To provide updated prevalence data on obesity trends among US children and adolescents aged 2 to 19 years from a nationally representative sample.
We used the NHANES for years 1999 to 2016. Weight status was determined by using measured height and weight from the physical examination component of the NHANES to calculate age- and sex-specific BMI. We report the prevalence estimates of overweight and obesity (class I, class II, and class III) by 2-year NHANES cycles and compared cycles by using adjusted Wald tests and linear trends by using ordinary least squares regression.
White and Asian American children have significantly lower rates of obesity than African American children, Hispanic children, or children of other races. We report a positive linear trend for all definitions of overweight and obesity among children 2–19 years old, most prominently among adolescents. Children aged 2 to 5 years showed a sharp increase in obesity prevalence from 2015 to 2016 compared with the previous cycle.
Despite previous reports that obesity in children and adolescents has remained stable or decreased in recent years, we found no evidence of a decline in obesity prevalence at any age. In contrast, we report a significant increase in severe obesity among children aged 2 to 5 years since the 2013–2014 cycle, a trend that continued upward for many subgroups.
Comments
RE: Different reference points, different conclusions. Time for a common reference point for severe childhood obesity studies
With great interest, we read the article by Skinner and colleagues1 who analyzed current data from the National Health and Nutrition Examination Survey (NHANES) to estimate the prevalence of severe obesity in the US. Their findings of a significant increase in severe obesity among some subgroups, and in particular among children ages 2 to 5 years since the 2013-2014 cycle, is indeed depressing. However, it comes to our attention that the conclusions from Skinner et al.'s study run somewhat contrary to a report released less than a fortnight ago, also from a highly respected journal with authors from highly regarded institutes. Using the same national surveys and an identical definition of severe obesity, Hales and colleagues2 found no significant increase in severe obesity among children ages 2 to 5 years.
The different conclusions between the two studies are largely explained by different reference periods selected by the authors to anchor their analyses. While Skinner et al.'s conclusion is based on comparing prevalence in 2015-2016 to 2013-2014, Hales et al. compared 2011-2012 to the 2007-2008 cycle. These differences in reference points are not unique to these studies as similar contrast was demonstrated by Skinner et al.3 and Ogden et al.4 two years ago. An additional explanation for these different conclusions from same national surveys is associated with the relatively small sample sizes of severe obesity among the pediatric population. With relatively small sample sizes of severe obesity among the pediatric population, there is a risk of insufficient statistical power. Skinner et al. also recommend "…that readers consider both the long-term trends as well as changes over 2-year cycles when considering the effects in specific populations."1 We agree with this recommendation and further caution that overweight and obesity prevalence rates have ebbed and flowed over the last 18 years making conclusions regarding short-term trends more challenging.
National reports of obesity, in particular among the pediatric population, influence health policy and carry considerable weight in the media and public perception. Public health strives to provide a consistent and united message. The chief concern with these contrasting conclusions is that it makes it challenging for the academic society, but more importantly, the general population, to understand whether current strategies used to control obesity are effective. No doubt, the results on any topic with national priority should be verified independently with different sources and from different researchers. It's advisable to present these reports with a robust statistical foundation with common benchmarks for better communication within the professional community and beyond.
References
1. Skinner A, Ravanbakht S, Skelton J, Perrin E, Armstrong S. Prevalence of Obesity and Severe Obesity in US Children, 1999-2016. Pediatrics. 2018;141(3):e20173459. doi:10.1542/peds.2017-3459.
2. Hales C, Fryar C, Carroll M, Freedman D, Ogden C. Trends in Obesity and Severe Obesity Prevalence in US Youth and Adults by Sex and Age, 2007-2008 to 2015-2016. JAMA. 2018. doi:10.1001/jama.2018.3060.
3. Skinner A, Perrin E, Skelton J. Prevalence of obesity and severe obesity in US children, 1999-2014. Obesity. 2016;24(5):1116-1123. doi:10.1002/oby.21497.
4. Ogden C, Carroll M, Flegal K. Prevalence of Obesity in the United States. JAMA. 2014;312(2):189. doi:10.1001/jama.2014.6228.
RE: Let’s Go!’s Response to Latest National Childhood Obesity Data
While new national data reported in Pediatrics by Skinner et al(1) show childhood obesity on the rise, rates in Maine and elsewhere offer a more promising story. In Maine, where Let’s Go!, The Barbara Bush Children’s Hospital at Maine Medical Center, MaineHealth, and dozens of other community organizations have led a coordinated approach to reducing childhood obesity, there are signs of success. In their discussion, Skinner et al write that “obesity prevalence remains high, with scant evidence that…efforts are counteracting the personal environmental forces that contribute to excess weight gain in children, at least on a national scope.” National may be the key word here.
Let’s Go! is a Maine-based obesity prevention program working with a network of local partners to promote and implement policy and environmental changes that facilitate healthy eating and active living in multiple settings across communities. Let’s Go! works with nearly 1,400 sites and each year trains hundreds of educators, school nutrition professionals, out-of-school counselors, program directors, and health care practice teams to create healthier places where children and youth spend time. In the decade-plus that Let’s Go! has been monitoring the prevalence of childhood obesity in Greater Portland, those rates have stabilized. Recent data also show a downward trend specifically among girls aged 3 to 18, from 13% in 2013 to 9.7% in 2016. Even more encouraging are data reported from the statewide Maine Integrated Youth Health Survey (MIYHS).(2) A significant decrease was seen in the percent of third graders who have obesity, from 20.8% in 2015 to 12.3% in 2017. Additionally, the MIYHS data appear to show a slight downward trend in obesity among fifth graders, from 22.6% in 2015 to 19.1% in 2017.
