Title of article:

Obesity screening in older women with the body mass index: a receiver operating characteristic (ROC) analysis

Authors: Sardinha LB, Teixeira PJ.
Journal: Science and Sports, Jul 2000; 15: 212-9

Abstract

Purposes. - To assess the usefulness of body mass index (BMI) for obesity screening in women over 50 years of age.
Materials and Methods. - The sample included 383 Caucasian women (60.5 ± 7.1 years; 64.9 ± 10.9 kg; 154.2 ± 6.2 cm). Percent body fat (%BF) was assessed by dual-energy x-ray absorptiometry and BMI (kg.m-2) was derived from weight (kg) and height (m2). Nonparametric receiver operating characteristic (ROC) analysis was used to assess the performance of the widely used international classification of obesity of BMI 30 kg.m-2 to correctly identify obese women with 35%BF estimated by DXA. ROC analysis was also used to identify the BMI value with the best tradeoff between sensitivity and specificity.
Results. - For the BMI 30 kg.m-2 a very low sensitivity (26.8) was found, indicating a high rate of false-negative cases (73.2), and the highest possible rate for specificity (100.0), with all nonobese women identified as nonobese by BMI. The best tradeoff between sensitivity and specificity was 25.5 kg.m-2 with a sensitivity of 74.4 and a specificity of 91.9.
Conclusion. - These results suggest that in older women, a BMI cutoff value of 30 kg.m-2 may not be adequate to screen for obesity, given the high rate of false negative cases.

Comments and Key points

This article thinks that a BMI of 30.0 is too high, for defining "obese" in older women over 60 years old. It thinks that a BMI of 25.5 is a better obesity threshold. It defines "true obesity" as having a body fat percentage over 35%. That's a common definition, but it is arbitrary nevertheless.

The article showed a graph of the BMI and body fat percentages of the older women in the study. Notice that the great majority of these women have a body fat percentage over 35%.

Is it natural and normal, or is it problem that the majority of older women have a body fat percentage over 35%? No-one knows for sure.

For younger and middle-aged adults, a body fat percentage over 35% is definitely associated with higher risks of disease, but studies of older adults hasn't been able to prove a significant extra risk.

In fact, older adults have a higher mortality rate² if they have a lower body mass index.

A different study¹, of adult women of all ages, showed that a BMI of 25.85 is the least-risky in terms of diseases related to being overweight or underweight.

So maybe it's not such a good idea afterall, to label the majority of older woman as obese, if their BMI is over 25.5, when might actually be reasonably healthy after all.

References

1. Allison D, Faith M, Heo M. Townsend-Butterworth D, Williamsom D. Meta-analysis of the effect of excluding early deaths on the estimated relationship between body mass index and mortality. Obes Res 1999; 7: 342-54.
2. Allison DB, Zhu SK, Plankey M, FaithMS, Heo M. Differential associations of body mass index and adiposity with all-cause mortality among men in the first and second National Health and Nutrition Examination Surveys (NHANES I and NHANES II) follow-up studies. Int J Obes Relat Metab Disord, Mar 2002;26(3):410-6

Review & comments by Steven B. Halls, MD, Modified on: 23-June, 2008, Copyright.
See other BMI studies.