University of Minnesota

National Diet Heart Study

Type Diet/Drug (Stage): Diet (1º) (Pilot)
Study Category: The Prevention Trials (1946-1973)
Year Begun: 1962
Location: Baltimore, Boston, Chicago, Minneapolis-St. Paul, and Oakland, USA
Principal Investigator(s): National Diet Heart Group
External Resource: View URL


In 1960 an executive committee on diet and heart disease concluded that a mass field trial of diet fat and cholesterol in the diet to reduce first attacks of coronary disease in middle-aged American men would require a population of up to 100,000 with a follow-up of 4 to 5 years to demonstrate a 20% change in rate. An enterprise of this size caused the experts to recommend a feasibility study and awards were made to six investigators in 1962 to study diet in a double-blind design using fat-modified foods in mental institutions and in another design with free-living general populations, including one on food bought in the ordinary fashion.


The experience and thoughtfulness of the committee and the investigators is evident in the numerous aspects of a mass field trial the feasibility study was designed to assess. These are listed because of their importance historically in the development of intervention studies in cardiovascular disease.

The National Diet-Heart Pilot Trial was designed to determine the following:

  1. Effectiveness of the investigator group and volunteers.
  2. The volunteer rate.
  3. The effectiveness of dietary modifications to reduce saturated fat, increase poly’s and reduce dietary cholesterol.
  4. The availability, tolerability, distribution and costs of specially prepared commissary foods.
  5. Indices of adherence to diet.
  6. The extent of reductions in defined lipoprotein levels.
  7. The dropout, non-adherence, and non-response rates.
  8. Feasibility of the double-blind design.
  9. Staff equipment and facilities needed.
  10. Center variations from a common protocol.
  11. Experience in pooling data.

Equally interesting were the questions added to an extended study decided on during the first year of the study to examine further the following issues:

  1. What proportion of participants for a pilot study would volunteer for an extended study?
  2. Could costs be reduced by buying foods on the open market?
  3. Could participants be instructed to select and trim meats and would they get adequate blood lipid responses?
  4. Could diet modification be made less demanding for participants and investigators?

Free-living populations were studied in five clinics: Baltimore, Boston, Chicago, Minneapolis-St. Paul, and Oakland and in a closed population of Faribault State Hospital in Minnesota. There was an experienced central laboratory and coordinating center. 225 married men ages 45-54 in good health were recruited to each center and it proved feasible to recruit men in large numbers for such a research undertaking.


Many innovations in trial design and operation developed from this intensive national effort. For example, men were checked for credit rating to ensure that they could afford the foods they had to pay for. A faint-of-heart period was designed in which people were put on experimental diets with a run-in period prior to randomization to help ensure loyal adherence. Information was obtained on the proportion of exclusions by various criteria, including overweight and electrocardiographic abnormalities. Exclusions due to too many meals eaten out and other measures were tqken to identify and eliminate irresolute and unsuitable volunteers prior to the crucial step of randomization. The statistical center developed approaches to stratified random assignments in 16 different strata, based on entry risk characteristics.

The study demonstrated the feasibility of obtaining consistent high-quality serum cholesterol analyses by use of a central laboratory. It demonstrated quantitative measurement of change in fatty acid compensation of red cells and adipose tissue and diets were analyzed quantitatively by a central laboratory. Much was learned about food center management and nutritional instruction and motivation and repeat 24-hour food records. It was determined that specially prepared fat modified foods could be satisfactorily procured and distributed and were acceptable, palatable, and indistinguishable between the diets compared, but that economies of scale would be necessary to make such food production self supporting. Pooling data was possible because of the insignificant differences between centers based on the common protocol and manual of operations.

Conclusions and Recommendations:

The investigative team, based on the national diet-heart pilot trial, recommended that large-scale studies on prevention of premature coronary heart disease by dietary means:

  1. Should be planned and put into operation as soon as possible because of the epidemic nature of coronary disease and the demonstration that serum cholesterol could be safely lowered by modification of the usual American diet in ways that are acceptable to large numbers of people and that volunteers can be readily recruited into such a program and because the small studies reported to date “do not settle the scientific and public health question”.
  2. A committee should be formed without delay to develop practical plans and protocol for such a large-scale long-term 5-year controlled experiment in the free-living American population, with a recommendation that it should be undertaken only with a non-double blind design, maintaining, however, the random allocation of participants and the objective monitoring of confounding variables and the objective blind assessment of endpoint events.

A significant alternative approach suggested is the first formal proposal we find of a multiple risk factor trial, which we consider historically worth quoting here: “as an alternative approach to a design involving only diet intervention in the experimental group, consideration should be given to recommendations to the experimental group not only with respect to diet for correction of hypercholesterolemia, but also with regard to controllable obesity, hypertension, cigarette smoking, and physical activity. This study would then involve testing the more general hypothesis that application of current knowledge concerning coronary risk factors can reduce incidence of the disease. And in the resolution of this matter, it should be kept in mind that a positive result in a two-group study on diet would be a major achievement even if it could not rigidly and unequivocally be inferred that the hypocholesterolemic diet alone – as distinct from the overall diet program – was responsible for the result.”

(To our knowledge, this is the first real suggestion of a multiple-risk factor trial and appeared in the March 1968 proposal of the Diet Heart Study Group.) (HB)


“The National Diet-Heart Study Final Report.” Circulation, 1968; 37(3 suppl): I1-I26.

Report of the Diet-Heart Review Panel of the National Heart Institute. Mass Field Trials and the Diet-Heart Question: Their Significance, Timeliness, Feasibility and Applicability. Dallas, Tex: American Heart Association; 1969, AHA Monograph no. 28.