Portal de Periódicos da CAPES

Rethinking the Use of Body Mass Index z-Score in Children and Adolescents with Severe Obesity: Time to Kick It to the Curb?

2017; Elsevier BV; Volume: 188; Linguagem: Inglês

10.1016/j.jpeds.2017.05.003

1097-6833

Aaron S. Kelly, Stephen R. Daniels,

Childhood Cancer Survivors' Quality of Life

See related article, p 50 See related article, p 50 As the prevalence of pediatric severe obesity continues to increase,1Skinner A.C. Perrin E.M. Skelton J.A. Prevalence of obesity and severe obesity in US children, 1999-2014.Obesity (Silver Spring). 2016; 24: 1116-1123Crossref PubMed Scopus (269) Google Scholar it has become apparent that some of the “tried and true” body mass index (BMI) metrics many researchers and clinicians have become accustomed to using, such as BMI z-score, are proving less reliable in certain circumstances and in many cases can be misleading.2Woo J.G. Cole T.J. Assessing adiposity using BMI z-score in children with severe obesity.Obesity (Silver Spring). 2017; 25: 662Crossref PubMed Scopus (9) Google Scholar Two reports from Freedman et al, the most recent published in the current volume of The Journal,3Freedman D.S. Butte N.F. Taveras E.M. Goodman A.B. Ogden C.L. Blanck H.M. The limitations of transforming very high body mass indexes into z-scores among 8.7 million 2- to 4-year-old children.J Pediatr. 2017; 188 (50–6.e1)Abstract Full Text Full Text PDF Scopus (51) Google Scholar raise important concerns about the use of BMI z-score in studies that include children with severe obesity. Using data from the United States Centers for Disease Control and Prevention, Pediatric Nutrition Surveillance System, the authors analyzed BMI values from 8.7 million children ages 2-4 years old obtained between 2008 and 2011. They demonstrated that the theoretical maximum BMI z-score varied by more than 3-fold across ages and in children with severe obesity, BMI z-score was only moderately correlated with BMI percent of the 95th percentile (ie, the percentage above or below the absolute BMI cutpoint associated with the age- and sex-specific 95th BMI percentile) and delta BMI percent of the 95th percentile (ie, the difference in absolute BMI units of the child's BMI from the 95th percentile BMI). They also demonstrated that BMI z-score values could differ by more than 1 SD among children with similar body size, simply owing to differences in age or sex.3Freedman D.S. Butte N.F. Taveras E.M. Goodman A.B. Ogden C.L. Blanck H.M. The limitations of transforming very high body mass indexes into z-scores among 8.7 million 2- to 4-year-old children.J Pediatr. 2017; 188 (50–6.e1)Abstract Full Text Full Text PDF Scopus (51) Google Scholar These findings extend recent observations from this same group showing almost no association between BMI z-score and measures of body fatness in children and adolescents with severe obesity.4Freedman D.S. Butte N.F. Taveras E.M. Lundeen E.A. Blanck H.M. Goodman A.B. et al.BMI z-scores are a poor indicator of adiposity among 2- to 19-year-olds with very high BMIs, NHANES 1999-2000 to 2013-2014.Obesity (Silver Spring). 2017; 25: 739-746Crossref PubMed Scopus (148) Google Scholar Together, these studies draw the same conclusion using 2 distinct datasets, namely, that BMI z-score (generated from the Centers for Disease Control and Prevention growth charts) is a poor metric when applied to youth with severe obesity because values do not correlate with adiposity level and can be highly misleading, potentially causing erroneous conclusions to be drawn from research studies and even in the clinical setting. Freedman et al3Freedman D.S. Butte N.F. Taveras E.M. Goodman A.B. Ogden C.L. Blanck H.M. The limitations of transforming very high body mass indexes into z-scores among 8.7 million 2- to 4-year-old children.J Pediatr. 2017; 188 (50–6.e1)Abstract Full Text Full Text PDF Scopus (51) Google Scholar cite a specific example of how a weight management intervention implemented among 2-year-old girls with severe obesity might seem to be effective based on change in BMI z-score when in fact the intervention may have had no impact on the true BMI trajectory, simply owing to mathematical limitations regarding how BMI z-score is calculated among girls in this age group. The authors contend that the exact opposite conclusion could be drawn if the same intervention were to be evaluated in 2-year-old boys.3Freedman D.S. Butte N.F. Taveras E.M. Goodman A.B. Ogden C.L. Blanck H.M. The limitations of transforming very high body mass indexes into z-scores among 8.7 million 2- to 4-year-old children.J Pediatr. 2017; 188 (50–6.e1)Abstract Full Text Full Text PDF Scopus (51) Google Scholar In these types of scenarios, BMI z-score is applied essentially in a way in which it was never intended to be used and, accordingly, generates spurious results. As an aside, it should be noted that BMI z-scores derived from international cohorts, such as from Britain and The Netherlands, may be less susceptible to the problems identified with BMI z-scores using the Centers for Disease Control and Prevention growth chart data, and therefore could potentially have utility in the context of severe obesity.5Woo J.G. Using body mass index Z-score among severely obese adolescents: a cautionary note.Int J Pediatr Obes. 2009; 4: 405-410Crossref PubMed Scopus (58) Google Scholar Freedman et al3Freedman D.S. Butte N.F. Taveras E.M. Goodman A.B. Ogden C.L. Blanck H.M. The limitations of transforming very high body mass indexes into z-scores among 8.7 million 2- to 4-year-old children.J Pediatr. 2017; 188 (50–6.e1)Abstract Full Text Full Text PDF Scopus (51) Google Scholar, 4Freedman D.S. Butte N.F. Taveras E.M. Lundeen E.A. Blanck H.M. Goodman A.B. et al.BMI z-scores are a poor indicator of adiposity among 2- to 19-year-olds with very high BMIs, NHANES 1999-2000 to 2013-2014.Obesity (Silver Spring). 