We agree that much more needs to be done on a national scale to address the many drivers of obesity, such as manipulative food marketing tactics and dated federal food policies. Yet local successes such as those found in Maine and those reported recently by Jernigan et al for the Childhood Obesity Declines Project (3) bolster a more optimistic view of the efforts to curb childhood obesity. According to Jernigan et al, “Each community was unique in its approach but all had cross-sector partnerships working together in coordinated comprehensive ways.”(3) Skinner et al(1) recommend the same kind of collaborative, coordinated strategy that Let’s Go! has employed since 2006, both with its earliest partners and most recently with its state collaborator, the Maine Center for Disease Control & Prevention. As local movements are taking root, changing environments, and demonstrating effectiveness, they need reinforcement from a national-level policy and strategy to grow stronger and remain sustainable, as suggested by both Skinner et al(1) and Dr. Ludwig’s Commentary.(4) But let’s not discount the impact that local programs and efforts have already made in leading the charge against this epidemic.
References
1. Skinner AC, Ravanbakht SN, Skelton JA, Perrin EM, Armstrong, SC. Prevalence of obesity and severe obesity in US children, 1999–2016. Pediatrics. 2018;141(3):e20173459
2. Maine Department of Health and Human Services and the Maine Department of Education. Maine Integrated Youth Health Survey Data. Maine CDC Website. Accessed April 2, 2018.
3. Ottley P, Dawkins-Lyn N, Harris C, et al. Childhood Obesity Declines Project: An exploratory study of strategies identified in communities reporting declines. Child Obes 2018;14:S12–S21.
4. Ludwig DS. Epidemic childhood obesity: Not yet the end of the beginning. Pediatrics 2018;141:e20174078.
RE: Severe Obesity Has not Increased in 2-5 yo Children
A recent Pediatrics report by Skinner et al “found no evidence of a decline in obesity prevalence at any age,” but “a significant increase in severe obesity among children aged 2 to 5 years since the 2013– 2014 cycle...1” Dr. Ludwig’s Commentary2 pointed to these data as evidence that “our public health approach has largely failed...” Skinner et al’s interpretation of their data appears misplaced. Although we agree with Ludwig that the efforts to prevent and treat obesity should be intensified, we believe that his conclusion is unnecessarily gloomy.
Data from the National Health and Nutrition Examination survey (NHANES) indicate that the prevalence of obesity in 2-5 yo children decreased from 13.4% in 2003-2004 to 8.3% in 2011-2012, increased to 9.3% in 2013-2014, and increased again to 13.7% in 2015-2016. Skinner et al showed that obesity, but not severe obesity (BMI > 120% of the 95th percentile), increased among 2-5 yo. The increase in obesity prevalence among 2-5 yo relied on a single 2015-2016 data point. Because of the small sample size, this change could be accounted for by an increase of 31 children with obesity between 2013-2014 and 2015-2016. The NHANES Estimation Procedures state that “at least 4 years of data must be combined to obtain an acceptable level of reliability.”3 In addition, the reliability of subpopulation estimates is in part impacted by the limited number of sites selected for each survey cycle and who was sampled in each site. For example, California and Massachusetts have reported substantial differences in prevalence across regions within the state. These observations suggest that the prevalence of obesity within the location sampled could have a disproportionate impact on the estimates of national prevalence.
We were encouraged by the decrease in the prevalence in 2-5 yo observed in NHANES 2011-2012, and the modest increase in 2013-2014. The likelihood that these decreases were real was buttressed by findings that both obesity4 and severe obesity in 2-4 yo children enrolled in WIC decreased significantly between 2010 and 2014, based on over three million children in each survey year. Because 30% of US children aged 2-4 yo are enrolled in WIC, decreases in obesity in this group could account for the decreases observed in the 2011-2014 NHANES samples. Reductions in whole milk and juice in the revised WIC package in 2010 amounted to almost 9000 kcal/month, which could easily explain the declines in prevalence among 2-4 yo children.
The prevalence changes in WIC participants indicate that national policy efforts can have an impact on obesity. However, in the current political climate, further changes at the federal level are unlikely. Initiatives at the state and local level began the movement to reduce smoking, and the decreases in obesity reported in six states and multiple communities5 suggest that local public health initiatives may be succeeding. Spreading and scaling successful initiatives at the state and local level must occur before decreases in obesity occur nationally. It is too early to state that “public health approaches have failed.” It is more appropriate to conclude that we have lacked an effective national approach to control obesity.
References
1. Skinner AC, Ravanbakht SN, Skelton JA, Perrin EM, Armstrong SC. Prevalence of Obesity and Severe Obesity in US Children, 1999-2016. Pediatrics 2018.
2. Ludwig DS. Epidemic Childhood Obesity: Not Yet the End of the Beginning. Pediatrics 2018.
3. Chen TC PJ, Clark J, Shin HC, Rammon JR, Burt VL. National Health and Nutrition Examination Survey: Estimation procedures.: National Center for Health Statistics, 2018.
4. Pan L, Freedman DS, Sharma AJ, et al. Trends in Obesity Among Participants Aged 2-4 Years in the Special Supplemental Nutrition Program for Women, Infants, and Children - United States, 2000-2014. MMWR Morb Mortal Wkly Rep 2016; 65(45): 1256-60.
5. Dietz WH, Economos CD. Progress in the control of childhood obesity. Pediatrics 2015; 135(3): e559-61.