In this guide
Dairy has been through every possible reputation in the past twenty years: essential, harmful, overrated, underrated. Fermented dairy (yoghurt, kefir, cheese and sour cream) has quietly become the favourite child, with study after study linking it to better health. This month a large US analysis added a twist: the apparent benefit seems to follow a curve, not a straight line, and it fades at higher intakes.
We read the coverage and the paper's reported results (published in Scientific Reports and summarised by News-Medical on 16 September), and tried to answer what a protein-conscious reader actually wants to know: should I be eating more of this, how much, and what's the catch?
The short version: in 37,366 US adults followed for a median of about 9.4 years, people who ate any fermented dairy had an 18% lower adjusted risk of death from any cause and a 20% lower risk of cardiovascular death than non-consumers. The relationship was J-shaped: lower mortality up to roughly 150 g a day, with no significant association at higher intakes. It's observational, based on 1–2 days of diet recall, and consumers differed from non-consumers in many ways, so it shows association, not proof that dairy extends life.
Why Fermented Dairy Keeps Making Headlines
Fermentation changes dairy. Bacteria convert lactose into lactic acid, the proteins partly break down, and some products contain live cultures. Researchers have spent years asking whether those changes make yoghurt and cheese behave differently from plain milk in the body. Several large cohort studies and meta-analyses point toward small benefits, and the new analysis from the National Health and Nutrition Examination Survey (NHANES) is one more data point in that direction, with a useful complication.
The NHANES Study at a Glance
| Item | Details |
|---|---|
| Journal and date | Scientific Reports, 2026 (Zhang and Li; doi:10.1038/s41598-026-71687-9); News-Medical summary 16 September 2026 |
| Data | NHANES, ten survey cycles from 1999 to 2018, linked to deaths through 31 December 2019 |
| Participants | 37,366 US adults aged 20+ |
| Follow-up | Median about 9.4 years; 5,646 deaths, including 1,791 from cardiovascular disease |
| Fermented dairy | Cheese, yoghurt, sour cream and buttermilk, from 24-hour dietary recalls (one recall in 1999–2002; two averaged afterwards) |
| Main result | Consumers vs non-consumers: adjusted hazard ratio 0.82 for all-cause and 0.80 for cardiovascular mortality |
| Dose-response | J-shaped; lower mortality below about 150 g/day; no significant association at 150 g/day or more |
| Stated limitations | Short dietary recall; no data on dairy fat content; kidney disease not adjusted for; residual confounding |
Who Ate Fermented Dairy, and How Much
Among consumers, the median intake was only about 28 g a day, roughly a thin slice of cheese or a few spoonfuls of yoghurt. Consumers were more often younger, female, non-Hispanic White, married and better educated. They were more physically active, had less hypertension, diabetes, heart disease and stroke, and drank more alcohol. They had slightly higher body mass index and total cholesterol, and higher HDL.
Crude death rates were lower in consumers (12.7% versus 18.1% for all causes; 3.9% versus 5.9% for cardiovascular disease), but since consumers were younger, the crude gap exaggerates the effect. The adjusted figures (18% and 20% lower risk) are the ones to use.
The Results
- Any versus none: adjusted hazard ratio 0.82 (about 18% lower) for all-cause death and 0.80 (about 20% lower) for cardiovascular death.
- J-shaped curve: mortality fell as intake rose from low levels, with the best-looking range below about 150 g a day.
- Higher intakes: at 150 g/day or more, no statistically significant association was found. That's different from "harmful"; it means any benefit couldn't be distinguished from chance.
- Consistency: results were similar across subgroups and in a sensitivity analysis adjusting for energy intake.
The 150-Gram Problem: Not All Grams Are Equal
This is the practical catch. The study pooled yoghurt, cheese, sour cream and buttermilk and measured them in grams. But 150 g means very different things across foods:
| Food | What 150 g looks like | Approx. calories |
|---|---|---|
| Plain yoghurt | One small pot | 90–130 kcal |
| Kefir | About 150 ml, a little over half a cup | 80–100 kcal |
| Cottage cheese | About ⅔ cup | 120–140 kcal |
| Cheddar | A large chunk, around five slices | About 600 kcal |
So "up to 150 g" can mean a daily pot of yoghurt or a heap of cheese with roughly five times the calories. Because the study can't separate them, the threshold isn't a target to hit. If you want to apply it, think in terms of portions of yoghurt, kefir or cottage cheese, with cheese as a smaller, flavour-forward addition.
The Healthy-User Problem
Observational dairy studies face a recurring trap. People who eat yoghurt tend to be health-conscious in other ways, and adjustment can't remove all of that. The reverse also happens: people with illness, lactose problems or special diets may avoid dairy, which can make non-consumers look worse. Note too that consumers in this study had more cancer and drank more alcohol, which cuts the other way, a reminder that these cohorts are messy in both directions.
The authors acknowledge that the downward slope below 150 g could partly reflect who the very low and non-consumers are, not what dairy does. The J shape itself may be partly an artefact of how sparse the data are at high intakes.
How It Fits With the Wider Evidence
A 2026 meta-analysis in Frontiers in Nutrition, pooling 50 cohort studies covering more than three million adults, reported modest associations of yoghurt (around 7% lower all-cause mortality) and smaller ones for cheese (around 3%). A Swedish cohort analysis has examined high-fat and low-fat fermented milk and cheese together with blood protein signatures, and other trials are exploring effects on inflammation and gut microbes. The overall picture is consistent but modest: small benefits for yoghurt-type foods, neutral to slightly favourable for cheese, and nothing that proves cause.
