Weight Status of Children Participating in the National Spina Bifida Patient Registry

The centralized data center collects deidentified data from each NSBPR center. After initial enrollment, demographic, social, clinical characteristics, and treatment history are updated annually or biennially. Data collection and transfer procedures were approved by each site’s institutional review board and appropriate assents and consents were obtained.

This descriptive cross-sectional study used NSBPR data collected from all eligible individuals attending participating centers from 2009 to 2018. Subjects were excluded from this study if they had missing height or weight (n = 145), an extremely low or high body mass index (BMI ≤ 10 kg/m² or ≥50 kg/m²; n = 46), or yearly gains in height or weight of more than 3 standard deviations (n = 3), because of the concern for data inaccuracy. The final sample included 7215 children aged 2 to 19 years (Fig 1).

Per the NSBPR protocol, weight was measured in kilograms on a calibrated scale for children able to stand independently while wearing minimal clothing and with all outerwear removed. For those unable to stand independently, weight was measured using a wheelchair scale. The weights of the chair and clothing or braces were subtracted from the total weight to obtain a “child-only” weight. Standing height (SH), for those who could stand, was measured in centimeters with a stadiometer. For those unable to stand, arm span (AS) was measured (cm), while sitting in a chair with arms extended outward laterally, as a straight line across the child’s back from furthest extension of fingers on 1 side of the body to furthest extension of fingers on the opposite side of the body. When SH or AS was not measurable, recumbent length (RL) was measured (cm) as the distance from the crown of the head to the heel of the foot and can be used directly as a proxy for SH.

Demographic and clinical characteristics were collected. See Fig 2 for categories and definitions.

The sample was divided by SB type (MMC versus NMMC) and children were placed into age groups used by the NHANES of 2 to 5, 6 to 11, and 12 to 19 years of age. To calculate age- and sex-specific BMI percentiles and determine weight status, we used the CDC’s equations coded in SAS.¹³  From these results, each measure from participants was classified to a weight status category: underweight (<fifth percentile), normal weight (5th to <85th percentile), overweight (85th to <95th percentile), and obese (≥95th percentile).

χ² tests assessed the univariate associations between weight status category and each demographic or clinical factor. We dichotomized weight status categories as under/normal weight and OW/OB. To account for the effect of repeated observations of weight status and other time-dependent variables from the same person, we used GEE models with logit link function.¹⁴  The GEE regression models also accounted for correlated data from participants clustered by clinic. Multiple GEE regression models were conducted to test the independent association between weight status and demographic and clinical factors; an independent correlation structure was specified. Statistical tests were all 2-sided, and P values <.05 were considered statistically significant; 95% confidence intervals (CIs) were calculated for odds ratio (OR) estimates. Statistical analyses were performed using SAS version 9.4 (Cary, NC, USA). Per NSBPR protocol, a secondary independent analyst replicated the statistical analysis.

Study participants (N = 7215) were aged 2 to 19 years (mean = 11.1, SE = 0.06) with children 12–19 years the largest group at 44.4%, 6–11 years 32.9%, and 2–5 years 22.6%. There were slightly more females (51.4%) than males. Most participants were non-Hispanic white (57.2%) followed by Hispanic or Latino (25.1%) and non-Hispanic Black (7.5%). More than three-fourths of the sample had MMC (76.3%) and the majority (60.2%) was classified as community ambulators (Table 2).

For the entire sample, 45.2% of children were OW/OB. However, OW/OB prevalence differed by SB type (MMC 49.2% vs NMMC 32%) (Table 3). Children with MMC in the 2- to 5-year age group had the highest prevalence of OW/OB (52.7%) followed by the 6- to 11- and 12- to 19-year age groups (47.4% and 48.9%, respectively) (Table 4). For the NMMC group, the opposite pattern emerged where the 12- to 19-year age group had the highest prevalence (37.5%) followed by 6- to 11- and the 2- to 5-year age groups (29.2% vs 27.4%) (Table 5).

