β-Carotene (Provitamin A Carotenoid) · Evidence-First Educational Reference

§1.1 · Chemical definition

β-Carotene (chemical name β,β-carotene · molecular formula C₄₀H₅₆ · molecular weight 536.87 g/mol) is a natural fat-soluble pigment in the carotenoid family, specifically in the carotene subgroup (carotenoids that contain only carbon and hydrogen, no oxygen). Its molecular structure has two defining features:

• A 22-carbon polyene chain with 11 conjugated carbon-carbon double bonds — the physical-chemical basis for both its strong yellow-orange color and its prominent singlet-oxygen quenching activity.
• Two β-ionone rings, one at each end of the polyene chain — this symmetric double-β-ring structure makes β-carotene theoretically the most efficient provitamin A carotenoid, because central symmetric cleavage can theoretically produce two retinaldehyde molecules. Other provitamin A carotenoids (α-carotene, γ-carotene, β-cryptoxanthin) have either one β-ionone ring or one β-ring with a structural variation, giving them roughly half the theoretical provitamin A activity of β-carotene.

§1.2 · Geometric isomers

β-Carotene exists in multiple geometric isomers that differ in bioavailability and tissue distribution because of cis / trans configuration at specific double bonds. The principal forms:

Key fact: Natural sources (particularly Dunaliella) supply a mixture of all-trans plus 9-cis; industrial chemical synthesis produces an all-trans dominant product (>95% all-trans). The two forms differ in bioavailability and tissue distribution, but whether that difference is sufficient to change long-term clinical endpoints — particularly the CARET / ATBC negative findings — remains hypothesis-level, with no large RCT directly validating the distinction. See §6.4.

§1.3 · Provitamin A — not preformed retinol

β-Carotene itself has no direct retinol activity. It must undergo central symmetric cleavage in the small-intestine enterocyte by β-carotene 15,15′-monooxygenase 1 (BCO1, formerly BCMO1), theoretically yielding two retinaldehyde molecules, which are then reduced to retinol — vitamin A.

Official RAE (Retinol Activity Equivalents) conversion (IOM 2001 and EFSA 2015 current standards):

1 μg preformed retinol = 1 μg RAE
1 μg dietary β-carotene = 0.083 μg RAE
(equivalently: 12 μg dietary β-carotene ≈ 1 μg RAE)

Supplement form (β-carotene dissolved in oil and encapsulated) is more efficiently absorbed:
1 μg supplement β-carotene ≈ 0.5 μg RAE
(equivalently: 2 μg supplement β-carotene ≈ 1 μg RAE)

Other provitamin A carotenoids (α-carotene / β-cryptoxanthin):
24 μg dietary ≈ 1 μg RAE

Educational principle (no exaggeration permitted): β-carotene is not an "equivalent plant substitute for retinol." The 12:1 dietary conversion means that 12 mg of dietary β-carotene supplies only 1 mg of retinol activity. For strict vegan / plant-based populations it is the only dietary vitamin A source; for the general population it is complementary, not primary.

§2.1 · Natural food sources (mg β-carotene per 100 g edible portion · USDA FoodData Central ranges)

Typical Western adult dietary β-carotene intake: 1.5-3 mg/day (NHANES United States survey, EFSA European survey data).

Typical East Asian dietary intake: 2-4 mg/day (higher vegetable consumption, though chlorophyll-masked dark-green-leafy sources are less recognized by consumers).

§2.2 · Supplement forms and typical doses

Order-of-magnitude reminder: Supplement doses and dietary intakes differ by an order of magnitude — a single 15 mg capsule contains roughly as much β-carotene as 3-5 carrots. The directional evidence for clinical endpoints differs between dietary and supplement intake (compare §4 against §5 / §6), and this is the most important educational boundary for the ingredient.

§3.1 · Absorption

β-Carotene is released from the food matrix, solubilized into bile-secreted mixed micelles, and actively transported into small-intestine enterocytes (predominantly duodenum and upper jejunum) via the SR-BI (scavenger receptor class B type 1) transporter.

