency "not known" [4].

Population evidence

Ehret 2006 [6]: retrospective, 167 methadone-maintained injection drug users versus 80 non-methadone injection drug users; QTc of 0.50 s or more in 16.2% versus 0%, torsades de pointes in 6 patients (3.6%); multivariable analysis attributed 31.8% of QTc variability to CYP3A4 drug interactions, hypokalemia and impaired liver function (abstract via OpenAlex reconstruction; hypokalemia is named, magnesium is not). Case series [10]: five patients with torsades on chronic methadone (mean dose 268 mg/day), all with contributing factors, one with potassium 2.0 mmol/L. Case report [11]: a 65-year-old man on methadone substitution with potassium 2 mmol/L at presentation, QTc 550 ms and torsades degenerating to ventricular fibrillation, treated with intravenous magnesium, potassium and lidocaine, with methadone stopped; the article attributes the event to hypokalemia and hypomagnesemia, but the magnesium values relayed by the summarizer (1.4 mmol/L at presentation, 2.5 mg/dL at discharge) are in inconsistent units and are not interpreted here. Serum magnesium [12]: 83 drug-addicted patients versus 81 healthy controls in Poland, mean serum magnesium 0.57 mmol/L (all patients below the reference range per the abstract), with no correlation with methadone treatment duration and no difference between methadone-treated and untreated patients (abstract via OpenAlex reconstruction).

Mechanism

Named: methadone blocks cardiac potassium channels (label 5.4 [1][2]); low extracellular potassium and magnesium reduce repolarization reserve. That is a mechanism for the electrolyte acting on the drug effect. No mechanism by which methadone lowers serum potassium or magnesium was found.

Why yellow, not orange

No retrieved study measured potassium or magnesium before and after methadone initiation or against a matched non-methadone group; the one treated-versus-untreated magnesium contrast was null [12]; hypomagnesemia in this population is present regardless of methadone (Hard Rule 6). The label listing is real but unquantified.

Counter-evidence held on record

Counter-evidence and confounding The electrolyte findings in torsades reports [10][11] come from patients with other contributing factors (QT-prolonging antibiotics, cocaine, high doses). Several controlled cohorts [7][9] found no electrolyte-QTc association, which weakens the claim that routine electrolyte shortfall explains the QT signal. Magnesium values in [11] are not compared numerically because of the unit inconsistency. Framing caution This is a risk-modifier and an arrhythmia warning, not a message that methadone depletes magnesium or potassium. Do not start potassium or magnesium supplements on this basis; measured levels, ECG findings and any replacement belong to a prescriber.

Limits on what this shows

• Null findings: Serum zinc no different [13]; urinary zinc no different [14]. Neither trial reports a baseline zinc level.
• At-risk subgroups: None defined. The trials were single-clinic Iranian cohorts.

Direction

Possible increased urinary loss during methadone detoxification (single preliminary report); no depletion shown in serum. Outcome gains from supplementation are not evidence of baseline deficiency.

Regulatory status

No zinc statement in the label text retrieved ("vitamin and mineral supplementation guidance" absent per the summarizer) [2].

Population evidence

Iyengar 1994 [15]: random urine samples from controls and methadone-treated drug users; urinary zinc 600 +/- 50 versus 300 +/- 30 micrograms per gram creatinine (p below 0.001); group sizes not captured; partial reconstruction of the abstract from an inverted index; the title says "a preliminary study during methadone detoxification". A 2015 narrative review [16] reports the same finding from that single methadone study (read through a summarizing tool). Kermanshah study [14]: cross-sectional, 67 methadone-treated patients and 62 healthy companions; urinary zinc not different (p=0.232), while urinary selenium, copper and iron were lower. Tenerife case-control [13]: serum zinc 1.17 (controls), 1.09 (maintenance) and 0.93 (detoxification) mg/L, no significant difference. Trial [17]: randomized, 68 methadone-maintained patients completing 3 months, zinc supplementation reduced relapse likelihood (p=0.01), craving (p=0.002), depression (p=0.01), anxiety (p below 0.001) and stress (p=0.001); zinc deficiency was not an inclusion criterion and serum zinc was not reported (full text via a Springer page summarizer). Trial [18]: 63 maintenance patients, 3 months, depression scores lower with zinc (p=0.038); anxiety inventory not different; serum zinc not mentioned (Brieflands page, summarized).

Retention

No retrieved study related zinc status to retention in maintenance treatment. [17] measured a relapse prediction scale and craving, not retention in treatment; [18] measured mood and quality of life.

Mechanism

Not established. Increased urinary zinc during detoxification is unexplained in the retrieved text.

Why yellow, not orange

Tier A measurements conflict (urinary excess [15] versus null [14]; serum null [13]), the single positive study is preliminary and was read at abstract level, and every cohort is confounded by opioid dependence, diet and drug use (Hard Rule 6). Benefit from supplementation in an unmeasured population is a treatment effect, not a depletion finding.

