Eggs: the evidence is weaker than either camp says

The strongest trial evidence about eggs is not about anybody’s cholesterol. Pooling seven randomised trials in 3673 children under five in sub-Saharan Africa, feeding eggs raised weight-for-age by 0.33 z-scores (95% CI 0.11 to 0.55) and weight-for-height by 0.30 (0.12 to 0.48), while height-for-age did not move at 0.05 (95% CI -0.05 to 0.14). Seven trials in one region, in children under five, is the whole of the grade A evidence here. For adults the picture is softer. An umbrella review of 14 meta-analyses published since January 2020 graded the overall quality of that literature as critically low. Every significant association in it was graded very weak: heart failure at RR 1.15 (95% CI 1.02 to 1.30), cancer mortality at 1.13 (1.06 to 1.20) and LDL cholesterol 7.39 mg/dl higher (5.82 to 8.95), while the same review reported no significant association with cardiovascular outcomes or with all-cause mortality. A controlled feeding trial shows where the argument actually sits. In 61 adults on three 5-week isocaloric diets, LDL rose with saturated fat intake (beta 0.35, P = 0.002) while dietary cholesterol showed no such link (beta -0.006, P = 0.42).

How much is actually in it

• Nearly the whole fat of an egg sits in the yolk: 28.8 g per 100 g against 0.0 g in the white, and the selenium follows it at 55.8 micrograms. — USDA Foundation Foods records for Eggs, Grade A, Large (n = 4 records, 24 samples) Evidence E source
  Fat 28.8 g per 100 g in yolk, 9.96 g in whole egg, 0.0 g in white; selenium 55.8 micrograms per 100 g in yolk, with 17.9 to 55.8 across the four records and 7 to 81.3 across 24 Foundation samples

What a real portion looks like

• One large egg without the shell weighs 50.3 g, of which 34.0 g is white and 17.0 g is yolk, which is how a per 100 g table turns into a breakfast. — USDA Foundation Foods portion records for Eggs, Grade A, Large (n = 3 records) Evidence E source
  1 egg whole without shell 50.3 g, 1 egg white 34.0 g, 1 egg yolk 17.0 g

What your body absorbs

• Genotype decides how much of an egg lands in the blood: in 91 adults given 640 mg of extra cholesterol a day for 30 days, carriers of the ABCG5 C/C variant raised LDL more than the other genotypes. — 40 men and 51 premenopausal women randomly assigned to an egg diet supplying 640 mg extra dietary cholesterol and 600 micrograms lutein plus zeaxanthin a day, or a substitute, for 30 days with a 3-week washout (n = 91) Evidence B source
  Greater rise in LDL cholesterol in C/C genotype carriers (P < 0.05) and a trend for lutein (P = 0.08) compared with C/G and G/G genotypes

What cooking and storage change

• Of 123 children with diagnosed egg allergy taken through a supervised boiled egg yolk challenge, 115 passed and went on eating yolk at home. — 123 children with egg allergy, median age 12 months, challenged with boiled egg yolk between 2019 and 2022 (n = 123) Evidence C source
  115 of 123 (93.4%) tolerant to boiled egg yolk; egg white tolerance followed a mean 4 months later in that group against 11 months in the non-tolerant group
• Baking changes what an allergic child can eat, and an egg white specific IgE at or below 7.39 kU/L marked the milder phenotype in 143 children. — 143 children with IgE-mediated immediate-type egg or milk allergy in infancy, 76 of whom became tolerant to the baked form before age two (n = 143) Evidence C source
  Egg white specific IgE cut-off 7.39 kU/L predicting the mild phenotype, area under the curve 0.703