For the MMC sample, 5.4% were underweight, 45.3% normal weight, 19.5% OW, and 29.8% OB with 19.8% in obesity class 1, 6.6% in severe obesity class 2, and 3.5% in severe obesity class 3. The NMMC sample had lower frequencies of overweight or obesity, with 5.8% underweight, 62.2% normal weight, 15.7% OW, and 16.2% OB with 9.9%, 3.9%, and 2.5%, respectively, in obesity class 1, 2, and 3 (Table 3). Both groups exceeded the expected 5% for obese category according to the US population charts: nearly 30% for MMC and more than 16% for NMMC.

There were differences in the univariate analyses for children with MMC in the weight categories based on demographic (age, sex, and race/ethnicity) and clinical variables (functional level of lesion, ambulation, and number of shunt surgeries) (Table 4). Fewer differences occurred for children with NMMC (Table 5). In the MMC group, more females and children with a ventriculoperitoneal shunt were OW/OB. Weight categories differed significantly by age, race/ethnicity, functional level of lesion, and ambulation in both MMC and NMMC, with the highest percentage of OW/OB in the 2 to 5 year olds (MMC) and 12 to 19 year olds (NMMC) and individuals identified as Hispanic or Latino, followed by non-Hispanic white and non-Hispanic Black. Children with lumbar lesions had the highest OW/OB for MMC (mid-lumbar) and NMMC (high-lumbar) groups. The prevalence of OW/OB differed by ambulation, with household ambulators (MMC) and therapeutic ambulators (NMMC) the highest. There was no difference in distribution of weight categories by clinic region (Tables 4 and 5).

The multivariable GEE model identified 4 demographic (age, sex, race/ethnicity, region of clinic) and 1 clinical variable (level of lesion) associated with OW/OB for those with MMC (Table 6). For age, the odds of OW/OB were significantly lower for the 2 older age groups when compared with the 2- to 5-year-old group (6 to 11 year olds: OR, 0.79; 95% CI, 0.72–0.88; P < .001; 12 to 19 years old: OR, 0.88; 95% CI, 0.78–0.99; P = .038). The odds of OW/OB in females were significantly higher than in males (OR, 1.19; 95% CI, 1.07–1.32; P = .002) and in those of Hispanic or Latino ethnicity than non-Hispanic white (OR, 1.65; 95% CI, 1.44–1.88; P < .001). In contrast, the odds of OW/OB for those whose race was “other” was significantly lower than non-Hispanic white (OR, 0.78; 95% CI, 0.63–0.96; P = .02) and those attending clinics in the western region of the US when compared with those in the Northeast (OR, 0.71; 95% CI, 0.59-0.84; P ≤ .001). Having a low- (OR, 1.16; 95% CI, 1.01–1.34; P = .03), mid- (OR, 1.32; 95% CI, 1.14–1.53; P ≤ .001) or high-lumbar (OR, 1.45; 95% CI, 1.18–1.78; P ≤ .001) lesion increased the odds of OW/OB significantly compared with those with sacral lesions.

A major finding of this study was the high prevalence of OW/OB among children with SB, especially those with MMC. In this study, almost one-half of the children with MMC were OW/OB, higher than the national OW/OB estimate of 35.1% among TD children 2–19 years of age.¹⁵  Obesity prevalence rates by age group for those with MMC were consistently higher than the general US population ages 2 through 5 (33.4% vs 13.9%), 6 through 11 (28.2% vs 18.4%) and 12 through 19 years (29.2% vs 20.6%) respectively.¹⁶  In addition to obesity-related comorbidities well-described in the general population, children with MMC face additional risks compounded by OW/OB, such as compromised mobility, self-care, and independence. The percentages of those with MMC in each obesity class were also higher than the prevalence rates in US population of children, respectively for class 1 (19.7% vs 12.5%), 2 (6.6% vs 4.1%), or 3 (3.5% vs 1.9%) obesity.¹⁷  The presence of severe obesity in childhood, coupled with diagnosis-specific concerns, may result in a deleterious cycle of further obesity and negative outcomes across the lifespan.

