Semi-skimmed milk is cow’s milk with about half of the fat taken out. By EU law it holds 1.5 to 1.8% fat, against at least 3.5% for whole milk. The US version is 2% milk, called reduced-fat.
The numbers on this page come from the USDA entry for 2% milk without added vitamins A and D. So they fit US 2% milk and sit very close to European semi-skimmed.
Digesting milk sugar as an adult is a recent genetic trait. Lactase persistence, the ability to digest milk sugar after childhood, is the most strongly selected single-gene trait of the past 10,000 years. That holds in European and many African, Middle Eastern and southern Asian populations.
Even so, people who call themselves lactose intolerant often cope with a cup. In a double-blind trial, 30 adults with self-reported severe intolerance had only trivial symptoms on 2% milk.
Most studies below cover milk or all dairy, and few separate semi-skimmed from whole, so what they show about milk fat is limited. Several have dairy industry money behind them, and the cards say so where the paper does.
Health benefits of semi-skimmed milk
The top findings from 31 studies, strongest evidence first
What the stars mean
- Strong. Several randomized trials, pooled in a meta-analysis, point the same way, or regulators (FDA or EFSA) have approved the claim.
- Moderate. Good randomized trials support it, but there are fewer of them, or their results vary.
- Limited. Small, short or single trials, or mixed results. Promising, but it needs confirmation.
Population and lab studies get no stars: they show a link or a mechanism, not proof.
How we rate evidence →- ★★Higher bone density with dairyIn children and teens, extra dairy raised whole-body bone mineral content by 3.0% in pooled trials. In postmenopausal women, dairy, mostly milk, raised hip and spine bone density.
- ★★Mixed results for height in childrenA pooled analysis of 17 randomized trials in ages 6 to 18 found no change in height. An older review of 12 controlled studies found about 0.4 cm more height a year with a cup of milk a day.
- ★★No weight change unless calories were cutIn 14 pooled trials in adults, high-dairy diets (dairy as a group) cut weight 1.29 kg more than usual weight-loss diets when calories were restricted. Waist size dropped 2.43 cm more, but without calorie restriction, weight did not change significantly.
- ★★Less milk allergy after early formula feedingIn 491 Japanese infants, newborn cow’s milk formula (at least 10 mL a day) cut milk allergy at 6 months from 6.8% to 0.8%. That was formula, not milk from a carton.
- ★Few extra symptoms in self-described lactose intoleranceIn 30 adults who reported symptoms from milk, a daily cup of regular milk was not significantly worse than lactose-free milk for a week. The 21 who could not absorb lactose well had 2.5 more episodes of gas a day on regular milk.
All 31 studies, strongest evidence first
Extra dairy raised whole-body bone mineral content 3.0% in children
This meta-analysis (a study pooling many trials) covered randomized trials of extra dairy in children and teenagers aged 3 to 18. Whole-body bone mineral density rose 1.8% and height 0.21 cm. The effects were generally consistent across sex, region and baseline calcium intake. It covers dairy as a group, not semi-skimmed milk alone, and three authors work for Yili, a dairy company.
Dairy, mostly milk, raised hip and spine bone density after menopause
This meta-analysis pooled 6 randomized trials with 618 healthy postmenopausal women, in which milk was the main dairy product. The standardized gain, a measure of effect size, was 0.21 at the lumbar spine and 0.37 at the total hip. The hip effect showed from 12 months of use. The trials measured bone density, not fractures.
Milk products did not change height in ages 6 to 18
This meta-analysis pooled 17 randomized trials of milk and milk products for 12 weeks or more, in 2,844 children who were not undernourished. Body weight rose 0.48 kg more and lean mass 0.21 kg more than in controls. The gain in percent body fat was 0.27 points smaller than in controls. The Dairy Farmers of Canada and the Dairy Research Cluster Initiative supported the study.
