Milk, bones and the iodine nobody counts

Plain milk supplies about 19 percent of all the calcium people in the United States eat, which makes it harder to replace than almost any other single food. The thing it is famous for is shakier than that. Milk looks protective for bone in cross-sectional and case-control studies, the protection vanishes in cohort studies, and in those cohorts every extra 200 g of milk a day went with a 9 percent higher risk of hip fracture. The health claim the FDA allows on the carton is about calcium and vitamin D building peak bone mass, and it says nothing about milk preventing a fracture. What genuinely swings in milk is iodine. Across 34 countries it ran from 5.5 to 49.9 micrograms per 100 g, winter milk carried more of it than summer milk, and conventional milk carried more than organic, which is the reverse of what most people buying organic expect.

The numbers, per 100 g

How much is actually in it

• The nutrient in milk that actually swings is iodine, not calcium: across 34 countries it ranged ninefold, and winter milk carries measurably more of it than summer milk. — 66 sources reporting milk-iodine concentration in 34 countries (n = 66 sources, 34 countries) Evidence B source
  Range 5.5 to 49.9 μg/100 g, median 17.3; winter minus summer mean difference 5.97 μg/100 g (p = 0.001); conventional minus organic 6.00 μg/100 g (p < 0.0001), significant only in summer
• Two farming variables move milk iodine in opposite directions from what shoppers assume: conventional milk carried more iodine than organic, and the gap appeared only in summer. — the subset of the 66 sources that reported season or farming system (n = not separately stated for each sub-analysis in the abstract) Evidence B source
  Winter higher than summer by 5.97 μg/100 g (p = 0.001); conventional higher than organic by 6.00 μg/100 g (p < 0.0001); organic-conventional difference significant only in summer (p = 0.0003)

What a real portion looks like

• Plain milk alone supplies about a fifth of all the calcium Americans eat, which is why replacing it is a bigger nutritional decision than replacing most single foods. — US population, What We Eat In America / NHANES 2009-2010 (n = national dietary survey) Evidence E source
  Milk and dairy contributed 37% of total calcium intake, of which about half (19% of the total) came from plain milk, including milk on cereal and in other beverages

What your body absorbs

• Calcium from a milk-derived supplement appeared in the blood more slowly and less completely than calcium from ordinary calcium carbonate, which is the opposite of what dairy-derived calcium is usually sold on. — 10 healthy postmenopausal women aged 50-65 (n = 10) Evidence B source
  Serum-ionized calcium higher after calcium carbonate than after milk-derived calcium permeate at 240 min (difference 0.039 mmol/L; 95% CI 0.017-0.061; P = 0.00078); 360-min urinary calcium higher after carbonate (difference 2.47 mmol/L; 95% CI 1.90-3.03; P = 0.0042)

What cooking and storage change

• Almost every question about whether pasteurised, UHT or raw milk digests differently rests on four studies and 57 people in total, and no two of them used the same heat treatment. — 57 healthy adults aged 20 to 68 in four trials from Finland, France and Sweden; in two trials subjects reported GI symptoms after milk but had no diagnosed lactose intolerance (n = 57 participants across 4 studies and 6 treatment groups) Evidence C source
  Three studies tested pasteurization, two tested UHT, one oven-heated the milk; none shared both a heat treatment and an outcome measure with another
• Heat treatment did not change how much fat from milk reached the blood, but it did change which fatty acids arrived. — healthy adults, 20-68 years, from the four included studies (n = 57 participants) Evidence C source
  Plasma triacylglycerol responses unaffected; higher plasma myristic and palmitic acid abundance at 2 h with pasteurized milk (n = 11) and at 4 h with UHT (n = 14); other differences inconsistent
• Fermented milk with no fat in it still raised the absorption of fat-soluble vitamins, which means the carrier effect of milk is not simply its fat. — 20 healthy male participants (n = 20) Evidence B source
  Co-ingestion with 75 g non-fat fermented milk versus 75 g water raised incremental AUC for retinyl palmitate, cholecalciferol and menaquinone-7 in total plasma, and for α- and γ-tocopherol in the TRL fraction; doses were 830 μg vitamin A, 230 μg vitamin D, 100.8 mg vitamin E, 16.8 mg vitamin K

Safety, limits and interactions

• Milk measurably blocks five named antibiotics, which is the one milk interaction with enough evidence behind it to name the drugs. — 109 studies from 101 reports, in adult volunteers; 60.5% open-label cross-over as recommended by FDA (n = 109 studies) Evidence B source
  Absorption of ciprofloxacin, levofloxacin, norfloxacin, pazufloxacin and prulifloxacin significantly diminished when taken with milk; the largest negative effect of milk was on norfloxacin

