1 . Anthropometric indices are presented for representative samples of elderly people in South Wales, based on over 1500 subjects seen during community surveys.2. Body mass index declined with age after 70 years in both men and women. Estimates of fat and muscle volume based on upper arm measurements also showed a decline with age, which was particularly steep for triceps skinfold thickness in women.3. These indices are in general similar to results that have been reported from other surveys within the UK; they suggest that Welsh old people have less fat and muscle than elderly Americans.Anthropometric percentiles have long been used in the diagnosis and surveillance of malnutrition in children. During recent years various anthropometric indices have been published for adults, including results from large surveys in the USA (Bishop et al. 1981; Frisancho, 198 1). In general, the information available for the elderly has been rather sparse, especially for persons aged over 75 years. Information has been collected from over 1500 old people in South Wales during community surveys, and these findings are presented here in the form of percentiles to provide a frame of reference for elderly people in Britain. SUBJECTS AND METHODSThe surveys were conducted in three areas of South Wales, two of which have an industrial history of coal mining. Names and addresses of the subjects were obtained from the lists of local general practitioners. In the largest of the three surveys, the sample was stratified by age to provide more subjects over 75 years of age. This survey also included residents of old people's homes (4%) and long-stay hospital wards (3%) whereas, in the other two areas, only people living at home were included. The values of the sampling ratios and other details of the surveys have already been published (Burr et al. 1974(Burr et al. , 1975.The heights and weights of the subjects were measured in indoor clothes without shoes. In the largest of the three surveys, the triceps skinfold thickness was measured with Harpenden callipers over the left triceps muscle midway between the acromion and the olecranon process; three measurements were made and the median reading taken. The mid-arm circumference was measured twice at the same point and the mean taken as the true value. The arm measurements were all made by the same observer (M. L. B.).Body mass index (Quetelet index) was calculated as weight (kg) divided by the square of the height (m". The arm muscle area (mm2) and arm muscle circumference (mm) were calculated using the following formulas derived from Gurney & Jelliffe (1973) :arm muscle circumference = arm circumference -7r (triceps skinfold thickness) (arm muscle circumference)2 4n arm muscle area = These formulas assume that the upper arm is cylindrical, and they ignore the contribution to arm volume made by the humerus. Nevertheless, the indices derived from them seem to be serviceable and are widely used.

Associations between smoking habit, social class, body mass index, and diet were examined in 493 men aged 45 to 59 yr, selected from the general population and who had completed a 7-day weighed dietary record. Smokers were lighter than nonsmokers and had a lower body mass index. There was no difference in energy intake, but in general, smokers had lower intakes of vitamins, minerals, and dietary fiber. Exsmokers had similar intakes to nonsmokers. Manual workers tended to be shorter, had a higher body mass index, higher intakes of energy and carbohydrates, and lower intakes of vitamins and minerals than nonmanual workers. Social class had a greater effect than smoking habit on intakes of energy and carbohydrates, whereas smoking habit had the greater effect on intakes of minerals and vitamins. Body mass index was associated negatively with sucrose intake and positively with protein intake, smoking habit, and social class being less important determinants.

THE Welsh Heart Programme was established in 1985 as a national demonstration project to promote health, and to reduce the risks of cardiovascular dis ease among the population of Wales. The develop ment of a youth health project has been an important element, and to provide information for planning, monitoring and evaluation, two surveys were under taken in 1986. The first provided information on children's health behaviour, knowledge and attitudes. The second, reported here, investigated curriculum development for health education, the school environment, and family and community links. From a random sample of 81 secondary schools, 75 responded to the survey by means of a self-com pletion questionnaire. More than three-quarters reported having a planned health education pro gramme, but the results suggest that many pupils are either receiving disjointed programmes, or no health education at all. Most of the schools with planned programmes reported having a designated member of staff as health education co-ordinator, most of whom had received in-service training. The results suggest that a reasonable base exists for the development of health promotion strategies aimed at young people, although there is clear scope for further development. Priority issues are attention to the timing of health education programmes, teaching methods, and continued in-service support for teachers. The wider development of supporting school policies, as well as closer and more active links with families and local communities are also advocated.

SUMMARY Surveys were conducted in four areas in Wales with differing degrees of environmental lead. In two areas the source of the lead was traffic and in one it was spoil from lead mining in the past. The fourth area, which served as a control, was a village remote from heavy traffic, industry, and lead mining. Various environmental samples were taken, and children aged 1-3 years and their mothers were studied. Blood lead concentrations were raised in the lead mining area, and within the areas defined by traffic flow the blood lead concentrations of the mothers showed a gradient. Pica in the children, assessed by a questionnaire, showed no relation with blood lead, but the amount of lead removed from the children's hands with 'wet wipes' was an important contributor to blood lead concentrations.

Concern about the levels of lead in blood is widespread. There is uncertainty, however, about the relative importance of the various environmental sources. Lead in petrol is widely assumed to be one of the most important sources and air and dust have been identified as the main routes to man. Water is regarded as an important source in areas with a plumbosolvent water supply, but of little or no importance in other areas. In order to evaluate the contribution to blood lead by various environmental sources, we have conducted surveys of random samples of women in areas of Wales chosen to represent very different levels of exposure to traffic. We report here that lead in air makes a small, but significant, contribution to blood lead but there is no evidence of any contribution from dust. Although in none of the areas were high levels of lead detected in water, water emerges as an important contributor to blood lead.