Efavirenz

Limits on what this shows

• Null and contradicting findings: Overton 2015 [13] (ACTG A5280, double-blind, 165 randomized, all started efavirenz/emtricitabine/tenofovir DF): in the 86 placebo participants median baseline 25(OH)D was 25.1 ng/mL (IQR 17.8 to 31.5) and the change at week 48 was +0.7 ng/mL (IQR -5.3 to 4.3), at week 24 -0.6 ng/mL, not significant; PTH rose 22% (p of 0.001 or less). The vitamin D3 plus calcium arm rose by a median 24.5 ng/mL, so efavirenz did not prevent a rise. Entry required 25(OH)D of 10 to under 75 ng/mL; enrolment September 2011 to February 2012; no arm without efavirenz. MACS, Zhang 2017 [14]: 640 HIV-infected men on HAART, current efavirenz use OR 1.13 (95% CI 0.72 to 1.77, p = 0.601) for 25(OH)D below 20 ng/mL, season and site adjusted; the only significant antiretroviral was ritonavir at OR 0.58 (0.35 to 0.99). Stored sera 1999 to 2008, cross-sectional, reference category not labeled. Poowuttikul 2013 [38]: retrospective routine-care data from 160 HIV-infected youth aged 2 to 26 (95% African American); the abstract states that tenofovir and/or efavirenz prescription did not show different 25(OH)D levels than other antiretrovirals (no OR, CI or p value given). Musarurwa 2018 [39]: 284 hospitalised adults with HIV in Zimbabwe, efavirenz-based cART reported as not associated with vitamin D status (no statistics in the abstract; tuberculosis-selected cohort). Foca 2012 [40]: 75 people randomized to tenofovir/emtricitabine with efavirenz (33) or atazanavir/ritonavir (42), 1,25-dihydroxyvitamin D stable over 48 weeks; this is the hormone, not 25(OH)D, and 25(OH)D by arm is not in the abstract, so it does not bear on this entry. A 2021 meta-analysis [41] found ART exposure OR 2.296 (95% CI 1.287 to 4.097, 3 studies) for vitamin D deficiency but did not stratify by efavirenz, so it is not counted.
• At-risk subgroups: People with higher baseline 25(OH)D (largest absolute falls) [4]; high plasma efavirenz concentration and high CYP3A activity [5]; people with low Karnofsky score, high viral load or TB co-infection already have lower baseline levels [5]; Black race, Hispanic ethnicity, lower UV exposure and higher BMI independently predicted low levels in the SUN cohort [6].

Direction

Depletion (circulating 25-hydroxyvitamin D lower). Hormone-axis markers and bone density are not the nutrient and are not graded here (Hard Rule 2).

Regulatory status

The US efavirenz label sections read contain no vitamin D, calcium or bone statement. They do state in vivo induction of CYP3A and CYP2B6 and in vitro inhibition of CYP2C9 and CYP2C19 [1].

