ABSTRACT Title of thesis: INDEPENDENT AND JOINT EFFECTS OF PARENTAL ATTITUDES AND SPECIAL HEALTH CARE NEEDS ON PHYSICAL ACTIVITY AND SCREEN TIME AMONG CHILDREN AND ADOLESCENTS IN THE UNITED STATES Janet Ann Gingold, Masters of Public Health, 2012 Thesis directed by: Professor Olivia Carter-Pokras Department of Epidemiology and Biostatistics Sedentary lifestyles pose a threat to the health of children, especially those with special health care needs (SHCN). Using data from the 2007 National Survey of Children's Health, this study examined relationships between parental attitudes and low physical activity and high screen time among 6- to 17-year-olds with and without SHCN. Perceived limitation was associated with increased likelihood of low physical activity (AOR, 1.339; 95%CI, 1.079-1.662). Parenting stress (AOR, 1.189; 95%CI, 1.052-1.344) and lack of trust (AOR, 1.243; 95%CI, 1.104-1.399) were associated with increased likelihood of high screen time. Perceived limitation modified the effect of special health care needs status on high screen time. The likelihood of combined low physical activity and high screen time was greatest among children with SHCN whose parents reported both functional limitations in the child and parenting stress (AOR, 2.659; 95%CI, 1.741- 4.060). Parental attitudes and SHCN should be addressed in interventions to promote active lifestyles. INDEPENDENT AND JOINT EFFECTS OF PARENTAL ATTITUDES AND SPECIAL HEALTH CARE NEEDS ON PHYSICAL ACTIVITY AND SCREEN TIME AMONG CHILDREN AND ADOLESCENTS IN THE UNITED STATES by Janet A. Gingold Thesis submitted to the Faculty of the Graduate School of the University of Maryland, College Park in partial fulfillment of the requirements for the degree of Masters in Public Health 2012 Advisory Committee: Professor Olivia Carter-Pokras, Chair Professor Brit Saksvig Professor Tong Tong Wu ?Copyright by Janet Ann Gingold 2012 ii ACKNOWLEDGEMENTS Many thanks to my committee, Professor Olivia Carter-Pokras, Professor Brit Saksvig and Professor Tong Tong Wu, and to my family and friends for their invaluable support. iii Table of Contents ACKNOWLEDGEMENTS ................................................................................................ ii List of Tables ..................................................................................................................... vi List of Illustrations ........................................................................................................... viii List of Abbreviations ......................................................................................................... ix I. INTRODUCTION ........................................................................................................... 1 II. RESEARCH QUESTIONS AND HYPOTHESES ....................................................... 3 Question #1 ..................................................................................................................... 4 Question #2 ..................................................................................................................... 4 Question #3 ..................................................................................................................... 5 Question #4 ..................................................................................................................... 5 III. BACKGROUND: REVIEW OF LITERATURE ......................................................... 7 1. Relationships between physical activity and screen time and childhood obesity ....... 7 2. Parental influence on children's physical activity and screen time ............................. 8 3. The independent variables ........................................................................................ 10 a. Identifying children with special health care needs .............................................. 10 b. Parental perceptions of child's limitations ............................................................ 13 c. Parental mental health and parenting stress .......................................................... 14 d. Parental social support and trust in neighbors ...................................................... 15 e. Parental perception of child's safety ...................................................................... 15 iv 4. Measuring physical activity and screen time ............................................................ 16 IV. RESEARCH DESIGN AND METHODS .................................................................. 19 1. Study Population ....................................................................................................... 19 2. Defining dependent variables ................................................................................... 20 3. Description of independent variables ........................................................................ 20 4. Data analysis ............................................................................................................. 22 5. Human Subjects ........................................................................................................ 23 V. RESULTS .................................................................................................................... 25 1. Descriptive statistics ................................................................................................. 25 a. Characteristics of the population ........................................................................... 25 b. Low moderate-to-vigorous physical activity ........................................................ 26 c. High screen time ................................................................................................... 28 d. Sedentary life style ................................................................................................ 29 e. Associations of covariates with special health care needs status .......................... 30 f. Correlations between covariates ............................................................................ 31 2. Hierarchical logistic regression models .................................................................... 31 a. Low moderate-to-vigorous physical activity ........................................................ 32 c. High screen time ................................................................................................... 34 c. Sedentary lifestyle ................................................................................................. 36 3. Testing for effect modification ................................................................................. 37 v a. Low moderate-to-vigorous physical activity ........................................................ 38 b. High screen time ................................................................................................... 39 c. Sedentary lifestyle ................................................................................................ 40 4. Joint effects ............................................................................................................... 42 VI. DISCUSSION ............................................................................................................. 44 1. Association between parental attitudes and low MVPA ........................................... 44 2. Association between parental attitudes and high screen time ................................... 44 3. Interaction between parental attitudes and special health care needs status ............. 44 4. Joint effects of parental attitudes and SHCN on sedentary lifestyle ......................... 46 5. Summary and implications ....................................................................................... 47 6. Strengths and Limitations ......................................................................................... 49 7. Public health significance ......................................................................................... 53 VII. APPENDIX 1: TABLES ........................................................................................... 56 VIII. APPENDIX 2: ILLUSTRATIONS ......................................................................... 81 IX. BIBLIOGRAPHY....................................................................................................... 95 vi List of Tables Table 1. Socio-demographic characteristics of the 2007 National Survey of Children's Health study population, with population percents. United States, 2007 Table 2. Frequency of specific diagnostic categories listed in National Survey of Children's Health 2007, with population estimates Table 3. Weighted prevalence of low physical activity, high screen time and sedentary lifestyle by selected demographic characteristics, parental attitude indicators and special health care needs status. United States, 2007 Table 4. Weighted prevalence of special health care needs status by selected demographic characteristics and parental attitude indicators Table 5. Prevalence of low moderate-to-vigorous physical activity among 6- to 17-year-olds by special health care needs status and parental attitude indicators. United States 2007 Table 6. Hierarchical logistic regression models for effects of demographic characteristics, special health care needs status and parental attitudes on low moderate-to-vigorous physical activity in 6-17 year old children. United States 2007. Table 7. Prevalence of high screen time among 6- to 17-year-olds by special health care needs status and parental attitude indicators. United States 2007 Table 8. Hierarchical logistic regression models for effects of demographic characteristics, special health care needs status and parental attitudes on high screen time in 6-17 year old children. United States 2007 Table 9. Prevalence of sedentary lifestyle among 6- to 17-year-olds, by health care needs status and parental attitude indicators. United States 2007 Table 10. Hierarchical logistic regression models for effects of demographic characteristics, special health care needs status and parental attitudes on sedentary lifestyle in 6-17 year old children. United States 2007 Table 11. Effects of interaction terms in models for effects of parental attitudes, adjusted for age, gender, race/ethnicity, education of respondent and special health care needs status Table 12. Effects of parental attitudes on odds of low moderate-to-vigorous physical activity, adjusted for age, gender, race/ethnicity and education of 56 57 58 60 62 63 65 66 68 69 71 72 vii respondent, stratified by special health care needs status Table 13. Effect of special health care needs status on adjusted odds of low moderate-to-vigorous physical activity, adjusted for age, gender, race/ethnicity and education of respondent, stratified by parental attitudes Table 14. Effects of parental attitudes on odds of exceeding 2 hours of screen- based leisure activity, adjusted for age, gender, race/ethnicity and education of respondent, stratified by special health care needs status Table 15. Effect of special health care needs status on odds of exceeding 2 hours of screen-based leisure activity, adjusted for age, gender, race/ethnicity, education of respondent, stratified by parental attitudes Table 16. Effects of parental attitudes on odds of sedentary lifestyle adjusted for age, gender, race/ethnicity and education of respondent, stratified by special health care needs status Table 17. Effect of special health care needs status on odds of sedentary lifestyle, adjusted for age, gender, race/ethnicity and respondent's education, stratified by parental attitude indicators Table 18. Joint effects of single attitudes with SCHN with or without EBD on the adjusted odds of sedentary lifestyle among 6- to 17-year-olds. United States 2007 Table 19. Joint effects of combinations of parental attitudes on the odds of sedentary life style among 6-17 year olds. United States, 2007 Table 20. Joint effects of constellations of parental attitudes and special health care needs status on the odds of sedentary lifestyle among 6-17 year olds. United States, 2007 73 74 75 76 77 78 79 80 viii List of Illustrations Figure 1. Conceptual framework 81 Figure 2. Special health care needs categories (percent of population) 82 Figure 3. Estimated population prevalence (percent) of select chronic conditions (NSCH 2007) 83 Figure 4. Change in prevalence of low MVPA, high screen time and both with age 84 Figure 5. Prevalence of low MVPA, high screen time and sedentary lifestyle in children and adolescents, by race and ethnicity. 85 Figure 6. Prevalence of low MVPA, high screen time and sedentary lifestyle by special health care needs status. 86 Figure 7. Relationships between parental attitudes and low MVPA 87 Figure 8. Relationships between parental attitudes and high screen time 88 Figure 9. Relationships between parental attitudes and sedentary lifestyle 89 Figure 10. The effect of perceived limitations on high screen time, stratified by special health care needs status 90 Figure 11. The effect of SHCN on high screen time, stratified by limitations 91 Figure 12. Joint effects of individual parental attitudes and SHCN with and without EBD on sedentary lifestyle 92 Figure 13. Joint effects of parental attitudes on sedentary lifestyle 93 Figure 14. Joint effects of constellations of parental attitudes and special health care needs on sedentary lifestyle 94 ix List of Abbreviations AAP American Academy of Pediatrics AOR Adjusted odds ratio BMI Body mass index CDC Centers for Disease Control and Prevention CI Confidence interval CSHCN Child with special health care needs EBD Emotional, behavioral and developmental conditions HRSA Health Resources and Services Administration MVPA Moderate-to-vigorous physical activity NCHS National Center for Health Statistics NHANES National Health and Nutrition Examination Survey NIS National Immunization Survey NSCH National Survey of Children's Health OR Odds ratio PA Physical activity SBLA Screen-based leisure activity SE Standard error SHCN Special health care needs SLAITS State and Local Area Integrated Telephone Survey YRBS Youth Risk Behavior Survey 1 I. INTRODUCTION All children need physical activity to maintain an appropriate energy balance and to promote healthy growth and development. Physical activity stimulates physiologic and anatomic adaptations that improve the strength of muscle and bone while building cardio- respiratory capacity. Fundamental physical skills learned in childhood pave the way for continued physical activity across the life course. Because people tend to become less active as they get older and inadequate activity is associated with numerous chronic conditions, promoting active lifestyles during childhood is essential for lifelong health.1,2 For children with special health care needs (SHCN), finding appropriate opportunities for physical activity creates extra challenges for parents.3 While most children with SHCN are able to be physically active, their health care needs might affect their choice of activities because of physical limitations or because of their own preferences and the expectations of the adults in their lives. For children with SHCN, it is especially important to optimize physical activity, not only to minimize the impact of the existing condition on quality of life, but also to decrease the likelihood of developing co- morbidities, such as obesity and diabetes.3,4 Children?s participation in various activities is highly dependent upon their parents? perceptions and attitudes about what they can and should do.5,6,7 The demands of parenting a child with SHCN create stresses that challenge parents? coping abilities and affect mental health.3,8,9,10,11 When there are many competing priorities, sedentary activities, including screen-based leisure activity (screen time), sometimes displace more physically active pursuits. Parental perceptions of their child?s abilities and limitations, confidence in their ability to make good decisions about their child's activities, and their 2 trust in other adults who might interact with the child are potentially modifiable factors that can affect children's opportunities to engage in appropriate activities. Better understanding of the relationships between parental attitudes and the daily activities of children can facilitate interventions that address the needs of the family so that children with and without SCHN can engage in activities that optimize their growth and development while preventing obesity and its co-morbidities. 3 II. RESEARCH QUESTIONS AND HYPOTHESES According to the 2008 Physical Activity Guidelines for Americans, children should engage in at least one hour of moderate-to-vigorous physical activity (MVPA) daily, and this should include at least 20 minutes of vigorous activity at least three times a week.1,2 However, only 18.4% of adolescents met this guideline in 2009.12 The American Academy of Pediatrics (AAP) recommends that children engage in no more than two hours of screen-based leisure activities (SBLA) daily.13,14 While estimates of current media use vary widely by how it is measured, a Kaiser Family Foundation report estimates that in 2009, the average 8-18 year old spent about four hours a day with television, videos, movies, video games and recreational computer use.15 Compared with children without SHCN, children SHCN have been found to spend less time engaged in MVPA, to spend more time engaged in SBLA, and to have a higher prevalence of overweight and obesity.4 Because parents play a central role in creating opportunities for their children to engage in activities that promote optimal growth and development, understanding the relationship between parental attitudes and children's activities is essential for developing family-based interventions for this high-risk subpopulation. Using data on 6- to17-year-old US children from the 2007 National Survey of Children's Health (NSCH),16 we investigated the independent and joint effects of various parental attitudes and special health care needs status (SHCN with emotional, behavioral or developmental condition (EBD), SHCN without EBD or no SHCN) on children?s engagement in physical activity and screen-based leisure activity. We predicted that the parent?s perception of the child?s limitations, the parent?s mental health and perceived 4 stress due to parenting, the parent?s perception of social support and trust in neighbors and the parent?s perception of the child?s safety would be associated with varying levels of physical activity and screen-based leisure activity, and that the strength of these associations would vary with special health care needs status. Question #1: Are parental attitudes (perceptions of their child?s limitations, the stress of parenting, social support, trust in neighbors and perception of child?s safety) associated with the child?s engagement in adequate physical activity? Null Hypothesis #1: The proportion of children who do not engage in adequate physical activity will not vary with parental attitudes. Hypothesis #1: The proportion of children who do not engage in adequate physical activity will be greater among households where parents? perceptions of the child?s limitations are greater, where parents report greater stress of parenting and poorer mental health, where parents report less social support and less trust in neighbors, and where parents perceive their child as less safe, compared with households with more positive attitudes. Question #2: Are parental attitudes (perceptions of their child?s limitations, the stress of parenting, social support, trust in neighbors and perception of child?s safety) associated with the child?s engagement in leisure-based screen activities? Null Hypothesis #2: The proportion of children who engage in excessive screen- based leisure activity will not vary with parental attitudes. Alternative Hypothesis #2: The proportion of children who engage in excessive screen-based leisure activity will be greater among households where parents? perceptions of the child?s limitations are greater, where parents report greater 5 stress of parenting and poorer mental health, where parents report less social support and less trust in neighbors, and where parents perceive their child as less safe, compared with households with more positive parental attitudes. Question #3: Do parental attitudes (perceptions of their child?s limitations, the stress of parenting, social support, trust in neighbors and perception of child?s safety) modify the effect of special health care needs status on children's engagement in physical activity and screen-based leisure activities? Null Hypothesis #3: The association between SHCN status and child?s engagement in adequate physical activity and excessive screen-based leisure activity will not vary with parental attitudes. Alternative Hypothesis #3: The presence of unfavorable parental attitudes will increase the proportion of children who engage in inadequate physical activity and excessive screen-based leisure activity for children in each of three special health care needs categories (no SHCN, SHCN without EBD and SHCN with EBD). The effect will be greatest for children with SHCN with EBD and least for those with no SCHN. Question #4: Which constellation of parental attitudes and special health care needs places children is associated with the greatest likelihood of combined inadequate physical activity and excessive screen-based leisure activity? Null Hypothesis #4: The odds of engaging in both inadequate physical activity and excessive screen-based leisure activity will not change with parental attitudes or SHCN status. 