Counter-evidence held on record

Counter-evidence and confounding The trials [17][18] come from one country and one clinic type, rely on self-report scales, and could not fully blind participants [17]. Whether the two Iranian trials share participants was not checked. The 1994 urinary data are from detoxification, a different state from stable maintenance. Framing caution Not a statement that methadone lowers zinc and not a recommendation to supplement zinc. No dose is given here. A mood or relapse benefit in a trial is not proof of deficiency.

Limits on what this shows

• Null findings: No change in vitamin D after one year on treatment [21] (the Hard Rule 6 test: if prevalence is similar before and after the drug, the drug is not the cause). No vitamin D difference attributable to methadone duration was reported in the commentary on [20] [25].
• At-risk subgroups: Older patients and Black or Hispanic patients had higher deficiency in [20], which is the general-population pattern.

Direction

Context flag from the population and indication, not a depletion by the drug. Hard Rule 2: low bone density is not vitamin D deficiency.

Regulatory status

No vitamin D statement in the label text retrieved [2][4].

Population evidence

Kim 2009 [20]: cross-sectional, 93 patients from a 350-patient Boston program; 25-hydroxyvitamin D below 20 ng/mL in 36% and 20 to 30 ng/mL in a further 16% (52% low overall; PubMed listing summary [20] confirms the 52%); risks were age 40 or over (about 3.5-fold) and Black or Hispanic ethnicity (about 3.3-fold); comparison was to historical healthy volunteers. Malik 2020 [21]: observational, 272 patients with vitamin D data, 112 with paired admission and one-year values; vitamin D "did not change over time", and patients with higher vitamin D had longer cumulative retention (8.1 versus 4.8 years); this is nutrient status predicting retention, not a drug effect (abstract via OpenAlex reconstruction). Kashan study [22]: cross-sectional, 500 patients; serum vitamin D was positively correlated with depression and anxiety scores (journal page, summarized). Bone: 2025 cohort [23] found methadone associated with lower lumbar T-scores, with vitamin D included as a covariate, and the association lost significance after adjusting for estradiol and free testosterone (abstract via OpenAlex reconstruction). A PubMed listing on methadone and bone density returned six records (lower BMD in men on methadone, 83% below normal in one series) that were not opened [24].

Mechanism

Absent for methadone lowering vitamin D. The bone pathway in [23] runs through sex hormones.

Why yellow

Directly measured in methadone cohorts (Tier A) but the finding is high baseline prevalence without a drug-attributable change; yellow is the context-flag grade.

Counter-evidence held on record

Counter-evidence and confounding The Boston comparison group was historical [20]; the Iranian trial reported in a news item (Tier D, not used) supplemented vitamin D in 64 methadone patients with cognitive and mood gains [26], which is a treatment effect, not evidence of depletion. The positive vitamin D-depression correlation in [22] runs opposite to the usual expectation and was read only as a summary. Framing caution Not a claim that methadone depletes vitamin D. Testing is a clinical decision for at-risk groups; no supplement dose is given here.

Limits on what this shows

• Null findings: Searches for methadone SIADH case reports surfaced a tramadol case and an ecstasy case, not methadone [54]. No methadone cohort reporting sodium was retrieved.
• At-risk subgroups: Unknown; label notes opioids and diuretic interactions only in the electrolyte context [2].

Direction

Excess or retention (water retention with dilutional low sodium). Hard Rule 5: this is a disturbance, not a dietary deficiency, and must never imply that sodium or salt intake should increase; supplementation would be the hazard.

Regulatory status

US label section 6, Renal: "antidiuretic effect, urinary retention or hesitancy" [1]. Hyponatremia and ADH were not found in the label text returned by the summarizers [1][2]. SmPC: fluid retention and weight increase "common" [4].

Population evidence

Human: none attributing. A PubMed listing for "methadone hyponatremia" returned two records, a case report of a man stable on methadone 100 mg daily with persistent QTc above 550 ms who was also treated for alcohol intoxication and hyponatremia (not attributed to methadone) [34] and a Prescrire SSRI-interaction item (not opened). Animal (Tier B): intravenous methadone raised plasma arginine vasopressin from below 3 pg/mL within 5 minutes in 9 conscious dogs (Elsevier journal, 1987; urine output not substantially reduced) [32]; a 2012 Veterinary Journal study in six dogs reported significantly increased vasopressin concentrations after methadone [33].

Mechanism

Present in animals: methadone-stimulated vasopressin release. Human relevance not shown.

Why yellow

Tier B mechanistic animal data plus a label term; no human attribution. Not graded higher.

Counter-evidence held on record

Counter-evidence and confounding The dog studies used intravenous dosing and anesthetic conditions in [33]. The 1987 abstract reported no substantial urine output change [32]. Hyponatremia in people with opioid dependence has many causes (alcohol, SSRIs, polydipsia). Framing caution This is a retention-direction risk, not a low-salt problem. Do not increase sodium or fluids on this basis; evaluation belongs to a prescriber.

Limits on what this shows

• Null findings: None specific to nutrients.

Direction

Drug adverse effect (reduced gut motility). Hard Rule 2: an outcome of the drug's pharmacology, not a nutrient deficiency.

