RESUMO
Skinfolds at five anatomic sites were measured on 47 subjects in a study of gallbladder disease in Starr County, Texas, by two methods, one and two handed. A comparison was made of the two methods as to bias and precision. There were statistically significant but biologically unimportant differences (on the order of 1-2 mm) at four of the five sites. At three of these, the two-handed method gave larger values (lower thigh, triceps, and subscapular), and at the other site (medial calf) this method produced smaller values on average. Measurement precision was improved using the two-handed method at three of the five sites, significantly so for the triceps skinfold. Measuring skinfolds with two hands improves the precision of measurement for some skinfold sites in obese individuals. It does not introduce important and systematic biases across skinfold sites. Adoption of this technique should be used when circumstances outweigh the cost of involving an extra observer in the measurement of skinfold thickness, such as when precision is very important and when subjects are obese. © 1995 Wiley-Liss, Inc.
