Explore Supplements Subscribe
The VerdictLOW CONVICTION

Eat your carrots and liver. Skip the capsule.

Ask yourself one question. Do you smoke, did you ever smoke, or were you exposed to asbestos? If yes, do not buy any beta-carotene supplement at any dose. If no, you still do not need one — get vitamin A from food and skip the capsule.

Vitamin A is a family of compounds your body needs for vision, immunity, and skin integrity. Your body uses an enzyme to convert plant beta-carotene from carrots into vitamin A, and it deliberately slows the conversion down when stores are full — built-in brakes. A beta-carotene capsule bypasses the brakes entirely, and in two large trials those un-braked doses caused more lung cancer in smokers, not less.

That's the general answer. Your stack is different.

Check your whole stack
SH
Dr. Seth Holbrook, DPT — Doctor of Physical Therapy • Coach to 300+ clients
I built The Verdict to cut through recycled health advice and show what the evidence actually supports.
Vitamin · Fat-Soluble · Antioxidant Carotenoid

Vitamin A & Beta-Carotene

Vitamin A is the family of fat-soluble compounds your body uses for vision, immunity, and epithelial maintenance. You get it two ways — preformed retinol from animal foods and provitamin-A carotenoids from plants. The food version is self-limiting. The capsule is not.

Ask yourself one question. Do you smoke, did you ever smoke, or were you exposed to asbestos? If yes, do not buy any beta-carotene supplement at any dose. If no, you still do not need one. Get vitamin A from food and skip the capsule.
In two large trials totalling 47,000 smokers, beta-carotene capsules at 20-30 mg per day increased lung cancer rather than preventing it. The harm signal persisted six years after they stopped taking it. For healthy non-smoking adults, every primary-prevention RCT has come back negative. The food version works because of conversion brakes the capsule bypasses.
Time to act: 30 seconds — check your multivitamin label for beta-carotene.
The Verdict

Eat your carrots and liver. Skip the capsule. If you ever smoked, that "skip" is a hard rule.

Your body uses an enzyme called BCMO1 to convert plant beta-carotene into vitamin A in the gut, and it deliberately slows the conversion down when stores are full. That brake is the safety feature that makes carrots a safe lifelong food. A beta-carotene capsule bypasses the brake entirely. The IOM conversion factors record this in numbers: 12 micrograms of dietary beta-carotene equals 1 microgram of vitamin A activity, but only 2 micrograms of supplemental beta-carotene does. The 12:1 vs 2:1 gap is not a typo — it is the consequence of the conversion brake. In two large trials those un-braked doses caused more lung cancer in smokers, not less.

  1. The VerdictBeta-carotene capsules at 20-30 mg per day increased lung cancer in 47,000 smokers across two independent trials. The harm signal persisted for years after the supplementation stopped.
  2. What Most People Get WrongPeople assume "natural antioxidant" plus "dietary inverse association with disease" means the supplement does the same thing. It does not. The food version works because of conversion brakes the capsule bypasses.
  3. The Protocol in Plain EnglishAdults: 900 micrograms vitamin A per day for men, 700 for women, from food — about one egg yolk, a serving of dairy, and a cooked orange or leafy vegetable with fat. No supplement needed.

Best For

Children in vitamin-A-deficient regions on WHO programs. Diagnosed AMD patients on the AREDS2 formula with beta-carotene removed, ophthalmologist-supervised. Cystic fibrosis patients with pancreatic insufficiency, specialist-supervised. None of these are retail consumer indications.

Skip If

You are a healthy adult buying for "antioxidant", "immune", "anti-aging", "skin", "longevity", or "cancer prevention". Skip absolutely if you smoke or ever smoked. Skip preformed retinol over 10,000 IU per day if pregnant. Skip "mixed carotenoid complex" capsules. Eat carrots and leafy greens with fat instead.

Keep scrolling for the protocol, safety profile, and full evidence

The Protocol

Vitamin A protocol

Vitamin A and the carotenoids are fat-soluble — absorption requires bile acids and dietary fat. The recommended daily allowance is 900 micrograms retinol activity equivalents per day for adult men and 700 micrograms for adult women. In food-secure adult populations, RDA is reachable through diet alone. Supplementation is indication-specific, not optimization.

