Dairy linked to fewer deaths and strokes in 21 countries

Overview

Dehghan M et al.. Lancet 2018 · PMID 30217460 · DOI 10.1016/S0140-6736(18)31812-9

More than 2 servings/day: 17% lower mortality, 34% lower stroke risk than no dairy.

Does the study support the specific claim? No · Methodological quality: moderate

Original publication

Plain-language summary

Scientific in-depth review

1. Full reference

Authors

Dehghan M et al.

Journal

Lancet

Year

2018

Identifier

PMID 30217460 DOI 10.1016/S0140-6736(18)31812-9

Verification

verified – PubMed efetch 2026-10-26

Retracted

no

Corrections

none

Extent checked

abstract only

Conflicts of interest and funding

Funding sources listed in full paper (not read); record lists Medical Research Council among grants.

2. Research question

Are total and specific dairy foods associated with mortality and major CVD?

3. Study design and population

Study design

Cohort study – Multinational prospective cohort; country-specific validated FFQ; Cox frailty models with centre random intercepts

Population

Adults 35-70 y in 21 countries (high-, middle- and low-income)

Country

21 countries

Period

2003-2018; median 9.1 y

Sample size

136,384

Inclusion and exclusion criteria

PURE participants with dietary data

4. Methods and endpoints

Exposure / intervention

Total dairy, milk, yogurt, cheese, butter; whole- vs low-fat

Comparator

No intake

Endpoints

Composite of death or major CVD; total, CV and non-CV mortality; MI; stroke

Statistical methods

Multivariable Cox frailty models

5. Actual results

Results

Total dairy >2 servings/day v none: composite HR 0.84 (0.75-0.94); total mortality 0.83 (0.72-0.96); non-CV mortality 0.86 (0.72-1.02); CV mortality 0.77 (0.58-1.01); major CVD 0.78 (0.67-0.90); stroke 0.66 (0.53-0.82); MI 0.89 (0.71-1.11). Milk >1 serving v none: composite 0.90 (0.82-0.99); yogurt 0.86 (0.75-0.99); cheese 0.88 (0.76-1.02). Butter intake low, not associated (1.09, 0.90-1.33). 10,567 composite events.

Absolute risks

not reported

Authors' conclusion

Dairy consumption was associated with lower risk of mortality and major cardiovascular disease events in a diverse multinational cohort.

6. Strengths

• Very large, multinational incl. low- and middle-income countries
• Standardised outcomes

7. Weaknesses

• Single baseline FFQ
• Residual confounding (dairy intake marks affluence in poorer countries)
• Funding/COI not checked in full text

8. Risk of bias

Orienting tool

ROBINS-E

Type of assessment

simplified assessment (abstract only)

Full formal assessment

no

Overall risk of bias

some concerns

Rationale

Confounding by socioeconomic status possible.

9. Assessment of the statistical interpretation

Relative reductions of about 15-20%; observational; CIs log-symmetric (stats/C-DRY-misc_ci_checks.py).

10. Relevance to the specific claim

In a multinational cohort including low- and middle-income countries, higher dairy, milk and yogurt intake was associated with lower, not higher, mortality and CVD.

11. Independent contrary findings

A-DRY-01 (Swedish women, very high milk intake) found higher mortality.

12. Transferability to Germany and Europe

Broad international population; more transferable to Egypt than Scandinavian cohorts.

13. Overall methodological judgement

Large, diverse cohort contradicting a general mortality harm of dairy; confounding cannot be excluded.

14. Open questions

Whether associations are causal.

15. Linked fact checks

• High milk consumption raises mortality (women drinking 3+ glasses/day had 93% higher mortality) and increases rather than prevents hip fractures. – Studies contradicting or limiting the claim
• Dairy products harm health and cause inflammation. – Studies contradicting or limiting the claim