Citation
Neurocognitive Findings in Prader-Willi Syndrome and Early-Onset Morbid Obesity

Material Information

Title:
Neurocognitive Findings in Prader-Willi Syndrome and Early-Onset Morbid Obesity
Creator:
MILLER, JENNIFER LYNNE ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Child psychology ( jstor )
Childhood ( jstor )
Chromosomes ( jstor )
Lesions ( jstor )
Magnetic resonance imaging ( jstor )
Morbid obesity ( jstor )
Obesity ( jstor )
Prader Willi syndrome ( jstor )
Siblings ( jstor )
White matter ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Jennifer Lynne Miller. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
6/30/2006
Resource Identifier:
109787824 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

NEUROCOGNITIVE FINDINGS IN PRADER-WILLI SYNDROME AND EARLY-ONSET MORBID OBESITY By JENNIFER LYNNE MILLER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Jennifer Lynne Miller

PAGE 3

iii ACKNOWLEDGMENTS I thank my mentor, Daniel J. Driscoll , Ph.D., M.D., for his guidance and support and my family for their patience.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii ABSTRACT.....................................................................................................................vi ii CHAPTER 1 INTRODUCTION........................................................................................................1 2 METHODS...................................................................................................................4 Participants................................................................................................................... 4 Weight of Participants..................................................................................................6 Cognitive and Achievement Tests................................................................................7 Behavioral Analysis......................................................................................................7 Socioeconomic Status...................................................................................................8 Statistical Analysis for Cognition and Behavior..........................................................8 MRI Scanning Procedure..............................................................................................9 3 COGNITIVE AND BEHAVIORAL FINDINGS......................................................11 Cognitive Results........................................................................................................11 Achievement...............................................................................................................11 Behavioral Findings....................................................................................................11 Socioeconomic Status.................................................................................................12 4 MRI RESULTS..........................................................................................................15 White Matter Lesions.................................................................................................15 Parietal-Occipital Lobe Abnormalities.......................................................................16 Sylvian Fissure Polymicrogyria..................................................................................16 Insula Closure.............................................................................................................17 5 CONCLUSIONS........................................................................................................20

PAGE 5

v LIST OF REFERENCES...................................................................................................27 BIOGRAPHICAL SKETCH.............................................................................................31

PAGE 6

vi LIST OF TABLES Table page 2-1 Characteristics of subjects undergoing MRI evaluation..........................................10 2-2 Characteristics of subjects undergoing cognitive, achievement and behavioral evaluation.................................................................................................................10 4-1 MRI findings............................................................................................................18 4-2 Relationship between presence of white matter lesions and clinical phenotype......18

PAGE 7

vii LIST OF FIGURES Figure page 2-1 Percentage of ideal bod y weight at time of eval uation. EMO patients were significantly heavier at the time of evalua tion than either PWS patients or than controls.....................................................................................................................10 3-1 Comparison of General Intellectual Abilities (GIA). EMO patients have a significantly lower GIA than controls (p <0.001; c )................................................13 3-2 Comparison of total achievement scor es. EMO individuals have lower total achievement than controls (p = 0.001), but are not significantly different than PWS patients............................................................................................................13 3-3 Comparison of adaptive sk ills and problem behaviors............................................14 3-4 Externalizing and internalizing be haviors as assessed by Parent BASC evaluation.................................................................................................................14 4-1 White matter lesions in subjects with early-onset morbid obesity...........................18 4-2 Abnormal parietal sulci............................................................................................19 4-3 Abnormal closure of insula......................................................................................19

PAGE 8

viii Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science NEUROCOGNITIVE FINDINGS IN PRADER-WILLI SYNDROME AND EARLY-ONSET MORBID OBESITY By Jennifer L Miller December 2005 Chair: Daniel J. Driscoll Major Department: Clinical Investigation This case-control study was undertaken to examine whether early-onset morbid obesity is associated with c ognitive impairment, neuropathol ogic changes, and behavioral problems. We compared head MRI scans and cognitive, achievement, and behavioral evaluation of subjects with Pr ader-Willi syndrome (PWS), early -onset morbid obesity of unknown etiology (EMO), and normal-weight sibli ng controls from each of these groups. Head MRIs were done on 17 subjects with PWS, 18 with EMO, and 21 normal-weight siblings, while cognitive, achievement and be havioral evaluation were performed on 19 patients with PWS, 17 individuals with EMO, and 24 normal-weight siblings. The study was conducted on the General Clinical Resear ch Center. Patients were identified as EMO based on a history of obesity (BMI > 95% or >150% of ideal body weight) before age 4 years. The Woodcock-Johnson Test of Cognitive Ability and Academic AchievementThird Edition was used to assess general intellectual ability and achievement. The

PAGE 9

ix Behavioral Assessment System for Children was used to obtain information regarding behavioral symptoms. The interpreters of the 3-dimensional (3-D) MRI scans were blinded to the diagnosis of the patients. Two-way t-tests were used to analyze differences in means across groups. We found that the mean General Intellect ual Ability (GIA) score of the EMO group was 77.4 17.8; PWS 63.3 14.2; and controls 106.4 13.0 (p=0.021 for EMO vs PWS; p<0.001 EMO vs sibs and PWS vs sibs). Achievement scores for the three groups were: EMO 78.7 18.8; PWS 71.2 17.0; and controls 104.8 17.0 (p=0.33 for EMO vs PWS, p<0.001 EMO vs PWS and PWS vs sibs). Si gnificant negative behaviors and poor adaptive skills were found in the EMO group. Intracranial abnormalities noted on MRI in individuals with PWS included ventriculome galy, parietal lobe abnormalities, sylvian polymicrogyria, incomplete insular closur e, and white matter lesions. All 17 PWS subjects examined had some, if not all, of these findings, while none of the normal weight control subjects had any of these findings. Five of the EMO subjects had white matter lesions, similar to those seen in patients w ith PWS, but none of the other intracranial abnormalities seen in subjects with PWS. We conclude that individuals with early-onset morbid obesity have significantly lower cognitive function and more behavioral problems than controls with no history of childhood obesity. Additionally, we found that EMO and PWS subjects have MRI abnormalities that have not been described previously.

