Am J C/in Nutr l991;53:879-85. Printed in USA. © 1991 American Society for Clinical Nutrition 879 Menstrual differences due to vegetarian and nonvegetarian diets13

ian premenopausal women whom we closely screened. The two groups were indistinguishable with respect to height, weight, body mass index, and menarche. The incidence of menstrual irreg-ularity was 4.9 % among nonvegetarians and 26.5 % among vegetarians (P = 0.009). The vegetarian group consumed sig-nificantly greater amounts of polyunsaturated fatty acids, car-bohydrates, vitamin B-6, and dietary fiber whereas the nonveg-etarians reported greater intakes ofsaturated fatty acids, protein, cholesterol, caffeine, and alcohol. Logistic-regression analyses showed that the probability of menstrual regularity among all study subjects (n = 75) was positively associated with increasing protein/Id and increasing cholesterol/kJ intakes. The probability of being menstrually regular was negatively associated with in-creasing dietary fiber/Id and increasing magnesium/Id intakes. These results are consistent with the notion that menstrual reg-ularity can be influenced by specific dietary nutrients that may have direct effects or exert their effects by modu-lating circulating sex steroid status. Am J C/in Nuir 199 l;53