Are Milk Products Bad for You? Depends Which One You Mean

Milk, yogurt and cheese are not one food, and the evidence stops pretending they are. The PURE study, a cohort of more than 136,000 adults across 21 countries, found higher dairy intake tracked with lower rates of death and major cardiovascular events over about nine years^[1]. A pooled analysis of three US cohorts and 14 prospective studies worldwide found yogurt specifically associated with a lower risk of type 2 diabetes, while total dairy and milk barely moved the needle^[2]. And one of the largest cohorts ever run on plain milk found the reverse in Swedish women: more of it tracked with higher death rates instead of longer ones^[3].

None of that is a contradiction. It is what happens when “is dairy bad for you” gets asked about foods that differ as much as a glass of milk, a tub of live-culture yogurt and a wedge of aged cheddar.

What the largest human studies found

Start with the best evidence available: decades of cohort data. Nobody randomizes tens of thousands of people to a lifetime of dairy or no dairy, so what exists is decades of tracking what people already eat.

The two largest findings:

• The PURE cohort followed 136,384 adults aged 35 to 70 across five continents for a mean of 9.1 years. People eating more than two servings of dairy a day had a lower risk of the combined outcome of death or major cardiovascular events (hazard ratio 0.84) compared with people eating none^[1].
• A pooled analysis of US and global cohorts, covering 459,790 participants in the meta-analysis arm, found total dairy had no appreciable link to type 2 diabetes risk (hazard ratio 0.99 per daily serving), but yogurt did: each additional daily serving tracked with a lower diabetes risk (hazard ratio 0.83 in the US cohorts, relative risk 0.82 in the pooled meta-analysis)^[2].

136,384 adults followed in the PURE dairy cohort across 21 countries PMID 30217460

The pattern worth remembering: across nearly every outcome studied, yogurt’s association looks better and more consistent than milk’s or total dairy’s. That shows up again below.

The milk study that undercuts the “builds strong bones” assumption

If one piece of dairy evidence should make you pause before another glass of milk, it is this one. A Swedish cohort of 61,433 women, tracked for a mean of 20.1 years, found no fracture protection from higher milk intake at all. Women who drank three or more glasses of milk a day had nearly double the all-cause mortality rate of those who drank less than one glass, and each additional glass was linked to a higher hip fracture risk as well, running counter to the assumption that milk protects bones^[3].

That is the opposite of what decades of “milk builds strong bones” messaging implies. It does not mean milk caused those deaths and fractures. This is observational data, and the authors who ran it flagged the same confounders and reverse-causation risks that apply to any diet-and-disease cohort: people who drink a lot of milk differ from people who don’t in ways a model can only partly adjust for.

The one real downside: a modest prostate cancer signal

Every additive page on this site has an honest downside, and dairy’s is real. A meta-analysis of prospective cohort studies (the Continuous Update Project, 32 studies) found higher intake of total dairy, milk and cheese each associated with a modestly elevated risk of total prostate cancer: a relative risk of 1.07 per 400 grams of total dairy a day, 1.03 per 200 grams of milk, and 1.09 per 50 grams of cheese^[4].

The leading explanation is hormonal. A systematic review and meta-analysis found moderate evidence that milk intake raises circulating insulin-like growth factor 1 (IGF-1), and that higher IGF-1 itself tracks with greater prostate cancer risk. The authors call IGF-1 “a potential mechanism,” explicitly short of a proven causal pathway^[5].

IGF-1 is the same growth hormone pathway that shows up anywhere cells are told to proliferate, including the one people chase when they talk about growth hormone and muscle growth. A pathway that helps build tissue in one context is not automatically dangerous in another, which is exactly why “raises IGF-1” counts as a lead on prostate cancer rather than a verdict.

An association in cohort data is not proof dairy causes prostate cancer. Diet quality, calcium intake and how often men get screened can all move both variables at once. What the honest version of this downside looks like:

• The risk increase per study is modest: 3% to 9% per typical serving increase, well short of a doubling.
• It applies to total prostate cancer, and the Continuous Update Project itself called for more detail on which prostate cancer subtypes are involved.
• It gives a man with a strong family history of prostate cancer a specific reason to moderate dairy. It doesn’t generalize to everyone else.