PAGE 10

1 CHAPTER 1 INTRODUCTION Childhood obesity is a major health probl em throughout the world, with increasing prevalence, severity, and appearance at younger ages. Data from the International Obesity Task Force indicate that 22 million of the world’s children under 5 years of age are overweight or obese (1). A number of environmental cha nges have occurred over the past decade, including reduced walking, increased portions sizes, and increased availability of inexpensive, high-fat/high-calorie foods. Such evidence suggests that the rise in obesity among most age groups in th e population is primarily due to external societal changes. In contrast, since infants and very young children have limited access to these environmental factor s, the development of obesity in the under-4-year-old age group is most likely due to genetic influe nces. Determination of the etiology and consequences of childhood obesity is crucial to the understanding a nd ultimate treatment of this condition. Prader-Willi syndrome (PWS) is the most commonly recognized genetic cause of childhood obesity, characterized by infantile hypot onia, mental retardation, short stature, hypogonadism, early-onset obesity, and hyperphagia. Typical psychiat ric and behavioral manifestations include self-injury (e.g., nail biting and skin picking), explosive outbursts, and obsessive-compulsive behaviors. PW S presents an opportunity to correlate neuropathologic abnormalities with the comple x neurobehavioral phenotype. Only a few small studies of brain structure in PWS patie nts have been published. One autopsy study reported cortical gyral malformations and he terotopic neurons in the cerebellar white

PAGE 11

2 matter in a 6-month-old girl with PWS (2). Leonard et al; using traditional magnetic resonance imaging (MRI), reported abnormal Sylvian fissure morphology in four patients with PWS (3). An abnormal cerebral gyral folding pattern was described using 3 dimensional (3-D) MRI in one infant with PWS (4). Abnormal pituitary findings have been well documented in patients with PWS (5, 6). Recent advances in MRI technology allow the brain to be visualized in 3 dimensions permitting evaluation of in vivo neuropathology. Approximately 70% of PWS cases are due to a genetic deletion on chromosome 15 (15q11-13), 25% of PWS cases are from a ma ternal uniparental disomy (UPD) of chromosome 15, and the remaining cases result from imprinting defects (7, 8). The early-onset morbid obesity is the most sign ificant health problem and the primary cause of morbidity and mortality in individuals with PWS. Adipose tissue produces various secretory pr oteins, including leptin, tumor necrosis factor-alpha (TNF), and adiponectin. Obesity-indu ced abnormal levels of these “adipokines” cause increased insulin resi stance, with resultant hyperinsulinemia, dyslipidemia, inflammation, and endothelial dysfunction. Childhood obesity results in longer exposure to the adipokines produced by the adipose tissue, putting obese children at risk for developing the me tabolic and cardiovascular co mplications of obesity at relatively younger ages. By 6 – 12 years of ag e, one in four typical overweight children has hyperinsulinemia with impaired glucose tolerance, and 60% of these children have evidence of hypertension, dyslipidemia, subclinical inflammation, and disturbed endothelial function (1, 9). However, little is known about the effects of early childhood obesity on the developing brain.

PAGE 12

3 Prader-Willi syndrome is an excellent mode l to help assess the causes and effects of early childhood morbid obesity, since it has a characteristic age of onset of obesity in addition to well-described learning difficultie s and behavioral problems (10-13). The most striking characteristic of patients with PWS is the development of morbid obesity by 2-3 years of age. However, many child ren also have early-onset morbid obesity (EMO), but, to date, do not have a discerni ble genetic abnormality to account for their obesity. This study was designed based on the hypothesis that the development of obesity early in life can cause damage to th e developing brain, result ing in intracranial abnormalities, as well as cognitive and behavioral impairments.

PAGE 13

4 CHAPTER 2 METHODS Participants We evaluated subjects with PWS, a we ll-defined cause of childhood-onset obesity, mental retardation, and behavioral problems, as well as subjects with early-onset morbid obesity (EMO) of unknown etiology, in order to compare such factors as birth history, medical problems, age of onset of obesity, nutritional assessment, cognitive function, MRI findings, and behavior. To reduce the effects of socioecono mic variables, the siblings of both of these obese groups se rved as the control group. This study was approved by the University of Florida Inst itutional Review Board, and all participants and/or their guardians provide d written informed consent. In this case-control study we comp ared head MRI scans and cognitive, achievement, and behavioral evaluation of subj ects with PWS, EMO, and sibling controls from each of these groups. The subjects with PWS served as an obesity comparison group for the children with EMO. The study was conducted over two days on the General Clinical Research Center. Molecular testing was done on all PWS subjects using previously described methods (7). Overall, we studied 25 subject s with PWS who were recruited from the Genetics Clinic at the University of Florid a. Ten PWS subjects with a deletion of the chromosomal 15q11-13 region and 7 with mate rnal uniparental disomy (UPD) of chromosome 15 (age 4 years to 39 years) had a 3-D MRI of their head (Table 2-1), while

PAGE 14

5 11 subjects with deletion and 8 with UPD (age 4 years to 39 years) underwent cognitive and behavioral evaluation (Table 2-2). All PW S subjects were obese before age 4 years. Subjects with EMO were recruited from the General Pediatric, Pediatric Endocrinology, and Pediatric Genetic s clinics at the University of Florida. Subjects were selected based solely on a history of wei ght >150% of ideal body weight (IBW) for height before the age of 4 years and no rec ognized syndromal cause of obesity. Growth charts documenting the weight, height, and h ead circumference from birth were required for entry into the study to ascertain that the onset of obesity was before the age of four years. The EMO subjects who underwent c ognitive/achievement/b ehavioral evaluation and MRI were age 4-22 years (Table 2-1, 2-2). Sibling controls who were closest in ag e to the proband and who did not have a history of childhood obesity or a known gene tic abnormality (e.g., Down syndrome) were recruited from both the PWS and the EMO families, although not every proband had a sibling. The sibling controls were age 3.5 years to 43 years (Table 2-1, 2-2). All participants in the EMO group had a nor mal chromosomal analysis, as well as a normal fluorescence in situ hybridization and DNA methylati on analysis using the small nuclear ribonucleoprotein N ( SNRPN ) probe located in the Prader-Willi region of chromosome 15 (7). Sequence analysis of the melanocortin 4 receptor (MC4R) gene did not reveal any mutations (14). Fragile X DNA testing done by polymerase chain reaction and Southern blot analysis ( 15) revealed no mutations. No EMO subjects were found to have a leptin deficiency by commercial te sting with a radioimmunoassay developed by Quest Diagnostics.