More dairy did not change weight significantly unless calories were cut
This meta-analysis pooled 14 randomized trials in adults, covering dairy as a group. On calorie-restricted diets, high-dairy diets led to 1.29 kg more weight loss than usual weight-loss diets. Waist size dropped 2.43 cm more. Without calorie restriction, weight, fat mass and waist did not change significantly.
Newborn formula cut milk allergy at 6 months from 6.8% to 0.8%
This randomized trial in Okinawa, Japan, followed 491 infants: 242 took cow’s milk formula and 249 avoided it. The formula group got at least 10 mL a day as newborns, and a food challenge checked allergy at 6 months. The test was infant formula from 1 to 2 months of age, not milk from a carton, and about 70% of babies were still breastfed. NHS advice is still no cow’s milk as a drink before age 1.
Daily milk added about 0.4 cm of height a year
This meta-analysis pooled 12 controlled studies from several countries, 7 of them run since the 1990s. They added about a cup (245 mL) of milk a day. Many lacked randomization or had groups that differed at the start, so the author rates the evidence as moderate quality. Milk seemed to work better than other dairy, and the effect was larger in teenagers and in children short for their age. The newer meta-analysis of randomized trials above, in ages 6 to 18, found no height change.
Self-described severely lactose-intolerant adults had only trivial symptoms on 2% milk
In this double-blind crossover trial, 30 adults reported symptoms after less than 240 mL of milk. Each drank a daily cup of regular or lactose-free milk for a week. Symptom scores did not differ significantly, and 21 of the 30 turned out to malabsorb lactose. Those malabsorbers had 2.5 more episodes of flatus (gas) a day on regular milk.
Two daily cups with meals caused no significant extra symptoms
In this double-blind crossover trial, people with low lactase drank regular or lactose-free milk for 7 days. They had 240 mL twice a day, the two doses taken with food. People who called themselves markedly lactose intolerant (19) reported more gas on both milks than those who did not (13). The authors read this as underlying gas wrongly blamed on lactose, and the trial was small.
Three daily servings of 1% milk left daytime blood pressure unchanged
In this controlled-feeding crossover trial, 55 healthy adults with high-normal blood pressure ate three 6-week diets. The diets had no dairy, 3 servings of 1% milk, or 1 serving of cheddar. Daytime blood pressure did not change, but on the milk diet triglycerides were 9.9% higher and HDL cholesterol 4.1% lower than with no dairy. C-reactive protein, an inflammation marker, and blood sugar markers did not differ.
Swedish women drinking most milk had nearly double the death risk
Two cohorts followed 61,433 women for 20.1 years and 45,339 men for 11.2 years. In women, the top group, three or more glasses a day, had a hazard ratio for death of 1.93 against less than one glass. Per glass, the hazard ratio for death was 1.15 in women and 1.03 in men. Hip fractures in women did not drop (1.09 per glass). The authors warn of residual confounding and reverse causation, and declared no outside support.
Daily milk was not linked to hip fracture risk in women
This meta-analysis pooled prospective cohorts, which follow people forward in time. In 6 studies of 195,102 women (3,574 hip fractures), the risk ratio per daily glass was 0.99. In 3 studies of 75,149 men (195 hip fractures) it was 0.91, not significant. The authors say more data are needed in men.
In US studies only, more milk went with 25% fewer hip fractures
This meta-analysis pooled prospective cohorts. Across all studies, the risk ratio for the highest against the lowest milk intake was 0.86, not significant. It was 0.75 in the USA and 1.00 in Scandinavia. The authors suggest vitamin D fortification of milk, more common in the US, may explain the gap.
Top dairy intake did not clearly lower fracture risk, except vertebral fractures
This meta-analysis pooled 10 cohort studies in Europeans and non-Hispanic white North Americans, covering dairy as a group. The hazard ratio was 0.95 for any fracture and 0.87 for hip fracture, a borderline result. Vertebral fractures were 18% less likely (0.82), also borderline. Results varied a lot between studies for any fracture, and bone density results were too mixed to pool.
