Beta-Carotene for Vitamin A Deficiency

Beta-carotene can correct and prevent vitamin A deficiency, because the body turns it into vitamin A, but it is a slower and less predictable route than preformed vitamin A. In 510 Senegalese children aged 2-15 with deficiency-related changes on the surface of the eye, beta-carotene and retinyl palmitate were statistically equivalent: 50.0% and 51.2% had normal eye cytology 7 weeks later (Senegal equivalence trial, 1993). In Bangladeshi men on a low-vitamin A diet, 60 days of synthetic beta-carotene raised total-body vitamin A about as much as retinyl palmitate (Bangladesh stores trial, 2004).

We rate it 3 out of 5, a step below vitamin A for deficiency, because the pregnancy trials were weaker. In Nepal, weekly beta-carotene cut night blindness in pregnancy less than vitamin A did (relative risk 0.83, not significant, vs 0.62) (Nepal night blindness trial, 1998). In Bangladesh it barely moved blood retinol (1.18 vs 1.13 µmol/L on placebo) (Bangladesh pregnancy trial, 2011). Its advantages are about safety: the body converts less once stores are full (Philippines conversion study, 2000), and unlike high-dose preformed vitamin A it is not known to cause birth defects (NIH ODS).

An enzyme in the gut, beta-carotene monooxygenase 1 (BCMO1), splits beta-carotene into vitamin A. Official conversion factors assume 2 mcg of beta-carotene from a supplement, or 12 mcg from food, gives 1 mcg of vitamin A as retinol activity equivalents (RAE) (NIH ODS). The gap reflects absorption: the body absorbs 10-30% of beta-carotene from most foods, and 8.7-65% from supplements (NIH ODS).

Trials that measured it directly found similar or worse ratios. In Bangladeshi men, synthetic beta-carotene converted at about 6:1 by weight, Indian spinach at about 10:1 and sweet potato at about 13:1 (Bangladesh stores trial, 2004). In Indonesian schoolchildren, orange fruit worked out at 12:1 and leafy vegetables with carrots at 26:1 (Indonesia fruit and vegetables trial, 1998).

Conversion has a brake. In Filipino children fed carotenoid-rich meals, the poorer their starting vitamin A status, the more their stores improved (Philippines conversion study, 2000). That limits the risk of overshooting, but it also means beta-carotene does least once you are close to normal.

Several things reduce conversion. Women carrying one common BCMO1 gene variant converted about 32% less beta-carotene, and those carrying two variants about 69% less (BCMO1 variants study, 2009). In Bangladeshi children with roundworm infection, 1.2 mg of beta-carotene a day raised serum retinol only when the children were also dewormed (Bangladesh deworming trial, 2010). In the one trial in cystic fibrosis, supplements raised blood beta-carotene but not vitamin A (Cochrane cystic fibrosis review, 2018). Average conversion factors can't tell you how well you personally convert.

First confirm the deficiency: night blindness is the first sign, and a blood retinol of 20 mcg/dL (0.70 µmol/L) or less often reflects moderate deficiency (NIH ODS). The trials used modest daily or weekly amounts. A wafer with 3.5 mg of beta-carotene a day for 12 weeks raised serum retinol by 0.32 µmol/L in Indonesian breastfeeding women (Indonesia wafer trial, 1995). Bangladeshi men took 750 mcg retinol equivalents a day as beta-carotene (4.5 mg at the study's 6:1 factor) for 60 days (Bangladesh stores trial, 2004). Pregnant Indonesian women took 4.5 mg a day (Indonesia zinc trial, 2004), and the Nepal and Bangladesh pregnancy trials gave 42 mg once a week (Nepal mortality trial, 1999; Bangladesh pregnancy trial, 2011).

To meet the adult RDA from a supplement, the 2:1 factor means about 1.8 mg a day for men (900 mcg RAE), 1.4 mg for women (700 mcg), 1.5 mg in pregnancy (770 mcg) and 2.6 mg when breastfeeding (1,300 mcg) (NIH ODS). Labels often use IU: 1 IU of supplemental beta-carotene is 0.3 mcg RAE. Only the preformed part of a product's vitamin A counts toward the upper limit, so check what share of the vitamin A in multivitamins comes from beta-carotene (NIH ODS).

The strongest direct test is a blinded equivalence trial in rural Senegal. 510 children aged 2-15 with abnormal eye cytology got either beta-carotene or retinyl palmitate in the formulation the World Health Organization recommended; at 7 weeks 50.0% and 51.2% had reverted to normal, and the two were statistically equivalent (one-tailed P=0.03) (Senegal equivalence trial, 1993). The abstract does not give the beta-carotene dose.

