This item is only available as the following downloads:
1 EXAMINING THE CONTRIBUTION OF HEALTH BEHAVIORS AND PSYCHOSOCIAL FUNCTIONING IN ANTI EPILEPTIC DRUG INDUCED WEIGHT GAIN AMONG CHILDREN WITH EPILEPSY By KATHERI NE WELLS FOLLANSBEE JUNGER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Katherine Wells Follansbee Junger
3 To my family
4 ACKNOWLEDGMENTS I thank my husband and family for their love and support throughout this process. I am al so grateful to my mentor for the excellent training he provided and his ongoing guidance.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 Anti Epileptic Medication and Weight Status ................................ .......................... 17 Valproate ................................ ................................ ................................ .......... 17 Carbamazepine ................................ ................................ ................................ 18 Vigabatrin ................................ ................................ ................................ ......... 18 Gabapentin ................................ ................................ ................................ ....... 19 Health Behaviors and Growth Velocity ................................ ................................ .... 20 Psychosocial Functioning and Epilepsy ................................ ................................ .. 23 Primary Aims and Hypotheses ................................ ................................ ................ 25 Aim 1 : To D escribe Weight Status in Youth with Epilepsy Who Are Prescribed AEDs. ................................ ................................ .......................... 25 Aim 2 : To Examine the Relationship Between Behavioral Health Factors (E.G., Caloric Intake, Energy Expenditure) and Weight Status in Youth on AEDs. ................................ ................................ ................................ ............ 26 Aim 3: To Assess the Relationship Between Weight Status and Psychosocial Functioning in Terms of Depressive Symptoms and Quality of Life in Youth with Epilepsy ................................ ................................ ........ 26 Exploratory Analyses ................................ ................................ .............................. 27 2 METHOD ................................ ................................ ................................ ................ 29 Participants ................................ ................................ ................................ ............. 29 Procedure ................................ ................................ ................................ ............... 29 Measures ................................ ................................ ................................ ................ 30 Questionnaires ................................ ................................ ................................ 30 Anthropometrics ................................ ................................ ............................... 33 Medical Records/Chart Review ................................ ................................ ........ 33 Statistical Analyses ................................ ................................ ................................ 33 Sample Size ................................ ................................ ................................ ..... 33 Preliminary Data Analyses ................................ ................................ ............... 34 Pr imary Analyses ................................ ................................ ............................. 35 Exploratory Analyses ................................ ................................ ........................ 36
6 3 RESULTS ................................ ................................ ................................ ............... 37 Preliminary Analyses ................................ ................................ .............................. 37 Aim 1 Weight Status Between Groups ................................ ................................ .. 40 Aim 2 Behavioral Health Factors Between Groups ................................ ............... 4 1 Aim 3 Psychosocial Functioning Between Groups ................................ ............... 42 Exploratory Analyses ................................ ................................ .............................. 43 4 DISCUSSION ................................ ................................ ................................ ......... 56 Aim 1 AEDs and Weight Status ................................ ................................ ............. 56 Behavioral Health Factors ................................ ................................ ....................... 59 Dietary Intake ................................ ................................ ................................ ... 59 Physical Activity ................................ ................................ ................................ 62 Psychosoci al Functioning ................................ ................................ ....................... 64 Limitations ................................ ................................ ................................ ............... 67 Summary ................................ ................................ ................................ ................ 71 APPENDIX: MEASURES ................................ ................................ .............................. 72 Parent Measures ................................ ................................ ................................ .... 72 Chart Review and Administrative Forms ................................ ................................ 79 LIST OF REFERENCES ................................ ................................ ............................... 83 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 91
7 LIST OF TABLES Table page 1 1 Hypothesized Trajectory of Growth Acceleration and Mechanisms Underlying Accelerated Weight Gain for Weight Positive AEDs ......................... 28 3 1 Demographic Characteristics Across Participants by AED Category ................ 44 3 2 Gender, Minority Status, and Age for All Participants Entered into the Frequency Match Procedure ................................ ................................ .............. 45 3 3 Demographics, Epilepsy Type, Time Since Diagnosis, Seizure Frequency, Medications, and BMI Z Scores by AED Group for Matched Sample ................. 46 3 4 Behavioral Health Factors by AED Group for Matched Sample ........................ 48 3 5 Pearson Product Correlations Among Psychosocial and W eight Status Variables for Matched Sample ................................ ................................ ............ 49 3 6 Psychosocial Functioning by Group for Matched Sample ................................ 50
8 LIST OF FIGURES Figure page 3 1 Participant Flow Chart ................................ ................................ ........................ 51 3 2 Proposed Mediation of AEDs and Change in Weight Status by Average Caloric Intake ................................ ................................ ................................ ...... 52 3 3 Proposed Mediation of AED and Change in Weight Status by Physical Activity ................................ ................................ ................................ ................ 53 3 4 Proposed Media tion Model for AED and Depressive Symptoms by Change in Weight Status ................................ ................................ ................................ ..... 54 3 5 Proposed Mediation Model for AED and Quality of Life by Change in Weight Status ................................ ................................ ................................ ................. 55
9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EXAMINING THE CONTRIBUTION OF HEALTH BEHAVIORS AND PSYCHOSOCIAL FUNCTIONING IN ANTI EPILEPTIC DRUG INDUCED WEIGHT GAIN AMONG CHILDREN WITH EPILEPSY By Katherine Wells Follansbee Junger August 2012 Chair: David Janicke Major: Psychology Anti epileptic drugs (AEDs) are the first line of treatment in pediatric epilepsy and successfully prevent seizure recurrence in the majority of patients. However, several of these medications have been linked to increased growth velocity, which increases risk for a range of medical and psychosocial comorbidities. The pathophysiology of changing growth status is unknown, but theories include increased appetite, reduced metabolism, and increased fatigue. The purpose of the present study was to evaluate the r ole of dietary intake and physical activity, the behavioral proxies of appetite and fatigue, in weight status change among children taking AEDs. Depressive symptoms and quality of life were also assessed. Participants included 49 youth, ages 8 17, and thei r parent/legal guardian who were assessed at baseline and 4 6 month follow up. Children on weight positive AEDs were compared to those taking weight negative, neutral, or no AEDs. Due to group size imbalance, frequency matching across age, race, and gender was used to select 7 cases in each group (14 participants total) to comprise the final sample for analysis. No differences emerged in weight status, dietary intake, physical activity, or depressive symptoms between groups at baseline or over
10 time. Those t aking weight positive AEDs had significantly lower quality of life at baseline compared to those not on weight positive AEDs. Across all participants at baseline, there were non significant trends between higher weight status and higher levels of depressiv e symptoms, and higher weight status and better quality of life. The results of this study are consistent with previous research showing that children with epilepsy are at greater risk for being overweight/obese and having a lower quality of life than the general population. Further research with larger sample sizes targeting children at diagnosis is needed to elucidate the role of dietary intake and physical activity in AED induced weight gain.
11 CHAPTER 1 INTRODUCTION Epilepsy is the most common, chronic n eurological disorder of childhood, with a prevalence rate of nearly 1% in children under 16 years of age (Shinnar & Pellock, 2002). It is a heterogeneous disorder with numerous etiologies including developmental brain malformations, trauma, illness, vascul ar defects, and metabolic disorders (Leonard & George, 1999). The International League Against Epilepsy developed a nosology for epilepsy that classifies seizures on three levels: (1) site of epileptic discharge (partial/localized, generalized, or indeterm inate); (2) group of syndromes (idiopathic, symptomatic, or crytogenic); and (3) specific syndrome (Shinnar & Pellock, 2002). Idiopathic refers to seizures associated with genetic defects and age dependent epilepsies (i.e., benign rolanic and absence). Sym ptomatic includes seizures associated with brain injury, lesions, or static encephalopathy. Cryptogenic is defined as seizures occurring in the absence of identifiable pathology and with unknown etiology (Leppik, 1998). Community based studies have found t hat among those with childhood onset epilepsy, partial or localized seizures accounted for approximately 60% of seizures, followed by generalized (~30%) and undetermined (~10%) (Berg, Levy, Testa & Shinnar, 1999; Sillanpaa, Jalava, & Shinnar, 1999). Develo pmental disabilities such as mental retardation, cerebral palsy, and autism are much more common among children with epilepsy compared to children in the general population, with comorbidity rates between 30 40% (Sillanpaa, 1992). Epilepsy occurs at the hi ghest rate among children with severe mental retardation and when co occuring, seizure onset tends to be earlier and prognosis less favorable compared to children of normal intelligence or those with less severe intellectual deficits (Leonard & George, 199 9).
12 Regarding the course of illness, the median age of seizure onset in children is between 5 and 6 years, (Shinnar & Pellock, 2002) while the highest incidence rate occurs in children under 1 (Pellock, 2004). After having 1 seizure, the recurrence rate i s approximately 40% by 2 years (Berg & Shinnar, 1991). It has been reported that approximately 55 65% of children who experience a seizure eventually become seizure free ( Holland & Glauser, 2007; Sillanpaa, Jalava, Kaleva, & Shinnar 1998), with many achi eving remission within 2 years of beginning anti epileptic drugs (AEDs) ( Berg, Shinnar, Levy, Testa, Smith, et al., 2001 ). Idiopathic epilepsies (such as benign rolandic and absence epilepsies), which comprise around 30% of cases in community samples, have the best prognosis followed by cryptogenic then remote symptomatic etiologies (Leppik, 1998; Shinnar & Pellock, 2002). Berg and colleagues (2001) followed 613 children with newly diagnosed epilepsy and classified outcomes at 2 years as good (in remissio n for at least 1 year at the 2 year follow up), intractable (failed 2 or more AEDs and experienced 1 or more seizures per month for more than 18 months), and indeterminate (did not meet criteria for the other categories). Results showed that 52% of patient s were in remission at 2 years, while 8% were categorized as intractable, and 38% were judged to have an indeterminate outcome at that time. Importantly, children in the indeterminate group had higher rates of discontinuous treatment and episodes of non co mpliance with the medical regimen compared to those in the other 2 categories. Two year outcomes were highly predictive of status at 4 years with 83% and 87% of children maintaining their classifications as in remission or intractable, respectively. In add ition, 54% of children who were indeterminate at 2 years achieved remission by 4 years. Etiology was strongly
13 associated with remission status in that idiopathic epilepsies were the least likely and remote symptomatic etiologies the most likely to be intra ctable. No association was found between cryptogenic etiologies and remission status. Children with epilepsy onset before age 1 were at the greatest risk for intractability. In summary, AED medications effectively prevent seizure recurrence in the majority of children; therefore, identifying factors that may compromise treatment adherence is essential to ensure the best outcomes for children with epilepsy. While AEDs confer clear benefits in terms of seizure reduction and disease management among children with epilepsy, it is well established that several of these medications are associated with clinically significant weight change (Biton, 2003a,b; Grosso, Mostardini, Piccini, & Balestri, 2009). Indeed, weight gain is the most frequently reported adverse si de effect of the widely used medication, valproate (Biton, Mirza, Montouris, Vuong, Hammer, & Barrett, 2001; Stephen, Sills, Leach, Butler, Parker, Hitiris, et al., 2007), occurring in up to 70% of adult patients (Corman et al., 1997). In children weight g ain has been reported to be slightly less prevalent, approximately affecting 45 60% of those taking the medication (Egger & Brett, 1981; Wirrell, 2003). Clinically significant weight change has generally been defined as a 5% or 5kg change in weight from ba seline in adults (Biton, 2003b). In children, who are still growing, excessive weight has been measured by an increase in BMI z score (Jallon & Picard, 2001; Novak, Maytal, Alshansky, Eviatar, Sy Kho et al., 1999), which adjusts for expected growth based o n age and gender. AEDs have been categorized according to their typical effects on weight status including as being associated with weight gain (e.g., valproate, carbamazepine,
14 gabapentin, and vigabatrin), weight loss (topiramate, felbamate, and zonsimide ), or as having no effect on weight (phenytoin, lamotrigine, and levetiracetam) (Biton, 2003b; Jallon & Picard, 2001; Vanina, Podolskaya, Sedky, Shahab, Siddiqui, et al., 2002). respectively. The majority of studies have been conducted on valproate simply because it has been in use the longest but an increasing body of evidence is growing to evaluate other weight positive AEDs (Biton, 2003b). Several reviews that hav e summarized the effects of AED on weight status (Biton, 2003b; Jallon & Picard, 2001; Vanina et al., 2002), indicate that valproate is most strongly associated with weight gain. The accelerated growth velocity attributed to valproate typically occurs with in the first 3 months of AED treatment, plateaus by month 6, and remains stable for the duration of treatment (see Biton, 2003b; Zimmerman, Kraus, Himmerich, Schuldi & Pollmacher, 2003). Similar data is not available for other weight positive AEDs in part because they are newer and have been the focus of fewer investigations. Therefore, documenting the trajectory of accelerated growth for the other weight positive AEDs is an important step in this line of research. It is not only imperative to know which A EDs cause weight gain and the course, but also to be able to preemptively identify patients at risk for experiencing inappropriate weight gain as a side effect. Predictors of accelerated growth are poorly understood and findings across studies have been in consistent (Biton, 2003b). For instance, the relationship between baseline weight status and change in growth velocity subsequent to AED therapy has been mixed with some studies showing that weight gain tends to be most severe among patients who were norma l weight status before initiation of AEDs
15 (Corman et al., 1997; Jallon & Picard, 2001), and others finding the greatest increase in BMI z score among children who were the most overweight prior to intervention (Novak et al, 1999; Wirrell, 2003). Similar in consistencies have been documented for gender (e.g., Isojarvi, Laatikainen, Knip, Pakarinen, Juntunen, et al., 1996; Novak et al., 1999). Importantly, most studies investigating weight gain have relied on small samples, thus assessment of risk factors has been limited due to a lack of power (Biton, 2003b). Clinically significant weight gain as a side effect of AED treatment may reduce tolerance for the medical regimen and lead to noncompliance (Egger & Brett, 1981; Zimmerman et al., 2003), consequently imp eding disease management (Wirrell, 2003; Zimmerman et al., 2003), especially among adolescents who tend to have more body concerns (Biton, 2003b). Children and adolescents may elect to terminate the AED, with witch to another medication despite having a positive response in terms of seizure control to avoid or reverse weight gain (Wirrell, 2003). Besides the direct negative influence on physical health in children with epilepsy via treatment noncompliance, chi ldhood obesity increases the risk for a range of chronic health conditions including type II diabetes, metabolic syndrome, polycystic ovaries, and other risk factors for cardiovascular disease including hypertension and hypercholesterolemia (Fagot Campagna Pettitt, Engelgau, Burrows, Geiss, Valdez et al. 2000; Strauss, 1999; Weiss, Dzuira, Burgert, et al., 2004). Indeed, one study conducted among 20 prepubertal girls found that valproate induced obesity was linked to the development of insulin sensitivity after one year, which was unrelated to serum valproate or baseline weight status (Verotti, Basciani, De Simone, Trotta, Morgese, et