PAGE 15

6 In addition to the extensive genetic tes ting described above, we also performed array-based comparative genomic hybrid ization (CGH) on all the EMO subjects searching for a chromosomal deletion or duplication that would explain their neurocognitive impairment and/or obesit y. A total of 174 genomic BAC (bacterial artificial chromosome) cl ones distributed across the length of the long arm of chromosome 15 were used for the microarray. Th e highest density of clones is across the approximately 10 Mb 15q11-q14 interval encompassing the Angelman/Prader-Willi critical region, including th e common deletion/duplication breakpoints. This particular microarray achieved a resolution of greater th an one clone per megabase for the entire chromosome 15. Additionally, over 130 cl ones (BACs and P1-derived artificial chromosomes) specific for the subtelomeric regions of all other chromosomes were included. Also included were clones specific for the genomic regions for the SIM1 gene (chromosome 6q16.3) and a number of genomic regions (17p11, 22q11, and 10q23.3) implicated in autism spectrum disorders. The validation of genomic clones, and production and analysis of array CGH expe riments were carried out as described previously (16, 17). Despite this extensive analysis, no segmental losses or gains were identified in any of the subjects in the EMO cohort. Only one EMO subject who underwent cogni tive and behavioral testing did not have fragile X, MC4R, or array CGH done due to insufficient DNA. However, his similarly affected maternal half brother did have this testing and was normal. Furthermore, fragile X testing on the mother was normal. Weight of Participants Age of onset of obesity for subjects with PWS was between 18 months and 36 months, although the majority of the PWS subjects became obese after 24 months. The

PAGE 16

7 range of weight at their time of testing va ried from ideal body weight (IBW) to 325% of IBW for height, and as a group were an average of 180% overweight. The EMO group had more variation in age of onset of obesity (2 months to 3 years) and was more consistently and severely obese, being an average of 225% overweight at the time of evaluation (Figure 2-1). However, percent IBW underestimates the degree of obesity in PWS due to their decreased musc le mass (18). None of the sibling controls were obese before age 10 years. Cognitive and Achievement Tests Cognitive and achievement testing was done using the Woodcock-Johnson Test of Cognitive Ability and Academic AchievementThird Edition (WJ-III) (19). The WJ-III is the most comprehensive test battery avai lable for the clinical assessment of children and youth. The Tests of Cognitive Ability on th e WJ-III were designed to measure all the broad cognitive abilities (i.e., Fluid Reasoning, Comprehension-Knowledge, Visual Processing, Processing Speed, Long-Term Retr ieval, and Short-Term Memory). The Tests of Achievement measure scholastic perf ormance in reading, mathematics, written language, knowledge, and academic skills. Co -norming of the Tests of Cognitive Ability and Achievement on the WJ-III permits identification of ability/achievement discrepancies on the same standardization sa mple. The WJ-III allows for testing of individuals from 2-79 years of age. Re sults for the general population typically approximate a score of 100 15. Behavioral Analysis Psychosocial adaptation of par ticipants was measured with the parent, teacher, and youth self-report forms of the Behavioral A ssessment System for Children (BASC) (20). The BASC is a nationally-standardized quest ionnaire designed to measure the adaptive

PAGE 17

8 and problem behaviors of children and adoles cents between the cognitive ages of 4-18 years in school, home, and community set tings. Items are grouped into broadand narrow-band scales (mean = 50, standard devi ation = 15). For the Behavioral Symptoms Index (BSI), which is an overall eval uation of negative behaviors (aggression, hyperactivity, withdrawal, depression, anxi ety, and somatization) , externalizing and internalizing problem behaviors scaled scores in the 41-59 range are average; in the 6069 range “at-risk”; and 70 a nd above “clinically significant” . For the Adaptive Skills, scaled scores in the 41-59 range are average; in the 30-40 range “at risk”; and 30 or below “clinically significant”.A ll individuals were given the parent, teacher, and youth Behavioral Assessment System for Children (B ASC) forms as applicable. Unfortunately, the majority of the teacher BASC forms were not returne d, and many individuals did not qualify to fill out the youth (self-report) BASC form, as the individual must be cognitively at least 8 years of age to complete this form. Since all of the patients who participated in the study were accompanie d by their parents, the most complete behavioral information was obtained from th e Parent Assessment portion of the BASC. Administrators of both the WJ-III and the BASC tests were blinded as to the diagnosis of the patients. Socioeconomic Status Socioeconomic status (SES) was assessed for every subject by having the parents fill out a questionnaire detailing parental education and annual in come prior to the visit. The answers were confirmed by verbal ques tioning at the time of the evaluation. Statistical Analysis for Cognition and Behavior This was a prospective nonrandomized study examining cognitive and behavioral scores in three different test groups – EMO, PWS, and sibl ing controls. Each group was

PAGE 18

9 compared in terms of scores on a standard ized test of cognitiv e abilities (GIA) and achievement (which evaluates the presence of learning disabilities). The primary analysis was a comparison via analysis of varian ce of the three groups (PWS vs. EMO vs. Control) on the basis of the (a) Age Standa rdized GIA and (b) Adaptive Skills (ADPT-Tscore) on the BASC controlling for gender. If significant at p<0.05, this was followed up by two sample two-sided t-tests on each pair of groups. Other variables were compared via two sample two-sided t-tests between the pairs of groups as s econdary analyses to provide pilot information for future researc h. Significant findings on secondary variables will need confirmation in a future study. MRI Scanning Procedure Subjects were not sedated for the MRI scans which were performed using a headdedicated 3 Tesla Allegro scanner (Siemens , Erlangen, Germany). Both Turbo Spin Echo (TSE) and Turbo FLASH pulse sequen ces were used for quick whole-brain scanning and high-resolution scanning focused on the subcortical nuclei. The anatomic MRI images included a 3D T1-weighted w hole brain scan and 2-dimensional (2-D) proton density weighted and T2 weighted MRI images. For the 2-D anatomic scan, a turbo spin echo (TSE) pulse sequence was used, which produces a protein density weighted image and T2 weighted image simu ltaneously. The slice thickness was 3.8mm, with a matrix of 256x256 and a total of 36 transa xial slices to cover the whole brain. For the 3-D scan, a 3D MPRAGE pulse sequence was used to collect T1 weighted 3-D anatomic MRI images. The matrix was 256x256, slice thickness was1.0~1.3mm, and 160 axial slices were used to cover the whole brain. The s cans were interpreted by two independent reviewers, one of whom is a boa rd-certified neuroradio logist and the other has published extensively in the field of neur oradiology. The reviewers were blinded to

