Obesity is associated with many severe and potentially life-threatening conditions, such as: type 2 diabetes; atherosclerotic cardiovascular disease; metabolic syndrome (insulin resistance-hypertension-dyslipidemia); polycystic ovary syndrome; and nonalcoholic fatty liver disease, to name a few. Although obesity per se is not the sole cause of these conditions,1 their association with excess weight is useful because it alerts the busy clinician that an overweight individual is at risk for this group of illnesses.
In fact, obesity often precedes other pathologic changes in these conditions and, thus, could be thought of as the “fever” of metabolic dysfunction.
In order to utilize excess weight as a tool for identifying individuals at risk for metabolic dysfunction in clinical practice, it would be helpful to quantify it in a manner that makes sense to both the practitioner and the patient. The body mass index (BMI), denoted as kg/m2, is widely used as a simple screening method that requires only two easily obtained measurements: the height (converted to meters [m] then squared [m2]); and weight (converted to kilograms [kg]). Practically, the BMI can be rapidly calculated with a computer or portable hand-held “BMI wheel” (using either metric or English units). Moreover, it can be obtained in both the doctor’s office as well as in schools2 and is useful in correlating the level of obesity with potential adverse outcomes.3 However, the problem with BMI in children is that normative values are sex- and age-dependent, precluding the use of absolute cut-off numbers for overweight and obesity, such as those that are used in adults (25 kg/m2 and 30 kg/m2, respectively).
In order to deal with this obstacle, the concept of “percentile for age” is used almost universally in the pediatric setting. This works well for research purposes, but it has many limitations in the clinical setting.
First, the statistical concept of “percentiles” is difficult for pediatric patients and their families to comprehend. After all, it does not make sense to many individuals how 32% of children and adolescents in the Unites States can currently be at or above the 85th percentile of the currently used growth charts (the 2000 sex- and age-specific Centers for Disease Control and Prevention [CDC] growth charts).4 Second, parents frequently do not identify their own children as overweight or obese,5 regardless of what “percentile” their child is in. Third, a large percentage of pediatricians do not use electronic medical records and, thus, do not get an instant read-out of each patient’s BMI percentile,6 making it difficult during a busy clinic visit to quickly tell a family the “normal” BMI for their child’s age.
The intent of the percentile system was to establish norms for BMI at various ages. However, to circumvent the widespread misunderstanding of what “percentiles” really mean to the lay public, we propose that the concept of “percentiles” be dropped in clinical practice in place of a system that is readily understood by practitioner and patient and/or family alike. But how can new “norms” be established that approximate the percentile norms that are in common use?
As it turns out, the currently used 2000 CDC growth chart curves for the 85th BMI-for-age percentiles (the cut-off point for overweight) for both boys and girls closely follow a fairly linear pattern between the ages of 7 and 15 years, which corresponds to the patient’s age plus 10 (age + 10). For practical purposes, the adult “85th percentile” norm of 25 kg/m2 can be used to denote overweight in adolescents older than 15 years of age (15-year-old + 10 = 25); before 7 years of age, the value of 17 kg/m2 (7-year-old + 10 = 17) can be used. Although there are minor deviations from the traditional CDC-defined 85th BMI-for-age percentiles using the “age + 10” calculation for those patients aged 7 to 15 years, they are neither adverse nor detract from the clinical utility of quickly approximating (without access to a BMI percentile chart) the “overweight” BMI cut-off for a patient in the busy clinical setting.
To avoid the shortcomings in clinical practice with using the traditional “percentile” system, we propose a new method to report BMI – a two-number notation in which the raw BMI score is compared with the upper limit of normal for age (corresponding to approximately the 85th BMI-for-age percentile). Using this system, the BMI would be written similar to a blood pressure measurement, with the raw BMI score written first (the top number) and the upper limit of normal written second (the bottom number). For example, a 9-year-old child (boy or girl) with a raw BMI score of 21 kg/m2 would have a BMI of 21/19 (pronounced “twenty-one over nineteen”), with the bottom value of 19 derived using the “age + 10” method of estimating the 85th BMI-for-age percentile (described above). Both the physician and parent (and child!) would easily be able to tell that 21 is a bit above the upper-normal BMI value of 19 and, thus, understand that the child is overweight. This method is quick, easy, and effective.
