Abstract
Purpose
The purpose of this study is threefold: Determine recent trends in several mental health problems in the USA, identify risk factors that may be responsible for the trends and evaluate intervention policies to reduce the consequences of these problems.
Design/methodology/approach
This study used data from the National Survey of Children's Health (NSCH), a nationally representative survey of children under the age of 17 that was conducted between 2016 and 2022. Prevalence rates in the data take into account the probability of selection and nonresponse. Because of the possible correlation in the longitudinal responses in the data, an appropriate extension of the generalized linear models (the marginal models) was used. Marginal models, also known as population-average models, do not require distributional assumptions for the observations, only a regression model for the mean response. The avoidance of distributional assumptions leads to the use of the generalized estimating equations (GEE) method.
Findings
The author found that the odds of children and adolescents experiencing mental health problems in the USA changed over a seven-year period, from 2016 to 2022. Anxiety and depression, in particular, have both increased, with anxiety increasing faster than depression; however, behavioral issues and attention deficit disorder/attention deficit hyperactivity disoder (ADD/ADHD) remained stable until 2020 (the start of COVID-19), when they began to rise. This paper also found a link between increased social media use and increased mental health problems, and bullying has a negative impact on the mental health of children and adolescents.
Originality/value
The NSCH, an annual representative survey, was used in this study to assess mental health problems among children and adolescents in the USA. Marginal models, which enable the capture of potential correlations among observations of the same subject, were used in conjunction with the GEE method. This study differs from previous research, which used other surveys, pre-COVID-19 data points and logistic regressions that assumed independence in repeated observations.
Keywords
Citation
Lachaab, M. (2024), "Trends, risk factors and interventions for some mental health problems in the US children and adolescents: evidence from the National Survey of Children’s Health, 2016-2022", Journal of Public Mental Health, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/JPMH-12-2023-0108
Publisher
:Emerald Publishing Limited
Copyright © 2024, Emerald Publishing Limited
1. Introduction
Problems with mental health in US children and adolescents pose a public health concern and burden and contribute to the leading causes of disability and death in this subpopulation (Fatori et al., 2013; Bor et al., 2014; Twenge et al., 2019). The economic costs of depression only are also very high in the USA (Pelham et al., 2007; Mrazek et al., 2014). Evaluating the time trends of these mental health problems can help guide decisions about resource allocation and other future government services and programs for improving the health and well-being of these children and adolescents and the whole nation.
Several studies have been drawn to this topic. Among the recent ones are Twenge et al. (2017, 2019), Mojtabai and Olfson (2020), Daly (2022) and Goodwin et al. (2022). However, their findings are still mixed, and the trends of these mental health issues are still debated. For instance, some found that some mental health disorders are on the rise, while others found that they are declining. Furthermore, most studies are based on data drawn from the National Survey on Drug Use and Health (NSDUH), and they have mostly focused on adolescents aged 12 to 17 years and have used pre-COVID-19 data.
In addition, the identification of the risk factors for these mental health problems can help determine the intervention policies, programs and practices needed to reduce these burdens, disabilities and deaths. Studies’ findings on risk factors of these mental health issues and intervention policies are mixed.
To our knowledge, the only study that used the most recent annual, representative data from the National Survey of Children's Health (NSCH) which focused on both children and adolescents is that of Lebrun-Harris et al. (2022). However, the data they used do not allow inferences about the post effects of the COVID-19 pandemic on children and adolescents’ health and well-being. The incorporation of NSCH-2021 and NSCH-2022 data in our study allows us to assess seven-year trends in children and adolescents' mental health problems and investigate the effects of the COVID-19 pandemic. Furthermore, the study of Lebrun-Harris et al. (2022) applied logistic regressions that assumed the independence of the clustered observations. In our study, we applied a general statistical approach (the marginal models/generalized estimating equations [GEE] method), which allows for possible correlations among observations of the same subject.
We herein examine the time trends of some mental health conditions (depression, anxiety, behavior problems and ADD/ADHD) using recent data from the NSCH from 2016 to 2022. We have considered the four mental health conditions that are commonly used in the literature and are available in the NSCH data set. We hypothesized that there would be a different pattern in the increase in these mental health conditions, and that this increase would be driven by risk factors including social media use and bullying victimization. Therefore, in addition to analysis of time trends of these mental health issues, we look at the potential risk factors and recommendation options for these mental health issues. The findings of our study have implications for public health and policy.
The rest of this paper is structured as follows: Section 2 provides a review of the literature related to trends in mental health issues, potential risk factors of these trends and effective interventions to reduce the consequences of these problems. In Section 3, we present the data, perform the statistical analysis and report the results. Section 4 is devoted to the discussion, and Section 5 concludes.