Population evidence

Havers 2014 [4] (cohort nested in the randomized PEARLS trial, 9 countries, 220 with week 0 and 24 results): mean change in 25-hydroxyvitamin D from week 0 to 24 was -7.94 ng/mL (95% CI -10.42 to -5.54) on efavirenz/lamivudine/zidovudine and -6.66 (-9.40 to -3.92) on efavirenz/emtricitabine/tenofovir DF, versus -2.29 (-4.83 to 0.25, not significant) on atazanavir/didanosine/emtricitabine without efavirenz. Arm A vs B p = 0.002, C vs B p = 0.022, A vs C p = 0.498. From week 24 to 48 levels were flat (arm A -0.22, arm B +1.52, arm C +0.63 ng/mL). Baseline median 32 ng/mL (IQR 25 to 39), 14% below 20 ng/mL; the highest baselines fell most; country and baseline season were significant (p Nylen 2016 [5] (prospective, efavirenz-based cART): in 89 HIV-only adults severe deficiency went from 27% at baseline to 76% (week 4), 79% (week 16) and 43% (week 48), with median 25(OH)D3 falling about 40%, 50% and 14%; in 102 TB-HIV adults 57% to 70% and 72% at weeks 4 and 16. Wohl 2014, ECHO [34] (second independent randomized dataset; secondary analysis of a randomized, double-blind Phase III trial, antiretroviral-naive adults, tenofovir DF/emtricitabine in both arms, 25(OH)D measured on stored samples at baseline and weeks 24 and 48): 690 randomized (rilpivirine 346, efavirenz 344), 582 with paired baseline and week 48 data (292 and 290). Mean change at week 48 was -2.5 ng/mL with efavirenz (P below 0.0001 vs no change) and -0.2 ng/mL with rilpivirine (P = 0.57). The between-arm difference remained significant after adjusting for season of randomization and race or ethnicity (adjusted figures and CIs not in the abstract). Severe deficiency (below 10 ng/mL): 5% in both arms at baseline, 9% with efavirenz vs 5% with rilpivirine at week 48 (P = 0.032); among people with insufficiency or deficiency at baseline, new severe deficiency 8% with efavirenz vs 2% with rilpivirine (P = 0.0079). Abstract read from the SAGE page, an OpenAlex record and a PubMed record. Brown 2010 [35] (observational, stored plasma from 87 antiretroviral-naive people in Cleveland, 51 starting efavirenz and 36 starting non-efavirenz regimens, 89% of them protease-inhibitor based; 25(OH)D at baseline and 6 to 12 months): adjusted for baseline level, race and season, 25(OH)D fell 12.7 +/- 3.7 nmol/L (mean +/- SE, P = 0.001) more on efavirenz, and the adjusted prevalence ratio for hypovitaminosis D (37.5 nmol/L or below) was 1.8 (95% CI 1.2 to 2.8, P = 0.007); baseline median 52.7 nmol/L. Conesa-Botella 2010 [36] (prospective Belgian cohort starting HAART, paired plasma before and after 12 months, participant number not stated in the abstract): 25(OH)D fell significantly on both nevirapine- and efavirenz-containing regimens but not on protease-inhibitor regimens; in the multivariate model an NNRTI-based regimen (P = 0.001) and lower body weight (P = 0.008) were associated with the decrease; the share below 30 ng/mL rose from 70.1% to 81.6%. Cook 2016 [12] (pilot randomized trial, efavirenz/emtricitabine/tenofovir DF vs raltegravir/darunavir/ritonavir; 35 randomized, 20 completed): the listing summary reports 25(OH)D rose in the raltegravir/darunavir/ritonavir arm (p = 0.0004) but not the efavirenz arm (p = 0.78); the abstract text reads as sustained reductions with efavirenz. MONET [11]: in 221 people switched from NNRTI- or PI-based HAART, lower screening 25(OH)D was associated with efavirenz use (p = 0.0062) and zidovudine use (p = 0.015), and rises to week 96 were larger in those who stopped efavirenz and/or zidovudine. Cross-sectional : Dao [6] (n = 672, efavirenz aOR 1.98, 95% CI 1.18 to 3.34 for 25(OH)D below 30 ng/mL); Allavena [7] (n = 2,994, 2,660 on ART; efavirenz the only antiretroviral associated with deficiency, aOR 1.89, 95% CI 1.45 to 2.47, with no effect of etravirine, darunavir or raltegravir); Cervero [8] (n = 352; efavirenz use independently associated with deficiency or insufficiency, 25(OH)D higher with PIs than efavirenz, effect size not in abstract); Kwan [9] (n = 463; efavirenz associated with deficiency, protease inhibitors with lower odds, no OR in abstract); Dave [10] (n = 444; efavirenz aOR 2.04, 95% CI 1.01 to 4.13 for deficiency); Welz [37] (n = 1,077, London clinic; among people on combination therapy current efavirenz use aOR 2.0, 95% CI 1.5 to 2.7 for severe deficiency below 10 ug/L; abstract read from a PubMed record and an OpenAlex record).