16 al., 2002). Another study among 16 women found that valproate induced obesity use was linked to the development of polycysti c ovaries, hyperadrogenism, abnormal lipid profiles, and hyperinsulinaemia, which resolved within a year after patients were switched to lamotrigine and lost weight (Isojarvi, Rattya, Myllyla, Knip, Koivunen, et al., 1998). What is especially troubling is that AED induced weight gain could not only exacerbate these complications among previously overweight children but produce new risk in children who move from normal to overweight status as a result of treatment. In addition, overweight children are more l ikely than non overweight youth to experience impaired psychosocial functioning, such as low self esteem (Erermis, Cetin, Tamar, Bukusoglu, Akdeniz, & Goksen, 2004; Pierce & Wardle, 1997), depression (Davison, & Birch, 2001; Must, 1996), stigmatization (Pu hl, & Latner, 2007) and negative body image (Pesa, Syre, & Jones, 2000). Children with epilepsy are already at much greater risk for suffering from psychological and behavioral disorders compared to healthy peers (Batzel, Dodrill, Dubinsky, Ziegler, Connol ly, Freeman, et al., 2007; Smith, M.L., Elliot, I.M., Lach, 2004), with psychiatric comorbidity rates between 30 50% (Salpekar & Dunn, 2007). Difficulties in this population include higher rates of academic problems, attention deficit/hyperactivity disorde r, depression, anxiety, peer victimization, bipolar disorder, impaired self esteem, lower self competence, poorer communication, lower quality of life, and more disruptive behavior disorders (see Leonard & George, 1999; Pellock, 2004; Salpekar & Dunn, 2007 ; Shinnar & Pellock, 2002 for review). Higher rates of psychosocial difficulties have also been documented among children with epilepsy compared to other chronic illness populations, including children with asthma, diabetes (reported in Salpekar & Dunn, 20 07), and juvenile rheumatoid arthritis
17 (Wirrell, Camfield, Camfield, Dooley, Gordon, et al., 1997). In light of these data, medications which contribute to overweight among children with epilepsy may compound psychosocial dysfunction which could act as bot h a barrier to compliance with epilepsy treatment as well as to adopting positive lifestyle behaviors associated with maintenance of a healthy body mass index. Anti Epileptic Medication and Weight Status Given that increased growth velocity resulting from AED treatment in children with epilepsy poses significant risks to both physical and mental health, delineating the mechanisms which support increased weigh status is imperative and is the first step to developing an intervention to reduce weight gain as a side effect of treatment and to mitigate the associated risks. While a substantial amount of research exists documenting the relationship between AEDs and weight status (Biton, 2003b; Jallon & Picard, 2001; Vanina et al., 2002), to date there is limited u nderstanding of the pathophysiological mechanisms to explain the propensity for accelerated growth velocity among weight positive AEDs. Several reviews (Biton, 2003b; Jallon and Picard, 2001; Zimmerman et al., 2003) are available that summarize what is cur rently known about the trajectory of AED induced weight gain and hypothesized mechanisms. An overview of these findings by medica tion will be provided below ( Table 1 1 ). Valproate Several different mechanisms have been offered to explain valproate induced weight gain including increased appetite and thirst, increased consumption of carbohydrates, increased secretion of insulin and proinsulin, decreased energy expenditure, reduced basal energy turnover, and decreased leptin (see Biton, 2003b; Jallon & Picar d, 2001; Zimmerman et al., 2003). In a study conducted among 100
1 8 children and adolescents, 38% of children who had been typically developing reached the 98%ile for weight gain velocity following initiation of valproate, which was explained by an increase i n appetite (Egger & Brett, 1981). In the same study, lethargy was reported among 9% of youth initiating valproate. However, among the patients who suffered from lethargy, decreased appetite and weight loss were also reported, suggesting that reduction in p hysical activity is an unlikely explanation for weight gain, at least among children. Importantly, the data in this study was gathered through a retrospective chart review without validated measures of dietary intake and physical activity. It appears that increased appetite was merely documented by anecdotal parent and child report. Carbamazepine Little research has investigated the cause of weight gain associated with carbamazepine treatment; however, proposed mechanisms include edema due to increased sec retion of antidiuretic hormones, and increased appetite and caloric intake (see Biton, 2003b; Jallon & Picard, 2001). Very little research has been conducted among children and trajectory of growth velocity change has not been documented. Vigabatrin Cont rolled clinical trials examining the tolerability and efficacy of vigabatrin have suggested a dose dependent relationship between the medication and weight gain reported as a side effect (see Biton, 2003b). An explanatory mechanism has not yet been offered (Biton, 2003b; Jallon & Picard, 2001) and description of the growth trajectory has not been reported.
19 Gabapentin No research has examined mechanisms causing weight gain for this drug; however, increased consumption of carbohydrates and reductions in ener gy expenditure have been documented for other GABAergic drugs (see Biton, 2003b). Again, the course of change in weight status subsequent to gabapentin is unknown. In summary, data examining the pathogenesis of AED induced weight gain is preliminary and w rought with limitations. Specifically many of the studies cited in these reviews relied on cross sectional or retrospective data, chart reviews, small sample sizes, non randomized designs, and clinical anecdotes (Biton, 2003b). While these types of studie s are important for generating hypotheses they cannot be used as evidence to support causal statements about the mechanisms underlying change. Furthermore, very few studies were conducted with the express purpose of investigating AED induced weight gain. In drug studies assessing tolerability and efficacy of medications, weight gain was often tracked only as an adverse side effect. This is problematic not only because data to assess proposed mechanisms of accelerated weight gain (e.g., assessment of metabo lic parameters, dietary intake, energy expenditure, appetite) were not included but also because the incidence of excessive weight gain may be underreported as an adverse side effect. In one trial of valproate and carbamazepine, excessive growth velocity a s determined by exceeding the 97 th centile for age and gender was documented in 26% and 29% of children, respectively. However, it was only reported as an adverse side effect by 9% of the patients on valproate and 4% of the patients on carbamazepine (Easte Tear, & Verity; 1997). The true incidence of AED induced weight gain can only be determined by using objective measures of height and weight over time.
20 This gap in knowledge is especially pronounced for the newer medications vigabatrin and gabap entin, and among children who have been the focus of far fewer studies. So far, investigations of vigabatrin and gabapentin have mostly utilized add on designs (i.e., adding the new AED to an existing regimen); thus, findings are confounded by potential in teractions related to polytherapy (Biton, 2003b). Properties of medications with similar mechanisms of action have also been extrapolated to the newer AEDs without evaluation of those properties in the AED of interest. In children retrospective chart revie ws have commonly been used to examine the relationship between measures height and weight and medications over time (e.g., Egger & Brett, 1981; Wirrell, 2003). This provides documentation of the relationship between AED and weight status change but offers little by way of explaining of the mechanisms involved. Health Behaviors and Growth Velocity To expand on existing knowledge and contribute to the growing body of research evaluating AED induced weight gain, the proposed mechanisms listed above could be evaluated to some extent by assessing for changes in measurable health behaviors. On a basic level, weight gain results from a positive imbalance between energy intake and energy expenditure, which in terms of behavior is regulated by caloric intake and ph ysical activity. The pathophysiology underlying increased growth velocity should manifest in detectable patterns in energy regulating behaviors, such that alterations in these behaviors subsequent to AED therapy may reveal the source of the energy imbalanc e. For instance, if the mechanism catalyzing weight gain is appetite stimulation (Luef, Abraham, Haslinger, Trinka, Seppi et al., 2002), then increases in caloric intake subsequent to initiation of AED should be observed. Some have proposed an increase in specific nutrients such as carbohydrates and sweets (see Zimmerman et al., 2003),
21 which could be also be evaluated through the prospective assessment of dietary intake. On the other hand, increasing growth velocity despite stable caloric intake and physica l activity could indicate depressed metabolism or a reduction in basal energy turnover as the culprit (Leibowitz, 1992; Zimmerman et al., 2003). Beyond providing data to support or dispute existing hypotheses about the source of accelerated weight gain in children on AEDs, delineating the role of modifiable health behaviors may provide a concrete target for clinical intervention. Children and families could be taught preventative strategies such as monitoring, meal planning, increasing fruit and vegetable i ntake, and problem solving techniques (Epstein, Paluch, Roemmich, & Beecher, 2007; Janicke, Sallinen, Perri Lutes, Huerta, Silverstein, & Brumback, 2008) to reduce weight gain as a side effect of treatment, with corresponding benefits in terms of treatment compliance, seizure reduction, and preclusion of obesity related physical and psychosocial complications. Physical activity plays an important role in maintaining a healthy weight status among children and is the only manipulatable variable influencing en ergy expenditure (Spear, Barlow, Ervin, Ludwig, Saelens, Schetzina, & Taveras, 2007). Research suggests that overweight children take approximately 3000 less steps per day than their normal weight peers (Tudor Locke, Williams, Reis, & Pluto, 2002), and inc reasing physical activity among overweight children is frequently a component of weight 2008; Wilfley, Tibbs, Van Buren, Reach, Walker, & Epstein, 2007). Increased levels of physical activity have been associated with lower cardiovascular risk factors, BMI, and body fat, as well as improved maintenance of healthy body weight (see Spear et al.,
22 2007). Previous studies investigating the impact of epilepsy on quality of life have indicated that children diagnosed with epilepsy have restricted access to physical activities (Van Empelen, Jennekens Schinkel, van Rijen, Helders, & van Nieuwenhuizen, 2005), compared to healthy peers or siblings, such as less time spent participating in organized athletics and extracurricular activities (Arida, Cavalheiro, de Silva, & Scorza., 2008; Wong & Wirrell, 2006). Individuals with epilepsy as well as their physicians and caretakers often harbor concerns that engaging in physical activity will cau se seizures ( Ablah, Haug, Konda, Tinius, Ram, Sadler, & Liow, 2009) The physical toll of inactivity in this population has been documented: compared to those in general population, people with epilepsy display lower levels of muscle strength, flexibility (Steinhoff, Neususs, Thegeder, & Reimers, 1996), endurance (Bjorholt, Nakken, Rohme, & Hansen, 1990) and maximum oxygen intake ( Nakken, Bjorholt, Johannessen, Loyning, & Lind, 1990 ). On the contrary, emerging data suggests that physical activity may not on ly facilitate cardiovascular and emotional health (Ericksen, H.R., Ellertsen, B., Gronningsaeter, H., Nakkan, K.O., Loyning, Y., & Ursin, H., 1994), but it may actually reduce seizure frequency in those with epilepsy (Arida et al., 2008; Eriksen et al., 19 94). Reduction of seizure frequency as a result of AEDs may offer affected youth new opportunities to become more active (Elliot, Lach, & Smith, 2000). Indeed, the American Academy of Pediatrics (1983) released a position statement acknowledging the import ance of athletic endeavors in children and adolescents and stated that participation in such activities should not be restricted when seizure control is good with the exception of activities that could reasonably result in head injuries or injury. The Inte rnational League Against Epilepsy later advised that the only activities
23 that sports that should be prohibited in epilepsy are sky diving and scuba diving (Commission of Pediatrics of the International League of Epilepsy, 1997). Thus, it is important to co nsider physical activity in health promotion among children with epilepsy. The studies described above that reported on rates of physical activity in individuals with epilepsy typically utilized self report and questionnaire data, which can be unreliable especially when assessing behaviors over extended periods of time and among younger children. Objective documentation of physical activity patterns in children on AED medications could provide important information about activity patterns, the influence of AEDs on activity levels, and help to explore whether individuals who achieve good seizure control are in fact able to embrace opportunities to become more active. Psychosocial Functioning and Epilepsy As stated above, an abundance of literature has docume nted higher rates of psychopathology among children with epilepsy (e.g., Leonard & George, 1999; Pellock, 2004; Salpekar & Dunn, 2007) and overweight children (e.g., Eremis et al., 2004; Zeller & Modi, 2008) individually compared to healthy peers, however no research has explicitly investigated the effect of overweight on psychosocial functioning among children with epilepsy. Children with epilepsy are at greater risk for psychosocial dysfunction than any other chronic illness population (Rutter, Graham, & Yule, 1970), with comorbidity rates reported to be between 30 60% (Salpekar & Dunn, 2007). Approximately 1 in 4 children with epilepsy report clinically significant depression, which presents a particular threat in this population due to the higher rates o f suicide compared to children without epilepsy (Plioplys, 2003). Higher levels of depressive symptomatology (Eremis et al., 2004), suicidal ideation, and suicide attempts (see Puhl
24 & Latner, 2007 for review) have also been reported among clinically referr ed overweight children compared to non overweight peers. Both children with epilepsy and overweight children also report impaired quality of life compared to healthy peers. Epilepsy is accompanied by a wide range of psychosocial stressors for the affected child and their family, including missed days at work and school, loss of control and unpredictable seizures, restrictions on reaching developmental milestones such as driving and achieving autonomy, family discord, financial burden, cognitive impairment, and stigmatization (Leonard & George, 1999; Salpekar & Dunn, 2007). These factors interact to reduce functioning across multiple domains including physical, social, and cognitive quality of life (Van Empelen et al., 2005). Similarly, one study comparing o verweight to non overweight children and adolescents found that overweight youth reported significantly worse quality of life at a level comparable to children undergoing chemotherapy (Schwimmer, Burwinkle, & Varni, 2003). Understanding the impact of AED induced weight gain among children with epilepsy in terms of depressive symptoms and quality of life is therefore imperative to evaluate the clinical significance of overlapping risk. It is possible that both having epilepsy and being overweight exacerbate risk for psychopathology in terms of prevalence, severity, course, and prognosis. In addition, impaired emotional and social al., 2004) and impede successful weig ht management. Separate bodies of literature underscore the importance of early detection and intervention for psychiatric comorbidities in each of these populations (e.g., Leonard & George, 1999; Puhl &
25 Latner, 2007) yet until it is known how these condit identify those a greatest risk and offer effective treatment will be limited. In summary, dietary intake and physical activity are critical but as yet uninvestigated factors in the expanding body of research seeking t o identify the pathogenesis of AED induced weight gain. Isolating contributory health behaviors and psychosocial correlates of abnormal weight gain in children with epilepsy could facilitate the development of novel clinical interventions to mitigate its o ccurrence as a side effect of epilepsy treatment, thereby reducing the additional risk for obesity related morbidity and mortality and the possible adverse effect of weight gain on compliance with the medical regimen. The current project will provide essen tial pilot data to support an application for a randomized clinical intervention trial in what would be a new area of epilepsy research. Primary Aims and Hypotheses Aim 1 : To D escribe Weight Status in Youth with Epilepsy Who Are Prescribed AED s. Hypothesis 1.1 : Children who are prescribed at least 1 weight positive AED (e.g., valproate, carbamazepine, gabapentin, vigabatrin) will have higher weight status at baseline relative to those not prescribed a weight positive AED (e.g., phenytoin, lamotrigine, levet iracetam, topirimate, felbamate, zonisamide; no medication). Hypothesis 1.2 : Children who are prescribed at least 1 weight positive AED will demonstrate an increase in weight status from baseline to follow up relative to those not prescribed a weight posi tive AED.
