Abstract

PIP: Knowledge is needed about what women generally experience (behavior and beliefs about sexual, personal, social, and dietary factors) during menstruation as baseline data. Data were obtained from a WHO non nationally representative sample of 5322 parous women from 14 cultural groups between 1973 to 1980 and a subsample of 500 women with detailed daily diaries from 10 countries (Egypt, India, Indonesia, jamaica, Korea, Mexico, Pakistan, Philippines, UK, and Yugoslavia). This study identified and analyzed 1) critical variables through principal component analysis and varimax rotation, 2) clusters of types of women with particular menstrual experiences and particular beliefs, and 3) the relationship between demographic variables and the 2 resultant cluster and the relationship between beliefs and experiences. Factor analysis resulted in the identification of 6 factors and 13 variables which accounted for more than 65% of the variance: amount of bleeding, activity during the last menstrual period, mood during last menstrual period, discomfort during last menstrual period, predictability, and blood characteristics (smell and color). Beliefs that explained more than 53% of the estimated variance were bathing behavior during menstruation, fertility and femininity issues, interpretations and implications of menstruation, and beliefs about not washing hair or body during menstruation. Alpha estimates of reliability for the belief ranged from .33 to .73, and for behavior the range was .40 to .59. The cluster analysis of type of persons identified 3 groups of women: type I (26.4%) who experienced low blood loss the first day and had a long duration of bleeding; type II (52.3%) who had a short duration of bleeding; and type III (21.2%) who had the heaviest bleeding and longest duration of bleeding. The cluster analysis of women's belief types indicated 9 profiles. For instance, Type 2 women tended to believe that menstruation is dirty but prefer more blood loss. Belief Profiles 3, 4, 5, 8, and 9 are consistently represented and profiles 1, 2, 6, and 7 show wide variability. 7 demographic factors were found to be significantly related to belief clusters: country of residence, religion, literacy, age, work environment, social status, and rural vs. urban area. There was evidence, for instance, that type 2 women were overrepresented in the use of modern methods, and that Belief Profile 1, which represents 10.23% of the sample, showed 17.86% using modern methods. The findings show that Belief Profile 1 persons who disagreed with many items were the most likely to use modern methods and have smaller family sizes. Beliefs, which reflect socialization according to demographic variables, appear to affect choice of methods and family size.