What it does to your health

• An umbrella review of 14 meta-analyses found every significant egg association rated very weak, including heart failure risk 1.15 and LDL cholesterol 7.39 mg/dl higher. — fourteen meta-analyses published since January 2020, ten of observational studies and four of intervention studies (n = 14 meta-analyses) Evidence C source
  Heart failure RR 1.15 (95% CI 1.02 to 1.30), cancer mortality RR 1.13 (1.06 to 1.20), LDL cholesterol WMD 7.39 (5.82 to 8.95), HDL cholesterol WMD 1.37 (0.49 to 2.25); no association with cardiovascular outcomes or all-cause mortality
• In 521,120 older Americans followed for a median 16 years, each extra half egg a day came with a 7% higher rate of death from any cause, cardiovascular disease and cancer alike. — 521,120 US adults aged 50 to 71 at recruitment in 1995 to 1996, followed to the end of 2011, 129,328 deaths including 38,747 from cardiovascular disease (n = 521,120) Evidence C source
  Hazard ratio 1.07 (95% CI 1.06 to 1.08) for all-cause mortality per additional half whole egg a day, 1.07 (1.06 to 1.09) for cardiovascular and 1.07 (1.06 to 1.09) for cancer mortality; cholesterol intake mediated 63.2% of the all-cause association
• Across 11 observational studies and 1,274,437 people, egg intake ran with lower glioma risk in Asian populations (RR 0.51) and with rising head and neck cancer risk above 50 g a day. — 1,274,437 participants in eleven observational studies, 4199 cases of glioma and 3164 of head and neck cancer (n = 1,274,437) Evidence C source
  Glioma RR 0.51 (95% CI 0.35 to 0.73) in Eastern populations; head and neck cancer risk rising gradually above 50 grams a day in non-linear dose-response
• Adding two eggs a day to a plant-based diet for 4 weeks raised HDL cholesterol and lifted plasma choline from 9.47 to 10.54 micromoles per litre in 24 adults with metabolic syndrome. — 24 adults with metabolic syndrome, mean age 49.3 years, on a plant-based diet for 13 weeks, randomised crossover of two eggs versus egg substitute with 70 g spinach at breakfast (n = 24) Evidence B source
  Higher HDL cholesterol (p < 0.025) and lower body weight (p < 0.02) on eggs; plasma choline 10.54 versus 9.47 micromoles per litre (p < 0.025); plasma zeaxanthin rose only on eggs
• In 20 prediabetic men, swapping a quarter of a 100 g glucose load for 1.5 whole eggs or seven egg whites blunted the fall in flow-mediated dilation at 30 to 60 minutes, and two yolks did nothing of the sort. — 20 prediabetic men in a randomised crossover of four isoenergetic 400 kcal meals: 100 g glucose alone, or 75 g glucose with 1.5 whole eggs, seven egg whites or two egg yolks (n = 20) Evidence B source
  Flow-mediated dilation area under the curve higher with whole eggs and egg whites only; glucose and insulin areas under the curve lower with those two arms; F2-isoprostanes lower with whole eggs and whites and unchanged by yolk

Safety, limits and interactions

• Pooled across Chinese surveys, 7.20% of eggs carried Salmonella, and the shell was nearly twice as often contaminated as the contents, 8.30% against 4.50%. — eggs sampled at farms and retail markets across Chinese provinces in studies pooled by meta-analysis Evidence C source
  Overall pooled prevalence 7.20%, eggshells 8.30% versus egg contents 4.50%, farm samples 9.60% versus retail 5.80%, with S. typhimurium and S. enteritidis predominant
• The classic English case-control study of 232 Salmonella enteritidis patients tied illness to raw shell egg products such as homemade mayonnaise, and to shop bought sandwiches containing mayonnaise or eggs. — 232 patients with confirmed primary sporadic Salmonella enteritidis phage type 4 infection in England in August and September 1988, 160 with questionnaire data, matched to up to three neighbourhood controls each (n = 232) Evidence C source
  Raw shell egg products matched p = 0.02, shop bought sandwiches with mayonnaise p = 0.00004 or with eggs p = 0.02, lightly cooked eggs unmatched p = 0.02
• A worldwide review of 30 studies found mercury above the limits set by USDA, China and Thailand in 12 of 17 countries, with the highest concentrations in chicken eggs from Colombia. — 30 eligible studies of mercury and methylmercury in poultry meat and eggs across sixteen countries (n = 30 studies) Evidence C source
  Limits of 0.05 mg per kg (USDA and China NHC) and 0.02 mg per kg (Thailand) exceeded in 12 of 17 countries, highest in Colombian chicken eggs
• In two clinical cohorts of 44 and 1993 patients on long-term atorvastatin, liver enzyme elevation was more common in those with greater dietary cholesterol intake. — two observational cohorts of patients on long-term atorvastatin, 44 assessed for individual cholesterol intake and 1993 assessed around holiday season meals, alongside hamster experiments (n = 2,037) Evidence C source
  Higher incidence of hepatic toxicity, measured as alanine aminotransferase elevation, with greater dietary cholesterol intake and with heavier holiday eating

