Full-fat or low-fat dairy: what trials and long-term studies show about weight, cholesterol and the heart

For two generations the advice at the dairy case has been to reach for the skim milk and the low-fat yogurt. Whole milk and full-fat cheese carry saturated fat, saturated fat raises cholesterol, and cholesterol clogs arteries: the logic seemed complete.

Heart guidelines still follow it. The American Heart Association’s current dietary guidance names low-fat or fat-free dairy products among its preferred sources of protein. Over the past decade, though, trials and long-term studies have repeatedly failed to find the harm that the logic predicts. A Canadian trial published this year, and widely reported as overturning the old advice, is the latest. It is also weaker than its coverage suggested, and a close reading shows why the question is not yet closed.

Why have guidelines favored low-fat dairy?

The reasoning starts from a single nutrient. Dairy fat is mostly saturated, and diets high in saturated fat raise LDL cholesterol, the kind associated with heart disease. An American trial team summarized the position before testing it: dietary guidelines traditionally recommend low-fat dairy because dairy’s high saturated fat content is thought to promote cardiovascular disease.

That is an inference about a nutrient, applied to a food. Whether milk, yogurt and cheese behave the way their fat content predicts is a question only studies of the foods themselves can answer.

What happened when adults ate three servings of full-fat dairy a day?

Harvey Anderson, a professor of nutritional sciences at the University of Toronto, ran a trial with colleagues in Toronto and Halifax. They enrolled adults aged 25 to 60 who were overweight or had obesity but were otherwise healthy, and 74 participants completed the study.

Everyone was counseled to follow Canada’s national food guide. One group cut 500 calories a day and kept dairy to a single low-fat serving at most. A second cut the same calories and then added them back as three servings of full-fat dairy. A third ate three servings of full-fat dairy with no calorie limit.

After 12 weeks, the measures that low-fat advice is meant to protect had not moved. In the paper’s words, neither treatment nor time affected waist circumference, fat and fat-free mass, resting metabolic rate, fasting blood cholesterol. Triglycerides, a blood fat, rose in every group a month in but returned to baseline by week 12. The calorie-cutting group lost about 0.7 kilograms and the group that swapped those calories for dairy gained about 0.35, a difference of roughly two pounds.

“Those that had three servings of dairy didn’t have adverse levels of blood cholesterol or lipids or evidence of insulin resistance,” Anderson said in a university statement.

Three details in the paper temper that. The first is who paid. The authors disclose that this research was supported by Dairy Research Cluster 3 (Dairy Farmers of Canada and Agriculture and Agri-Food Canada), and state that the funders had no role in publication. The second is size. The team had planned for 50 people in each group over 24 weeks. The pandemic and the end of funding left them with half the participants and half the time, a shortfall they describe as limiting the recruitment of the targeted sample size of 50 participants per group.

The third matters most for the headline claim. No group ate low-fat dairy in the same amounts. The comparison was between full-fat dairy and very little dairy, so the trial shows that adding full-fat dairy did no measurable harm in 12 weeks. It does not show that full-fat is as good as or better than low-fat.

Has full-fat dairy been compared directly with low-fat?

It has, by a separate team. Researchers at the Fred Hutchinson Cancer Center in Seattle recruited 72 adults with metabolic syndrome, a cluster of risk factors that includes high blood pressure, high blood sugar and excess weight around the waist. After a month on minimal dairy, participants were assigned to stay on that diet or to eat a little over three servings a day of either low-fat or full-fat milk, yogurt and cheese for 12 weeks.

The result matched the Canadian one and answered the question the Canadian trial left open. In the authors’ conclusion, a diet rich in full-fat dairy had no effects on fasting lipid profile or blood pressure compared with either of the other diets. These were people at raised risk, for whom a rise in cholesterol would have mattered most.

What do long-term studies show about dairy and heart disease?

Twelve-week trials measure risk markers. Whether people who eat dairy fat go on to have more heart attacks takes years and very large numbers, and there the evidence comes from observational studies.

The largest is PURE, which recorded the diets of 136,384 people in 21 countries and followed them for about nine years. Its authors concluded that dairy consumption was associated with lower risk of mortality and major cardiovascular disease events in a diverse multinational cohort. The pattern held for milk and yogurt. Butter was the exception: butter intake was low and showed no protective link.

A meta-analysis from the University of Reading and partners pooled studies from Europe, North America and Asia. A total of 29 cohort studies were available for meta-analysis, covering more than 900,000 people. The verdict was duller than PURE’s and arguably more reassuring. The combined data demonstrated neutral associations between dairy products and cardiovascular and all-cause mortality.

One difficulty with food questionnaires is that answers are often inaccurate. A consortium based at the University of Cambridge got around that by measuring fatty acids in blood that come largely from dairy fat. The team pooled the findings from 16 prospective cohort studies and found that adults with higher levels of these markers developed type 2 diabetes less often. The authors add a caution of their own: the biomarkers do not distinguish between different food sources of dairy fat, so cheese, yogurt and butter cannot be told apart.

Why might dairy fat behave differently from other saturated fat?

The leading explanation is that a food is more than its nutrient label. In cheese and yogurt, fat is bound up with protein, calcium and the products of fermentation, and this structure, often called the food matrix, may change how the fat is absorbed. Anderson’s team invokes it to explain why their participants did not gain weight.

A group of nutrition scientists made the wider case in the Journal of the American College of Cardiology in 2020. They argued that the health effects of foods cannot be predicted by their content in any nutrient group, and that for whole-fat dairy the totality of available evidence does not support further limiting the intake of such foods. That paper is a reassessment by researchers who favor changing the advice. The heart association’s guidance, published a year later, kept its low-fat wording.

Small, short and partly industry-funded: the limits of the dairy evidence

No randomized trial has followed people long enough to count heart attacks or strokes on full-fat versus low-fat dairy. The trials that exist last about three months and measure cholesterol and blood pressure.

The long-term studies have the usual weakness of observational research. Those who eat more dairy differ from those who eat less in income, in the rest of their diet and in where they live, and statistical adjustment only goes so far. PURE also spans countries with very different diets and incomes, which makes like-for-like comparison hard.

The Canadian volunteers were also a narrow group. People with high blood pressure, raised blood sugar or chronic illness were excluded, as were women past menopause. Full-fat dairy carries more calories than low-fat, and how it fits into the diet of someone with high cholesterol or heart disease depends on their overall risk, which their doctor assesses.

Trials and cohorts have not found the harm from full-fat dairy that its saturated fat content predicts, and guidelines have not yet changed in response.