Factors influencing the absorption rate:

• Co-ingested dietary fat: when the meal contains less than 3 g of fat, absorption is essentially zero; at ≥10-15 g of fat per meal the absorption rate rises substantially (a feature shared by all fat-soluble vitamins).
• Food matrix: free form (e.g. supplement oil-suspension) > cooked and cell-wall-disrupted vegetables > raw whole-cell-wall vegetables. For example, raw whole carrot chewed without further processing has an absorption rate of only 1-3%; cooked, minced, and oil-prepared carrot can reach 30-40%.
• Dose saturation: the higher the single dose, the lower the per-unit absorption rate (typical saturation: supplement form 6-33%, dietary form 5-15%).
• Co-ingestion of other carotenoids: β-carotene, lutein, zeaxanthin, and lycopene share the SR-BI transport channel — single high-dose supplementation competitively inhibits absorption of the other carotenoids. This is one of the core arguments for why single-high-dose chronic supplementation is non-physiological (see §10).

§3.2 · BCO1 central symmetric cleavage and conversion

Once inside the enterocyte, roughly 60-70% of β-carotene is acted on by BCO1 (β-carotene 15,15′-monooxygenase 1) via central symmetric cleavage, theoretically yielding two retinaldehyde molecules. Retinaldehyde reductase then reduces them to retinol, and the retinol is esterified to retinyl ester for incorporation into chylomicrons and lymphatic uptake. The remaining 30-40% enters chylomicrons as intact β-carotene for systemic circulation and tissue distribution.

Critical negative-feedback mechanism: BCO1 expression and activity are negatively regulated by vitamin A status — when hepatic retinol stores are adequate, the intestinal transcription factor ISX (intestine-specific homeobox) is induced and suppresses BCO1 transcription. This negative feedback explains why high oral doses of β-carotene do not cause hypervitaminosis A — the core safety advantage of β-carotene relative to preformed retinol.

§3.3 · BCO1 genetic polymorphism (a substantial inter-individual variation)

Approximately fifteen years of research have revealed that multiple common SNPs in the BCO1 gene meaningfully alter β-carotene → retinaldehyde conversion efficiency:

• R267S (rs12934922) and A379V (rs7501331) double-SNP carriers: Leung and colleagues in female volunteers showed a substantial reduction in postprandial plasma retinyl-ester / β-carotene ratio in double-SNP carriers, implying an enzymatic-activity reduction of 30-40% (Leung WC et al. FASEB J 2009; 23:1041 · PMID 19103647).
• Upstream regulatory SNPs (rs6420424 / rs11645428 / rs6564851): Lietz and colleagues in a larger female cohort confirmed that upstream SNPs reduce BCO1 catalytic activity by 48-59% (Lietz G et al. J Nutr 2012; 142:161S · PMID 22113863).

Clinical implication: roughly 30-45% of the population are "low converters" (poor responders). For this subset:

• Relying on β-carotene as the sole vitamin A source (e.g. strict vegan diets) creates a real risk of subclinical vitamin A deficiency;
• The "vitamin A equivalent" assumption when supplementing β-carotene needs to be individualized.

Educational takeaway: the 12 μg β-carotene = 1 μg RAE relationship is a population mean, not an individual constant. BCO1 polymorphism testing is not yet a routine clinical screen, but it explains why serum retinol responses to the same dietary or supplement dose can vary by 4- to 5-fold between individuals.

§4.1 · Dietary β-carotene and lung cancer risk (inverse association)

Several large prospective cohort studies consistently show that dietary β-carotene intake from vegetables and fruit is inversely associated with lung cancer risk:

• High-intake group (highest tertile or quintile vs lowest) lung cancer relative risk approximately 0.7-0.8 (a 20-30% risk reduction).
• The association is most stable in non-smokers; in smokers the observational association is weaker or inconsistent.
• The association strength is lower than that for other vegetable / fruit components (lutein, lycopene) — suggesting β-carotene itself is probably not a sufficient protective factor and is more likely a biomarker for an overall healthy dietary pattern (high vegetable / fruit consumption) rather than an independently separable active component.