Milk-raising gene variants were not linked to fewer hip fractures
This Mendelian randomization study in 97,811 Danes used inherited lactase persistence as a stand-in for lifelong milk drinking. Each variant copy added about 0.58 glasses of milk a week, yet the hip fracture hazard ratio was 1.01. A meta-analysis with other Northern European studies found no link to fracture overall. It found slightly higher femoral neck (upper thigh bone) density.
Dairy and milk showed no link with death, heart disease or stroke
This dose-response meta-analysis pooled 29 cohorts with 938,465 people. Fermented dairy such as yogurt and cheese showed slight inverse links (risk ratio 0.98 per 20 g a day). These faded after one large Swedish study was removed. The last author received funding for this analysis from the Global Dairy Platform, Dairy Research Institute and Dairy Australia.
Daily milk went with 10% lower risk of death or heart events
This cohort in 21 countries followed 136,384 people aged 35 to 70 for 9.1 years. More than 1 serving of milk a day, against none, gave a hazard ratio of 0.90 for death or major heart events. The trend test was borderline. More than 2 servings of total dairy a day went with 0.84 for the same outcome and 0.66 for stroke.
Each 200 g of daily milk went with 7% lower stroke risk
This dose-response meta-analysis pooled 18 cohort studies with 762,414 people and 29,943 strokes. The risk ratio was 0.82 in East Asian studies and 0.98 in Western ones, so the link was mainly in East Asia. Results varied a lot, and high-fat milk was directly linked to stroke in a few studies. The Dutch Dairy Association partly funded it with an unrestricted grant.
Milk was not linked to type 2 diabetes, but yogurt was
This dose-response meta-analysis of 22 cohorts covered 579,832 people and 43,118 cases. Total dairy had a small inverse link (risk ratio 0.97 per 200 g a day), and low-fat dairy a similar one that was not significant. Other types, including milk, showed none. Yogurt at about 80 g a day had a risk ratio of 0.86.
Milk-raising gene variants were not linked to lower type 2 diabetes risk
This Mendelian randomization study covered 97,811 Danes. The odds ratio for diabetes with lactase persistence was 0.96, and for overweight or obesity 1.06. Any milk intake against none had a diabetes hazard ratio of 1.10, also not significant.
High-fat milk was linked to a 15% higher risk of death
This meta-analysis pooled prospective cohorts. The higher risk was for death from any cause; high-fat milk also went with 1.09 for cardiovascular death and 1.17 for cancer death. Total dairy went with lower cardiovascular death (0.93). These are links in observational data and cannot show cause.
Canadian children on whole milk had lower obesity odds than skim drinkers
The CHILD cohort covered 2,043 children at age 5 and 1,574 at age 8. At age 5, whole milk (3.25%) went with a BMI z-score 0.34 lower than skim, and an obesity odds ratio of 0.22. Most children drank 2% (48.9%) or whole (23.9%) milk. The authors report no conflicts of interest.
More milk, even low-fat, was linked to slightly higher prostate cancer risk
This meta-analysis of 32 prospective studies was part of the Continuous Update Project. Per 200 g a day, the risk ratio was 1.03 for total milk and 1.06 for low-fat milk. Total dairy was 1.07 per 400 g and dietary calcium 1.05 per 400 mg. Supplemental calcium was not linked to total prostate cancer, only to fatal cases.
Cohort studies found no link between milk and prostate cancer
This meta-analysis pooled 45 observational studies with 26,769 cases. In cohorts the risk ratio for milk was 1.06, not significant. Pooled case-control studies of dairy gave 1.14, a borderline result. Case-control studies of milk differed too much to combine.
Whole milk went with fatal progression in men with prostate cancer
This cohort followed 21,660 US physicians for 28 years, with 2,806 cases and 305 deaths. For fatal cancer in the whole cohort, whole milk gave a borderline hazard ratio of 1.49. Among men already diagnosed, whole milk gave 2.17 for progression to fatal disease. Skim and low-fat milk went with low-grade, early stage, screen-detected cancers.
