Over the past 40 to 50 years, researchers have investigated the relationship between childhood ADHD and a number of demographic factors, including race/ethnicity and socioeconomic status (SES). Although there is some evidence to suggest that the rate of ADHD diagnosis, treatment and access to care may be affected by these demographic variables, studies have yielded inconsistent results and are often based only on parental report. To address these limitations, this study used broader ADHD diagnostic determination, including case-definition diagnoses (i.e. diagnoses determined based on structured data from both parents and teachers), to examine these relationships in a multisite elementary-school–based sample (Bax et al. 2019).
This secondary analysis was based on data from the ADHD Prevalence Study and the Tics/ADHD Prevalence Study, which were two separate but related studies that determined the prevalence of ADHD and/or tic disorders in children aged 4–12 years.* In both studies, a two-stage sampling design was employed, which first recruited teachers and later students and their parents/caregivers for sampling. Sample design weights for each child were then constructed for both teacher and parent/caregiver samples. The analysis considered a number of measures, including race/ethnicity and SES risk factor variables, and ADHD diagnosis, treatment, case-definition and access to care outcome variables. The study hypotheses were investigated through comparison of simple bivariate and multipredictor logistic regression coefficients within and across three outcomes: parent-reported diagnosis, treatment with medication, and diagnoses meeting ADHD study case definitions. Results from the study are summarised below.
- 51.9% of the study participants were male, 51.3% were non-Hispanic white, and 46.0% were between 8 and 10 years old.
- 53.1% of participants had private health insurance, yet 53.9% qualified for free or reduced school lunches.
- The majority of primary parents/caregivers reported attending some college or graduating from college (72.2%); however, more than half of families had an annual income of less than $45,000 (60.4%).
- Most study participants lived in an urban setting (78.8%) and in two-parent households (72.6%).
- Of the total sample, 10% of participants had a parent-reported ADHD diagnosis, 8.8% had used ADHD medication in the previous 4 weeks, and 8.3% met the study’s case-definition criteria for ADHD.†
Parent-reported ADHD diagnosis outcome
- Simple bivariate regression demonstrated that with the exception of the No Insurance indicator, all SES indicators were significantly associated with children having a parent-reported clinical ADHD diagnosis:
- Binned Income: Odds ratio (OR) 0.92 (95% confidence interval [CI] 0.87–0.98; p < 0.01)
- Reduced Lunch: OR 1.98 (95% CI 1.15–3.41; p < 0.05)
- Primary Parent/Caregiver Education: OR 0.83 (95% CI 0.70–0.98; p < 0.05)
- Medicaid Recipient: OR 2.32 (95% CI 1.34–4.01; p < 0.01)
- None of the SES indicators reached significance when all five variables were included simultaneously or when controlling for all covariates (i.e. SES, race/ethnicity, other).
- The multivariable model demonstrated that, in comparison to the referent white group, all other race/ethnicity groups individually evidenced lower odds of having a parent-reported ADHD diagnosis (p < 0.05), with the exception of the American Indian indicator.
- In the multivariable model, males and children living in a single-parent/caregiver household were also significantly more likely to have a parent-reported diagnosis (males: OR 4.34 [95% CI 2.53–7.46; p < 0.001]; single-parent/caregiver household: OR 1.93 [95% CI 1.10–3.40; p < 0.05]).
Treatment with medication outcome
- Simple bivariate analysis demonstrated that only the Medicaid Recipient (OR 2.07 [95% CI 1.19–3.60; p < 0.01]) and No Insurance (OR 0.08 [95% CI 0.02–0.36; p < 0.001]) indicators were significantly associated with medication use.
- These effects remained significant when controlling for all other covariates (Medicaid Recipient: OR 2.16 [95% CI 1.07–4.35; p < 0.05]; No Insurance: OR 0.10 [95% CI 0.02–0.52; p < 0.01]).
- In the multivariable model, the race/ethnicity effect for ADHD medication suggested lower rates of use for non-white groups of children, and differences between the white group and both the black and other groups reached statistical significance (p < 0.05).
- The multivariable model also demonstrated that males were significantly more likely to be using medication (OR 4.60 [95% CI 2.80–7.57; p < 0.001]).
ADHD case-definition outcome
- Similar to the parent-reported ADHD diagnosis model, bivariate regression demonstrated that all SES indicators except No Insurance were predictive of the case-definition outcome:
- Binned Income: OR 0.91 (95% CI 0.87–0.96; p < 0.001)
- Reduced Lunch: OR 1.99 (95% CI 1.31–3.02; p < 0.001)
- Primary Parent/Caregiver Education: OR 0.85 (95% CI 0.75–0.97; p < 0.05)
- Medicaid Recipient: OR 2.18 (95% CI 1.40–3.40; p < 0.01)
- None of the SES indicators remained significant when all five covariates were considered simultaneously or in models that controlled for all covariates.
- The race/ethnicity variable was also less predictive of the case-definition outcome and did not reach significance in the bivariate or covariate model.
- The multivariable model demonstrated that, compared with the referent white group, Hispanic participants were significantly less likely to meet case-definition criteria (OR 0.31 [95% CI 0.12–0.79; p < 0.05]).
- In the multivariable model, males and children living in a single-parent/caregiver household were also significantly more likely to meet ADHD case-definition criteria (males: OR 2.78 [95% CI 1.72–4.50; p < 0.001]; single-parent/caregiver household: OR 1.86 [95% CI 1.07–3.23; p < 0.05]).
This analysis is one of the few studies to have investigated ADHD-associated risk factors using a rigorous ADHD case definition determined through analysis of Diagnostic and Statistical Manual of Mental Disorders (DSM) symptom criteria that is based upon both parent and teacher ratings. Despite this, the study was limited by several factors. First, although data collection procedures were designed to facilitate the development of an ADHD case definition, they do not technically reflect actual clinically derived diagnoses. Parent interviews were also conducted approximately 1 year after the teacher reports were obtained, which may weaken the validity of the case definition. Additionally, the study only evaluated a select number of SES and race/ethnicity covariates, meaning that other potential confounders not controlled for in the study may have impacted results. Finally, there were small sample sizes in the Hispanic and American Indian groups, meaning the null findings for these groups should be interpreted with caution.
The authors concluded by stating that children who are white, male, have health insurance (particularly Medicaid) and live in single-parent/caregiver households are more likely to have an ADHD diagnosis and/or medication treatment compared with their counterparts. Furthermore, children being raised by a single parent/caregiver were also at a higher risk of a false-positive diagnosis, which requires further investigation. The authors stated that these findings may facilitate an improved understanding of the complex interaction between sociodemographic variables and ADHD diagnosis and treatment, which may aid clinicians in providing more appropriate care.
*The ADHD Prevalence Study was a Centres for Disease Control and Prevention (CDC)-funded project involving two sites, including Oklahoma (three school district samples) and South Carolina (one school district sample). The Tics/ADHD Prevalence study was a second CDC-funded project comprising one site in Oklahoma (two school district samples), which included the use of a tic screener following the ADHD screener during the first phase of sample recruitment; the tic data were not factored into this analysis
†Of the total sample, 5.7% of participants were taking ADHD medication but did not meet the study’s case-definition criteria for ADHD
Bax AC, Bard DE, Cuffee SP, et al. The association between race/ethnicity and socioeconomic factors and the diagnosis and treatment of children with attention-deficit hyperactivity disorder. J Dev Behav Pediatr 2019; 40: 81-91.