Mechanism

Named, Tier B plus indirect human support. Efavirenz induces CYP3A and CYP2B6 in vivo [1]. In primary human dermal fibroblasts efavirenz, stavudine and ritonavir lowered CYP2R1 (the 25-hydroxylase) and raised CYP24A1 (the catabolic 24-hydroxylase) expression, with lower cellular 1,25-dihydroxyvitamin D3 [15]. In vivo, high CYP3A activity (4beta-hydroxycholesterol/cholesterol ratio), high plasma efavirenz concentration and low cholesterol predicted the early fall [5]. Limits: rifampicin, a stronger CYP3A inducer, had no significant effect on 25(OH)D3 in the same cohort [5], so general CYP3A induction does not explain it alone; 1,25-dihydroxyvitamin D was not measured in the cohorts that show a fall.

Cohort test, co-drugs

Efavirenz was in every cohort cited, always with a nucleoside backbone. Two comparisons hold the backbone constant: in ECHO both arms took tenofovir DF/emtricitabine and only the third agent differed (efavirenz vs rilpivirine, randomized, double-blind) [34], and in Havers the efavirenz arms had two different backbones [4]. Brown compared efavirenz starters with mostly protease-inhibitor starters, so the comparator is an active regimen, not non-use [35]. Conesa-Botella pooled efavirenz and nevirapine in its multivariate NNRTI term, but reported the efavirenz fall on its own in the abstract [36]. In Havers the two efavirenz arms carry different backbones (lamivudine/zidovudine and emtricitabine/tenofovir DF) and fell by similar amounts, while the arm with emtricitabine but no efavirenz did not [4], which points at efavirenz and away from the backbone. Allavena found efavirenz the only antiretroviral associated with deficiency among 2,660 treated patients, with no effect of etravirine, darunavir or raltegravir [7]. In MONET efavirenz and zidovudine could not be separated [11]; in Overton and Cook every efavirenz participant also took emtricitabine/tenofovir DF [13][12]. CLASS EXTRAPOLATION (capped at yellow, not counted toward this grade): an urban cohort of 62 men found median 25(OH)D 42.4 nmol/L on an NNRTI vs 64.9 nmol/L on a PI (OR for deficiency 6.62, 95% CI 1.91 to 22.89); efavirenz was not named in the abstract, so efavirenz was not confirmed to be in that cohort [16].

Why red (changed from orange on 2026-10-10)

Red needs at least two independent Tier A sources with efavirenz in the cohort, direct measurement of the nutrient, consistent direction, clinically meaningful size, a named mechanism and no substantial contradicting Tier A evidence. Independent sources: the PEARLS substudy [4] and the ECHO trial [34] are different randomized trials in different populations, each with an arm that did not take efavirenz, plus an adjusted comparison of efavirenz against other regimens [35], a prospective NNRTI cohort [36] and the Ethiopian cohort [5]. The ECHO design rules out the backbone as the explanation, because both arms took tenofovir DF/emtricitabine. Direction is consistent in every one of those, and size reaches thresholds that matter (severe deficiency 27% to 76% at week 4 in [5]; 5% to 9% in [34]; falls of 7 to 8 ng/mL in [4]). Mechanism is named (CYP3A and CYP2B6 induction in the label [1], lower CYP2R1 and higher CYP24A1 expression in human fibroblasts [15], and high CYP3A activity and efavirenz concentration predicting the fall in vivo [5]). The three contradicting datasets are judged not substantial: the ACTG A5280 placebo arm [13] had no efavirenz-free comparator, enrolled people with baseline 10 to under 75 ng/mL and measured at the same season a year later, and its regimen is the one that fell by a mean of 2.5 ng/mL in ECHO [34], so a small mean change with a wide spread is not inconsistent; MACS [14] is a cross-sectional stored-serum analysis (OR 1.13, 95% CI 0.72 to 1.77) against four cross-sectional cohorts with aOR 1.89 to 2.04; the youth cohort [38] is small and retrospective with no effect size. Limits that remain: ECHO's mean fall is small, effect size varies with baseline, season and country, 1,25-dihydroxyvitamin D was not measured where a fall is shown, and a stronger CYP3A inducer (rifampicin) did not change 25(OH)D3 in [5]. This is a circulating-level finding only; an editor decides whether size justifies the red label.