PopulationDoseTiming / FormSource
Healthy adult male (18-65), food-secureRDA 900 µg RAE/d (≈3,000 IU equivalents) from diet aloneWith fat-containing meals · liver / dairy / eggs / oily fish / orange & leafy vegetablesIOM 2000 [cite-unverified]; NIH ODS
Healthy adult female (non-pregnant)RDA 700 µg RAE/d (≈2,333 IU) from diet aloneWith fat · foodIOM 2000 [cite-unverified]
Pregnant womenRDA 770 µg RAE/d; prenatal vitamin should use β-carotene rather than high-dose preformed retinol; hard ceiling 10,000 IU/d preformed retinolWith fat · β-carotene + dietary retinolIOM 2000 [cite-unverified]; NIH ODS
Lactating women (food-secure)RDA 1,300 µg RAE/d from dietWith fat · foodCochrane PMID 27012320 — insufficient evidence for routine VAS in non-deficient women
Lactating women (VAD-prevalent LMIC regions)Single high-dose VAS per WHO programPer WHO protocol · retinyl palmitateCochrane PMID 27012320 (GRADE low)
Children 6 mo - 5 y in VAD-prevalent LMIC regions100,000 IU 6-monthly (6-11 mo); 200,000 IU 4-6 monthly (12-59 mo)Per WHO schedule · retinyl palmitate oil · periodic high-doseWHO; Imdad Cochrane 2022 PMID 35294044
Cystic fibrosis with pancreatic insufficiencyPer CF center protocol (5,000-10,000 IU/d typical, individualised)With pancreatic enzyme replacement + fat · water-miscible retinyl preparationCochrane PMID 30091146
AMD patients (intermediate / advanced unilateral)AREDS2: lutein 10 mg + zeaxanthin 2 mg + vit C 500 mg + vit E 400 IU + zinc 80 mg + cupric oxide 2 mg (β-carotene EXCLUDED)Daily, with food, indefinite, under ophthalmologist supervisionCochrane AMD 2023 PMID 37702300
Heavy smokers / former smokers / asbestos-exposedNOT INDICATED for β-carotene supplementation at any doseAvoid any isolated β-carotene supplement; if AMD diagnosed, use AREDS2 (no β-carotene)ATBC [cite-unverified]; CARET [cite-unverified]; USPSTF 2022 PMID 35727272
Healthy adults seeking "antioxidant" / "immune" / "anti-aging" / "skin" / "cancer prevention"NOT INDICATED — diet-only adequacy is the evidence-based actionNo isolated supplement adds documented benefitBjelakovic Cochrane 2012 PMID 22419320; USPSTF 2022

Forms Comparison

Whole food β-carotene ⭐
~10-20% raw · ~30-50% cooked + fat
Recommended source for healthy adults. Self-limiting via BCMO1 brake. 12:1 RAE conversion.
£0 (already in varied diet)
Whole food preformed retinol ⭐
~70-90% with dietary fat
Liver, dairy, eggs, oily fish. Direct retinol supply within food matrix; UL ceiling not reached at normal intake.
£0 (varied diet)
Synthetic β-carotene capsule
~50% · 2:1 RAE
Not recommended for healthy adults. Bypasses BCMO1 brake. Smoker harm signal at 20-30 mg/d.
£8-20/mo · skip
Retinyl palmitate (preformed supplement)
~70-90% with fat
Cheapest preformed form. CF / WHO VAS programs. Hepatotoxicity ceiling ≥10,000 IU/d adults chronic.
£6-15/mo · indication-restricted
Mixed carotenoid complex / full-spectrum
Variable per carotenoid
Marketing position. No outcome RCT shows superiority over diet.
£15-40/mo · skip
AREDS2 (β-carotene EXCLUDED)
Standardised ophthalmology formula
Diagnosed AMD only, ophthalmology-supervised. Lutein + zeaxanthin replaced β-carotene; matched outcomes, removed smoker harm.
£20-45/mo · clinician-led
Red palm oil (food)
Good in palm-oil cooking matrix
VAD prevention in LMIC where palm oil is dietary staple. RR 0.55 VAD reduction in 9-RCT MA.
Low (regional food)
Cod liver oil
Food-derived preformed retinol + vit D + omega-3
~3,000-5,000 IU/d preformed retinol typical. Watch the cumulative retinol if combined with multivitamin.
£8-15/mo

