How accurately are height, weight and leg length reported by the elderly, and how closely are they related to measurements recorded in childhood?

Background This paper examines (1) the accuracy of self-reported height, leg length and weight in a group of subjects aged 56-78; (2) whether recent measurement of height and weight influences the accuracy of self-reporting and (3) associations between childhood and adult height, leg length and BMI measured in old age. Methods All 3182 surviving members of the Boyd Orr cohort were sent postal questionnaires in 1997-1998 and a sub-sample (294) was also clinically examined. Results Self-reported height was overestimated and body mass index (BMI), based on reported height and weight, underestimated. The mean difference between self-report and measured values were for height: 2.1 cm in males and 1.7 cm in females; for BMI the difference was -1.3 kg/m(2) in males and -1.2 kg/m(2) in females. Shorter individuals and older subjects tended to over-report their height more than others. The overweight under-reported their weight to a greater extent. Recent measurement appeared to decease over-reporting of height but not weight. Correlations between self-report and measured height and BMI were generally over 0.90, but weaker for leg length (r = 0.70 in males and 0.71 in females). Adult height and leg length were quite closely related to their relative values in childhood (correlation coefficients ranged from 0.66 to 0.84), but associations between adult and childhood BMI were weak (r = 0.19 in males and 0.21 in females). Conclusions Self-reported measures of height and weight may be used in studies of the elderly although systematic reporting errors may bias effect estimates, As overweight individuals tend to under-report and the short and underweight tend to over-report, studies investigating associations of disease with height and weight using self-reported measures will underestimate effects. The weak associations between childhood and adult BMI indicate that associations between childhood adiposity and adult cardiovascular disease found in this cohort may reflect the specific effect of childhood overweight, rather than its persistence into adulthood. This suggests that avoidance of adiposity may be as important in childhood as in adulthood.