26 Aim 2 : To Examine the Relationship Between Behavioral Health Factors (E.G., Caloric Intake, Energy Expenditure) a nd Weight Status in Youth on AED s. Hypothesis 2.1 : At baseline, children on a weight positive AED will report higher caloric intake than children not taking a weight positive AED. Hypothesis 2.2 : Average caloric intake across baseline and follow up will mediate the relationship between type of AED medication at baseline and change in weight status from baseline to follow up. Hypothesis 2.3 : At baseline, children on a weight positive AED will demonstrate lower levels of physical activity than children not on a weight positive AED. Hypothesis 2.4 : Average level of physical activity from baseline to follow up will mediate the rel ationship between type of AED medication at baseline and change in weight status from baseline to follow up. Aim 3: To Assess the Relationship Between Weight Status and Psychosocial Functioning in T erms of Depressive Symptoms and Quality o f Life in Youth with Epilepsy Hypothesis 3.1 : At baseline, those on a weight positive AED will report higher levels of depressive symptoms than those not on a weight positive AED. Hypothesis 3.2 : At follow up, those on a weight positive AED will report higher levels of de pressive symptoms than those not on a weight positive AED. Hypothesis 3.3 : At baseline, higher weight status will be associated with more depressive symptoms across groups. Hypothesis 3.4 : Change in weight status from baseline to follow up will mediate the relationship between type of AED at baseline and depressive symptoms at follow up.
27 Hypothesis 3.5 : At baseline, those on a weight positive AED will report lower quality of life than those not on a weight positive AED. Hypothesis 3.6 : At follow up, those on a weight positive AED will report lower quality of life than those not on a weight positive AED. Hypothesis 3.7 : At baseline, higher weight status will be related to lower quality of life across groups. Hypothesis 3.8 : Change in wei ght status from baseline to follow up will mediate the relationship between type of AED at baseline and quality of life at follow up. Exploratory Analyses Aim 4: E xamine whether the type o f AED (weight positive versus not weight positive) is associated wit h differences i n key nutrition variab les (e.g., fat, carbohydrates) that are related t o weight gain
28 Table 1 1. Hypothesized Trajectory of Growth Acceleration and Mechanisms Underlying Accelerated Weight Gain for Weight Positive AEDs Weight Positive AED Accelerated Growth Trajectory Hypothesized Mechanisms Underlying Weight Gain Valproate Increase rapidly for 3 months, plateau by 6 months Increased appetite Increased thirst Increased consumption of carbohydrates Decreased energy expenditure Reduced basal energy turnover Carbamazepine Unknown Edema Increased appetite Vigabatrin Unknown Unknown Gabapentin Unknown Unknown
29 CHAPTER 2 METHOD Participants Participants were 49 youth and their parent or legal guardian recruited at their regularly scheduled visit in the pediatric neurology clinic at Shands Hospital at the University of Florida. Eligible participants met the following inclusion criteria: (1) be between the ages of 8 17; (2) diagnosed with seizures; (3) fluent in English; and (4) accompanied to the appointment by a parent or legal guardian. Parent inclusion criteria were follows: (1) able to read, write, and speak English; (2) lived in the same h ousehold as the participating child at least 50% of the time; and (3) did not have any plans to move out the area for the next 9 months. Exclusion criteria included: (1) any child medical condition that would impact weight status (e.g., Prader Willi syndr ome); (2) use of a wheelchair or assisted walking device (e.g., walker); (3) serious psychopathology or other medical or behavioral condition (i.e., schizophrenia, bipolar disorder) that would interfere with the ability to complete study measures; and (4) being on the ketogenic diet. Some children were not able to complete self report measures due to a cognitive delay; in these cases, parents still completed parent report measures. Procedure The current protocol was approved by the governing IRB. A member of the research team was present in the pediatric neurology clinic on designated days for recruitment. Potential participants were identified by a review of their medical records, and those m eeting initial screening criteria were approached as they waited in private patient rooms during their regularly scheduled visit. Parent and child dyads were asked if they would like to hear more about a study examining health status in children with
30 epile psy. Families who expressed interest were given more information about the study and completed the informed consent/assent. Assessments took place at baseline (i.e., the day of recruitment following completion of consent and assent protocol), and at the ch appointment in the neurology clinic between 4 6 months post baseline. For each assessment, families completed a packet of questionnaires that took approximately 45 minutes. At the end of each assessment visit the child was given a n accelerometer and instructed to wear it for 7 consecutive days. Families were given a prepaid mailer to return the accelerometer to the research team. If families were unable to attend their follow up appointments at the neurology clinic or a team member was unable to meet them at this visit, they were provided the opportunity to complete study measures at the them. Measures Questionnaires Child and parent participants c ompleted the following questionnaires and anthropometric measurements at each assessment (see Appendix for all measures): Dietary i ntake The Block Kids 2004 is a 77 item questionnaire that assessed the week The f ood list for this questionnaire was developed from the NHANES 1999 2002 dietary recall data. The nutrient database was developed from the USDA Nutrient Database for Dietary Studies, version 1.0. The child completed this measure with the help of their paren t. Energy e xpenditur e Children wore a Sensewear (Bodymedia, Inc, Pittsburgh, PA) armband accelerometer for 7 days following each assessment visit. The Sensewear
31 armbands objectively evaluated total energy expenditure and physical activity energy expenditu re, steps taken, and intensity expressed as metabolic equivalents (METS) among the children recruited to participate in this study. Children and their parents were instructed in the proper usage of an accelerometer and were given a prepaid mailer for its r eturn. Participants were blinded to the data generated by the armband, which recorded and stored data until it was downloaded to a designated computer with appropriate licensed software. Participants were asked to wear the accelerometer for 7 days only ta king the device off to bathe or swim. In line with the extant literature suggesting that 10 hours of daytime wear is considered adherent to assess daytime levels of physical activity (Ekelund, Laun, Sherer, Esliger, Griew, & Cooper, 2012), participants wer e included if they had at least 10 hours of data for 2 weekdays and 1 weekend day. The first 2 eligible weekdays and first eligible weekend day were selected. Next, minutes spent in sedentary (0 2.9 METs), moderate (3.0 5.9 METs), vigorous (6.0 8.9 M ETs), and very vigorous (9 METs and above) activity, and average METs was calculated for the 3 selected days. Minutes in moderate, vigorous, and very vigorous activity were summed to determine minutes in physical activity (Ekelund, Luan, Sherer, Esliger, G riew, & Cooper, 2012). Since some individuals wore the accelerometer over night while others did not, minutes spent in sedentary activity was not comparable. Focusing on minutes spent in physical activity (which is not achieved during sleep) therefore targ ets waking hours is comparable between participants. Child health related quality of life The PedsQL is a 23 item scale that measures health related quality of life in healthy children and those with acute and chronic conditions. Participants are asked to rate the extent to which items have been a
32 problem for them in the past month on a 5 point Likert scale with anchors of 0 = never and 4 = always. The PedsQL is comprised of 4 subscales physical, emotional, social, and academic functioning as well as an o verall quality of life index. Items are reverse scored so that never = 100 and always = 0 with 25 point intervals between response options. Averages are then computed within each subscale and across the 23 items for the overall score. Higher scores indicat e better functioning, with 100 signaling no difficulties in functioning. Overall quality of life was used in the present study. The measure has been reported to have excellent internal consistency, clinical validity, and factor analytic support for subscal es (Varni, Seid, & Kurtin, 2001). Both child and parent proxy forms were administered. Child b ehavior and p sychosocial f unctioning The Behavioral Assessment System for Children (BASC) is a broadband measure of child behavioral and psychosocial functionin g. Parents are asked to rate the frequency of each behavior on a 4 point Likert scale (never, sometimes, often, almost always). The child version (BASC PRS C) has 134 items and the adolescent version (BASC PRS A) has 150 items. The BASC yields T scores for internalizing and externalizing domains, additional clinical scales, and activities of daily living, as well as subdomains within these areas. The depression subscale within the internalizing domain was used to assess for depressive symptoms in children. In line with the manual for this measure, a T score of 65 or greater based on age and gender norms were considered to be clinically elevated. The measure was completed by the participating parent. Validity scales were examined to ensure valid response patterns.
33 Demographic i nformation This questionnaire obtained family background information such as age, gender, race, marital status, education, and family income. This questionnaire also collected i nformation regarding parent and child medical history. Anthropometrics Height and w eight chart at each assessment. In the event that the follow up assessment was completed outside of a medical appointment, anthropometric measurements were taken by the Medical Records/Chart Review Medical i nformation Starting from the date of the epilepsy diagnosis, children are typically seen in the pedia tric neurology clinic every 4 6 months. Information on height and weight, the type of seizure disorder (i.e., generalized, partial, or unclassified), time since diagnosis, anti epileptic medication and other medication usage, dosage of medication, medical comorbidities, and psychiatric comorbidities was collected from the up assessment. Statistical Analyses All statistical analyses were conducted using the Statistical Package for the Social Sciences, (SPSS, v ersion 17.0, SPSS Inc., Chicago, Illinois). Sample Size This was the first project investigating the impact of health behaviors on weight status change in children on AED medication, and thus there was no existing data to guide power analyses. A sample s ize of 50 child/parent dyads gave .94 power to detect an effect size of .5 at follow up.
34 Preliminary D ata A nalyses Height and weight were converted first into BMI using the Quetelet formula [BMI= weight (kg)/ height (cm 2 )], and then into BMI z scores, wh ich plots child weight status relative to other children who are the same age and gender. BMI z score change was then calculated by subtracting their weight status at baseline from their weight status at follow up. A positive BMI z score change indicates i ncreased weight status from baseline to follow up while a negative change score indicates reduced weight status over the study period. For descriptive purposes, children were also classified into underweight, healthy weight, overweight, and obese weight st atus categories based on widely accepted cutoffs (e.g., Cole, Bellizzi, Flegal & Dietz, 2000). Children were separated into three groups: (1) children prescribed at least 1 weight positive AED; (2) children not prescribed a weight positive AED (i.e., chil dren taking weight neutral or negative AEDs or those not taking any AED); (3) children who were placed on, or taken off a weight positive AED during the course of the study. Those children in category 3 were excluded from analyses because the relationship between category of AED and change in weight status, behavioral health factors, and psychosocial functioning could not be assessed consistently over time. Upon examination of the medication status of child participants, it was discovered that only 11 child ren used weight positive AED, while 34 used non weight positive AED. To adjust for the significa nt imbalance between groups frequency matching was used to select a subset of participants from the non weight positive group who were similar to the weight p ositive group in age, gender, and race/ethnicity. The decision to match on these characteristics was guided by literature indicating that caloric intake and physical activity vary across these parameters (USDA, 2010). For instance, average level of
35 physica l activity declines during adolescence, especially among females (Troiano, Berrigan, Dodd, Masse, Tilert, & MacDowell, 2008; Zametkin, Zoon, Klein, & Munson, 2004). In addition, one of the critical developmental periods for childhood onset obesity occurs d uring early adolescence, especially for females (Dietz, 1994). Children from minority backgrounds are more likely to have a higher weight status than their non minority peers (Ogden, Carroll, Kit, & Flegal, 2012; Zametkin et al., 2004). Most importantly, m atching participants on demographic characteristics provided a way to control for these confounds a priori since the small sample size precluded the inclusion of these variables as covariates (Ahmed, Fatmi, Siddiqui, & Sheikh, 2011; Christoffel, Donovan, S hofer, Wills, & Levine, 1996). Preliminary data analyses included independent t tests to determine whether participants that completed both baseline and time 2 assessment (completers) differed from those that did not complete time 2 assessment (non completers) on age, weight status, caloric intake, or minutes spent in physical activity. Following the matching procedure whereby a subset of individuals from the non weight po sitive group were selected to match cases in the weight positive group on gender, age, and race/ethnicity, independent t tests were again used to assess for differences between those cases selected for analysis and those not selected for analysis across in dependent and dependent variables. Primary Analyses I ndependent samples t tests were used to assess for baseline differences between the weight positive and non weight positive group in weight status, total caloric intake, minutes spent in any physical ac tivity (i.e, > 3 METs) quality of life, and depressive symptoms Analysis of variance (ANOVA) was used to test for group
36 differences in change in weight status over time. Pearson product moment correlations were calculated between base line BMI z score, an d depressive symptoms and quality of life at baseline and follow up to determine the direction and strength of the relationships. Then, multiple regression analysis was used to test for mediation. The steps to test mediation outlined by Baron and Kenny (19 86) are as follows: establish a significant relationship between the predictor (AED at baseline) and the outcome (weight status at follow up); (2) establish a relationship between the predictor and the mediator (average caloric intake across baseline and f ollow up); (3) show that after controlling for the relationship between the predictor and the mediator, the mediator predicts the outcome; and (4) show that the reduction in the strength of association between the predictor and the outcome is significant i n the presence of the mediator using a Sobel test. Psychosocial measures were examined for missing data. On the PedsQL, if >80% of data points were present on individual subscales, mean substitution based on subscale scores was used to impute missing data points. Exploratory A nalyses Independent samples t tests were used to test for group differences in fat intake and carbohydrate intake at baseline.