2. Literature review
Over the past few years, a growing body of literature has focused on the prevalence of mental health problems in the US population. Twenge et al. (2019) assessed age, period and birth cohort trends in mood disorders and suicide-related outcomes between 2005 and 2017 using data from the NSDUH, and using hierarchical linear modeling to separate the effects of age, time period and cohort/generation effects. They found that the rates of major depressive episodes, serious psychological distress and suicide-related outcomes (suicidal ideation, plans, attempts and deaths by suicide) increased during that period. They also found that cultural trends (electronic communication, digital media and a decline in sleep duration) contribute to an increase in mood disorders and suicidal thoughts and behaviors.
Using data from the same survey (NSDUH) from 2005 to 2018, and logistic and negative binomial regressions, Mojtabai and Olfson (2020) examined national trends in the care of different mental health problems and in different treatment settings among adolescents. Their findings show that a greater proportion of adolescents received care for internalizing problems (depression or suicidal ideations), whereas a smaller proportion received care for externalizing and relationship problems. Furthermore, Daly (2022) studied the temporal trends in adolescent depression by sociodemographic characteristics using 11 years of data from the same NSDUH survey and logistic regression analysis with cluster robust standard errors. He showed a sustained increase in depression among adolescents (especially girls) in the USA between 2009 and 2019. In addition, Goodwin et al. (2022) found that depression increased most rapidly among adolescents and young adults and increased among nearly all sex, racial/ethnic, income and education groups, using the NSDUH survey of adolescents from 2015 to 2020 and Poisson regression with robust standard errors point estimates.
A study based on different surveys (Monitoring the Future and the Risk Youth Behavior Surveillance System) and logistic regressions was conducted by Twenge et al. (2017), who found that new media screen time (including reported use of electronic devices, social media and reading news online) significantly increased the likelihood of short sleep duration, and time spent on these screen activities increased between 2009 and 2015.
The preceding studies used data from the NSDUH survey and other surveys, as well as pre-COVID-19 data. We are aware of only one study that used NSCH, that of Lebrun-Harris et al. (2022), who investigated trends in children's health-related measures. Their findings, based on logistic regressions, show significant increases in children's diagnosed anxiety and depression, as well as decreases in physical activity, between 2016 and 2020. However, Lebrun-Harris et al. (2022) applied logistic regressions that assumed the independence of the clustered observations. In our study, we applied a general statistical approach (the marginal models/GEE method), which allows for possible correlations among observations of the same subject.
As regards the causes of some mental health problems, findings from previous studies present a diversity of psychosocial factors. Factors are gender, age, living with a single parent, community violence, witnessing physical violence, family drug involvement, negative family interaction, school disengagement, child physical abuse, social economic status and maternal anxiety or depression.
Some of these previous studies explained how these factors impact the mental health of children and adolescents. Gender difference in depression, for example, is explained by the increased pressure on girls and boys to conform to normal gender norms during adolescence. Girls enter adolescence before boys, resulting in physical and hormonal changes, and this has been associated with depressive symptoms. School disengagement puts students at risk for dropping out, academic failure and other negative psychological outcomes such as depression. Furthermore, children with poor socioeconomic position frequently have less access to education and social activity, resulting in more health problems, particularly mental health, than their peers with higher socioeconomic status. In addition, children were more likely to have ADHD symptoms or display aggressive or hostile behavior if their mothers had higher worry, sadness or stress during pregnancy.
However, previous studies’ findings on these risk factors are still mixed. In this study, we focus on some possibly responsible causes of these mental health problems: Childhood and adolescence adversities (home violence, child lived with someone alcoholic and child lived with someone mentally ill), family structure (divorce, death and jail or prison), neighborhood factors (neighborhood violence and bullying) and other cultural and environmental factors (time spent with social media and hours of sleep).
Childhood and adolescent adversities are risk factors for some mental health problems. In fact, when children and adolescents witness domestic violence, they are more likely to react negatively. They may fight or miss school. They may also be more likely to experience negative emotions such as anxiety and live in anticipation for the next time a physical or verbal assault occurs in their household. As regards family structure, children and adolescents, whose parents are imprisoned, died or divorced may experience low self-esteem, depression and disturbed sleeping patterns. In addition, living in neighborhoods with violence can affect children and adolescents' development by changing the way that a part of the brain detects and responds to potential threats, which could lead to poorer mental health.
The effects of bullying also have negative impacts on children and adolescents’ mental health in many ways (Chou et al., 2020). It can cause feelings of rejection, exclusion, isolation, low self-esteem and some can develop depression and anxiety as a result. Children and adolescents experiencing bullying may feel that they are not worth helping or that nobody likes them. They may feel self-conscious or embarrassed lots of the time. They may also feel scared, sad or overwhelmed, and find it difficult to sleep or eat. Many children and adolescents who have been bullied find it hard to ever feel safe or confident in anything they do, leading them to isolate themselves from others and to give up the things they enjoy doing.
3. Data and statistical analysis
3.1 Data
The data that we use come from the NSCH, a nationally representative survey of children younger than 17 years old living in the 50 States and the District of Columbia. The NSCH is funded and directed by the Health Resources and Services Administration, Maternal and Child Health Bureau (2023) and fielded by the U.S. Census Bureau. Since 2016, this survey has been repeated annually, making trend analysis possible. This survey has many questions including sociodemographics, risk factors and interventions of mental health problems.
