Obesity in Young Children with Autism Spectrum Disorders: Prevalence and Associated Factors

Abstract

Background:

The purpose of this study was to identify rates of overweight and obesity in young children with autism spectrum disorders (ASD) and factors related to overweight.

Methods:

Retrospective chart reviews were conducted for 273 children with ASD [i.e., autistic disorder, Asperger's disorder, pervasive developmental disorder not otherwise specified (PDD-NOS)] after receiving outpatient services with a developmental pediatrician or the developmental team at a children's hospital. Information on child demographics, height and weight, medications prescribed, and adaptive functioning was collected from charts.

Results:

Rates of overweight and obesity in children with ASD were found to be above nationally representative prevalence estimates for children. Among children with autistic disorder, 17.16% had a body mass index (BMI) percentile in the overweight range and 21.89% had a BMI percentile in the obese range. For children with Asperger's disorder/PDD-NOS, 12.50% were considered overweight and 10.58% were considered obese. Neither psychotropic medications prescribed nor adaptive functioning was found to be related to whether the child was overweight or obese.

Conclusions:

Children with ASD are at risk for overweight and obesity, and children with autistic disorder are at greater risk for weight problems than children with Asperger's disorder/PDD-NOS. Further research is needed to identify factors related to overweight in children with ASD.

Introduction

Recent estimates suggest approximately 32% of United States children and adolescents are overweight or obese, as defined by a body mass index (BMI) over the 85^th percentile.¹ Even for younger children specifically (e.g., between 2 and 5 years of age), approximately 21% of children are considered either overweight or obese.¹ Overweight in children has both short-term health consequences, such as high cholesterol or blood pressure,² as well as long-term health consequences. Childhood BMI has been associated with both adult BMI and adult adiposity, and young children with a BMI over the 95^th percentile are four times more likely to be overweight as adults than their normal weight peers.³ Additionally, psychosocial difficulties, including low self-esteem⁴ and decreased quality of life,^4,5 often accompany overweight and obesity. Total direct and indirect economic costs of obesity have been reported to be nearly 10% of total medical expenditures, and the estimated yearly cost rose to $147 billion in 2008.⁶ With rates of obesity increasing at younger ages, the lifetimes costs are drastically increasing and health complications are occurring at younger ages.

There have been few studies of obesity in children with an autism spectrum disorder (ASD), despite the fact that the Center for Disease Control and Prevention (CDC) estimates this condition affects 1 in 88 children in the United States.⁷ ASD, also known as the pervasive developmental disorders (PDD) in the Diagnostic and Statistical Manual of Mental Disorders, 4^th edition (DSM-IV), is a group of neurodevelopmental disorders characterized by varying deficits in communication, impairments in social interactions, or patterns of repetitive or stereotypical behavior or interests. These diagnoses include autistic disorder, Asperger's disorder, and PDD not otherwise specified (PDD-NOS).

Previous research suggests children with ASD may be at as great a risk or even greater risk for overweight and obesity in comparison to their typically developing peers. For example, prevalence data from 1992 to 2003 that were obtained through a chart review methodology indicate children with ASD have prevalence rates of overweight and obesity slightly higher than the general population, with approximately 36% of children having a BMI over the 85^th percentile and 19% of children having a BMI over the 95^th percentile.⁸ Other research has demonstrated rates of obesity in this population to be elevated in comparison to typically developing children.⁹ For example, Curtin and colleagues reported the prevalence of obesity in children with autism to be approximately 30% in comparison to approximately 24% in children without autism.⁹ Obesity is also a common co-morbidity among adults with ASD, with 42% having a BMI above the recommended range.¹⁰

Unique barriers to healthy lifestyles and weight management for children with ASD may be responsible for this high risk of overweight. For example, children with ASD may display frequent food refusal,¹¹ problematic eating behaviors,¹² motor impairments,¹³ and cognitive difficulties.¹⁴ Additionally, numerous barriers to physical activity in children with ASD have been noted.¹⁵ For example, intrapersonal factors, such as time spent playing video games or computers or watching TV or listening to music, and interpersonal factors, such as a lack of peer exercise partners or parents not having enough time, are barriers to children with ASD participating in after-school physical activity.¹⁵ Parents of children with ASD face a multitude of barriers to getting their child to eat a healthy diet and obtain the recommended amount of physical activity, some of which are specific to children with special needs conditions.

