¹Institut National Agronomique Paris-Grignon, Laboratoire de Biologie et Nutrition Humaine, 16 rue Claude Bernard 75005 Paris, France

²Clinique du poids, 12 rue de Ponthieu, 75008 Paris, France

³Institut National de la Recherche Agronomique, Direction Scientifique NHSA, 147 rue de l'Université 75007 Paris, France

Correspondence to: J Ch Leblanc, Institut National Agronomique Paris-Grignon, Laboratoire de Biologie et Nutrition Humaine, 16 rue Claude Bernard 75005 Paris, France.

Contributors: JChL was responsible for coordination of the project and produced the final draft. HY was responsible for statistical analysis and contributed to writing the draft. AK was responsible for conversion of foods units into weight and PhV is the head of the laboratory. All four contributors are guarantors for the integrity of the article as a whole. Reprints: JCh Leblanc, Institut National Agronomique Paris-Grignnon, Laboratoire de Biologie et Nutrition Humaine, 16 rue Claude Bernard 75005 Paris, France.

Objective: To assess food behaviour and determine nutritional intakes of various vegetarian populations in France.

Design: A five-day self-administered food record which was mailed to members of the three principal French vegetarian organisations.

Subjects: 145 subjects, aged 7-87 y; 94 classical vegetarians (19% of those contacted), 34 Hindu lactovegetarians (17% of those contacted) and 17 macrobiotic (34% of those contacted).

Setting: The survey was conducted between March 1997 and July 1997 in France.

Results: Vegetarianism in France is represented by three main classes of food behaviour: ovolactovegetarian (AAV), lactovegetarian (KRI) and macrobiotic (MMK). The geometric mean intakes ranged from 1952 kcal/d (KRI), 2051 kcal/d (MMK) to 2384 kcal/d (AAV) for males and from 1302 kcal/d (MMK), 1675 kcal/d (AAV) to 1804 kcal/d (KRI) for females, after adjusting for age and BMI. The energy consumption in the MMK group was significantly lower than that in the AAV (P<0.05) and KRI groups (P<0.01), respectively. A difference among groups was observed for females (P=0.0002), but not for males. The MMK group consumed less lipid than the other two vegetarian groups, 46 g/d for men and 38 g/d for women vs 80 g/d for men and 61 g/d for women in the AAV group and 93 g/d for men and 81 g/d for women in the KRI group, respectively. Differences with AAV and MMK were statistically significant (P<0.001 for men and women for both groups). Mean protein consumption ranged from 60 g/d (AAV), 64 g/d (KRI) to 77 g/d (MMK) for males and from 46 g/d (MMK), 50 g/d (AAV) to 58 g/d (KRI) for females. Mean carbohydrate intakes ranged from 247 g/d (AAV), 321 g/d (KRI) to 338 g/d (MMK) in males and from 209 g/d (MMK), 228 g/d (AAV) to 242 g/d (KRI) in females. There were no significant differences in protein and carbohydrate intakes between the groups. Median calcium intakes ranged from 758.2 mg/d (MMK), 863 mg/d (AAV) to 989.3 mg/d (KRI) for the men and from 500.8 mg/d (MMK), 863 mg/d (AAV) to 934 mg/d (KRI) for the women. In the men, there was no differences in daily calcium intakes between the three vegetarian groups. However, we found a significant difference for women (P=0.0041). The women in the MMK group presented significantly lower daily calcium intakes than the women in the AAV (P=0.013) and KRI (P=0.0032) groups. The AAV and KRI groups consumed dairy products supplying respectively 36% and 53% for the men and 39% and 59% for the women of total calcium against 0% for men and women in the MMK group. Median iron intakes ranged between 12.5 mg/d (KRI), 13.2 mg/d (AAV) and 22.5 mg/d (MMK) for the men and between 11.2 mg/d (KRI), 14.6 mg/d (AAV) and 16.9 mg/d (MMK) for the women. MMK (men P=0.0172 and women P=0.0131) and AAV (only in men P=0.037) groups consumed significantly higher quantities of iron than did the KRI group. The heme iron median intake in males and females of the three vegetarian groups was very low (<0.5%). Overall, the female vegetarians consumed 58.1 (MMK), 109 (AAV) and 127.4 (KRI) mg of vitamin C per day and the males 76.3 (MMK), 150.4 (AAV) and 150.4 (KRI) mg per day. Median vitamin B9 intakes ranged from 247.5 µg/d (KRI), 312 µg/d (MMK) to 390.4 µg/d (AAV) for the men and from 188.3 µg/d (MMK), 266.9 µg/d (KRI) to 323.8 µg/d (AAV) for the women. Vitamin B12 consumption ranged from 0.2 µg/d (MMK), 1.5 µg/d (AAV) to 1.7 µg/d (KRI) in the women and from 0.6 µg/d (MMK) to 1.0 µg/d (AAV and KRI) in the men. No differences in consumption were observed in the males. On the other hand, the women in the MMK group consumed significantly less vitamin C and B12 than did the women in the AAV (P=0.0006) and KRI (P=0.0396) groups, while it was at the limit of significance for the females (P=0.0715) for vitamin B9.

Conclusion: Our results suggest that vegetarians have a better understanding of dietary requirements than does the general population. We observed that the more restrictive the vegetarian diet, the more likely were there to be deficiencies in minerals and vitamins, especially heme iron, diary calcium and vitamin B12 intakes. This was a particularly true for macrobiotic diets and especially for women, which represents a group at particular risk for mineral and vitamin deficiency.

Sponsorship: The study was sponsored by a non-profit organisation (Vegetarian association, Center of Nutritional Information Research, Center of Meat Information) and was coordinated by the Center of Foch Research.

European Journal of Clinical Nutrition (2000) 54, 443-449

vegetarian; food behaviour; nutritional intakes; 5-day diet records