Absorption Tips

  • Cook your vegetables — heat releases carotenoids from the cellular matrix and triples bioavailability vs raw
  • Eat with fat — olive oil drizzle on greens, butter on sweet potato, avocado in the salad, oily fish, nuts
  • Zinc deficiency impairs retinol-binding protein synthesis; chronic low zinc can produce functional vit A deficiency even with adequate stores
  • Fat-blockade drugs (orlistat) and bile-acid sequestrants (cholestyramine, colestipol) reduce fat-soluble vitamin absorption — separate timing or specialist review
  • Severe protein malnutrition limits transport via retinol-binding protein

Safety & Interactions

Vitamin A safety profile

Vitamin A safety is dose- and form-stratified. The two highest-severity signals are isotretinoin co-administration (additive toxicity, hard contraindication) and β-carotene in smokers (lung cancer harm signal, hard contraindication). Pregnancy adds a teratogenic ceiling above 10,000 IU/d preformed retinol.

Drug & Population Interactions

Isotretinoin / Acitretin / Bexarotene (systemic retinoids) SEVERE

Additive vitamin A toxicity — hepatotoxicity, hypervitaminosis A, intracranial hypertension, teratogenicity, mucocutaneous toxicity. Absolute contraindication on retinol-supplement co-administration. Isotretinoin is pregnancy category X.

Current / former smokers + asbestos-exposed × β-carotene SEVERE

β-carotene 20-30 mg/d increased lung cancer incidence in ATBC (RR 1.18, N=29,133 male smokers) and CARET (RR 1.28 incidence, RR 1.17 all-cause mortality, N=18,314 smokers / former smokers / asbestos-exposed). CARET stopped early for harm. Harm persisted 6 years post-intervention.

Pregnancy (first trimester) × preformed retinol >10,000 IU/d SEVERE

Teratogenic — craniofacial, cardiac, CNS malformations documented. Hard ceiling 10,000 IU/d preformed retinol in pregnancy. β-carotene does not share this teratogenic risk because BCMO1 down-regulation prevents retinoid overload from carotenoid sources.

Warfarin and vitamin K antagonists Moderate

High-dose preformed vit A can potentiate hepatic vitamin K antagonism. Avoid vit A supplementation without INR-monitoring protocol from prescriber.

Pre-existing liver disease × preformed retinol Moderate-Severe

Hepatic stellate cells store >50% of body vit A. Cirrhosis, hepatitis, alcoholic liver disease amplify vit A hepatotoxicity at doses safe in healthy livers. Avoid preformed retinol supplementation.

Tetracycline / doxycycline / minocycline × high-dose vit A Moderate

Risk of benign intracranial hypertension (pseudotumour cerebri). Avoid the combination.

Postmenopausal women on bisphosphonates × chronic preformed retinol ≥3,000 µg RAE/d Moderate

Inverse association with bone-mineral density and increased hip-fracture risk in observational data. Mechanism: osteoclast activation. Avoid chronic high-dose preformed retinol.

Orlistat & bile-acid sequestrants (cholestyramine, colestipol) Moderate

Fat-blockade and bile-acid sequestration reduce fat-soluble vitamin absorption; can produce iatrogenic vit A deficiency over chronic use. Separate timing; consider water-miscible retinyl preparation if clinician-indicated.

Contraindicated Populations

Upper Limits

Tolerable Upper Intake Level (preformed retinol, adult): 3,000 µg RAE/d (10,000 IU). β-carotene has no formal UL — BCMO1 down-regulation provides self-limiting protection — but ATBC / CARET dosing 20-30 mg/d in smokers establishes a practical harm threshold well below absence-of-UL. Pregnancy hard ceiling: 10,000 IU/d preformed retinol. Acute toxicity: single dose >300,000 IU adult (folk-medicine fish-liver / bear-liver poisoning). Chronic toxicity: ≥25,000 IU/d preformed retinol within weeks-months. (IOM 2000 [cite-unverified]; NIH ODS)

Conviction · LOW overall — endpoint-stratified
Healthy-adult prevention DEBUNKED + smoker β-carotene NET HARM HIGH-supported. LMIC pediatric VAS HIGH. AMD progression diagnosed MODERATE (AREDS2, β-carotene EXCLUDED). Pregnancy teratogenic ceiling HIGH-supported. Dietary intake the only HIGH-supported consumer action.
What would change this?