6 Alternative Hypothesis #4: The odds of engaging in both inadequate physical activity and excessive screen-based leisure activity will be significantly greater for those children who have both SCHN with EBD and parents who perceive their child's limitations as greater, who have poorer mental health and greater stress due to parenting, who have less social support and trust in neighbors, and who perceive their child as less safe. We aimed to shed light on parental factors that could be modified in future family- based interventions for promoting more active lifestyles to optimize growth and development. 7 III. BACKGROUND: REVIEW OF LITERATURE 1. Relationships between physical activity and screen time and childhood obesity Between 1980 and 2008, the prevalence of obesity among 6- to 11-year-olds in the US almost tripled.17 Childhood overweight and obesity have been linked to numerous health risks, including hypertension, hyperlipidemia, type 2 diabetes, sleep disturbance, orthopedic problems and psychological problems in childhood. Obese children, particularly adolescents, have been found to be more likely to become obese adults with related chronic diseases.17 Both physical activity and screen-based leisure activity have been shown to be associated with childhood overweight and obesity.17 Failure to meet guidelines for both physical activity and screen time increased the risk of overweight for boys by a factor of 4.5 and for girls by a factor of 3, compared with those who met both guidelines.18 Less active children tend to become less active adults.19 Sisson's (2010) analysis of 2003 NSCH data revealed that for both boys and girls, the odds of everyday physical activity decreased as hours of TV/video watching increased, and the combined influence of low levels of physical activity and high levels of TV/video watching increased the odds of being overweight.20 While evidence does not support the idea that TV viewing directly displaces physical activity,21,22 excessive TV viewing has been linked with overweight, irregular sleep, and mental health problems.23,24 A growing body of evidence has linked greater screen time with increased risk of poor dietary habits,25,26 obesity,27,28,29 metabolic syndrome and cardiovascular risk 8 factors. Using data on 2964 children in the National Health and Nutrition Examination Survey (NHANES), 2001-2004 Anderson et al22 estimated that 37.3% of US 4- to11- .year-old children engaged in active play less than seven days a week, 65% engaged in more than 2 hours of screen time daily, and 26.3% had both low active play and high screen time. Combined low activity and high screen time was associated with BMI greater than the 95th percentile, female gender and non-Hispanic black race/ethnicity.22 The most successful interventions for prevention and treatment of childhood obesity involve changing parental behavior to affect the behavior of the child.17,30,31,32 Studying the interaction between parent attitudes and special health care needs is important because children with SHCN are at greater risk for obesity, low physical activity levels and excessive use of screen-based leisure activities,4 and the special stress of having a child with SHCN can impact parental ability to channel the child's activity appropriately.3,7,9,33 2. Parental influence on children's physical activity and screen time Among the numerous correlates of physical activity that have been investigated, parental support emerges as a consistent positive association.17,20,34,35,36,37,38 Parenting practices have also been associated with screen time.1,17 In a recent review of 103 studies of parental influence on children's physical activity, Trost and Loprinzi39 found consistent association between parent support (informational, emotional, appraisal, instrumental or combined) and physical activity, with somewhat stronger associations found for adolescents than for younger children. They found few studies that examined parenting style as an influence on physical activity, and only 2 of 7 showed positive significant association with authoritative parenting style. Mixed findings 9 of 8 studies that examined family cohesion and physical activity led to a conclusion that the evidence was "inconclusive." Regarding the association of child and parent physical activity, just 19 of 46 studies in children 6-12 years old and 8 of 27 studies in adolescents 13-18 years old showed significant positive associations.39 Welk et al explored mechanisms of parental influences on physical activity in 994 children in grades 3-6 using child self-report and parental questionnaires to measure both direct and indirect effects of parents on child physical activity (PA).40 They measured four different dimensions of parental support (role modeling, encouragement, involvement and facilitation) as well as a composite "parental influence." They found that parental influence affects child PA directly and through mediation by child intrapersonal factors (enjoyment of PA and perceived PA competency). Facilitation and overt encouragement were most strongly associated, but all of the tested scales contributed significantly to predicted PA.40 Heitzler et al41 used structural equation modeling to study relationships among interpersonal variables (parent MVPA, parent support, peer support), intrapersonal variables (self-efficacy, enjoyment, barriers) and MVPA measured by accelerometer in 720 10-17 year olds. They found that perceived social support from both parents and peers were significantly related to intrapersonal factors that promote physical activity, but that peer support was more strongly correlated with MVPA than was parent support. Parental MVPA, reported by parents through a detailed activity questionnaire, was significantly associated with youth MVPA.41 Most studies of parental influence on children's physical activity have focused on parental support for PA and parental physical activity, without looking at factors that 10 might affect the parent's ability to provide appropriate support and engage in active play with their child, such as parent's mental health, parenting stress and social support. While several studies have investigated the associations between specific chronic conditions in childhood and physical fitness or obesity-related behaviors,42,43,44,45 little is known about the determinants of physical activity and screen-based leisure activity in CSHCN as a group, or how CSHCN with and without EBD differ from children without special health care needs. Parenting practices have also been associated with screen time. Household rules about television watching are associated with decreased screen time15 and having a television in the bedroom is associated with increased screen time.46 3. The independent variables a. Identifying children with special health care needs Approximately 20% of children in the US have at least one chronic condition that requires special health care, educational services, counseling or therapy. Comorbidities are common: 3.9% of US children have two chronic conditions and 4.8% have three or more chronic conditions.45 Children with chronic medical conditions are up to three times more likely than the general population to have a coexisting emotional, behavioral or developmental condition.10 Van Cleave et al48 studied three cohorts of children for six years, from age 2-8 though age 8-14 and found that while the prevalence of chronic conditions in children is increasing with time, many chronic conditions are dynamic. Many children who had a chronic condition at the outset did not have the same condition at the end of the six year study period; most of the chronic conditions present at the end developed during the 6 11 year study period. For example, just 42% of those who had asthma at the outset still had asthma after six years, while 78 percent of the children who had asthma at the end of follow-up did not have asthma at the outset. Similarly, just 37% of children who were obese at the outset were still obese at the end of follow-up, while 67% of the children who were obese at the end of follow-up were not obese at the outset.48 To better plan for the needs of children with chronic conditions, the Health Resources Services Administration (HRSA) Maternal and Child Health Bureau has recently developed a non-categorical approach to identifying them, rather than relying on condition-specific prevalences. The Child With Special Health Care Needs (CSHCN) Screener is a brief questionnaire that identifies children with SHCN by consequences and service needs rather than by diagnosis.49,50 To be identified as a child with SHCN, a child must have at least one of these five special needs due to any medical, behavioral or other health condition that has lasted or is expected to last more than twelve months: 1. Child needs prescription medicine other than vitamins 2. Child needs or uses more medical care, mental health or educational services than is usual for most children of the same age 3. Child is limited or prevented in any way in his/her ability to do the things most children of the same age can do 4. Child needs or get special therapy, such as physical occupational or speech therapy 5. Child has any kind of emotional, developmental or behavioral problem for which (he/she) needs treatment or counseling 12 The CSHCN screener has been validated by comparison with the Questionnaire for Identifying Children with Chronic Conditions--Revised, and it was found to be equally reliable when used in telephone surveys or in self-administered mail questionnaires.47,48 In a national sample of 17,985 children, Bethell et al50 found that 15.3% met at least one screener criterion. They found that the proportion of children who met at least one of the CSHCN screener criteria changed with age. The percent of children meeting CSHCN criteria was 8.0 in preschoolers, 17.2% in 5-9 year olds, 17.9% in 10-14 year olds and 18.4% in 15-18 year olds. Males (17.7%) were significantly more likely than females (12.8%) to meet CSHCN screener criteria. Among 0-13 year olds, 12.8% of Hispanic children, 15.1% of non-Hispanic white children, 14.6% of non-Hispanic black children and 9.7% of children from "other" racial/ethnic groups met CSHCN screener criteria.50 Data from the 2003 NSCH51 showed that children with SHCN (identified by the CSHNC screener) were more likely than those without SHCN to have unemployed parents and live in poverty. Children 12- to 17-years-old were more likely to have SCHN than those 6- to 11-years-old. While the overall prevalence of children with SHCN in adolescents was not significantly different from the prevalence of children with SHCN in pre-teens, 12- to 17-year olds were more likely than 6- to 11-year-olds to have SHCN with frequent headaches and with depression or anxiety. Males were more likely than females to have SHCN. While non-Hispanic black preschoolers were more likely to be identified as children with SHCN than non-Hispanic white or Hispanic children, among 6- to 17-years-olds, non-Hispanic white children were more likely to be identified as children with SHCN than non-Hispanic black or Hispanic children. Children whose parents had fair or poor mental health were more likely to have SHCN than children 13 whose parents had excellent mental health. Children in two-parent households were less likely to have SCHN with EBD than children without two parents in the home.51 Children from households where the primary language is English were three times as likely as those from households primarily using other languages to be identified as children with SHCN. Parents who attended college were more likely to have children with SHCN than parents who did not attend college.51Newacheck et al point out that these differences may arise because parents who are better able to navigate the medical care system are more likely to obtain diagnoses and services.51 b. Parental perceptions of child's limitations Illness during infancy or childhood can cause parents to perceive the child as especially vulnerable, even after the illness abates. This "Vulnerable Child Syndrome", as described by Green and Solnit,52 can distort the parent-child relationship, resulting in child behavior problems, difficulty with separation, infantile behavior, hypochondriasis and academic underachievement.8,52,53 Parents who perceive their child as vulnerable have been described as unnecessarily restricting their children's physical activity.33 While parental perception of child vulnerability and parental over-protectiveness have been investigated as determinants of child adjustment and academic achievement,7,8,54 we have not found any recent population-based studies that examined this construct with regard to obesity-related behaviors in children with SHCN with and without EBD. This association is of interest because of theoretical links between intrapersonal characteristics (outcome expectancy, self-efficacy, and perceived competence) and health behaviors, and the influence of parental perceptions on opportunities and encouragement that promote development of these characteristics in children.5,39,54 In a study of asthmatic 14 children, for example, Pianosi and Davis found that the child's perceived competence at physical activity was correlated with aerobic fitness, but asthma severity was not.43 Similarly, Fong et al found that children with developmental coordination disorder were more likely to be more active when they perceived themselves to have more motor ability.42 c. Parental mental health and parenting stress Having a child with SHCN puts special stresses on the family. Considerable work has been done in clinical settings to evaluate the relationship between chronic illness in childhood and family adjustment.9 Most of these studies have focused on small groups with a particular diagnosis (cystic fibrosis, cancer, limb deficiency, sickle cell anemia), but less is known about the impact of chronic conditions in general at the population level. Wallander and Varni developed a conceptual model of child and family adjustment to pediatric chronic physical disorders designed to be "generic," that is, to address the psychosocial issues that are common to children with children with chronic conditions, independent of their specific diagnosis.9 This model illustrates the interplay of various intrapersonal and social-ecological factors with factors related to disease and disability as related to the mental, social and physical adjustment of the child. Notably, the adjustment of family members and social support provided by the family affects cognitive appraisal and coping strategies that enable affected children to deal with the stresses of condition- related problems, daily hassles and major life events in a way that promotes their appropriate "development into autonomous, healthy, and well-functioning adults."9 Among parents of 2- to 17-year-old children with SHCN with EBD, 42.8% report coping "very well" with parenting compared with 57.2% of CSHCN without EBD4 and 15 60% of all parents surveyed.47 In a study of multiple social risks on children's general health using data from the 2003 NSCH, Larson et al found that low maternal mental health increased the odds of that the child would be overweight and that the parents would rate the child's general health as less than ?very good.?55 We have not found any previous analysis of population-level data regarding the association of parental mental health and parenting stress with obesity-related behaviors in children with SHCN. Data from the 2003 NSCH indicated that children with SHCN were more likely than children without SHCN to be in families that deal with conflict by arguing or shouting and families that eat fewer meals together.51Among 6-17 year olds, children whose parents reported close relationships with their children were less likely to have SHCN with a behavioral/conduct problem than those who did not have close relationships.51 d. Parental social support and trust in neighbors Using data from 2003 NSCH, Singh et al found that low social capital was significantly associated with increased risk of physical inactivity even after adjusting for other factors.56 This study did not look for differences between the general population and the subpopulation of children with SHCN. Children living in supportive neighborhoods were less likely to have SHCN with EBD than children in less supportive neighborhoods.51 e. Parental perception of child's safety In the 2003 NSCH, children whose parents reported less neighborhood safety were more likely to have frequent headaches, developmental problems and behavior/conduct problems.51 Larson found that perception of the neighborhood as unsafe increased the odds of overweight,55 but Singh found no association between neighborhood safety and 16 physical inactivity.56 Using data from the 2007 NSCH, Danielson57 found that children with EBD conditions were more likely to live in neighborhoods perceived as unsafe and more likely to have inadequate activity levels compared with those without EBD conditions. There was significant interaction between EBD status and perceived neighborhood safety.57 4. Measuring physical activity and screen time The literature includes studies that measure physical activity levels by self-report, proxy report, direct observation, and objective measurement by pedometer or accelerometer, and variations in measuring physical activity complicates comparisons across studies.21 Self-report methods include single questions, multiple questions, 24 hour recall and 3-day recall. Both self-report and proxy reports have been shown to lack validity when compared with objective measurements.58,59,60,61 Because of cognitive limitations in young children, proxy reports by parents are used in studies of young children where objective measures are not feasible. Murphy et al found that a single multiple choice question to elicit a description of the child?s overall activity level was a good predictor of child fitness levels.62 Measures of vigorous activity have been found to be more reliable than measures of moderate activity.60 Many studies dichotomize physical activity levels based on whether the reported level of physical activity does or does not meet current guidelines. Current guidelines recommend that children get at least 60 minutes of moderate-to-vigorous physical activity daily, including at least 20 minutes of vigorous activity at least three days a week.2 In the 2007 NSCH parents were asked "During the past week, on how many days did (child) exercise, play a sport, or participate in physical activity for at least 20 minutes 17 that made him/her sweat and breathe hard??63 Both moderate and vigorous activity cause sweating and increased respiration; they are distinguished by intensity, and how much sweat production and respiratory increase they cause. Therefore, while the intensity description captures both moderate and vigorous activity, the duration (20 minutes) is better aligned with guidelines for vigorous activity. The analysis of Singh et al56, 64,65,66 and the NSCH chartbook47 use a 3-day cut point to define those who engage in "regular" physical activity, whereas Anderson22 et al use a 6-day cut point when analyzing a similar question from NHANES. Objective measures of children's physical activity have shown that children often engage in short bursts of vigorous activity67 which may not be included in the answer to the NSCH question. As young people become more autonomous and spend more time away from home, parental report of their unstructured activity levels might become less reliable.60 Nonetheless, Singh notes that NSCH parental reports are similar to youth self-reports about physical activity in the Youth Risk Behavior Survey (YRBS).56 Until recently, studies of sedentary activities have emphasized television viewing, but in the last decade computer-based leisure activities may have displaced some television viewing for some children. Therefore, measures of screen-based leisure activity in more recent studies include both TV and computer use.68 The 2007 NSCH includes a question about TV, videos and video games, and a separate question about non-school related computer use.63 Together these questions allowed calculation of total minutes of screen- based leisure activity, which we dichotomized using a 2-hour cut point consistent with the AAP guidelines. 