Epidemiological Profile of Health and Behaviors in Middle Childhood

We analyzed data from the 2016 and 2017 NSCH. The NSCH is a cross-sectional, address-based household survey of children aged 0 to 17 years that is directed and funded by the Health Resources and Services Administration’s Maternal and Child Health Bureau and conducted by the US Census Bureau. Beginning in 2016, the NSCH has been fielded annually and administered via a Web- or paper-based survey to respondents who are parents or caregivers knowledgeable of the sample child’s health and health care. The overall weighted response rates were 40.7% and 37.4% in 2016 and 2017, respectively, whereas the interview completion rates (the proportion of households known to include children who completed the topical questionnaire) were 69.7% and 70.9%, respectively. Specific information regarding the NSCH methodology and design may be found elsewhere.¹¹ 

The NSCH allows for data from multiple years to be appended to obtain multiyear estimates. The 2016 and 2017 NSCH contained information from completed surveys of 50 212 and 21 599 children, respectively. The data sets include survey weights to ensure the generalizability of estimates to the noninstitutionalized population of US children. To obtain population-based estimates and account for nonresponse, for each survey year, a complex weighting scheme is employed that includes nonresponse adjustments and iterative raking to population controls derived from same-year American Community Survey data.^12,13  The current study was exempt from institutional review board review because we used publicly available data sets.

The NSCH consists of a screener questionnaire (to determine eligible households with children at sampled addresses) and age-specific topical questionnaires for children aged 0 to 5, 6 to 11, and 12 to 17 years. The screener identifies demographic information for age-eligible children in a household and information used to identify and categorize children with special health care needs (CSHCN).^14,15  The more detailed topical questionnaire collects information specific to the randomly-selected sample child, including questions on age-appropriate developmental milestones and behaviors. Although many of the variables included in this analysis originated from the topical questionnaire administered for children aged 6 to 11 years, there were also items that appeared on all 3 sets of topical questionnaires. Survey items selected to create this epidemiological profile cover 3 categories. Sociodemographic, family, and neighborhood characteristics include sex, race and/or ethnicity, indicators of household socioeconomic status, family structure, and neighborhood conditions. Family poverty level, as reported in the NSCH Public Use File, is derived from the US Census Bureau’s poverty thresholds and is capped at 400% to protect the confidentiality of respondents.¹⁶  The NSCH contains several questions about the presence of neighborhood characteristics, grouped here as “amenities” (ie, sidewalks or walking paths, parks or playground, recreational or community centers, libraries or bookmobiles) or “detractions” (ie, litter and/or garbage on streets and sidewalks, poorly-maintained or rundown housing, vandalism). Health status indicators include general health status, oral health status, presence of chronic or genetic health conditions, and emotional and behavioral disorders. Parents were also asked about their child having been diagnosed with several chronic, genetic, and emotional and behavioral conditions through the following questions: “Has a doctor or other health care provider ever told you that [your child] has…[condition]?” and “If yes, does [your child] currently have the condition?” The third category covers several behaviors, including school delinquency, physical activity, screen time, and sleep habits. Variables of interest (Supplemental Table 5) were primarily selected: (1) variables associated with known differences in other age groups (eg, emotional and behavioral disorders) and (2) variables that are known risk factors for certain health issues among children (eg, activity engagement, sleep behaviors).

Survey weights were combined from the 2016 and 2017 NSCH data sets and adjusted before study analyses.¹⁶  Weighted descriptive analyses were conducted to determine the composition of the middle-childhood population in the United States and the distribution of the aforementioned characteristics among the population. This age group was defined as 6 to 11 years, which is consistent with the Current Population Reports published by the US Census Bureau.¹⁷  We also conducted bivariate analyses, applying χ² tests of independence, to examine differences among children by age subgroup (ie, 6–7, 8–9, and 10–11 years) for select variables. Several demographic variables were imputed for the purpose of raking during weighting procedures, including sex, race and/or ethnicity, and household poverty. Child demographic variables were imputed by using hot-deck imputation, and multiple imputation was applied to the poverty level. Stata version 15 (Stata Corp, College Station, TX) was used to analyze data for this study; statistical significance was determined as P < .05.

Our study sample from the 2016 and 2017 NSCH included parent or caregiver responses about 21 539 children aged 6 to 11 years, yielding a weighted estimate of 24.8 million children in the United States. Sociodemographic characteristics are shown in Table 1. Most of these children lived in 2-parent (married or unmarried) households (75.9%) in which the highest household education was more than high school (70%). A majority of children were born in the United States (96%) and lived in an English-speaking household (85%). Approximately 70% of children came from families with household incomes <400% of the federal poverty level, and 41% of households received free or reduced-cost meals in schools. In addition, ∼37% of children were covered by some form of public insurance (eg, Medicaid or State Children’s Health Insurance Program), and 6% were reported to have no health insurance.