Counter-evidence held on record

Counter-evidence on record The two null datasets are described in full above [13][14]. Competing explanations: season (Havers found baseline season significant [4]; Overton enrolled September to February and measured at the same season a year later [13]; ECHO adjusted for season of randomization [34]), country and sunlight, disease activity and viral load at baseline [5], tenofovir and zidovudine co-exposure in several cohorts [11], the active comparator being a protease-inhibitor regimen in [35], and confounding by indication in the cross-sectional studies [37][7][6][10] (cross-sectional ORs cannot separate pre-existing from drug-induced differences). Framing caution This is a monitoring prompt about circulating 25-hydroxyvitamin D after starting efavirenz-based therapy, not a finding that efavirenz causes bone disease and not a supplement recommendation. Any vitamin D decision belongs to a prescriber.

Limits on what this shows

• Null and contradicting findings: Jacobson 2024 [18] (IMPAACT 2010/VESTED randomized, 340 mothers, 90% on folic acid): authors found no clinically meaningful difference in red-cell folate between dolutegravir and efavirenz arms; adjusted geometric mean ratio for the change from entry to delivery, DTG+FTC/TAF vs EFV/FTC/TDF, 1.03 (95% CI 1.00 to 1.06). This compares arms in pregnancy and uses red-cell folate; it is not a comparison against baseline off therapy and is not a direct refutation of the serum fall.
• At-risk subgroups: Only women were analysed; in 26 women pregnant before week 24 (efavirenz n = 7; dolutegravir arms n = 10 and 9) the efavirenz arm changed -3.3 +/- 8.1 nmol/L at week 12 against rises of 2.4 and 2.3, and by week 24 it was +2.8 +/- 13.4 [17]. The label's first-trimester fetal-harm statement is the pregnancy context for any folate question [1].

Direction

Depletion of circulating (serum) folate; mechanism unknown. Not a dietary-shortfall claim.

Regulatory status

No folate statement in the label sections read [1].

Population evidence

Chandiwana 2021 [17]: ADVANCE substudy, 406 women (TAF/FTC + dolutegravir n = 148, TDF/FTC + dolutegravir n = 133, TDF/FTC/efavirenz n = 125), mean age 31.5, serum folate in nmol/L. Efavirenz arm: baseline 23.4 +/- 7.6, week 12 17.4 +/- 7.5 (change -5.9 +/- 8.1), week 24 18.5 +/- 7.7 (change -4.8 +/- 9.0). TDF/FTC + dolutegravir: 24.2 to 22.4 to 24.1 (changes -1.8 and -0.1). TAF/FTC + dolutegravir: 22.8 to 26.8 to 28.5 (changes +4.0 and +5.7). Low folate (below 14 nmol/L): efavirenz arm 12 (10%) at baseline, 50 (40%) at week 12, 38 (30%) at week 24; TDF/FTC + dolutegravir 14%, 23%, 13%; TAF/FTC + dolutegravir 18%, 5%, 6%. Efavirenz versus both dolutegravir arms p < 0.001 at weeks 12 and 24; paired within-arm change significant for efavirenz and TAF arms (p < 0.001) and not for TDF/FTC + dolutegravir (p = 0.917). No confidence intervals reported.