What people believe that the data does not support

• The claim that milk protects against fractures survives only in the weakest study designs: in cohort studies it disappears, and every extra 200 g of milk a day went with a 9 percent higher hip-fracture risk. — cross-sectional, case-control and cohort studies of milk and dairy intake with osteoporosis or hip fracture outcomes (n = number of included studies not stated in the abstract) Evidence B source
  Milk and osteoporosis overall RR 0.79 (95% CI 0.57-1.08), not significant; hip fracture RR 0.75 (95% CI 0.57-0.99) in cross-sectional and case-control only; in cohort meta-regression every extra 200 g milk/day associated with 9% greater hip fracture risk
• The link between milk and acne is the one popular belief about milk that the data supports, and skim milk is associated with it more strongly than whole milk. — observational studies of dairy, milk, yogurt and cheese intake and acne, across children, adolescents and adults (n = number of included studies not stated in the abstract) Evidence C source
  Highest versus lowest intake: dairy OR 2.61 (95% CI 1.20-5.67), total milk OR 1.48 (1.31-1.66), low-fat milk OR 1.25 (1.10-1.43), skim milk OR 1.82 (1.34-2.47); no association for yogurt or cheese
• The health claim the FDA actually allows on a milk carton is about calcium and vitamin D building peak bone mass, not about milk preventing fractures. — US labelling of foods bearing the calcium, vitamin D and osteoporosis claim Evidence E source
  The authorised relationship runs through low peak bone mass as one of many risk factors, not through fracture reduction by any particular food

Who this works differently for

• For an infant, cow's milk is a poor iron vehicle twice over: it carries little iron, and it blocks the absorption of iron eaten with it. — 37 studies, 1531 participants: preterm infants, full-term infants and children up to 24 months (n = 1531 participants across 37 studies) Evidence C source
  Iron absorption 49% from breast milk versus 10% from home-made cow's milk formula at 6 months; cow's milk iron content 0.5 mg/L; cow's milk inhibits inorganic iron absorption but not haemoglobin iron
• Lactose intolerance is not a fixed verdict on milk: probiotics reduced the symptoms it causes in adults across twelve trials. — adults with lactose intolerance (n = 12 studies) Evidence B source
  Symptoms including abdominal pain, diarrhoea and flatulence reduced; largest effect on breath-hydrogen AUC (SMD -4.96; 95% CI -6.92 to -3.00); monostrain probiotics reduced abdominal pain and total symptoms

What is still unknown

• US federal advice on milk fat reversed in 2025: whole milk and full-fat dairy are now named among the recommended sources of fat, while the WHO guideline capping saturated fat at 10 percent of energy still stands. — US general population aged 2 and over Evidence E source
  Whole milk and full-fat dairy named among whole-food fat sources the guidance directs people toward, reversing the previous low-fat dairy direction

The bottom line

If you take ciprofloxacin, levofloxacin, norfloxacin, pazufloxacin or prulifloxacin, keep milk away from the dose and let the pharmacist tell you how long to leave between them. Those five are the drugs with enough pharmacokinetic evidence behind them to name, and milk measurably cut how much of each one got in. Two groups have a separate reason to think about milk. Babies absorb little iron from cow's milk and absorb less of the iron eaten alongside it, so milk before the first birthday is a question for a doctor rather than a shopping decision. People with lactose intolerance have more room than they are usually given: across twelve trials, probiotics eased the abdominal pain, diarrhoea and flatulence. Acne is the one popular belief about milk that the data supports, with odds of 1.48 for milk overall and 1.82 for skim, so the low-fat carton is not the safer one here. On milk fat itself, two authorities now say opposite things. The Dietary Guidelines for Americans 2025-2030 name whole milk and full-fat dairy among recommended sources of fat, while the WHO guideline capping saturated fat at 10 percent of energy still stands. Nobody has reconciled those two, and we would rather show you the disagreement than pick the side we like.