Adventist men with high dairy intake had 27% higher prostate cancer risk
This cohort of 28,737 Adventist men, many with very little dairy, had 1,254 cases over 7.8 years. Men at the 90th percentile of dairy intake (430 g a day) were compared with men at the 10th (20.2 g a day). This was dairy as a group, not milk alone, and risk rose most at low intakes. Nondairy calcium was not linked (hazard ratio 1.16).
Milk-raising gene variants were not linked to prostate cancer
This two-sample Mendelian randomization used UK Biobank, FinnGen and cancer consortia data. Per milk-raising variant, the odds ratio was 1.01 for prostate cancer and 0.95 for colorectal cancer, a lower risk. One cohort, FinnGen, did show a higher prostate cancer risk (1.07). The authors declared no competing interests.
Daily milk went with 9% lower colorectal cancer risk per 200 g
Meta-analysis of 19 cohort studies. The risk ratio was 0.91 per 200 g of milk and 0.83 per 400 g of total dairy, with inverse links restricted to colon cancer. Cheese showed no link.
Heavy milk drinkers were more likely to have acne, skim milk most
This meta-analysis pooled observational studies. For the highest against the lowest intake, the odds ratio was 1.48 for milk, 1.25 for low-fat milk and 1.82 for skim milk. Yogurt and cheese showed no link. Observational data cannot show that milk causes acne.
Just 0.54% of European children had proven milk allergy by age 2
The EuroPrevall birth cohort followed 12,049 babies in 9 countries, 9,336 of them to age 2. Allergy was suspected in 358 children and confirmed in 55 by a double-blind food challenge. A year later, 69% of those retested tolerated milk, so most outgrew it. This included all children with non-IgE allergy and 57% of those with IgE-type allergy, the antibody kind.
In referral patients, 79% outgrew IgE-mediated milk allergy by age 16
This study followed 807 patients from a large referral population. Resolution was 19% by age 4, 42% by age 8 and 64% by age 12. Higher cow’s milk IgE (allergy antibody) levels went with persistent allergy, and asthma and allergic rhinitis also predicted the outcome. The authors say outcomes were worse than earlier reports, which may reflect a referral population.
Babies given cow’s milk had 3.76 times the iron-deficiency anemia risk
This systematic review covered 23 studies in infants up to 3 years old: 1 randomized trial, 4 non-randomized trials, 8 case-control studies and 10 cohorts. The anemia risk ratio, against follow-on formula, came from 4 pooled studies. For type 1 diabetes, 6 of 7 case-control studies found no difference by age at first cow’s milk.
Who should be careful with semi-skimmed milk
Most people can drink semi-skimmed milk freely. Take a little care if one of these applies:
- Babies under 12 months. The NHS (the UK’s health service) says cows’ milk should not be given as a drink until a baby is 1 year old. In a review, babies who drank it had 3.76 times the risk of iron-deficiency anemia of those on follow-on formula. The NHS also says not to give skimmed or 1% milk as a drink to children under 5.
- Cow’s milk allergy. This is an immune reaction, not lactose intolerance, and it mostly affects infants. In Europe, 0.54% of children had challenge-proven allergy by age 2, and 69% of those retested tolerated milk a year later. The NHS says to call 999 for swollen lips, mouth or tongue, or fast or difficult breathing.
- Lactose intolerance. It is a gut reaction to lactose, not an allergy. The NIDDK, a US health institute, says most people with it can have some lactose. Many can have 12 g, about the amount in 1 cup of milk, with no or mild symptoms. Lactose-free milk is another option.
How to choose, store and eat semi-skimmed milk
- Semi-skimmed or 2%: one cup (246 g) gives 123 kcal, 4.9 g of fat and 295 mg of calcium. The EU band for semi-skimmed is 1.5 to 1.8% fat.
