Intended for healthcare professionals

CCBYNC Open access

Re-evaluation of the traditional diet-heart hypothesis: analysis of recovered data from Minnesota Coronary Experiment (1968-73)

BMJ 2016; 353 doi: (Published 12 April 2016) Cite this as: BMJ 2016;353:i1246

Re: Re-evaluation of the traditional diet-heart hypothesis: analysis of recovered data from Minnesota Coronary Experiment (1968-73)

Ramsden et al’s intriguing findings from this famous experiment raises a fundamental question in cardiovascular disease prevention and treatment. The institutionally-imposed change in dietary fat quality produced lowering of serum cholesterol, but paradoxically the greater the reduction in serum cholesterol the greater was the excess in all-cause mortality. The authors recognise the challenge to explain this paradox. To state the obvious, statins are a triumph of allopathic medicine and cholesterol-lowering is the mode of action.(1) Could it be then that controlling one’s blood lipids using polyunsaturated oils as part of a healthy diet increases rather than reduces coronary risk?

An analysis in the Whitehall II cohort identified dietary patterns specifically linked to serum total and HDL-cholesterol and serum triglycerides, and then related these diets to future risk of myocardial infarction or coronary death.(2) The dietary patterns linked to an adverse blood lipid profile were associated with increased coronary risk over the 15 years of follow-up, after controlling for demographic and behavioural variables. Reported average intake of polyunsaturated margarine differed by a factor of almost 3 across quartiles of dietary pattern scores (healthy versus unhealthy: 14.4 and 5.4 g/d). Intakes of other sources of polyunsaturated fats likely varied in a similar manner. So, the paradoxical findings in the Minnesota Coronary Experiment (MCE) in the 1960s were not replicated in Whitehall II around the millennium.

We would like to suggest there may have been a ‘period effect’ at work in the MCE. The table of characteristics of the MCE cohort does not report the prevalence of smoking, but it is likely to have been almost universal at the time of the study. Those who smoked the most, and those who were sickest as a result of their habit, are likely to have experienced the largest decline in cholesterol. This mechanism is a form of reverse causation, and it might explain the paradox in question.

(1) Baigent C, Keech A, Kearney PM et al. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet 2005 October 8;366(9493):1267-78.
(2) McNaughton SA, Mishra GD, Brunner EJ. Food patterns associated with blood lipids are predictive of coronary heart disease: the Whitehall II study. Br J Nutr 2009 August;102(4):619-24.

Competing interests: No competing interests

21 April 2016
Eric Brunner
Professor of epidemiology
Professor Gita Mishra
University College London
1-19 Torrington Place