We analyzed the data from 2016 through 2022. For the year 2021, the NSCH survey was done from June 2020 to January 2021, and therefore the data collection was not disrupted by the COVID-19 pandemic during that year. The combined sample size from 2016 to 2022 included 275,133 children (individuals are not counted more than once), with an annual range from 21,599 in 2017 to 54,103 in 2022. We have removed the missing values for the selected variables and considered only repeated measures of two or more on each subject, which left us with 61,385 observations. We finally pooled seven years of data into a single data file, which included a variable for the survey year.
3.2 Statistical analysis
The mental health problems considered in this analysis are: depression, anxiety, behavior problems and ADD/ADHD disorders. The survey questions asked about whether or not the child has a mental health condition and the severity of the condition. Following the study’s three main objectives, I have pursued one goal in each stage. The fourth stage was included to investigate the severity of mental health disorders, which is ordinal. In the four stages of the analysis, appropriate extensions of the generalized linear models were applied.
In the first stage, the temporal trends of mental health disorders were assessed across survey years using marginal models along with the GEE method (Liang and Zeger, 1986), which were adjusted for sex, age, race, ethnicity and family income. The second stage involved investigating potential risk factors for these mental health using similar marginal/GEE models. In the third stage, the severity (mild, moderate or severe) of the mental health problems was also assessed using cumulative logit models adjusted for the risk factors. And in the final stage, we investigate the potential intervention measures for these mental health problems.
3.2.1 First stage analysis: temporal trends.
We consider a marginal model for analyzing the longitudinal data. Marginal models extend generalized linear models to longitudinal data and they have frequently been used in biomedical and health sciences. We let Yij = 1 if the ith child experienced a mental health problem in the jth year and Yij = 0 otherwise, for j = 2016, …, 2022. The distribution of each Yij is Bernoulli and the probability of success is modeled using a logit link function. We are interested in relating the changes in E(Yij) to some covariates. That is, associated with each response, Yij, there is a vector of covariates, Xij, i = 1, … N; j = 1, …, ni. The vector of responses, Yi, are assumed to be independent of one another, but the repeated measures on the same subject are not assumed independent.
The probability of having mental health problem is modeled as a logistic function:
The outcome variable is a binary indicator of mental health disorder in a child for which they need treatment or counseling. Ethnicity is categorized as Hispanics or latino, not Hispanics or latino. Race is categorized as black or African American, white and other. Family poverty ratio is categorized as 50%–100%, 100%–199%, 200%–299% and 299%–400% of the federal poverty, hard to cover basics is categorized as never, rarely, somewhat often and very often, and age is categorized as 0–5, 6–12 and 13–17 years old.
The term marginal indicates that the model for the mean response at each occasion depends only on the covariates of interest, and not on any random effects or previous responses. However, the specification of the mean and the covariance does not determine the joint distribution of Yi. As a result, the method of maximum likelihood cannot be used for the estimation of the parameters in the marginal model without further distributional assumptions. The GEE approach provides a convenient alternative to maximum likelihood estimation. The essential idea behind the GEE approach is to extend the usual likelihood equations for a generalized linear model by incorporating the covariance matrix of the vector of responses.
The GEE method is often referred to as the “population-average model,” where the target of inference is the population. The primary goal is to make inferences about population means. The within-subject association among repeated responses is regarded as a nuisance characteristic of the data that must be accounted for to make correct inferences about changes in the population response mean. For a given estimate of α, Liang and Zeger (1986) show that the GEE estimate of the parameter β is the solution of an equation that has no closed-form solution; instead the solution requires an iterative algorithm (scheme), which switches between estimating β for fixed value of α and estimating α for fixed values of β. This scheme yields a consistent estimate for β. That is, with very high probability, the estimator is close to the population parameter in large samples.
We use the R software to estimate the above marginal model by the GEE. The function that performs this estimation is geeglm() in the geepack package. The estimation results are presented in Table 1. These results provide evidence that the log odds of mental health problems experienced by children changed over the seven-year period, 2016–2022. For example, the log odds of having depression increased, with an increase in the log odds of 0.085 (or 0.091–0.006) in 2016, 0.158 (or 2 * 0.091–4 * 0.006) in 2017, 0.219 (or 3 * 0.091–9 * 0.006) in 2018, 0.268 (or 4 * 0.091–16 * 0.006) in 2019, 0.305 (or 5 * 0.091–25 * 0.006) in 2020, 0.33 (or 6 * 0.091–36 * 0.006) in 2021, and 0.343 (or 7 * 0.091–49 * 0.006) in 2022. These increases in log odds correspond to odds of 1.089 (e0.085) in 2016, 1.171 (e0.158) in 2017, 1.245 (e0.219) in 2018, 1.307 (e0.268) in 2019, 1.356 (e0.305) in 2020, 1.391 (e0.33) in 2021 and 1.409 (e0.343) in 2022.
