Childhood Obesity An Optimistic Review, Part 1: Diagnosis
This first article of a 3-part series lays the framework for an optimistic approach to the growing problem of childhood obesity. It offers a critical look at how obesity is defined and practical advice about assessing BMI percentile, approaching the history and physical examination, and ordering the most appropriate diagnostic tests.
The second part of the series outlines childhood obesity treatment strategies in 6 areas using the mnemonic, “Find And Shape More Complete Management”—food, activity, stress, monitoring, comorbidities, and medication. Successful intervention depends on partnering with children and families to design a plan in each of these areas.
Part 3 counters common myths about pediatric obesity with facts. Biologic factors such as genetics and hormones, physical factors such as sedentary lifestyles and unhealthy diets, and social factors such as the media and marketing influence the prevalence of and perceptions about childhood obesity.
Why another article on obesity?
On a typical day, I review a chart before entering the examination room. The patient is a 7-year-old girl who is here for a growth evaluation. In addition to seeing children in my diabetes clinics, I see them for other endocrine concerns such as short stature, failure to thrive, precocious puberty, and congenital adrenal hyperplasia with rapid growth. As I enter the room, I immediately recognize that this appointment is not likely to involve these typical endocrine issues. The child sitting on the exam table clearly is obese.
Childhood obesity is a growing problem. Glance into any waiting room, and you are likely to see a child or a parent for whom chairs with armrests are a physics problem, not just furniture.
In June 2013, the American Medical Association declared obesity a disease in the hopes that defining it as such, instead of as a lifestyle or behavioral choice, will help change approaches to research on, policy about, and treatment of obesity and will prompt health insurers to reimburse providers for treating patients with it.1
So, why write another article on childhood obesity?
My objective is to go beyond a recitation of the facts and to provide an optimistic framework of information about the approach to pediatric obesity, without stigmatizing the children and families who have placed trust in our care and recommendations.
Scope of the Problem
Based on current data from the Third National Health and Nutrition Examination Survey (NHANES), the prevalence of obesity in U.S. children and adolescents aged 2 to 19 years is 17%, or 1 in 6 children.2 Prevalence varies geographically, across income quartiles, and among racial and ethnic groups. Peak prevalence is in the Southern states, in low-income families, and in African American, Mexican American and American Indian ethnic groups.3
It is essential to note that these statistics may not correlate with any individual medical professional’s personal experience, and research shows that clinicians give more weight to personal experience than to statistics. A closer look at the NHANES data shows that the peaks in prevalence are just the highlights of the overall data. There is no region of the country, income group, or ethnic group whose members are not affected by childhood obesity. Put another way, if a child is obese, his or her risk is 100%, not 17%. When taking care of an individual child, mistakes can be made if conclusions are drawn based on assumed risk and not on individual characteristics.
A clear definition of obesity offers a place to start the process of diagnosis, treatment, and discussion of health and disease status with children and their families. Currently accepted definitions of childhood overweight and childhood obesity are based on body mass index (BMI) percentile for age and sex as described in the 2000 growth charts from the Centers for Disease Control and Prevention (CDC) (Table 1).4-7
In the pediatric and adult medical literature, BMI and BMI percentile are criticized for failing to identify body component (muscle versus fat). However, alternative methods of body fat measurement (eg, waist circumference, waist-to-hip ratio, skin-fold thickness, dual-energy X-ray absorptiometry) have not achieved the acceptance that BMI percentile has. These alternatives either are harder to reproduce or cost more time and money than does calculating BMI percentile.
If BMI percentile is flawed, why use it? Why not rely on the clinical observation skills instilled in every medical provider? In a study of the ability of medical students, family and pediatric residents, and community physicians to identify, by photo alone, the BMI category of 3 preschool-aged children, only 15% correctly identified the heaviest child as having an obese BMI percentile; 21% correctly identified the child with an overweight BMI percentile; and 86% correctly identified the child with a normal BMI percentile.8 Students’ and physicians’ level of education and field of training had no bearing on the accuracy of identification.
Because health care providers as a group appear not to be good at visually identifying obesity, assessing BMI percentile is the important first step in my optimistic framework. That said, while accurately determining BMI percentile is necessary, it is not sufficient — it can categorize a child as overweight or obese, but it cannot reveal the child’s health or disease state.9 The necessary focus is on early identification and prevention or treatment of ill health and disease, not remaking children into a particular size and shape.