Mechanism

Absent as a tested mechanism. The authors state it is unclear whether efavirenz or TAF drives the arm differences and cite a general lowering of micronutrients on ART; they hypothesize dolutegravir may block cellular folate uptake, which would raise serum folate while lowering tissue folate, untested in humans [17]. The flat tenofovir/emtricitabine plus dolutegravir arm argues against the nucleoside backbone as the cause, but a dolutegravir effect that masks a backbone fall cannot be excluded.

Cohort test, co-drugs

Efavirenz was always given with tenofovir DF and emtricitabine [17]. Hemoglobin was reported as 12.2 g/dL in the efavirenz group (timepoint not stated in the text returned); MCV and B12 were not reported.

Second-source hunt, 2026-10-10

PubMed listings "efavirenz folate" (5 records: the two trials already cited, a Brazilian pregnancy outcomes cohort reporting folic acid use of 44% in efavirenz-exposed vs 48% in dolutegravir-exposed women with no folate level, a nanoparticle paper and an Iranian paper on B12 and folic acid supplements improving blood counts) and "efavirenz homocysteine" (1 record, a rilpivirine metabolism paper) found no second trial or cohort measuring folate on efavirenz against baseline [46]. VESTED [18] is the only other randomized comparison and uses red-cell folate in pregnancy: the dolutegravir/emtricitabine/TAF arm rose 3% more than the efavirenz arm (adjusted geometric mean ratio 1.03, 95% CI 1.00 to 1.06), which the authors judged not clinically meaningful; a second record of the abstract was read from a CDC stacks page and a Duke scholars page. The IMPAACT CROI 2023 poster and the Botswana Harvard page appeared in search results but were not opened for numbers.

Why orange, not red

One randomized trial, serum folate only (a short-term, non-fasting marker per the authors), no red-cell folate, no named mechanism, no second efavirenz cohort measuring folate against baseline, and a pregnancy trial with red-cell folate that found no meaningful difference. Boundary in the title: women, first 24 weeks, serum level.

Counter-evidence held on record

Counter-evidence on record Staple foods are folic-acid fortified in South Africa, so few women were deficient, and supplement intake was not accounted for [17]. Serum folate can move with recent intake. The red-cell result in pregnancy [18] did not show a clinically meaningful difference. Framing caution This is a functional monitoring prompt, not a finding that efavirenz causes folate deficiency, and it makes no supplement recommendation. Pregnancy and any folic acid decisions belong to the prescriber.

Limits on what this shows

• Null and contradicting findings: In the listing-level results efavirenz lost less spine BMD than atazanavir/ritonavir, a similar amount at the hip, and a similar total BMD loss to lopinavir/ritonavir [33]. A bone-loss outcome in a regimen that always contains tenofovir cannot be attributed to efavirenz.

Direction

Outcome (BMD loss). Not a nutrient entry; no calcium level was shown to fall (Hard Rule 2).

Regulatory status

No bone statement in the label sections read [1].

Population evidence

Overton [13]: in the placebo arm of 86 on efavirenz/emtricitabine/tenofovir DF total hip BMD fell a median 3.19% (IQR -5.12 to -1.02) and lumbar spine 2.91% by week 48; vitamin D3 plus calcium reduced the hip loss to 1.46% (p = 0.001), lumbar spine difference p = 0.085 complete-case. Dave [10]: total hip BMD 0.909 g/cm2 on ART vs 0.956 in ART-naive (p = 0.0001); efavirenz-based ART independently associated with lower lumbar BMD. Cook [12]: BMD fell in the efavirenz/emtricitabine/tenofovir DF arm (total hip p = 0.002 and femoral neck p = 0.004 in the listing summary) but not in the raltegravir/darunavir/ritonavir arm. Listing-level only [33]: in ACTG A5224s spine BMD fell 1.7% with efavirenz vs 3.1% with atazanavir/ritonavir (p = 0.035) and hip 3.1% vs 3.4% (p = 0.61); in a randomized comparison total BMD fell 2.3% with efavirenz vs 2.5% with lopinavir/ritonavir (p = 0.86).