Similarly, we computed the temporal odds for the other different mental health problems (anxiety, behavior problems and ADD/ADHD). Figure 1 shows the trends for the four mental health problems. Depression and anxiety have both increased during this time span, with anxiety increasing more rapidly than depression. However, behavior issues and ADD/ADHD remained steady until about 2020 (beginning of COVID-19) and then increased.
Examining whether these mental health issues vary on gender, race/ethnicity and income level, to determine which groups are most affected, we found that females, non-Hispanics and whites have higher log odds of having a mental health issue. We also found that children and adolescents living in poverty (low socioeconomic class) have higher odds of having mental health problems than those who were not in poverty, and adolescents are more likely to have depression and anxiety.
Finally, the estimate of the within-subject association is about 0.10, indicating a positive association. That is, the odds of having a mental health problem in one year is approximately 1.11 exp(0.10) higher if the child had a mental health problem in the previous year (this suggests that a mental health problem lasts more than one year). These results were obtained using the unstructured covariance matrix, as observations closer together in time may be more heavily correlated. When we fitted a model with an exchangeable working covariance structure, we got comparable findings.
3.2.2 Second stage analysis: risk factors.
Mental health disorders are linked to different social, cultural or environmental factors (including life experiences) (Bor et al., 2014). For example, reasons for the increase in depression include increases in bullying and victimization (Chou et al., 2020), and the use of social media and technology (Bhandari et al., 2017; Li et al., 2017).
We consider the following risk factors that are commonly used in the literature and are available in the NSCH data set: Parent or guardian divorced or separated; parent or guardian served time in jail or prison; parent or guardian died; child lived with anyone who had a problem with alcohol or drugs; child lived with anyone who was mentally ill (the source is the NSCH data set), suicidal or severely depressed; child saw or heard parents or adults slap, hit, kick or punch one another at home; child was a victim of violence or witnessed violence in their neighborhood; time spent with social media; hours of sleep the child gets on most weeknights; and child bullied, picked on or excluded by other children. We consider the following logit model:
Table 2 reports the estimation results of some mental health risk factors. We found a positive association between the increase in time spent on social media and the increase in depression. We also found that the increase in sleep time has contributed to a decrease in depression, anxiety and behavior problems. Furthermore, our results show that bullying victimization has adverse outcomes for children and adolescents with mental health problems. Indeed, as the frequency of bullying increases, the odds ratio of mental health problems increases. Our study also shows that life events (divorce or separation of parents, death in family, jail or prison of family members) affect mental health impairment. In addition, child experiences with home or neighborhood violence, living with alcoholic persons or living with mentally ill people can cause mental health issues.
3.2.3 Third stage analysis: severity of mental health problems.
Mental health impairment is ordinal, with categories: mild, moderate and severe. When the categories are ordered, the logits can use the ordering. This results in models that have simpler interpretations and greater power than baseline-category logit models, which treat the response variable as nominal rather than ordinal. We relate mental health problems to some explanatory variables, some of which are measured as binary and some have multicategories. The cumulative logit model under consideration takes the following form:
The estimation results of model (4) are presented in Table 3. With three response categories, the model has only two intercepts. These intercepts allow for computing the response probabilities. The results show that the estimated odds of the severity of mental health problems below any fixed level are higher when the child has parents who are divorced, dead or in jail or prison. Furthermore, these cumulative odds are higher when the child experienced home or neighborhood violence or lived with someone who was alcoholic or mentally ill.
In addition, the estimated odds of the severity of mental health problems below any fixed level increase as the time spent using social media increases, especially when the time spent is four or more hours. And these odds increase with the frequency of bullying but decrease with the number of sleep hours.
3.2.4 Fourth stage analysis: intervention measures.
Given the burden of mental health disorders among children and adolescents in the USA, it is essential that effective interventions and preventions are identified and implemented. We attempt to evaluate the effectiveness of interventions and preventions for improving adolescent health and well-being.
We consider the following variables from the NSCH: the number of days the child exercises, plays a sport or participates in physical activity for at least 60 min; whether the child participates in a sports team or takes sports lessons after school or on weekends; whether the child participates in any clubs or organizations after school or on weekends; whether the child needs or gets special therapy, such as physical, occupational or speech therapy; and whether the child uses any type of alternative health care or treatment. Alternative health care can include relaxation therapies, herbal supplements and others. Therefore, the model is:
Table 4 presents the estimates of these intervention measures. As shown in the table, exercising, playing a sport or participating in physical activity for at least 60 min for the child was found to be effective in reducing depression, anxiety and behavior problems with no impact on ADD/ADHD. Similar results were found for participating in a sports team, taking sports lessons after school or on weekends or participating in any clubs or organizations after school or on weekends, and receiving special therapy.
4. Discussion
Information on recent trends in mental health problems among children and adolescents in the USA is needed to help guide decisions about resource allocation and other future government services and programs aimed at improving their health. Furthermore, identifying the risk factors for these mental health problems can aid in the development of intervention policies, programs and practices that will reduce the public burden and costs associated with these issues.