Differences in children’s weight status by demographics, age, race/ethnicity, and sex were seen among those with MMC. Somewhat unexpectedly, higher levels of OW/OB were in the 2- to 5-year MMC age group. Toddlers with SB often achieve motor milestones later than TD peers,^18,19  which may contribute to larger percentages of OW/OB. Rates of OW/OB may then decrease because these children age and become more mobile. Inconsistent with findings in TD children, females with MMC in all age groups were more likely to be OW/OB than males.²⁰  Although TD postpubertal females in the second decade of life are more commonly OW/OB, no differences by sex have been identified in prepubertal children. Some females with MMC enter puberty early, as early as 8 to 10 years of age.²¹  This could contribute to pubertal storage of fat cells and increase in OW/OB when compared with same-age males. The differences by age and sex have implications for anticipatory guidance including appropriate nutrition and promotion of early mobility through therapies and at-home physical activities.

Individuals who identified as Hispanic or Latino had a higher prevalence of OW/OB than their non-Hispanic white or Black counterparts for MMC and NMMC groups. Similarly, CDC data reveal that Hispanic or Latino children have the highest rate of obesity in the general population.²²  However, unlike our study, the CDC data on non-Hispanic Black children reveals a higher prevalence of obesity than the non-Hispanic white population.²²  Because of the small numbers, our data may not accurately represent the non-Hispanic Black population with SB. The increased prevalence of obesity in the Hispanic/Latino population is consistent with other SB studies that demonstrate disparities in clinical outcomes related to race/ethnicity.²³  The lower odds of being OW/OB in the other race/ethnicity category may, in part, be explained by a large proportion of individuals classified as Asian (38.8%), a group with lower rates of obesity in the general population.²⁴  Understanding how race/ethnicity is associated with OW/OB may have been clarified if measures of socioeconomic status were available (eg, income) because lower socioeconomic status and obesity have been related in the general population.²⁵ 

Similar to age, sex, and ethnicity, geographic location was significantly associated with OW/OB in the multivariable analysis. Those attending clinics in the West were less likely to be OW/OB when compared with those in the Northeast, Midwest, and South. This difference may be explained by the western climate and culture that support year-round activity for individuals with a physical disability.

Functional level of lesion, ambulation, and shunt status are indicators of SB severity, but only functional level of lesion was related to weight status in the multivariable analysis. However, those with higher (eg, thoracic, high lumbar) lesions are more likely to have the most limited ambulation and a shunt. These severity indicators were related to weight status in the univariate analysis. Because these variables are moderately related to each other, it is possible that ambulation and shunt status did not add significant explanatory power to level of lesion.²⁶  In addition to weight management challenges encountered by all children,^27,28  these findings support further investigation into unique SB-related challenges. For example, preliminary evidence suggests that individuals with SB have a decreased total energy expenditure consequently influencing the daily energy intake required.²⁹  In addition, sedentary activity and screen time may be increased from lack of recreational options and time-consuming medical regimens.^3,30  Finally, comfort eating may occur, and role modeling of healthy habits may be challenged because of caregiving-related stress or decreased access to healthier foods.^30,31 

The prevalence of OW/OB and related demographic factors in the NMMC sample resemble TD children, generally increasing from youngest to oldest and highest in the Hispanic or Latino participants. Similarly, the risk for obesity-related health problems is present and indicates the need for increased monitoring and interventions throughout childhood.

There are several limitations to our study. The use of BMI to assess adiposity in children with SB has been questioned. Individuals with SB have reduced lean mass; thus, BMI may underestimate weight status.^11,32  However, without alternative options and because of its cost-effectiveness and ease of use in a clinical setting, BMI is widely accepted as a screening tool for OW/OB.³³  Although AS is the recommended alternative for SH in the NSBPR for those unable to stand, it required GEE modeling to create an adjustment for use in the BMI calculation. This is also a strength of the study, but because the equation was based on a limited subsample, it may not be representative of the population and needs validation in future analyses. Overall, the challenges associated with the measurement of height (eg, lack of standardized assessment) for individuals who are unable to stand indirectly affects the reporting of and accuracy of prevalence rates of OW/OB for this population.

Although the NSBPR has specific protocols for anthropomorphic measurements, methodological variations may exist among clinics because of orthopedic impairments of participants. Further, although the NSBPR represents most participants attending centers, the percentage of those participating differs by site. Similarly, the centers may not represent the SB population in the United States, limiting generalizability of the findings. However, the registry provides the largest population of children with SB to date. Longitudinal analysis of data from multiple clinic visits offers a more complete assessment of weight status and its association with covariates. The use of a GEE model enabled us to examine the associations between covariates and weight status while simultaneously capturing the influence of between- and within-subject variation.