What people believe that the data does not support

• In a controlled feeding trial in 61 adults, LDL cholesterol tracked saturated fat intake with a coefficient of 0.35, while dietary cholesterol showed no such link at -0.006. — 61 adults, mean age 39 years, BMI 25.8, baseline LDL below 3.5 mmol/L, on three isocaloric diets of 5 weeks each in a randomised crossover (n = 61) Evidence B source
  Saturated fat versus LDL beta 0.35 (P = 0.002); dietary cholesterol versus LDL beta -0.006 (P = 0.42); two eggs a day on a low saturated fat diet lowered LDL from 109.3 to 103.6 (P = 0.02)
• The United States health claim rule attaches risk to blood cholesterol rather than to any food, drawing its lines at 240 mg/dL total and 160 mg/dL LDL. — United States food labelling, health claims on dietary saturated fat and cholesterol Evidence E source
  High risk thresholds of 240 mg/dL total cholesterol and 160 mg/dL LDL cholesterol, with borderline ranges of 200 to 239 and 130 to 159

Who this works differently for

• Across 21 studies and 99,453 people, egg intake closest to 50 g a day, about one egg, was longitudinally linked to a lower risk of cognitive decline, pooled risk ratio 0.91, on a U-shaped curve whose minimum sits at 50 to 60 g a day. — 99,453 individuals with 18,893 cases in 13 longitudinal and 8 cross-sectional studies, mainly from Asian countries and the United States (n = 99,453) Evidence C source
  Pooled RR 0.91 (95% CI 0.83 to 0.99) at intake closest to 50 g a day, with a U-shaped dose-response whose minimum falls at 50 to 60 g a day
• Pooling seven randomised trials in 3673 children under five in sub-Saharan Africa, egg feeding raised weight-for-age by 0.33 z-scores and weight-for-height by 0.30, and left height-for-age alone. — 3673 children under five years old in seven randomised controlled trials published between 2013 and 2023 in sub-Saharan Africa (n = 3673) Evidence A source
  Weight-for-age mean difference 0.33 (95% CI 0.11 to 0.55), weight-for-height 0.30 (0.12 to 0.48), height-for-age 0.05 (-0.05 to 0.14)
• FDA puts two numbers on egg safety, a refrigerator at 40 degrees Fahrenheit or below and cooking until yolk and white are both firm, and names children, older adults, pregnant women and immunocompromised people as the ones for whom it matters most. — United States consumers, with a named higher risk group of children, older adults, pregnant women and people with weakened immune systems Evidence E source
  Storage at 40 degrees Fahrenheit or below, cooking until yolk and white are firm
• Bird-egg syndrome runs the other way from ordinary egg allergy: sensitisation starts with chicken serum albumin from feathers and droppings, and the reaction lands on the yolk rather than the white. — children and adults sensitised to bird antigens who later react to egg, described in a review with one illustrative paediatric case (n = 1 case) Evidence D source
  Sensitisation to chicken serum albumin, alpha-livetin, with respiratory and gastrointestinal symptoms after egg intake or bird contact

What is still unknown

• Mapping 37 human trials of fortified eggs, the reviewers found the biomarkers moved and the health outcomes did not follow. — 37 human intervention studies of fortified, enriched or bio-enhanced eggs published between 2005 and 2025 (n = 37 studies) Evidence C source
  Consistent improvement in nutrient status biomarkers, with downstream health outcomes limited, inconsistent and not comparable across nutrients
• The strongest statement the evidence on eggs supports is about itself: across 14 meta-analyses the overall quality was rated critically low. — fourteen meta-analyses of egg consumption and health outcomes appraised with AMSTAR-2 and NutriGrade (n = 14 meta-analyses) Evidence C source
  Substantial variability and high heterogeneity, overall study quality rated critically low