Despite an abundance of empirical research regarding obesity trends in typically developing children, minimal research has been dedicated to factors related to obesity in children with autism. This research has produced inconsistent findings regarding which factors are related to obesity in children with ASD and has suggested that the factors related to obesity (e.g., diet or physical activity) may not be the same as those associated with weight problems in typically developing children. For example, research has produced mixed results regarding the relationship between dietary intake (e.g., consumption of certain food groups) and weight problems in children with ASD.^16,17 Additionally, minimal research has focused on the relationship between physical activity and overweight specifically in children with ASD, although it has been suggested that barriers to physical activity in this group may be related to elevated risk for obesity.¹⁸ Similarly, research has not produced consistent findings on whether autism severity is related to overweight. Ho and colleagues found diagnosis severity to be positively associated with obesity,¹⁷ while Memari and colleagues found overall symptom severity, as well as severity of symptoms in the specific areas of sociability and health and physical behavior, was inversely related to BMI.¹⁹

The relationship between psychotropic medications used to treat symptoms associated with ASD and BMI is also unclear. Children with ASD are treated with a variety of psychotropic medications to address their emotional and behavioral symptoms and can include stimulants, serotonin-specific reuptake inhibitors (SSRIs), atypical antipsychotics, and combinations of these medications. Memari and colleagues found the number of psychotropic medications to be inversely related to BMI among girls, whereas a weak positive correlation was found between medications and BMI among boys.¹⁹ Additionally, risperidone has been related to initial increases in BMI among young children or children with lower baseline BMI, but increases in BMI did not continue across time.²⁰ Overall previous research has failed to yield consistent factors associated with weight status for children with ASDs.

The present study aimed to assess the prevalence of overweight and obesity among children diagnosed with ASD, as well as to examine whether psychotropic medications prescribed and adaptive functioning are related to weight. It should be noted that adaptive functioning was examined as a proxy for autism severity because adaptive functioning has been shown to be associated moderately to strongly with autism severity in previous research.²¹ Given the inconsistencies in past research on whether these factors are related to overweight, a descriptive and exploratory approach was taken without a priori hypotheses.

The findings of this study have the potential to contribute to an area of research where results have been inconsistent and unclear. Discrepancies in previously published results may be related to limitations in data collection methods. For example, data on height and weight were often collected from parent report rather than being measured consistently in a medical clinic. Additionally, at times, wide age ranges were examined and discrepancies in findings may be due to different processes occurring at developmental stages (e.g., hormone changes impacting growth and weight gain in adolescence). To help eliminate possible confounding factors, data in the current study were collected from a narrow age range and information was obtained from review of children's medical records, therefore allowing objective assessment of study variables without the bias that can occur when parent-report information is used.

Methods

Participants

A total of 273 participants [M age=3.89, SD=0.91] were identified through a retrospective chart review and were approximately three-quarters male and predominantly Caucasian. Approximately two-thirds of participants were diagnosed with autistic disorder, and approximately one-third of participants were diagnosed with either PDD-NOS or Asperger's disorder. The categories of PDD-NOS and Asperger's disorder were collapsed due to the nature of billing codes. Approximately 95% of children were taking no psychotropic medication to treat symptoms related to their ASD. Detailed demographic and medical information is presented in Table 1.

Table 1.

Demographic and Medical Information for Total Sample

	Number (Percentage)
Gender
Male	222 (81)
Female	51 (19)
Diagnosis
Autistic disorder	169 (62)
Asperger's disorder/PDD-NOS	104 (38)
Race/ethnicity
Caucasian	198 (72)
African American	49 (18)
Hispanic	16 (6)
Other race	10 (4)
Psychotropic medications prescribed^a
No psychotropic medication	260 (95)
Stimulant only	3 (1)
Atypical antipsychotic only	2 (1)
Anticonvulsant only	4 (2)
Alpha-2 adrenergic agonist only	2 (1)
Multiple medications	2 (1)

Due to rounding, the percentages total 101% for the Psychotropic medication prescribed category.

PDD/NOS, pervasive developmental disorder/not otherwise specified.

Procedures

Children with ASD were identified through attendance at a medical appointment with a developmental pediatrician or the developmental team at a midwestern hospital. Medical charts of children attending an appointment in the clinic between November 1, 2005, and May 1, 2009, were reviewed. Inclusion criteria included being between the ages of 2.5 and 5 years, being diagnosed with ASD, and recent height and weight readings. Participants were excluded if they had a diagnosis or impairment that is known to affect weight (e.g., Prader–Willi, Down syndrome, being wheelchair bound; n=2), if the patient did not receive oral feedings (n=0), or if incomplete medical chart information was available (n=10). Due to the retrospective nature of medical chart review, parental consent was not required in accordance with the Institutional Review Board policies of the hospital at which data collection took place. This research was approved by the Institutional Review Board at Children's Mercy Hospitals and Clinics.