PAGE 19

10 the identity and diagnosis of the subjects. Although 43 subjects underwent both the cognitive/behavioral testing and an MRI, so me subjects only cons ented to participate only in cognitive/behavioral testing (n = 17) or MRI scanning (n = 13). Table 2-1: Characteristics of subjects undergoing MRI evaluation PWS EMO Controls N 17 18 21 Age (yrs) .9-39 (mean = 11.4) 4-22 (mean = 10.9) 3.5-43 (mean = 19.5) Gender 11 M, 6 F 9 M, 9 F 8 M, 13 F Race 1 Hispanic, 16 White 2 Hispanic, 16 White 21 White M= Males F=Females Table 2-2: Characteristics of subjects under going cognitive, achievement and behavioral evaluation PWS EMO Controls N 19 17 24 Age (yrs) 4-39 (mean = 24.6) 4-22 (mean = 13.4) 4-43 (mean = 21.0) Gender 13 M, 6 F 8 M, 9 F 10 M, 14 F Race 2 Hispanic, 17 White 3 Hispan ic, 2 Black, 12 White 24 White M = males F = females Figure 2-1. Percentage of id eal body weight at time of evaluation. EMO patients were significantly heavier at the time of evalua tion than either PWS patients or than controls. The white area encompasse s weights >150% of ideal body weight (IBW). The boxplots graph data as a box representing statistical values. The boundary of the box closest to zero indicates the 25th percentile, a line within the box marks the median, the cross within the box indicates the mean, and the boundary of the box furthest from zero indicates the 75th percentile. The whiskers above and below the box indicate the 90th and 10th percentiles.

PAGE 20

11 CHAPTER 3 COGNITIVE AND BEHAVIORAL FINDINGS Cognitive Results The three way ANOVA comparing the Gene ral Intellectual Ability (GIA) of the three groups, adjusting for gender, was signi ficant at P<0.001. EMO subjects were lower than the control siblings (p<0.001), with average GIA for EMO subjects being 77.4 .8 and the average GIA for sibling controls be ing 106.4 .9 (Figure 3-1). The mean GIA for the PWS group was 63.3 .2, which was lower than controls (p <0.001). The PWS group had a significantly lower mean than the EMO group (P=.020). Additionally, the EMO subjects had lower thinking ability (79.4 .9 vs 109.3 12.3), cognitive efficiency (79.3 .6 vs 103.2 .0), phonemic awareness (90.1 .3 vs 114.7 .5), and working memory (85.1 .2 vs 106.9 .2 ) than controls (p<0.001 for all measures). Achievement EMO subjects had a total achievement score (78.7 18.8) on the WJ-III that was similar to the PWS (71.2 17.0) group (p= 0.33), but lower than the controls (104.8 17.0; p<0.001) (Figure 3-2). All broad area s of achievement measured by the WJ-III (math, reading, and written language) were sign ificantly lower in EMO subjects than in the controls (p =0.017, <0.001, and 0.0014, respectively). Behavioral Findings The three groups differed significantly on th e standardized adaptive skills score, adjusted for gender (p=0.0025). The EMO subjects had inferior adaptive skills (38.5

PAGE 21

12 9.3) than the control group (50.8 10.5, p= 0.0020) (Figure 3-3). Individuals with EMO had a slightly higher (59.6 15.7) score on th e Behavioral Symptoms Index (BSI) than controls (51.6 12.9), although this differe nce was not significant (p=0.14). Subjects with EMO did not have signifi cantly worse externalizing behaviors (e.g., aggression and hyperactivity) than either cont rols (p = 0.10) or PWS subjects (p= 0.17). However, the EMO subjects had more inte rnalizing behaviors (including anxiety, depression, and somatization) than either the PWS (p=0.020) or the control group (p= 0.059) (Figure 3-4). Socioeconomic Status Although the PWS parents had a higher pa rental education (p = 0.02) and a higher annual income (p= 0.054) than the EMO pare nts, pooling the siblings from each group reduced any SES difference between the cont rols and either the PWS or EMO groups. Sub-analysis of the control group showed th at there were no significant differences in GIA (p= 0.67) or achievement (p= 0.60) be tween the EMO siblings and the PWS siblings, suggesting that SES differences were probably not responsible for the disparity in GIA and achievement scores betw een the EMO patients and controls.

PAGE 22

13 Figure 3-1. Comparison of Gene ral Intellectual Abilities (GIA). EMO patients have a significantly lower GIA than controls (p <0.001; c ). PWS patients also have a lower GIA than controls (p <0.001; b). PWS patients have a lower GIA than EMO patients (p=0.020; a). The shaded gray area (GIA = 85-115) represents the normal range for GIA, while the white area below 85 represents the borderline and mentally retarded ra nge, and the white area above 115 represents the above average range. Figure 3-2. Comparison of total achievement scores. EMO individuals have lower total achievement than controls (p = 0.001), but are not significantly different than PWS patients. PWS patients also had achievement scores that were lower than controls (p=0.001). The shaded gray area (achievement score 85-115) represents the normal range.

PAGE 23

14 Figure 3-3. Comparison of adaptiv e skills and problem behaviors. EMO patients had worse adaptive skills than controls (p = 0.002; (a) on the boxplot). EMO patients did not have significantly worse problem behaviors than controls (p=0.14), nor did PWS patients (p=0.86). Plain gray shading is clinic ally significant and the stippled gray area is “at risk”. Figure 3-4. Externalizing and internalizi ng behaviors as assessed by Parent BASC evaluation. EMO patients had more internaliz ing problem behaviors than the PWS group (p=0.020; (a) on boxplot) and controls (p=0.059; (b) on boxplot).