18 Summary of the conceptual model Drawing from Wallander's model9 based on family dynamics and Singh's model65 of social and behavioral determinants of childhood obesity, Figure 1 shows a conceptual model based on the interrelationships described above. We posit that parental attitudes, including their perception of a child's physical limitations, their mental health and ability to cope with day-to-day demands of parenting, their perception of available social support, their ability to trust their neighbors and their perception of their child's safety, affect their ability to provide appropriate opportunities for, and to set appropriate limits on, behaviors that affect growth and development. Parental attitudes also influence the child's self concept and perceived competence, which in turn affect the child's choice of activities. Furthermore, we posit that the presence of special health care needs can have both direct and indirect effects on both sedentary behavior and obesity. Some conditions interfere with mobility, precluding physical activity, and some require medications that cause excess weight gain. SHCN can also affect the parent's ideas about what the childe can and should do, while also affecting the child's self concept and perceived competence. The focus of the current study is the association of parental attitudes and SHCN with variation in physical activity and screen time. 19 IV. RESEARCH DESIGN AND METHODS For this cross-sectional study of a nationally representative sample of 6- to 17-year- old boys and girls in the United States, we performed a secondary analysis of the publicly available dataset from the 2007 National Survey of Children's Health. 1. Study Population The 2007 National Survey of Children's Health (NSCH) is a module of the State and Local Area Integrated Telephone Survey (SLAITS), conducted by the National Center for Health Statistics (NCHS), Centers for Disease Control and Prevention (CDC), funded by the Maternal and Child Health Bureau of the Health Resources and Services Administration.16 SLAITS uses random-digit dialing of land-line phones to identify households for the National Immunization Survey (NIS), and households from this NIS sample that include children less than 18 years old are eligible for the NSCH. When screening questions indicate that the household includes more than one child, one child is randomly selected from the household to be the subject of the interview. The respondent was the adult in the household who knows the most about the child's health. A total of 91,642 interviews were completed in 2007 and 2008, surveying approximately 1,700 households in each of the 50 states and the District of Columbia. For this study we limited our focus to all 64,076 children ages 6-17, including 15,049 children with SHCN and 49,027 without SHCN. The respondents were 74.6% mothers, 18.8% fathers and 6.6% others. (We will subsequently refer to the respondents as "parents.") 20 2. Defining dependent variables We based our determination of physical activity on responses to the question, ?During the past week, on how many days did (child) exercise, play a sport, or participate in physical activity for at least 20 minutes that made him/her sweat and breathe hard??63 We categorized children as having low MVPA if the parent reported physical activity on five or fewer days per week. We based our determination of screen-based leisure activity on responses to these questions: ? On an average weekday, about how much time does (child) use a computer for purposes other than schoolwork? ? On an average weekday, about how much time does (child) usually watch TV, watch videos or play video games?63 We computed the sum of the minutes for recreational computer use plus the minutes for watching TV and videos and playing video games. We categorized children as having high screen time if the total was greater than 120 minutes. We categorized children as having a sedentary lifestyle if they had both low MVPA (20 minutes of MVPA on five or fewer days/week) and high screen time (more than 120 minutes of screen time per day.) 3. Description of independent variables Children were categorized by special health care needs status (SHCN) based on responses to the CSHCN screener and questions about specific EBD conditions. Children who met no CSHCN screener criteria were classified as "without SHCN." If children met at least one CSHCN screener criterion and the parent indicated that the child needed counseling or therapy for an EBD condition or the parent indicated that the child 21 currently had one of the specific EBD conditions in Table 2, they were classified as "SHCN with EBD." If children met at least one CSHCN screener criterion but did not need counseling or therapy for an EBD condition and did not currently have EBD condition listed in Table 2, they were classified as "SHCN without EBD." To measure the parent's perception of the child's limitations, we used the dichotomous answer to the question, "Is child limited or prevented in any way in his/her ability to do the things most children of the same age can do?"63 To provide a summary measure of parent's mental health and parenting-related stress, we created an index ranging from 0-5 that sums dichotomized answers to questions asking parents to rate ? Their general mental and emotional health ? How well they are coping with the demands of parenthood ? How often they feel the child is much harder to care for than most children his/her age ? How often he/she does things that really bother them ? How often they felt angry with him/her63 Because questions about the other parent were not asked in single-parent households, and because the respondent was the parent who knows the most about the child and his/her medical issues, we used only information about the mental health of the respondent.61 For the logistic regression modeling, we collapsed this into two categories to compare those with a score of zero to those with a score of 1 through 5. Parental perception of social support was measured by the answer to the question, "Is there someone you can turn to for day-to-day emotional help with parenthood/raising children?"63 To measure the parent's perception of the child's safety, we summed the (dichotomized) answers to questions about safety at school and safety in the neighborhood,63 for a scale ranging from 0 (usually or always safe at both school and in 22 the neighborhood) to 2 (usually or always safe in neither school nor neighborhood.) This was collapsed into two categories ("zero" vs "1 through 2") for modeling. As a measure of the parent's trust in neighbors, or their perception of social capital in their community, we used an index ranging from 0-4 that sums dichotomized answers regarding whether they agree or disagree with the statements ? People in this neighborhood help each other out ? We watch out for each other's children in this neighborhood ? There are people I can count on in this neighborhood ? If my child were outside playing and got hurt or scared, there are adults nearby who I trust to help my child63 This was collapsed into two categories ("zero" vs "1 through 4") for modeling. Additional covariates considered as potential confounders included age, gender, race/ethnicity, body mass index (BMI) classification, parental education, poverty level,50,58 severity of condition and number of conditions.4 4. Data analysis Data were analyzed using SAS 9.2 (SAS Institute, Inc.) survey procedures, applying appropriate sampling weights because of the complex survey design. PROC SURVEYFREQ was used to describe frequencies of each outcome for each parental attitude variable (child's limitations, mental health/parenting stress, social support, trust, safety), SHCN and covariate (age, gender, race/ethnicity, parental education, poverty level, BMI classification, severity of condition). Rao-Scott ?2 statistics were used to test for significant associations. (Table 3). We also used PROC SURVEYFREQ to determine frequencies and prevalence of each parental attitude variable and covariate for each 23 SHCN category and used Rao-Scott ?2 statistics to test for significant associations. (Table 4). Where covariates were significantly associated with both the outcomes and SHCN category, they were considered potential confounders. Multicolinearity was evaluated using PROC CORR to compute Pearson correlation coefficients. Observations with data missing for the relevant variables were excluded from the analysis. "Don't know" and "refused" responses were recoded as missing. PROC SURVEYLOGISTIC was used to create logistic regression models to determine the effects of each of the parental attitude variables, SHCN status and covariates on each of three outcome variables: (1) low physical activity (20 minutes of physical activity less than six days/week) (2) high screen time (television, videos, video games and recreational computer use greater than 120 minutes/day), and (3) sedentary lifestyle (both low physical activity and high screen time). Adjusted odds ratios were calculated for children SHCN with and without EBD using children without SHCN as the reference group. To evaluate interactions between the predictor variables, we tested interaction terms for significant effects and created models stratified by special health care needs status and by attitude indicators to examine effects separately in different subgroups. To analyze joint effects, we created variables for various constellations of parental attitudes and special health care needs status and used logistic regression to determine adjusted odds ratios for sedentary life style using those without any of the characteristics in the constellation as the reference group. 5. Human Subjects Respondents to the NSCH were informed that participation was voluntary and gave informed consent.16 We used de-identified data in a publicly available dataset. No attempt 24 was made to link any data to individuals. The proposal was submitted to the University of Maryland Institutional Review Board for approval, and was declared exempt because the data set does not include personal identifiers. 25 V. RESULTS 1. Descriptive statistics a. Characteristics of the population The 2007 NSCH included interviews about 64,076 school-aged children 6-17 years of age. Table 1 shows the distribution of various sociodemographic characteristics in the study population. The sample included 33,292 boys and 30,693 girls. There were 27,792 children (6-11 years old) and 36,284 adolescents (12-17 years old). The respondents included 46,750 mothers (including adoptive and step-mothers), 13,388 fathers (including adoptive and step-fathers), and 3926 ?others? acting in a parenting capacity. Table 2 shows the frequency of specific diagnostic categories in the sample and population estimates computed using sampling weights to adjust for the complex sampling design. While an estimated 23% of the population were identified by the CSHCN screener as children with SHCN, an estimated 24% of the population had one of the listed chronic conditions, 11% had two, 5% had three and 7% had four or more. An estimated 4.7% of children had one of the listed conditions which the parent described as severe and 1.6% had two or more severe conditions. The sample included 49,027 children without special health care needs, 7,527 children with SHCN without EBD, and 7,522 children with SHCN with EBD. Using appropriate weights, this indicates that 77.1% (SE, 0.41) of the US population of 6- to 17-year-olds, or an estimated 37,997,602 children and adolescents have no SHCN. An estimated 11.2% (SE, 0.31) or 5,531,804 have SHCN without EBD and an estimated 11.7% (SE, 0.31) or 5,749,742 have SCHN with EBD (Figure 2).As seen in Figure 3, the most 26 common of the specific diagnoses listed were respiratory allergies (19.0%; SE 0.40%), asthma (10.3%; SE, 0.30%)), learning disabilities (9%; SE, 0.30), and attention deficit disorder (8.2%; SE, 0.26%). The prevalence of all three outcomes (low MVPA, high screen time and sedentary lifestyle) varied significantly with gender, age (Figure 4), BMI classification (for those 10 and older), race/ethnicity (Figure 5), respondent?s education, household poverty ratio, and special health care needs status (Figure 6). Table 3 shows the prevalence of each outcome among different subpopulations. b. Low moderate-to-vigorous physical activity As shown in Table 3, we found that 64.2% (SE, 0.5%) of US 6-17 year olds get 20 minutes of moderate-to-vigorous physical activity less than 6 days a week (low MVPA). A significantly higher proportion of girls than boys had low MVPA. A significantly higher proportion of adolescents than of younger children had low MVPA. The prevalence of low MVPA also varied significantly with race/ethnicity, with the highest prevalence of low MVPA less than 6 days a week being among Hispanics and the lowest prevalence being among non-Hispanic multi-racial children and adolescents. The prevalence of low MVPA was significantly greater among children from poor households than among the more affluent. There was also a significant association between low MVPA and respondent?s education, with higher prevalence among children of the less educated. The prevalence of low MVPA varied by state, from a low of 56% in North Carolina to a high of 73% in Connecticut. Maryland ranked eighteenth, with 64% and the District of Columbia ranked fiftieth with 71%. 27 As shown in Table 3, the prevalence of low MVPA did not differ significantly by special health care needs status (p=.1714). The prevalence of low MVPA was significantly higher among children with perceived limitations (70.2%; 95%CI, 66.72- 73.79) than among children without perceived limitations (63.7%; 95%CI, 62.74-64.74). The prevalence of low MVPA was lower among children whose parents reported that they had social support (63.2%; 95%CI, 62.21-64.25) than among children whose parents did not report that they had social support (70.1%; 95%CI, 67.77-73.39). Children whose parents indicated less trust in their neighbors had higher prevalence of low MVPA than those indicating greater trust, but this difference was not significant (p=.0648). Those who had zero scores on the trust index had a significantly lower prevalence of low MVPA (63.41%; 95%CI, 62.31-64.50) than those with scores of 1-4 (67.0%; 95%CI, 64.81-69.25). Prevalence of low MVPA was least among children whose parents considered them usually or always safe both in school and in their neighborhood, and greatest among those whose parents considered them usually or always safe in neither school nor neighborhood. For the parental the mental/health stress index, the chi square test indicated significant differences in low MVPA (p=.035); the highest prevalence was among those with a score of 3 out of 5 and the lowest prevalence was among those few with scores of 4 and 5. Among children whose parents had a zero score on the mental health/stress index the prevalence of low MVPA (63.2%; 95%CI, 62.14-64.28) was significantly less (p=.0006) than among children whose parents had a score of 1-5 (67.6%; 95%CI, 65.41-69.84). 28 c. High screen time Overall, 48.4% (SE, 0.5%) of 6-17 year olds exceeded 2 hours of screen time daily (high screen time). The prevalence of high screen time was significantly greater among adolescents than among younger children, and significantly greater among boys than girls. Non-Hispanic black children had a significantly higher prevalence (63.7%; SE, 1.19%) of high screen time than non-Hispanic white children (44.8%; SE, 0.56%), Hispanic children (48.2%; SE, 1.57%) or non-Hispanic children of other races (43.8%; SE, 2.89%). Lower household income and lower parental education were associated with significantly higher prevalence of high screen time. Among the 50 states and the District of Columbia, Vermont had the lowest prevalence of high screen time (36%) and Florida had the highest prevalence (57%). Maryland ranked thirty-second at 49% and the District of Columbia ranked forty-second at 52%. There were significant differences in the proportions of children with high screen time by SHCN status, with the highest prevalence among children with SHCN with EBD (53.2%) and the lowest prevalence among those with no SHCN (48.0%). The proportion of children with high screen time was greater among children with perceived limitations than among those without perceived limitations. The proportion of children with high screen time was lower among children whose parents had social support than among those whose parents did not have social support. There was also a significant difference in screen time related to perceived safety; 47% of children who were considered safe both in school and their neighborhoods and 59.7% of those considered safe in neither school nor neighborhood had high screen time. The proportion of children with high screen time also varied significantly by scores on the parental mental health/stress index, with highest 29 prevalence among those with scores of 4 and 5 and lowest among those with a score of zero. Higher scores on the trust in neighbors index were also associated with higher prevalence of high screen time. d. Sedentary life style Overall, an estimated 33.3% of US 6-17 year olds had sedentary lifestyles (combined low MVPA and high screen time). The prevalence of the sedentary lifestyle combination was higher among adolescents than among younger children. While girls had higher prevalence of low MVPA and boys had higher prevalence of high screen time, the combination was significantly more prevalent among girls (35.1%) than among boys (31.5%). Race/ethnicity was also associated with significant differences in the prevalence of sedentary lifestyle, with the highest prevalence among non-Hispanic black children (42.2%). Prevalence of sedentary lifestyle was significantly lower among children whose parents had more than twelve years of education than among those with twelve years or less. Household income was also associated with significant differences in prevalence of sedentary lifestyle, with the lowest prevalence among the most affluent. Among the 50 states and the District of Columbia, Vermont had the lowest prevalence of sedentary lifestyle (24%) and Florida had the highest (39%). Maryland ranked twenty- second with 32% and the District of Columbia ranked fiftieth with 38%. Special health care needs status was significantly associated with sedentary lifestyle, with prevalence of 31.5% among children with no SHCN, 39.1% among children with SHCN without EBD and 41.6% among children with SHCN with EBD. Children with perceived limitations had a significantly higher prevalence of sedentary lifestyle than those without perceived limitations. There was a lower prevalence of sedentary lifestyle 30 among children whose parents had social support than among those whose parents did not have social support. Among children who were considered safe both in school and in their neighborhood, the prevalence of sedentary lifestyle was significantly lower than among those considered safe in neither school nor neighborhood. Scores on the trust index were also associated with significant differences in sedentary lifestyles, with lowest prevalence (32.3%) among those with a score of zero and highest prevalence (43.3%) among those with a score of 4. Scores on the mental health/stress index were also associated with significant differences in prevalence of sedentary lifestyle, with the highest prevalence among those with a score of 3 out of 5 (44.4%) and lowest prevalence among those with scores of zero (31.5%) and five (31.1%). e. Associations of covariates with special health care needs status As shown in Table 4, Rao-Scott chi square tests indicated significant associations between special health care needs status and gender, race/ethnicity, respondent, respondent's education, household poverty ratio, number of conditions and number of severe conditions. All of the parental attitudes of interest were also significantly associated with special health care needs status. Among children with no perceived limitations, 81.6% had no SHCN, 10.3% had SHCN without EBD and 8.2% had SHCN with EBD. Among children with perceived limitations, 19.1% had no SHCN, 23.6% had SHCN without EBD and 57.2% had SHCN with EBD. There was a higher prevalence of SHCN with EBD and a lower prevalence of SHCN without EBD among those without social support, compared with those with social support. There was lower prevalence of SHCN with EBD among those considered safe both in school and in the neighborhood compared with those not 31 considered usually or always safe in school, neighborhood or both. Non-zero scores on the trust index were associated with higher prevalence of SHCN with EBD and lower prevalence of no SHCN than zero scores. Among children whose parents' reports yielded scores greater than one on the mental health/stress index, there was a higher prevalence of children with SHCN with EBD and a lower prevalence of no SHCN and SHCN without EBD than among those with scores of zero or one. f. Correlations between covariates To evaluate colinearity between variables we used proc corr to generate a correlation matrix. Because respondent's education was highly correlated with poverty ratio (Pearson correlation coefficient (rho=.417), and the poverty ratio had more missing data, we chose to use only the respondent's education to indicate socio-economic status in the models. We also noted correlations between SHCN status and perceived limitations (rho=.404), SHCN and mental health/stress (rho=.246) SHCN status and number of conditions (rho= .665) and SHCN status and severity (rho=.337). Number of conditions was also correlated with condition severity (rho=.430) and mental health/stress (rho=.223). Therefore we did not include number or severity of conditions in the models. Among the parental attitudes, we noted correlation between perceived safety and trust in neighbors (rho=.276) and between perceived limitations and mental health/stress (rho=.209). Other combinations had correlation coefficients less than 0.2. 