Does area of residence affect body size and shape? A Ellaway 1 , A Anderson 2 and S Macintyre 1 1 MRC Medical Sociology Unit, 6 Lilybank Gardens, Glasgow, G12 8RZ; and 2 Department of Human Nutrition, University of Glasgow, Royal In®rmary, Queen Elizabeth Building, Glasgow G31 2ER OBJECTIVE: To examine whether neighbourhood or residence is associated with body size and shape (height, weight, BMI, waist circumference and waist-hip ratio). DESIGN: Analysis of data collected in fact to face interviews at the second wave of longitudinal health survey of two adult age cohorts in the West of Scotland. SETTING: Four socially contrasting urban neighbourhoods in Glasgow City, Scotland. SUBJECTS: A total of 691 subjects: 142 males and 176 females aged 40 at interview; and 167 males and 206 females aged 60 at interview. All had been resident in their current neighbourhood for at least four years. MEASUREMENTS: height, weight, BMI, waist circumference and waist-hip ratio. RESULTS: Neighbourhood of residence was signi®cantly associated with height, BMI, waist circumference and waist- hip ratio after controlling for individual characteristics such as gender, age, social class, smoking behaviour and material deprivation (an index comprising income, housing tenure and car ownership). Individuals living in the most deprived neighbourhood were signi®cantly shorter, and had bigger waist circumferences, waist-hip ratios and BMIs. CONCLUSIONS: If Health of the Nation targets on reducing the proportion of overweight individuals in the population are to be met, public health policy should focus on places as well as people. Keywords: body size; area of residence; deprivation Introduction The problem of overweight in the UK is large and getting larger. In 1980, 39% of males and 32% of females in England had BMIs over 25 kg/m 2 ; by 1993 this had increased to 57% males and 48% females. 1,2 Similar problems are found in Scotland, it being estimated that, in the 40±59 age group, over half of Scottish men and 45% of women are overweight. 3 Targets have been set to reduce the proportion of overweight people in the population by the year 2005. 3,4 Overweight is linked with increased mortality and contributes to a wide range of conditions, including ischaemic heart disease, hypertension, stroke, certain cancers, and gall bladder diseases. Risk of disease grows with increasing BMI and is particularly marked at high BMI. 5 Overweight has been shown to be responsible for around 40% of the incidence of coronary disease in women 6 and to account for up to 95% of the incidence of non-insulin-dependent diabetes mellitus. 7 The health bene®ts from correcting overweight include a reduction in the likelihood of non-insulin-dependent diabetes mellitus and an increase in life expectancy in individuals who already have the disease. 8 In economic terms, a lowering of the rates of CVD, cancer and strokes would result in signi®cant reductions in the amount spent on drugs and social care required to manage these diseases and their effects. 3 Excess body weight results from a chronic excess of dietary energy intake compared to energy expen- diture. Therefore, two main causes predominateÐ extra energy intake (most commonly due to a diet high in fat) and low physical activity levels. In most cases, both causes co-exist. The prevalence of over- weight increases with age, 3 and varies by social class with fewer overweight people in higher social classes than in lower social classes. 9 There are also associa- tions between overweight and a wide range of life events including marriage, pregnancy and retirement. 5 The potential in¯uence of area of residence as a possible contribution to the development of over- weight has rarely been addressed (although a study in the West of Scotland found that waist hip ratio was independently associated with postcode sector of residence after controlling for a number of individual characteristics 10 ). Other health outcomes have been shown to vary by area of residence after adjustment for other factors such as smoking, social class and age (for example, blood pressure and respiratory symp- toms 11±14 and mortality rates 15 ). It has been shown that opportunities for physical recreation and eating according to current healthy diet guidelines vary between different types of areas, 16±20 that diet varies Correspondence: Ms A Ellaway. Received 9 September 1996; revised 24 December 1996; accepted 3 January 1997 International Journal of Obesity (1997) 21, 304±308 ß 1997 Stockton Press All rights reserved 0307±0565/97 $12.00