LONGITUDINAL WEIGHT CHANGES OVER 4 YEARS AND ASSOCIATED HEALTH FACTORS IN 629 MEN AND WOMEN AGED OVER 65

Objectives: (1) To assess changes in body weight longitudinally over 4 years in a representative group of men and women aged over 65 years, living in their own homes. (2) To assess whether initial weight (or we ight in proportion to skeletal size) was associated with health 4 year s later. (3) To assess whether any changes found in body weight were a ssociated with changes in physical or mental health. (4) To compare th e cross-sectional age-related changes in weight found in the initial s urvey with the longitudinal changes found over 4 years. Design: A larg e randomised age-stratified survey using a structured questionnaire an d measurements of body weight and demispan (for skeletal size), with a 4-year follow-up. Setting: The survey was conducted in the respondent s' own homes. Subjects: 958 subjects age over 65 years, who were recru ited from the Nottingham general practitioners' lists, took part in th e initial survey; 629 of these subjects completed the second survey 4 years later. Results: (1) The mean 4-year change in body weight was a small but significant loss; in women (n = 385) it was 1.56 kg (P < 0.0 01, 95% CI 1.02-2.10) and in men (n = 244) 0.85 kg (P = 0.010, 95% CI 0.21-1.49). (2) Initial weight did not predict mortality, new morbidit y nor health 4 years later. (3) There were no robust associations betw een weight change and either absolute measures of physical health or c hanges in these measures. (4) The longitudinal change in weight was si milar to that predicted by the cross-sectional data. Conclusions: (1) Ageing, in old age, is associated with loss of body weight, but with a large intra-individual variation. (2) Neither initial body weight nor the change, was associated with mortality or morbidity over 4 years i n a large representative sample of old people living in their own home s in a food-rich country.