PAGE 24

15 CHAPTER 4 MRI RESULTS All of the individuals with PWS had mild ventriculomegaly. Abnormalities were also seen in the parietal lobe, the Sylvian fi ssure, and the insula in a subset of subjects with PWS. These abnormalities were not seen in any of the EMO subjects or sibling controls (Table 4-1). White matter lesions were noted in subjects with PWS over age 18 years of age and in EMO subjects over 11 years of age, but in none of the sibling controls. White Matter Lesions White matter lesions were seen in 6 of 8 subjects with PWS over age 18 (Figure 41a), but in none of the younger subjects with PW S or in any of the control subjects. Five of the 18 EMO subjects had white matter lesion s noted on brain MRI (Figure 4-1b). All 5 with white matter lesions were over age 11 years of age at the time of the study (age range 12 years to 22 years). Only one of th e six EMO subjects over 11 years of age did not have any white matter lesions.No relati onship was found between the presence of white matter lesions and waist to hip circumfe rence, duration of obesity, fasting insulin, glucose, or triglyceride levels (Table 4-2), or with fasting cholesterol, uric acid, and postprandial insulin, glucose, or triglyceride le vels. Patients with PWS had an average of 3 white matter lesions on MRI (range 1-7), while the patients with EMO had an average of 2 lesions (range 1-5). More white matter lesi ons tended to appear w ith a greater degree of obesity (p=.09) in the EM O subjects (Table 4-2). Accurate growth records,

PAGE 25

16 particularly from early childhood, were not ava ilable for many of the older patients with PWS to determine their highest weight or duration of morbid obesity. Several of the EMO and PWS subjects had previous head MRIs at younger ages which did not show any evidence of white ma tter lesions, demonstrating that the lesions appear with increasing age. These white ma tter lesions, which may be due to ischemia, vasculitis, or demyelination, were diffuse and found in different area s of the brain from subject to subject, however were consistently seen in the orbitofrontal cortex in both groups of patients Parietal-Occipital Lobe Abnormalities Subjects with PWS showed decreased vol ume of gray matter in the parietaloccipital lobes resulting in broadening of the local sulci (Figure 4-2). Six of the seventeen PWS patients had bila teral parietal-occipital lobe abnormalities, while 4 had isolated right-sided findings. No obvious relationship was seen between the laterality of parietal-occipital lobe abnormalities and molecula r class of subjects with PWS. No such abnormalities were found in EMO patients, normal weight controls, or in any of the five PWS patients younger than 3 years of age. The parietal-occip ital lobe, which is necessary for somatosensory awareness and in tegration, has been found to be abnormal in individuals with obsessive-compulsive behavi ors (21), which is a common feature of individuals with PWS. Sylvian Fissure Polymicrogyria Sylvian fissure polymicrogyria is associated with language disorders (22). The left Sylvian fissure allows integration of speech and language while the right Sylvian fissure is more involved in speech articulation (23, 24). We found that 13 of 17 subjects with PWS had Sylvian fissure polymicrogyria. Nine subjects with PWS ha d bilateral Sylvian

PAGE 26

17 fissure abnormalities, all of whom had a deletion of 15q11-13. One subject with a deletion had an isolated right Sylvian fissu re abnormality. Normal Sylvian cortical development was only seen in 2 subjects wi th a deletion and 2 subjects with UPD, although all 5 of our UPD PWS subjects had a normal right Sylvian fissure. PWS individuals with UPD have fewe r problems with articulation th an patients with a deletion of chromosome 15 (25), which fits with our observation that 5/5 UPD PWS subjects had a normal right Sylvian fissure as opposed to only 2/12 deletion subjec ts. All of the EMO and control subjects had norma l Sylvian fissure morphology. Insula Closure The portion of the cortex located deep in th e Sylvian fissure is the insula. Most of the PWS subjects in our study showed failure of the insula to close completely (Figure 43). Eight subjects had bilatera l insular abnormalities, while 5 had unilateral failure of the insula to close (3 right-sides, 2 left-sided ). Thus far, we have not observed any relationship between failure of the insula to close completely and c linical features (e.g. degree of obesity, appetite, etc) or the molecular class of the PWS subjects. We also found no relationship between the age of the patient and abnormalities in either the Sylvian fissure or the insula. None of the subjects with EMO or controls had insular abnormalities. Interestingly, pain perception and autonomic control are functions ascribed to the insula (26), and these are frequently abnormal in PWS (18).

PAGE 27

18 Table 4-1: MRI findings Diagnosis VentriculomegalyParietaloccipital abnormalities Sylvian fissure polymicrogyria Failure of insula closure White matter lesions PWS 17/17 10/17 13/17 13/17 6/17 EMO 0/18 0/18 0/18 0/18 5/18 Controls 0/21 0/21 0/21 0/21 0/21 Table 4-2: Relationship between presence of white matter lesions and clinical phenotype Patients BMI SDS Waist:hip Ratio Fasting insulin mcIU/mL Fasting glucose mg/dL Fasting triglycerides mg/dL Duration of obesity (mos) + WM lesions 7.29.0 0.920.008 16.522.6 82.9 14.6 155.7119.5 151.829.5 WM lesions 4.06.0 0.9910.00113.758.79 92.4 23.6 115.858 111.938.2 EMO t test p=0.09 p=0.30 p= 0.76 p=0.32 p=0.43 p=0.63 + WM lesions 1.91.2 0.95 0.007 4.173.49 82.8 18.9 97.554.1 Data not available WM lesions 3.14.0 0.95 0.007 11.77.6 93.5 25.7 103.3 52.2 Data not available PWS t test p=0.16 p=0.91 p=0.052 p=0.43 p=0.85 N/A All data is expressed as mean standard deviation. BMI: Body mass index; SDS: Standard deviation score Figure 4-1. White matter lesions in subjects w ith early-onset morbid obesity. (a) Marked white matter lesions, including a lacuna r infarct (arrow) in a 33 year old patient with PWS with a history of lo ng-standing obesity. (b) Multiple white matter lesions in a 15 year old EMO patie nt with morbid obesity since age 9 months.

PAGE 28

19 ab ab Figure 4-2. Abnormal parietal sulci. The control subject (a ) has normal parietal lobes, while the PWS patient (b) has decreased gray matter in the parietal lobes, resulting in increased sulci size. INSabc INSabc Figure 4-3. Abnormal closure of insula. Th e control subject (a) has a normal insula, while the PWS subject (b and c) has an insula that has failed to close completely.

