Variation in height and weight in the Norwegian population.

Since 1962, as an international collaborative project, studies have been in progress of disease among migrants and native populations in Great Britain, Norway, and the United States. The background for the first studies and data on various demographic characteristics, residential history, smoking history, and cardiorespiratory symptoms for migrants and native-born in the United States have been published elsewhere (Reid, 1966; Reid et al., 1966). For the men in the sample of Norwegian-born residents of the United States and in the study groups in Norway additional data have been collected on dietary habits. The items of information that one attempted to obtain for all individuals under study included height and weight. Reasons for considering height and weight were the known and suspected association of body size, as indexed by height, weight, or a function of the two, with morbidity and mortality from a number of diseases and the variation of body size with living conditions and habits of the individuals. The main objective of this paper is to present the pattern of variation in height, weight, and weight adjusted for height in a cross-section of the adult Norwegian population. For this purpose only the data pertaining to individuals in a probability sample of the general population will be considered. The index weight divided by the square of height (W/H2), sometimes called the Davenport index, here referred to as the bulk index, was chosen as a simple measure of weight adjusted for height. This is in accordance with Wedervang (B0e, Humerfelt, and Wedervang, 1957) who, in his analysis of height and weight measurements in the adult population of the city of Bergen, Norway, found that (W/H2) in both males and females was more nearly uncorrelated with height than was either (W/H) or (W/H8). Several recent reports have dealt with the choice of indices of bulk, or obesity, derived from height and weight (Khosla and Lowe, 1967; Evans and Prior, 1969; Florey, 1970; Benn, 1971). Florey (1970), who did not control for age, concluded that, based on the criterion that the index for weight adjusted for height should be uncorrelated with height, (W/H) might be better than (W/H2) in western female populations. Consequently, the variation of weight (W/H) and (W/H2), with height within sex and age in our sample will be examined.