Mechanism

Not isolated for efavirenz. Tenofovir exposure, ART-initiation bone loss common to all regimens and any vitamin D effect are overlapping candidates; none is tested for efavirenz alone in the sources read.

Counter-evidence held on record

Counter-evidence on record Fall risk, tenofovir co-exposure, HIV itself and weight regain at ART start are competing explanations. Serum calcium was not reported in the sources opened. Framing caution BMD loss is an outcome. Do not read it as low calcium or low vitamin D, and do not infer a supplement need.

Limits on what this shows

• Null and contradicting findings: Barroso could not separate efavirenz from tenofovir/emtricitabine, nor drug effect from return to health after viral suppression; the authors list both as limitations [19].

Direction

Excess (lipids higher). Outcome entry; not a nutrient deficiency (Hard Rule 2).

Regulatory status

Label sec 5.11: treatment with efavirenz has resulted in increases in total cholesterol and triglycerides; test before starting and periodically. Non-fasting total cholesterol of at least 240 mg/dL in 28 to 54% and at least 300 mg/dL in 4 to 20% of patients across trial arms; grade 3 to 4 triglycerides (at least 751 mg/dL) in 6 to 17% (summarizer-read) [1].

Population evidence

Barroso [19]: 29 on efavirenz/emtricitabine/tenofovir DF for a mean 22.3 months vs 33 antiretroviral-naive: higher total cholesterol (p < 0.001), LDL (p = 0.004), HDL (p = 0.018), triglycerides (p = 0.048), glucose (p = 0.010); total cholesterol/HDL ratio not different. Group means were in figures only.

Mechanism

Not established in the sources opened. Li 2018 (listing level only, not opened) reports efavirenz perturbing 13 coding genes in a human T-cell line, 11 of them lipid or cholesterol biosynthesis genes.

Counter-evidence held on record

Counter-evidence on record Untreated HIV lowers cholesterol, so part of the rise is recovery. Barroso is cross-sectional with partial longitudinal data [19]. Framing caution A lipid result for the clinician. It is not a vitamin or mineral finding.

Limits on what this shows

• Null and contradicting findings: Transepithelial electrical resistance was unchanged with efavirenz alone at 10 uM and only trended up with 1, 10 or 50 uM manganese, not reported as significant. The tested concentrations (10 to 20 uM) exceed the label's single-dose peak plasma range of 1.6 to 9.1 uM [1], and the article does not address clinical relevance. Zinc and iron, which SLC39A8 also transports at different affinities, were not tested with efavirenz [43].

Direction

Excess or retention direction (increased cellular uptake of a trace metal and a toxic metal), mechanism only. Not a deficiency, and supplementation would be the hazard, not the remedy (Hard Rule 5). No manganese supplement is suggested.

Regulatory status

No manganese, zinc or transporter statement in the label sections read [1].

Population evidence

None. The article reports no in vivo or human data for efavirenz and metal levels; its Discussion only speculates that some efavirenz side effects could relate to manganese homeostasis [43].

Mechanism

Named, Tier B (cell lines). Damm-Ganamet 2025 [43]: efavirenz was identified in a library screen as a potentiator of cadmium uptake in SLC39A8-inducible HEK293-T-REx cells; at 20 uM it increased the apparent maximal uptake rate (Vmax) and slightly reduced cadmium affinity (Km). Efavirenz also increased manganese-induced calcein quenching in the same cells (1 mM manganese; no fold change given). It rescued cadmium uptake in six disease-associated SLC39A8 variants. The potentiation was selective: no enhancement in SLC39A4 or SLC39A14 cells, and it acted on mouse SLC39A8 but not chicken, frog or zebrafish. In primary murine intestinal epithelial monolayers, the active enantiomer at 20 uM for 24 hours raised MT1 expression about 1.5-fold. Full text opened through a summarizing fetch tool; the abstract was also read from an OpenAlex record.