Are there children with a normal BMI who are at risk for or who already have metabolic dysfunction? Absolutely, and this is a drawback of using BMI as the sole screening tool for metabolic abnormalities.7 Primary care physicians, thus, need to examine other factors associated with metabolic dysfunction (such as family history, a sedentary lifestyle, and a diet that includes sugar-sweetened beverages).8 Are there children with an abnormal BMI (above the upper-normal for age) who do not have metabolic dysfunction? Yes, and although the prevalence of metabolic abnormalities is higher among obese children than normal-weight children,9 not every obese child has metabolic syndrome or one of its associated conditions. However, given that weight gain during childhood and adolescence influences a child’s risk of becoming overweight or obese later in life,10 quickly identifying a child as overweight in the context of a busy clinical practice using the two-number system — and effectively conveying that information to the patient and/or family in an easily understood manner — is paramount. First, it would provide easy-to-understand objective data supporting the provider’s recommendation for a healthier diet and lifestyle (a treatment with no potential adverse effects); second, it may potentially increase patient/family adherence to weight management strategies.
The BMI is easily calculated — in both the inpatient and outpatient setting — using a patient’s height and weight. In children, the norms are constantly changing with age, but the upper limit of normal follows a fairly linear pattern between the ages of 7 and 15 years in both sexes, which approximately equals the child’s age + 10. Younger than the age of 7 years, the upper-normal BMI is approximately 17 kg/m2; older than the age of 15 years, the upper-normal BMI remains at the adult level of 25 kg/m2. We propose a simple way to report BMI in pediatric patients that includes both the raw BMI number as well as an approximation for the upper norm for age, giving the pediatrician and the patient and/or family a better perspective on what the BMI should be for any given age. Thus, we suggest that the BMI — at least in the clinical setting — be reported as two numbers, similar to a blood pressure reading, with the raw calculation as the top number and the approximated upper norm for age as the bottom number.
- Sheehan MT, Jensen MD. Metabolic complications of obesity. Pathophysiologic considerations. Med Clin North Am. 2000;84(2):363–385. doi:10.1016/S0025-7125(05)70226-1 [CrossRef]
- Nihiser AJ, Lee SM, Wechsler H, et al. BMI measurement in schools. Pediatrics. 2009;124(Suppl 1):S89–S97. doi:10.1542/peds.2008-3586L [CrossRef]
- Kirk S, Zeller M, Claytor R, Santangelo M, Khoury PR, Daniels SR. The relationship of health outcomes to improvement in BMI in children and adolescents. Obes Res. 2005;13(5):876–882. doi:10.1038/oby.2005.101 [CrossRef]
- Ogden CL, Carroll MD, Flegal KM. High body mass index for age among US children and adolescents, 2003–2006. JAMA. 2008;299(20):2401–2405. doi:10.1001/jama.299.20.2401 [CrossRef]
- Eckstein KC, Mikhail LM, Ariza AJ, Thomson JS, Millard SC, Binns HJ. Parents’ perceptions of their child’s weight and health. Pediatrics. 2006;117(3):681–690. doi:10.1542/peds.2005-0910 [CrossRef]
- DesRoches CM, Campbell EG, Rao SR, et al. Electronic health records in ambulatory care--a national survey of physicians. N Engl J Med. 2008;359(1):50–60. doi:10.1056/NEJMsa0802005 [CrossRef]
- St-Onge MP, Janssen I, Heymsfield SB. Metabolic syndrome in normal-weight Americans: new definition of the metabolically obese, normal-weight individual. Diabetes Care. 2004;27(9):2222–2228. doi:10.2337/diacare.27.9.2222 [CrossRef]
- Bremer AA, Auinger P, Byrd RS. Relationship between insulin resistance-associated metabolic parameters and anthropometric measurements with sugar-sweetened beverage intake and physical activity levels in US adolescents: findings from the 1999–2004 National Health and Nutrition Examination Survey. Arch Pediatr Adolesc Med. 2009;163(4):328–335. doi:10.1001/archpediatrics.2009.21 [CrossRef]
- Weiss R, Dziura J, Burgert TS, et al. Obesity and the metabolic syndrome in children and adolescents. N Engl J Med. 2004;350(23):2362–2374. doi:10.1056/NEJMoa031049 [CrossRef]
- Dietz WH. Health consequences of obesity in youth: childhood predictors of adult disease. Pediatrics. 1998;101(3 Pt 2):518–525.