37 CHAPTER 3 RESULTS Preliminary Analyses Demographic information is presented by group in Table 3 1. Overall 49 families were recruited into the study. Participants were 7 17 years old (M=12.3; SD=2.8). The majority of child participants was Caucasian (67.3%), followed by African American (16.3%), Hispanic (4.1%), Bi racial (4.1%), Asian (2.0%), and Other/Not r eported (6.1%). The vast majority of participating adults were mothers (79.6%), though fathers (14.3%), grandparents (2.0%), and step parents (2.0%) were also included. The majority of caregivers were currently married (69.4%). Median income was $40,000 $4 9,999. Across the whole sample, 38 families (78%) completed the follow up assessment. Independent t tests were used to test for differences between completers and non completers. Results indicated that the groups did not differ in terms of age, caloric int ake, minutes spent in physical activ ity, quality of life, depressive symptoms or BMI z score at baseline. Of the total sample, 11 children were on weight positive medication and 34 were in the non weight positive category (i.e., they were taking weight neutral or weight negative medication, or not taking any medication) ( Figure 3 1 ). Four individuals were put or taken off of a weight positive AED during the course of the study and were therefore excluded from analyses. Of the 34 youth in the non weight positive group, 5 did not complete the dietary intake questionnaire at baseline and 1 did not complete the que stionnaire at follow up and were also excluded from further analysis, reducing that group size to 28. Reasons for incomplete data collection included having to leave early, failing to return the measures via mail, and administrative error. Three of the rem aining
38 individuals in the weight positive group and 2 in the non weight positive group did not complete follow up, reducing the eligible number of participants in each group to 8 and 26, respectively. Of the 8 children in the weight positive group who com pleted follow up, 1 was non adherent to the accelerometer at baseline while 2 more were non adherent to the accelerometer at follow up. This meant that only 5 families in the weight positive group had complete physical activity data across follow ups. To p rotect the small sample size from further loss, it was decided that families who completed the entire baseline assessment, including having physical activity data, and who completed the follow up assessment with or without physical activity data would be m aintained. As a result, analysis of physical activity data would be limited to baseline. On the basis of these revised inclusion criteria, there were 7 eligible families in the weight positive group. Six a children did not have accelerometer data at baseli ne in the non weight positive group, leading to a sample size of 20. To allow for the control of covariates and balance grou p size, it was decided that frequency matching would be used to select a subset of participants in the non weight positive group for analysis. After obtaining several statistical consultations, this approach was judged to be the most scientifically sound given the small sample size and unbalanced groups. Seven cases from the non weight positive group, out of 20 eligible, were selected to match the frequency of age, gender, and minority status variables in the weight positive group (Table 3 2). Only 4 individuals out of the eligible 20 in the non weight positive group were of minority status, thus those 4 were selected for inclusion to m atch the frequency ( i.e., 4) of minorities in the weight positive group.
39 Three out of those 4 selected participants were male, which also fulfilled the frequency of males needed to match the weight positive group. Lastly, 3 more individuals were selected f rom the eligible non weight positive sample to match the distribution of ages in the weight positive group. To ensure systematic selection, the first 3 individuals based on date of recruitment who were of the needed ages were selected for inclusion. Indep endent t tests were used to assess whether the selected group of matched cases from the non weight positive group differed from the larger group from which they were drawn on any of the variables of interest. Results suggested no significant differences in caloric intake, minutes spent in physical activ ity, quality of life, depressive symptoms or BMI z score at baseline. Nor were there significant differences in change in BMI z score from baseline to follow up. Demographic characteristics, health status, e pilepsy type, time since diagnosis, and number of AED medications for the final s ample are presented in Table 3 3 In the weight positive group, 4 youth had generalized, 1 had partial, and 2 had unclassified seizures. In the other group, all 7 had generali zed seizures. Etiology of seizures in the weight positive group was more varied with 5 having idiopathic epilepsy and 2 having symptomatic epilepsy. In the non weight positive group, all 7 had idiopathic epilepsy. While not tested statistically, visual ins pection of the data suggests that youth in the weight positive group were less likely to have well controlled epilepsy. Within that group, 3 participants had daily or weekly seizures, 2 had seizures monthly, 1 had seizures yearly, and only 1 had seizures l ess than yearly. By comparison, in the non weight positive group, 2 individuals had daily or weekly seizures, 1 had seizures monthly, 1 had seizures yearly, and 3 participants had seizures less than yearly. The majority of
40 participants in both conditions w ere on monotherapy (6 weight positive; 4 non weight positive), and there was 1 individual in each group on polytherapy. The remaining 2 participants in the non weight positive group were not taking AED medication. The majority of participants in each group had been diagnosed greater than 2 years earlier (71% in each group). Only one participant (14%) in each group had been diagnosed with epilepsy within 6 months of baseline. Finally, based on parent report and review of medical records, 6 participants (86%) of individuals in the weight positive group had been diagnosed with a developmental disorder (e.g., autism, developmental delay) while only 2 individuals (29%) in the non weight positive group had a developmental problem. Two individuals in the non weight positive group were diagnosed with ADHD and Oppositional Defiant Disorder, respectively. No additional psychiatric diagnoses were documented for the weight positive group. In terms of additional health conditions, 5 individuals in the weight positive grou p had comorbid medical diagnoses that included gingival hyperplasia, genetic disorder, NOS, antiphospholipid syndrome, cerebral palsy, static encephalopathy, subdural hematoma, and encephalitis. Only one person in the non weight positive group had addition al medical diagnoses (i.e., borderline diabetes, obesity, and obstructive sleep apnea). Aim 1 Weight Status Between Groups Independent samples T test was used to assess for differences in baseline weight status between groups [t(12) = 1.41, p = .18]. Th ose in the weight positive group (Mean BMI z score = .28, SD = 1.86, range 3.18 1.73 ) did not have a significantly higher weight status as baseline than those in the non weight positive group (Mean BMI z score = .98, SD = .54, range 1.10 2.90 ) ( Tabl e 3 3 ). For descriptive purposes participants were categorized into widely accepted weight status categories based on
41 BMI percentile. In the weight positive group, 2 children were classified as underweight, 3 as healthy weight, 1 as overweight, and 1 as ob ese. In the non weight positive group, 3 youth were classified as healthy weight and 4 as obese. While not tested statistically, there appeared to be more variability in categorical weight status in the weight positive group, and a greater tendency towards being obese in the non weight positive group. Analysis of variance was used to examine the change in weight status (i.e., change in BMI z score) from baseline to follow up between those on weight positive AEDs and those not on weight positive AEDs. No dif ferences in change in weight status from baseline to follow up emerged between the weight positive group (M = .52; SD = 1.26) and the other group (M = .07, SD = .31) [F(1,12) = .86, p = .377]. Aim 2 Behavioral Health Factors B etween Groups Information about behavioral health factors by gro up is presented in Table 3 4 Independent samples t test was used to assess for differences in caloric intake at baseline between groups. Those on weight positive AEDs (M= 1797.88, SD = 586.54) did not report significantly greater caloric intake than those not on weight positive AEDs (M = 1740.95, SD = 1080.64) [t(12) = .12, p = .90]. Multiple regression analysis was used to examine whether average caloric intake mediated the relationship between type of AED a t baseline and change in weight status. In line with procedures outlined by Baron and Kenny (1986), the path between the exogenous (group membership) and endogenous (weight status) variable was tested first. Type of AED at baseline did not significantly pr edict change in weight status (R 2 = .2 6, F[1,12] = .86, p = .37) ( Figure 3 2). Therefore, mediation was not possible. Next, independent samples t tests were used to examine baseline differences in minutes in physical activity between groups. Those in the weight positive group (M =
42 365.42, SD = 313.27) did not exhibit significantly lower levels of physical activity than those in the non weight positive group (M = 424.29, SD = 207.90) (t = .41, p = .69). As stated above, group membership did not predict change in weight status from baseline to follow up (R 2 = .26, F[1,12] = .86, p = .37); therefore, minutes spent in physical activity could not assessed as a me diator of this relationship ( Figure 3 3). Aim 3 Psychosocial Functioning Between Groups Upon examination of the PedsQL data, it was discovered that 1 participant failed to answer 1 item (out of 8) on the physical subscale of the parent proxy report at follow up. Therefore, the mean response was calculated using the rest of the items on that s ubscale and substituted for the missing value. Pearson product moment correlations were calculated to determine the strength and direction of relationships between weight status and psychosocial vari ables at baseline (Table 3 5 ). At baseline, moderate but non significant positive correlations were found betw een weight status and depressive symptoms (r = .48) and weight status and quality of life (r = .40). This means that higher weight status was related to both higher levels of depressive symptoms and bet ter quality of life. Correlations between baseline weight status and follow up levels of depressive symptoms (r = .1 7) and quality of life (r = .05 ) were weak. Significant correlations were observed between depressive symptoms at baseline and follow up (r = .59, p < .05 ). While not significant, moderate stability was also found between serial measurem ents of quality of life (r = .45 ). Finally depressive symptoms at baseline strongly predicted quality of li fe at follow up (r = .74 p < .01) meaning that h igher levels of depressive symptoms at baseline were related to lower quality of life at follow up.
43 Levels of psychosocial variables by group are presented in Table 3 6 Independent samples t tests revealed that the weight positive group (M = 59.94, SD = 1 1.29) reported significantly worse quality of life than the non weight positive group (M = 80.43, SD = 11.24) at baseline (t = 3.40, p < .01). At follow up, the groups did not differ in quality of life (t = 1.66, p = .12 ). No group differences em erged in level of depressive symptoms at baseline (t = .43, p = .67) or follow up (t = .72, p = .49). The average level of depressive symptoms across groups and assessment points was in the non depressed range. Two individuals (28.6%) in the weigh t positive group and 1 individual (14.3%) in the non weight positive group reported levels of depressive symptoms above the clinical cutoff of T = 65 at both baseline and follow up. Finally, linear regression was used to determine whether group membership predicted depressive symptoms and quality of life at follow up. Results suggested that group status was not significantly predictive of depressive symptoms (R 2 = 04, F[1,12] = .52, p = .49) ( Figure 3 4) or quality of life (R 2 = .19, F[1,12] = .2.74, p = 12 ) ( Figure 3 5) at follow up. As a result of these non significant paths, further investigation as change in weight status as a mediator for these relationships was not conducted. Exploratory Analyses Independent samples T tests were used to determi ne whether groups significantly differed in terms of key nutrients at baseline. Those in the weight positive group did not exhibit significantly different intake of calories from fat (M= 69.86, SD = 20.72) than those in the non weight positive group (M = 71 .78, SD = 48.89) [t(12) = .10, p = .93]. Similarly, no group differences emerged in calories from carbohydrates (weight positive M = 222.04, SD = 81.30; non weight positive M = 220.66, SD = 138.97) [ t(12) = .02, p = .98].