Most children lived in neighborhoods with ≥2 amenities (78%) or 0 detractions (75%), yet >10% of parents reported that their neighborhoods had 0 amenities and ≥2 detractions. Many parents reported that they felt their children were not safe in their neighborhoods (36.5%) and in school (19.5%).

As shown in Table 2, a majority of children aged 6 to 11 years were reported to be in excellent or very good physical health (89%) and oral health (73%). An estimated 21% met the criteria for having special health care needs. Analyses of those with current conditions revealed that allergies were the most prevalent physical condition among children (21%), followed by asthma (9.5%) and frequent headaches or migraines (2.6%). When examining emotional, behavioral, and developmental disorders, the most prevalent conditions were attention-deficit disorder (ADD) or attention-deficit/hyperactivity disorder (ADHD) and behavioral or conduct problems (reported among 9% and 8% of children, respectively), followed by anxiety (6.5%). Approximately 3% of children in this age group had a current diagnosis of autism spectrum disorder (ASD). Additionally, nearly 8% had a current learning disability, 6% had a current speech or language disorder, and 6% had a developmental delay. Table 3 reveals significant differences in the prevalence of emotional and behavioral conditions across middle-childhood age subgroups. For example, the prevalence of ADD or ADHD increased from 5.4% among children aged 6 to 7 years to 11.7% among children aged 10 to 11 years (P < .0001). Older age was also associated with an increased prevalence of diagnosed anxiety (P < .0001). We found significant variation for several conditions (ie, behavioral or conduct problems, developmental delay, and speech or language disorder); specifically, prevalence peaked at 8 to 9 years old and decreased among those aged 10 to 11 years.

Behaviors and reported activities among children aged 6 to 11 years are shown in Table 4. More than 22% of parents reported that their child had been bullied, picked on, or excluded by other children, and 7% of children exhibited bullying behavior to others. In the NSCH, parents were also asked a set of questions on schooling and activities: “Compared to other children his or her age, how much difficulty does this child have making or keeping friends?” Nearly 20% of parents reported that their child had difficulty making or keeping friends compared with other children their own age. Approximately 22% of children had missed ≥4 days of school because of illness or injury during the past year, and most parents (88%) reported never or only once having been contacted regarding a problem with their child at school during the past year. More than 95% of children reportedly met indicators associated with school performance (eg, showing interest or curiosity in learning new things, caring about doing well in school).

Most children aged 6 to 11 years were active: 58% participated in sports during the past 12 months (ie, on a sports team or in a sports lesson), and 56% participated in exercise, a sport, or another physical activity for at least 60 minutes most days of the week. Regarding screen time, nearly one-half of all children reportedly watched >2 hours of television per weekday, whereas nearly 40% spent >2 hours per weekday using devices for purposes other than schoolwork. Overall, parents reported that more than one-third of children had ≥4 hours per weekday of total screen time (ie, watching television and/or using devices). In the NSCH, parents were also asked several questions about children’s bedtime and sleep habits. Nearly 60% of children usually or sometimes went to bed at the same time every night, and two-thirds (66%) reportedly received an average of ≥9 hours per day of sleep.

Table 3 shows how screen time and sleep behaviors varied by children’s age and sex. The percentage of children spending ≥4 hours per weekday using devices for purposes other than schoolwork significantly increased from 4.8% of those aged 6 to 7 years to 8.2% of those aged 10 to 11 years (P < .0001). A similarly significant increase was seen regarding total average screen time of ≥4 hours per weekday (from 28.0% of younger children to 40% of older children; P < .0001). Conversely, there was a significant decrease in the percentage of children receiving the recommended minimum of 9 hours of sleep per day; 71.8% of those aged 6 to 7 years reported ≥9 hours of sleep per day, yet this proportion dropped to 58.7% of older children (P < .0001). Patterns in screen time and physical activity were also notably different for boys compared with girls at different ages, with a 15-percentage-point increase in the proportion of girls averaging ≥4 hours of total screen time per weekday between ages 6 to 7 and 10 to 11 years (compared with a 9-percentage-point increase among their male counterparts) and a 10-percentage-point decline in the proportion of girls meeting physical activity guidelines between these ages (compared with a 5-percentage-point decline for boys of the same age).

With this epidemiological profile of middle childhood, we are the first to examine this population using a comprehensive range of indicators from a large nationally representative data set. Our analysis included a range of factors and confirmed that the population of children aged 6 to 11 years in the United States is generally healthy according to most traditional indicators of health and well-being. Most were in good health, despite the fact that 4 in 10 experienced ≥1 chronic conditions, and had access to health-affirming family- and community-level factors.