All information collected from medical records was originally obtained as part of the child's standard medical care appointment or developmental assessment. Specific information gathered included child demographics (e.g., gender, ethnicity/race), height and weight (to calculate BMI percentile and BMI z-score), diagnosis (i.e., autistic disorder or Asperger's disorder/PDD-NOS), and current medications prescribed. Diagnostic information was obtained by recording the child's billable diagnosis according to the International Classification of Diseases, 9th Revision (ICD-9). The ICD-9 collapses both Asperger's disorder and PDD-NOS into one billable diagnosis and, therefore, a child's diagnosis was coded as either autistic disorder or Asperger's disorder/PDD-NOS. Height and weight were converted to BMI z-scores and percentiles while taking into account the child's age and gender. This was accomplished by using the Baylor College of Medicine Children's BMI-percentile-for-age Calculator (www.bcm.edu/cnrc/bodycomp/bmiz2.html/) which is based on the 2000 CDC BMI-for-Age Growth Charts for girls and boys ages 2–20.²² These measures were used because they are more comparable across child age and gender.

For the children receiving an assessment from the developmental team (n=264), information on adaptive level, as measured by the Vineland Adaptive Behavior Scales, Second Edition (VABS),²³ was also collected. The VABS is a normed survey interview administered as a semistructured interview to parents to assess the frequency at which the child performs various adaptive behaviors through open-ended questions. The VABS provides scores for overall adaptive functioning, as well as adaptive functioning in the specific domains of communication, daily living skills, socialization, and motor skills. Medication status was collected from the medical record and reviewed. The psychotropic medication used by youth in this study included the following known pharmacologic classes—stimulants, atypical antipsychotics, anticonvulsants, alpha-2 adrenergic agonists, or some combination of these medications, referred to as “multiple medications” (see Table 1).

Statistical Plan/Preliminary Analyses

To examine for differences in weight categories, frequency counts and percentages were calculated to assess the number of children with ASD with a BMI over the 85^th or 95^th percentile and therefore met suggested cutoff scores for overweight or obesity, respectively.²⁴ A 2×3 Pearson chi-squared test was used to determine whether the number of children with ASD categorized as under/normal weight, overweight, or obese in our current sample differed from the number of children within these categories in a sample of children and adolescents representative of the overall population. The comparison sample was the subset of children 2–5 years old in the nationwide prevalence study by Ogden and colleagues¹ (Fig. 1). This age range is comparable to the 2.5–5 years age range of children in our ASD sample.

Figure 1.

Percentage of children in each sample with BMI between 85^th and 95^th percentile or over 95^th percentile. (White bars) ASD sample (ages 2–5. ^a15.4% of study participants had a BMI between the 85^th and 95^th percentile (95% CI 11–19.5%). ^c17.6% of study participants had a BMI >95^th percentile (95% CI 13.0–21.9%). (Grey bars) Ogden et al. (2010) national sample¹ (ages 2–5). ^b10.8% of Ogden et al. national sample had a BMI between the 85^th–95^th percentile (95% CI 6.9–14.7%). ^d10.4% of Ogden et al. national sample had a BMI >95^th percentile (95% CI 7.6–13.1%). ASD, autism spectrum disorders; CI, confidence interval.

When chi-squared statistics are significant, individual terms or cells can be examined to identify which ones contribute the most to the significant chi-squared statistic.²⁵ This is done by computing the difference between the observed and expected value for each term and standardizing this difference by dividing it by the square root of the expected value to create standardized residuals.^25,26 Therefore, post hoc comparisons of adjusted standardized residuals to the critical Z-value of 1.96 for each cell were used to determine which values were responsible for the significant chi-squared statistic. Pearson chi-squared tests were also used to examine whether child BMI z-scores categories varied by diagnostic category (i.e., autistic disorder or Asperger's disorder/PDD-NOS) or by level of adaptive functioning. For the analyses assessing the relationship between medications prescribed and weight status, multiple cells had expected values below 5, and therefore, the Fisher exact test was used instead because of its utility with small sample sizes.