With respect to trends, we found a significant increase in depression and anxiety during the time span 2016–2022, with anxiety increasing more rapidly than depression. However, behavior problems and ADD/ADHD remained steady until the beginning of COVID-19 and then increased. These results are consistent with and extend findings from earlier studies that have found increases in depression and anxiety and decreases in behavior problems among children and adolescents in the USA (Bor et al., 2014; Fink et al., 2015; Mojtabai et al., 2016; Mojtabai and Olfson, 2020; Twenge et al., 2019; Keyes et al., 2019; Daly, 2022; Goodwin et al., 2022).
In fact, using 11 years of the NSDUH of adolescents aged 12–17 years old, Daly (2022) found that the prevalence of past-year major depression increased by 7.7 percentage points from 8.1% to 15.8% between 2009 and 2019. He also found that major depression increased by 12 percentage points from 11.4% to 23.4% among girls, and the gender difference in the prevalence of major depression increased from 6.4% to 14.8% between 2009 and 2019. His results also show that black participants experienced a comparatively small increase in depression (4.1%). Using the same NSDUH of adolescents from 2015 to 2020, Goodwin et al. (2022) found that depression increased most rapidly among adolescents and young adults and increased among nearly all sex, racial/ethnic, income and education groups.
We also found varying patterns of mental health problems among subgroups. Specifically, we found that females, non-Hispanics and whites have higher log odds of having a mental health issue, while children and adolescents living in poverty have higher odds of having mental health problems than those who do not live in poverty, and adolescents are more likely to suffer from depression and anxiety. These demographic profiles align with previous research findings (Twenge et al., 2019).
In terms of risk factors for mental health problems, we found a link between increased social media use and an increase in depression. Our findings are consistent with those of Riehm et al. (2019), who discovered that adolescents who spend more than three hours per day on social media are at a higher risk for mental health problems, particularly internalizing problems.
Our findings are also in line with the results of Twenge and Martin (2020), who used three large surveys of 13- to 18-year-old adolescents in the USA and UK to investigate the relationship between daily digital media use and several measures of psychological well-being. They found that the associations between moderate or heavy digital media use and low psychological well-being and mental health issues were generally larger for girls than for boys. They also found that, regardless of gender, heavy users of digital media were often twice as likely as low users to be low in well-being or have mental health issues, including risk factors for suicide.
Our study findings also confirm previous reports that the increase in sleep time contributes to a decrease in depression, anxiety and behavior problems (Twenge et al., 2017) and that bullying victimization has adverse outcomes for children and adolescents with mental health problems (Pontes et al., 2018; Pengpid and Peltzer, 2019; and Chou et al., 2020).
As regards the effectiveness of some intervention policies to reduce the consequences of these problems, our results show that sports activities were found to be effective in reducing depression, anxiety and behavior problems. This result is consistent with previous results that show that physical activity is associated with decreased symptoms of depression and anxiety (Larun et al., 2006; Lubans et al., 2012; Cooney et al., 2013). In addition, it was found that attending specialized therapy, joining a club or group after school or on the weekends, playing sports and taking sports instruction all helped to lessen these mental health issues.
5. Conclusion
In this paper, we used recent NSCH data from 2016 to 2022 and a general statistical approach to examine the time trends of some mental health problems in children and adolescents in the USA. We found different patterns in the increase of these mental health conditions. We also looked at the risk factors and recommended treatment options for these mental health issues. We found a link between increased social media use and an increase in mental health problems, and bullying has negative consequences for children and adolescents with mental health issues. Our findings also show that physical activity and participation in clubs and organizations are associated with fewer mental health problems.
The findings of the study have implications for public health and policy. First, public health interventions should consider reducing electronic device use as a target of intervention to improve the health of children and adolescents. Second, given the impact of bullying on mental health problems, more work in policies, programs and practice is needed to reduce bullying. Third, the study finding emphasizes the critical importance of directing prevention and intervention efforts toward females. Finally, the findings indicate a need for additional research to better understand the role that risk factors such as social media, bullying, violence and sleep disruption may play in mental health issues.
There exist certain limitations to this study. First, potential variations in patterns within subpopulations are not depicted in the presented overall trends. Consequently, more research is required to determine how much the differences among the sociodemographic groups have evolved during the period 2016–2022. Second, further study is intended to investigate the degree of differences between gender and socioeconomic groups since evaluating the outcomes of interventions based on gender and socioeconomic status is crucial. Third, additional analysis that considers other indicators of mental health problems in children and adolescents, such as eating disorders and post-traumatic stress disorder, is needed to inform policy priorities and interventions in the USA.