Ill health can be defined as any condition that limits function and enjoyment. Ill health in an obese child may include a lack of stamina on the playground, pain or musculoskeletal effects from poorly fitting shoes, maladaptive interpersonal/social relationships with peers, or intermittent depression. Further concerns include obesity-related diseases: prediabetes and diabetes mellitus; hypertension; hyperlipidemia; polycystic ovary syndrome; nonalcoholic fatty liver disease and cholelithiasis; asthma and obstructive sleep apnea (OSA); and tibia vara and slipped capital femoral epiphysis.
Assessing BMI Percentile
The use of electronic health records has made BMI percentile assessment simultaneously easier and harder. It is easier if the data are collected and entered correctly but harder when data-entry errors occur—entering weight in pounds but height in centimeters, for example, will result in most children being classified as obese.
Before age 2 years, BMI is not assessed; weight for length is preferred. Once qualification for measurement of standing height (not recumbent length) and the age of 2 years have both been achieved, BMI percentile should be considered another vital sign, assessed along with pulse, respiration, blood pressure, temperature, and pain level.
Obtaining a Meaningful History
Well-framed, open-ended questions that require detailed answers, as opposed to yes-or-no questions, provide a more meaningful assessment and often suggest a plan of action in management. With new patients, it can be helpful to get a sense of their daily dietary routine. Ask broad questions, such as the following (despite the notorious inaccuracy of patients’ dietary recall):
• How many meals and snacks a day? Four? Six? Too many to count?
• Any recent changes in diet?
• What do you drink?
• How many hours a day are no-food hours (either formally declared with a closed kitchen, or informal/incidentally)?
• What’s your favorite fruit and vegetable? (A child who cannot name a favorite probably isn’t getting any.)
• Who serves portions?
• What happens if food is left on the plate?
I generally leave the detailed diet history to a dietitian, or I send families home with a 3-day intake form, with blanks to complete for brand name and measured portion sizes.
Questions about activity and sedentary time also are important:
• What activities do you enjoy? (Regularly scheduled and just-for-fun activities count.)
• Do you tire out before other kids?
• How much screen time do you have in a day? A week?
• How long before bedtime does screen time stop?
• How many hours do you usually sleep?
• Do you need sleep-in mornings on weekends? (Poor sleeping is associated with increased weight gain.)
Ask about good and ill health. A family history of obesity-related disease and the age of onset can guide screening.
• What are some of the things that are hard for you to do?
• Tell me about your friends.
• What do you do at recess or after school?
• What size clothes fit you?
• How often does your clothing size change?
The review of systems should include any reported snoring, daytime napping, skin changes and infections, musculoskeletal problems, and nocturia with or without enuresis, and it should include screening for depression and for teasing or bullying.
Careful Observation During the Exam
Before beginning a physical examination, I ask myself to intentionally consider whether the child appears to have any of the syndromes associated with obesity. Look at the child sitting, standing, and moving. Notice how much jiggle there is. Adipose jiggles; muscle doesn’t. This “jiggle-assessment” has become as routine a part of my observational examination as skin tone, hair pattern, eye contact, and posture. My examination is split between looking for signs of the causes of obesity (eg, papilledema, goiter, central adiposity with peripheral muscle wasting, low muscle tone or volume, stigmata of syndromes) and the effects of obesity (eg, acanthosis nigricans, hepatomegaly, leg bowing, xanthomas, xanthelasmas).
Thoughtful Use of Diagnostic Tests
Screening and specific laboratory tests should be based on the assessment of a patient’s risk and evidence-based standards of care. This is easier said than done. We practice in a don’t-want-to-miss-something climate, and the pressure is high to check for one more possibility. Table 2 lists my set of reasonable routine diagnostic tests for all obese children, and extra tests for obese children with specific concerns.
Fasting plasma glucose. Screen every 2 years for diabetes in overweight or obese children aged 10 years or older (younger if puberty has begun) with at least 2 risk factors (ethnicity, family history, signs of insulin resistance). The normal value is 99 mg/dL or lower; 100 to 125 mg/dL is considered impaired, and a level of 126 mg/dL or above is diagnostic for diabetes, if repeated on a second day or corroborated with other abnormal glucose levels.10
Fasting lipid panel. Screen all children once between ages 9 and 11 years and again between ages 17 and 21 years.11 Table 3 lists pediatric reference ranges for fasting lipid panel test results. Screen between ages 2 and 8 and ages 12 and 17 only if the child’s family history is positive (eg, a parent with dyslipidemia or early myocardial infarction or angina) or if the child has a high-risk condition (eg, diabetes mellitus, hypertension, obesity, end-stage renal disease, cardiac transplant, history of Kawasaki disease with current aneurysms), a moderate-risk condition (eg, Kawasaki disease with regressing aneurysms, chronic inflammatory disease, HIV, nephrotic disease), or smokes tobacco.11
Liver screening. Transaminase measurement is useful to detect subclinical nonalcoholic fatty liver disease. Levels of alanine aminotransferase and aspartate aminotransferase that are double the laboratory upper limit in the absence of acute infection should trigger further evaluation.