§4.2 · Dietary β-carotene and cardiovascular disease

The evidence pattern is similar to that for lung cancer: observational studies show a modest inverse association between dietary β-carotene intake and coronary heart disease, stroke, and cardiovascular mortality, but the single-nutrient causal interpretation is directly challenged by the negative findings from ATBC, PHS, and WHS RCTs — i.e. "more dietary intake is associated with better outcomes, but supplement intervention is either ineffective or harmful" (see §6).

§4.3 · Dietary vs supplement — the scientific consensus statement

The NIH Office of Dietary Supplements states in its β-carotene fact sheet (paraphrased here, with the original framing preserved):

"Diets high in foods containing carotenoids (including β-carotene) are associated with lower risk of several chronic diseases. However, high-dose β-carotene supplements have not shown the same benefit, and in smokers and asbestos-exposed workers, supplements have caused harm. The protective effect appears to come from the whole food matrix rather than from β-carotene alone."

This is the first framing principle for any educational writing on the ingredient:

★ Whole-food protective ≠ Single-nutrient supplement protective ★

§5.1 · ATBC trial (Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study · 1994)

• PMID 8127329 · The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group · N Engl J Med 1994; 330(15):1029-1035.
• Design: randomized, double-blind, placebo-controlled, 2×2 factorial · southwestern Finland · 29,133 men aged 50-69 who smoked at least 5 cigarettes per day for at least 26 years.
• Four intervention arms: α-tocopherol 50 mg/day · β-carotene (synthetic all-trans) 20 mg/day · combination · placebo.
• Duration: 5-8 years (median 6.1 years).
• Primary endpoint: lung cancer incidence.
• Key results:
  • β-carotene arm showed a +18% increase in lung cancer incidence (NEJM original RR 1.18, 95% CI 1.03-1.36, p=0.01) versus participants not assigned to β-carotene.
  • +8% increase in all-cause mortality (RR 1.08, 95% CI 1.01-1.16).
  • α-tocopherol did not significantly change lung cancer incidence but showed a 32% reduction in prostate cancer incidence (see the tocopherols sub-page).
• The trial was not terminated early (unlike CARET), but the 1994 NEJM publication established the clinical recommendation that "β-carotene should not be used as a cancer chemoprevention agent in heavy smokers."

§5.2 · CARET trial (Beta-Carotene and Retinol Efficacy Trial · 1996)

• PMID 8602180 · Omenn GS, Goodman GE, Thornquist MD, et al. · N Engl J Med 1996; 334(18):1150-1155.
• Design: multi-center, randomized, double-blind, placebo-controlled · 6 US research centers · 18,314 individuals at high lung cancer risk (heavy smokers, former smokers, and individuals with occupational asbestos exposure).
• Intervention: β-carotene (synthetic all-trans) 30 mg/day plus retinyl palmitate 25,000 IU/day vs placebo.
• Duration: planned 5 years · actually terminated in year 4 because of clear harm, 21 months ahead of schedule.
• Key results:
  • +28% increase in lung cancer incidence (RR 1.28, 95% CI 1.04-1.57).
  • +17% increase in all-cause mortality (RR 1.17, 95% CI 1.03-1.33).
  • +26% increase in cardiovascular mortality.
• The Data and Safety Monitoring Board (DSMB) voted to terminate the trial in January 1996; the NEJM published the full results in May of the same year.

§5.3 · CARET same-year risk-factor analysis (clarification: NOT a 12-year follow-up)

• PMID 8901853 · Omenn GS, Goodman GE, Thornquist MD, et al. · J Natl Cancer Inst 1996; 88(21):1550-1559. "Risk factors for lung cancer and for intervention effects in CARET, the Beta-Carotene and Retinol Efficacy Trial."
• This paper was published in November of the same year as the CARET primary-results paper, and further analyzes lung cancer risk factors and intervention-effect subgroup distributions within the CARET cohort. It is NOT a long-term follow-up paper. (This topic has a long history of PMID misbinding — this page transparently discloses that earlier drafts had labeled this paper as a "12-year follow-up", which is incorrect and has been corrected against the actual PubMed record.)