Figures
Figure 1
Trends in mental health problems of the US children, 2016–2022
Odds ratios from the marginal model/GEE of the children and adolescents mental health problems in the USA from 2016–2022
| Variable | Depression | Anxiety | Behavior Pbs | ADD/ADHD |
|---|---|---|---|---|
| Time | ||||
| Time | 1.095*** | 1.077*** | 0.946*** | 0.849*** |
| Time2 | 0.993*** | 1.001*** | 1.013*** | 1.021*** |
| Gender | ||||
| Female | 1.000 | 1.000 | 1.000 | 1.000 |
| Male | 0.701*** | 0.878*** | 0.883*** | 0.798*** |
| Age in years | ||||
| 0–5 | 1.000 | 1.000 | 1.000 | 1.000 |
| 6–12 | 1.322*** | 1.128*** | 0.847*** | 1.289*** |
| 13–17 | 1.129*** | 1.994** | 0.429*** | 0.439** |
| Ethnicity | ||||
| Non-Hispanics | 1.000 | 1.000 | 1.000 | 1.000 |
| Hispanics | 0.743** | 0.697*** | 0.843*** | 0.914*** |
| Race | ||||
| African-American | 1.000 | 1.000 | 1.000 | 1.000 |
| White | 1.282*** | 1.012*** | 1.253*** | 1.403*** |
| Other | 1.172*** | 0.894** | 1.024*** | 1.347*** |
| Poverty ratio | ||||
| 50–99% | 1.000 | 1.000 | 1.000 | 1.000 |
| 100–199% | 0.973*** | 0.972*** | 0.832*** | 0.934*** |
| 200–299% | 0.997*** | 0.748*** | 0.900*** | 0.994** |
| 300–400% | 0.888*** | 0.664** | 0.722*** | 0.988** |
| Hard to cover basics | ||||
| Never | 1.000 | 1.000 | 1.000 | 1.000 |
| Rarely | 1.022** | 0.992*** | 1.033** | 1.022*** |
| Somewhat often | 1.341*** | 0.893** | 1.542*** | 1.113*** |
| Very often | 1.701** | 1.022** | 2.002** | 1.092** |
***is 1% significant and **is 5% significant
Source: Table by author
Odds ratios from the marginal model and GEE of the children mental health problems in the USA from 2016–2022
| Variable | Depression | Anxiety | Behavior Pbs | ADD/ADHD |
|---|---|---|---|---|
| Parent divorced or separated | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.150*** | 0.931*** | 0.991*** | 0.781*** |
| Parent spent time in jail or prison | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 0.977*** | 1.141*** | 1.242*** | 1.235*** |
| Parent died | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.483*** | 1.122*** | 1.078*** | 0.853*** |
| Child lived with alcoholic person | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 0.932*** | 0.867*** | 0.944*** | 0.972*** |
| Child lived with mentally ill person | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.376**** | 1.741*** | 1.144*** | 1.122*** |
| Child experienced home violence | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.171*** | 0.899*** | 1.345*** | 1.281*** |
| Child experienced neighborhood violence | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.229*** | 1.105*** | 0.754*** | 1.051*** |
| Time spent with social media | ||||
| Less than 1 h | 1.000 | 1.000 | 1.000 | 1.000 |
| 1 h | 1.462*** | 1.085*** | 1.694*** | 1.031*** |
| 2 h | 1.440*** | 1.293*** | 1.671*** | 1.034*** |
| 3 h | 1.225** | 1.769*** | 1.563*** | 1.174*** |
| 4 or more hours | 1.606*** | 1.890*** | 1.620*** | 1.264*** |
| Average sleep hours | ||||
| Less than 6 h | 1.000 | 1.000 | 1.000 | 1.000 |
| 6 h | 0.873*** | 0.514*** | 0.258*** | 0.902*** |
| 7 h | 0.712*** | 0.313*** | 0.290*** | 0.976** |
| 8 h | 0.677*** | 0.331*** | 0.332*** | 0.959** |
| 9 h | 0.591*** | 0.336*** | 0.357*** | 0.881*** |
| 10 h | 0.631*** | 0.444*** | 0.357** | 0.767** |
| Bullied, picked on or excluded | ||||
| Never (in the past 12 months) | 1.000 | 1.000 | 1.000 | 1.000 |
| 1–2 times (in the past 12 months) | 1.191*** | 1.168*** | 1.209*** | 1.520*** |
| 1–2 times per month | 1.157*** | 1.980*** | 1.171** | 1.526*** |
| 1–2 times per week | 2.613*** | 2.147*** | 2.355*** | 2.121*** |
| Almost every day | 3.586*** | 2.674*** | 3.960*** | 3.693*** |
***and ** indicate that the statistics are significant at the 1% and 5% respectively
Source: Table by author
Results for fitting the cumulative logit model (odds numbers)
| Variable | Depression | Anxiety | Behavior Pbs | ADD/ADHD |
|---|---|---|---|---|
| Intercepts | ||||
| Mild|moderate | 0.618*** | 0.614*** | 0.703*** | 0.460** |
| Moderate|severe | 2.322*** | 2.862*** | 2.261*** | 1.010** |
| Parent divorced or separated | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 2.003*** | 1.988** | 1.223*** | 0.997** |
| Parent died | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.667** | 1.231*** | 1.032** | 1.211*** |
