How valid are self-reported height and weight? A comparison between CATI self-report and clinic measurements using a large cohort study

Objective: To examine the relationship between self-reported and clinical measurements for height and weight in adults aged 18 years and over and to determine the bias associated with using household telephone surveys. Method: A representative population sample of adults aged 18 years and over living in the north-west region of Adelaide (n=1,537) were recruited to the biomedical cohort study in 2002/03. A computer-assisted telephone interviewing (CATI) system was used to collect self-reported height and weight. Clinical measures were obtained when the cohort study participants attended a clinic for biomedical tests. Result: Adults over-estimated their height (by 1.4 cm) and under-estimated their weight (by 1.7 kg). Using the self-report figures the prevalence of overweight/obese was 56.0% but this prevalence estimate increased to 65.3% when clinical measurements were used. The discrepancy in self-reported height and weight is partly explained by 1) a rounding effect (rounding height and weight to the nearest 0 or 5) and 2) older persons (65+ years) considerably over-estimating their height. Conclusion: Self-report is important in monitoring overweight and obesity; however, it must be recognised that prevalence estimates obtained are likely to understate the problem. Implications: The public health focus on obesity is warranted, but self-report estimates, commonly used to highlight the obesity epidemic, are likely to be under-estimations. Self-report would be a more reliable measure if people did not round their measurements and if older persons more accurately knew their height.