§5.4 · Does the harm persist after stopping? Two long-term follow-ups

• CARET 6-year post-discontinuation follow-up · PMID 15572756 · Goodman GE, Thornquist MD, Balmes J, et al. · J Natl Cancer Inst 2004; 96(23):1743-1750. Conclusion: the harm signal (elevated lung cancer incidence and all-cause mortality) from β-carotene plus retinyl palmitate persisted 6 years after stopping supplementation — the magnitude attenuated but the direction did not change.
• ATBC 18-year post-discontinuation follow-up · PMID 24338499 · Virtamo J, Taylor PR, Kontto J, et al. · Int J Cancer 2014; 135(1):178-185. Conclusion: the β-carotene lung cancer risk signal continued to attenuate year by year over 18 years of follow-up but had not returned to baseline; the α-tocopherol prostate cancer protective signal also faded within 5-10 years.

§5.5 · Non-smoker-dominant population control evidence: PHS (Physicians' Health Study)

• PMID 8602179 · Hennekens CH, Buring JE, Manson JE, et al. · N Engl J Med 1996; 334(18):1145-1149. "Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease."
• Design: 22,071 US male physicians aged 40-84 (11% current smokers · 39% former smokers · 50% never smokers) · β-carotene 50 mg every other day (≈25 mg/day equivalent) · median follow-up 12 years.
• Results: no significant difference in total cancer incidence, cardiovascular events, or all-cause mortality. The current-smoker subgroup analysis did not replicate the ATBC / CARET harm signal — but the subgroup sample was small, the follow-up was shorter than ATBC, and the dose was lower than CARET.

Key interpretation: PHS provides neutral evidence that "long-term β-carotene supplementation at ≈25 mg/day in a non-smoker-dominant population shows neither clear harm nor clear benefit"; it cannot be used to refute the clear ATBC / CARET risk signal in the smoker / high-risk population.

§5.6 · Systematic reviews and the Cochrane position

• Druesne-Pecollo et al. 2010 meta-analysis · PMID 19876916 · Int J Cancer 127:172-184. Pooled 9 RCTs (including ATBC, CARET, PHS):
  • High-dose β-carotene 20-30 mg/day overall lung cancer risk +16%;
  • Smoker subgroup +20%;
  • Gastric cancer risk overall +34%;
  • Asbestos-exposed subgroup +54%.
  • Conclusion: "β-carotene supplementation showed no protective effect in cancer prevention."
• Bjelakovic Cochrane 2012 · PMID 22419320 · Cochrane Database Syst Rev 2012(3):CD007176. Antioxidant supplements (β-carotene / vitamin A / vitamin C / vitamin E / selenium) and all-cause mortality Cochrane review:
  • β-carotene + vitamin A combination: RR 1.29 (95% CI 1.14-1.45) significantly increases all-cause mortality;
  • β-carotene + vitamin E combination: RR 1.10 (1.01-1.20) significantly increases all-cause mortality.
  • Cochrane explicit conclusion: β-carotene / vitamin A / vitamin E supplements may increase mortality and are not recommended for general-population preventive use.

§5.7 · Regulatory / public-health institution positions (5 authorities · convergent guidance)

§5.8 · Chapter summary

• Smokers / former smokers / asbestos-exposed workers: the risk signal from high-dose (≥20 mg/day) β-carotene supplements is consistent, replicated across studies, and persists after stopping.
• Non-smoker-dominant populations: high-dose supplements show no clear benefit (PHS primary analysis neutral · USPSTF Grade D).
• General-population preventive use: prefer dietary vegetables and fruit; high-dose supplements are not recommended (Cochrane 2012 · NIH-ODS · USPSTF).
• This is one of the most important "antioxidant nutrient myth refuted by RCTs" cases in the history of evidence-based medicine, alongside the vitamin E cardiovascular RCT negative findings and the SELECT prostate cancer negative findings.

§6.1 · WHS (Women's Health Study)

A WHS substudy in 39,876 women evaluated β-carotene 50 mg every other day for 2.1 years (short-term): no significant effect on cancer or cardiovascular events. The WHS β-carotene arm was discontinued early following the publication of ATBC / CARET results on the recommendation of the institutional review board. (For factual-checking rigor this page mentions the study narratively without assigning a specific PMID, avoiding the high risk of PMID misbinding among the multiple WHS primary analyses.)