2. Hierarchical logistic regression models For each outcome, we used PROC SURVEYLOGISTIC to compute crude odds ratios for each of the study variables (special health care needs status, perceived limitations, 32 trust in neighbors, parental mental health/stress, social support and perceived safety) and covariates (gender, race/ethnicity, respondent's education and age). Model 1 adjusts for gender race/ethnicity, respondent's education and age. Model 2 adjusts for gender, race/ethnicity, respondent's education, age and special health care needs status. Model 3 is the fully adjusted model, which adjusts for gender, race/ethnicity, respondent's education, child's age, special health care needs status, perceived limitations, mental health/stress, trust in neighbors, social support and perceived safety. a. Low moderate-to-vigorous physical activity Table 5 shows the prevalence of low MVPA for each of the SHCN categories and dichotomized parental attitude variables. The differences between the SHCN categories are not statistically significant. There are significant differences in prevalence for the all of the parental attitude indicators, with higher prevalence associated with presence of perceived limitations, lack of social support, and non-zero scores on the mental health/stress, trust in neighbors and safety indices.. Table 6 shows the crude odds ratios and three models for low MVPA. Unadjusted odds ratios for all of the variables of interest except special health care needs status show significant effects (Wald chi square with p<.05). Model 1 shows that adjusting for gender, race/ethnicity, respondent's education and child's age caused little change in odds ratios for those variables. Model 2 shows that after adjusting for the sociodemographic variables, children with SHCN with EBD had significantly greater odds of low MVPA than children with no SHCN (, 1.241; 95%CI, 1.092-1.411). In the fully adjusted model, the effect of SHCN status was no longer significant, and the effects of mental health/stress, trust in neighbors, social support and perceived safety were no longer 33 significant (Wald chi square with p>.05). Children with perceived limitations were significantly more likely than those without perceived limitations to have low MVPA after adjusting for demographic characteristics, special health care needs status and the other parental attitude variables (, 1.339; 95%CI, 1.079-1.662). Figure 7 summarizes the results of logistic regression for the effects of the five parental attitude variables on low MVPA in each of the following models: ? Unadjusted ? Model 1: Adjusted for demographic variables (age, gender, race/ethnicity, and respondent's education ? Model 2: Adjusted for demographic variables and SHCN status ? Model 3: Adjusted for demographic variables and the other attitudes While unadjusted logistic regression shows significant effects on the odds of low MVPA for all of the parental attitude variables, only perceived limitations remains significant in the fully adjusted model. 34 c. High screen time Table 7 shows the prevalence of high screen time for each category by SHCN status and dichotomized parental attitude indicators. The proportion of children with high screen time is significantly higher among those with SHCN with EBD than those without SHCN. The proportion with high screen time among those with SHCN without EBD is not significantly different from the other two SHCN categories. All of the parental attitude indicators show significant differences, with higher prevalence among those with perceived limitations, lack of social support and non-zero scores on mental health/stress, trust in neighbors and perceived safety indices. Table 8 shows the crude odds ratios and adjusted odds ratios computed from three logistic regression models for high screen time. For high screen time, all of the demographic covariates, special health care needs status and parental attitudes variables showed significant effects on the unadjusted odds of high screen time (Wald chi square with p<.05). Model 1 (adjusting for gender, race/ethnicity, education of respondent and age of child) resulted in little change in these effects. Model 2 showed that after adjusting for gender, race/ethnicity, education of respondent and age of child, children with SHCN with EBD were more likely than children without SHCN to have high screen time (, 1.172; 95%CI, 1.031-1.333). In the fully adjusted model which included the parental attitude variables, this relationship was no longer significant. After adjusting for all the other covariates, effects of special health care needs status, perceived limitations, social support and perceived safety on the odds of high screen time were not significant (Wald chi square with p>.05). The fully adjusted model also showed 35 that girls were significantly less likely than boys to have high screen time (AOR, 0.887; 95%CI, 0.815-0.966), non-Hispanic black children were twice as likely as likely as non- Hispanic white children to have high screen time (AOR, 2.007; 95% CI, 1.77-2.276) and children of high school graduates were more likely than children whose parents have more than 12 years of education to have high screen time (AOR, 1.45; 95% CI, 1.280- 1.564). The child?s age in years also had a significant positive effect on the odds of sedentary life style. Children whose parents scored 1 through 5 on the mental health/stress index were significantly more likely to have high screen time than those whose parents scored zero (AOR, 1.189; 95%CI, 1.052-1.344). Children whose parents scored 1-4 on the trust in neighbors index were more likely than those whose parents scored zero to have high screen time (AOR, 1.243: 95%CI, 1.104-1.399). Adding attitudes to the model lowered the AOR for non-Hispanic black children by 6%. Figure 8 summarizes the results of logistic regression models for high screen time for effects of each of the parental attitude variables ? Unadjusted ? Model 1: Adjusted for demographic variables (age, gender, race/ethnicity, and respondent's education ? Model 2: Adjusted for demographic variables and SHCN status ? Model 3: Adjusted for demographic variables and the other attitudes Similar to the analysis for low MVPA, all of the attitude variables show significant effects in the unadjusted regression. However, only the mental health/stress and trust in neighbors variables show significant effects after adjusting for the other covariates. 36 c. Sedentary lifestyle Table 9 shows the prevalence of sedentary lifestyle (both less than six days/week with 20 minutes of exercise and more than 120 minutes/day of screen time) for each category by SHCN status and dichotomized parental attitude indicators. The prevalence of sedentary lifestyle is significantly greater among children with SHCN with EBD than among children without SHCN, but the prevalence among children with SHCN without EBD is not significantly different from the other two categories. All of the parental attitude indicators are associated with significant differences in prevalence, with higher prevalence among those with perceived limitations, lack of social support and non-zero scores on the mental health/stress, trust in neighbors and perceived safety indices.. Table 10 shows crude odds ratios and adjusted odds ratios computed in three logistic regression models for sedentary lifestyle. Unadjusted logistic regression showed significant effects for all of the sociodemographic characteristics, special health care needs status and parental attitudes variables (Wald chi square with p<.05). In Model 1, adjusting for gender, race/ethnicity and age slightly attenuated the effect of respondent's education. In Model 2, the effect of special health care needs status was significant after adjustment for gender, race/ethnicity, respondent's education and child's age, with children with SHCN with EBD (AOR, 1.267; 95%CI, 1.111-1.445) and children with SHCN without EBD (AOR, 1.177; 95%CI, 1.020-1.358) more likely to have sedentary lifestyle than children without SHCN. After adjusting for all the other covariates in Model 3, the effect of special health care needs status was no longer significant. Adding attitudes to the model decreased the AOR for non-Hispanic black children by 5%. In the fully adjusted model, the effects of social 37 support and perceived safety on the odds of sedentary life style were no longer significant (Wald chi square with p>.05). Children with perceived limitations were more likely to have sedentary lifestyles than children without perceived limitations (AOR, 1.245; 95% CI, 1.018-1.522). Children whose parents scored 1 - 5 on the mental health/stress index were more likely to have sedentary lifestyles than those whose parents scored zero (AOR, 1.206; 95%CI, 1.068-1.363). In the fully adjusted model, children whose parents scored 1 - 4 on the trust in neighbors index were also more likely to have sedentary lifestyles than those whose parents scored zero (AOR, 1.149; 95%CI, 1.02-1.295). Figure 9 summarizes the results of logistic regression for the effects of each of the parental attitude variables on sedentary lifestyle, giving unadjusted odds ratios and AOR for each of the three models as described before. As for low MVPA and high screen time, all of the attitude variables had significant effects on sedentary lifestyle. However, only the perceived limitations, mental health/stress and trust in neighbors variables had significant effects after adjusting for the covariates. 3. Testing for effect modification To test whether parental attitudes modify the effect of special health care needs on the outcomes, we used PROC SURVEYLOGISTIC to produce logistic regression models that included interaction terms. For each of the outcomes, we created models that included gender, race/ethnicity, education of respondent, age of child, special health care needs status, the attitude variable and the interaction term. None of the interaction terms had significant effects on odds of low MVPA or sedentary lifestyle. Only the interaction term for perceived limitations and special health care needs had a significant effect on the 38 odds of high screen time (Table 11). Tables 12-15 list results of logistic regression using domain analysis to stratify by SHCN status and level of parental attitude indicators. a. Low moderate-to-vigorous physical activity Table 12 lists results of separate logistic regression models predicting low MVPA for each attitude variable adjusted for demographic characteristics, both for the overall population and for each subpopulation defined by SHCN status. Among all 6- to 17-year- olds, those with perceived limitations were more likely to have low MVPA than those without perceived limitations (AOR, 1.445; 95%CI, 1.182-1.765) This relationship was essentially the same for children with SHCN without EBD; the AOR point estimate increased for children without SHCN and decreased for children with SHCN with EBD, but because of wider confidence intervals (smaller numbers) those differences were not significant. There was no significant change in the adjusted odds ratios in the SHCN status subpopulations for the other attitude variables. Table 13 lists results from logistic regression models that use domain analysis to examine changes in the adjusted odds of low MVPA with changes in SHCN status when stratified by the attitude variables. Among all 6- to 17-year-olds, after adjusting for gender, race/ethnicity, respondent's education and age of child, the odds of low MVPA for children with SHCN without EBD was not significantly different from the odds of low MVPA for children with no SHCN, but the odds of low MVPA for children with SHCN with EBD was significantly greater (AOR, 1.241; 95%CI, 1.092-1.411). Among children whose parents had zero scores on the mental health/stress index, this relationship was unchanged, but among children whose parents had scores of 1 - 5 on the mental health/stress index, there was no significant difference among the SHCN categories. 39 Similar patterns were seen for the other parental attitude indicators, with significantly different adjusted odds for children with SHCN with EBD among those with more "favorable" attitudes, but no significant effect of SHCN status on low MVPA among those with less trust, less perceived safety, greater perceived limitations and less social support. b. High screen time Table 14 shows how the adjusted odds ratios for the effects of parental attitudes on high screen time change with stratification by SHCN status. Among all 6- to 17-year- olds, there was no significant difference between those children who had perceived limitations and those who did not in the odds of high screen time. Stratifying by special health care needs status, we found that among children with SHCN without EBD, those with perceived limitations had significantly greater odds of high screen time than those without perceived limitations (AOR, 1.494; 95%CI, 1.135-1.967). The odds of high screen time did not differ significantly by perceived limitations among children without SHCN or among children with SHCN with EBD (Figure 10). Among all 6- to 17-year- olds, the adjusted odds of high screen time was significantly higher for those whose parents scored 1 - 4 on the trust in neighbors index than for those who scored zero (AOR, 1.287; 95%CI, 1.149-1.442). The stratified analysis showed similar results for children without SHCN, but the relationship was attenuated among those with SHCN without EBD (AOR, 1.042; 95%CI, 0.735-1.473) and amplified among those with SHCN with EBD (AOR, 1.509; 95%CI, 1.172-1.942). The stratified models for mental health/stress score, perceived safety and social support showed similar results in all three special health care needs strata. 40 Table 15 shows how the adjusted odds ratios for the effect of special health care needs status on the odds of high screen time changes with stratification by the parental attitude indicators. Among all 6- to 17-year-olds, the likelihood of high screen time was slightly, but significantly, higher among children with SHCN with EBD than among those without SHCN (AOR, 1.172; 95%CI, 1.031-1.333). Stratifying by presence/absence of perceived limitations, we found a similar relationship among children with perceived limitations: those with SHCN with EBD were significantly more likely than those without SHCN to have high screen time (AOR, 1.237; 95%CI, 1.171-1.428). This difference was not found among those without perceived limitations. Among children without perceived limitations, the odds of high screen time for children with SHCN without EBD was 1.452 (95%CI, 0.902-2.337) times the odds of high screen time for children without SHCN (p=.0798). Stratifying by scores on the trust in neighbors index, we found a slightly amplified relationship among those who indicated some lack of trust in their neighbors (score 1 - 4): children with SHCN with EBD were more likely than those without SHCN to have high screen time (AOR, 1.409; 95%CI, 1.100-1.804). Among those who did not indicate lack of trust there was no significant difference in odds of high screen time by SHCN status. c. Sedentary lifestyle Table 16 shows how the adjusted odds ratios for effects of parental attitude indicators on sedentary lifestyle change with stratification by special health care needs status. Among all 6- to 17-year-olds, the odds of sedentary lifestyle was significantly higher among those with perceived limitations than among those without perceived 41 limitations (AOR, 1.407; 95%CI, 1.17-1.69). Stratifying by SHCN status, we found that this relationship persisted among children with SHCN without EBD, (AOR, 1.565; 95%CI, 1.173-2.0870). This difference was attenuated and no longer significant among children without SHCN and among children with SHCN with EBD. Among all 6- to17- year-olds, the odds of sedentary lifestyle was significantly higher for those whose parents indicated some lack of trust in neighbors (AOR, 1.191; 95%CI, 1.061-1.338) and point estimates were similar in the SHCN subpopulations. Among all 6-17 year olds, children whose parents indicated some mental health/stress problem were more likely than those whose parents indicated no mental health/stress problem to have sedentary lifestyles (AOR, 1.280; 95%CI, 1.058-1.402). The point estimate for this relationship was higher among children with SHCN without EBD, but the confidence interval was wider (AOR, 1.395 95%CI, .964-2.019). To determine if parental attitudes had different effects on sedentary lifestyle depending on special health care needs status, we stratified by parental needs indicators and examined adjusted odds ratios for children with SHCN without EBD and SHCN with EBD (Table 17). Among all 6- to 17-year-olds the odds of sedentary life style was slightly higher for children with SHCN without EBD (AOR, 1.177; 95%CI, 1.02-1.358) and SCHN with EBD (AOR, 1.267; 95%CI, 1.111-1.445) than for those without SHCN. Among those without perceived limitations, children with SHCN with EBD had significantly greater odds of sedentary lifestyle than children without SHCN (AOR, 1.233; 95%CI, 1.066-1.425) but this difference was not found among those with perceived limitations (AOR, 0.883; 95%CI, 0.546-1.428). 42 Among children whose parents had social support, children with SHCN with EBD had a significantly greater likelihood of sedentary lifestyle (AOR, 1.33; 95%CI, 1.154-1.533) than those without SHCN, but this difference was not found among those without social support. For parental mental health/stress, perceived safety and trust in neighbors, AORs were similar for the ?zero? score strata and the non-zero score strata for those with SHCN with EBD. Among children whose parents reported some mental health/stress problem, the point estimate for AOR for children with SHCN without EBD was 30% higher than among children whose parents reported no mental health/stress problem but confidence intervals for these estimates overlap. Similarly, among children whose parents have social support, the odds of sedentary lifestyle was greatest among those with SHCN with EBD (AOR, 1.330; 95% CI, 1.154-1.533). Among children whose parents lack social support the differences in odds of sedentary lifestyle by SHCN status was not significant. 4. Joint effects To evaluate joint effects of attitudes and special health care needs status on the likelihood of sedentary lifestyle, we created combination variables involving the parental attitude variables that showed significant associations in the fully adjusted model in Table 10 (perceived limitations, mental health/stress and trust in neighbors). We calculated the odds ratios adjusted for age, gender, race/ethnicity and respondent's education, using children with none of the characteristics in the combination as the reference category. Table 18 and Figure 12 shows AORs for combinations of a single parental attitude variables with each SHCN status, with children without SHCN and without that parental attitude as the reference category. 