- Whole or lower fat: the US guidelines for 2025 to 2030 say to include full-fat dairy, while the NHS suggests lower-fat milk for adults. One pooled analysis linked high-fat milk to a 15% higher risk of death. Canadian children on whole milk had lower odds of obesity.
- Lactose-free or smaller servings: the NIDDK says lactose-free milk is just as healthy as regular milk. People with self-reported severe intolerance had only trivial symptoms on a cup of 2% milk a day.
Choose pasteurized milk, keep it in the fridge and use it by the date on the carton. The USDA entry here has no added vitamin D, but milk is fortified with vitamin D in the United States, so check the label.
Sources
- Milk and dairy in your diet. NHS.
- Food allergies. NHS.
- Lactose intolerance. NHS.
- Definition and facts for lactose intolerance. National Institute of Diabetes and Digestive and Kidney Diseases.
- Eating, diet and nutrition for lactose intolerance. National Institute of Diabetes and Digestive and Kidney Diseases.
- Dietary Guidelines for Americans, 2025 to 2030. US Department of Health and Human Services and US Department of Agriculture.
- Council Regulation (EC) No 1234/2007, Annex XIII: drinking milk categories (repealed).
- Griebler U, Bruckmüller MU, Kien C, et al. Health effects of cow’s milk consumption in infants up to 3 years of age: a systematic review and meta-analysis. Public Health Nutr. 2016.
- Schoemaker AA, Sprikkelman AB, Grimshaw KE, et al. Incidence and natural history of challenge-proven cow’s milk allergy in European children: EuroPrevall birth cohort. Allergy. 2015.
- Suarez FL, Savaiano DA, Levitt MD. A comparison of symptoms after the consumption of milk or lactose-hydrolyzed milk by people with self-reported severe lactose intolerance. N Engl J Med. 1995.
- Bischoff-Ferrari HA, Dawson-Hughes B, Baron JA, et al. Milk intake and risk of hip fracture in men and women: a meta-analysis of prospective cohort studies. J Bone Miner Res. 2011.
History of semi-skimmed milk
People were milking cattle by the seventh millennium BC. The earliest direct evidence is traces of milk fat in more than 2,200 pottery vessels from the Near East and southeastern Europe. They date to the fifth to seventh millennia BC, and milking looks especially important in northwestern Anatolia.
Drinking milk as an adult needed a genetic change. Around 7,000 pottery fat residues from more than 550 sites show that milk use was widespread in Europe from the Neolithic onward.
In the UK Biobank, the lactase persistence variant was only weakly linked to how much milk people drink today. The authors of a 2022 study suggest the trait spread because milk was risky for people without it in famine or heavy infection.
Heating milk to make it safer came late. Louis Pasteur showed in 1864 that brief heating kept wine from spoiling, and Franz von Soxhlet suggested pasteurizing milk in 1886. New York City made it mandatory in 1910, and the first US state law followed in 1947. A federal rule for milk in interstate trade came in 1973.
Taking the fat out got easy with the centrifugal separator, first made by Gustaf de Laval in the late 19th century. Today most skim milk is made by spinning whole milk in a centrifuge, and lower-fat milks can be made by mixing skim and whole milk. Skim milk was not always for people: a 1930 Oregon bulletin called it the very best supplement for growing pigs.
Rules then sorted milk into fat bands. Until 1 January 2008, milk outside the EU ranges could not be sold as milk. That kept 1% milk off British shelves until lobbying changed it.
In 1998, British retailers adopted red for skimmed, green for semi-skimmed and blue for whole milk. John Morris of the British Retail Consortium proposed the scheme.
The advice has swung back and forth since. The US guidelines for 2025 to 2030 say to include full-fat dairy with no added sugars and to aim for 3 servings a day. They also say to keep saturated fat under 10% of daily calories.
Sources
- Evershed RP, Payne S, Sherratt AG, et al. Earliest date for milk use in the Near East and southeastern Europe linked to cattle herding. Nature. 2008.