§6.2 · Elderly immune RCT

Santos et al. 1996 (PMID 8901800) in a very small sample (n=21 elderly men) using β-carotene 50 mg every other day for 10-12 years suggested an improvement in NK cell activity · but the sample size is too small to constitute an independently generalizable conclusion.

§6.3 · Long-term cognitive PHS-II substudy

• PMID 17998490 · Grodstein F, Kang JH, Glynn RJ, Cook NR, Gaziano JM · Arch Intern Med 2007; 167(20):2184-2190.
• In 4,052 PHS male physicians on long-term supplementation (β-carotene 50 mg every other day · mean 18 years), total cognitive score and verbal memory were significantly better than placebo.
• However, short-term 1-year intervention was ineffective, and the finding applies only to a non-smoker-dominant population and cannot be extrapolated.

§6.4 · The natural 9-cis hypothesis (strictly controlled · Tier C · must not be over-stated)

Researchers have proposed the hypothesis that the synthetic all-trans β-carotene (>95% all-trans) used in CARET and ATBC lacks the approximately 40% 9-cis isomer present in Dunaliella natural product, and that 9-cis is selectively retained in liver and subcutaneous tissue and may have a different pro-/anti-oxidant balance. If this hypothesis were correct, then natural mixed-isomer products (Dunaliella-sourced) might not reproduce the lung cancer risk signal of synthetic all-trans in smokers.

§7.1 · AREDS1 (Age-Related Eye Disease Study · 2001) · The original antioxidant compound

• PMID 11594942 · AREDS Research Group · Arch Ophthalmol 2001; 119(10):1417-1436.
• Design: 3,640 patients with intermediate age-related macular degeneration (AMD) · 2×2 factorial randomization:
  • ① Antioxidant compound (β-carotene 15 mg + vitamin C 500 mg + vitamin E 400 IU);
  • ② Zinc (80 mg) + copper (2 mg);
  • ③ Combination of both;
  • ④ Placebo.
• Result: the combination arm (antioxidant compound + zinc) reduced the risk of progression from intermediate to advanced AMD by 25% (5-year follow-up).
• Clinical significance: AREDS1 established the clinical evidence base for the "antioxidant + zinc" compound supplementation as the standard intervention for intermediate AMD.
• Important caveat: AREDS1 enrollment excluded current smokers (based on the already-published ATBC / CARET results), but former smokers continued to be enrolled and subsequent subgroup analyses suggested that the β-carotene lung cancer risk signal in former smokers persisted within the study arms.

§7.2 · AREDS2 (2013) · β-Carotene replaced by lutein + zeaxanthin

• PMID 23644932 · AREDS2 Research Group · JAMA 2013; 309(19):2005-2015.
• Design: 4,203 AMD patients · 2×2 factorial randomization:
  • First dimension: on top of the AREDS1 base formula, add ① lutein 10 mg + zeaxanthin 2 mg; ② ω-3 LCPUFA (DHA 350 mg + EPA 650 mg); ③ both; ④ neither.
  • Second dimension (separate concurrent randomization): ① remove β-carotene (substitute additional lutein / zeaxanthin); ② retain β-carotene 15 mg.
• Key results:
  • On top of the original AREDS formula, lutein 10 mg + zeaxanthin 2 mg effectively substitutes for β-carotene — AMD progression protection is equivalent or modestly better.
  • In sub-arms retaining β-carotene, former smokers showed a significantly elevated lung cancer incidence (consistent with the ATBC / CARET signal).
  • AREDS2 therefore formally recommends: all AMD patients should switch to the lutein / zeaxanthin substituted version (the "AREDS2 formula").
• Clinical implication: from 2013 onward, the American Academy of Ophthalmology (AAO) formally adopted the AREDS2 formula (lutein + zeaxanthin · without β-carotene) as the standard recommendation for intermediate AMD. β-carotene exited mainstream ophthalmology AMD intervention.