44 Table 3 1. Demographic Characteristics Across Participants by AED Category Characteristic Weight Positive Non Weight Positive Excluded N 11 34 4 Child Age 14.0 (2.5) 11.9 (2.8) 11.5 (2.1) Boys/Girls (n) 6 / 5 13 / 21 3 / 1 Adult Participant Mother 9 27 3 Father 2 4 1 Other 0 2 0 Two parent households (%) 55 74 75 Child Race/Ethnicity Caucasian 5 24 4 African American 5 3 0 Hispanic 0 2 0 Bi racial 0 2 0 Asian 0 1 0 Other 0 1 0 Not reported 1 1 0 Family Income Below $19,999 3 5 0 $20,000 $59,999 5 16 3 Over $60,000 3 11 1 Not reported 0 2 0 Completed follow up (n) 8 29 1 BMI z score baseline 0.1 (1.8) 1.0 (1.1) 1.1 (1.5) BMI z score follow up 0.4 (2.6) 0.9 (1.2) N/A
45 Table 3 2. Gender, Minority Status, and Age for All Participants Entered into the Frequency Match Procedure Participant Group Selection Status Gender Age Minority Status 1 Weight positive Selected F 16 Minority 2 Weight positive Selected F 16 Minority 3 Weight positive Selected F 11 Minority 4 Weight positive Selected M 15 Caucasian 5 Weight positive Selected F 8 Minority 6 Weight positive Selected M 14 Caucasian 7 Weight positive Selected M 13 Caucasian 8 Non weight positive Selected M 11 Minority 9 Non weight positive Selected F 17 Caucasian 10 Non weight positive Selected M 12 Minority 11 Non weight positive Selected F 15 Caucasian 12 Non weight positive Selected F 16 Caucasian 13 Non weight positive Selected F 11 Minority 14 Non weight positive Selected M 9 Minority 15 Non weight positive Not selected F 11 Caucasian 16 Non weight positive Not selected F 12 Caucasian 17 Non weight positive Not selected F 8 Caucasian 18 Non weight positive Not selected F 10 Caucasian 19 Non weight positive Not selected F 8 Caucasian 20 Non weight positive Not selected M 13 Caucasian 21 Non weight positive Not selected F 11 Caucasian 22 Non weight positive Not selected F 12 Caucasian 23 Non weight positive Not selected M 17 Caucasian 24 Non weight positive Not selected F 12 Caucasian 25 Non weight positive Not selected F 10 Caucasian 26 Non weight positive Not selected F 11 Caucasian 27 Non weight positive Not selected F 11 Caucasian
46 Table 3 3 Demographics Epilepsy Type, Time Since Diagnosis, Seizure Frequency, Medications, and BMI Z Scores by AED Group for Matched Sample Characteristic Weight Positive Non Weight Positive N 7 7 Child Age 13.3 (2.9) 13.0 (3.0) Boys/Girls (n) 3/ 4 3/ 4 Child Race/Ethnicity Caucasian 3 3 African American 4 1 Hispanic 0 2 Bi racial 0 0 Asian 0 1 Family Income Below $19,999 2 2 $20,000 $59,999 3 3 Over $60,000 2 2 Not reported 0 0 Type of Epilepsy Generalized 4 7 Partial 1 0 Unclassified 2 0 Etiology Idiopathic 5 7 Symptomatic 2 0 Cryptogenic 0 0 Time Since Diagnosis Within 6 months 1 1 Within 2 years 1 1 Greater than 2 years 5 5 Seizure Frequency at Baseline Daily/Weekly 3 2 Monthly 2 1 Yearly 1 1 Less than Yearly 1 3 Number of AED Medications Zero 0 2 One 6 4 Two 1 1 Co morbid Developmental Disorder 6 (86%) 2 (29%) Co morbid Medical Diagnosis 5 1 BMI z score baseline 0.3 (1.9) 1.0 (1.4) Underweight 2 0 Healthy (n) 3 3 Overweight (n) 1 0 Obese (n) 1 4
47 Table 3 3 Continued Characteristic Weight Positive Non Weight Positive BMI z score follow up 0.8 (2.6) 0.9 (1.5) BMI z score range, follow up 4.12 1.87 1.48 2.31 BMI z score change 0.52 (1.26) 0.07 (.31)
48 Table 3 4 Behavioral Health Factors by AED Group for Matched Sample Health Behavior Weight Positive (N=7) Non Weight Positive (N=7) Dietary Intake Total caloric intake, baseline 1797.88 (586.54) 1740.95 (1080.64) Total caloric intake, follow up 1397.68 (556.18) 1494.79 (62.72) Calories from fat, baseline 69.86 (20.72) 71.78 (48.89) Calories from carbohydrates, baseline 222.04 (81.30) 220.66 (138.97) Physical Activity Minutes in sedentary activity, baseline 3443.29 (348.40) 3690.57 (225.55) Minutes in moderate activity, baseline 320.71 (264.79) 386.86 (186.48) Minutes in vigorous activity, baseline 43.43 (53.48) 36.00 (40.51) Minutes in very vigorous activity, baseline 1.29 (2.36) 1.43 (1.90) Average METs, baseline 1.58 (0.31) 1.64 (0.20)
49 Table 3 5 Pearson Product Correlations Among Psychosocial and Weight Status Variables for Matched Sample Variable 1 2 3 4 5 1. BMI z score baseline -2. BASC depression, baseline .48 -3. BASC depression, follow up .17 .59* -4. QOL, baseline .40 .20 .23 -5. QOL, follow up 05 .74 ** 31 45 -* p <.05; ** p <.01
50 Table 3 6 Psychosocial Functioning by Group for Matched Sample Psychosocial factor Weight Positive (N=7) Non Weight Positive (N=7) BASC Depressive symptoms, baseline (t score) 54.43 (12.90) 51.29 (14.19) N at or above clinical cutoff, baseline 2 (28.6%) 1 (14.3%) Depressive symptoms, follow up (t s core) 47.43 (25.10) 54.71 (9.27) N at or above clinical cutoff, follow up 2 (28.6%) 1 (14.3%) PedsQL Quality of life, baseline 59.94 (11.29) 80.43 (11.24) Quality of life, follow up 63.20 (16.79) 78.42 (17.58 )
51 Participants Recruited: N=49 Participants Excluded: N=4 starting or stopping a weight positive drug during course of the study. Weight Positive: N=11 Non Weight Positive: N=34 Participants Excluded: N=6 did not complete dietary intake Participants lost to follow up: N=2 Weight Positive: N=8 Participants lost to follow up: N=3 Non weight Positive: N=26 Participants Excluded: N=1 No baseline accelerometer data Participants Excluded: N=6 No baseline accelerometer data Weight Positive: N=7 Non weight Positive: N=20 Non weight Positive Participants Not Selected /Matched for Analysis : N=13 Non weight Positive Participants Matched for Analysis: N=7 F igure 3 1 Participant Flow Chart
52 Figure 3 2. Proposed Mediation of A ED s and Change in Weight Status by Average Caloric Intake Type of AED Change in Weight Status from Baseline to Follow up Average Caloric Intake Across Assessments
53 Figure 3 3. Proposed Mediation of AED and Change in Weight Status by Physical Activity Type of AED Change in Weight Status from Baseline to Follow up Physical Activity at Baseline
54 Figure 3 4. Proposed Media tion Model for AED and Depressive Symptoms by Change in Weight Status Type of AED Depressive Symptoms at Follow up Change in Weight Status from Baseline to Follow up
55 Figure 3 5. Proposed Mediation Model for AED and Quality of Life by Change in Weight Status Type of AED Quality of Life at Follow up Change in Weight Status from Baseline to Follow up
56 CHAPTER 4 DISCUSSION Anti epileptic medications are the first line of treatment in pediatric epilepsy. Up to 65% of children on AEDs eventually become seizure free, often within in the first 2 years of medical therapy. A few of these medications have been associated with accel erated weight gain (see Biton, 2003b for review), which could increase the risk for obesity related medical and psychosocial comorbidities. Significant weight gain as a side effect of treatment could contribute to the high rates of non adherence in this po pulation (Modi, Ingerski, Rausch, & Glauser, 2011), especially among body conscious adolescents (Daniels, Nick, Liu, Cassedy, & Glauser, 2009). The etiology of increased growth velocity remains unknown but has been theorized to include increased appetite, decreased energy expenditure, and reduced basal energy turnover (Biton, 2003b). To our knowledge, this is the first study that assessed the relationship between AEDs and critical, modifiable health behaviors that regulate caloric balance and therefore wei ght status (USDA, 2010). Both youth with epilepsy and those of overweight and obese weight status are at greater risk for psychosocial impairment than their healthy, normal weight peers (e.g., Eremis et al., 2004; Leonard & George, 1999; Pellock, 2004; Zel ler & Modi, 2008). We also assessed depressive symptomatology and quality of life to determine whether AED induced weight gain leads to additional psychosocial difficulties in youth with epilepsy. Aim 1 AEDs and Weight Status It was hypothesized that chi ldren taking a weight positive AED would demonstrate higher weight status at baseline and greater increases in weight status over time compared to children on weight neutral or negative AEDs or those not taking AEDs.
57 Neither of these predictions was suppor ted by the present data. Although there were no significant differences in mean BMI z score between groups at baseline, there appeared to be a greater tendency towards overweight or obesity in the non weight positive group when looking at weight status fro m a categorical perspective. Specifically, at baseline 2 individuals (29%) were overweight or obese in the weight positive group compared to 4 participants (57%) in the non weight positive group, contrary to expectations. No significant differences emerged between groups over time. Given that the majority of participants had been diagnosed with epilepsy greater than 2 years ago, our inability to detect group differences in growth velocity over time is not surprising. Existing research on the weight related impact of valproate (Depakote) has generally suggested that the period of increased growth velocity occurs within the first 3 months of treatment, stabilizes by 6 months, and then plateaus for the duration of treatment (Biton, 2003b; Demir & Aysun, 2000; E gger & Brett, 1981; Novak et al., 1999). However, one study found that in younger children the period of increased growth velocity can span 16 months (Grosso et al., 2009). The trajectory of accelerated weight gain has not been established for the other AE Ds most consistently associated with weight gain (i.e., carbamazepine, gabapentin, and vigabatrin). If children had been taking weight positive medication for longer than 6 months, it is very likely that this important window of accelerated growth velocity was not captured by the present study. Therefore, children in weight positive group would not be gaining weight more rapidly than those in the other group as a result of medication side effects. Under the assumption that the accelerated weight gain had al ready taken place, it was anticipated that children on weight positive AEDs at baseline would still exhibit a
58 higher weight status compared to those on weight neutral, negative, or no AEDs, since extant literature suggests that increased weight status is m aintained over the duration of treatment (Biton, 2003b). Yet, those on weight positive AEDs at baseline did not demonstrate higher weight status than those in the non weight positive group. There are several potential explanations for these findings. First even when focusing on valproate, the AED most strongly associated with weight gain, only 45 60% of youth experience this side effect (Egger & Brett, 1981; Wirrell, 2003). The prevalence of accelerated weight gain amongst the other weight positive AEDs ra nges from 5 25% (Biton, 2003b; Easter et al., 1997 ). Therefore, the base rate for affected children in our study could potentially be quite low given the sample size and render it difficult to assess. Second, children with epilepsy are more likely to prese nt at diagnosis with higher rates of both underweight and overweight compared to healthy controls (Daniels et al., 2009) and to have a higher weight status over time than healthy age matched peers (Wong & Wirrell, 2006). While we did not capture weight sta tus at diagnosis, by baseline, 43% of youth in our study had BMIs in the overweight or obese range which exceeds the national rate of 33% in children aged 6 19 years (Ogden et al., 2012). Indeed, less than half of the children in our study fell in the heal thy weight status range. Armed with this information, clinicians may be inclined to consider potential weight effects when deciding between equally effective medications to promote weight gain among individuals with preexisting underweight and weight loss in those in the overweight or obese range, which over time may theoretically produce less variability in weight status between children on different medications over time.
59 Etiology has been related to weight status among children with epilepsy (Daniels at el., 2009). Children with symptomatic seizures are at greater risk to have a weight status under the 10 th percentile while those with idiopathic seizures are at greater risk for being overweight. Additionally, children with intractable epilepsy are more l ikely than healthy peers to experience growth failure, especially those with comorbid neurological conditions (Bergqvist, Trabulsi, Schall, & Stallings, 2008). While not tested statistically, there were more individuals with symptomatic and uncontrolled ep ilepsy in the weight positive group, as well as more individuals with comorbid developmental problems. It is possible that one or more of these factors was suppressing growth velocity accelerations or that AED induced weight gain had helped to normalize th eir weight status relative to peers. By excluding individuals on the ketogenic diet and those who were non ambulatory, we likely eliminated many others who would be on the lower end of the weight spectrum. A larger trial would be necessary to investigate p otential interactions between seizure type, etiology, comorbid neurological problems, and AED induced weight gain. Behavioral Health Factors Dietary Intake Increased appetite has been proposed as one of the mechanisms underlying accelerations in growth vel ocity for weight positive AEDs. The behavioral proxy of appetite is dietary intake which should theoretically increase when appetite is stimulated. If increased appetite is indeed a factor contributing to changes in weight velocity, it was hypothesized th at this could be captured via level of caloric intake such that those on weight positive AEDs would report higher caloric intake at baseline compared to those not on weight positive medications. We also predicted that average
60 caloric intake from baseline t o follow up would mediate the relationship between type of AED at baseline and weight status change over time, meaning that higher levels of caloric intake explain why appetite stimulating medications are associated with increasing weight status. Contrary to expectations, those taking weight positive AEDs did not report higher caloric intake at baseline compared to those on weight neutral, negative, or no AEDs. In addition, type of AED at baseline was not related to weight status change over time precludin g examination of caloric intake as a mediator of growth velocity. The average level of daily caloric intake at baseline in the weight positive group (M = 1798, S.D. = 587) and non weight positive group (M = 1741; S.D. = 1080) was slightly lower than the na tionally reported intake in children and adolescents (1897 2218 kcal/day) in the NHANES III survey (McDowell, Briefel, Alaimo, Bischof, et al., 1994). Only 4% of calories per day were reported to come from fat and only 12% of calories from carbohydrates. These figures are well below the nationally reported intake of fat (~35%), as well as in conflict with data finding that carbohydrates constitute the greatest source of calories in the American diet (USDA, 2010). At follow up, average daily caloric intake fell to between 1400 and 1500 calories per day across groups. Comparing these results to those from NHANES III database suggests that current estimates of intake may be low. The lack of significant group differences in caloric intake is not surprising gi ven that the groups did not differ in weight status at baseline or over time. Again, given the small sample size used for data analysis, potentially low prevalence of weight gain as a side effect, and time since diagnosis, it is possible that dietary intak e is a contributory
61 factor in accelerated growth velocity but was not adequately captured in the current study. Differences in intake over time may reflect the assessment environment. At baseline, the child and participating adult completed the dietary int ake measure together under the supervision of a research assistant. At follow up, many of the families completed the assessments remotely. It is possible that the nutritional surveys were completed independently by the child rather than in collaboration wi th the parents at follow up, and consequently that the reported variety and/or portion sizes of foods consumed were less accurate. Future studies seeking to elucidate the pathogenesis of AED induced weight gain should measure dietary intake pre and post AED initiation and consider using the gold standard of nutritional assessment, 3 day dietary recalls completed with the child and parent, if time and resources allow. Individuals on the ketogenic diet (excluded from the present study) have been the virtua l sole focus of studies assessing dietary intake in youth with epilepsy. The high fat, low carbohydrate regimen can sometimes reduce seizures in those with medically refractory disease (Neal, Chaffe, Schwartz, Lawson, Edwards et al., 2008). Research has sh own that intractable epilepsy is associated with lower caloric intake, poorer nutritional status, and growth retardation (Volpe, Schall, Gallagher, Stallings, & Bergqvist, 2007). The ketogenic diet in itself produces declining weight status (Liu, Williams, Basualdo Hammond, Stephens, & Curtis, 2003). In contrast to the expanding body of research focusing on children at the bottom of the growth curve, little attention has been paid to assessing nutritional intake and growth among youth with epilepsy at the o ther end of the weight spectrum. Considering the high rates of overweight and obesity among youth (Ogden et al., 2012), the increased tendency towards overweight
