Purpose: We evaluated associations between excess body fat (%Fat) and various indices of obesity calculated from height and weight data.
Methods: In 147 adult males, %Fat was measured by the underwater-weighing method, and obesity indices were generated by the following 5 approaches: the Broca-Katsura (Katsura method), the Kato-Wataya (Kato method), Japan Society for the Study of Obesity (BMI method; based on the body weight at which the BMI is 22), and the Meiji Life Insurance Co. methods, and the Tables and Figures for Assessment of Obesity and Leanness published by the Ministry of Health and Welfare (MHW method).
Results: %Fat was 20% or more (obese) in 67 males (45.6%), 15-20% in 39 (26.5%), 10-15% in 35 (23.8%), and less than 10% in 6 (4.1%). The correlation coefficients between the obesity indices and %Fat were 0.612 for the Katsura method, 0.590 for the Kato-method, 0.611 for the BMI method, 0.612 for the Meiji Life Insurance Co. method, and 0.550 for the MHW method, being significant in each case (P < 0.01). When the cut-off point was set as 110% for each obesity index, sensitivity was highest with the Kato-method (82.1%), and specificity was highest with the Meiji Life Insurance Co. method (93.8%). With the MHW method, the receiver operating characteristic (ROC) curve was slightly farther from the point of sensitivity of 100% and 1-specificity of 0% than the others.
Conclusion: Excess fat accumulation can not be accurately assessed by obesity indices calculated from body height and weight data. Validity was similar among obesity indices examined.