§7.3 · Educational implication of the paradigm shift

AREDS1 (2001)              →    AREDS2 (2013)
─────────────────────────────────────────────────────────
β-carotene 15 mg            →    Lutein 10 mg + zeaxanthin 2 mg
(provitamin A carotene)         (macular pigment · selective foveal deposition)
─────────────────────────────────────────────────────────
Driver                      →    Former smoker lung cancer signal + equivalent or superior alternative
Scientific posture          →    "Evidence replacement" rather than "refuting the prior generation of research"

This paradigm shift exemplifies the core spirit of evidence-based medicine: when a long-term safety signal is clear and an equivalent or superior alternative exists, mainstream guidance should actively update. The lesson for consumer education is that evidence-based nutrition positions evolve with long-term data — this is scientific honesty, not "experts changing their minds again."

See sibling pages: Lutein and Zeaxanthin.

§8.1 · The β-carotene safety advantage relative to preformed retinol

Because of the vitamin A status negative-feedback regulation of BCO1, oral high-dose β-carotene almost never causes hypervitaminosis A (in contrast to preformed retinol / retinyl ester intake). This is the core safety advantage of β-carotene as a provitamin A source.

§8.2 · Carotenodermia (the yellowing-skin condition · benign and reversible)

• Definition: long-term high-dose β-carotene intake (typically ≥30 mg/day for weeks to months) can result in substantial β-carotene deposition in the stratum corneum and subcutaneous fat, producing visible orange-yellow staining at sites with thicker stratum corneum (palms, soles, nose, nasolabial fold).
• Clinical significance: entirely benign, fully reversible (resolves within 2-6 weeks of stopping supplementation), with no pathological consequence. Often misidentified as jaundice — the differentiation point: carotenodermia does NOT affect the sclera (jaundice presents first in the sclera because bilirubin binds elastin in the sclera; β-carotene does not bind elastin).
• Public-health note: infants and toddlers consuming large amounts of carrot puree, sweet potato puree, or pumpkin puree during the complementary feeding period often develop an "orange-yellow" skin tint — this is typical carotenodermia and requires no intervention.

§8.3 · Tolerable Upper Intake Level (UL)

§8.4 · Drug interactions

• Orlistat: blocks fat absorption, simultaneously lowering β-carotene and other fat-soluble vitamin (A / D / E / K) absorption.
• Bile-acid sequestrants (cholestyramine / colestipol): reduce fat-soluble vitamin absorption via a similar mechanism.
• Concurrent high-dose preformed retinol / retinyl ester: BCO1 negative feedback is suppressed by preformed vitamin A, further reducing β-carotene conversion (physiological, not a harmful interaction).
• High-dose vitamin E: observational studies suggest a mild reduction in serum β-carotene levels; mechanism unclear; clinical significance limited.

§8.5 · Pregnancy and breastfeeding

Because of the BCO1 self-limiting mechanism, β-carotene does not carry the teratogenic risk associated with preformed retinol (particularly retinyl ester) in pregnancy (retinol ≥10,000 IU/day in the first trimester can be teratogenic · see the vitamin A reference page).

• Dietary β-carotene intake: safe.
• Supplement form: recommended dose ≤6 mg/day.

§8.6 · Children

WHO / UNICEF use β-carotene or preformed VA capsules widely in vitamin A deficiency (VAD) supplementation programs in high-prevalence low-income countries, with age-adjusted dosing (Imdad A et al. Cochrane Database Syst Rev 2017 · PMID 28282701; Mayo-Wilson E et al. 2011 · PMID 21868478). Dietary β-carotene (particularly sweet potato, pumpkin, and orange-fleshed mango) is the cornerstone of VAD prevention in low-income settings.

§8.7 · Photoprotection background (Köpcke 2008 meta-analysis · 7 RCTs)

A meta-analysis by Köpcke W and Krutmann J 2008 (PMID 18086246) pooled 7 randomized controlled trials on β-carotene oral supplementation and minimal erythema dose (MED) — a small but consistent photoprotection signal was observed after at least 10 weeks of supplementation; the effect size is modest and is best treated as adjunctive to topical sunscreen, not a substitute. Earlier drafts erroneously reported "12 RCTs"; the verified PubMed record is 7 RCTs and this page uses that correct figure. See also the broader carotenoid photoprotection review by Stahl W and Sies H 2012 (PMID 23053552).