43 Table 19 and Figure 13 show how different combinations of these three parental attitude variables are related to the odds of sedentary lifestyle after adjustment for demographic factors. Children whose parents perceived them as having limitations and also reported mental health/stress problems had significantly increased odds of sedentary lifestyle, compared with those with neither perceived limitations nor parental mental health/stress problems (AOR, 1.42; 95%CI, 1.082-1.864). The combination of parental mental health/stress problems and lack of trust is also associated with a significant increase in the odds of sedentary life style (AOR, 1.614; 95%CI, 1.337-1.948). The combination of perceived limitations and lack of trust is also associated with a significant increase in the odds of sedentary life style (AOR, 1.557; 95%CI, 1.163-2.086). These adjusted odds ratios are somewhat greater than the AORs for these factors separately, as seen in Figure 9. The AOR for mental health/stress problems alone was 1.28 (95% CI, 1.139-1.438). The AOR for lack of trust alone was 1.191 (95%CI, 1.061-1.338). The AOR for perceived limitations alone was 1.407 (95%CI, 1.117-1.690). Table 20 shows the effects of combinations of multiple parental attitudes and special health care needs on the odds of sedentary life style. Two constellations increased the odds of sedentary lifestyle over two-fold: parental mental health/stress, limitations and SHCN without EBD (AOR, 2.659; 95%CI, 1.741-4.06) and lack of trust, limitations and SHCN without EBD ( 2.434; 95%CI, 1.436-4.126). From Table 16 in the subpopulation who had SHCN without EBD, the AOR for perceived limitations was 1.565 (95%CI, 1.173-2.087), for parental mental health/stress was 1.395 (95%CI, 0.964-2019) and for lack of trust was 1.064 (95%CI, 0.747-1.517). As shown in Figure 14, the three constellations with the highest AORs all involve children with SHCN without EBD. 44 VI. DISCUSSION 1. Association between parental attitudes and low MVPA Our first research question asked if there was an association between parental attitudes and low MVPA. Consistent with our hypothesis, we found that perception of functional limitations in the child, lack of social support, perceived lack of safety, mental health/stress problems and lack of trust were all associated with an increased prevalence of low MVPA. However, after adjusting for demographic factors and special health care needs status, only the presence of perceived limitations was associated with significantly increased odds of low MVPA. 2. Association between parental attitudes and high screen time Our second research question asked if there is an association between parental attitudes and high screen time. Consistent with our hypothesis, we found that perception of functional limitations in the child, lack of social support, perceived lack of safety, mental health/stress problems and lack of trust were all associated with an increased prevalence of high screen time. However, after adjusting for demographic factors and special health care needs status, only mental health/stress problems and lack of trust were associated with significantly increased odds of high screen time. 3. Interaction between parental attitudes and special health care needs status Our third research question concerned whether the effects of special health care needs on low MVPA and high screen time were modified by parental attitudes. The effects of special health care needs on low MVPA and high screen time were small. We found no 45 evidence of multiplicative interaction for mental health/stress, perceived safety, social support or trust in neighbors. We did find support for interaction between the effects of perceived functional limitations and special health care needs on high screen time. However, contrary to our expectations, the odds of high screen time was amplified among children with SHCN without EBD, but not among children with SHCN with EBD. Nonetheless, the results concerning the interaction between parental perception of the child's limitations and SHCN status need to be interpreted with caution. In the stratified analysis, we compared the adjusted odds of high screen time in children with perceived limitations who had SHCN with the adjusted odds of high screen time in children with perceived limitations who had no SHCN. This reference group (children without SHCN who have perceived limitations) is relatively small, and might be quite variable. This group is comprised of 717 sample children, representing a population prevalence of 1.77% (SE, 0.17) of the children without SHCN. Of these, 54.0% (SE, 5.0) had high screen time. Both the point estimate and the standard error for this group are higher than for the general population (48.4%, SE, 0.50) and for all children without SHCN (47.6%; SE, 0.58). Among children with perceived limitations and SHCN with EBD, on the other hand, the prevalence of high screen time is 51.7% (SE, 2.66). Of the children whose parents perceive them has having functional limitations, 80.8% meet at least one criterion in the CSHCN screener. Those who do not may have a limiting condition expected to last less than 12 months, or they may lack access to appropriate care for diagnosis and treatment. Alternatively, they may have health belief systems that cause them not to identify the child's limitation as due to "a medical, emotional, 46 behavioral or developmental condition." Hence, the apparent attenuation of the association between SHCN with EBD and high screen time may be due to higher risk of high screen time for this atypical reference group rather than due to lower risk of high screen time for children with SHCN with EBD. 4. Joint effects of parental attitudes and SHCN on sedentary lifestyle We found that some combinations of parental attitudes and special health care needs increased the odds of sedentary lifestyle above the expected effects of the individual factors. However, contrary to our expectations, combined effects were greater for children with SHCN without EBD than for children with SHCN with EBD. The constellation of factors with the greatest likelihood of sedentary lifestyle was parental mental health/stress along with perceived limitations and SHCN without EBD (AOR, 2.659; 95%CI, 1.741-4.06). The constellation of perceived limitations along with lack of trust and SHCN without EBD was also associated with a greater than two-fold increase in the odds of sedentary lifestyle (AOR, 2.434; 95%CI, 1.436-6.126). We expected that the greatest likelihood would be among those with all five "unfavorable" attitudes and SHCN with EBD. Only 40 sample children had parents with all five unfavorable attitudes and only 9 of these had SHCN with EBD. Of these, 3 had sedentary lifestyles and 6 did not. The adjusted odds ratio for this group (with the reference group having all favorable attitudes and no SHCN) was 0.509 (95%CI, 0.115-2.25), but with less than 30 observations in this group, estimates are not considered reliable.16 Other constellations with four or more factors also had too few children in each category for reliable estimates. 47 5. Summary and implications In the fully adjusted models, we found significant associations between perceived limitations and low MVPA. Parental mental health/stress problems and lack of trust were associated with a significant increase in the likelihood of high screen time. Perceived limitations, parental mental health/stress problems and lack of trust were associated with significantly increased likelihood of sedentary lifestyle. The joint presence of two of these three factors further increased the odds of sedentary lifestyle, as did the presence of SHCN without EBD. Our analysis supports the expectation that children with SHCN are more likely to be perceived as having functional limitations than children without SHCN. However, as described by Green and Solnit 52 and Perrin53 parents' perception of their children's limitations are not always realistic. From this survey, we cannot tell how many of these children have physical limitations that would preclude moderate-to-vigorous physical activity. However, most physical disabilities need not preclude physical activity when appropriate modifications are made.17 The association found between perceived limitations and the likelihood of low MVPA and sedentary lifestyle, especially in children with SHCN without EBD, suggests that further attention to perceived barriers to participation in active play and other organized activities among parents in this group is indicated. The 2007 NSCH asks three additional questions about parents' perceptions of their child?s limitations: whether they are limited in their ability to attend school regularly, to participate in sports and other activities and to make friends.63 We did not include these responses in our analysis because the questions were only asked when children met at 48 least one of the CSHCN screener criteria. However, responses to these questions indicate that there are significant differences between children with SHCN without EBD and children with SHCN with EBD. Limited ability to attend school regularly was reported by parents of 9.1% (95%CI, 7.53-10.63) of children with SHCN without EBD and 16.8% (95%CI, 14.68-19.00) of children with SHCN with EBD. Limited ability to participate in sports and other activities was reported for 14.8% (95%CI, 13.00-16.58) of children with SHCN without EBD and 25.8% (95%CI, 23.22-28.31) of children with SHCN with EBD. Children with SHCN with EBD (28.5%; 95%CI, 26.10-30.95) were ten times more likely than those with SHCN without EBD (2.5%; 95%CI, 1.80-3.35) to be limited in their ability to make friends. These differences underscore the variability of concerns that parents have about their children and the potential impact of these perceived limitations on the opportunities that parents provide for their children to be physically active and productively engaged. When parents are unable to trust in their neighbors? ability or willingness to look out for their children or help them if they are hurt or scared, children may have fewer opportunities for unstructured outdoor play or participation in neighborhood sports and other activities. The association of such concerns with high screen time and sedentary lifestyles indicates a need to address concerns about the neighborhood when developing a plan for a more active lifestyle. The association of mental health and parenting stress issues with high screen time and sedentary lifestyles further raises the importance of considering parental mental health and coping styles when advising changes in a child?s routine. Allowing children to engage in screen time often affords stressed parents an important respite. Guidance about 49 screen time limits might be more effective if it is coupled with advice about stress- relieving, non-competitive physical activities that families can enjoy together. 6. Strengths and Limitations The NSCH collects information about a variety of parental perceptions that might affect parental influence on children's activities. The large sample size in the NSCH allows detection of differences between groups even when stratified. The CSHCN screener contained within the NSCH allowed us to identify a subpopulation of children whose medical, behavioral, emotional and behavioral characteristics might impact their parents' attitudes about the importance of various activities for their growth and development. As described by van der Lee et al, the comparison of findings about children with chronic conditions is complicated by a wide variety of definitions involving functional limitations, service needs and duration of conditions.69 While the CSHCN screener provides some clarity in separating those without SHCN from those with SHCN, our attempt to separate those with and with emotional, behavior and developmental conditions was complicated by lack of information. The incomplete list of specific diagnoses limited our ability to separate children who have only an EBD condition or only a physical condition from those who have both a physical condition and an EBD condition. Furthermore, our ability to adjust for the severity of the child's condition was limited because information about the severity of the condition was only available for listed diagnoses; children with rare but severe conditions that were not listed (e.g. congenital heart disease, sickle cell disease, cancer, HIV) were therefore counted among those having SHCN because they met at least one CSHCN screener criterion, but would not be 50 counted among those having severe conditions. Lack of a global indicator of the severity of the child?s condition interfered with the ability to adjust for severity. Because this is a secondary data analysis, we were limited to the questions that were asked. We did not have direct questions about perceived vulnerability or parent's concerns about the effect of PA on their child's condition. Our analysis was limited to the attitudes of the respondent, who was designated as the adult who knows the most about the child's health.63 However, the other parent or other adults in the household may also influence the child's activities. Our measures of attitudes have not been validated. For our index of parental mental health and parenting stress we included parent's self-rated mental health as well as responses to questions about parenting stress. Through this measure we aimed to capture parents whose ability to cope with day-to-day demands might impact their ability to provide opportunities for their children to engage in appropriate activities. However, this measure does not enable us to determine whether the observed score is attributable to a depressed, withdrawn parent or a difficult, oppositional child. We only know about the SHCN status of the sample child. The parent's experience with other children in the household (with or without SHCN or EBD) might impact their attitudes, especially with respect to their mental health/stress index or their perception of functional limitations. The measurement of physical activity by a single-question parental report is particularly problematic because parents don't necessarily know how much activity their children do while away from home. Moreover, the wording of the physical activity question is not well aligned with current guidelines. The guidelines call for 60 minutes of 51 moderate-to-vigorous activity daily, including at least 20 minutes of vigorous activity at least three days a week.2 The NSCH question asks about the number of days per week when the child was active for 20 minutes and the activity description in the question (enough to make him/her sweat and breathe hard) can be interpreted as either moderate or vigorous activity.63 Using a ?less than six days? cutoff to define low physical activity improves upon previous work that used a 3-day cutoff, but does not accurately identify all children who fail to meet the current physical activity guidelines. Unlike some previous analyses of the data from 2007 NSCH,70 we have included both recreational computer use and time spent with television, videos and video games in our measurement of screen time. This is better aligned with the AAP guidelines, and indicates that a higher proportion of the population exceeds the guidelines. However, the 2007 NSCH asks only about media use ?on a typical weekday,? and this might underestimate the average daily screen time including weekends. Our estimate of mean screen time was considerable lower than the Kaiser Family Foundation estimates which included weekends as well.15 There may be some question about how accurate parent reports are, especially with older children. The health implications of screen-based leisure activity lying on a sofa between the remote and a bowl of snack food can be quite different from accessing the same content while on the go. The 2007 NSCH predates widespread use of smart phones and tablet computers; further research will be needed to examine changing patterns of media use with greater availability of more portable devices. Moreover, new ways of accessing viewing content has changed both the programming and the advertising that children are exposed to. New tools such as ecologic momentary assessment coupled with 52 accelerometry data would provide a more complete assessment of physical activity and sedentary behavior.71 BMI is a potentially important covariate because it is associated with low MVPA, high screen time, sedentary lifestyle and special health care needs.4,26,65,66,68 In the 2007 NSCH, the BMI classification was based on parental reports of height and weight. Because the BMI classification was found to lack reliability for children less than 10 years old, it was not included in the publicly available data set for children under 10. We did not include BMI in our models because it was no available for the younger children and because of concerns about the validity of parental reports of heights and weights.72 Parental report of special health care needs and specific diagnoses in this study are not confirmed by physician diagnosis or other objective documentation of the child's needs and condition. The survey did not provide information about physical mobility issues that might interfere with physical activity. Because of the limited list of specific diagnostic categories, we were unable to define a category of those with emotional or behavioral problems without a diagnosis of a physical problem. Because of these issues, the implications of the findings about parental perception of the child?s limitations are unclear. Elucidation of the factors that contribute to low MVPA and high screen time, especially among those with functional limitations, will require more specific surveys to explore perceived barriers to participation in more active pursuits. Random-digit dialing does not capture households without landline phones. While sampling weights include adjustments for non-response and lack of phone lines, we cannot know how respondents differ from non-respondents in terms of the key variables. 53 In our analysis, we did not make adjustments for multiple comparisons. Because we made many comparisons, this increases the probability that some of the "significant" associations may have occurred by chance. Because the data are cross-sectional, causal inferences cannot be made. The cross- sectional data also do not adequately capture the dynamic nature of chronic conditions in childhood. 7. Public health significance Advances in technology have enabled more children to survive to live with the consequences of prematurity, congenital anomalies, cystic fibrosis, sickle cell anemia, cancer, HIV/AIDS and other conditions which previously caused death during childhood.72 With the availability of better pharmaceutical treatments, identification of children with more common, less lethal conditions, such as allergies, asthma and attention deficit disorder has increased.73 However, this growing population of children with special health care needs strains the resources of families9 and the health care delivery system.50 Optimizing the health and well-being of the subpopulation of children who have SHCN is important, not only to improve their quality of life and decrease their need for expensive health care and therapies, but also to decrease the development of comorbidities. Common chronic conditions of adulthood, such as obesity, type 2 diabetes, and resultant cardiovascular consequences have their roots in childhood.74,75 Physical activity and screen-based leisure activity are modifiable factors that can have ramifications for health throughout the life course. The American Academy of Pediatrics recommends that pediatricians routinely ask screening questions about physical activity 54 and screen time during well child visits.1 When children do not meet the current guidelines, health care providers can help parents reevaluate priorities and find ways to work more activity into daily routines. The advice of health professionals is especially important for children whose parents perceive them as having functional limitations, so that inappropriate restriction of activity can be avoided and appropriate activities can be recommended. Health care professionals and educators should help families optimize their children's engagement in developmentally appropriate activities within the ever- changing constraints of their special health care needs. Schools and communities play an essential role in providing opportunities for all children to be physically active. CDC school health guidelines call for inclusive physical education programs with appropriate modifications so that all children can be more active.17 School-based programs that monitor physical fitness and provide fitness report cards76 can help parents and school personnel recognize physical activity and physical fitness as important priorities as they decide how children should be spending their time. Because of the recent trends in childhood obesity, improved surveillance of common obesity-related behaviors in children is warranted. To determine whether progress is being made toward the achievement of Healthy People 2020 objectives regarding physical activity, the National Survey of Children?s Health should revise the questions for parents of school-aged children to better identify whether or not children get at least 60 minutes of moderate-to-vigorous physical activity daily, at least 20 minutes of vigorous activity 3 days a week and muscle strengthening activity three days a week. Further study of the impact of media use on activity levels, including accelerometer data 55 and ecological momentary assessment, will be important as children spend more time with an ever-expanding array of electronic devices. Changing the behavior of children requires changing the behavior of parents. Our data indicate a significant association between sedentary lifestyles and parental perception of functional limitations in the child, parenting stress and parents' trust in their neighbors. Better understanding of parents' ideas about what their children can and should do, as well as their perceived barriers to participation in active endeavors, can aid in the development of interventions to promote lifestyles that optimize the growth and development of all children, including those with special health care needs. 