PAGE 29

20 CHAPTER 5 CONCLUSIONS We have shown that early-onset obesity is associated with lower general intellectual ability, lower achievement, increased incidence of behavi oral issues, and the occurrence of white matter lesions. The lower cognitive function and behavioral abnormalities are independent of socioeconomic status. These findings are even more striking when the EMO and the PWS groups are compared. PWS is a syndrome that is well-known to result in mental retardation a nd behavioral problems. Nonetheless, the EMO group had similar total achievement scores and behavioral scores that were similar or worse than subjects with PWS. Although the PWS patients had a later age of onset of obesity than the EMO patients, PWS is a con tiguous gene syndrome that results in the loss of expression of several imprinted genes on chromosome 15 which are normally expressed in the brain (7), which would explain the more severe total cognitive impairment seen in the PWS patients. Our cognitive and behavioral findings for the PWS group are similar to the results found by other groups (10-13) We hypothesize that the lower cognition and achievement and increased behavioral problems in the non-PWS individuals with early -onset obesity were due to the abnormal hormonal and metabolic milieu that occurs with obesity. Abnormal accumulation of certain metabolites, such as phenylalanine or galactose, either in utero or in early childhood, can result in neurological impairme nt and mental retardation which becomes obvious when the child is older (27-29). Th ese substances, which are naturally found in the body, can result in devastating neurological impairment if they are present in excess

PAGE 30

21 during early development. Elevated levels of these metabolites and their by-products may cause delayed or absent myelination of ne urons in the brain (30) . Mental retardation is a known consequence of unt reated children with phenylke tonuria and galactosemia. Poorly controlled phenylketonuria has been show n to result in white matter lesions in the brain, which are thought to be associated with learning disabili ties and behavioral problems (30). Therefore, we postulate that a lterations in levels of adipokines (such as leptin, adiponectin, or TNF), hormones (such as insulin), or neurotransmitters (such as neuropeptide-Y) could cause damage to the developing brain if pr esent early in life before myelination of the brain is complete, similar to what is seen in phenylketonuria and other inborn erro rs of metabolism. Excess adipose tissue leads to an abnormal hormonal milieu including dyslipidemia, most often hypertriglyceridemia , as well as both leptin and insulin resistance, with subsequent hyperleptinemia a nd hyperinsulinemia (9). Hyperinsulinemia has been related to cognitive impairment and to the development of Alzheimer’s disease (31). Recent studies in mice show that hypertri glyceridemia prevents the entry of leptin to the brain, causing elevated peri pheral leptin levels (32). Leptin has been shown to be a neuroprotective hormone, so resistance to le ptin at the blood-brain barrier may be harmful to brain development (33). Studies in leptin resistant mice show white matter cysts and degenerative lesions on MRI (33). Previous studies in PWS have noted an a bnormal cerebral gyral fo lding pattern (4), abnormal Sylvian fissure morphology (3), and pituitary gland abnormalities (5, 6), but these studies have studied fewer than 5 subjec ts. Our study is the la rgest to date and is the first to identify parietal-occipital lobe abnormalities and insu la aberrations in

PAGE 31

22 PWS.We are also the first to identify white matter lesions in individuals with PWS and EMO. The white matter lesions, as well as the other morphological brain abnormalities, were not seen in any of the sibling cont rols, none of whom had a history of childhood obesity. The two PWS subjects over the age of 16 who did not have white matter lesions had both been receiving growth hor mone treatment during childhood, and had had long-term control of their obesity (150% and 175% IBW), with no significant fluctuations in weight over time. White matter lesions were seen in EMO patients as young as 12 years of age. Therefore, we postula te that the white matter lesions are due to the duration and/or severity of obesity dur ing early childhood. We will have to follow our cohort of young patients longitudinally to identify the pathogenesis and progression of the white matter lesions, however, we hypot hesize that obesity-der ived perturbations in adipokines, hormones, or neurotransmitters may be responsible for white matter lesions, as well as the cognitive, and/or behavioral impairment. Additional evidence for the hypothesis that early-onset obesity may contribute to decreased cognitive function comes from patients with PWS. Anecdotally, patients whose weight was well cont rolled in childhood seem to do better cognitively than patients with PWS whose weight was poorly controlled during early childhood. One study has shown that patients with PWS who were diagnosed in infancy before becoming obese, and who adhered to a dietary plan that prevented the early onset of obesity, had a statistically significant mean IQ score that wa s 20 points higher than those patients with PWS who were obese early in childhood (34). The EMO patients we studied were a heterogeneous group with no evidence of any known syndromal cause of obesity. Extens ive genetic testing for known causes of

PAGE 32

23 cognitive impairment (e.g., chromosomal, PWS, and Fragile X syndrome, as well as array CGH) and obesity (MC4R and leptin mutations ) were excluded. Th e common factor in this group was the development of early-onset morbid obesity. While we suspect that almost all of our EMO patients will eventually be found to have a genetic explanation for their early-onset obesity we feel that most of these patients will have a single gene defect causing the obesity with the c ognitive impairment and behavi oral problems occurring as secondary effects. Sleep apnea, which is common in individua ls with obesity, has negative effects on cognition and behavior, most commonly causing executive dysfunctioning, impaired working memory, and anxiety/hyperactivity in children (35). A magnetic resonance spectroscopy study in adults with a mean age of 48 year s showed that absolute concentrations of N-acetylaspartate and cholin e were significantly reduced in the frontal white matter of patients with sleep apnea, lead ing the authors to conclude that the sleep apnea may be contributing to the development of cognitive executive deficits (36). We have found that all patients with PWS have some degree of sleep apnea beginning in infancy , which is usually mild and has b een treated as warranted (37). We are conducting further studies to dete rmine the role of the sleep apnea in our subjects with PWS and EMO in the development of white ma tter lesions and cogni tive and behavioral issues, but at the current time we feel that sl eep apnea is not playing a major role in most of our patients. Our study is the first to find a difference in GIA and behavior irrespective of SES between individuals with EMO and contro ls who were at normal weight during childhood. A longitudinal study of individuals age 65-85 years with obesity showed a

PAGE 33

24 correlation between high waist-to-hip ratio and memory decline, but no relationship between BMI and cognitive function (38). Cross-sectional studies of adult males showed that intelligence test scores and educational level are invers ely related to obesity (39, 40). Several Danish studies evaluating intellect ual performance in males who underwent an obligatory draft board examination concluded th at severe obesity was strongly associated with diminished intellectua l performance (39-42). Additionally, a study of 102 children found that those who were severely obese (> 50% overweight) at an average age of 9.8 years had a significantly lower performance IQ than control children (43). However, the authors of all of these studies concluded th at they were unable to distinguish whether reduced intellectual performance can be ascribed solely to obe sity or if it is due to the association of obesity with lower socioeconomic status a nd lower educational level. Indeed, one recent study found that the differenc es in test scores between overweight and normal weight children could be explained by differences in socioeconomic variables (44). However, none of the previous studi es exclusively analyzed a group of subjects whose obesity began as early or who we re as severely obese as our EMO group. Another consideration is that none of the IQ tests currently in use, including the popular Wechsler scales for children (W ISC-III, 1991) and adults (WAIS-R, 1981), measures all the broad cognitive abilities (45) with the exception of the WoodcockJohnson tests (19). The WJ-III, which was us ed in this study, was designed to sample various cognitive abilities and academic achievement for individuals ranging in developmental levels from 2 years to 79 y ears. The WJ-III is the most comprehensive test battery available for the clinical assessm ent of children and youth. In contrast to the WJ-III, the Wechsler IQ scores, which have typically been more widely used for IQ