Several notable patterns were observed with respect to age. We found a significant increase in the prevalence of diagnosed ADHD with age, which is consistent with findings reported by Visser at al,¹⁸  who found the average age at ADD or ADHD diagnosis among children to be ∼7 years. We also found significant differences in the prevalence of developmental delays; specifically, a peak prevalence at 8 to 9 years of age, followed by a drop before early adolescence. This could be due to the receipt of routine screening among school-aged children and early intervention services, which have been reported to occur primarily for children ages 9 to 12 years.¹⁹ 

Results also indicated a significant increase in the prevalence of diagnosed anxiety problems as children aged (from 4.2% of children aged 6 to 7 years to 7.7% of children aged 10 to 11 years) as well as higher diagnoses of anxiety among boys than girls. Earlier onset of anxiety, compared with other mental health conditions, has been documented with increasing prevalence across adolescence.^20,21  Our results are among the few to reveal similar findings in the preceding age group of middle childhood and could reflect an increasing prevalence of early-onset mental health conditions or enhanced screening or symptom recognition among providers or caregivers. Our findings on sex differences in anxiety diagnoses (ie, greater prevalence of anxiety among boys) are surprising given that research has shown girls to be at greater risk for exhibiting anxiety symptoms during early adolescence.^22–24  These results suggest that additional risk factors beyond sex may impact anxiety prevalence among children and may inform several opportunities in a child’s life course for targeted screening and intervention.

Results regarding sleep, screen time, and physical activity behaviors are concerning and merit further consideration. Although most children aged 6 to 11 years are reported to get the recommended 9 to 12 hours of sleep per day, the proportion of children obtaining this duration of sleep significantly decreases as children age. Sleep insufficiency in children has been shown to be associated with behavior and learning problems, emotional disturbances, and increased risk for several physical and emotional and behavioral conditions developing into adolescence.^25,26  We observed similar patterns regarding parent-reported screen time and electronic device use for children. More than one-third of children had ≥4 hours per weekday of total screen time using devices for non-schoolwork purposes; both screen time and time using electronic devices significantly increased with increasing age. These findings align with past work that has shown increasing use of devices and digital media among children at younger ages.^4,27,28  Although there are no longer firm guidelines for screen time in middle childhood, parents are advised to limit use of digital media for entertainment purposes and prioritize that children get recommended daily amounts of sleep and physical activity.²⁹  This guidance extends into adolescence, supported by studies that have found associations between increased screen time at younger ages and insufficient sleep and sleep disturbances in adolescence.^30,31  Finally, although more than half of children were reported to have participated in 60 minutes of physical activity most days of the week, <30% did so every day, as recommended in the Physical Activity Guidelines for Americans.³²  More importantly, the proportion of children doing so every day significantly decreases with age and varies by sex, with significant declines observed among girls. The interplay between declining rates of physical activity and increasing rates of screen time among girls highlights the importance of tracking such indicators for younger children and the opportunity to intervene at earlier ages.

This study has limitations. First, the cross-sectional data only allow us to draw associations, rather than causal relationships, between variables. Second, response bias may have been introduced because the NSCH contained parent-reported data. Although research has revealed strong agreement between parental report and clinical records for diagnoses made in the past year, there is little research examining the validity of parent-reported information for behavioral outcomes, including sleep and activity. Despite this dearth, our methodology and reliance on parental report is consistent with most existing studies in which similar outcomes in children have been examined.^3,27  In addition, the weighted response rate of ∼40% for both survey years and an interviewed sample that appears to live in areas associated with higher levels of income and education suggest the possibility of nonresponse bias. However, extensive nonresponse bias analyses conducted at several different phases of the survey indicate that there was no strong or consistent evidence of nonresponse bias.^12,13 

This study lends credence to the perception that children in middle childhood are generally healthy and thriving but highlights the fragility of this picture by sounding an early warning around the emergence and progression of selected health-risk behaviors. Findings establish a benchmark from which we can examine how health status and behaviors among these children evolve over time (through longitudinal or trend analyses). Further research could add an important layer to existing work on adolescent health and health behaviors; for example, focusing on the association of risk behaviors in middle childhood with the likelihood of certain behaviors and health outcomes in adolescence. This may inform targeted interventions for middle-childhood subgroups, with implications for optimal screening and service delivery for children at younger ages. With this study, we also lay the groundwork for future researchers examining differences by sociodemographic characteristics beyond age and sex (eg, differences by family and neighborhood factors), and by health care access and use. Such efforts should prioritize the interplay between and among these factors to elucidate which factors (alone and in combination) may serve as significant risk and protective forces within the middle-childhood population.

ADD

attention-deficit disorder

ADHD

attention-deficit/hyperactivity disorder

ASD

autism spectrum disorder

CSHCN

children with special health care needs

NSCH

National Survey of Children’s Health