Carrots: the food and the pill point opposite ways

This is the clearest example in the whole database of why a nutrient is not its food. People who ate the most carrots had about 10 percent lower cancer incidence across 50 prospective studies and 52000 cases. Beta-carotene, the compound carrots are famous for, was then given as a supplement in 8 randomised trials covering 167141 people, and lung cancer risk went up by 16 percent, and by 21 percent among smokers and asbestos workers. A separate pooling of 31 trials and 216734 people found no effect on death from any cause and a rise in lung cancer deaths. The carrot reviewers open their own paper by noting that beta-carotene failed in controlled trials, which is why they look elsewhere for an explanation and call the mechanism unresolved.

The numbers, per 100 g

What a real portion looks like

• Eight milligrams of carotenoids a day, about a large glass of orange-carrot juice, raised blood carotenoids by six tenths in six weeks. — 162 adults from four US racial and ethnic groups (25% Black, 25% Asian, 27% non-Hispanic White, 23% Hispanic), randomised to 177 mL apple juice with negligible carotenoids, 177 mL orange-carrot juice with 4 mg carotenoids a day, or 355 mL with 8 mg a day; 158 completed (n = 162) Evidence B source
  Skin carotenoid score at 6 weeks was 414.0 in the high-dose arm and 369.7 in the moderate-dose arm against 305.2 in the control; skin carotenoids rose 42% from baseline in the high-dose arm and plasma carotenoids rose 61%

What it does to your health

• People who ate the most carrots had about a tenth lower cancer incidence than those who ate the least, across fifty prospective studies. — 50 prospective studies recording carrot intake with 52000 cancer cases, and 30 prospective studies with 9331 cases reporting plasma alpha-carotene (n = 52000 cases) Evidence C source
  Relative risk 0.90 (95% CI 0.87-0.94, p < 0.00004) for carrot intake and 0.80 (0.72-0.89, p < 0.00006) for plasma alpha-carotene; heterogeneity moderate, interaction with cancer type not significant, dose-response significant (p < 0.01)
• Across 31 trials and 216 thousand people, beta-carotene supplements did not prevent a single category of death, and lung cancer deaths went up. — 216734 adults in 31 trials, 108622 supplemented and 108112 control (n = 216734) Evidence A source
  All-cause mortality risk ratio 1.02 (95% CI 0.98-1.05, I2 42%), with no preventive effect on cancer, cardiovascular, cerebrovascular or other causes; lung cancer mortality rose (RR 1.14; 1.02-1.27; I2 3%) and HIV-related mortality fell (RR 0.55; 0.33-0.92)

What people believe that the data does not support

• Beta-carotene taken as a supplement raised the risk of lung cancer rather than lowering it, and more so in smokers and asbestos workers. — 167141 participants across 8 randomised controlled trials (n = 167141) Evidence A source
  Lung cancer risk ratio 1.16 (95% CI 1.06-1.26) overall, 1.21 (1.08-1.35) among smokers and asbestos workers and 1.18 (1.07-1.29) among non-medics; meta-regression found no relationship between dose and the size of the harm
• The one thing everyone knows about carrots is the one thing Europe refused to let anyone claim: that they help eyesight. — — Evidence E source
  Status Non-authorised; all four register entries mentioning carrot are non-authorised

What is still unknown

• The doses that caused that harm were 15 to 30 milligrams a day, far above what food delivers, and every trial gave other substances alongside. — 8 trials ranging from 4060 to 39876 participants; five gave beta-carotene daily and three on alternate days; two included only health professionals, two people with cardiac risk factors, two only smokers, one only asbestos workers and one a nutrient-inadequate general population (n = 167141) Evidence A source
  The dose of beta-carotene varied from 15 to 30 mg daily, and all the studies also administered other active substances

The bottom line

Two details keep the comparison fair. The trial doses were 15 to 30 mg of beta-carotene a day alongside other active substances, which is far above what a vegetable delivers, and nobody in those trials was randomised to eat a carrot. For a sense of scale in the other direction, 8 mg of carotenoids a day as juice for six weeks raised carotenoids in skin by 42 percent and in blood by 61 percent. That trial was built to validate a measuring device, so raised carotenoids were the endpoint and no benefit was tested. Europe refused the claim everyone expects, the one about eyesight and dark adaptation. And the advice that cooking carrots frees more carotene is something we could not source: our corpus returns no randomised studies on cooking and carotenoid uptake at all. One practical line sits under all of this. If you smoke, or worked with asbestos, a beta-carotene supplement is the one form to leave alone, and that is a conversation for whoever looks after your health rather than a shelf decision.

Sources

1. car-dr-05 · randomised controlled feeding trial
   In the high-dose group, there was a 42% change in skin carotenoids from baseline (mean = 290.4) to a 6-wk follow-up (increase of 123, 123/290 = 42.4%). There was a 61% change in the plasma carotenoids in the high-dose group.
   J Nutr, 2023 · checked 2026-09-22
2. car-dr-01 · systematic review and meta-analysis of prospective cohort studies
   Of 198 observational studies, in 50 prospective studies with 52000 cases recording carrot intake, the cancer-risk was substantially reduced (RR 0.90, 95% CI 0.87-0.94, p ˂ 0·00004).
   Crit Rev Food Sci Nutr, 2025 · checked 2026-09-22
3. car-dr-04 · systematic review and meta-analysis of randomised controlled trials
   In a random-effects meta-analysis of all 31 trials, beta-carotene supplements were found to have no preventive effect on mortality (risk ratio 1.02, 95% confidence interval 0.98-1.05, I 2 = 42%).
   Front Med (Lausanne), 2022 · checked 2026-09-22
4. car-dr-02 · systematic review and meta-analysis of randomised controlled trials
   The findings indicate that beta-carotene supplementation was associated with an increased risk of lung cancer (RR = 1.16, 95% CI = 1.06-1.26). This effect was even more noticeable among smokers and asbestos workers (RR = 1.21, 95% CI = 1.08-1.35)
   Nutrients, 2022 · checked 2026-09-22
5. car-dr-06 · regulatory register entry
   POL-HC-7011 Daucus carota (Carrot) For eye health - promotes maintenance of vision apparatus functions Improves dark adaptation Strengthens eye capillaries Reduces eye tiredness in case of vision exertion Non-authorised
   EU Register on nutrition and health claims, заява POL-HC-7011, знімок 2026-09-21 · checked 2026-09-22
6. car-dr-03 · systematic review and meta-analysis of randomised controlled trials
   The dose of beta-carotene varied from 15 to 30 mg daily. All the studies assumed the administration of other active substances in addition to beta-carotene
   Nutrients, 2022 · checked 2026-09-22