- Evershed RP, Davey Smith G, Roffet-Salque M, et al. Dairying, diseases and the evolution of lactase persistence in Europe. Nature. 2022.
- Pasteurization. Wikipedia.
- Cream separator. Wikipedia.
- Fat content of milk. Wikipedia.
- Skimmed milk. Wikipedia.
- Council Regulation (EC) No 1234/2007, Annex XIII: drinking milk categories (repealed).
- Dietary Guidelines for Americans, 2025 to 2030. US Department of Health and Human Services and US Department of Agriculture.
Frequently asked questions about semi-skimmed milk
Is semi-skimmed milk good for you?
For most people it can fit well, but the evidence for health benefits is mixed. One cup (246 g) has 123 kcal, 8.1 g of protein, 295 mg of calcium and 35% of the Daily Value of vitamin B2. In a meta-analysis of 29 cohorts, total dairy and milk showed no link with death, heart disease or stroke. No study here reaches strong evidence, and people with a cow’s milk allergy should avoid it.
Is semi-skimmed milk the same as 2% milk?
Nearly. In the EU, semi-skimmed milk holds 1.5 to 1.8% fat by law, while US 2% milk, called reduced-fat, holds 2%. The USDA entry behind this page is “Milk, reduced fat, fluid, 2% milkfat, without added vitamin A and vitamin D”. So a cup of European semi-skimmed has a little less fat than the 4.9 g listed here.
What is the difference between semi-skimmed and whole milk?
Fat: whole milk must hold at least 3.5% fat in the EU, against 1.5 to 1.8% for semi-skimmed. One cup of this 2% milk has 4.9 g of fat, 3.1 g of it saturated. Experts differ: the 2025 to 2030 US guidelines say to include full-fat dairy, and the NHS suggests lower-fat milk for older children and adults. The studies here point both ways: high-fat milk was linked to a 15% higher risk of death in one pooled analysis. Whole milk was linked to lower obesity odds (0.22) than skim milk in Canadian 5-year-olds.
How much protein and calcium does semi-skimmed milk have?
One cup (246 g) has 8.1 g of protein and 295 mg of calcium, about 16% and 23% of the daily values. Per 100 g that is 3.3 g of protein and 120 mg of calcium, and the protein scores 100/100 for essential amino acids. This USDA entry lists no vitamin D, but milk is fortified with vitamin D in the United States, so check the label.
Does milk make your bones stronger?
In trials it adds a little bone density, but the cohorts do not clearly show fewer fractures. Dairy raised whole-body bone mineral content by 3.0% in children and bone density at the hip and spine in postmenopausal women. Those trials measured density, not fractures. In cohorts, the risk ratio for hip fracture was 0.99 per daily glass in women, and a Swedish cohort found 1.09 per glass. US studies alone showed a 25% lower hip fracture risk.
Can you drink semi-skimmed milk if you are lactose intolerant?
Many people can, in small amounts: the NIDDK says most people with lactose intolerance can have some lactose. About 12 g, the amount in around 1 cup of milk, often causes no or only mild symptoms. This cup has 12.3 g of lactose, and lactose-free milk is another option. In a double-blind trial, 30 adults with self-reported severe intolerance had only trivial symptoms on a cup of 2% milk a day.
Is milk bad for you?
Not for most people, though a few links are worth knowing. Pooled cohorts found no link between milk and type 2 diabetes, and a gene-based study found none either. Prostate cancer is less clear: a meta-analysis of 32 studies found 1.03 per 200 g of milk a day and 1.06 for low-fat milk. Another of 45 studies found no link in cohorts, and these small links cannot show cause.
Is semi-skimmed milk good for weight loss?
It can fit in a calorie-controlled diet, but milk is not a weight-loss food. In a meta-analysis of 14 trials, more dairy without calorie limits did not change weight significantly. On calorie-restricted diets, high-dairy diets lost 1.29 kg more. A cup has 123 kcal, so it counts toward the day’s total.