Results

Weight Categories

BMI percentiles and z-scores were calculated for age and gender and examined. Ten children (3.66%) were underweight with a BMI below the 5^th percentile, and 173 children (63.37%) were in the normal weight range (BMI in the 5^th to 85^th percentile). Forty-two (15.38%) children were overweight with a BMI between the 85^th and 95^th percentile, and 48 (17.58%) children were obese with a BMI over the 95^th percentile. Overall, 32.96% of our sample was either overweight or obese. These rates of overweight/obesity were found to be significantly different than the rates reported by Ogden and colleagues¹ as estimates for the population of children ages 2–5 [χ² (2)=16.29, p<0.001]. Post hoc comparisons identified that our ASD sample had a higher rate of children with a BMI over the 95^th percentile and with a BMI between 85^th and 95^th percentile in comparison to the normative sample as represented by an adjusted standardized residual greater than 1.96 and corresponding to an alpha of less than 0.05.

Differences in BMI z-Scores by Diagnosis, Medication Status, or Adaptive Functioning

Child BMI category was found to vary significantly by diagnostic group [χ² (2)=8.05, p=0.02], with children with autistic disorder having higher rates of overweight and obesity than children with Asperger's disorder/PDD-NOS, as evidenced by adjusted standardized residuals greater than 1.96. For children with Asperger's disorder/PDD-NOS, 12.50% (n=13) had a BMI between the 85^th and 95^th percentile, whereas 10.58% (n=11) had a BMI over the 95^th percentile. For children diagnosed with autistic disorder, 17.16% (n=29) had a BMI between the 85^th and 95^th percentile, whereas 21.89% (n=37) had a BMI over the 95^th percentile.

Regarding medication, 260 (95%) children were not prescribed a psychotropic medication. Given the low frequency of specific psychotropic medications prescribed, medication status was collapsed into a categorical variable of “prescribed psychotropic medication” (n=13) or “not prescribed psychotropic medication” (n=260). All youth who were prescribed a psychotropic medication were combined into the medication group (n=13). Medication status was found not to be related to child BMI z-score (p=0.57). Given that stimulant medications may have a different impact on appetite- or weight-related variables than other psychotropic medications, further analyses were conducted by excluding the children who were prescribed stimulants (n=3). Medication status without stimulants was also found not to be related to child BMI z-score (p=0.86).

Eighty-nine (33.71%) children had no or only mild impairment in adaptive functioning, whereas 112 (42.42%) children had moderate impairment, and 63 (23.86%) had severe impairment. Weight status was not related to overall adaptive functioning [χ² (4)=1.50, p=0.83], or functioning in the specific domains of motor skills [χ² (4)=5.17, p=0.27], socialization [χ² (4)=7.06, p=0.13], daily living skills [χ² (4)=3.89, p=0.42], and communication [χ² (4)=1.39, p=0.85].

Discussion

The current study supports previous findings that children with ASD are at greater risk for overweight and obesity, as approximately 15% were considered overweight and approximately 18% were obese. Overall, approximately 33% of children with ASD were either overweight or obese with a BMI over the 85^th percentile. Furthermore, our results demonstrated that the percentage of children with ASD who had a BMI in the overweight/obese range was significantly greater than in the normative sample.¹ Importantly, the racial/ethnic background (i.e., Caucasian, African American, Latino) of youth in this sample is similar to the racial/ethnic distribution found in the United States population. This suggests that some other demographic characteristics known to be associated with overweight (e.g., ethnicity) were not responsible for the difference of overweight rates in the current sample in comparison to the national normative data. Higher rates of overweight among children with ASD demonstrate the importance of supporting healthy lifestyles in these children and investigating barriers to healthy living. Children with ASD experience unique barriers to healthy lifestyles, such as frequent use of food as rewards in behavior modification programs due to the ineffectiveness of social reinforcements or selectivity in diets and high rates of food refusal. In our study, children with autistic disorder were found to be at greater risk for overweight/obesity than children meeting criteria for Asperger's disorder/PDD-NOS. Children with autism may be likely to have greater impairment or disability, which is a more significant barrier to a healthy lifestyle. Interestingly, adaptive functioning in the domains of communication, daily living skills, socialization, and motor skills was not found to be related to overweight. This suggests that barriers to healthy living that result from the impairments associated with autism and increased risk for overweight are not present in these domains.