PAGE 34

25 testing, measure only a narrow range of cognitiv e abilities. Furthermore, the Wechsler IQ scores cannot be compared across groups w ith a wide age span because they consist of three different tests (one for preschool ch ildren, one for school-aged children and youth, and one for adults) and their scores are de rived from different samples. All of the previous studies evaluating the relationship between cognitive function and obesity used either the WISC-III or the WAIS-R, neither of which is as comprehensive as the WJ-III, which was used in our study. One limitation of our study is the small sa mple size with few EMO/sibling matched pairs. However, the GIA of our control sibli ngs fell well within th e standardized normal range, and we saw a highly si gnificant difference between EMO subjects and normal controls despite the small sample size. In the 4 EMO/sibling pairs studied thus far, we have found a difference in GIA greater than 2 standard deviations between the EMO patient and the sibling control. Sibling matching is a strength of our study because it allows us to reduce the effects of genetic and socioeconomic factors as a cause of the difference in GIA. Additional EMO/sibling pa irs will be recruited to extend this study. Another limitation is the cross-sectiona l nature of this study. We do not know whether the lower cognitive function, behavi oral problems, and white matter lesions are due to the onset, duration, or the degree of obesity, and/or to the hormonal milieu associated with obesity. Likewise, we are not certain whether the white matter lesions are related to the findings of cognitive dysfunction. These questions will be addressed in future longitudinal studies with annual MRI of the brain and asse ssments of cognition, achievement, and behavior. Further studies are also needed to correlate the GIA,

PAGE 35

26 achievement, and behavioral findings with specific obesity-related metabolic and hormonal derangements. In conclusion, we have found a strong co rrelation between early-onset obesity and cognitive impairment, achievement, and behavi oral problems, as well as the presence of white matter lesions. This relationship is in dependent of any of the known causes of poor cognitive function and achievement, such as genetic abnormalities, low socioeconomic status, or other environmental factors. Therefore, early childhood obesity alone may compromise cognitive ability and achievem ent, adding to the public health concern surrounding the epidemic of obesity in chil dhood and further emphasizing the need for early intervention.

PAGE 36

27 LIST OF REFERENCES 1 du Toit G, van der Merwe MT. The epidemic of childhood obesity. S Afr Med J 2003;93(1):49-50. 2. Hayashi M, Itoh M, Kabasawa Y, Ha yashi H, Satoh J, Morimatsu Y. A neuropathological study of a case of the Prad er-Willi syndrome with an interstitial deletion of the proximal long arm of chromosome 15. Brain Dev 1992;14(1):58-62. 3 Leonard CM, Williams CA, Nicholls RD, Agee OF , Voeller KK , Honeyman JC , Staab EV . Angelman and Prader-Willi syndrome: a magnetic resonance imaging study of differences in cerebral structure. Am J Med Genet 1993;46(1):26-33. 4 Yoshii A, Krishnamoorthy KS, Grant PE . Abnormal cortical development shown by 3D MRI in Prader-Willi syndrome. Neurology 2002;59(4):644-5. 5 Miller L, Angulo M, Price D, Taneja S. MR of the pituitary in patients with PraderWilli syndrome: size determination and imaging findings. Pediatr Radiol 1996;26(1):43-7. 6 Cacciari E, Zucchini S, Carla G, Pirazzoli P, Cicognani A , Mandini M , Busacca M , Trevisan C . Endocrine function and morphologi cal findings in patients with disorders of the hypothalamo-pituitary area : a study with magnetic resonance. Arch Dis Child 1990;65(11):1199-202. 7 Glenn CG, Driscoll DJ, Thomas PY, Nic holls RD. Genomic imprinting: potential function and mechanisms revealed by the Prader-Willi and Angelman syndromes. Mol Hum Reprod 1997;3:321-32. 8 Nicholls RD, Knepper JL. Genome or ganization, function and imprinting in Prader-Willi and Angelman syndromes. Annu Rev Genomics Hum Genet. 2001;2:153-75. 9 Tounian P, Aggoun Y, Dubern B, Varille V, Guy-Grand B, Sidi D, Girardet JP, Bonnet D. Presence of increased stiffn ess of the common carotid artery and endothelial dysfunction in severely obese children: a prospective study. Lancet 2001 358:1400-4. 10 Dykens EM, Cassidy SB. Correlates of malada ptive behavior in children and adults with Prader-Willi syndrome. Am J Med Genet 1995;60(6):546-9.

PAGE 37

28 11 Holland AJ, Whittington JE, Butler J, Webb T, Boer H, Clarke D. Behavioural phenotypes associated with specific geneti c disorders: evidence from a populationbased study of people with Prader-Willi syndrome. Psychol Med 2003;33(1):14153. 12 Dimitropoulos A, Feurer ID, Butler MG, Thompson T. Emergence of compulsive behavior and tantrums in children with Prader-Willi syndrome. Am J Ment Retard 2001;106(1):39-51. 13 Whitman BY, Accardo P. Emotional symptoms in Prader-Willi syndrome adolescents. Am J Med Genet 1987;28(4):897-905. 14 Farooqi IS, Yeo GS, Keogh JM, Aminian S, Jebb SA, Butler G, Cheetham T , O'Rahilly S . Dominant and recessive inheritance of morbid obesity associated with melanocortin 4 receptor deficiency. J Clin Invest 2000;106(2):271-9. 15 Pergolizzi RG, Erster SH, Goonewardena P, Brown WT. Detection of full fragile X mutation. Lancet 1992;339(8788):271-2. 16 Yu W, Ballif BC, Kashork CD, Heilstedt HA , Howard LA , Cai WW , White LD , Liu W , Beaudet AL , Bejjani BA , Shaw CA , Shaffer LG . Development of a comparitive genomic hybridization microarray and demonstration of its utility with 25 well-characterized 1p36 deletions. Hum Mol Genet 2003;12:145-52. 17 Sahoo T, Shaw CA, Young AS, Whitehouse NL , Schroer RJ , Stevenson RE , Beaudet AL . Array-based comparitive genomic hybr idization analysis of recurrent chromosome 15q rearrangements. Am J Med Genet 2005; 139(2): 106-13. 18 Goldstone AP. Prader-Willi syndrome: a dvances in genetics, pathophysiology and treatment. Trends Endocrinol Metab 2004;15(1):12-20. 19 Woodcock RW, McGrew, K.S., and Mather, N. Woodcock-Johnson PsychoEducational Battery III. Itasca, IL : Riverside Publishing Company; 2001. 20 Doyle A, Ostrander R, Skare S, Crosby RD, August GJ. Convergent and criterionrelated validity of the Be havior Assessment System for Children-Parent Rating Scale. J Clin Child Psychol 1997;26(3):276-84. 21 Anderson KE, Savage CR. Cognitive and neurobiological findings in obsessivecompulsive disorder. Psychiatr Clin North Am 2004;27(1):37-47. 22 Guerreiro MM, Hage SR, Guimaraes CA, et al. Developmental language disorder associated with polymicrogyria. Neurology 2002;59(2):245-50. 23 Hickok G, Buchsbaum B, Humphries C, Mu ftuler T. Auditory-motor interaction revealed by fMRI: speech, music, and working memory in area Spt. J Cogn Neurosci 2003;15(5):673-82.

