University of Minnesota

Ducimetiere on The French Paradox

Several European studies together provide evidence with which to address the more dramatic question today in French epidemiology: the so-called “French Paradox.” The term is given to an apparent anomaly of national death certification in Europe, where sharply lower age-specific death rates are ascribed to ischemic heart disease in France compared with many other Western European countries, the UK, and the U.S., in the presence of a clearly affluent lifestyle and risk factor levels (Ducimetiere 2004).

Pierre Ducimetiere, statistician-epidemiologist at INSERM in Paris, attributes the term to cardiologist Jean Richard’s 1981 article and to a 1987 article “Le paradoxe francais,” based on early results in a Paris study among policemen (Richard et al., 1981;1987). Ducimetiere now directs the French enterprise in CVD epidemiology and with his group is responsible for the Paris (Policemen’s) Prospective Study as well as for collaborations with Belgian and Irish studies and for the three French centers organized in the WHO MONICA Project: Lille, Strasbourg, and Toulouse.

Historically, Jean Lenegre, the grandest of the grand patrons of French cardiology post-World War II, was skeptical about reports of low coronary death rates in France.1. He attributed the apparent phenomenon to a diagnostic rigor that required certain evidence, particularly from necropsy and beyond the clinical picture, for certification of myocardial infarction, either as a clinical diagnosis or a cause at death certification. His opinion was forthright and clear: “Diagnoses in the case of cardiac deaths are, in France, either imaginary or erroneous!” (Lenegre 1958). [ed. To the contrary, the U.S. physician’s certification of a coronary cause on the death certificate is a commonplace and based on no systematic criteria.]

Lenegre’s young colleague, Jean Himbert, in early 1971, showed examples of many death certificates ascribed to “cardiosclerose,” a generic, wastebasket label for a cardiac death not meeting the French criteria for myocardial infarction, and not then coded to the ischemic rubrics of the International Vital Statistics of WHO. Thus, the combined opinions of these distinguished cardiologists in the early 1970s was that there was no French Paradox, only conservative French diagnostic criteria and diagnostic custom.

The Paris Prospective Study, nevertheless, found coronary incidence, by agreed criteria, one-third lower than group data from the U.S. Pooling Project of cohort studies. Moreover, MONICA, the multi-national CVD surveillance research coordinated by WHO beginning in the 1980s, produced relevant evidence that:

  1. Coronary death certification in France is systematically lower than that found in many European areas, after careful hospital documentation and standardized validation of reports;
  2. France has a distinct North-South gradient just as does all Europe, with higher coronary death rates in Lille than in Strasbourg and than in Toulouse;
  3. Rates in Lille resemble those in nearby Belgian Flanders; in Strasbourg they resemble those in neighboring Switzerland and Germany; while those in Toulouse in Provence resemble those in Spain and Italy (MONICA 2003).

French coronary death rates are, therefore, not atypical, according to Ducimetiere, but rather they are comparable to those among their contiguous countries at the same latitude. Moreover, the risk factor-disease associations are consistent with those found outside France.

Furthermore, Ducimetiere points to the still-lacking documentation that French lifestyle and diet are as luxurious as commonly thought. For example, few Frenchmen eat regularly in the gourmet restaurants of France. The daily household fare may, in fact, be considerably more varied, light, fresh, and less meat-and-dairy-oriented than customarily envisioned from afar (Ducimetiere 2004). In fact, the animal fat consumption in France is comparable to that in other European cohorts of the Seven Countries Study (Richard et al, 1987).

Thus, France’s lower-than-average coronary death rates appear appropriate to the regional cultures (and cuisines and risk characteristics) and to death rates in its contiguous geography. According to Ducimetiere: “There is now much evidence that the southern European diet and other lifestyle factors play a part and may modulate the effect of cholesterol and fat in the etiology of coronary heart disease. We conclude that the time has come to relieve epidemiology of the French paradox. Much more attention should be paid to collecting reliable data to produce more satisfactory explanations for the complex causes of heart disease” He summarizes: “that CHD rates are not so low in France, animal fat intake not so high, and the diet-heart concept not so unique that the existence of a ‘French paradox’ might be sustained any more, except for satisfying cultural fantasy or business marketing.” (Ducimetiere 2008) (Henry Blackburn)

Footnote 1.In post-war France, the dean of French cardiology, Professor Jean Lenegre, recognized early the advancing coronary epidemic. In the 1970s, he enlisted junior colleagues in implementation and created the political will and support for a CVD epidemiology and prevention enterprise in France. He did this at the center of the francophone medical universe, in Paris, which by tradition recognized little of merit in the provinces and bestowed its enlightenment grudgingly upon them.

Lenegre early commissioned young cardiologists Jean Richard and Jean Himbert to develop studies of CVD incidence and culture-specific risk. He found Daniel Schwartz, biometrician at the French “NIH,” INSERM, to develop the needed study design and methods. Schwartz, in turn, enlisted his young statistical colleague, Pierre Ducimetiere, to lead the planning, field work, and analysis.


Ducimetiere, P., in an interview conducted by Henry Blackburn and Kalevi Pyorala. October 2004, History of Cardiovascular Epidemiology, University of Minnesota.

Ducimetiere, P. 2008. Dialogues in Cardiovascular Medicine.

Lenegre, J. 1958. Heart diseases caused by coronary atherosclerosis, angina pectoris, myocardial infarct, cardiac insufficiency and various complications, sudden death. Revue du Praticien 8: 1717.

MONICA. (Ed. H Tunstall-Pedoe) 2003. World’s Largest Study of Heart Disease, Stoke, Risk Factors, and Population Trends 1979-2002.

Richard, J.L. Cambien, R., Ducimetiere, P. 1981. Particularites epidemiologiques de la maladie coronaire en France. Nouvelle Presse Medicale; 10: 1111-1114.

Richard JL. 1987. Les facteurs de risque coronarien. Le paradoxe français. Archives des Maladies du Cœur N° spécial Avril :17-21.