| Parent spent time in jail or prison | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.788** | 1.998*** | 1.170*** | 1.345*** |
| Child lived with an alcoholic person | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.433*** | 1.566*** | 1.921** | 1.103** |
| Child lived with a mentally ill person | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.483*** | 1.456*** | 1.266*** | 1.325*** |
| Child experienced home violence | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.775** | 1.422** | 1.703*** | 1.007** |
| Child experienced neighborhood violence | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.284*** | 1.335*** | 1.502** | 1.075*** |
| Time spent with social media | ||||
| Less than 1 h | 1.000 | 1.000 | 1.000 | 1.000 |
| 1 h | 1.203** | 0.875** | 1.984** | 0.798** |
| 2 h | 1.109*** | 0.910*** | 1.233*** | 0.992*** |
| 3 h | 1.566*** | 0.958*** | 1.332*** | 1.022*** |
| 4 or more hours | 1.677*** | 1.213*** | 1.403*** | 1.293*** |
| Average sleep hours | ||||
| Less than 6 h | 1.000 | 1.000 | 1.000 | 1.000 |
| 6 h | 0.578*** | 0.559*** | 0.692*** | 0.635*** |
| 7 h | 0.478*** | 0.452*** | 0.537*** | 0.538*** |
| 8 h | 0.452*** | 0.441*** | 0.553*** | 0.553*** |
| 9 h | 0.406*** | 0.428*** | 0.488*** | 0.552*** |
| 10 h | 0.488*** | 0.448*** | 0.519*** | 0.595*** |
| Bullied, picked on or excluded | ||||
| Never (in the past 12 months) | 1.000 | 1.000 | 1.000 | 1.000 |
| 1–2 times (in the past 12 months) | 0.961*** | 1.015*** | 0.878*** | 1.142*** |
| 1–2 times per month | 1.039*** | 1.082*** | 0.930*** | 1.139*** |
| 1–2 times per week | 1.448*** | 1.669*** | 1.313*** | 2.127*** |
| Almost every day | 2.022*** | 2.890*** | 2.159*** | 3.139*** |
***and ** indicate that the statistics are significant at the 1% and 5% respectively
Source: Table by author
Odds ratios from the marginal model/GEE of the children mental health problems in the USA from 2016–2022
| Variable | Depression | Anxiety | Behavior Pbs | ADD/ADHD |
|---|---|---|---|---|
| Physical activity per week | ||||
| 0 days | 1.000 | 1.000 | 1.000 | 1.000 |
| 1–3 days | 0.744*** | 0.974*** | 0.873*** | 1.035*** |
| 4–6 days | 0.526*** | 0.667*** | 0.848*** | 1.459*** |
| Every day | 0.724*** | 0.712*** | 0.938*** | 1.574*** |
| Sports team | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 0.878*** | 0.836*** | 0.889*** | 0.960** |
| Clubs/organizations | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 0.894*** | 0.797*** | 0.651*** | 0.975** |
| Special therapy | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 0.933*** | 0.532*** | 0.831*** | 0.995** |
| Alternative health care | ||||
| No | 1.000 | 1.000 | 1.000 | 1.000 |
| Yes | 1.057** | 1.402** | 1.137** | 1.093** |
*** and **indicate that the statistics are significant at the 1% and 5% level, respectively
Source: Table by author
References
Bhandari, P.M., Neupane, D., Rijal, S., Thapa, K., Mishra, S.R. and Poudyal, A.K. (2017), “Sleep quality, internet addiction and depressive symptoms among undergraduate students in Nepal”, BMC Psychiatry, Vol. 17 No. 1, p. 106.
Bor, W., Dean, A.J., Najman, J. and Hayatbakhsh, R. (2014), “Are child and adolescent mental health problems increasing in the 21st century? A systematic review”, Australian & New Zealand Journal of Psychiatry, Vol. 48 No. 7, pp. 606-616.
Chou, W.J., Wang, P.W., Hsiao, R.C., Hu, H.F. and Yen, C.F. (2020), “Role of school bullying involvement in depression, anxiety, suicidality, and low self-esteem among adolescents with high-functioning autism spectrum disorder”, Frontiers in Psychiatry, Vol. 11, p. 9.
Cooney, G.M., Dwan, K., Greig, C.A., Lawlor, D.A., Timer, J., Waugh, F.R., McMurdo, M. and Mead, G.E. (2013), “Exercise for depression”, Cochrane Database of Systematic Reviews, Vol. 2013 No. 9, p. CD004366.
Daly, M. (2022), “Prevalence of depression among adolescents in the U.S. from 2009 to 2019: analysis of trends by sex, race/ethnicity, and income”, Journal of Adolescent Health, Vol. 70 No. 3, pp. 496-499.
Fatori, D., Bordin, I.A., Bartira, M.C. and de Paula, C.S. (2013), “Influence of psychosocial risk factors on the trajectory of mental health problems from childhood to adolescence: a longitudinal study”, BMC Psychiatry, Vol. 13 No. 1, p. 31.