Counter-evidence held on record

Counter-evidence on record Engineered cell lines, one report, concentrations above clinical peak plasma levels, no human exposure data. Added in the 2026-10-10 repair from a listing that returned the record while hunting zinc and iron signals. Framing caution A cell-culture mechanism for metal uptake. It is not evidence that people on efavirenz accumulate manganese or cadmium, and it supports no supplement or dietary action.

Limits on what this shows

• Null and contradicting findings: No cohort measuring B12 on efavirenz was retrieved, so there is no null or positive population result. Confounding by indication (HIV, diet, malabsorption) is presumed (Hard Rule 6).

Direction

Depletion state in one patient (coincident). Neuropathy and optic neuritis are outcomes (Hard Rule 2).

Regulatory status

No B12 statement in the label sections read [1].

Population evidence

Lubega 2024 [28]: 33-year-old woman with HIV, serum B12 78 pmol/L, MCV 101 fL, hemoglobin 12.9 g/dL, one week of vision loss, earlier peripheral neuropathy treated with B12 tablets about 3 years before; vision recovered over about 5 months of B12 injections. The record states no cause and does not mention diet. A PubMed listing for efavirenz and B12 returned only this report and an Iranian cross-sectional study of B12 and folic acid supplements in 100 people on HAART with the regimen not stated [33].

Mechanism

Absent. The authors offer no malabsorption, diet or enzyme-induction explanation.

Counter-evidence held on record

Counter-evidence on record Single case; HIV itself and diet are alternative explanations. Framing caution A case report, not a causal finding.

Limits on what this shows

• Null and contradicting findings: None found. Capsule and sprinkle instructions and a 2-hour fasting window for sprinkled capsule contents appear only in the Tier D page and are not used here.

Direction

Nutrient affects drug (Hard Rule 4). The drug is not shown to deplete anything here.

Regulatory status

Label sec 12.3: with a high-fat, high-calorie meal the 600 mg tablet's AUC rose 28% and Cmax 79% vs fasting. Sec 2.2 and 2.3: take on an empty stomach, preferably at bedtime [1]. A Drugs.com consumer page (Tier D) repeats the advice and says food raises the amount of drug in the body [31].

Population evidence

Label food-effect figures (study design not in the text returned) [1]. No separation interval in hours is stated in the label text returned.

Mechanism

Not stated in the label text returned; only the exposure figures are available.

Counter-evidence held on record

Counter-evidence on record Label figures were read through a summarizing tool, not the raw label. Framing caution A timing instruction for the prescriber or pharmacist to confirm, not a reason to change diet.

Limits on what this shows

• Why no grade: Dual condition met: no opened Tier B or C source for depletion of either mineral (the opened human record is a small null that supports no color) and no circulating claim found. Disclosed gap, not reassurance. The unopened rat study is a tooling-limited lead, not evidence.

What was searched

PubMed "efavirenz zinc" (15 results, 10 shown): reviews of anti-HIV agents, a rat study of zinc given with an efavirenz/lamivudine/tenofovir HAART for hepatic lipids (Hamed 2022, listing-level, not opened; zinc given as a treatment, no depletion result in the listing), an antioxidant profile study on generic HAART in India with regimen not named (Sundaram 2008, listing-level), and unrelated chemistry. PubMed "efavirenz selenium" (1 record): pre-ART selenium as an outcome predictor in PEARLS [30]. No source measured zinc or selenium change on efavirenz. Combined boolean query "efavirenz zinc selenium" returned an error and was rerun separately.