Regulatory status

Label section 2.8 (summarized): "During prolonged administration of methadone, monitor patients for persistent constipation and manage accordingly" [2]. Section 12.2 (medlibrary mirror): effects on stomach, duodenum and colon motility with resulting constipation [3]. SmPC 4.8: constipation common; nausea and vomiting very common [4]. Section 7: laxatives are listed with diuretics as drugs capable of inducing electrolyte disturbances, a QT-risk consideration [2].

Population evidence

No frequency estimate for methadone constipation was retrieved beyond the SmPC category. No study linking methadone constipation to a nutrient measurement was found.

Mechanism

Present: opioid-receptor effects on gut motility (label 12.2 [3]).

Why yellow

Label-level context flag. The only nutrient-relevant link is indirect: managing constipation with agents that shift electrolytes can interact with the drug's QT risk.

Counter-evidence held on record

Counter-evidence and confounding Shares the label QT and electrolyte statements with the potassium and magnesium entry and the gut motility mechanism with the microbiome entry (see overlap map). Framing caution Constipation is not a sign of magnesium or fiber deficiency from this evidence. Management choices, including laxatives, belong to a prescriber who knows the QT context.

Limits on what this shows

• Null findings: No dose-independent effect could be separated from recovery in the retrieved data.

Direction

Weight gain (excess direction over years) after a baseline of illness-linked undernutrition. Hard Rule 2: weight change is an outcome. Hard Rule 6: baseline undernutrition precedes methadone.

Regulatory status

US label section 6: "weight gain" under Metabolic; "anorexia" is also listed [2]. SmPC: weight increase and fluid retention common [4].

Population evidence

Iran 2012 [47]: 55 opioid users starting maintenance (47 followed 8 weeks); weight 61.4 to 65.3 kg and BMI 21.4 to 23 kg/m2 (p below 0.01); underweight prevalence 27.3% to 12.7% (Brieflands page, summarized). UKY chart review [48]: 74 patients retained 3 years; overweight or obese categories 42% to 88%; more gain at higher doses (summarized). Malik 2020 [21]: significant BMI gain over a year. Tenerife [13]: 106 opiate addicts (87 on maintenance) had lower red cell count, albumin, folate, retinol, alpha-tocopherol, potassium and selenium than 186 controls; the authors attributed the vitamin and selenium gaps to low intake of vegetables, legumes and fish, and noted maintenance patients ate more meals than heroin patients in detoxification (PDF, summarized).

Mechanism

Not established in the retrieved sources (fluid retention and appetite change are listed label effects).

Why yellow

Context flag: a labeled effect and a disease-recovery effect in the same population, with no nutrient measured as deficient because of the drug. Folate, retinol and selenium differences in [13] are attributed by the authors to diet and are not drug-attributed.

Counter-evidence held on record

Counter-evidence and confounding Baseline undernutrition, stimulant use, hepatitis, smoking and housing all affect weight and micronutrient status in this population. The 8-week and 3-year cohorts are small and uncontrolled. Framing caution Not a nutrient deficiency claim. Weight and diet assessment are part of routine care; the evidence here does not support a specific supplement.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or C evidence supports even a yellow grade (the measured results are null for drug attribution), and no circulating consumer claim for methadone and B12 was traced to rebut as white. Mixed result: a literature absence in the records opened, with a tooling limit on the consumer-claim audit. The null findings describe what was measured; they are not a finding of safety. Re-check recommended.

What was searched

WebSearch: "methadone maintenance vitamin B12 homocysteine folate levels patients study" (returned a methamphetamine B12 study, B vitamin status of heroin users, and unrelated pages); "methadone depletes vitamin B12 nutrient depletion drug mugger methadone" (returned book listings for a consumer depletion handbook, which were not opened, plus the Norwegian B12 cohort). PubMed listing "methadone vitamin B12" (6 records: one methadone cohort opened as [21], one multivitamin trial for mood in maintenance patients seen as a title only, a 2004 discriminant-analysis paper from the Tenerife group seen as a title only, and three irrelevant items). Opened: Malik 2020 [21] (260 maintenance patients with B12; low B12 associated with stimulant misuse, not with methadone); Western Norway cohort [49] (672 people, 2,451 B12 measurements, 39% on methadone at baseline; opioid agonist therapy dose ratio effect 19 pmol/L, 95% CI -12 to 50, not significant; 1.2% below 175 pmol/L); Tenerife [13] (B12 582 in maintenance versus 523 in controls, ng/L, no significant difference). Excluded cohorts: methamphetamine addicts [50] (not methadone), B1, B2 and B6 enzyme coefficients in heroin users and early methadone patients [51] (not B12). Consumer page [52] mentions no nutrient. Label word check: no B12 statement in text retrieved [2][4].

Channel outcomes

PubMed listing worked after rate-limit waits; PubMed and PMC article pages returned bot challenges; WebSearch ran out of its shared budget after the two queries above, so the consumer-claim side of the search was limited to those two queries and the Drugs.com consumer page.