A pre-registered, independent, double-blind RCT (N ≥ 5,000) of healthy adult never-smokers given β-carotene 20-30 mg/d × ≥5 years with all-cause mortality + lung cancer incidence as PRIMARY endpoints, showing null harm signal — would upgrade the all-population conviction from "moderate net harm" to "neutral in never-smokers." The smoker harm conviction (ATBC + CARET, N ≈ 47,000 combined, persistent post-intervention) would not be touched. A modern AREDS3 trial returning β-carotene in a never-smoker subgroup showing AMD-progression benefit not matched by lutein/zeaxanthin — would change ophthalmology default. Unlikely. A causal Mendelian randomisation finding showing genetic β-carotene status causally reduces a hard clinical endpoint — would re-open the supplemental hypothesis. Current MR (Wang 2024 PMID 38566359) shows null causal effects, consistent with the supplemental-RCT null. Field STABLE per Perplexity preflight; no in-flight trial is positioned to overturn the prevention conviction direction in 3-5 y.

Worth Your Money?

Estimated weekly cost
£2-5/week for a β-carotene or preformed retinol capsule that is not indicated for healthy adults. £0/week if you cover RDA from food. £5-12/week for an AREDS2 formula (β-carotene EXCLUDED) if your ophthalmologist has prescribed it for diagnosed AMD.
Worth it if
Your ophthalmologist has prescribed AREDS2 for diagnosed intermediate or advanced unilateral AMD (β-carotene REMOVED), or your specialist is supervising supplementation for cystic fibrosis with pancreatic insufficiency, or you live in a WHO-defined VAD-prevalent region where pediatric or post-partum VAS is a public-health program.
Lower priority if
You are a healthy adult buying for "antioxidant" / "immune" / "anti-aging" / "skin" / "longevity" / "cancer prevention". Your money goes further first on vit D (#4), magnesium (#5), omega-3 (#6), creatine (#1), or a competent multivitamin only if dietary inadequacy is documented (#14). For most healthy adults, an extra serving of cooked leafy greens with olive oil beats every β-carotene capsule on the shelf.
Skip for healthy adults Conditional yes for AREDS2 (β-carotene EXCLUDED) + CF specialist + LMIC WHO program
Get the next review free
Want to stop wasting money on supplements that do not work and quietly avoid the ones that cause harm? The Verdict reviews one supplement every week — evidence-graded, honest, free.
Subscribe to The Verdict