Sources

1. milk-b-01 · systematic review and meta-analysis of composition data
   Milk‑iodine concentration was available for 34 countries (from 66 sources) and ranged from 5.5 to 49.9 μg/100 g (median 17.3 μg/100 g).
   Walsh et al., Food Chemistry, 2024 · checked 2026-09-21
2. milk-b-02 · meta-analysis of composition data
   Meta-analyses identified that iodine concentration is significantly higher in: (i) winter than summer milk (mean difference 5.97 μg/100 g; p = 0.001), and (ii) in conventional than in organic milk (mean difference 6.00 μg/100 g; p < 0.0001).
   Walsh et al., Food Chemistry, 2024 · checked 2026-09-21
3. milk-b-13 · dietary survey analysis
   Milk and dairy products contributed over one-third (37%) to total calcium intake, of which about half (19%) was from plain milk.
   FSRG Dietary Data Briefs, USDA Agricultural Research Service, WWEIA/NHANES 2009-2010 · checked 2026-09-21
4. milk-b-06 · randomised controlled cross-over trial
   Postprandial calcium absorption from CP was lower than that of CC, and concurrently, urinary concentration reflected increased serum appearance by CC compared with CP, highlighting different metabolic responses.
   Acute RENEW investigators, The Journal of Nutrition, 2023 · checked 2026-09-21
5. milk-b-03 · rapid systematic review of randomised trials
   Three studies measured the effects of pasteurization (15–17), two measured the effects of UHT treatment (16, 17) and in one study the milk was heated in an oven (18).
   Vasilopoulou et al., Frontiers in Nutrition, 2021 · checked 2026-09-21
6. milk-b-04 · rapid systematic review
   Overall plasma triacylglycerol responses were unaffected by milk heat treatments reported, but plasma fatty acid composition differed.
   Vasilopoulou et al., Frontiers in Nutrition, 2021 · checked 2026-09-21
7. milk-b-07 · randomised crossover trial
   The incremental area under the curve for retinyl palmitate, cholecalciferol, and menaquinone-7 concentrations in the total plasma fraction, and α-tocopherol and γ-tocopherol concentrations in the TRL fraction of plasma were significantly higher for the group of the co-ingestion of fat-soluble vitamins and non-fat fermented milk, compared with the co-ingestion of fat-soluble vitamins and water.
   Journal of Nutritional Science and Vitaminology, 2026 · checked 2026-09-21
8. milk-b-09 · systematic review with meta-analyses of pharmacokinetic studies
   The data concerning the impact of beverages were limited to 10 quinolones, and the absorption of 5 (namely ciprofloxacin, levofloxacin, norfloxacin, pazufloxacin, and prulifloxacin) was significantly diminished when consumed with milk.
   Clinical Pharmacokinetics, 2024 · checked 2026-09-21
9. milk-b-10 · meta-analysis of observational studies
   Despite an inverse association between milk and dairy intake and risk of osteoporosis and hip fracture in cross-sectional and case-control studies, no such association was seen in cohort studies.
   Malmir et al., Critical Reviews in Food Science and Nutrition, 2020 · checked 2026-09-21
10. milk-b-11 · meta-analysis of observational studies
    Highest compared with lowest category of dairy (OR: 2.61, 95% CI: 1.20 to 5.67), total milk (OR: 1.48, 95% CI: 1.31 to 1.66), low-fat milk (OR: 1.25, 95% CI: 1.10 to 1.43) and skim milk (OR: 1.82, 95% CI: 1.34 to 2.47) intake significantly was associated with the presence of acne.
    Aghasi et al., Clinical Nutrition, 2019 · checked 2026-09-21
11. milk-b-12 · regulation, 21 CFR 101.72(a)
    An inadequate intake of calcium or calcium and vitamin D contributes to low peak bone mass, which has been identified as one of many risk factors in the development of osteoporosis.
    21 CFR 101.72, дозволена заява про здоров'я: кальцій, вітамін D і остеопороз · checked 2026-09-21
12. milk-b-05 · systematic review of isotopic absorption studies
    At 6 months of age, full-term infants fed breast milk attained greater iron stores than did those fed a home-made cow's milk formula due to a higher iron absorption rate from breast milk (49%) than from home-made cow's milk formula (10%) [45].
    Nutrients, 2024 · checked 2026-09-21
13. milk-b-08 · meta-analysis of intervention studies
    The results showed that probiotic administration alleviated the symptoms of lactose intolerance, including abdominal pain, diarrhea, and flatulence.
    Oak & Jha, Journal of Dairy Science, 2023 · checked 2026-09-21
14. milk-b-14 · agency guidance
    The guidance calls for receiving the bulk of fat from whole food sources, such as meats, poultry, eggs, omega 3–rich seafood, nuts, seeds, full-fat dairy, olives, and avocados.
    Dietary Guidelines for Americans 2025-2030, фактшит USDA/HHS · checked 2026-09-21