 |
|
Efficacy of a Weight Management Program in Overweight Children—Reply
In Reply: We agree with Dr Young that the NNT can be useful for evaluating the effect of an intervention. The NNT is calculated as the reciprocal of the absolute risk reduction and describes the average number of participants who must be treated to prevent 1 occurrence of an event. The NNT is particularly effective when outcomes are discrete events.1 For continuous outcomes, such as those measured in our study, threshold cutoffs for "clinically significant" changes need to be defined. While the use of different thresholds will necessarily result in different NNTs, clinically significant thresholds for the pediatric population are not well established.
Despite these shortcomings, NNT is a useful additional measure of the utility of our intervention. At enrollment, all of our participants had a BMI greater than the 95th percentile for age, which was our definition of obesity. In the control group, 0 of 69 resolved obesity during follow-up vs 8 of 105 (7.6%) in the intervention group (P = .02 by Fisher exact test), resulting in an NNT of 13 (ie, 13 children needed to be treated with the weight management program for 1 participant to resolve obesity). Nevertheless, 75% of our population had BMIs above the 99th percentile, and clinically meaningful changes may have occurred without dropping below the 95th percentile. When the 97th percentile was used as a threshold, 0 of 66 in the control group resolved vs 15 of 99 (15.2%) in the weight management program (P < .001), for an NNT of 7.
While all children were overweight at study baseline, not all participants were insulin resistant; some participants developed insulin resistance during the study. The NNT thus can measure the number needed to resolve insulin resistance or the number needed to prevent insulin resistance, depending on the resistance status of the participants at baseline. Although there is no standard definition of insulin resistance, Lee et al2 have suggested that a homeostasis model assessment of insulin resistance of greater than 4.39 be used. Using this cutoff, we observed that 21 of 59 participants (35.6%) with insulin resistance at baseline resolved in the intervention group compared with only 7 of 48 (14.6%) in the control group (P = .01); the NNT to resolve insulin resistance was 4.8. Similarly, 6 of 46 (13.0%) without insulin resistance at baseline developed resistance in the intervention group compared with 6 of 21 (28.6%) in the control group (P = .17); the NNT for prevention of insulin resistance was 6.4.
Unlike adults, children do not have an ideal percentage body fat. While mean change in both percentage body fat and estimated body fat mass in our study were notable for the intervention group, there is no threshold to determine NNT for such outcome measures.
Financial Disclosures: None reported.
Mary Savoye, RD, CD-N, CDE
mary.savoye{at}yale.edu
James Dziura, PhD
Yale Center for Clinical Investigation
Yale University School of Medicine
New Haven, Connecticut
1. Altman DG, Schulz KF, Moher D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med. 2001;134(8):663-694.
FREE FULL TEXT
2. Lee JM, Okumura MJ, Davis MM, Herman WH, Gurney JG. Prevalence and determinants of insulin resistance among U.S. adolescents: a population-based study. Diabetes Care. 2006;29(11):2427-2432.
FREE FULL TEXT
Letters Section Editor: Robert M. Golub, MD, Senior Editor.
JAMA. 2007;298:1860-1861.
 CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati
What's this?
RELATED LETTER
Efficacy of a Weight Management Program in Overweight Children
Paul C. Young
JAMA. 2007;298(16):1860.
EXTRACT
| FULL TEXT
|