The lack of significant findings may mean that in our sample of young children impairments related to ASD symptomatology, such as motor impairments or fewer opportunities for physical activities with peers, have not had the cumulative impact on obesity status that they might in older children. The rates of overweight and obesity found in the current study are lower than those identified in some prior research, which demonstrated as many as 30% of children with ASD have a BMI score over the 95^th percentile.⁹ However, the present study focuses specifically on young children, which is in contrast to prior research including older children. Younger children have been demonstrated to have lower rates of overweight and obesity in typically developing samples,¹ and our findings may represent a similar trend in children with ASD. Additionally, the present study used a health care–seeking sample and obtained height and weight information from medical chart reviews, which may have contributed to these lower rates. For example, Curtin and colleagues identified a much higher prevalence of obesity in this population when using a community sample and obtaining height and weight information from parent report.⁹ It may be the case that children actively receiving medical care treatment show lower rates of obesity in comparison to community samples. Additionally, parent report of height and weight has been demonstrated to be inaccurate, which may mean study methodology can impact rates identified. When using a retrospective chart review procedure for obtaining height and weight information for children with ASD attending a tertiary clinic, Curtin and colleagues found 19% of the sample to be above the 95% percentile, which is more consistent with the 21% prevalence found in the Autistic disorder group in the current study.⁸

Current study findings contribute to a larger body of research examining factors related to overweight in children with ASD. Psychotropic medication status at this young age was found not to be related to child BMI z-score. The fact that rates of overweight were still high, even in the absence of psychotropic medication, is significant, because this indicates that other factors beyond medication are responsible for the increased risk of this population. Furthermore, avoiding use of psychotropic medication does not serve as a protective factor against overweight, and families of children with ASD are in need of prevention programs for overweight even if medicines are not prescribed. However, there is the possibility that no relationship was found due to the small percentage (only 5%) of children in our sample who were on a psychotropic medication. The children in the sample were young, and therefore more recently diagnosed with ASD; older children with ASD may be more likely to have been prescribed medication due to failed behavior treatment attempts or growing behavioral concerns. Prior research has found varying effects of medication on weight in children with autism with factors such as duration of medication and gender impacting this effect.^19,20 Therefore, additional research examining the impact of medication on weight in older children is needed.

Through use of retrospective medical chart reviews, the current study was able to contribute information regarding diagnosis, level of adaptive functioning, and medication use as they are related to overweight in young children with ASD. Young children with ASD were found to be at significantly greater risk for overweight than a normative sample, and children with autistic disorder were at even greater risk than children diagnosed with Asperger's disorder/PDD-NOS. This highlights the importance of providing weight management programs specifically for children with ASD. We also identified that medication status and adaptive functioning were not related to BMI status in a retrospective study. Given the high risk of overweight in children with ASD, further research is necessary to identify important factors that should be addressed in weight management programs for these children and their families. Future research should examine predictors that may be related to overweight in this population, such as diet and physical activity, and longitudinally evaluate the factors that were examined in this study.

While the use of retrospective medical chart reviews in the current study allowed for objective measurement of BMI, this methodology only captures functioning and weight at one period of time and does not allow for a longitudinal examination of the relationship between study variables and BMI z-score. Future research should monitor the growth trajectories for youth with ASD to assess factors related to weight over time. Another limitation of collecting information from a clinical setting is that a strict research-level protocol for assessment and diagnosis was not used. Furthermore, diagnostic information was obtained through recording of child billable diagnosis, which collapses across Asperger's disorder and PDD-NOS. Therefore, differences in prevalence rates of overweight between these two groups could not be assessed. Last, the sample of children in the current study was young, and therefore our findings may not be generalizable to children of other ages.

Conclusion

The current study identified that young children with ASD are at greater risk for overweight than normative peers, and that children diagnosed with autistic disorder are at greatest risk. Medications prescribed and level of adaptive functioning were found not to be related to weight in children with ASD. This knowledge contributes to the body of literature examining factors related to overweight in this population and establishes a research base for future research and development of appropriate interventions for meeting the needs of these families.

Footnotes

Acknowledgments

A subset of study findings was presented previously as poster presentations at the 2008 Obesity Society Meeting, the 2009 Midwest Conference on Pediatric Psychology, and the 2010 Pediatric Academic Societies Meeting. Funding source: None

Author Disclosure Statement

No competing financial interests exist.

References

Ogden

, Carroll

, Curtin

et al. Prevalence of high body mass index in US children and adolescents, 2007–2008. JAMA, 2010; 303:242–249.

Freedman

, Mei

, Srinivasan

et al. Cardiovascular risk factors and excess adiposity among overweight children and adolescents: The Bogalusa Heart Study. J Pediatr, 2007; 150:12–17.