Fermented Dairy as a Protein Source
For readers who track protein, the real value of fermented dairy is practical. It's compact, widely available, and rich in high-quality protein, mostly casein and whey, with a high leucine content that supports muscle protein synthesis. Approximate values:
| Food (approximate serving) | Calories | Protein (g) | Notes |
|---|---|---|---|
| Plain Greek yoghurt, 170 g, low-fat | 100–130 | 15–17 | Check labels for added sugar |
| Skyr, 150 g | 90–100 | 15–17 | High protein, low fat |
| Plain kefir, 240 ml | 110–150 | 8–11 | Drinkable; contains live cultures |
| Cottage cheese, 1 cup (225 g) | 180–220 | 24–25 | Often high in sodium |
| Cheddar, 28 g | 115 | 7 | About 180 mg sodium and 5–6 g saturated fat |
| Parmesan, 28 g | 110–120 | 10 | Intense flavour; 330–440 mg sodium |
Use the protein intake calculator to set your daily target, and see our guide to protein sources for muscle gain for how dairy compares with eggs, fish, poultry and plant options.
Choosing Well: Sugar, Salt and Saturated Fat
- Sugar. Flavoured and fruit yoghurts can carry 12–20 g of sugar per pot, much of it added. Choose plain and add fruit, nuts or a little honey.
- Sodium. Cheese and cottage cheese can be salty. If you watch blood pressure, compare labels.
- Saturated fat. Full-fat cheese is energy-dense. The study had no information on dairy fat, so it can't tell you whether low-fat or full-fat is better; dietary guidance generally favours moderate portions of whichever you choose.
- Live cultures. Not all yoghurts contain them after processing; check for "live" or "active cultures" if that matters to you.
- Calories. Fermented dairy is nutritious but not free. Use the recipe nutrition calculator to see what a yoghurt bowl or cheesy dish adds up to.
Who May Want to Skip or Limit It
People with a milk allergy should avoid dairy altogether. Many people with lactose intolerance tolerate yoghurt and kefir, and aged cheeses, better than milk, because fermentation lowers lactose, but tolerance varies. People with kidney disease, those on sodium-restricted diets and those managing high LDL on medical advice should check portions of salty or high-saturated-fat cheeses with their clinician. And if you don't eat dairy, nothing here says you need to start: fish, eggs, legumes, tofu and nuts cover protein well.
The Bottom Line
This is a large, carefully adjusted observational analysis finding that people who eat some fermented dairy have lower mortality, with the strongest signal at moderate intakes and nothing significant above about 150 g a day. It's consistent with other cohorts, and also subject to healthy-user bias, crude dietary measurement and unknown dairy fat. It doesn't prove dairy extends life, and the J-shape shouldn't be read as a precise optimum.
For your plate, the sensible reading is gentle: if you enjoy yoghurt, kefir, skyr or cottage cheese, a daily portion is a convenient, protein-rich habit that fits the evidence. Choose plain versions, treat cheese as a flavour, and don't expect any single food to do the work of the whole diet.
Frequently Asked Questions
Does eating yoghurt and cheese lower your risk of dying?
In a US study of 37,366 adults, people who ate fermented dairy had about 18% lower adjusted all-cause mortality and 20% lower cardiovascular mortality. It's observational, so it shows association, not cause.
How much fermented dairy is best?
The study found lower mortality up to roughly 150 g a day, then no significant association at higher intakes. That figure mixes yoghurt and cheese, so it isn't a precise target. A daily portion of yoghurt, kefir or cottage cheese is a sensible reading.
What counts as fermented dairy?
In this study, cheese, yoghurt, sour cream and buttermilk. Kefir and skyr are also fermented dairy products, though not specifically listed in the summary.
Is cheese as healthy as yoghurt?
Meta-analyses suggest a smaller association for cheese than for yoghurt. Cheese is also higher in calories, sodium and saturated fat per serving, so portion size matters.
Does the J-shaped curve mean too much dairy is harmful?
Not necessarily. At 150 g/day or more, no statistically significant association was found, which means the benefit couldn't be distinguished from chance, not that harm was shown. Data are also sparser at high intakes.
Is this good protein for building muscle?
Yes. Yoghurt, skyr, kefir and cottage cheese provide high-quality, leucine-rich protein. Greek yoghurt and cottage cheese are particularly protein-dense per calorie.
Should I start eating dairy for my health?
Not on the strength of one observational study. If you don't eat dairy, you can meet protein needs with fish, eggs, legumes, tofu and nuts. If you already enjoy it, plain fermented options are a reasonable choice.
References & Sources
- Zhang, F. & Li, W. (2026). Fermented dairy consumption and all-cause and cardiovascular mortality in US adults: NHANES 1999–2018. Scientific Reports. doi:10.1038/s41598-026-71687-9; News-Medical (16 September 2026).
- Fermented dairy and mortality: summaries of the NHANES analysis (AgriMoon, September 2026).
- Yogurt, cheese and chocolate and mortality: meta-analysis of 50 cohort studies (2026). Frontiers in Nutrition; News-Medical (17 March 2026).
- High-fat and low-fat fermented milk and cheese intake, proteomic signatures, and risk of all-cause and cause-specific mortality (Swedish cohort analysis).
- MacrosMeasure (2026). Free Protein Intake Calculator.
Calculate your own numbers — free.
18 science-backed calculators covering everything in this article and more.