62 among children with epilepsy (Daniels et al., 2009; Modi, Ingerski, Rausch, & Glauser, 2011), and the association of certain AEDs with increased weight gain (Biton, 2003a,b), endocrine (Aydin, Serdologlu, Okuyaz, Bideci, Gucuyener, 2005; Verotti et al., 2002 ), and biochemical changes ( Isojarvi et al., 1998) this population may be particularly vuln erable to unhealthy weight status and consequently to obesity related medical and psychosocial comorbidities. Interventions targeting dietary intake may be more effective in reducing weight status among youth compared to attempts to increase physical activ ity (Spear et al., 2007). Understanding nutritional intake in this population is the necessary first step to developing interventions to promote healthy weight status. Physical Activity Physical activity also regulates weight status through its impact on energy balance, and thus we hypothesized that lower levels of physical activity among youth on weight positive AEDs secondary to fatigue or lethargy could also explain increased growth velocity. This was not supported by the present data. Youth on weight p ositive AEDs did not spend significantly more minutes in physical activity over the course of 3 days (M = 367) than those on non weight positive AEDs (M = 426). Since the groups did not differ in weight status change over time, physical activity was not ex plored as a mediator. One important finding was that, across groups, youth on average spent more than 2 hours per day in physical activity (>3 METs), more than doubling the national guidelines for physical activity of 60 minutes per day in this age group ( USDA, 2010). These results are somewhat inconsistent with literature on activity patterns in children with epilepsy suggesting this population may engage in less physical activity than healthy peers or siblings (Van Empelen et al., 2005; Wong & Wirrell, 20 06). However, these previous studies used parent and self reports of physical activity instead of
63 objective measures of energy expenditure such as accelerometry, which was used in the current study. Additionally, children with epilepsy have been found to e ngage in fewer group based physical activities which confer benefits not only through physical conditioning but also increased opportunity for peer socialization (Wong & Wirrel, 2006). The latter is important as youth with epilepsy report feeling socially withdrawn (McEwan, Espie, Metcalfe, Brodie, & Wilson, 2004).We did not document the types of physical activities that youth were engaged in, whether individual versus group based, or recreational versus organized sports. It will be important for future stu dies to not only measure energy expenditure objectively but also record how children spend their time. Research on physical activity and physical fitness in children has shown that overall levels of daily physical activity are significantly correlated to c ardiorespiratory fitness measured by V0 2peak, and that this relationship is stronger for intensity in the vigorous range (Dencker, Thorsson, Karlsson, Linden, Svenssen, et al., 2006). A recent meta analysis of studies published from the International Chil Accelerometry Database representing more than 20,000 children found that more time cardiometabolic profile (Ekelund, Luan, Sherar, Esliger, Griew, & Cooper, 2012), which reduces risk for obesity related comorbidities such as type 2 diabetes and coronary disease. Existing literature suggesting that children with epilepsy are less active than healthy peers is concerning in part because they may not be achieving healthy fitn ess levels. Indeed, studies of physical fitness in adults with epilepsy have indicated lower levels of endurance, strength, flexibility, and cardiorespiratory fitness compared to the general population (Nakkan et al., 1990; Steinhoff et al., 2005). Similar data is not
64 available for children with epilepsy. Results from the present study are encouraging in that they show that at least among those in our sample, children on average are spending enough time in sufficiently intense activities to potentially reap these benefits. However, we did not measure fitness or assess biochemical markers such as cholesterol or insulin resistance. Future studies examining physical activity in this population should include objective measures of cardiorespiratory and metabolic functioning. Our physical activity findings must be interpreted with caution. First, it is possible that that youth who were compliant with the accelerometer may have been more active than their non adherent counterparts. Additionally, there is a tendenc y for studied behaviors to increase (Campbell, Maxey, & Watson, 1995; Vehmas, 1997), so that youth may have been inclined to be more active than usual on days that they were wearing the accelerometer. Due to a lack of power, we could not look at predictors of physical activity, such as seizure control. Lastly, our small sample size makes it difficult to generalize any findings. Psychosocial Functioning Across all individuals in our study, higher levels of depressive symptoms at baseline were significantly c orrelate d to higher levels of depressive symptoms and worse quality of life at follow up. Concurrent ratings of depressive symptoms and quality of life at baseline and follow up were not related As depressive symptoms intensify over time, changes in mood, activity level, concentration, energy, sleep, and appetite may translate into impairments in HRQOL. Chronic c omorbid internalizing disorders have been strongly tied to poorer HRQOL in long term follow ups of those with childhood onset epilepsy, a relation ship that persists long after remission is established,
65 and is more robust than disease related factors (Baca, Vicrey, Caplan, Vassar, & Berg, 2011). The rate of c linically significant depressive symptoms in our sample (21%) was relatively consistent with previous studies (25%) (Plioplys, 2003), and is particularly concerning due to the 3 fold increase in suicide among people with epilepsy compared to the general population ( Christensen, Vestergaard, Mortensen, Sidenius, & Agerbo, 2007 ). Taken together, t hi s information may be clinically meaningful if timely identification and treatment of individuals with depression lead to better outcomes in HRQOL and reduces risk of self harm. Baseline weight status showed moderate but non significant correlations with co ncurrent depressive symptoms (r = .48) and quality of life (r = .40). Considering the strength of the correlations, the lack of significance likely reflected the small sample size. The positive relationship between weight status and depressive symptoms was in the expected direction given extant literature suggesting that overweight and obese youth have higher rates of depression than their healthy weight peers (Eremis et al., 2004; Puhl & Latner, 2007). With research also showing that children and adolescen ts with epilepsy are at increased risk for depression ( Plioplys, 2003 ), it is possible that the combination of these conditions compounds risk. A s prevalence rates of clinically elevated depressive symptoms in this sample are similar to those previously re ported among youth with epilepsy, our findings do not appear to bear out this concern; however, this was not explicitly tested. We hypothesized that HRQOL would be negatively associated with weight status in line with extant literature documenting lower H RQOL in obese children compared to healthy weight peers (Doyle, le Grange, Goldschmidt, & Wilfley, 2007; Zeller & Modi,
66 2006). However, the relationship between HRQOL and weight status in the present study was in the opposite direction, such that increasin g weight status was related to better HRQOL. It is possible that the individuals on the lower end of the weight spectrum in our sample may have had poorer disease control or additional medical or neurological comorbidities that increased their susceptibility to both being underweight and experiencing a reduced quality of life. For example, children with refractory epilepsy and comorbid cognitive and physical developmental problems experience high rates of feeding difficulties (e.g.,chewing, swal lowing, self feeding) that interfere with nutritional and weight status (Bertoli, Cardinali, Veggiotti, Trentani, Testolin, et al., 2006). If this was true in our sample, lower weight status may have corresponded to more severe medical or neurodevelopmenta l issues that impacted daily functioning. On the other hand, this finding may be idiosyncratic to this small group of children. It would be important to replicate this finding in larger studies before interpreting the significance of the current results. Overall, findings from the present study were consistent with the well established body of literature showing that children with epilepsy have lower quality of life than the normative population (Ingerski et al., 2010; Modi, King, Monahan, Koumoutsos, Mori ta, et al., 2009). Peer interactions around disease disclosure and barriers to obtaining developmentally expected autonomy negatively impact quality of life in adolescents with epilepsy (McEwan et al., 2004). Other factors which have been shown to adversel y impact HRQOL in this population include seizure frequency (Camfield, Breau, & Camfield, 2001; Sabaz, Cairns, Bleasel, Lawson, Grinton, et al., 2003) and AED side effects (Benavente Aguilar, Morales Blanquez, Rubio, & Rey, 2004; Modi et al., 2011).
67 Our ex amination of factors related to quality of life were restricted to AED type and weight status in the present study, and it is a limitation that we could not assess the influence of disease or treatment factors on HRQOL due to the small sample size. In part icular, this line of investigation may have shed light on the confusing relationship found between weight status and HRQOL. The only group differences in psychosocial functioning at either time point was that the weight positive group had significantly wo rse HRQOL at baseline than the non weight positive group. This could not be explained by weight status as predicted, since the groups did not differ in BMI z scores at baseline or over time. However, group differences did exist in the prevalence of develop mental delay. Eighty six percent of the children in the weight positive group had been diagnosed with a developmental delay compared to only 29% in the non weight positive group. Neuropsychological functioning has been inversely associated with psychosocia l functioning among children with epilepsy (Baca et al., 2011; Leonard & George, 1999), and the presence of a developmental delay has been demonstrated to have global negative impacts on quality of life, and particularly adverse effects on social and schoo l functioning (Modi et al., 2009). The greater prevalence of developmental delays among children in the in the weight positive group may therefore help to explain why they were reported to have lower HRQOL at baseline than those in the non weight positive group; however, the fact that this difference disappears over time despite the chronic nature of developmental problems weakens this hypothesis. Limitations This study had several notable limitations. First, the small sample size and imbalance in groups r educed the ability to detect significant findings and precluded
68 more sophisticated statistical analyses (e.g., structural equation modeling to assess mediation). Beyond recruiting a larger sample, ensuring balance between groups during the recruitment phas e will be important for future studies to better assess for AED related patterns in health behaviors. Additionally, we controlled for age, race, and gender given their established relationship to health behaviors (Odgen et al., 2012; Spear et al., 2007). G iven our small sample size, we were unable to look at health behavior patterns by age group. As children age they have more control over their environment and therefore their health behaviors may be more responsive to changes in appetite or energy (e.g., t hey have increased independent access to food) compared to children whose intake or physical activity is more closely monitored by parents. It would be advisable to explore these as predictors of unhealthy weight status in future studies rather than contro l for them, as they may help to identify individuals at higher risk who would be more likely to benefit from preventative interventions. Existing literature suggests that increased weight velocity often occurs within in the first 6 months of AED initiatio n then stabilizes for the duration of treatment (Demir & Aysun, 2000; Egger & Brett, 1981; Novak et al., 1999). Our sample included many children who have been on AEDs for several years, dramatically reducing our ability to ascertain the impact of initial treatment on changes in growth velocity and consequently insight into causal mechanisms. Effect sizes for group differences in weight status and behavioral health factors were uniformly small and likely reflected, at least in part, the lengthy time since d iagnosis. The ideal study would target individuals at epilepsy diagnosis and follow them over the first year of treatment to determine whether AED
69 initiation in fact produces changes in dietary intake and physical activity that explain accelerated weight g ain. A few individuals in the study were taking stimulant medications for ADHD, a common comorbidity in pediatric epilepsy (Salpekar & Dunn, 2007). These types of medications are known for their appetite suppressing properties (Sonuga Barke, Coghill, Wiga l, DeBacker, & Swanson, 2009) and therefore are a confounding factor. Given the small sample size of the present study, it was not possible to control for stimulant medication. It remains unknown how these medications interact with AEDs to influence dietar y intake and weight status. The questionnaires used to assess psychosocial functioning were parent report. The choice to use parent proxy was deliberate given the high rate of developmental delays among youth with epilepsy that prevented some children fro m being able to fill out their own measures. Depression is an internalizing condition and may be under recognized by caregivers compared to externalizing problems. Similarly, the quality of ctioning across a range of domains. While physical functioning may be more overtly observed in daily interactions, difficulties in social or emotional functioning may be less apparent to outside observers or have less opportunity to be witnessed, especiall y as children get older (Eiser & Morse, 2001). Comparison of parent and child reports of HRQOL indicate that parents tend to endorse a lower HRQOL than their children (Baca et al., 2010; Verhey, Kulik, Ronen, Rosenbaum, Lach et al., 2009). HRQOL in pediat ric epilepsy is DelosReyes, Phillips et al., 2003), which was not assessed in our study.
70 Our sample was heterogenous in terms of seizure etiology, type, and control. While this i s positive for generalization to the greater epilepsy population it also makes interpreting results confusing as these factors may interact with medication to produce variability in side effect profiles, behavioral health factors, and psychosocial function ing. Future investigations with more homogenous samples might provide some clarification on these issues or identify a subgroup that would benefit the most from obesity intervention. For example, idiopathic epilepsies (e.g., absence, benign rolandic) have the best prognosis in terms of long term medical and psychosocial functioning (Shinnar & Pellock, 2004), but have also been associated with the highest rates of overweight and obesity (Daniels et al., 2009). Obesity as a comorbidity for these children may consequently be perceived as more impairing if it compromises their relatively high levels of functioning. Furthermore, a small proportion (14%) of the youth in our study was on polytherapy. This introduces a confound as to whether interactions between med ications differentially affect health behaviors (e.g., a weight positive and weight neutral drug wash out changes in appetite). Our inclusion of those individuals is consistent with other studies assessing the impact of AEDs on weight status, especially am ong newer AEDs that are often tested as add ons (Biton, 2003b), and is a limitation of the broader literature. On the other hand, polytherapy is common among youth with intractable epilepsy and so understanding the impact of medications both in isolation a nd in combination is clinically relevant. Finally, our hospital is a large tertiary care center with a wide catchment area. The distance to treatment likely impacted our study in two significant ways. First, many individuals from farther away opted to comp lete the follow up via mail which meant that
71 we were unable to verify objective measures of height and weight. Secondly, updated information (e.g., medication, comorbidities) was not available in their medical charts. Summary In closing, our study was the first to assess the role of behavioral health factors in AED induced weight gain among children with epilepsy. While methodological issues limited our ability to properly assess how dietary intake and physical activity related to weight status change durin g initial medical treatment when AED induced accelerations in growth velocity occur, information gleaned here is still important in documenting behaviors that are related to overall physical and psychosocial health. Future studies in larger samples targeti ng youth at diagnosis will be better able to elucidate the contribution of health behaviors as potentially explanatory mechanisms in the pathogenesis of AED induced weight gain. Our study was also important in highlighting the high prevalence of overweight /obesity as a comorbidity in this population, along with elevated rates of neurodev elopmental disorders, depressive symptoms and impaired quality of life compared to healthy peers. Continued efforts to understand the etiology of these issues in combinatio n with the development of evidence based interventions to mitigate these difficulties will be imperative to promote the best medical and psychosocial outcomes in this highly vulnerable population.