PAGE 38

29 24 Demonet JF, Price C, Wise R, Frackowia k RS. Differential activation of right and left posterior sylvian regions by sema ntic and phonological tasks: a positronemission tomography study in normal human subjects. Neurosci Lett 1994;182(1):25-8. 25 Cassidy SB, Forsythe M, Heeger S, et al. Comparison of phenotype between patients with Prader-Willi syndrome due to deletion 15q and uniparental disomy 15. Am J Med Genet 1997;68(4):433-40. 26 George MS, Nahas Z, Bohning DE, et al . Vagus nerve stimulation therapy: a research update. Neurology 2002;59(6 Suppl 4):S56-61. 27 Waisbren SE, Azen C. Cognitive and behavioral development in maternal phenylketonuria offspring. Pediatrics 2003;112(6 Pt 2):1544-7. 28 Antshel KM, Waisbren SE. Timing is ev erything: executive functions in children exposed to elevated levels of phenylalanine. Neuropsychology 2003;17(3):458-68. 29 Manis FR, Cohn LB, McBride-Chang C, Wolff JA, Kaufman FR. A longitudinal study of cognitive functioning in patients with classical gal actosaemia, including a cohort treated with oral uridine. J Inherit Metab Dis 1997;20(4):549-55. 30 Huttenlocher PR. The neuropathology of phenylketonuria: human and animal studies. Eur J Pediatr 2000;159 Suppl 2:S102-6. 31 Strachan MW. Insulin and cognitive function. Lancet 2003;362(9392):1253. 32 Banks WA, Coon AB, Robinson SM, Moinuddi n A, Shultz JM, Nakaoke R, et al. Triglycerides induce leptin resistance at the blood-brain barrier. Diabetes 2004;53(5):1253-60. 33 Dicou E, Attoub S, Gressens P. Neuroprot ective effects of leptin in vivo and in vitro. Neuroreport 2001;12(18):3947-51. 34 Crnic KA, Sulzbacher S, Snow J, Ho lm VA. Preventing mental retardation associated with gross obesity in the Prader-Willi syndrome. Pediatrics 1980;66(5):787-9. 35 El-Ad B, Lavie P. Effect of sleep apnea on cognition and mood. Int Rev Psych 2005;17(4):277-82. 36 Alachanatis M, Deligiorgis N, Zias N, Amfilochiou A , Gotsis E , Karakatsani A , Papadimitriou A . Frontal brain lobe impairment in obstructive sleep apnoea: a proton MR spectroscopy study. Eur Respir J 2004;24(6):980-6. 37 Miller J, Silverstein J, Shuster, J, Driscoll DJ, Wagner M. Short-term effects of growth hormone on sleep abnormalities in Prader-Willi syndrome. JCEM 2005; 1279-1282.

PAGE 39

30 38 Jagust W , Harvey D , Mungas D , Haan M . Central obesity and the aging brain. Arch Neurol 2005 Oct;62(10):1545-8. 39 Halkjaer J, Holst C, Sorensen TI. Intell igence test score and educational level in relation to BMI changes and obesity. Obes Res 2003;11(10):1238-45. 40 Halkjaer J, Sorensen TI. Psychosocial and demographic determinants of regional differences in the prevalence of obesity. J Biosoc Sci 2004;36(2):141-52. 41 Sorensen TI, Sonne-Holm S, Christensen U, Kreiner S. Reduced intellectual performance in extreme overweight. Hum Biol 1982;54(4):765-75. 42 Sorensen TI, Sonne-Holm S. Intelligence te st performance in obesity in relation to educational attainment and parental social class. J Biosoc Sci 1985;17(4):379-87. 43 Li X. A study of intelligence and persona lity in children with simple obesity. Int J Obes Relat Metab Disord 1995;19(5):355-7. 44 Datar A, Sturm R, Magnabosco JL . Childhood overweight and academic performance: national study of ki ndergartners and first-graders. Obes Res 2004;12(1):58-68. 45 Carroll JB. Human cognitive abilities: A survey of factor-analytic studies. New York: Cambridge University Press; 1993.

PAGE 40

31 BIOGRAPHICAL SKETCH I received my Bachelor of Science and Master of Science degrees concurrently from Emory University. My thesis projec t was investigating the effects of various chemotherapeutic agents on DNA. After comple ting these degrees I decided to go on to medical school. Initially, in medical school I continued basic scien ce research in a lab studying retinitis pigmentosa in Xenopus frogs, but then decided to enter the arena of clinical research. During my residency in the Department of Pe diatrics at the University of Florida I began working on the HIV 076 Trial interviewing the patients and developing the clinical database which was then used to track data from the trial for many years. However, during my second year of residency I d ecided to pursue a career in pediatric endocrinology. I therefore began to focus my interest on pedi atric obesity research. My current thesis project was built on the foundatio n of investigating the causes and effects of early-onset morbid obesity. I completed my board certification in pediatrics in 2001 and my board certification in pediatric endocrinology in 2005. I am currently an Assistant Professor in Pediatric Endocrinology doing clinical research investigating the causes and effects of early-onset obesity.