A New Weight-for Height Standard based on British Anthropometric Data

It was concluded in a previous paper (Billewicz, Kemsley, and Thomson, 1962) that the ratio of observed to standard weight is the simplest and most reliable index of adiposity, provided that the standard is suitably chosen. This ratio has obvious advantages: it is easy to calculate, the values ob tained do not depend on the units of measurement, and if the mathematical model presented in the previous paper gives a good approximation to reality the ratio should also yield results which are clinically and statistically consistent. The idea of using the ratio of observed to standard weight as an index to adiposity is, of course, not new. Indices of this type have been used in the past, though lately there has been a tendency to use instead various formula indices, notably the Ponderal Index. We suspect that the reluctance of research workers to use standard weights stems from the difficulty of finding suitable standards and from the idea that the use of standards implies the acceptance of standards as ideal or desirable weights. In this paper we present British weight-for-height standards, examine the properties of the ratio of observed to standard weight, and consider the applicability of this ratio to a variety of situations. The weight-for-height standard most commonly used in Britain is derived from the actuarial data collected in America towards the end of the 19th century (Medico-Actuarial Investigation, 1912). In the absence of generally-accepted standards based on British data, Davidson, Meiklejohn and Passmore (1959) recommended the continued use in Britain of this standard, in preference to more recent American actuarial data, on the grounds that it agrees more closely with recent British measurements. These actuarial standards include the weight of cloth ng and the height of footwear, but the appro priate allowances for them are not usually specified. T e data were derived from part of the American population, namely from those whose lives were propo ed for insurance. They were, therefore, sub ject to some measure of self-selection, although it is not k own whether or how far this has affected the s andards, for example, by producing a bias towards certain income groups. In the latest set of American standards (Metropolitan Life Insurance Company, 1959) ranges of desirable weights are given for ind viduals of a given height and aged 25 years and over. It is clear from the text that the desi able weights are those associated with lowest mortality experience, but no definition of large, medium, or small frame is given.