The bottom line

Where a number comes from decides what it means. In 521,120 US adults aged 50 to 71 followed to 2011, each additional half egg a day ran with a 7% higher rate of death from any cause (HR 1.07, 95% CI 1.06 to 1.08), and the same for cardiovascular and cancer death. Those intakes came from a food frequency questionnaire, and the authors name residual confounding as a limitation. In that analysis, people eating egg whites and egg substitutes had lower mortality, which points at the yolk cholesterol rather than at the egg. Blood does not answer the same in everyone. Among 91 adults given an extra 640 mg of dietary cholesterol a day for 30 days, carriers of the ABCG5 C/C variant raised LDL more than the other genotypes (P < 0.05), in what the authors call a small sample. The rest of the human trials are small and short. In 24 adults with metabolic syndrome eating a plant-based diet, two eggs a day for 4 weeks raised HDL cholesterol and lifted plasma choline from 9.47 to 10.54 micromoles per litre. LDL, triglycerides, glucose and blood pressure did not move in that crossover, which lost 6 of its 30 people. In 20 prediabetic men, whole eggs or egg whites taken with a glucose load blunted the three-hour dip in vessel function, and yolk alone did none of that. Observational data on the ageing brain draws a U: across 99,453 people, intake closest to 50 g a day, about one egg, went with a 9% lower risk of cognitive decline in the longitudinal studies (RR 0.91, 95% CI 0.83 to 0.99), and the curve turns upward past that. An egg is two foods in one shell. USDA records put 28.8 g of fat per 100 g in the yolk, 0.0 g in the white and 9.96 g in whole egg. Selenium reads 55.8 micrograms per 100 g in the yolk, with a spread from 7 to 81.3 across 24 sampled Foundation records. A large egg out of the shell weighs 50.3 g, of which 34.0 g is white and 17.0 g is yolk, so any per 100 g figure is about two eggs. Cooking here is a safety instruction before it is a preference. The FDA tells consumers to store eggs at 40 degrees Fahrenheit or below and to cook until yolk and white are both firm, without stating how much illness either step prevents. It names children, older adults, pregnant women and people with weakened immune systems as those at greater risk. In England in 1988, 232 people with confirmed Salmonella enteritidis phage type 4 infection were more likely than matched neighbours to have eaten raw shell egg products such as homemade mayonnaise (matched P = 0.02). Shop bought sandwiches containing mayonnaise carried the strongest link (P = 0.00004). That is pre-vaccination data, so it describes the pattern rather than today’s absolute risk. In eggs sampled in China, pooled Salmonella prevalence was 7.20%, higher on shells at 8.30% than in contents at 4.50%, and that paper says the level is above United States and European figures, so it should not be read onto another country’s supply. Egg allergy has a shape worth knowing. Of 123 children with egg allergy, median age 12 months, taken through a supervised boiled yolk challenge at one centre, 115 tolerated it and went on eating yolk at home. That is a retrospective record of supervised challenges and not a plan for the kitchen table. In 143 children with immediate-type egg or milk allergy, an egg white specific IgE at or below 7.39 kU/L marked the milder baked-tolerant phenotype, with an area under the curve of 0.703, which classifies weakly. Bird-egg syndrome runs the opposite way from the usual advice about the white: sensitisation to chicken serum albumin from feathers and droppings lands the reaction on the yolk, described in a review with one illustrative case. A review of 30 studies found mercury above the limits set by USDA, China and Thailand in poultry and eggs in 12 of 17 countries, highest in Colombian chicken eggs, with no pooled estimate and uneven country coverage. On medicines, the only evidence we found is two observational cohorts of patients on long-term atorvastatin, in whom alanine aminotransferase elevation was more common with higher dietary cholesterol intake, reported without an effect size. Fortified eggs move the biomarkers of the nutrient they carry across 37 human trials, and the review that mapped them found the health outcomes limited and inconsistent. Read every figure on this page against the umbrella review’s own verdict that the overall quality of egg research is critically low.