56 VII. APPENDIX 1: TABLES Table 1: Socio-demographic characteristics of the 2007 National Survey of Children's Health study population, with population percents. United States, 2007 Number in sample Population Percent (weighted) Total 64076 Gender 63985 Boys 33292 51.14 Girls 30693 48.86 Age in years 64076 6 to 11 27792 48.53 12 to 17 36284 51.47 Race/ethnicity 62985 Non-Hispanic white 43789 57.22 Non-Hispanic black 6450 15.06 Hispanic 7357 19.36 Non-Hispanic multiracial 2776 3.79 Non-Hispanic other race 2613 4.57 Respondent's relationship to child 64064 Mother 46750 74.61 Father 13388 18.83 Other 3926 6.57 Respondent's education 63248 < 12 years 5269 12.33 High school graduate 13075 25.70 More than high school 44904 61.96 Household poverty ratio 58700 <= 100% 6113 16.60 >100 and <=200% 9623 20.61 >200 and <=300% 10787 18.38 >300 and <=400% 9469 14.18 >400% 22708 30.22 Special Health Care Needs Status 64076 No SHCN 49027 77.11 SHCN without EBD 7527 11.23 SHCN with EBD 7522 11.67 Note: Number of sample children for varies because observations with missing data were excluded from analysis. Data from National Survey of Children's Health, 2007 back to text 57 Table 2. Frequency of specific diagnostic categories listed in National Survey of Children's Health 2007, with population estimates Number in sample Population Prevalence (percent) (weighted) Estimated population frequency Emotional Behavioral and Developmental Conditions Attention deficit disorder 5338 8.18 4,010,749 Depression 1662 2.49 1,224,562 Anxiety 2530 3.52 1,731,070 Behavior or conduct disorder 2182 4.00 1,967,373 Autism spectrum disorder 759 1.16 569,154 Developmental delay 1983 3.49 1,716,259 Tourette's Syndrome 147 0.19 92,087 Other specific conditions Learning disability 5477 8.98 4,408,694 Asthma 6357 10.31 5,066,022 Diabetes 329 0.55 270,170 Speech problem 1844 3.37 1,661,124 Hearing problem 995 1.54 759,984 Vision problem (not correctable with glasses) 887 1.56 766,903 Seizure disorder 422 0.75 370,748 Brain injury/concussion 196 0.43 209,487 Bone, muscle or joint problem 1824 2.73 1,343,779 Respiratory allergy 13238 18.99 9,340,978 Food allergy 2925 4.28 2,103,831 Skin allergy 6986 11.08 5,457,175 Migraine headaches 3579 5.29 2,606,054 Recurrent ear infections 2217 3.86 1,898,390 back to text 58 Table 3: Weighted prevalence of low physical activity, high screen time and sedentary lifestyle by selected demographic characteristics, parental attitude indicators and special health care needs status. United States, 2007 Low MVPA High screen time Sedentary lifestyle No. in sample Percent (weighted) p value Percent (weighted) p value Percent (weighted) p value Total 64076 64.2 48.4 33.3 Gender .0001 .0038 .0001 Boys 33292 57.9 49.8 31.5 Girls 30693 70.8 46.9 35.1 Age in years 0.0001 0.0001 0.0001 6 4447 52.8 31.9 17.6 7 4520 53.4 34.7 19.2 8 4521 56.3 37.1 22.0 9 4554 54.9 38.4 21.8 10 4903 60.8 44.8 29.6 11 4641 65.5 47.1 33.3 12 5246 69.0 49.5 34.4 13 5332 66.3 56.5 38.2 14 5793 69.0 56.9 43.3 15 6008 73.0 59.4 45.2 16 6632 73.6 61.0 46.8 17 6837 73.8 59.7 44.9 Race/ethnicity 0.0001 0.0001 0.0001 Non-Hispanic white 43789 61.6 44.8 30.4 Non-Hispanic black 6450 64.9 63.7 42.2 Hispanic 7357 71.5 48.2 36.3 Non-Hispanic multiracial 2776 55.0 48.7 30.1 Non-Hispanic other race 2613 69.2 43.8 31.2 Respondent's relationship to child 0.2859 0.0001 0.0023 Mother 46750 64.0 46.5 32.3 Father 13388 65.7 52.7 36.2 Other 3926 62.4 57.0 36.1 Respondent's education 0.0001 0.0001 0.0001 < 12 years 5269 71.5 50.7 37.5 High school graduate 13075 66.0 56.2 39.3 More than high school 44904 62.0 44.6 30.1 Household poverty ratio 0.1182 0.0001 0.0001 <= 100% 6113 66.9 50.2 34.7 >100 and <=200% 9623 64.7 54.0 36.8 >200 and <=300% 10787 63.4 52.4 36.2 >300 and <=400% 9469 64.0 47.4 32.5 >400% 22708 62.8 42.5 29.6 Special Health Care Needs Status 0.1714 0.0006 0.0005 No SHCN 49027 63.8 47.6 32.6 SHCN without EBD 7527 64.8 48.3 34.9 SHCN with EBD 7522 66.5 53.5 37.8 59 Table 3 (cont'd): Weighted prevalence of low physical activity, high screen time and sedentary lifestyle by selected demographic characteristics, parental attitude indicators and special health care needs status. United States, 2007 Low MVPA High screen time Sedentary lifestyle No. in sample Percent (weighted) p value Percent (weighted) p value Percent (weighted) p value Child with perceived limitations 0.0009 0.0026 0.0001 no 59143 63.7 48.0 32.7 yes 4208 70.3 54.0 41.1 Parental mental health/parenting stress 0.035 0.0001 0.0001 zero 52003 63.2 46.4 31.5 one 8399 67.0 54.3 39.1 two 1816 69.9 58.8 41.6 three 747 71.4 57.4 44.4 four 234 61.2 61.9 41.8 five 54 62.1 61.5 31.1 Social Support 0.0001 0.006 0.0048 yes 57312 63.2 47.7 32.7 no 6490 70.6 52.4 37.2 Trust in neighbors 0.0648 0.0001 0.0001 zero 51736 63.4 46.6 32.2 one 4916 66.5 53.3 36.7 two 2368 65.4 45.3 34.5 three 1522 68.6 54.7 39.0 four 1420 69.1 61.2 43.6 Perceived safety 0.0001 0.0001 0.0001 zero 52658 63.1 47.0 32.3 one 6957 68.1 52.9 36.2 two 2063 71.4 59.6 43.3 Number of conditions 0.3831 0.0001 0.0002 none 33583 63.5 46.0 31.4 one 15720 64.8 50.6 35.2 two 7211 66.3 52.9 36.8 three 3272 63.3 49.1 34.9 four or more 4075 65.1 52.0 35.8 Severity of condition 0.0535 0.7223 0.407 none severe 59625 64.4 48.8 333.2 one severe 2976 59.5 49.9 33.6 more than one severe 1027 68.1 48.3 37.9 BMI classification (ages 10-17 years) 44101 0.0001 0.0001 0.0001 <5 2186 66.4 55.0 37.9 5 to 85 29121 66.7 51.9 36.9 85-95 6754 72.8 58.8 44.8 >95 6040 72.9 63.7 47.7 Data from the 2007 National Survey of Children's Health back to text 60 Table 4: Weighted prevalence of special health care needs status among 6-17 year olds by selected demographic characteristics and parental attitude indicators. United States 2007 Without SHCN SHCN without EBD SHCN with EBD No. in sample Percent (weighted) Percent (weighted) Percent (weighted) p value Total 64076 77.11 11.23 11.67 Gender 0.0001 Boys 33292 73.53 11.06 15.41 Girls 30693 80.83 11.41 7.76 Age 6 4447 79.07 12.20 8.74 7 4520 78.74 12.18 9.08 8 4521 78.01 10.35 11.64 9 4554 74.95 11.53 13.52 10 4903 77.00 11.05 11.95 11 4641 77.49 11.07 11.44 12 5246 77.49 11.64 10.88 13 5332 77.55 10.18 12.27 14 5793 76.32 11.59 12.09 15 6008 74.50 11.09 14.41 16 6632 77.56 11.06 11.38 17 6837 76.66 10.77 12.57 Race/ethnicity 0.0001 Non-Hispanic white 43789 75.27 12.31 12.41 Non-Hispanic black 6450 75.96 10.69 13.34 Hispanic 7357 81.95 9.22 8.83 Non-Hispanic multiracial 2776 73.24 11.51 15.24 Non-Hispanic other race 2613 86.09 7.98 5.94 Respondent's relationship to child 0.0001 Mother 46750 76.06 11.95 12.00 Father 13388 82.37 9.37 8.26 Other 3926 74.00 8.38 17.62 Respondent's education 0.001 < 12 years 5269 79.48 8.34 12.18 High school graduate 13075 76.82 10.48 12.69 More than high school 44904 76.57 12.20 11.23 Household poverty ratio 0.0001 <= 100% 6113 72.74 10.24 17.01 >100 and <=200% 9623 77.11 9.93 12.96 >200 and <=300% 10787 77.63 11.47 10.84 >300 and <=400% 9469 76.68 13.48 8.84 >400% 22708 77.34 12.53 10.13 Data from the 2007 National Survey of Children's Health 61 Table 4 (cont'd): Weighted prevalence of special health care needs status among 6-17 year olds by selected demographic and characteristics and parental attitude indicators. United States 2007 Without SHCN SHCN without EBD SHCN with EBD No. in sample Percent (weighted) Percent (weighted) Percent (weighted) p value Child with perceived limitations 0.0001 no 59723 81.558 10.2771 8.1671 yes 4266 19.1714 23.6061 57.2224 Mental health/parenting stress score 0.0001 zero 52003 80.27 11.43 8.29 one 8399 70.99 11.67 17.34 two 1816 53.69 8.86 37.45 three 747 42.76 8.38 48.86 four 234 37.71 3.86 58.44 five 54 34.36 21.48 44.16 Social Support 0.0001 yes 57312 77.16 11.63 11.22 no 6490 76.43 8.66 14.90 Trust in neighbors score 0.0001 zero 51736 77.93 11.33 10.74 one 4916 75.71 11.85 12.44 two 2368 71.46 11.21 17.33 three 1522 68.54 11.05 20.42 four 1420 71.20 11.58 17.22 Perceived safety score zero 52658 77.65 11.34 11.01 0.0028 one 6957 74.99 10.78 14.24 two 2063 73.69 11.15 15.15 Number of conditions 0.0001 none 33583 96.28 3.03 0.69 one 15720 74.08 18.07 7.85 two 7211 50.94 27.17 21.89 three 3272 30.15 31.67 38.18 four or more 4075 12.01 12.11 75.88 Severity of conditions 0.0001 none severe 59625 80.76 10.40 8.85 one severe 2976 43.21 23.15 33.64 more than one severe 1027 4.65 12.82 82.53 BMI classification (ages 10-17) Total in sample 44101 0.0002 <5%ile 2186 76.743 9.713 13.544 5-85%ile 29121 78.167 10.745 11.088 85-95%ile 6754 74.088 12.590 13.322 >95%ile 6040 71.926 12.492 15.582 back to text 62 Table 5. Prevalence of low moderate-to-vigorous physical activity among 6- to 17-year-olds by special health care needs status and parental attitude indicators. United States 2007 No. in sample Percent (weighted) 95% CI p value Special health care needs status 63434 0.1714 Without SHCN 48531 63.7906 62.6749 64.9063 SHCN without EBD 7466 64.8183 62.0599 67.5768 SHCN with EBD 7437 66.4606 63.9236 68.9977 Child with perceived limitations 63351 0.0009 no 59143 63.743 62.7433 64.7427 yes 4208 70.2561 66.7227 73.7896 Mental health/stress score 62656 0.0006 zero 51599 63.2157 62.1466 64.2848 1 to 5 11057 67.6297 65.4147 69.8446 Trust in neighbors score 61404 0.0047 zero 51330 63.4062 62.3139 64.4986 1 to 4 10074 67.0288 64.8054 69.2523 Social Support 63170 <.0001 yes 56812 63.2357 62.2147 64.2566 no 6358 70.5818 67.7743 73.3893 Child's perceived safety score 63434 <.0001 zero 54566 63.1618 62.1085 64.2151 1 to 2 8868 68.9655 66.6577 71.2734 p values indicate the probability of the observed Rao Scott chi square if there are no true differences between catetories. Data from 2007 National Survey of Children's Health back to text 63 Table 6: Hierarchical logistic regression models for effects of demographic characteristics, special health care needs status and parental attitudes on low moderate-to-vigorous physical activity in 6-17 year old children. United States 2007 Unadjusted Model 1* OR 95% CI p value AOR 95% CI p value Gender <.0001 <.0001 Boys reference Girls 1.748 1.6 1.909 1.77 1.62 1.93 Race/ethnicity <.0001 <.0001 Non-Hispanic white reference Non-Hispanic black 1.149 1.019 1.295 1.11 0.978 1.263 Hispanic 1.577 1.356 1.834 1.51 1.293 1.765 Non-Hispanic multiracial 0.791 0.639 0.979 0.82 0.652 1.04 Non-Hispanic other race 1.362 1.047 1.772 1.46 1.12 1.897 Respondent's education <.0001 0.0008 < 12 years 1.566 1.334 1.838 1.34 1.132 1.588 High school graduate 1.18 1.068 1.303 1.12 1.012 1.241 More than high school reference Age in years <.0001 <.0001 1.101 1.087 1.115 1.1 1.09 1.119 Special Health Care Needs Status 0.2627 No SHCN reference SHCN without EBD 1.05 0.918 1.2 SHCN with EBD 1.107 0.974 1.258 Child with perceived limitations 0.0007 no reference yes 1.379 1.145 1.662 Mental health/stress score 0.0007 zero reference 1 - 5 1.224 1.089 1.375 Trust in neighbors score 0.0027 zero reference 1 - 4 1.189 1.062 1.331 Social Support <.0001 yes reference no 1.384 1.19 1.609 Child's perceived safety score 0.0027 zero reference 1 - 2 1.391 1.237 1.565 *Model 1 adjusts for gender, race/ethnicity, respondent's education and child's age 64 Table 6 (cont'd): Hierarchical logistic regression models for effects of demographic characteristics, special health care needs status and parental attitudes on low moderate-to-vigorous physical activity Model 2** Model 3*** AOR 95% CI p value AOR 95% CI p value Gender <.0001 <.0001 Boys Girls 1.798 1.646 1.964 1.8 1.645 1.962 Race/ethnicity <.0001 <.0001 Non-Hispanic white Non-Hispanic black 1.113 0.98 1.264 1.06 0.932 1.207 Hispanic 1.525 1.306 1.782 1.47 1.254 1.727 Non-Hispanic multiracial 0.82 0.649 1.035 0.81 0.637 1.019 Non-Hispanic other race 1.483 1.139 1.929 1.42 1.091 1.857 Respondent's education 0.001 0.013 < 12 years 1.336 1.127 1.583 1.27 1.068 1.507 High school graduate 1.118 1.01 1.238 1.09 0.987 1.213 More than high school Age in years <.0001 <.0001 1.104 1.09 1.118 1.1 1.089 1.118 Special Health Care Needs Status 0.0028 0.2397 No SHCN SHCN without EBD 1.097 0.961 1.252 1.06 0.923 1.214 SHCN with EBD 1.241 1.092 1.411 1.12 0.977 1.288 Child with perceived limitations 0.008 no yes 1.34 1.079 1.662 Mental health/stress score 0.8181 zero 1- 5 0.99 0.868 1.118 Trust in neighbors score 0.2766 zero 1 - 4 1.07 0.949 1.201 Social Support 0.0963 yes no 1.15 0.976 1.345 Child's perceived safety score 0.1514 zero 1 - 2 1.1 0.965 1.257 **Model 2 adjusts for gender, race/ethnicity, respondent's education, child's age and special health care needs status ***Model 3 adjusts for gender, race/ethnicity, respondent's education, child's age, special health care needs status, perceived limitations, parental mental health/stress, trust in neighbors, social support and perceived safety back to text 65 Table 7. Prevalence of high screen time among 6- to-17-year-olds by special health care needs status and parental attitude indicators. United States 2007 No. in sample Percent (weighted) 95% CI p value Special health care needs status 64076 0.0006 Without SHCN 49027 47.584 46.446 48.722 SHCN without EBD 7527 48.2617 45.3813 51.1421 SHCN with EBD 4151 53.5214 50.7685 56.2743 Child with perceived limitations 63989 0.0026 no 59723 47.9562 46.9301 48.9824 yes 4266 53.9768 50.2132 57.7404 Mental health/stress score 63253 <.0001 zero 52003 46.4086 45.3145 47.5027 1 to 5 11250 55.477 53.1526 57.8014 Trust in neighbors score 61962 <.0001 zero 51736 46.5649 45.4583 47.6715 1 to 4 10226 55.1882 52.8082 57.5681 Social Support 63802 0.006 yes 57312 47.7374 46.7002 48.7746 no 6490 52.363 49.2375 55.4885 Child's perceived safety score 64076 <.0001 zero 55056 46.945 45.8751 48.015 1 to 2 9020 54.6341 52.0852 57.183 p values indicate the probability of the observed Rao Scott chi square if there are no true differences between catetories Data from 2007 National Survey of Children's Health back to text 66 Table 8: Hierarchical logistic regression models for effects of demographic characteristics, special health care needs status and parental attitudes on high screen time among 6-17 year old children, United States 2007 Unadjusted Model 1* OR 95% CI p value OR 95% CI p value Gender 0.004 0.0036 Boys reference Girls 0.887 0.817 0.962 0.881 0.809 0.959 Race/ethnicity <.0001 <.0001 Non-Hispanic white reference Non-Hispanic black 2.173 1.938 2.438 2.144 1.9 2.418 Hispanic 1.164 1.014 1.337 1.14 0.982 1.323 Non-Hispanic multiracial 1.138 0.926 1.399 1.218 0.996 1.489 Non-Hispanic other race 0.926 0.729 1.175 1.028 0.805 1.314 Respondent's education <.0001 <.0001 < 12 years 1.296 1.111 1.511 1.215 1.03 1.434 High school graduate 1.581 1.435 1.743 1.462 1.323 1.615 More than high school reference Age in years <.0001 <.0001 1.13 1.115 1.144 1.13 1.115 1.144 Special health care needs status No SHCN 0.0026 SHCN without EBD 1.031 0.907 1.173 SHCN with EBD 1.242 1.098 1.404 Child with perceived limitations 0.0007 no reference yes 1.379 1.145 1.662 Mental health/stress score zero reference 0.0007 1 through 5 1.224 1.089 1.375 Trust in neighbors score zero reference 0.0027 1 through 4 1.189 1.062 1.331 Social Support 0.0057 yes reference no 1.223 1.06 1.41 Child's perceived safety score <.0001 zero reference 1 through 2 1.391 1.237 1.565 *Model 1 adjusts for gender, race/ethnicity, respondent's education and child's age 67 Table 8 (cont'd): Hierarchical logistic regression models for effects of demographic characteristics, special health care needs status and parental attitudes on high screen time in 6-17 year old children, United States 2007 Model 2* Model 3* OR 95% CI p value OR 95% CI p value Gender 0.0081 0.0057 Boys reference reference Girls 0.891 0.818 0.971 0.887 0.815 0.966 Race/ethnicity <.0001 <.0001 Non-Hispanic white reference reference Non-Hispanic black 2.146 1.902 2.421 2.007 1.77 2.276 Hispanic 1.147 0.988 1.332 1.079 0.927 1.256 Non-Hispanic multiracial 1.213 0.992 1.484 1.178 0.961 1.443 Non-Hispanic other race 1.04 0.814 1.33 1.005 0.786 1.286 Respondent's education reference <.0001 reference <.0001 < 12 years 1.213 1.028 1.431 1.127 0.952 1.334 High school graduate 1.46 1.322 1.613 1.415 1.28 1.564 More than high school Age in years reference <.0001 reference <.0001 1.13 1.115 1.144 1.129 1.115 1.144 Special health care needs status 0.0472 0.4332 No SHCN reference reference SHCN without EBD 1.057 0.929 1.202 1.042 0.91 1.193 SHCN with EBD 1.172 1.031 1.333 1.093 0.951 1.257 Child with perceived limitations 0.8005 no reference yes 1.025 0.846 1.242 Mental health/stress score 0.0056 zero reference 1 - 5 1.189 1.052 1.344 Trust in neighbors score 0.0003 zero reference 1 - 4 1.243 1.104 1.399 Social Support 0.6128 yes reference no 1.041 0.89 1.219 Child's perceived safety score 0.6551 zero reference 1 - 2 1.03 0.904 1.173 **Model 2 adjusts for gender, race/ethnicity, respondent's education, child's age and special health care needs status ***Model 3 adjusts for gender, race/ethnicity, respondent's education, child's age, special health care needs status, perceived limitations, parental mental health/stress, trust in neighbors, social support and perceived safety back to text 68 Table 9. Prevalence of sedentary lifestyle among 6- to 17-year-olds by special health care needs status and parental attitude indicators. United States 2007 No. in sample Percent (weighted) 95% CI p value Special health care needs status 63434 0.0005 Without SHCN 48531 32.3991 31.3391 33.4591 SHCN without EBD 7466 34.9007 32.053 37.7484 SHCN with EBD 7437 37.8302 35.2155 40.4448 Child with perceived limitations 63351 <.0001 no 59143 32.7374 31.7746 33.7002 yes 4208 41.0508 37.4058 44.6958 Mental health/stress score 62656 <.0001 zero 51599 31.5243 30.5101 32.5386 1 to 5 11057 39.9264 37.6468 42.2059 Trust in neighbors score 61404 <.0001 zero 51330 32.1966 31.1634 33.2298 1 to 4 10074 37.6902 35.3964 39.9841 Social Support 63170 yes 56812 32.7054 31.7364 33.6743 0.0048 no 6358 37.1527 34.1393 40.1662 Child's perceived safety score 63434 <.0001 zero 54566 32.2592 31.2541 33.2643 1 to 2 8868 38.0467 35.6366 40.4568 p values indicate the probability of the observed Rao Scott chi square if there are no true differences between categories. Data from 2007 National Survey of Children's Health back to text 69 Table 10 Hierarchical logistic regression models for effects of demographic characteristics, special health care needs status and parental attitudes on sedentary lifestyle in 6-17 year old children, United States 2007 Unadjusted Model 1* OR 95% CI p value AOR 95% CI p value Gender 0.001 0.0006 Boys reference reference Girls 1.155 1.06 1.259 1.167 1.068 1.275 Race/ethnicity <.0001 <.0001 Non-Hispanic white reference reference Non-Hispanic black 1.651 1.475 1.848 1.61 1.427 1.818 Hispanic 1.326 1.149 1.531 1.297 1.115 1.509 Non-Hispanic multiracial 0.998 0.803 1.241 1.073 0.861 1.337 Non-Hispanic other race 0.975 0.758 1.253 1.09 0.852 1.394 Respondent's education <.0001 <.0001 < 12 years 1.416 1.207 1.66 1.275 1.074 1.513 High school graduate 1.503 1.361 1.66 1.392 1.257 1.541 More than high school reference reference Age in years <.0001 <.0001 1.151 1.136 1.166 1.151 1.136 1.167 Special Health Care Needs Status 0.0008 No SHCN reference SHCN without EBD 1.129 0.983 1.298 SHCN with EBD 1.255 1.107 1.422 Child with perceived limitations no reference <.0001 yes 1.45 1.227 1.713 Mental health/stress score <.0001 zero reference 1 - 5 1.506 1.349 1.681 Trust in neighbors score <.0001 zero reference 1 - 4 1.286 1.152 1.437 Social Support 0.002 yes reference no 1.259 1.088 1.457 Child's perceived safety score <.0001 zero reference 1 - 2 1.318 1.173 1.481 *Model 1 adjusts for gender, race/ethnicity, respondent's education and child's age back to text 70 Table 10 (cont'd): Hierarchical logistic regression models for effects of demographic characteristics, special health care needs status and parental attitudes on sedentary lifestyle among 6-17 year old children, United States 2007 Model 2** Model 3*** AOR 95% CI p value AOR 95% CI p value Gender 0.0001 0.0002 Boys reference reference Girls 1.187 1.087 1.298 1.184 1.084 1.294 Race/ethnicity <.0001 <.0001 Non-Hispanic white reference reference Non-Hispanic black 1.616 1.432 1.824 1.529 1.351 1.73 Hispanic 1.313 1.128 1.527 1.253 1.072 1.464 Non-Hispanic multiracial 1.066 0.856 1.327 1.04 0.832 1.3 Non-Hispanic other race 1.112 0.869 