Why yogurt and whole milk part ways

Fermentation is the likely reason yogurt keeps showing up differently in these cohorts. A review of yogurt’s role in metabolic health proposes that its live bacterial cultures, altered food matrix, and concentrated calcium and protein may explain its more consistent link to lower diabetes risk, through effects on gut microbiota, satiety and glycemic response. The authors are explicit that these remain proposed mechanisms still being worked out rather than settled fact^[6].

Whole-fat dairy’s saturated fat gets blamed for raising cardiovascular risk, and yet the cohort data does not always back that up. The Framingham Offspring Study found adults with the highest intake of dairy-derived saturated fat had a less atherogenic lipid profile, including larger LDL particle size, than those with the lowest intake, with no such pattern for saturated fat from non-dairy sources^[7].

That is a genuinely surprising result, and it is also why dietary guidance has started separating “saturated fat” as a single nutrient from the specific food it arrives in.

The side effect most people have: lactose intolerance

Forget disease risk for a moment. The most common adverse reaction to milk products is not a cohort statistic, it is a digestive one, and it affects most of the world’s adults.

68% of adults worldwide have reduced lactase activity and cannot fully digest lactose PMID 28690131

A systematic review and meta-analysis covering 62,910 participants across 89 countries found that figure, with sharp regional variation: about 28% in western, southern and northern Europe compared with about 70% in the Middle East, and other regional surveys reporting similarly high rates across East Asia^[8].

That variation tracks ancestry the same way alcohol flush reaction tracks the ALDH2 gene variant common in East Asian populations: a single inherited trait, independent of diet or willpower. We’ve covered the chemistry of lactose itself in more depth, including why “lactose-free” milk still isn’t vegan.

Cow’s milk protein allergy is a separate condition from lactose intolerance, driven by the immune system rather than a missing enzyme, and it is mostly a childhood issue. European pediatric gastroenterology guidelines report that children managed with strict avoidance develop tolerance in more than 75% of cases by age 3 and more than 90% by age 6^[9].

What this means if you’re deciding whether to drink milk

The evidence at a glance:

Dairy and health, by product and outcome

Human cohort and meta-analysis findings covered above.

A few practical lines follow directly from this evidence, rather than from stretching it:

• If metabolic health is the goal, the more consistent evidence favors yogurt over large volumes of liquid milk, since yogurt’s lower diabetes risk shows up across far more cohorts than any benefit tied to milk itself^[2].
• If you have a strong family history of prostate cancer, that is a specific, documented reason to moderate dairy intake. It does not generalize to a man with no such history making a routine grocery choice^[4].
• If you get bloating, gas or diarrhea after milk, you are one of roughly two in three adults worldwide dealing with an extremely common digestive response^[8].

Bottom line: no current evidence shows that moderate milk product intake shortens the life of someone without a specific risk factor, lactose intolerance, a milk allergy, or a family history of prostate cancer. What the cohorts show is a set of modest, product-specific signals rather than a single verdict on an entire food category.

If you are managing a dairy allergy or just trying to spot hidden milk solids and whey on a label, that is exactly the blind spot a barcode scanner misses and a label reader does not. The EatMatrix app reads the ingredient panel itself, which matters more for dairy than for most categories given how many names milk protein hides behind. For a similar evidence-by-dose pattern in a different nutrient, our look at vitamin D found the same story: the ingredient isn’t the danger, the specifics are.

This article summarises publicly available regulatory assessments and peer-reviewed research about food ingredients. It is general information and is not intended to diagnose, treat, cure, or prevent any disease, and it is not personalised dietary or medical advice. Regulatory positions change; the EatMatrix app reflects the current assessments in our database. If you have an allergy, a medical condition, or are pregnant, talk to a qualified clinician or dietitian before changing what you eat.