Low tierSystematic ReviewCitation verified
Meta-analysis of prospective cohort studies evaluating the association of saturated fat with cardiovascular disease
Patty W Siri-Tarino, Qi Sun, Frank B Hu, Ronald M Krauss - The American Journal of Clinical Nutrition, 2010
This pooled analysis of 21 prospective cohort studies found no significant association between dietary saturated fat intake and coronary heart disease, stroke, or total cardiovascular disease when comparing the highest and lowest intake groups. It is one of the most frequently skeptic-cited null results on saturated fat. The result must be read alongside its funding: the paper declared "no conflicts of interest" yet was supported by the National Dairy Council, with one co-author funded by a Unilever fellowship.
Key findings
- No significant association between saturated fat intake and CHD: pooled RR 1.07 (95% CI 0.96-1.19; P = 0.22) comparing extreme quantiles.
- No significant association with stroke: RR 0.81 (95% CI 0.62-1.05; P = 0.11).
- No significant association with total CVD: RR 1.00 (95% CI 0.89-1.11; P = 0.95).
- Authors concluded there is no significant evidence that dietary saturated fat is associated with increased CHD or CVD risk.
Effect measures
- Other: RR 1.0795% CI 0.96-1.19; P = 0.22
- Other: RR 0.8195% CI 0.62-1.05; P = 0.11
- Other: RR 1.0095% CI 0.89-1.11; P = 0.95
Why this evidence tier (Low)
- Risk of bias:
- Pools observational prospective cohorts only; subject to confounding (diet quality, lifestyle), measurement error in dietary self-report, and residual confounding that no cohort design can fully remove.
- Precision:
- Large pooled sample (347,747) gives reasonably tight intervals, but the confidence intervals for each endpoint cross 1.0, so a modest true effect cannot be excluded.
- Directness:
- Directly addresses the saturated-fat-to-CVD question, but tests association of intake levels, not the effect of replacing saturated fat with a specific nutrient, which is what intervention guidance turns on.
- Consistency:
- A frequently cited null that sits in tension with long-term replacement RCTs (e.g. the Hooper Cochrane review), so the body of evidence is not internally consistent.
- Funding / COI:
- The paper states verbatim that "no conflicts of interest were reported," yet it was supported by the National Dairy Council (authors PWS-T and RMK) and co-author QS was supported by a Unilever Corporate Research postdoctoral fellowship. Dairy and food-industry sponsorship of a skeptic-favourable null on saturated fat is a material conflict that the formal disclosure did not capture. Stated plainly: industry-funded despite a "no conflicts" declaration.
Low certainty. An observational null with confounding and an undisclosed industry-funding conflict; informative for the debate but weak ground for causal conclusions.
- Population:
- Pooled data from 21 prospective epidemiologic cohort studies; 347,747 subjects followed 5-23 years, of whom 11,006 developed coronary heart disease (CHD) or stroke.
- Conflicts of interest:
- The paper declared "no conflicts of interest were reported," but was funded by the National Dairy Council and one co-author held a Unilever postdoctoral fellowship - a material industry conflict for a saturated-fat null result that the formal disclosure did not record.
- Funding:
- National Dairy Council (PWS-T and RMK); co-author QS supported by a Unilever Corporate Research postdoctoral fellowship; FBH by NIH grant HL60712; additional NIH/NCRR support.
Limitations
- Observational cohorts only; associations are subject to confounding and cannot establish causation.
- Dietary intake measured by self-report (food frequency questionnaires), with attendant measurement error.
- Tests intake level rather than nutrient replacement, so it does not show what happens when saturated fat is swapped for unsaturated fat or carbohydrate.
- Funded by the National Dairy Council with a co-author Unilever fellowship despite a formal "no conflicts" declaration.