1.424 1.078 0.839 1.384 Respondent's education <.0001 <.0001 < 12 years 1.274 1.074 1.511 1.181 0.992 1.407 High school graduate 1.392 1.257 1.541 1.352 1.219 1.498 More than high school reference reference Age in years <.0001 <.0001 1.151 1.136 1.167 1.151 1.135 1.166 Special Health Care Needs Status 0.0004 0.1412 No SHCN reference reference SHCN without EBD 1.177 1.02 1.358 1.131 0.973 1.314 SHCN with EBD 1.267 1.111 1.445 1.114 0.967 1.284 Child with perceived limitations no reference 0.0329 yes 1.245 1.018 1.522 Mental health/stress score 0.0026 zero reference 1 - 5 1.206 1.068 1.363 Trust in neighbors score 0.0227 zero reference 1 - 4 1.149 1.02 1.295 Social Support 0.5997 yes reference no 1.043 0.892 1.219 Child's perceived safety score 0.8895 zero reference 1 - 2 0.991 0.87 1.129 **Model 2 adjusts for gender, race/ethnicity, respondent's education, child's age and special health care needs status ***Model 3 adjusts for gender, race/ethnicity, respondent's education, child's age, special health care needs status, perceived limitations, parental mental health/stress, trust in neighbors, social support and perceived safety. Data from the 2007 National Survey of Children's Health 71 Table 11. Effects of interaction terms in models for effects of parental attitudes, adjusted for age. gender, race/ethnicity, education of respondent and special health care needs status SHCN without EBD SHCN with EBD p for effect ? coefficient p value ? coefficient p value Low MVPA limitations*SHCN status -0.123 0.7457 -0.3011 0.4221 0.5772 mental health/stress*SHCN status 0.1739 0.3781 -0.0356 0.806 0.6201 trust in neighbors* SHCN status 0.0296 0.8659 -0.2191 0.1537 0.3287 social support*SHCN status 0.2312 0.3429 -0.1747 0.3595 0.3478 perceived safety*SHCN status -0.0755 0.7069 -0.2811 0.0855 0.2269 High screen time limitations*SHCN status 0.0135 0.2728 0.0024 0.9605 0.0252 mental health/stress*SHCN status 0.1672 0.3866 -0.0404 0.7757 0.6181 trust in neighbors* SHCN status -0.0893 0.6252 0.152 0.2925 0.7217 social support*SHCN status 0.3062 0.3199 -0.2273 0.2278 0.2422 perceived safety*SHCN status 0.1528 0.4379 -0.017 0.9195 0.7217 Sedentary lifestyle limitations*SHCN status 0.1928 0.5154 -0.1227 0.676 0.2973 mental health/stress*SHCN status 0.0141 0.9248 -0.0177 0.7755 0.9519 trust in neighbors* SHCN status 0.0219 0.9076 0.0011 0.9943 0.9932 social support*SHCN status 0.2453 0.4382 -0.2955 0.1151 0.1758 perceived safety*SHCN status 0.1989 0.3314 -0.096 0.5705 0.4817 back to text 72 Table 12: Effects of parental attitudes on odds of low moderate-to-vigorous physical activity, adjusted for age, gender, race/ethnicity and education of respondent, stratified by special health care needs status All children Children without SHCN AOR 95% CI p value AOR 95% CI p value Child with perceived limitations 0.0003 0.1697 no reference reference yes 1.445 1.182 1.765 1.602 0.818 3.141 Mental health/stress score 0.2888 0.5918 zero reference reference 1 - 5 1.067 0.946 1.204 1.107 0.763 1.608 Trust in neighbors score zero reference 0.065 reference 0.078 1 - 5 1.115 0.993 1.252 1.132 0.986 1.299 Social Support 0.0464 0.0892 yes reference reference no 1.172 1.003 1.371 1.174 0.976 1.412 Child's perceived safety score zero reference 0.0415 reference 0.0231 1 - 2 1.143 1.005 1.301 1.191 1.024 1.385 Children with SHCN without EBD Children with SHCN with EBD AOR 95% CI p value AOR 95% CI p value Child with perceived limitations 0.0231 0.0917 no reference reference yes 1.423 1.05 1.93 1.272 0.962 1.684 Mental health/stress score 0.5918 0.6843 zero reference reference 1 - 5 1.107 0.763 1.608 1.053 0.82 1.354 Trust in neighbors score zero reference 0.6203 reference 0.7549 1 - 5 1.084 0.789 1.489 0.958 0.73 1.256 Social Support 0.2489 0.8832 yes reference reference no 1.292 0.836 1.997 1.027 0.72 1.464 Child's perceived safety score zero reference 0.7831 reference 0.8074 1 - 2 0.947 0.641 1.399 0.963 0.713 1.302 p values indicate the probability of the observed Wald chi square if there is no true difference in odds of low MVPA by level of parental attitude indicator. Data from the 2007 National Survey of Children's Health back to text 73 Table 13. Effect of special health care needs status on adjusted odds of low moderate-to-vigorous physical activity, adjusted for age, gender, race/ethnicity and education of respondent, stratified by parental attitudes Children without SHCN Children with SHCN without EBD Children with SHCN with EBD AOR 95% CI AOR 95% CI p value Overall (age 6-17) 1.097 0.961 1.252 1.241 1.092 1.411 0.0028 Stratified by Mental health/stress zero reference 1.047 0.908 1.206 1.249 1.068 1.46 0.0194 1 - 5 reference 1.242 0.884 1.746 1.154 0.917 1.453 0.284 Trust in neighbors score zero reference 1.07 0.924 1.24 1.285 1.111 1.487 0.0031 1 - 4 reference 1.118 0.819 1.526 1.013 0.786 1.305 0.7809 Perceived safety score zero reference 1.078 0.936 1.242 1.271 1.104 1.463 0.0033 1 - 2 reference 1.088 0.763 1.552 1.045 0.782 1.396 0.8739 Child with perceived limitations no reference 1.035 0.895 1.196 1.179 1.026 1.354 0.0651 yes reference 0.823 0.446 1.518 0.703 0.402 1.229 0.3527 Social support yes reference 0.703 0.402 1.229 1.26 1.099 1.444 0.0036 no reference 1.296 0.797 2.106 0.975 0.696 1.364 0.5522 p values indicate the probability of the observed Wald chi square if there is no true difference in odds of low MVPA among SHCN categories back to text 74 Table 14: Effects of parental attitudes on odds of exceeding 2 hours of screen-based leisure activity, adjusted for age, gender, race/ethnicity and education of respondent, stratified by special health care needs status All children Children without SHCN AOR 95% CI p value AOR 95% CI p value Child with perceived limitations 0.146 0.7617 no reference reference yes 1.14 0.955 1.361 0.932 0.59 1.472 Mental health/stress score 0.0002 0.0044 zero reference reference 1 through 5 1.249 1.113 1.403 1.239 1.069 1.435 Trust in neighbors score <.0001 0.0003 zero reference reference 1 through 4 1.287 1.149 1.442 1.283 1.121 1.469 Social Support 0.1826 0.1687 yes reference reference no 1.11 0.952 1.293 1.132 0.949 1.351 Child's perceived safety score 0.0904 0.1552 zero reference reference 1 through 2 1.114 0.983 1.263 1.112 0.961 1.287 Children with SHCN without EBD Children with SHCN with EBD AOR 95% CI p value AOR 95% CI p value Child with perceived limitations 0.0042 0.5347 no reference reference yes 1.494 1.135 1.967 0.918 0.701 1.202 Mental health/stress score 0.25 0.1864 zero reference reference 1 through 5 1.234 0.862 1.765 1.171 0.927 1.48 Trust in neighbors score 0.8157 0.0014 zero reference reference 1 through 4 1.042 0.737 1.473 1.509 1.172 1.942 Social Support 0.3444 0.4704 yes reference reference no 1.332 0.735 2.413 0.889 0.646 1.223 Child's perceived safety score 0.6998 0.55 zero reference reference 1 through 2 1.081 0.727 1.608 1.095 0.813 1.474 p values indicate the probability of observed Wald chi square if there is no true difference between levels of the parental attitude indicators. Data from the 2007 National Survey of Children's Health back to text 75 Table 15. Effect of special health care needs status on odds of exceeding 2 hours of screen-based leisure activity, adjusted for age, gender, race/ethnicity and education of respondent, stratified by parental attitudes Children without SHCN Children with SHCN without EBD Children with SHCN with EBD AOR 95% CI AOR 95% CI p value Overall 1.057 0.929 1.202 1.172 1.031 1.333 0.0472 Stratified by Mental health/stress zero reference 1.015 0.886 1.163 1.152 0.976 1.359 0.2461 1-5 reference 1.148 0.813 1.621 1.067 0.857 1.328 0.676 Trust in neighbors score zero reference 1.072 0.935 1.229 1.104 0.954 1.278 0.296 1 -4 reference 0.938 0.679 1.296 1.409 1.1 1.804 0.0167 Perceived safety score zero reference 1.041 0.909 1.192 1.195 1.037 1.377 0.0473 1-2 reference 1.09 0.769 1.545 1.155 0.868 1.538 0.5797 Child with perceived limitations no reference 0.985 0.856 1.135 1.237 1.071 1.428 0.0128 yes reference 1.452 0.902 2.337 1.047 0.662 1.656 0.0798 Social support yes reference 0.703 0.402 1.229 1.26 1.099 1.444 0.0036 no reference 1.296 0.797 2.106 0.975 0.696 1.364 0.5522 p values indicate the probability of the observed Wald chi square if there is no true difference in odds of high screen time among the SHCN categories back to text 76 Table 16: Effects of parental attitudes on odds of sedentary lifestyle adjusted for age, gender, race/ethnicity and education of respondent, stratified by special health care needs status All children Children without SHCN AOR 95% CI p value AOR 95% CI p value Child with perceived limitations 0.0003 0.3073 no reference reference yes 1.407 1.17 1.69 1.296 0.788 2.133 Mental health/stress score <.0001 0.0094 zero reference reference 1 - 5 1.28 1.139 1.438 1.212 1.048 1.402 Trust in neighbors score 0.0031 0.0148 zero reference reference 1 - 4 1.191 1.061 1.338 1.187 1.034 1.363 Social Support 0.1441 yes reference 0.2127 reference no 1.102 0.946 1.283 1.139 0.957 1.356 Child's perceived safety score 0.3512 0.1826 zero reference reference 1 - 2 1.062 0.936 1.206 1.054 0.91 1.221 Children with SHCN without EBD Children with SHCN with EBD AOR 95% CI p value AOR 95% CI p value Child with perceived limitations 0.0023 0.3462 no reference reference yes 1.565 1.173 2.087 1.14 0.868 1.497 Mental health/stress score zero reference 0.0774 reference 0.0601 1 - 5 1.395 0.964 2.019 1.259 0.99 1.601 Trust in neighbors score 0.7311 zero reference reference 0.1381 1 - 4 1.064 0.747 1.517 1.214 0.94 1.568 Social Support 0.5542 0.3465 yes reference reference no 1.199 0.657 2.19 0.86 0.627 1.178 Child's perceived safety score 0.7709 0.963 zero reference reference 1 - 2 1.062 0.709 1.591 1.007 0.742 1.367 p values indicate the probability of the observed Wald chi square if there is no true difference in odds of sedentary life style by level of parental indicator. back to text 77 Table 17. Effect of special health care needs status on odds of sedentary lifestyle, adjusted for gender, age, race/ethnicity and respondent's education, stratified by parental attitude indicators Children without SHCN Children with SHCN without EBD Children with SHCN with EBD AOR 95% CI AOR 95% CI p value Overall reference 1.177 1.02 1.358 1.267 1.111 1.445 0.0004 Stratified by Mental health/stress zero reference 1.096 0.939 1.28 1.197 1.014 1.412 0.0682 1 - 5 reference 1.408 0.996 1.991 1.185 0.951 1.477 0.0812 Trust in neighbors score zero reference 1.164 0.994 1.363 1.251 1.076 1.454 0.0043 1 -4 reference 1.165 0.843 1.61 1.245 0.962 1.611 0.2047 Perceived safety score zero reference 1.135 0.973 1.323 1.274 1.102 1.474 0.0023 1 -2 reference 1.324 0.927 1.892 1.21 0.91 1.609 0.1665 Child with perceived limitations no reference 1.092 0.93 1.283 1.233 1.066 1.425 0.0138 yes reference 1.195 0.734 1.945 0.883 0.546 1.428 0.1791 Social support yes reference 1.151 0.999 1.327 1.33 1.154 1.533 0.0002 no reference 1.348 0.744 2.44 0.908 0.658 1.254 0.4701 p values indicate the probability of the observed Wald chi square if there is no true difference in the odds of sedentary lifestyle among SHCN categories back to text 78 Table 18. Joint effects of single attitudes with SCHN with or without EBD on the adjusted odds of sedentary lifestyle among 6- to 17-year-olds. United States 2007 Population Percent (weighted) Prevalence of Sedentary lifestyle AOR 95% CI p value Percent 95% CI No stress and no SHCN 74.506 31.036 29.903-32.17 reference Stress and SHCN without EBD 1.975 45.277 37.68-52.874 1.777 1.287-2.454 0.0005 Stress and SHCN with EBD 1.606 40.744 36.655-44.832 1.498 1.234-1.82 <.0001 All others 21.914 35.995 34.181-37.81 1.183 1.071-1.307 0.0009 No limitations and no SHCN 75.798 32.196 31.129-33.262 reference Limitations and SHCN without EBD 3.244 45.675 40.113-51.238 1.761 1.384-2.242 <.0001 Limitations and SHCN with EBD 1.687 37.854 32.853-42.854 1.383 1.087-1.758 0.0082 All others 4.057 35.848 33.681-38.014 1.165 1.039-1.306 0.009 No lack of trust and no SHCN 62.574 31.332 30.169-32.495 reference Lack of trust and SHCN without EBD 2.270 39.847 32.930-46.764 1.437 1.062-1.944 0.0189 Lack of trust and SHCN with EBD 3.095 40.474 35.398-45.55 1.55 1.22-1.97 0.0003 All others 32.061 35.921 34.187-37.655 1.213 1.102-1.336 <.0001 p value indicates the probability of the observed Wald chi square if there is no true difference between adjusted odds of sedentary lifestyle in this category and the reference category back to text 79 Table 19. Joint effects of combinations of parental attitudes on the odds of sedentary life style among 6-17 year olds. United States, 2007 Population Percent (weighted) Prevalence of Sedentary lifestyle AOR 95% CI p value Percent 95% CI No stress and no limitations 75.2 30.9 29.8-31.9 reference Stress and limitations 3.4 40.3 34.8-45.8 1.42 1.082-1.864 0.0114 All others 22.6 40.8 38.5-43.1 1.386 1.237-1.554 <.0001 No stress and no lack of trust 53.2 30.5 29.3-31.7 reference Stress and lack of trust 6.9 44.8 40.5-49.0 1.614 1.337-1.948 <.0001 All others 26.9 36.2 34.3-38.1 1.197 1.078-1.328 .0007 No limitations and no lack of trust 75.4 31.7 30.6-32.7 reference Limitations and lack of trust 2.0 43.9 37.3-50.5 1.557 1.163-2.086 0.0029 All others 22.6 37.7 35.6-39.8 1.228 1.100-1.371 .0003 No stress, no limitations and no lack of trust 11.9 32.8 30.1-35.5 reference Stress and limitations and lack of trust 1.2 40.0 30.6-49.4 1.322 0.842-2.076 0.2251 All others 87.0 33.2 32.3-34.3 .991 .863-1.137 .8954 p value indicates the probability of the observed Wald chi square if there is no true difference between adjusted odds of sedentary lifestyle in this category and the reference category back to text 80 Table 20. Joint effects of constellations of parental attitudes and special health care needs status on the odds of sedentary lifestyle among 6-17 year olds. United States, 2007 Population Percent (weighted) Prevalence of Sedentary lifestyle AOR 95% CI p value No stress, no limitations and no SHCN 62.3 30.8 29.6-31.9 reference Stress and limitations and no SHCN 0.6 35.7 19.5-51.9 0.857 .341-2.155 0.7425 Stress and limitations and SHCN without EBD 0.6 57.4 48.2-66.6 2.659 1.741-4.06 <.0001 Stress and limitations and SHCN with EBD 2.3 37.0 30.9-43.2 1.397 1.041-1.874 0.0258 All others 34.3 37.3 35.6-39-0 1.270 1.156-1.395 <.0001 No stress, no lack of trust and no SHCN 53.2 30.5 29.3-31.7 reference Stress and lack of trust and no SHCN 4.3 46.3 40.7-51.9 1.551 1.221-1.969 0.0003 Stress and lack of trust and SHCN without EBD 0.8 44.8 30.6-59.0 1.544 0.84-2.835 0.1616 Stress and lack of trust and SHCN with EBD 1.9 41.3 34.3-48.4 1.498 1.096-2.048 0.0112 All others 13.9 35.0 33.5-36.6 1.146 1.044-1.258 .0041 No limitations, no lack of trust and no SHCN 61.7 31.2 30.0-32.3 reference Limitations and lack of trust and no SHCN 0.3 49.8 35.4-64.3 1.619 0.925-2.833 0.0917 Limitations and lack of trust and SHCN without EBD 0.5 56.4 45.0-67.8 2.434 1.436-4.126 0.0009 Limitations and lack of trust and SHCN with EBD 1.2 37.1 28.2-46.1 1.307 0.86-1.984 0.2097 All others 36.3 36.3 34.6-37.9 1.207 1.100-1.325 <.0001 No stress, no lack of trust, no limitations and no SHCN 52.7 30.3 29.0-31.5 reference Stress and lack of trust and limitations and no SHCN 0.2 56.5 32.5-80.5 2.223 0.837-5.907 0.109 Stress and lack of trust and limitations and SHCN without EBD 0.2 53.1 36.7-69.4 2.11 1.033-4.309 0.0404 Stress and lack of trust and limitations and SHCN with EBD 0.8 34.3 22.9-45.7 1.277 0.713-2.287 0.4105 All others 46.1 36.6 35.1-38.1 1.141 1.136-1.166 <.0001 p value indicates the probability of the observed Wald chi square if there is no true difference between adjusted odds of sedentary lifestyle in this category and the reference category back to text 81 VIII. APPENDIX 2: ILLUSTRATIONS Obesity Screen Time Physical Activity Intrapersonal Factors Self concept, physical limitations, self efficacy outcome expectancies perceived competence Perception of Limitations Social Support Coping with Parenting Trust in neighbors Perceived Safety Parental Support Instrumental Informational Encouragement Special Health Care Needs Status Adapted from Wallander et al and Singh et al Figure 1. Conceptual framework showing relationships among attitudes, parental support, intrapersonal factors, behaviors and obesity. Special health care needs can affect the dynamics. (Combining ideas from models of Wallander9 and Singh65) back to text 82 Figure 2. Prevalence of special health care needs with and without emotional, behavioral and developmental conditions. Data from 2007 National Survey of Children's Health. back to text 83 Figure 3. Prevalence of specific chronic conditions. Data from 2007 National Survey of Children's Health. back to text 84 Figure 4. The prevalence of low MVPA, high screen time and sedentary lifestyle rise with age. Data from Table 3. back to text 85 Figure 5. Sedentary behaviors vary by race and ethnicity. High screen time and sedentary lifestyle are most prevalent among non-Hispanic black children. Low MVPA is most prevalent among Hispanic children and children of other races. Data from Table 3. *p<.05 back to text 86 Figure 6. Prevalence of low MVPA does not vary significantly by SHCN status. High screen time and sedentary lifestyle are more prevalent among children with special health care needs with emotional, behavior and developmental conditions. Data from Table 3. *p<.05 back to text 87 Figure 7. Relationships between parental attitudes and low MVPA. Model 1 adjusts for demographic factors (gender, race/ethnicity, education of respondent and child's age). Model 2 adjusts for demographic factors and special health care needs status. Model 3 adjusts for demographic factors, special health care needs status and the five attitudes. Only perceived limitations has a significant association after adjustment. The numbers shown have p values less than .05. See Table 6. back to text 88 Figure 8. Relationship between parental attitudes and high screen time. Model 1 adjusts for demographic factors (gender, race/ethnicity, education of respondent and child's age). Model 2 adjusts for demographic factors and special health care needs status. Model 3 adjusts for demographic factors, special health care needs status and the five attitudes. Only MH/stress and trust variables have significant associations after adjustment. The numbers shown have p values less than .05. (See Table 8.) back to text 89 Figure 9. Relationship between parental attitudes and sedentary lifestyle. Model 1 adjusts for demographic factors (gender, race/ethnicity, education of respondent and child's age). Model 2 adjusts for demographic factors and special health care needs status. Model 3 adjusts for demographic factors, special health care needs status and the five attitudes. Perceived limitations, MH/stress and trust variables have significant associations after adjustment. The numbers shown have p values < .05. (See Table 10) back to text 90 Figure 10. Parental perception of perceived limitations has a significant positive association with high screen time among children with SHCN without EBD but not among children without SHCN or with SHCN with EBD. (See Table 14.) *p<.05 back to text 91 Figure 11. SHCN with EBD has a significant positive association with high screen time among those without perceived limitations. Among those with limitations, the differences for children with SHCN with and without EBD are not statistically significant. (See Table 15.) *p<.05 92 Figure 12. When mental health/stress, perceived limitations or lack of trust is combined with SHCN with or without EBD, the odds of sedentary lifestyle is significantly greater than when neither the attitude nor the SHCN is present. (See Table 18.) *p < .05 back to text 93 Figure 13. Combinations mental health/stress, lack of trust and perceived limitations are associated with increased likelihood of sedentary lifestyle. (See Table 19.) *p < .05 back to text 94 Figure 14. Several constellations of attitudes and special health care needs are associated with increased likelihood of sedentary lifestyle. The three constellations with greatest odds ratios (red bars) are among children with SHCN without EBD. (See Table 20.) *p < .05 back to text 95 IX. BIBLIOGRAPHY 1. American Academy of Pediatrics, Committee on Nutrition. Policy Statement: Prevention of pediatric overweight and obesity. Pediatrics [Internet]. 2003 [Cited 2012 Apr 11];112(2):424-431. Available from http://pediatrics.aappublications.org/content/112/2/424.full 2. U.S. Department of Health and Human Services. 2008 Physical Activity Guidelines for Americans. ODPHP Publication No. U0036. Washington, DC 2008 3. Minihan PM, Must A, Anderson B, Popper B, Dworetsky B. Children with special health care needs: acknowledging the dilemma of difference in policy responses to obesity. Prev Chronic Dis [Internet]. 