72 APPENDIX MEASURES Parent Measures Information about your Family Male / Female Caucasian African American Asian American Hispanic Bi racial Other (please specify):_________________ ________ 4. _____/_____/_____ _______ 6. Your (Parent) Gender (please circle): Male / Female 7. Your race (please circle) Caucasian African American Asian American Hispanic Bi racial Other (please specify):_________________ Mother _____ Father _____ Step Mother _____ Step Father _____ Grandparent _____ Other Legal Guardian _____ Age______ Height______ Weight______ 10. Please indicate your current living arrangement/marital status (please check one): Currently Married _____ Single, Co Habitating _____ Single, Divorced _____ Single, Widowed _____ Single, Never Married _____
73 11. Including yourself, how many adults live in your home: ________________ 12. Including your child, how many children live in your home: ______________ 13. What is the highest level (grade) of school you completed? Middle school _____ Some college _____ Some high school _____ Graduated college _____ Graduated high school _____ Post Graduate school _____ 14. What is your current occupation: _____________________________________________ 15. Estimat ed Family Income per Year (please check one). Below $9,999 _____ $50,000 $59,999 _____ $10,000 $19,999 _____ $60,000 $69,999 _____ $20,000 $29,999 _____ $70,000 $79,999 _____ $30,000 $39,999 _____ Over $80,000 _____ $ 40,000 $49,999 _____
74 1. Has a doctor ever told you that you have a heart condition and you should only do physical activity recommended by a physician? Yes_____ No ______ If yes, please specify: __________________________ ___________________ _____________________________________________ _____________________________________________ 2. Did a doctor ever say that you had hypertension or high blood pressure? (Not including high blood pressure only when you were pregnant.) Yes_____ No______ If yes, please answer the following: a. Did you ever take pills for high blood pressure? Yes_____ No ______ b. Do you now take pills for high blood pressure? Yes_____ No ______ 3. Did a doctor ever say that you h ad diabetes or high blood sugar? (Not including diabetes only when you were pregnant.) Yes_____ No______ If yes, please answer the following: a. Did you ever take pills or insulin for diabetes? Yes_____ No ______ b. Do you now take pills for diabetes? Yes_____ No ______ c. Do you now take insulin for diabetes? Yes_____ No ______ 4. Are you currently participating in another weight control program? Yes_____ No______ If yes, please list the name of the program (s) below:
75 _______________________________________________ 5. Have you previously undergone bariatric surgery? Yes_____ No ______ 6. If you are a female, please answer the following questions: a. Are you currently pregnant? Yes______ No______ b. Do you plan on becoming pregnant within the next year? Yes____ No____ 7. Has you ever been diagnosed with a depression, bipolar disorder, an anxiety disorder, schizophrenia or mental illness? Yes_____ No_____ a. If yes, please specify: ______________________________________________________ 8. Are you currently taking any medications? Yes_____ No_____ a. If yes, please list the medications below: ____________________________________________________________ ___________________________________________________________________ 9. How would you describe your weight Very underweight Slightly underweight About right Slightly overweight Very overweight
76 o ry 1. Has a doctor told you that your child has any of the following conditions (please mark yes or no for each condition). a. Epilepsy Yes_____ No _____ b. Chronic lung disease that limits physical activity Yes_____ No _____ c. Osteoporosis (weak, thin, or brittle bones ) Yes_____ No _____ d. Bone or muscle injury that limits physical activity Yes_____ No _____ 2. Did a doctor ever say that your child had diabetes or high blood sugar? Yes_____ No _____ If yes, please answer the following: a. Did your child ever take pills or insulin for diabetes? Yes_____ No _____ b. Does your child now take pills for diabetes? Yes_____ No _____ c. Does your child now take insulin for diabetes? Yes_____ No _____ 3. Has a doctor ever told you that your ch ild had heart problems? Yes_____ No _____ If yes, please answer the following questions: a. Heart valve problems (tight or narrowed valves or leaky valves) Yes_____ No _____ b. Heart valve surgery (artificial valve or repair of valve) Yes_____ N o _____ c. Atrial fibrillation (irregular beating of the heart requiring pills for treatment) Yes_____ No _____ d. If other, please describe: __________________________________________________ ______________________________________________________________________ ______________________________________________________________________
77 4. Did a doctor ever say that your child had/has hypertension or high blood pressure? Yes_____ No______ I f yes, please answer the following questions: a. Does your child currently take pills for high blood pressure? Yes_____ No _____ 5. Has your child ever been diagnosed with a developmental delay, autism, bipolar disorder, childhood schizophrenia or other major psychiatric disorder? Yes_____ No_____ a. If yes, please specify: ______________________________________________________ ______________________________________________________ 6. Is your child currently taking any medicati ons? Yes_____ No_____ a. If yes, please list them below: ________________________________________________________________ ______________________________________________________________________ 7. How would you describe weight? Very underweight Slightly underweight About right Slightly overweight Very overweight 8. Not concerned A little concerned Concerned Very concerned
78 9. Currently, how frequently does your c hild have seizures? Multiple times per day Daily Weekly Monthly Yearly Less than once per year
79 Chart Review and Administrative Forms
80 Retrospective Medical Chart Review Study ID: ____________________________________ Date of epilepsy diagnosis: _______________________ Seizure Type: __________________________________ Baseline Visit Date: _____________ Child Height: _________ Child weight: __________ Comorbid Medical Diagnoses 1) Dx: __________________________________ 2) Dx: __________________________________ 3) Dx: __________________________________ 4) Dx: __________________________________ 5) Dx: __________________________________ Psychiatric Diagnoses 1) Dx: ___________________________________ 2) Dx: ___________________________________ 3) Dx: ___________________________________ Medications 1) Name: _______________________ Dosage: ___________________ Date of initiation: __________ __ Date of termination: _________ Reason for termination: ______________________________________________ 2) Name: _______________________ Dosage: ___________________ Date of initiation: __________ ___ Date of termination: _________ Re ason for termination: ______________________________________________ 3) Name: _______________________ Dosage: ___________________ Date of initiation: __________ ___ Date of termination: _________ Reason for termination: _____________________ _________________________ 4) Name: _______________________ Dosage: ___________________ Date of initiation: __________ ___ Date of termination: _________ Reason for termination: ______________________________________________ 5) Name: _______________________ Dosage: ___________________ Date of initiation: __________ ___ Date of termination: _________ Reason for termination: ______________________________________________ 6) Name: _______________________ Dosage: ___ ________________ Date of initiation: __________ ___ Date of termination: __________
81 Reason for termination: ______________________________________________
82 Prospective Medical Chart Review Study ID: ___________________________ Prospective Visit 1 (approximately 4 6 months post baseline) Date: _____________ Child Height: _________ Child weight: __________ Comorbid Medical Diagnoses 6) Dx: __________________________________ 7) Dx: _________________ _________________ 8) Dx: __________________________________ 9) Dx: __________________________________ 10) Dx: __________________________________ Psychiatric Diagnoses 4) Dx: ___________________________________ 5) Dx: ___________________________________ 6) Dx: ___________________________________ Medications 7) Name: _______________________ Dosage: ___________________ Date of initiation: __________ ___ Date of termination: _________ Reason for termination: _________________________________________ _____ 8) Name: _______________________ Dosage: ___________________ Date of initiation: __________ ___ Date of termination: _________ Reason for termination: ______________________________________________ 9) Name: _______________________ Dosa ge: ___________________ Date of initiation: __________ ___ Date of termination: _________ Reason for termination: ______________________________________________ 10) Name: _______________________ Dosage: ___________________ Date of init iation: __________ ___ Date of termination: _________ Reason for termination: ______________________________________________ 11) Name: _______________________ Dosage: ___________________ Date of initiation: __________ ___ Date of termination: _________ Reason for termination: ______________________________________________ 12) Name: _______________________ Dosage: ___________________ Date of initiation: __________ ___ Date of termination: __________ Reason for termination: ______________________________________________
83 LIST OF REFERENCES Ablah, E., Haug, A., Konda, K., Tinius, A.M., Ram, S., Sadler, T., & Liow, K. (2009). Exercise and epilepsy: a survey of Midwest epilepsy patients. Epilepsy & Behavior, 14(1), 162 166. Ahmed, B., Fatmi, Z., Siddiqui, A.R., Sheikh, A.L. (20 11). Predictors of unintentional poisoning among children 5 years of age in Karachi: a matched case control study. Injury Prevention, 17, 27 32. American Academy of Pediatrics Committee on Children with Handicaps and Committee on Sports Medicine (1983). Sports and the child with epilepsy. Pediatrics, 72, 884 885. Arida, R.M., Cavalheiro, E.A., da Silva, A.C., & Scorza, F.A. (2008). Physical activity and epilepsy: proven and predicted benefits. Sports Med icine, 38(7), 607 615. Aydin, K., Serdaroglu, A., Okuyaz, C., Bideci, A., & Gucuyener, K. (2005). Serum insulin, leptin, and neuropeptide y levels in epileptic children treated with valproate. Journal of Child Neurology, 20, 848 851. Baca, C.B., Vickrey, B.G., Calplan, R., Vassar, S.D., & Berg, A.T. (2011). Psychiatric and medical comorbidity and quality of life outcomes in child onset epilepsy. Pediatrics, 128, e1532 1543. Baron & Kenny (1986). The moderator mediator variable distinction in social psychol ogical research: conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173 1182. Batzel, L.W., Dodrill, C.B., Dubinsky, B.L., Ziegler, R.G., Connolly, J.E., Freeman, R.D., et al. (2007). An objective m ethod for the assessment of psychosocial problems in adolescents with epilepsy. Epilepsia, 32(2), 202 211. Benavente Aguilar, I., Morales Blanquez, C., Rubio, E.A., & Rey, J.M. (2004). Quality of life of adolescents suffering from epilepsy living in the co mmunity. Journal of Pediatrics and Child Health, 40(3), 110 113. Berg, A.T., & Shinnar, S. (1991). The risk of seizure recurrence following a first unprovoked seizure. Neurology, 41, 965. Berg, A.T., Shinnar, S., Levy, S.R., Testa, F.M., Smith Rapaport S., Beckerman, B., & Ebrahimi, N. (2001). Defining early seizure outcomes in pediatric epilepsy: the good, the bad, and the in between. Epilepsy Research, 43, 75 84. Bergqvuist, C., Trabulsi, J., Schall, J.I., & Stallings, V.A. (2008). Growth failure in children with intractable epilepsy is not due to increased resting energy expenditure. Developmental Medicine & Children Neurology, 50, 439 444.
84 Biton, V. (2003). Lamotrigine versus valproate monotherapy associated weight change in adolescents with epileps y: results from a post hoc analysis of a randomized, double blind clinical trial. Journal of Child Neurology, 18, 133 139. Biton, V. (2003). Effect of antiepileptic drugs on bodyweight: overview and clinical implications for the treatment of epilepsy. CNS Drugs, 17, 781 791. Biton, V., Mirza, W., Montouris, G., Vuong, A., Hammer, A.E., & Barrett, P.S. (2001). Weight change associated with valproate and lamotrigine monotherapy in patients with epilepsy. Neurology, 56, 172 177. Bjorholt, P.G., Nakken, K.O., R ohme, K., & Hansen., H. (1990). Leisure time habits and physical fitness in adults with epilepsy. Epilepsia, 31, 83 87. Camfield, C., Breau, L., & Camfield, P. (2001). Impact of pediatric epilepsy on the family: a new scale for clinical and research use. E pilepsia, 42(1), 104 112. Campbell, J.P., Maxey, V.A., & Watson, W.A. (1994). Hawthorne effect: implications for prehospital research. Annals of Emergency Medicine, 26(5), 590 594. Christensen, J., Vestergaard, M., Mortensen, P. B., Sidenius, P., & Agerbo E. (2007). Epilepsy and risk of suicide: a population based case control study. Lancet Neurol, 6(8) 693 698. Christoffel, K.K., Donovan, M., Schofer, J., Wills, K., & Lavigne, J.L. (1996). Psychosocial factors in childhood pedestrian injury: a matched c ase control study. Pediatrics, 97(1), 33 42. Cole, T.J., Bellizi, M.C., Flegal, K.M., & Dietz, W.H. (2000). Establishing a standard definition for child overweight and obesity worldwide: international survey. British Medical Journal, 320, 1 6. Corman, C L., Leung, N.M., Guberman, A.H. (1997). Weight gain in epileptic patients during treatment with valproic acid: a retrospective study. Canadian Journal of Neurological Sciences, 24(3), 240 244. Daniels, Z.S., Nick, T.G., Liu, C., Cassedy, A., & Glauser, T.A (2009). Obesity is a common comorbidity for pediatric patients with untreated, newly diagnosed epilepsy. Neurology, 73, 658 664. Davison, K.K., & Birch, L.L. (2001). Weight status, parent reaction, and self concept in five year old girls Pediatrics, 1 07 46 53. Demir, E., & Aysun, S. (2000). Weight gain associated with valproate in childhood. Pediatric Neurology, 22(5), 361 364.
85 Dencker, M., Thorsson, O., Karlsson, M.K., Linden, C., Svenssen, J., Wallmer, P., & Andersen, L.B. (2006). Daily physical act ivity and its relation to aerobic fitness in children aged 8 11 years. European Journal of Applied Physiology, 96(5), 587 592. Dietz, W.H. (1994). Critical periods in childhood for the development of obesity. The American Journal of Clinical Nutrition, 59, 955 959. Doyle, A.C., le Grange, D., Goldschmidt, A., & Wilfley, D.E. (2007). Psychosocial and physical impairment in overweight adolescents at high risk for eating disorders. Obesity, 15(1), 145 154. Tear, & Verity, C. (1997). Weight gain with valproate or carbamazepine: a reappraisal. Seizure, 6(2), 121 125. Ekelund, U., Luan, J., Sherar, L.B., Eslinger, D.W., Griew, P., & Cooper, A. (2012). Moderate to vigorous physical activity and sedentary time and cardiometabolic risk factors in children and adolescents. Journal of the American Medical Association, 307(7), 704 712. Egger, J., & Brett, E.M. (1981). Effects of sodium valproate in 100 children with special reference to weight. British Medical Journal, 283, 577 581. Eiser, C. & Morse, Results of a systematic review. Quality of Life Research, 10, 347 357. Elliot, I.M., Lach, L., & Smith, M.L. (2000). Adolescent and maternal perspectives on HRQOL and neuropsycholog ical status following epilepsy surgery. Epilepsy and Behavior, 1(6), 406 417. Epstein, Paluch, Roemmich, & Beecher (2007). Family based obesity treatment, then and now: Twenty five years of pediatric obesity treatment. Health Psychology, 26(4), 381 391. Er icksen, H.R., Ellertsen, B., Gronningsaeter, H., Nakkan, K.O., Loyning, Y., & Ursin, H., (1994). Physical exercise in women with intractable epilepsy. 35(6), 1256 1264. Erermis, S., Cetin, N., Tamar, M., Bukusoglu, N., Akdeniz, F., & Goksen, D. (2004). Is obesity a risk factor for psychopathology among adolescents? Pediatrics International, 46, 296 301. Fagot Campagna A., Pettitt D.J., Engelgau M.M., Burrows N.R., Geiss L.S., Valdez R., et al. (2000). Type 2 diabetes among North American children and adol escents: An epidemiologic review and a public health perspective. Journal of Pediatrics, 136, 664 672.
