Re-evaluation of the traditional diet-heart hypothesis: analysis of recovered data from Minnesota Coronary Experiment (1968-73).
Level 1 - systematic review of randomized trials
Systematic review and meta-analysis of randomized controlled trials
PubMed 27071971 · doi:10.1136/bmj.i1246
What was done
Authors recovered and analyzed previously unpublished raw data and documents from the Minnesota Coronary Experiment (MCE, 1968-73), a double-blind randomized controlled trial conducted in one nursing home and six state mental hospitals in Minnesota. The study evaluated 9,423 men and women aged 20-97, with longitudinal serum cholesterol measured in 2,355 participants exposed to assigned diets for one year or more, alongside 149 completed autopsy files. The intervention replaced saturated fat with linoleic acid (corn oil and polyunsaturated margarine) versus a control diet high in saturated fat. Measured outcomes included all-cause mortality, associations between cholesterol changes and death, and autopsy-detected coronary atherosclerosis and myocardial infarcts. A systematic review and meta-analysis of randomized controlled trials (5 trials, n=10,808) substituting saturated fat with linoleic acid without confounding interventions was also conducted.
What was found
The intervention lowered serum cholesterol compared with controls (mean change from baseline -13.8% vs -1.0%; P<0.001). Kaplan-Meier analyses showed no mortality benefit for the intervention group in the randomized cohort or any prespecified subgroup. In covariate-adjusted Cox regression, each 30 mg/dL (0.78 mmol/L) reduction in serum cholesterol was associated with a 22% higher risk of death (hazard ratio 1.22, 95% confidence interval 1.14 to 1.32; P<0.001). Autopsy data showed no evidence of benefit for coronary atherosclerosis or myocardial infarcts. In the meta-analysis of 5 randomized controlled trials, linoleic acid interventions showed no significant reduction in coronary heart disease mortality (HR 1.13, 95% CI 0.83 to 1.54) or all-cause mortality (HR 1.07, 95% CI 0.90 to 1.27).
Why it matters
These findings challenge the classic diet-heart hypothesis by demonstrating that lowering serum cholesterol via linoleic acid substitution does not translate into reduced coronary or all-cause mortality. They highlight how selective non-publication of trial data historically led to overestimation of the clinical benefits of vegetable oils rich in linoleic acid.
Limits
The primary trial sample was restricted to institutionalized patients in mental hospitals and a nursing home, limiting generalizability. Longitudinal cholesterol follow-up of one year or more was available for only 2,355 of the 9,423 randomized participants, and autopsy data were limited to 149 cases. The analysis relies on historical records recovered decades after trial completion.
Cited by
- context In the Minnesota Coronary Experiment, for every 30-point drop in LDL cholesterol, the risk of heart attack increased by 22%.
- supports Approximately 75% of subjects dropped out of the Minnesota Coronary Experiment in the first year.