Laboratory OSA assessment. Detection of increased bicarbonate on an electrolyte panel can indicate respiratory acidosis that is metabolically compensated. Keep in mind, however, that OSA in its earlier and more treatable stages may exist in the presence of a normal level of bicarbonate. In this case, a thorough history can be elucidating.
Thyroid screening. Assess antithyroid antibodies (antithyroglobulin antibodies and antithyroid peroxidase antibodies) if there is a family history of Hashimoto thyroiditis or Graves disease. Measurement of thyroid stimulating hormone (TSH) with free thyroxine is reasonable if a child has high cholesterol or a goiter. Routine TSH screening is not indicated, because a slightly elevated level in an obese child does not signify a disease process or identify a cause of obesity.12
Cortisol screening. This is necessary only if there is severe linear growth retardation (which is virtually always present in Cushing disease or Cushing syndrome) with increasing BMI percentile. A useful screening tool is an 11 pm to midnight salivary cortisol level. The normal level during this time interval is less than 0.09 µg/dL.13
Puberty assessment. Obesity is associated with earlier (or earlier for family pattern) pubertal development. In girls younger than 8 and boys younger than 9 years of age, assessment should begin with a radiographic bone-age assessment. Bone age often is advanced 2 to 3 years or more in obesity, and puberty timing is associated more closely with biologic age than with chronologic age. Further laboratory testing (including assessment of luteinizing hormone, follicle-stimulating hormone, estradiol, testosterone, 17-hydroxyprogesterone, and dehydroepiandrosterone sulfate) should occur in coordination with a pediatric endocrinologist whenever possible, since many laboratories do not have pediatric and sexual-maturity rated normal ranges with which to evaluate the results.
1.Moyer CS. Delegates declare obesity a disease. American Medical News. July 8, 2013:1-2,4.
2.Ogzden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of Obesity in the United States, 2009–2010. NCHS Data Brief. 2012;(82):1-8.
3.Centers for Disease Control and Prevention (CDC). Obesity among low-income preschool children. http://www.cdc.gov/obesity/downloads/PedNSSFactSheet.pdf. Accessed October 21, 2013.
4.Centers for Disease Control and Prevention (CDC). 2 to 20 years: boys: body mass index-for-age percentiles. http://www.cdc.gov/growthcharts/data/set1clinical/cj41l023.pdf. Published May 30, 2000. Updated October 21, 2000. Accessed October 3, 2013.
5.Centers for Disease Control and Prevention (CDC). 2 to 20 years: girls: body mass index-for-age percentiles. http://www.cdc.gov/growthcharts/data/set1clinical/cj41l024.pdf. Published May 30, 2000. Updated October 21, 2000. Accessed October 3, 2013.
6.Centers for Disease Control and Prevention (CDC). Basics about childhood obesity. http://www.cdc.gov/obesity/childhood/basics.html. Updated April 27, 2012. Accessed October 21, 2013.
7.Pan L, Blanck HM, Sherry B, Dalenius K, Grummer-Strawn LM. Trends in the prevalence of extreme obesity among US preschool-aged children living in low-income families, 1998-2010. JAMA. 2012;308(24):2563-2565.
8.Ahlers-Schmidt CR, Kroeker D, Chesser A, Hart T, Brannon J. Visual recognition of child body mass index by medical students, resident physicians, and community physicians. Kansas J Med. 2010;3(5):7-14. http://archie.kumc.edu/handle/2271/892. Accessed October 21, 2013.
9.Katz DL. Childhood obesity trends in 2013: mind, matter, and message. Child Obes. 2013;9(1):1-2.
10.American Diabetes Association. Consensus statement: type 2 diabetes in children and adolescents. Diabetes Care. 2000;23(3):381-389.
11.Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: summary report. Pediatrics. 2011;128(suppl 5):S213-256.
12.Dekelbab BH, Abou Ouf HA, Jain I. Prevalence of elevated thyroid-stimulating hormone levels in obese children and adolescents. Endocr Pract. 2010;16(2):187-190.
13.Beck NL, Morales AE, Buchmann RF, Birusingh RJ. Obese 11-month-old female with round facies. Pediatr Rev. 2013;34(4):185-191.
Dr Moltz is director of pediatric diabetes services at Children’s Hospital of Michigan and an assistant professor of pediatrics at Wayne State University School of Medicine in Detroit.