§9.1 · China (NMPA + GB standards)

• Food additive: GB 2760-2024 allows β-carotene as a colorant in multiple food categories.
• Nutrient fortifier: GB 14880-2012 allows β-carotene as a vitamin A fortifier (counted by RAE equivalents).
• Novel food ingredient: Dunaliella salina has been approved as a novel food ingredient.
• Health food (Blue Cap registration pathway): may be registered for "vitamin A supplementation" or "immune support" functional claims.

§9.2 · United States (FDA + DSHEA)

• GRAS food ingredient: 21 CFR 184.1245 lists β-carotene as GRAS.
• Color additive: 21 CFR 73.95 / 73.1095 / 73.2095 (permanently certification-exempt for food / drug / cosmetic use since 1963).
• Dietary supplement: regulated under DSHEA 1994; may be labeled as a vitamin A source with %DV; may carry Structure / Function Claims (e.g. "supports immune health"), but may not make disease prevention or treatment claims.
• Smoker warning: FDA and NIH-ODS converge on requiring consumer-facing educational materials to clearly carry the smoker warning (not a mandatory label warning but a strong industry-self-regulatory norm).

§9.3 · Brazil (ANVISA)

• Food colorant: RDC 239/2018 allows β-carotene as a natural colorant.
• Nutrient fortification and dietary supplement: IN 28/2018 Anexo IV specifies the adult supplement maximum dose for vitamin A sources (β-carotene counted by 12:1 RAE equivalent).
• Smoker warning: ANVISA explicitly requires labels of high-dose β-carotene supplements to indicate that smokers should not use them, consistent with the ATBC / CARET historical data.
• Pregnancy / breastfeeding / smoker high-dose marketing restrictions: regulatory prohibition.

§10.1 · Priority pyramid

Top (best):  Dietary vegetable / fruit β-carotene intake
  ├─ Orange-red vegetables: carrot, sweet potato, pumpkin, orange-flesh cantaloupe
  ├─ Dark leafy greens: spinach, kale, collards (chlorophyll-masked but high content)
  └─ Cooking guidance: mincing + oil-based cooking substantially raises absorption
                       (from raw whole 1-3% to 30-40%)

Middle (specific contexts): low- to mid-dose supplementation
  ├─ Suitable population: strict vegan / plant-based individuals (sole vitamin A source)
  ├─ Recommended dose: 3-6 mg/day (do not exceed 15 mg/day)
  └─ Form preference: Dunaliella natural mixed isomers ≈ synthetic all-trans
                       (regulators treat them as equivalent · see §6.4)

Bottom (not recommended): high-dose supplements ≥20 mg/day
  ├─ Smoker absolute contraindication (ATBC / CARET / Cochrane convergent position)
  └─ General population shows no evidence-based benefit (USPSTF Grade D)

§10.2 · Population suitability matrix

§10.3 · Form selection reference

• Dunaliella natural mixed isomers (all-trans + 9-cis ≈ 1:1): positioned as "natural source" in marketing. The 9-cis hypothesis provides only a mechanistic differentiation narrative — note: this hypothesis has not been validated by large RCTs and should not be used as a safety claim basis (see §6.4).
• Synthetic all-trans (chemical synthesis or Blakeslea trispora fermentation): cost-effective, but lacks the natural-isomer diversity.
• Compound carotenoid formulations: β-carotene 3-6 mg + α-carotene + lutein + zeaxanthin + lycopene · mimics the dietary carotenoid network and avoids the single-high-dose competitive absorption inhibition (§3.1).
• AMD patients: use the AREDS2 formula (no β-carotene · with lutein + zeaxanthin), see §7.2.

§10.4 · Label-reading guidance

• Check whether the daily dose is ≥20 mg (smokers must avoid absolutely).
• Check whether other fat-soluble vitamins (A / D / E / K) appear together at high doses — be alert to the Bjelakovic 2012 Cochrane meta-analysis combination signals ("β-carotene + vitamin A" and "β-carotene + vitamin E" both increase all-cause mortality · §5.6).
• Check whether the label includes BCO1 SNP / individual variation guidance (leading self-regulating brands have begun adding this educational labeling).