This study documents a high prevalence of OW/OB in children with SB and identifies demographic and clinical factors associated with the higher prevalence. The negative consequences of OW/OB in this population underscore the need for early identification, anticipatory guidance, prevention, and treatment of OW/OB. Further investigation of how unique SB-related challenges influence weight management are warranted.

The authors extend a special acknowledgment to Rodolfo Valdez for his assistance with the conception of the manuscript. The authors thank the many individuals with spina bifida and their family members who participated in this research, without whom the National Spina Bifida Patient Registry (NSBPR) would not be possible. The NSBPR has also been successful because of the contributions of the Centers for Disease Control and Prevention, the Spina Bifida Association, and all members of the NSBPR Coordinating Committee. Members of this Committee during the collection of the data reported are listed in alphabetical order and were Richard Adams, Texas Scottish Rite Hospital for Children, Dallas, Texas; Pat Beierwaltes, Children’s Hospital of Michigan, Detroit, Michigan; Timothy Brei, Riley Hospital for Children, Indianapolis, Indiana; Robin Bowman, Ann and Robert H. Lurie Children’s Hospital of Chicago, Chicago, Illinois; Heidi Castillo, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, and Texas Children’s Hospital, Houston, Texas; James Chinarian, Children’s Hospital of Michigan, Detroit, Michigan; Mark Dias, Hershey Medical Center, Hershey, Pennsylvania; Brad Dicianno, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania; Nienke Dosa, Upstate Golisano Children’s Hospital, Syracuse, New York; Carlos Estrada, Boston Children’s Hospital, Boston, Massachusetts; Kurt Freeman, Oregon Health and Science University, Portland, Oregon; David Joseph, Children’s Hospital of Alabama, Birmingham, Alabama; Lynne Logan, Upstate Medical University, Syracuse, New York; Pamela Murphy, District Medical Group Children’s Rehabilitative Services, Phoenix, Arizona; Jacob Neufeld, Children’s Hospital and Research Center at Oakland, Oakland, California, University of California at San Francisco Benioff Children’s Hospital, San Francisco, California, and St. Luke’s Boise Medical Center, Boise, Idaho; Joseph O’Neil, Riley Hospital for Children, Indianapolis, Indiana; Michael Partington, Gillette Children’s Specialty Healthcare, St. Paul, Minnesota; Paula Peterson, Primary Children’s Medical Center, Salt Lake City, Utah; Elaine Pico, Children’s Hospital and Research Center at Oakland, Oakland, California, and University of California at San Francisco Benioff Children’s Hospital, San Francisco, California; Karen Ratliff-Schaub, Nationwide Children’s Hospital, Columbus, Ohio; Kathleen Sawin, Children’s Wisconsin, Milwaukee, Wisconsin; Kathryn Smith, Children’s Hospital Los Angeles, Los Angeles, California; Katherine Steingass, Nationwide Children’s Hospital, Columbus, Ohio; Stacy Tanaka, Monroe Carell Jr. Children’s Hospital at Vanderbilt, Vanderbilt, Tennessee; Jeffrey Thomson, Connecticut Children’s Medical Center, Hartford, Connecticut, and Shriners Hospitals for Children Springfield, Springfield, Ohio; David Vandersteen, Gillette Specialty Clinics, St. Paul, Minnesota; William Walker, Seattle Children’s Hospital, Seattle, Washington; John Wiener, Duke University Medical Center, Durham, North Carolina; Pamela Wilson, Children’s Hospital Colorado, Denver, Colorado; and Hadley Wood, Cleveland Clinic, Cleveland, Ohio.

AS

arm span

BMI

body mass index

CDC

Centers for Disease Control and Prevention

CI

confidence interval

GEE

generalized estimating equation

MMC

myelomeningocele

NMMC

nonmyelomeningocele

NSBPR

National Spina Bifida Patient Registry

OW/OB

overweight and obesity

OR

odds ratio

RL

recumbent length

SB

spina bifida

SH

standing height

TD

typically developing