Fink, E., Begeer, S., Peterson, C.C., Slaughter, V. and de Rosnay, M. (2015), “Friendlessness and theory of mind: a prospective longitudinal study”, British Journal of Developmental Psychology, Vol. 33 No. 1, pp. 1-17.
Goodwin, R.D., Dierker, C.L., Wu, M., Galea, S., Hoven, C. and Weinberger, A. (2022), “Trends in U.S. depression prevalence from 2015 to 2020: the widening treatment gap”, American Journal of Preventive Medicine, Vol. 63 No. 5, pp. 726-733.
Health Resources and Services Administration, Maternal and Child Health Bureau (2023), National Survey of Children’s Health, Healthy Weight: Foundations in Early Childhood, Data Brief.
Keyes, K.M., Gary, D., O'Malley, P.M., Hamilton, A. and Schulenberg, J. (2019), “Recent increases in depressive symptoms among US adolescents: trends from 1991 to 2018”, Social Psychiatry and Psychiatric Epidemiology, Vol. 54 No. 8, pp. 987-996.
Larun, L., Nordheim, L.V., Ekeland, E., Hagen, K.B. and Heian, F. (2006), “Exercise in prevention and treatment of anxiety and depression among children and young people”, Cochrane Database of Systematic Reviews, Vol. 3 No. 3, p. CD004691.
Lebrun-Harris, A.L., Ghandour, M.R., Kogan, D.M. and Warren, D.M. (2022), “Five-year trends in US children’s health and well-being, 2016-2020”, JAMA Pediatrics, Vol. 176 No. 7, p. e220056.
Li, J.B., Lau, J.T.F., Mo, P.K.H., Su, X.F., Tang, J., Qin, Z.G. and Gross, D.L. (2017), “Insomnia partially mediated the association between problematic internet use and depression among secondary school students in China”, Journal of Behavioral Addictions, Vol. 6 No. 4, pp. 554-563.
Liang, K.Y. and Zeger, S.L. (1986), “Longitudinal data analysis using generalized linear models”, Biometrika, Vol. 73 No. 1, pp. 13-22.
Lubans, D.R., Plotnikoff, R.C. and Lubans N.J. (2012), “A systematic review of the impact of physical activity programmes on social and emotional well-being in at-risk youth”, Child and Adolescent Mental Health, Vol. 17 No. 1, pp. 2-13.
Mojtabai, R. and Olfson, M. (2020), “National trends in mental health care for US adolescents”, JAMA Psychiatry, Vol. 77 No. 7, pp. 703-714.
Mojtabai, R., Olfson, M. and Han, B. (2016), “National trends in the prevalence and treatment of depression in adolescents and young adults”, Pediatrics, Vol. 138 No. 6, p. e20161878.
Mrazek, D.A., Hornberger, J.C., Altar, C.A. and Degtiar, I. (2014), “A review of the clinical, economic, and societal burden of treatment-resistant depression: 1996–2013”, Psychiatric Services, Vol. 65 No. 8, pp. 977-987.
Pelham, W.E., Foster, E.M. and Robb J.A. (2007), “The economic impact of attention deficit/hyperactivity disorder in children and adolescents”, Journal of Pediatric Psychology, Vol. 32 No. 6, pp. 711-727.
Pengpid, S. and Peltzer, K. (2019), “Bullying victimization and externalizing and internalizing symptoms among in-school adolescents from five ASEAN countries”, Children and Youth Services Review, Vol. 106 No. 3, p. 104473.
Pontes, N.M., Ayres, C.G., Lewandowski, C. and Pontes, M.C. (2018), “Trends in bullying victimization by gender among US high school students”, Research in Nursing & Health, Vol. 41 No. 3, pp. 243-251.
Riehm, K.E., Feder, K.A., Tormohlen, K.N., Crum, R.M., Young, A.S., Green, K.M., Pacek, L.R., La Flair, L.N. and Mojtabai, R. (2019), “Associations between time spent using social media and internalizing and externalizing problems among US youth”, JAMA Psychiatry, Vol. 76 No. 12, pp. 1266-1273.
Twenge, M.J., Krizan, Z. and Martin, G.N. (2020), “Gender differences in associations between digital media use and psychological well-being: evidence from three large datasets”, Journal of Adolescence, Vol. 79 No. 1, pp. 91-102.
Twenge, M.J., Krizan, Z. and Hisler, G. (2017), “Decreases in self-reported sleep duration among U.S. adolescents 2009-2015 and association with new media screen time”, Sleep Medicine, Vol. 39, pp. 47-53.
Twenge, M.J., Cooper, A.B., Joiner, E.T., Duffy, M.E. and Binau S.G. (2019), “Age, period, and cohort trends in mood disorder indicators and suicide-related outcomes in a nationally representative dataset, 2005–2017”, Journal of Abnormal Psychology, Vol. 128 No. 3, pp. 185-199.
Further reading
Agresti, A. (2006), An Introduction to Categorical Data Analysis, John Wiley & Sons.