Freedman

, Khan

, Serdula

et al. The relation of childhood BMI to adult adiposity: The Bogalusa Heart Study. Pediatrics, 2005; 115:22–27.

Griffiths

, Parsons

, Hill

. Self-esteem and quality of life in obese children and adolescents: A systematic review. Int J Pediatr Obes, 2010; 5:282–304.

Williams

, Wake

, Hesketh

et al. Health-related quality of life of overweight and obese children. JAMA, 2005; 293:70–76.

Finkelstein

, Trogdon

, Cohen

et al. Annual medical spending attributable to obesity: Payer-and service-specific estimates. Health Aff (Millwood), 2009; 28:w822–w831.

Autism and Developmental Disabilities Monitoring Network Surveillance Year 2008 Principal Investigators; Centers for Disease Control and Prevention (CDC). Prevalence of autism spectrum disorders—Autism and Developmental Disabilities Monitoring Network, 14 Sites, United States 2008. MMWR Surveill Summ, 2012; 61:1–24.

Curtin

, Bandini

, Perrin

et al. Prevalence of overweight in children and adolescents with attention deficit hyperactivity disorder and autism spectrum disorders: A chart review. BMC Pediatr, 2005; 5:48.

Curtin

, Anderson

, Must

et al. The prevalence of obesity in children with autism: A secondary data analysis using nationally representative data from the National Survey of Children's Health. BMC Pediatr, 2010; 10:11.

10.

Eaves

, Ho

. Young adult outcome of autism spectrum disorders. J Autism Dev Disord, 2008; 38:739–747.

11.

Bandini

, Anderson

, Curtin

et al. Food selectivity in children with autism spectrum disorders and typically developing children. J Pediatr, 2010; 157:259–264.

12.

Provost

, Crowe

, Osbourn

et al. Mealtime behaviors of preschool children: Comparison of children with autism spectrum disorder and children with typical development. Phys Occup Ther Pediatr, 2010; 30:220–233.

13.

Page

, Boucher

. Motor impairments in children with autistic disorder. Child Lang Teach Ther, 1998; 14:233–259.

14.

Nicholas

, Charles

, Carpenter

et al. Prevalence and characteristics of children with autism-spectrum disorders. Ann Epidemiol, 2008; 18:130–136.

15.

Obrusnikova

, Cavalier

. Perceived barriers and facilitators of participation in after-school physical activity by children with Autism Spectrum Disorders. J Dev Phys Disabil, 2011; 23:195–211.

16.

Evans

, Must

, Anderson

et al. Dietary patterns and body mass index in children with autism and typically developing children. Res Autism Spectr Disord, 2002; 6:399–340.

17.

, Eaves

, Peabody

. Nutrient intake and obesity in children with autism. Focus Autism Other Dev Disabl, 1997; 12:187–192.

18.

Keeton

, Kennedy

. Update on physical activity including special needs populations. Curr Opin Pediatr, 2009; 21:262–268.

19.

Memari

, Kordi

, Ziaee

et al. Weight status in Iranian children with autism spectrum disorders: Investigation of underweight. Res Autism Spectr Disord, 2012; 6:234–239.

20.

Demb

, Valicenti-McDermott

, Navarro

et al. The effect of long-term use of risperidone on body weight of children with an autism spectrum disorder. J Clin Psychopharmacol, 2011; 31:669–670.

21.

Perry

, Flanagan

, Geier

et al. Brief report: The Vineland Adpative Behavior Scales in young children with Autism Spectrum Disorders at different cognitive levels. J Autism Dev Disord, 2009; 39:1066–1078.

22.

Kuczmarski

, Ogden

, Guo

et al. 2000 CDC growth charts for the United States: Methods and development. Vital Health Stat, 2002; 11:1–190.

23.

Sparrow

, Cicchetti

, Balla

. Vineland Adaptive Behavior Scales, second edition, Survey Interview Form/Caregiver Rating Form. Pearson Assessments: Livonia, Minnesota, 2005.

24.

Barlow

. the Expert Committee. Expert committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: Summary report. Pediatrics, 2007; 120,Suppl:S164–S192.

25.

Sterns

, Yates

, Moore

. The Practice of Statistics, 4^th. W.H. Freeman: New York, 2010.

26.

Agresi

. Categorical Data Analysis, 2^nd. John Wiley & Sons: Hoboken, NJ, 2002.