Three further trials found it raised vitamin A status. In Bangladeshi men, stores rose by 0.062 mmol with synthetic beta-carotene and 0.065 mmol with retinyl palmitate, compared with the control group, over 60 days (Bangladesh stores trial, 2004). In 175 Indonesian breastfeeding women, a beta-carotene wafer with 4.4 g of fat per daily dose raised serum retinol (+0.32 µmol/L, 95% CI 0.23-0.40) and breast-milk retinol (Indonesia wafer trial, 1995). In 244 Bangladeshi preschoolers with roundworm, 1.2 mg a day for 6 months improved serum retinol only alongside deworming (Bangladesh deworming trial, 2010).

In Nepal, 9,932 first pregnancies were followed in wards randomised to weekly vitamin A, beta-carotene (both 7,000 mcg retinol equivalents) or placebo. Vitamin A cut night blindness in pregnancy from 10.7% to 6.7% (RR 0.62, 95% CI 0.45-0.85); beta-carotene did less (RR 0.83, 0.63-1.11). In the second 3 months after birth, both reduced night blindness by about 50% (Nepal night blindness trial, 1998). In a sub-study of 298 pregnant women, dark adaptation was better with vitamin A than with beta-carotene, which did not differ from placebo (Nepal dark-adaptation study, 2000).

The mortality results conflict. In the Nepal trial of 44,646 women, pregnancy-related deaths fell by 49% with beta-carotene (RR 0.51, 0.30-0.86) and 40% with vitamin A (Nepal mortality trial, 1999). A larger Bangladesh trial with the same weekly doses found no reduction in maternal, fetal or infant deaths (beta-carotene RR 1.21, 0.81-1.81), and beta-carotene barely changed plasma retinol (1.18 vs 1.13 µmol/L) or night blindness (8.9% vs 9.2%) (Bangladesh pregnancy trial, 2011). Zinc may matter: in 170 pregnant Indonesian women, beta-carotene plus zinc improved infants' vitamin A status, with deficiency more than 30% less frequent than in the other groups (Indonesia zinc trial, 2004).

Beta-carotene in food is a weaker tool. An extra daily portion of dark-green leafy vegetables, with the same 3.5 mg of beta-carotene as the wafer, did not improve vitamin A status (Indonesia wafer trial, 1995). Orange fruit did better (Indonesia fruit and vegetables trial, 1998), and 125 g of cooked orange-fleshed sweet potato on 53 school days improved liver vitamin A stores in South African children (South Africa sweet potato trial, 2005). In night-blind pregnant women in Nepal, 6 weeks of amaranth leaves or carrots restored dark adaptation as well as retinyl palmitate (Nepal dark-adaptation trial, 2005).

Most of this evidence comes from low-income countries where poor diets cause deficiency. There are no good trials in adults whose deficiency comes from malabsorption or weight-loss surgery: the only cystic fibrosis trial (24 people) raised blood beta-carotene but not vitamin A (Cochrane cystic fibrosis review, 2018). For those groups, the evidence and guidelines on vitamin A for deficiency apply.

At the doses that correct deficiency, beta-carotene is well tolerated. No upper limit has been set for it, and it is not known to cause birth defects; the most common effect of a lot over time is carotenodermia, a harmless yellow-orange skin that reverses on stopping (NIH ODS). In one trial it appeared only at 30 mg a day, after 25-42 days, and not at 12 mg (carotenodermia trial, 1988).

The serious risk is in smokers. In 29,133 Finnish male smokers, 20 mg a day for 5-8 years raised lung cancer incidence by 18% and total mortality by 8% (ATBC trial, 1994). In 18,314 smokers, former smokers and asbestos workers, 30 mg a day plus 25,000 IU retinyl palmitate raised lung cancer risk (RR 1.28) (CARET trial, 1996). In AREDS2, lung cancers were more common with 15 mg a day (23 vs 11), mostly in former smokers (NIH ODS). The Food and Nutrition Board advises against beta-carotene supplements for the general population, except as a source of vitamin A to prevent deficiency (NIH ODS).

Preformed vitamin A is the opposite case in pregnancy: intakes above the upper limit can cause birth defects, and experts advise women who are or might be pregnant not to take more than 3,000 mcg RAE (10,000 IU) a day of vitamin A supplements (NIH ODS).

Few medicines interact with beta-carotene. Orlistat, a weight-loss drug, lowers absorption of beta-carotene and other fat-soluble vitamins, and its makers advise a multivitamin with vitamin A and beta-carotene (NIH ODS). In the other direction, zinc appears to help beta-carotene do its job (Indonesia zinc trial, 2004).