Meta-analysisindependent cross-checkabstract checked I-FAT-104 standard review
Cohorts did not link saturated fat to heart disease; trans fat was linked
Intake of saturated and trans unsaturated fatty acids and risk of all cause mortality, cardiovascular disease, and type 2 diabetes: systematic review and meta-analysis of observational studies.
Overview
de Souza RJ et al.. BMJ 2015 · PMID 26268692 · DOI 10.1136/bmj.h3978
SFA and CHD RR 1.06 (0.95-1.17); trans fat and mortality RR 1.34.
Does the study support the specific claim? Partly · Methodological quality: high
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Summarise saturated and trans fat studies.
- How was the study done?
Meta-analysis of cohorts.
- How many people, animals or samples were studied?
Up to 339,090 people.
- What was actually found?
No clear SFA link; trans fat harmful.
- How large was the observed effect?
SFA CHD RR 1.06.
- What do the results mean in everyday life?
The main proven danger is trans fat; SFA effects depend on what replaces it.
- What are the study's limitations?
Very low certainty; replacement not modelled.
- What does it say about Elawady's specific claim?
Partly supports the claim.
- Do the results apply to Germany and Europe?
International.
- What can explicitly not be concluded from this study?
Cannot show ghee is safe for arteries.
Scientific in-depth review
1. Full reference
- Authors
- de Souza RJ et al.
- Journal
- BMJ
- Year
- 2015
- Identifier
- PMID 26268692 DOI 10.1136/bmj.h3978
- Verification
- verified – PubMed efetch 2026-10-26
- Retracted
- no
- Corrections
- none
- Extent checked
- abstract only
- Conflicts of interest and funding
Canadian Institutes of Health Research (record).
2. Research question
Are saturated and trans fats associated with mortality, CVD, CHD, stroke, T2D?
3. Study design and population
- Study design
- Meta-analysis – Systematic review and meta-analysis of prospective cohorts with GRADE
- Population
Adults in cohort studies
- Country
Multinational
- Period
to May 2015
- Sample size
SFA: 3-12 cohorts per outcome; 90,501-339,090 participants
- Inclusion and exclusion criteria
Observational studies
4. Methods and endpoints
- Exposure / intervention
Saturated fat; trans fat (industrial/ruminant)
- Comparator
Lowest intake
- Endpoints
All-cause mortality, CVD, CHD, stroke, T2D
- Statistical methods
Pooled multivariable RR
5. Actual results
- Results
Saturated fat: all-cause mortality RR 0.99 (0.91-1.09); CVD mortality 0.97 (0.84-1.12); total CHD 1.06 (0.95-1.17); ischaemic stroke 1.02 (0.90-1.15); T2D 0.95 (0.88-1.03); CHD mortality 1.15 (0.97-1.36). Certainty for saturated fat 'very low'. Trans fat: all-cause mortality 1.34 (1.16-1.56); CHD mortality 1.28 (1.09-1.50); industrial trans fat and CHD 1.42 (1.05-1.92).
- Absolute risks
not reported
- Authors' conclusion
Saturated fats are not associated with all-cause mortality, CVD, CHD, stroke or T2D, but the evidence is heterogeneous with methodological limitations; trans fats are associated with mortality and CHD.
6. Strengths
- GRADE
- Separates industrial and ruminant trans fat
7. Weaknesses
- Observational
- Does not model replacement nutrient (replacement by refined carbohydrate may mask harm)
- Very low certainty
8. Risk of bias
- Orienting tool
- AMSTAR 2
- Type of assessment
- simplified assessment (abstract only)
- Full formal assessment
- no
- Overall risk of bias
- low
- Rationale
High-quality review of low-quality evidence.
9. Assessment of the statistical interpretation
Null associations with very low certainty; not equivalent to proof of no effect.
10. Relevance to the specific claim
The strongest observational support for the view that saturated-fat-rich foods are not clearly linked to heart disease; does not address replacement effects shown in RCTs.
11. Independent contrary findings
I-FAT-01 (RCTs of SFA reduction), I-FAT-02 (LDL causal).
12. Transferability to Germany and Europe
International.
13. Overall methodological judgement
Honest supporting evidence of very low certainty; outweighed by RCT and genetic evidence on LDL.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | high |
| Risk of bias | low |
| Precision of results | moderate |
| Directness | indirect |
| Reproducibility | unknown |
| Consistency with the overall evidence | mixed |
| Regional transferability | moderate |
| Adequacy of the interpretation (by the archive) | not applicable |
14. Open questions
Food-specific effects.