Sources

1. eggs-r4-21 · USDA FoodData Central Foundation Foods, dataset
   Eggs, Grade A, Large, egg yolk (FDC 748236): fat 28.8 g, energy 334 kcal per 100 g; Eggs, Grade A, Large, egg white (FDC 747997): fat 0.0 g, energy 55 kcal per 100 g; Eggs, Grade A, Large, egg whole (FDC 748967): fat 9.96 g, energy 148 kcal per 100 g; selenium 55.8 ug per 100 g in yolk, 17.9-55.8 across 4 records, 7-81.3 across 24 Foundation samples
   USDA FoodData Central, Foundation Foods, release 2026-04-30 · checked 2026-09-24
2. eggs-r4-22 · USDA FoodData Central Foundation Foods, dataset
   Eggs, Grade A, Large, egg whole (FDC 748967): portion 1.0 egg whole without shell = 50.3 g; Eggs, Grade A, Large, egg white (FDC 747997): portion 1.0 egg white = 34.0 g; Eggs, Grade A, Large, egg yolk (FDC 748236): portion 1.0 egg yolk = 17.0 g
   USDA FoodData Central, Foundation Foods, release 2026-04-30 · checked 2026-09-24
3. eggs-r4-09 · randomised crossover trial in humans
   Individuals possessing the C/C genotype experienced a greater increase in both LDL-C (P < 0.05) and a trend for lutein (P = 0.08) during the EGG period compared with those individuals with the C/G (heterozygote) or G/G genotypes (homozygotes).
   J Nutr, 2006 · checked 2026-09-24
4. eggs-r4-10 · retrospective cohort in children
   Of the 123 egg yolk OFCs, 115 (93.4%) patients were tolerant and consumed egg yolk at home.
   Eur Ann Allergy Clin Immunol, 2026 · checked 2026-09-24
5. eggs-r4-11 · retrospective cohort in children
   Receiver operating characteristic analysis determined cut-off values for specific-Immunoglobulin E (sIgE) levels (kU/L) predicting mild phenotype at first admission: egg white-sIgE ≤ 7.39, milk-sIgE ≤ 5.99, and casein-sIgE ≤ 4.99, with AUC values of 0.703, 0.716, and 0.749, respectively.
   Eur Ann Allergy Clin Immunol, 2026 · checked 2026-09-24
6. eggs-r4-01 · umbrella review of observational syntheses
   The level of evidence was very weak for all the significant associations: risk of heart failure (RR 1.15; 95%CI: 1.02-1.30), cancer mortality (RR 1.13; 95%CI 1.06-1.20), higher levels of LDL cholesterol (WMD 7.39; 95%CI 5.82-8.95)...Due to the critically low strength of studies, insufficient evidence is available to discourage egg consumption, suggesting eggs can be part of a healthy diet.
   Nutr Metab Cardiovasc Dis, 2025 · checked 2026-09-24
7. eggs-r4-03 · prospective cohort in humans
   In multivariable-adjusted models, the hazard ratios (95% confidence intervals) associated with each intake of an additional half of a whole egg per day were 1.07 (1.06-1.08) for all-cause mortality, 1.07 (1.06-1.09) for CVD mortality, and 1.07 (1.06-1.09) for cancer mortality.
   PLoS Med, 2021 · checked 2026-09-24
8. eggs-r4-05 · meta-analysis of observational studies
   We found that egg consumption was associated with a reduced risk of glioma, particularly in Eastern populations (i.e. Asian populations) (RR: 0.51, 95% CI: 0.35-0.73). Additionally, in the non-linear dose-response analysis, egg consumption above 50 grams per day (>1 egg/day) gradually increased the risk of HNC.
   Nutr Neurosci, 2026 · checked 2026-09-24
9. eggs-r4-06 · randomised crossover trial in humans
   When we compared individuals consuming the EGG versus the SUB breakfast, we observed a lower body weight (p < 0.02) and a higher HDL cholesterol (p < 0.025) after the EGG diet.
   Nutrients, 2022 · checked 2026-09-24
10. eggs-r4-08 · randomised crossover trial in humans
    Thus, replacing a portion of a glucose challenge with whole eggs or egg whites, but not yolks, limits postprandial impairments in VEF by attenuating increases in glycaemia and lipid peroxidation.
    Br J Nutr, 2018 · checked 2026-09-24
11. eggs-r4-12 · meta-analysis of cross-sectional prevalence surveys
    The overall pooled prevalence of Salmonella in Chinese eggs was 7.20%, which is higher than the levels reported in the United States and Europe. Contamination rates were higher on eggshells (8.30%) was than those in egg contents (4.50%), and prevalence was greater in farm samples (9.60%) than that in retail market samples (5.80%)
    Int J Food Microbiol, 2026 · checked 2026-09-24
12. eggs-r4-13 · case-control study in humans