Repair search, 2026-10-10

The Indian antioxidant study (Sundaram 2008 [42]) was opened this time (OpenAlex record plus PubMed record): in 26 people on efavirenz-containing HAART plus 9 untreated and 20 uninfected controls, zinc did not differ significantly from controls over 12 months; in the untreated group zinc went from 79.0 to 74.5 ug/dl (p = 0.052). Selenium was not measured. A null in a very small, regimen-mixed (stavudine or zidovudine backbones) cohort supports no color and is not reassurance. The listing "efavirenz AND micronutrient" (67 results, 10 shown) surfaced Damm-Ganamet 2025 [43], an in vitro finding that efavirenz potentiates SLC39A8 (a transporter for manganese, zinc and iron); efavirenz's effect on zinc uptake was not tested, and the manganese finding is graded as its own yellow entry. Further listings "efavirenz serum zinc levels" and "efavirenz plasma selenium" returned fetch errors that were not confirmed to be empty-result pages [46].

Channel outcomes

PubMed listings worked after retries; the rat study was not opened (second-hand listing only). A genuine gap for efavirenz as far as these listings show; the rat study abstract could change this if it reports zinc status and was not retrievable here.

Limits on what this shows

• Why no grade: Dual condition met on what was opened: no confirmed efavirenz inhibition (so no Tier B support) and no circulating claim. If the full text names efavirenz as an inhibitor this would become a yellow in vitro entry (placental uptake, not systemic carnitine). Disclosed gap, tooling-limited.

What was searched

PubMed "efavirenz carnitine" (6 records): Karahoda 2019 tested 16 antiretrovirals including 3 NNRTIs against OCTN2-mediated carnitine uptake; 7 inhibited, named strongest were ritonavir, saquinavir and elvitegravir, and efavirenz is not stated among the inhibitors [29]. Other records were acylcarnitine studies in people with HIV with no efavirenz stratification, a transporter paper for emtricitabine and an ototoxicity paper with L-carnitine rescue.

Repair search, 2026-10-10

A research-results record for Karahoda 2019 [45] relays that the abstract lists the NNRTIs tested as rilpivirine, efavirenz and etravirine, so efavirenz was tested; it names ritonavir, saquinavir and elvitegravir as the most potent of seven inhibitors and does not say whether efavirenz was one of the other four. A Charles University master's thesis record (Markova, 2019) states the aim of testing rilpivirine and efavirenz (among other drugs) on placental carnitine transport, but the abstract text was truncated before results [45]. The Elsevier landing page returned no content, the thesis handle page was blocked by robots rules, Europe PMC returned HTTP 429, and a final web search was refused for budget. The efavirenz result remains unretrieved after these channels [46].

Channel outcomes

OpenAlex record for PMID 30735677 had no abstract; the PubMed uid record gave a paraphrased abstract without the drug list; the full text was not opened. This is tooling-limited: the drug list may sit in the full text.

Limits on what this shows

• Why no grade: Dual condition met: no Tier B or C source and no circulating claim retrieved. Disclosed gap, not reassurance.

What was searched

PubMed "efavirenz pyridoxine" (4 records): a rifabutin interaction case report, a delamanid population PK model (efavirenz as a co-medication), a T-cell phenotype study and a tuberculosis guideline mentioning pyridoxine in pregnancy and isoniazid; none measure B6 on efavirenz [33].

Repair search, 2026-10-10

PubMed listing "efavirenz vitamin B6" returned 1 record (the T-cell phenotype paper already seen, where B6 is only an indexing term). Listings "efavirenz AND (vitamin B6 OR pyridoxal OR pyridoxal phosphate OR B vitamins)", "efavirenz pyridoxal" and a quoted-phrase variant returned fetch errors that were not confirmed to be empty-result pages, and were not retried further; the earlier "efavirenz pyridoxine" listing found no B6 measurement [46]. No circulating consumer claim was found in the pages opened [31][32].

Channel outcomes

Single listing, worked. Neuropsychiatric outcomes on efavirenz (label: any nervous system symptom 52.7% vs 24.6%) are drug toxicity and no opened source links them to B6 [1]. Genuine gap for the term search; open-web search unavailable.