2011 [Cited 2012 Apr 11];8(5)A95. http://www.cdc.gov/pcd/issues/2011/sep/10_0285.htm 4. U.S. Department of Health and Human Services, Health Resources and Services Administration, maternal and Child Health Bureau. Children with Special Health Care Needs in Context: A Portrait of the States and the Nation 2007. US Department of Health and Human Services, Rockville, Maryland, 2011. 5. American Academy of Pediatrics, Task Force on the Family. Family Pediatrics: Report of the task force on the family. Pediatrics [Internet]. 2003 [Cited 2011 Dec 1];111(6):1541-1571 Available from http://pediatrics.aappublications.org/content/111/Supplement_2/1541.full.pdf+ht ml 6. Spurrier NJ, Sawyer MC, Staugas R, Martin AJ, Kennedy D, Streiner DL. Association between parental perception of children's vulnerability to illness and management of children's asthma. Pediatric Pulmonology [Internet]. 2000[Cited 2012 Apr 11];29:88-93 Available from http://onlinelibrary.wiley.com/doi/10.1002/%28SICI%291099- 0496%28200002%2929:2%3C88::AID-PPUL2%3E3.0.CO;2- D/abstract;jsessionid=51861D84009CAF5C8D9BC9BB04531F41.d01t01 7. Thomasgard M, Shonkoff JP, Metz WP, Edelbrock C. Parent-child relationship disorders. Part II. The vulnerable child syndrome and its relation to parental overprotection. Developmental and Behavioral Pediatrics [Internet]. 1995 [Cited 2012 Apr 11];16 (4):251-256. Available from http://www.ncbi.nlm.nih.gov/pubmed/7593660 8. Anthony KK, Gil KM, Schanberg LE. Brief Report: Parental perceptions of child vulnerability in children with chronic illness. Journal of Pediatric Psychology [Internet]. 2001[Cited 2012 Apr 11]; 28(3):185-190. DOI: 10.1093/jpepsy/jsg005 Available from http://jpepsy.oxfordjournals.org/content/28/3/185.full.pdf 96 9. Wallander JL, Varni JW. Effects of pediatric chronic physical disorders on child and family adjustment. Journal of Child Psychology and Psychiatry [Internet]. 1998 [Cited 2012 Apr 11]; 39(1);29-46. http://www.ncbi.nlm.nih.gov/pubmed/9534085 10. Blackman JA, Gurka, MJ. Developmental and behavioral comorbidities of asthma in children. Journal of Developmental and Behavioral Pediatrics. 2007 [Cited 2012 Apr 11];28(2):92-99. DOI:10.1097/01.DBP.0000267557.80834.e5 11. Blackman JA, Gurka MJ, Gurka KK, Oliver MN. Emotional, developmental and behavioural co-morbidities of children with chronic health conditions [Internet]. Journal of Paediatrics and Child Health. 2011 [Cited 2012 Apr 11];47:742-747. DOI:10.1111/j.1440-1754.2011.02044.x 12. U.S. Department of Health and Human Services. Physical Activity. Healthy People 2020 Objectives. PA 3.1 Aerobic Physical Activity. [Internet] [Cited 2012 Apr 11]. Available from http://www.healthypeople.gov/2020/topicsobjectives2020/objectiveslist.aspx?topi cId=33 13. American Academy of Pediatrics, Committee on Public Education. Children, adolescents and television. Pediatrics [Internet].2001 [Cited 2012 Apr 11]. 107(2):423-427 Available from http://pediatrics.aappublications.org/content/107/2/423.full.pdf+html 14. American Academy of Pediatrics, Council on Communication and Media. Policy Statement--Media Violence. Pediatrics [Internet]. 2009 [Cited 2012 Apr 11] ;124(5):1495-1503. doi:10.1542/peds.2009-2146 Available from http://pediatrics.aappublications.org/content/124/5/1495.full.pdf+html 15. Rideout VJ, Foehr UG, Roberts DF. Generation M2: Media in the Lives of 8- to 18-year olds. Kaiser Family Foundation. 2010. [Cited 2012 Apr 11] Available from http://www.kff.org/entmedia/upload/8010.pdf 16. Blumberg SJ, Foster EB, Frasier AM, Satorius J, Skalland BJ, Nysse-Carris KL, Morrison HM, Chowdhury SR, O'Connor KS, Design and Operation of the National Survey of Children's Health, 2007. Vital Health and Statistics, Series 1: Program and Collection Procedures. National Center for Health Statistics, Hyattsville , 2009. [Cited 2012 Apr 11] Available from ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/slaits/nsch07/2_Methodology_Repo rt/NSCH_Design_and_Operations_052109.pdf 17. Centers for Disease Control and Prevention. School health guidelines to promote healthy eating and physical activity. MMWR.2011;60(5):1-75. 18. Laurson, K.R., Eisenmann, J.C., Welk, G.J., Wickel, E.E., Gentile, D. A. & Walsh, D. A. Combined influence of physical activity and screen time 97 recommendations on childhood overweight. Journal of Pediatrics [Internet]. 2008 [Cited 2012 Apr 11]; 153 209-215. doi: 10.1016/j.jpeds.2008.02.042 Available from http://www.ncbi.nlm.nih.gov/pubmed/18534231 19. Craigie AM, Lake AA, Kelly SA, Adamson AJ, Mathers JC. Tracking of obesity- related behaviours from childhood to adulthood: A systematic review. Maturitas [Internet]. 2011[Cited 2012 Apr 11]; 70:266-284. Available from http://www.maturitas.org/article/S0378-5122%2811%2900296-9/abstract 20. Sisson SB, Broyles ST, Baker BL, Katzmarzyk PT. Screen time, physical activity and overweight in U.S. youth: National Survey of Children's Health 2003. Journal of Adolescent Health [Internet]. 2010 [Cited 2012 Apr 26] :47:309-311 21. Sallis JF, Prochaska JJ, Taylor WC. A review of correlates of physical activity of children and adolescents. Medicine and Science in Sports and Exercise [Internet]. 2000:23(5):963-975. Available from http://www.ncbi.nlm.nih.gov/pubmed/10795788 22. Anderson SE, Economos CD, Must A. Active play and screen time in US children aged 4 to 11 years in relation to sociodemographic and weight status characteristics: a nationally representative cross-sectional analysis. BMC Public Health [Internet]. 2008 [Cited 2012 Apr 12];8:366 DOI: 10.1186/1471-2458-8- 366 Available from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2605460/pdf/1471-2458-8- 366.pdf 23. Maniccia DM, Davison KK, Marshall SJ, Manganello JA, Dennison BA. A meta- analysis of interventions that target children's screen time for reduction. Pediatics [Internet]. 2011 [Cited 2012 Apr 11];128:e193. DOI: 10.1542/peds.2010-2353 Available from http://pediatrics.aappublications.org/content/128/1/e193.full 24. Page AS, Cooper AR, Griew P, Jago R. Children's screen viewing is related to psychological difficulties irrespective of physical activity. Pediatrics [Internet]. 2010 [Cited 2012 Apr 11];126:e1011. DOI: 10.1542/peds.2010.1154 Available from http://pediatrics.aappublications.org/content/126/5/e1011.full 25. Strasburger, VC, Jordan AB, Donnerstein E. Health Effects of media on children and adolescents. Pediatrics [Internet]. 2010 [Cited 2012 Apr 11]; 125:756-767. DOI: 10.1542/peds.2009-2563 Available from http://pediatrics.aappublications.org/content/125/4/756.full 26. Pearson N, Biddle SJH. Sedentary behavior and dietary intake in children, adolescents, and adults: A systematic review. American Journal of Preventive Medicine [Internet]. 2011 [Cited 2012 Apr 11];41(2):178-188. DOI:10.1016/jamepre.2011.05.002 http://www.ajpmonline.org/article/S0749- 3797%2811%2900299-6/abstract 98 27. Boone JE, Gordon-Larsen P, Adair LS, Popkin BM. Screen time and physical activity during adolescence: longitudinal effect on obesity in young adulthood. International Journal of Behavioral Nutrition and Physical Activity [Internet]. 2007 [Cited 2012 Apr 11]; 4:26-35. DOI:10.1186/1479-5868-4-26. Available from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1906831/ 28. Mark AE, Janssen I. Relationship between screen time and metabolic syndrome in adolescents. Journal of Public Health [Internet]. 2008 [Cited 2012 Apr 11];30(2):153-160. DOI:10.1093/pubmed/fdn022. Available from http://jpubhealth.oxfordjournals.org/content/30/2/153.full 29. Vandewater EA, Shim M, Caplovitz AG. Linking obesity and activity level with children's television and video game use. Journal of Adolescence [Internet]. 2004 [Cited 2012 Apr 11]; 27:71-85. DOI: 10.1016/j.adolescence.2003.10.003 http://www.sciencedirect.com/science/article/pii/S0140197103000903 30. Epstein LH, Paluch RA, Roemmich JN, Beecher MD. Family-based obesity treatment, then and now: Twenty-five years of pediatric obesity treatment. Health Psychology [Internet]. 2007 [Cited 2012 Apr 11];26(4):381-391. DOI: 10.1037/0278-6133.26.4.381 Available from http://psycnet.apa.org/journals/hea/26/4/381/ 31. Golan M, Crow S. Parents are key players in the prevention and treatment of weight-related problems. Nutrition Reviews [Internet]. 2004[Cited 2012 Apr 14];62(1):39-50. DOI:10.1301/nr.2004.jan.39-50 Available from http://www.ncbi.nlm.nih.gov/pubmed/14995056 32. Golan M, Weizman A. Familial approach to the treatment of childhood obesity: Conceptual model. Journal of Nutritional Education [Internet].2001 [Cited 2011 Dec 2];33:102-107 Available from http://www.sciencedirect.com/science/journal/00223182/33/2 33. Bergman AB, Stamm SJ. The morbidity of cardiac nondisease in school children. New England Journal of Medicine [Print]. 1967;276:1008-1013 34. Beets MW, Cardinal BJ, Alderman BL. Parental social support and the physical- activity-related behaviors of youth: A review. Health Education and Behavior [Internet]. 2010 [Cited 2012 Apr 14];37(5)621-644. Available from http://heb.sagepub.com/content/37/5/621.short 35. Brockman R, Jago R, Fox KR. Children's active play: self-reported motivators, barriers and facilitators. BMC Public Health [Internet]. 2011 [Cited 2012 Apr 14];11:461. Available from http://www.biomedcentral.com/1471-2458/11/461 36. Craggs C, Corder K, van Sluijs WMF, Griffin SJ. Determinants of change in physical activity in children and adolescents. Am J Prev Med [Internet]. 99 2011[Cited 2012 4 Apr 12];40(6):645-658. DOI: 10:1016/j.amepre.2011.02.025 Available from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3100507/ 37. Sallis JF, Prochaska JJ, Taylor WC, Hill JO, Geraci JC. Correlates of physical activity in a national sample of girls and boys in grades 4 through 12. Health Psychology[Internet]. 1999 [Cited 2010 Dec 2];18(4):410-415. 38. Heitzler, C.D., Martin, S.L., Duke, J., & Huhman, M. Correlates of physical activity in a national sample of children aged 9-13 years. Preventive Medicine [Internet]. 2006 [Cited 2010 Dec 2]; 42(4) 254-260. doi:10.1016/j.ypmed.2006.01.010 Available from http://www.sciencedirect.com/science/article/pii/S0091743506000119 39. Trost SG, Loprinzi PD. Parental influences on physical activity behavior in children and adolescents: A brief review. American Journal of Lifestyle Medicine[Internet].2011 [Cited 2012 Apr 14];5(2):171-181. DOI: 10.1177/1559827610387236 Available from http://ajl.sagepub.com/content/5/2/171.short 40. Welk GJ, Wood K, Morss G. Parental influences on physical activity in children: An exploration of potential mechanisms. Pediatric Exercise Science [Internet].2003 [Cited 2012 Apr 14]:15:19-33 Available from http://ajl.sagepub.com/content/5/2/171.short 41. HeitzlerCD, Lytle LA, Erikson D, Barr-Anderson D, Sirard, JR, Story M. Evaluating a model of youth physical activity. American Journal of Health Behavior [Internet] 2011 [Cited 2010 Dec 10]; 34(5) 593-606. Available from http://ukpmc.ac.uk/articles/PMC3086379/reload=0;jsessionid=FneJ1yH1CdUOJP eAuOfm.24 42. Fong SSM, Lee VYL., Chan NNC, Chan RSH, Chak WK, Pang MYC. Motor ability and weight status are determinants of ouut-of-school activity participation for children with developmental coordination disorder. Research in Developmental Disabilities [Internet] 2011 [Cited 2012 Apr 14];32:2612-2623. DOI: 10.1016/j.ridd.2011.06.013 Available from http://www.sciencedirect.com/science/article/pii/S0891422211002496 43. Pianosi PT, Davis HS. Determinants of physical fitness in children with asthma. Pediatrics [Internet]. 2004 [Cited 2012 Apr 14];113(3) e225-e229. Available from http://www.pediatricsdigest.mobi/content/113/3/e225.full 44. Stevens GD, Pickering TA, Laqui SA. Relationship of medical home quality with school engagement and after-school participation among children with asthma. Journal of Asthma [Internet]. 2010 [Cited 2012 Apr 14];47:1001-1010. DOI: 10.3109/02770903.2010.514636 Available from http://www.ncbi.nlm.nih.gov/pubmed/20831470 100 45. Waring ME, Lapane KL. Overweight in children and adolescents in relation to attention deficit/hyperactivity disorder: Results from a national sample. Pediatrics[Internet]. 2008 [Cited 2012 Apr 14];122(1):e1-e6. DOI:10.1542/peds.2007-1955 Available from http://www.pediatricsdigest.mobi/content/122/1/e1.full 46. Sisson SB, Broyles ST, Newton RL, Baker BL, Chernausek SD.TVs in the bedrooms of children: does it impact health and behavior? Preventive Medicine [Internet] . 2011 [Cited 2012 Apr 14]; 52(2):104-108. Available from http://www.sciencedirect.com/science/article/pii/S0091743510004743 47. U.S. Department of Health and Human Services. The Health and Well-being of Children: A portrait of states and the nation 2007. National Survey of Children's Health 2007. U.S. Department of Health and Human Services, Rockville, 2009. 48. VanCleave J, Gortmaker SL, Perrin JM. Dynamics of obesity and chronic health conditions among children and youth. JAMA [Internet]. 2010 [Cited 2012 Apr 14]:101(7)623-630.Available from http://jama.ama- assn.org/content/303/7/623.full 49. Bethell CD, Read D, Neff J, Blumberg SJ, Stein REK, Sharp V, Newacheck PW. Comparison of the Children with Special Health Care Needs Screener to the Questionnaire for Identifying Children with Chronic Conditions--Revised. Ambulatory Pediatrics [Internet]. 2002 [Cited 2012 Apr 14] ;2(1):49-57. Available from http://www.ambulatorypediatrics.org/article/S1530- 1567%2805%2960082-2/abstract 50. Bethell CD, Read D, Stein REK, Blumberg SJ, Wells N, Newacheck PW. Identifying children with special health care needs: Development and evaluation of a short screening instrument. Ambulatory Pediatrics [Internet]. 2002 [Cited 2012 Apr 14]; 2(1)38-48. Available from http://www.sciencedirect.com/science/article/pii/S1530156705600810 51. Newacheck PW, Kim SE, Blumberg SJ, Rising JP. Who is at risk for special health care needs: Findings from the National Survey of Children's Health. Pediatrics [Internet]. 2008 [Cited 2012 Apr 14];122(2):347-359. DOI: 10.1542/peds.2007-1406 Available from http://pediatrics.aappublications.org/content/122/2/347.full 52. Green M, Solnit AJ. Reactions to the threatened loss of child: A vulnerable child syndrome. Pediatrics. 1964 [Cited 2012 Apr 26];34:58-66 Available from http://pediatrics.aappublications.org/content/34/1/58.short 53. Perrin EC, West PD, Culley BS. Is my child normal yet? Correlates of vulnerability. Pediatrics [Internet].1989 [Cited 2012 Apr 14]; 83(3):355-363. http://pediatrics.aappublications.org/content/83/3/355.short 101 54. Bandura, A. Health promotion by social cognitive means. Health Education and Behavior [Internet]. 2004 [Cited 2012 Apr 14];31(2): 143-164 doi: 10.1177/1090198104263660 Available from http://heb.sagepub.com/content/31/2/143.short 55. Larson K, Russ SA, Crall JJ, Halfon N. Influence of multiple social risks on children's health. Pediatrics [Internet]. 2008 [Cited 2012 Apr 14]; 121(2):337-344. DOI:10.1542/peds.2007-0447 Available from http://www.pediatricsdigest.mobi/content/121/2/337.full 56. Singh GK, Kogan MD, Siahpush M, van Dyck PC. Independent and joint effects of socioeconomic, behavioral and neighborhood characteristics on physical inactivity and activity levels among US children and adolescents. J Community Health [Internet]. 2008 [Cited 2012 Apr 14];33:201-216. DOI: 10.1007/s10900- 008- 9094-8 Available from http://www.springerlink.com/content/v86787428v362118/ 57. Danielson M. Relationship between perceived neighborhood safety and physical activity among youth with a mental/emotional/behavioral condition. Paper presented at: American Public Health Association Annual Meeting; November 1, 2011; Washington, D.C. 58. Pate RR, Freedson PS, Sallis JF, Taylor WD, Sirard J, Trost SG, Dowda M. Compliance with physical activity guidelines: Prevalence in a population of children and youth. Annals of Epidemiology [Internet] 2002 [Cited 2012 Apr 14];12:303-308. Available from http://www.sciencedirect.com/science/article/pii/S1047279701002630 59. Rice MN, Howell CC. Measurement of physical activity, exercise and physical fitness in children: Issues and concerns. Journal of Pediatric Nursing [Internet]. 2000 [Cited 2012 Apr 14]; 15(3):148-155. Available from http://www.sciencedirect.com/science/article/pii/S0882596300800032 60. Telford A, Salmon J, Jolley D, Crawford D. Reliability and validity of physical activity questionnaires for children: The Children's Leisure Activities Study Survey (CLASS). Pediatric Exercise Science [Internet]. 2004 [Cited 2012 Apr 14]:16:64-78. http://www.acaorn.org.au/streams/activity/CLASS_validity_leisure_activities.pdf 61. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Medicine and science in sports and exercise [Internet].2008 [Cited 2012 Apr 14];40(1):181-188. Available from http://journals.lww.com/acsm- msse/Abstract/2008/01000/Physical_Activity_in_the_United_States_Measured_b y.25.aspx 102 62. Murphy JK, Alpert BS, Christman JV, Willey ES. Physical fitness in children: A survey method based on parental report. American Journal of Public Health [Internet]. 1988 [Cited 2012 Apr 14];78(6)708-710. Available from http://ajph.aphapublications.org/doi/pdf/10.2105/AJPH.78.6.708 63. National Center for Health Statistics. National Survey of Children's Health: CATI Instrument. State and Local Area Integrated Telephone Survey. Centers for Disease Control and Prevention, Hyattsville 2007. [Cited 2012 Apr 11]. Available from ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/slaits/nsch07/1a_Survey_Instrumen t_English/NSCH_Questionnaire_052109.pdf 64. Singh GK, Kogan MD, vanDyck PC. A multilevel analysis of state and regional disparities in childhood and adolescent obesity in the United States. J. Community Health [Internet]. 2008 [Cited 2012 Apr 14];33:90-102. DOI: 10.1007/s10900-007-9071-7 Available from http://www.springerlink.com/content/cj226034u56074l7/ 65. Singh GK, Kogan MD, VanDyck PC, Siahpush M. Racial/ethnic, socioeconomic, and behavioral determinants of childhood and adolescent obesity in the United States: Analyzing independent and joint associations. Ann Epidemiol [Internet]. 2008 [Cited 2012 Apr 14];18:682-695. DOI: 10.1016/j.annepidem.2008.05.001 Available from http://www.sciencedirect.com/science/article/pii/S1047279708001129 66. Singh GK, Yu SM, Siahpush M, Kogan MD. High levels of physical inactivity and sedentary behaviors among US immigrant children and adolescents. Arch Pediatr Adolesc Med [Internet].2008 [Cited 2012 Apr 14]; 162(8)756-763. Available from http://archpedi.ama-assn.org/cgi/content/abstract/162/8/756 67. Rowlands AV, Eston RG. The measurement and interpretation of children's physical activity. Journal of Sports Science and Medicine [Internet]. 2007[Cited 2012 Apr 14];6:270-276 Available from http://www.jssm.org/vol6/n3/1/v6n3- 1pdf.pdf 68. Must A, Tybor DJ. Physical activity and sedentary behavior: a review of longitudinal studies of weight and adiposity in youth. International Journal of Obesity [Internet]. 2005 [Cited 2012 Apr 14];29:s84-s96. DOI: 10.1038/sj.ijo.0803064 Available from http://www.nature.com/ijo/journal/v29/n2s/full/0803064a.html 69. Van der Lee JH, Mokkink LB, Grootenhuis MA, Heymans HS, Offringa M. Definitions and measurement of chronic health conditions in childhood: A systematic review. Journal of the American Medical Association [Internet]. 2007 103 [Cited 2012 Apr 14];297:2741-2751 Available from http://jama.ama- assn.org/content/297/24/2741.short 70. National Center for Health Statistics. Screen time. Health Indicators Warehouse. [Cited 2012 Apr 11] Available from http://www.healthindicators.gov/Indicators/Screentime-Children6- 14years_1328/Profile/Data 71. Dunton GF, LiaoY, Intille SS, Spruijt-Metz D, Pentz M. Investigating children?s physical activity and sedentary behavior using ecological momentary assessment with mobile phone. Obesity [Internet]. 2011 [Cited 2012 Apr 14]: 19(6):1205-12. DOI:10.1038/oby.2010.302 Available from http://www.nature.com/oby/journal/v19/n6/abs/oby2010302a.html 72. Dubois L, Girad M. Accuracy of maternal reports of pre-schoolers' weights and heights as estimates of BMI values. International Journal of Epidemiology [Internet]. 2007 [Cited 2012 Apr 14];36:132-138. DOI: 10.1093/ije/dy1281 Available from http://ije.oxfordjournals.org/content/36/1/132.short 73. Thompson RJ, Gustafson KE. Adaptation to chronic childhood illness. American Psychological Association. Washington, DC, 1996. 74. Braveman P, Barclay C. Health disparities beginning in childhood: A life-course perspective. Pediatrics [Internet]. 2009 [Cited 2012 Apr 14]; 124:s163-s175. Available from http://www.pediatricsdigest.mobi/content/124/Supplement_3/S163.full 75. Juonala M, Magnussen CG, Berenson GS, Venn A, Burns TL, Sabin MA, Srinivasan SR, Daniels SR, Davis PH, Chen W, Sun C, CheungM, Viikari JSA, Dwyer T, Raitakari OT. Childhood adiposity, adult adiposity and cardiovascular risk factors. NEJM [Internet]. 2011 [Cited 2012 Apr 14];365:1876-1855. Available from http://www.nejm.org/doi/full/10.1056/NEJMoa1010112 76. Welk, G. J. & Meredith, M.D. (Eds.). (2008). Fitnessgram / Activitygram Reference Guide [Internet]. Dallas, TX: The Cooper Institute. [Cited 2012 Apr 11] Available from http://www.cooperinstitute.org/reference-guide