86 Grosso, S., Mostardini, R., Piccini, B. & Balestri, P. (2009). Body mass index and serum lipid changes during treatment with valproic acid in children w ith epilepsy. The Annuals of Pharmacotherapy, 43, 45 50. Holland, K. D., & Glauser, T. A. (2007). Response to carbamazepine in children with newly diagnosed partial onset epilepsy. Neurology, 69 (6), 596 599. Ingerski, L.M., Modi, A.C., Hood, K.K., Pai A.L., Zeller, M., Piazza Waggoner, C., Driscoll, K.A., Rothenberg, M.E., Franciosi, J., & Hommel., K.A. (2010). Health related quality of life across pediatric conditions. Journal of Pediatrics, 156, 639 644. Isojarvi, J., Laatikainen, T., Knip, M., Paka rinen, A., Juntunen, K., & Myllyla, V. (2004). Obesity and endocrine disorders in women taking valproate for epilepsy. Annals of Neurology, 39(5), 579 584. Isojrvi, J.I., Rtty, J., Myllyl, V.V., Knip, M., Koivunen, R., Pakarinen, A.J., Tekay, A., & Tap anainen, J.S. (1998). Valproate, lamotrigine, and insulin mediated risks in women with epilepsy. Annals of Neurology, 43(4), 446 451. Jallon, P. & Picard., F. (2001). Bodyweight gain and anticonvulsants: a comparative review. Drug Safety, 24(13), 969 978. Janicke, Sallinen, Perri, Lutes, Huerta, Silverstein, & Brumback (2008). Comparison of parent only vs family based interventions for overweight children in underserved rural settings. Archives of Pediatric and Adolescent Medicine, 162(12), 1119 1125. Leib owitz, S.F. (1992). Neurochemical neuroendocrine systems in the brain controlling macronutrient intake and metabolism. Trends in Neurosciences, 15, 491 497. Leonard & George (1999). Psychosocial and neuropsy chological function in children with epilepsy. Pe diatric Rehabilitation, 3(3), 73 80. Leppik, I.E. (1998). Contemporary diagnosis and management of the patient with epilepsy. Newtown, Pennsylvania: Handbooks in Health Care. Leuf, G., Abraham, I., Haslinger, M., Trinka, E., Seppi, K., Unterberger, I., Alg e, A., Windisch, J., Lechleitner, M., & Bauer, G., (2002). Polycystic ovaries, obesity and insulin resistance in women with epilepsy. Journal of Neurology, 249, 835 841. Liu, Y.M.C., Williams, S., Basualdo Hammond, C., Stephens, D., & Curtis, R. (2003). A prospective study: growth and nutritionals status of children treated with the ketogenic diet. Journal of the American Dietetic Association, 103(6), 707 712. McDowell, M.A., Briefel, R.R., Alaimo, K., Bischof, A.M., Caughman, C.R., Carroll, M.D., Loria, C. M., & Johnson, C.L. (1994). Energy and macronutrient intakes of persons 2 months and over in the United States: Third National Health and Nutrition Survey, Phase 1, 1988 91. Advance Data, 255, 1 24.
87 McEwan, M.J., Espie, C.A., Metcalfe, J., Brodie, M.J., & Wilson, M.T. (2004). Quality of life and psychosocial development in adolescents with epilepsy: a qualitative investigation using focus group methods. Seizure, 13, 15 31. Modi, A.C., King, A.S., Monahan, S.R., Koumoutsos, J.E., Morita, D.A., & Glauser, T. A. (2009). Even a single seizure negatively impacts pediatric health related quality of life. Epilepsia, 50(9), 2110 2116. Modi, A.C., Ingerski, L.M., Rausch, J.R., & Glauser, T.A. (2011). Treatment factors affecting longitudinal quality of life in new ons et pediatric epilepsy. Journal of Pediatric Psychology, 36(4), 466 475. Must, A. (1996). Morbidity and mortality associated with elevated body weight in children and adolescents. American Journal of Clinical Nutrition, 63 445S 447S. Nakken, K.O., Bjorh olt, P.G., Johannessen, S.I., Loyning, T., & Lind, E. (1990). Effect of physical training on aerobic capacity, seizure occurrence, and serum level of antiepileptic drugs in adults with epilepsy. Epilepsia, 31, 88 94. Neal, E.G., Chaffe, H., Schwartz, R.H., Lawson, M.S., Edwards, N., Fitzsimmons, G., Whitney, A., & Cross, J.H. (2008). The ketogenic diet for the treatment of childhood epilepsy: a randomised controlled trial. The Lancet Neurology, 7(6), 500 506. Novak, G.P., Maytal, J., Alshansky, A., Eviatar, L., Sy Kho, R., Siddique, Q. (1999). Risk of excessive weight gain in epileptic children treated with valproate. Journal of Child Neurology, 14, 490 495. Ogden, C.L., Carroll, M.D., Kit, B.K., & Flegal, K.M. (2012). Prevalence of Obesity and Trends in Bod y Mass Index Among US Children and Adolescents, 1999 2010. Journal of the American Medical Association, 307(5), 483 490. Pellock (2004). Defining the problem: psychiatric and behavioral comorbidity in children and adolescents with epilepsy. Epilepsy & Beha vior, 5, S3 S9. Pesa, J.A., Syre, T.R., & Jones, E. (2000). Psychosocial differences associated with body weight among female adolescents: the importance of body image. Journal of Adolescent Health, 26 330 337. Pierce & Wardle (1997). Cause and effect b eliefs and self esteem of overweight children. Journal of Child Psychology and Psychotherapy, 38 645 650. Plioplys, S. (2003). Depression in children and adolescents with epilepsy. Epilepsy & Behavior, 4, S39 S45. Puhl, R.M. & Latner, J.D. (2007). Stigma obesity, and the health of the nation s children. Psychological Bulletin, 133(4) 557 580.
88 Rutter, M., Graham, P., & Yule, W. (1970). A neuropsychiatric study in childhood. Clinics in Developmental Medicine. Nos. 35 36. Heinneman/S.I.M.P.: London. Sabaz, M., Cairns, D.R., Bleasel, A.F., Lawson, J.A., Grinton, B., Scheffer, I.E., & Bye, A.M.E. (2003). The health related quality of life of childhood epilepsy syndromes. Journal of Pediatrics and Child Health, 39(9), 690 696. Salpekar & Dunn (2007). Psychiatr ic and psychosocial consequences of pediatric epilepsy. Seminars in Pediatric Neurology, 14, 181 188. Schwimmer, Burwinkle, & Varni (2003). Health related quality of life of severely obese children and adolescents. Journal of the American Medical Associati on, 289(14), 1813 1819. Shinnar, S. & Pellock, J.M. (2002). Update on the epidemiology and prognosis of pediatric epilepsy. Journal of Child Neurology, 17, S4 S17. Sillanp M. (1992). Epilepsy in children: prevalence, disability, and handicap. Epilepsia 33(3), 44 449. Sillanpaa, M., Jalava, M., Kaleva, O., & Shinnar, S. (1998). Long term prognosis of seizures with onset in childhood. New England Journal of Medicine, 338 (24), 1715 1722. Sillanp, M., Jalava, M., & Shinnar, S. (1999). Epilepsy syndromes in patients with childhood onset seizures in Finland. Pediatric Neurology, 21 533 537. Smith, M.L., Elliot, I.M., Lach, L. (2004). Cognitive, psychosocial, and family function one year after pediatric epilepsy surgery. Epilepsia, 45(6), 650 660. Sonuga B arke, E.J.S., Coghill, D., Wigal, T., DeBacker, M., & Swanson, J. (2009). Adverse reactions to methylphenidate treatment for attention deficit/hyperactivity disorder: structure and associations with clinical characteristics and symptom control. Journal of Child and Adolescent Psychopharmacology, 19(6), 683 690. Spear, Barlow, Ervin, Ludwig, Saelens, Schetzina, & Taveras (2007). Recommendations for treatment of child and adolescent overweight and obesity. Pediatrics, 120(4), S254. Steinhoff, B.J., Neussiiss, K., Thegeder, H., & Reimers, C.D. (2005). Leisure time activity and physical fitness in patients with epilepsy. Epilepsia, 37(12), 1221 1227. Stephen, L.J., Sills, G., Leach, J.P. Butler, E., Parker, P., Hitiris, N., et al. (2007). Sodium valproate versus lamotrigine: A randomized comparison of efficacy, tolerability and effects on circulating androgenic hormones in newly diagnosed epilepsy. Epilepsy Research, 75, 122 129. Strauss, R.S. (1999). Childhood obesity. Current Problems in Pediatrics, 29, 1 29.
89 T roiano, R.P., Berrigan, D., Dodd, K.W., Masse, Tilert & MacDowell (2008). Physical activity in the United States measured by accelerometer. Medicine and Science in Sports and Exercise, (40)1, 181 188 Tudor Locke, C., Williams, J.E., Reis, J.P., & Pluto, D. (2002). Utility of pedometers for assessing physical activity: convergent validity. Sports Medicine, 32(12), 795 808. U.S. Department of Agriculture and U.S. Department of Health and Human Services. Dietary Guidelines for Americans, 2010. 7 th Edition, Washington, DC: U.S. Government Printing Office, December 2010. Van Empelen, R., Jennekens Schinkel, A., van Rijen, P.C., Helders, P.J.M. & van Nieuwenhuizen, O. (2005). Health related quality of life and self perceived competence of children assessed befo re and up to two years after epilepsy surgery. Epilepsia, 46(2), 258 271. Vanina, Y., Podolskaya, A., Sedky, K., Shahab, H., Siddiqui, A., Munshi, F., Lippmann, S. (2002). Bodyweight changes associated with psychopharmacology. Psychiatric Services, 53(7), 842 847. Varni, J.W., Seid, M., & Kurtin, P.S. (2001). PedsQL(TM) 4.0: Reliability and Validity of the Pediatric Quality of Life Inventory(TM) Version 4.0 Generic Core Scales in Healthy and Patient Populations. Medical Care, 39(8), 800 812. Vehmas, T. (19 97). Hawthorne effect: shortening of fluoroscopy times during radiation measurement studies. The British Journal of Radiology, 70, 1053 1055. Verhey, L.H., Kulik, D.M., Ronen, G.M., Rosenbaum, P., Lach, L., & Streiner, D.L. (2009). Quality of life in child hood epilepsy: what is the level of agreement between youth and their parents? Epilepsy & Behavior, 14(2), 407 410. Verotti, A., Basciani, F., De Simone, M., Trotta, D., Morgese, G., & Chiarelli, F. (2002). Insulin resistance in epileptic girls who gain we ight after therapy with valproic acid. Journal of Child Neurology, 17, 265 268. Volpe, S.L., Schall, J.I., Gallagher, P.R., Stallings, V.A., Bergqvist, C. (2007). Nutrient intake of children with intractable epilepsy compared to healthy children. Journal o f the American Dietetic Association, 107, 1014 1018. Weiss, R., Dzuira, J., Burgert, T.S., Tamborlane, W.V., Taksali, S.E., Yeckel, C.W., Allen, K., Lopes, M., Savoye, M., Morrison, J., Sherwin, R.S., & Caprio, S. (2004). Obesity and metabolic syndrome in children and adolescents. New England Journal of Medicine, 350, 2362 2374. Wilfley, Tibbs, Van Buren, Reach, Walker, & Epstein (2007). Lifestyle interventions in the treatment of childhood overweight: A meta analytic review of randomized controlled trials. Health Psychology, 26(5), 521 532.
90 Williams, J., Steel, C., Sharp, G.B., DelosReyes, E., Phillips, T., Bates, S., Lange, B., & Gabriel, M.L. (2003). Parental anxiety and quality of life in children with epilepsy. Epilepsy and Behavior, 4(5), 483 486. Wirr ell, E.C. (2003). Valproic acid associated weight gain in older children and teens with epilepsy. Pediatric Neurology, 28(2), 126 129. Wirrell, E.C., Camfield, C.S., Camfield, P.R., Dooley, J.M., Gordon, K.E. & Smith, B. (1997). Long term psychosocial outc ome in typical absence epilepsy: sometimes a Archives of Pediatric and Adolescent Medicine, 151(2), 152 158. Wong, J. & Wirrell, E. (2006). Physical activity in children/teens with epilepsy compared to that in their siblings witho ut epilepsy. Epilepsia, 47(3), 631 639. Zametkin, A.J., Zoon, C.K., Klein, H.W., & Munson, S. (2004). Psychiatric aspects of child and adolescent obesity: A review of the past 10 years. Journal of the American Academy of Child & Adolescent Psychiatry, 43(2 ), 134 150. Zeller & Modi (2008). Psychosocial factors related to obesity in children and adolescents. In E. Jelalian and R. Steele (Eds.), Handbook of Child and Adolescent Obesity. Zimmerman, U., Kraus, T., Himmerich, H., Schuld, A., & Pollmacher, T (2003). Epidemiology, implications and mechanisms underlying drug induced weight gain in psychiatric patients. Journal of Psychiatric Research, 37, 193 220.
91 BIOGRAPHICAL SKETCH Katherine Wells Follansbee Junger graduated Summa Cum Laude from the University of Vermont with a Bachelor of Arts in p sychology. Following a 3 year post baccalaureate position at Brown University, she came to the University of Florida to study pediatric p s ychology. She received her Ph D. in clinical p sychology in the summe r of 2012.