The National Survey of Children's Health is a cross-sectional random digit dial telephone survey of parents and guardians (2003, n = 102,353; 2007, n = 91,642; and 2011–2012, n = 95,677).¹⁶ The 2003 and 2007 surveys used landlines only; a sample of cell phone numbers was added in 2011–2012. In 2011–2012, the proportion of households known to include children whose parents completed the interview, known as the interview completion rate, was 54.1% and 41.2% for landline and cell phone samples, respectively. The overall response rate was 23% in 2011–2012, 46.7% in 2007, and 55.3% in 2003. ^16–18 The analyses were restricted to children aged 6 to 17 years with valid data on sex and anxiety and depression questions (n = 65,411 of 65,680 in 2011–2012; n = 63,874 of 64,076 in 2007; and n = 68,786 of 69,031 in 2003).

To compare the prevalence of diagnosed anxiety or depression over time for children aged 6 to 17 years, parent-reported data from the 3 survey years were analyzed. In 2003, parents responded to a single question: “Has a doctor or health care provider ever told you that [CHILD] has depression or anxiety?” In 2007 and 2011–2012, anxiety and depression were asked about independently. Parents were asked the following questions: “Has a doctor or other health care provider ever told you that [CHILD] had anxiety problems?” and “Has a doctor or other health care provider ever told you that [CHILD] had depression?” If yes, parents were asked, “Does [CHILD] currently have anxiety problems” and “Does [CHILD] currently have depression?” Parents were asked to rate current anxiety or depression as mild, moderate, or severe.

Sociodemographic characteristics included child's sex, age, race, Hispanic ethnicity, household income relative to the federal poverty level, parental educational attainment, region, and health insurance status and type (Table 1).

The parent report of 11 other specified neurobehavioral disorders and 9 specified chronic health conditions (Table 2) used the same question structure as that for anxiety and depression (i.e., whether a doctor or other health care providers ever told them their child had [condition], and if yes, if their child currently had the condition). Body mass index (BMI) was calculated using parent-reported height and weight; a BMI in the 95th percentile or higher was considered obese.¹⁹ BMI data for children younger than 10 years were not available because of concerns about the validity of the height data; thus, BMI was excluded from Table 2 and as a control variable.

Selected indicators of impact included the following: whether the child had “special health care needs” (CSHCN; defined as having a chronic condition resulting in functional limitations or service use or need beyond that generally required by children of the same age, assessed with a validated screener) ²⁰; whether their child had any kind of emotional, developmental, or behavioral problem for which he/she needs treatment or counseling; parenting aggravation (high/low; past month, calculated from the parent report of how often they felt angry with their child, that it was harder to care for their child than for other children of the same age, and that their child did things that really bothered them)²¹; and if the child had repeated a grade other than kindergarten. Indicators for the past 12 months included whether the parent was contacted about school problems and whether the child had received preventive medical care; received treatment or counseling from a mental health professional; taken medication because of difficulties with emotions, concentration, or behavior; saw a specialist other than a mental health professional; had a medical home, based on a series of 19 questions assessing medical home criteria²²; received needed care coordination, 1 component of the medical home⁹; and had unmet health care needs, including medical and/or mental health services. Access to a medical home required the child to have a personal doctor or nurse, a usual source of care, family-centered care, and receive referrals and care coordination, if needed. Care coordination was based on parent responses about whether they received help, if needed, arranging or coordinating the child's care among doctors or services.

SAS-callable SUDAAN (version 11.0.0, RTI International, Cary, NC) was used to account for the complex sampling design. Weighted prevalence estimates and 95% confidence intervals were calculated to determine the prevalence of parent-reported diagnosed anxiety and depression.

Using 2011–2012 data, prevalence ratios (PRs) were calculated using logistic regression to compare the prevalence of parent-reported anxiety and depression among children aged 6 to 17 years across demographic subgroups and the prevalence of co-occurring conditions relative to children without anxiety or depression. Adjusted PRs controlled for child's sex, age (6–11 vs 12–17 years), race and ethnicity (white non-Hispanic vs other), household education, and household income. To clarify the specific relationship between anxiety and depression (as opposed to other conditions) with indicators of impact, these models were further adjusted for co-occurring conditions. When considering the relationship between having a medical home or effective care coordination and impact, the models were further adjusted for whether the child had severe depression or anxiety. Household income was imputed for those (9.3%) without valid income data; this variable was provided in the public use data set. All estimates are generalizable to the population of noninstitutionalized US children aged 6 to 17 years.