    Illness due to S enteritidis phage type 4 was significantly associated with consumption of raw shell egg products (homemade mayonnaise, ice cream, and milk drinks containing eggs) (matched p = 0.02) and shop bought sandwiches containing mayonnaise (matched p = 0.00004) or eggs (matched p = 0.02).
    BMJ, 1989 · checked 2026-09-24
13. eggs-r4-15 · systematic review of contamination surveys
    Levels above those permitted by regulatory organizations such as the United States Department of Agriculture (USDA), China's National Health Commission (NHC), and Thailand's Ministry of Public Health - set at 0.05, 0.05, and 0.02 mg.kg-1 respectively - were observed in 12 of the 17 countries, with the highest concentrations detected in chicken eggs from Colombia.
    Biol Trace Elem Res, 2026 · checked 2026-09-24
14. eggs-r4-16 · observational cohort study with animal experiments
    In the human studies, we observed that the incidence of hepatic toxicity in patients receiving long-term ATV treatment was higher in patients with greater dietary cholesterol intake and in patients who consumed more food during Chinese holidays.
    Cardiovasc Drugs Ther, 2021 · checked 2026-09-24
15. eggs-r4-02 · randomised crossover trial in humans
    Across all diets, saturated fat intake was positively correlated with LDL cholesterol (β = 0.35, P = 0.002), whereas dietary cholesterol was not (β = -0.006, P = 0.42).
    Am J Clin Nutr, 2025 · checked 2026-09-24
16. eggs-r4-18 · regulation, health claim
    High blood total- and low density lipoprotein (LDL)- cholesterol levels are major modifiable risk factors in the development of coronary heart disease. High coronary heart disease rates occur among people with high blood cholesterol levels of 240 milligrams/decaliter (mg/dL) (6.21 millimoles per liter (mmol/L)) or above and LDL-cholesterol levels of 160 mg/dL (4.13 mmol/L) or above.
    21 CFR 101.75, Health claims: dietary saturated fat and cholesterol and risk of coronary heart disease · checked 2026-09-24
17. eggs-r4-04 · meta-analysis of observational studies
    Egg consumption closest to 50 g/day was longitudinally linked to a lower risk of cognitive decline (pooled RR = 0.91; 95%CI = 0.83-0.99), and a 'U-shaped' dose-response relationship was also observed at the lowest risks occurring at 50-60 g/day of egg.
    J Hum Nutr Diet, 2026 · checked 2026-09-24
18. eggs-r4-07 · meta-analysis of RCTs in children
    Seven studies involving 3673 children met the inclusion criteria. Egg-based intervention significantly improved WAZ (MD: 0.33; 95% CI: 0.11-0.55) and WHZ (MD: 0.30; 95% CI: 0.12-0.48). However, no significant effect was observed on HAZ (MD: 0.05; 95% CI: -0.05-0.14).
    J Nutr Sci, 2026 · checked 2026-09-24
19. eggs-r4-14 · regulator guidance page
    Store promptly in a clean refrigerator at a temperature of 40° F or below...Cook eggs until both the yolk and the white are firm...Certain people are at greater risk for severe illness and include children, older adults, pregnant women, and people with weakened immune systems (such as transplant patients and individuals with HIV/AIDS, cancer, and diabetes).
    U.S. Food and Drug Administration, What You Need to Know About Egg Safety · checked 2026-09-24
20. eggs-r4-17 · case report
    This association is known as bird-egg syndrome, which is caused by sensitization to chicken serum albumin (alpha -livetin) and is characterized by the development of respiratory and gastrointestinal symptoms after egg intake or after contact with bird antigens.
    Allergol Immunopathol (Madr), 2003 · checked 2026-09-24
21. eggs-r4-19 · scoping review in humans
    Thirty-seven human intervention studies met the eligibility criteria...Evidence for downstream health-related outcomes was more limited, inconsistent, and difficult to compare across nutrient categories and study populations.
    Nutrients, 2026 · checked 2026-09-24
22. eggs-r4-20 · umbrella review of observational syntheses
    The wide range of outcomes, with substantial variability and high heterogeneity, indicated a lack of robust evidence. The overall quality of studies was critically low.
    Nutr Metab Cardiovasc Dis, 2025 · checked 2026-09-24