2017; 25: 739-746Crossref PubMed Scopus (148) Google Scholar and others6Flegal K.M. Wei R. Ogden C.L. Freedman D.S. Johnson C.L. Curtin L.R. Characterizing extreme values of body mass index-for-age by using the 2000 Centers for Disease Control and Prevention growth charts.Am J Clin Nutr. 2009; 90: 1314-1320Crossref PubMed Scopus (320) Google Scholar, 7Kelly A.S. Barlow S.E. Rao G. Inge T.H. Hayman L.L. Steinberger J. et al.Severe obesity in children and adolescents: identification, associated health risks, and treatment approaches: a scientific statement from the American Heart Association.Circulation. 2013; 128: 1689-1712Crossref PubMed Scopus (654) Google Scholar have suggested that alternative metrics, such as BMI percent of the 95th percentile and delta BMI percent of the 95th percentile, are much more flexible and robust in the context of severe obesity as compared with BMI z-score. Although these measures seem to be more appropriate from an arithmetic perspective and likely preferred to BMI z-score, further research is needed to evaluate their usefulness by characterizing the association of changes in these measures with changes in adiposity in the context of longitudinal studies and interventional trials, as well as in identifying what magnitude of reduction is necessary to achieve clinically meaningful improvements in obesity-related cardiometabolic risk factors and comorbidities. For example, in the adult realm, it is now widely accepted that 3%-5% body weight reduction is clinically meaningful,8Jensen M.D. Ryan D.H. Apovian C.M. Ard J.D. Comuzzie A.G. Donato K.A. et al.2013 AHA/ACC/TOS guideline for the management of overweight and obesity in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and The Obesity Society.Circulation. 2014; 129: S102-38Crossref PubMed Scopus (1738) Google Scholar whereas an equivalent threshold of weight loss (BMI reduction) has yet to be identified for children or adolescents, perhaps because so many different BMI metrics have been used with no consensus on which is the preferred measure(s). So, what are researchers and clinicians to make of this and how should they respond? Although BMI z-score remains useful for obesity classification and BMI tracking among youth ranging from normal weight to moderate obesity, a convincing case has now been made to move away from the use of BMI z-score as a primary weight loss outcome in research involving youth with severe obesity and as a metric of tracking change among individual patients with high BMI values in the clinical setting. Researchers should carefully select which BMI measures to incorporate into study protocols based on the degree of obesity of the target population9Kelly A.S. Fox C.K. Rudser K.D. Gross A.C. Ryder J.R. Pediatric obesity pharmacotherapy: current state of the field, review of the literature and clinical trial considerations.Int J Obes (Lond). 2016; 40: 1043-1050Crossref PubMed Scopus (50) Google Scholar and health systems should implement alternative ways to track BMI trajectories within the electronic medical record that more accurately reflect body size fluctuations in youth with high BMI.10Gulati A.K. Kaplan D.W. Daniels S.R. Clinical tracking of severely obese children: a new growth chart.Pediatrics. 2012; 130: 1136-1140Crossref PubMed Scopus (110) Google Scholar In this time of relative uncertainty as to which BMI measures are the most accurate and useful, researchers should err on the side of reporting more, not fewer, BMI-derived outcomes in their publications to offer readers more flexibility in how to interpret findings from individual studies and allow for better comparison of results across studies. Peer reviewers and journal editors should push back on authors who report BMI z-score as a primary outcome in studies involving participants with severe obesity and insist that alternative BMI metrics be presented in publications. In response to the increasing number of children and adolescents with severe obesity, a logical next step for the scientific and clinical community is to work toward consensus on preferred BMI-based measures to be used uniformly in research studies involving participants with high BMI values and for systematic implementation in the clinical setting for classification and tracking purposes. Flexible metrics that can be applied to the entire range of the BMI spectrum are desirable so that direct comparisons can be drawn across different studies involving individuals with widely varying BMI levels. Currently, the field is hamstrung by a “Tower of Babel” phenomenon in which too many different “BMI languages” are being spoken, some of which (eg, BMI z-score) should no longer be uttered in certain settings. The field needs to establish a common dialect, shared by the research and clinical community alike, in an effort to maximize efficiencies and accelerate progress for one of the most pressing medical and public health issues of our time. The Limitations of Transforming Very High Body Mass Indexes into z-Scores among 8.7 Million 2- to 4-Year-Old ChildrenThe Journal of PediatricsVol. 188PreviewTo examine the associations among several body mass index (BMI) metrics (z-scores, percent of the 95th percentile (%BMIp95) and BMI minus 95th percentile (ΔBMIp95) as calculated in the growth charts from the Centers for Disease Control and Prevention (CDC). It is known that the widely used BMI z-scores (BMIz) and percentiles calculated from the growth charts can differ substantially from those that directly observed in the data for BMIs above the 97th percentile (z = 1.88). Full-Text PDF