Sources

  1. Imdad A et al. (2022). Vitamin A supplementation for preventing morbidity and mortality in children from six months to five years of age. Cochrane Database Syst Rev. PMID 35294044. 47 trials, N≈1.2M children in LMIC; RR 0.88 all-cause mortality.
  2. O'Connor EA et al. (2022). Vitamin and Mineral Supplements for the Primary Prevention of CVD and Cancer: USPSTF Evidence Report. JAMA. PMID 35727272. β-carotene → increased lung cancer + CV mortality; vit A / multivitamins no primary-prevention benefit.
  3. Evans JR, Lawrenson JG (2023). Antioxidant vitamin and mineral supplements for slowing the progression of age-related macular degeneration. Cochrane Database Syst Rev. PMID 37702300. AREDS2 (β-carotene replaced with lutein + zeaxanthin) is current ophthalmology standard.
  4. Bjelakovic G et al. (2012). Antioxidant supplements for prevention of mortality in healthy participants and patients with various diseases. Cochrane Database Syst Rev. PMID 22419320. β-carotene RR 1.05, vit A RR 1.16 in low-RoB trials.
  5. Bjelakovic G et al. (2013). Meta-regression + trial sequential analyses of β-carotene, vit A, and vit E on all-cause mortality. PLOS ONE. PMID 24040282. TSA ruled out "no harm" hypothesis.
  6. Bjelakovic G et al. (2004). Antioxidant supplements for prevention of gastrointestinal cancers. Lancet. PMID 15464182. β-carotene + vit A combined RR 1.29 mortality.
  7. Rutjes AW et al. (2018). Vitamin and mineral supplementation for maintaining cognitive function. Cochrane Database Syst Rev. PMID 30556597. No cognitive maintenance benefit.
  8. O'Neil SE et al. (2018). Vitamin A and β-carotene supplementation for cystic fibrosis. Cochrane Database Syst Rev. PMID 30091146. Insufficient RCT evidence for dose; supplementation justified by deficiency biology.
  9. McCauley ME et al. (2012). Vitamin A and carotenoids during pregnancy and maternal, neonatal and infant health outcomes. Paediatr Perinat Epidemiol. PMID 22742601. GRADE-variable per outcome.
  10. Wang J et al. (2024). Diet-derived circulating antioxidants, periodontitis and dental caries: A Mendelian randomization study. J Periodontal Res. PMID 38566359. Null causal effect of circulating β-carotene, retinol, lycopene.
  11. Mullan K et al. (2018). Plasma Antioxidant Status in Patients with Alzheimer's Disease. J Alzheimers Dis. PMID 29439339. Lower plasma α/β-carotene, lycopene, lutein, vit A/C/E in AD.
  12. Dong S et al. (2017). The Effect of Red Palm Oil on Vitamin A Deficiency: A Meta-Analysis of RCTs. Nutrients. PMID 29186779. VAD RR 0.55.
  13. Vlieg-Boerstra B et al. (2022). Nutrient supplementation for prevention of viral respiratory tract infections in healthy subjects. Allergy. PMID 34626488.
  14. Yao N et al. (2021). Carotenoids and subjects with overweight or obesity. Food Funct. PMID 33977977. Observational MA — diet quality marker.
  15. Ashor AW et al. (2014). Antioxidant vitamin supplementation reduces arterial stiffness. J Nutr. PMID 25098780. Biomarker only; no CV outcome benefit.
  16. Psaltopoulou T et al. (2018). Micronutrient Intake and Risk of Hematological Malignancies in Adults. Nutr Cancer. PMID 30288994. Inconsistent direction.
  17. Oliveira-Menegozzo JM et al. (2016). Vitamin A supplementation for postpartum women. Cochrane Database Syst Rev. PMID 27012320. GRADE low; insufficient evidence for routine VAS in non-deficient women.
  18. ATBC Study Group (1994). Effect of vitamin E and beta carotene on incidence of lung cancer in male smokers. N Engl J Med. [cite-unverified] N=29,133; β-carotene 20 mg/d × 5-8y; lung cancer +18%.
  19. Omenn GS et al. (1996). CARET trial — effects of beta carotene and vitamin A on lung cancer and CVD. N Engl J Med. [cite-unverified] N=18,314; β-carotene 30 mg/d + retinyl palmitate 25,000 IU/d × ~4y; lung cancer +28%, all-cause mortality +17%, stopped early for harm.
  20. Goodman GE et al. (2004). CARET 6-year post-intervention follow-up. J Natl Cancer Inst. [cite-unverified] Harm persisted.
  21. AREDS Research Group (2001). AREDS report no. 8 — antioxidants + zinc for AMD. Arch Ophthalmol. [cite-unverified] Original AMD formulation, superseded by AREDS2.
  22. AREDS2 Research Group (2013). Lutein + zeaxanthin and omega-3 for AMD. JAMA. [cite-unverified] β-carotene replaced; AMD outcomes maintained; smoker harm removed.
  23. Institute of Medicine (2001/2000). Dietary Reference Intakes for Vitamin A. National Academies Press. [cite-unverified] RDA 900 / 700 µg RAE/d; UL 3,000 µg RAE/d preformed retinol; pregnancy ceiling 10,000 IU/d; RAE conversion 1:1 retinol / 12:1 dietary β-carotene / 2:1 supplemental β-carotene.

Get the complete dosing protocol

Evidence-scored dosing, timing, forms, and who should skip it. One page, no fluff.

Get the protocol

Related free research

Supplements
Plant Sterols & Stanols — The Verdict
 Supplements
Exogenous Ketones — The Verdict
 Supplements
Serrapeptase — Does the Silkworm Enzyme Actually Work?

There are 424 more inside

Conviction-scored verdicts on supplements, nutrition, training, physio, and recovery.

Explore all Get weekly verdicts