Among children aged 6 to 17 years, the prevalence of ever being diagnosed with anxiety or depression increased from 5.4% (confidence interval [CI]: 5.1–5.7) in 2003 to 7.8% (CI: 7.3–8.4) in 2007 and to 8.4% (CI: 8.0–8.9) in 2011–2012; this is a 56% increase in diagnosed prevalence from 2003 to 2011–2012. Similarly, current anxiety or depression increased 13% from 4.7% (CI: 4.3–5.1) in 2007 and to 5.3% in 2011–2012 (p < 0.05). Separate estimates for individual diagnoses were as follows: current anxiety increased 19% from 3.5% (95% CI: 3.2–3.9) in 2007 to 4.1% in 2011–2012 (p < 0.01); ever-diagnosed anxiety increased 17% from 5.5% (CI: 5.1–6.0) to 6.4% (p < 0.01); current depression did not change significantly from 2007 (2.5%, CI: 2.2–2.8) to 2011–2012 (2.7%, p = 0.18, p > 0.05); and ever-diagnosed depression did not change significantly from 4.7% (CI: 4.3–5.2) in 2007 to 4.9% (CI: 4.5–5.3) in 2011–2012 (p > 0.05). These estimates correspond with approximately 2 million children aged 6 to 17 years in 2011–2012 with current anxiety, 1.4 million children with current depression, 2.6 million with current anxiety or depression, and 760,000 children with both.

Anxiety and depression among children aged 6 to 17 years varied by sociodemographic characteristics (Table 1). Current depression and current anxiety were more common among older children and those living in households with lower income. Current depression was more common among children living in households with lower parental education. Current anxiety and either anxiety or depression were more common among white non-Hispanic children and less common among children without health care coverage and children living in the West compared with the Northeast. The sociodemographic patterns were similar for ever-diagnosed depression or anxiety with 1 exception; higher parental education was associated with ever-diagnosed anxiety ( χ12=5.7, p < 0.05).

Overall, sex was not significantly associated with anxiety or depression. Age-stratified post hoc analyses were conducted to identify more specific relationships between sex and internalizing disorders, given previously documented associations.¹² Boys aged 6 to 11 years were more likely than girls of the same age to have depression ( χ12=9.5, p < 0.01), anxiety ( χ12=9.4, p < 0.01), or either ( χ12=13.3, p < 0.001); no differences were noted for children aged 12 to 17 years. Among children 15 to 17 years, current anxiety or depression was more common among girls (8.6% vs 6.7%; p < 0.05); there were no sex differences for either individual condition.

Children with anxiety or depression were more likely than those without either condition to have co-occurring conditions; this difference was significant for each individual condition (Table 2). Children with current anxiety or depression were more likely than those without anxiety or depression to be obese ( χ12=10.5, p < 0.01, 21.2% vs 15.3%). Specifically, 25.4% of children with current depression were obese (adjusted PR [aPR] 5 1.7, CI: 1.3–2.2), whereas 20.5% of children with current anxiety were obese (aPR = 1.4, CI: 1.1–1.7).

After adjusting for sociodemographic factors, anxiety or depression was associated with indicators of impact with 1 exception: receiving preventive medical care (data not shown). After further adjusting for co-occurring conditions (Table 3), the findings remained significant, except for the associations between depression and seeing a specialist and having an unmet health care needs including mental health services, between anxiety and unmet medical care needs, and between either anxiety or depression and repeating a grade. Considering children with anxiety or depression, 25% did not receive mental health treatment or counseling; this decreased to 19% when including medication for emotion, concentration, or behavior. Children with anxiety or depression were less likely than those with other chronic health conditions to have a medical home (aPR = 0.7, CI: 0.6–0.8) or effective care coordination (aPR = 0.7, CI: 0.6–0.8).

Among children aged 6 to 17 years, parents described current depression as mild for 53.5%, moderate for 35.7%, and severe for 10.7%; current anxiety was mild for 48.0%, moderate for 36.8%, and severe for 15.2%. There was no association between depression severity and having a medical home; however, children with a medical home or effective care coordination were more likely to have mild anxiety, rather than moderate or severe (data not shown). After adjusting for the severity of both conditions, children with anxiety or depression, and without effective care coordination and a medical home, were more likely than those with a medical home or effective care coordination to have parents with high parenting aggravation and unmet health care needs, including mental health services (Fig. 1). Findings were similar when considering effective care coordination and medical home separately.