The Relationship between Pediatric Attention-Deficit/ Hyperactivity Disorder Symptoms and Asthma Management

Abstract

Purpose:

Children with co-morbid Attention-Deficit/ Hyperactivity Disorder (ADHD) and asthma are at an increased risk for adverse health outcomes and reduced quality of life. The objective of these analyses was to examine if self-reported ADHD symptoms in children with asthma are associated with asthma control, asthma controller medication adherence, quick relief medication use, pulmonary function, and acute healthcare utilization.

Methods:

We analyzed data from a larger study testing a behavioral intervention for Black and Latinx children with asthma ages 10–17 and their caregivers. Participants completed the Conners-3AI self-report assessment for ADHD symptoms. Asthma medication usage data were collected for 3 weeks following baseline via electronic devices fitted to participants’ asthma medications. Other outcome measures included the Asthma Control Test, self-reported healthcare utilization, and pulmonary function measured by spirometry testing.

Results:

The study sample consisted of 302 pediatric participants with an average age of 12.8 years. Increased ADHD symptoms were directly associated with reduced adherence to controller medications, but no evidence of mediation was observed. Direct effects of ADHD symptoms on quick-relief medication use, health care utilization, asthma control, or pulmonary function were not observed. However, the effect of ADHD symptoms on ER visits was mediated by controller medication adherence.

Conclusions:

ADHD symptoms were associated with significantly reduced asthma controller medication adherence and indirectly with ER visits. There are significant potential clinical implications to these findings, including the need for the development of interventions for pediatric asthma patients with ADHD.

Asthma is a chronic condition that is characterized by airway inflammation and symptoms including wheezing, chest tightness, coughing, and shortness of breath(1). As of 2018 there were 24.8 million people (7.7%) diagnosed with asthma in the United States (US), with children ages 5–17 years having the highest prevalence rate of all age groups (8.9%)(2). With proper treatment and management, asthma can be well-controlled and allow for participation in daily activities with minimal restriction(1). Yet there are high rates of asthma-related ambulatory care visits and hospitalizations in the US, indicating that many children suffer from asthma that is not well controlled(3). Poorly controlled pediatric asthma places a significant economic, health, and quality of life burden on affected families and society at large(4). Children from minoritized populations are at a particularly high risk for poorly controlled asthma and adverse asthma outcomes(5), with Black and Latinx children having high rates of asthma-related hospitalizations and emergency department visits in the US(3).

Poor pediatric asthma control has been attributed to numerous behavioral and psychological factors, including the high rate of non-adherence to daily asthma controller medications such as inhaled corticosteroids (ICS) (6). Besides worse asthma control, poor medication adherence is associated with decreased quality of life, increased health care utilization, and increased risk for mortality(6). Poor adherence to daily controller medication has also been associated with the over-use of quick-relief (QR) medications(7). Overuse of QR medications over time can increase asthma-associated risks, as the medication may be less effective when there is a life-threatening emergency(8). Accordingly, it is critical to examine the factors that may be contributing to both poor asthma controller medication adherence and the overuse of QR medications in pediatric populations.

Attention-Deficit/ Hyperactivity Disorder (ADHD) is a neurodevelopmental disorder characterized by chronic patterns of inattention and/or hyperactivity that interfere with a child’s daily functioning(9). ADHD is the most common behavioral condition and the second most common chronic illness in children in the United States(10), with 6.1 million children ages 2–17 having an ADHD diagnosis(11). Rates of ADHD have been found to be even higher in low-income urban teenagers with asthma, with a recent school-based survey finding that 28% of this population met criteria for ADHD(12). Comorbid asthma and ADHD have been associated with reduced pediatric quality of life(13) and increased asthma severity(14). Given that ADHD symptoms often include executive functioning deficits such as disorganization, impulsivity, and difficulty planning, it has been hypothesized that these tendencies may impact adherence to ADHD stimulant medications(15). Whether ADHD symptoms may thereby also contribute to poor asthma controller medication adherence is not known.

In these analyses, we examined the direct relationship between ADHD symptoms with pediatric asthma controller medication adherence in Black and Latinx children, hypothesizing that increased ADHD symptoms would be associated with poor medication adherence. We also investigated the direct effects of ADHD symptoms on asthma control, pulmonary function, QR medication use, and acute healthcare utilization (i.e., ER and Urgent Care visits, hospitalizations). We hypothesized that children with increased ADHD symptoms would have worse asthma control and pulmonary function, be more likely to use their QR medications, and have increased acute healthcare utilization. We also hypothesized that controller medication adherence would mediate the relationship between ADHD symptoms and these outcomes.

Methods

Design and Participants

This was a cross-sectional secondary analysis of data from a randomized controlled trial (NCT #02702687) examining the effect of peak flow prediction with feedback on asthma symptom perception, medication adherence, asthma control, and health care utilization in Black and Latinx children ages 10–17 years. Study participants were recruited from pediatric primary care and asthma clinics in the Bronx, NY. Participant inclusion criteria were having a physician documented diagnosis of asthma and daily asthma controller medication prescription, experiencing breathing problems in the past year, and being English and/or Spanish-speaking. Additionally, caregivers participating in the study needed to have primary or equal responsibility for the child and live with them. Exclusion criteria included the caregiver or child having a cognitive disability that would hinder their capacity to complete the self-report measures, and the child having a pulmonary condition aside from asthma.

Procedures

The study protocol was approved by the Albert Einstein College of Medicine Institutional Review Board. Data were collected between August 2016 and September 2020. Following recruitment and screening, study participants and their caregivers provided assent and informed consent, respectively, for their study participation. At the initial study session, trained research assistants facilitated the completion of all study measures electronically using MediaLabv2016 software in English or Spanish based on participant preference. Pulmonary function testing was also conducted using a handheld spirometer that accompanies a computerized program (nSpire, Longmont, CO), and study participants were provided with medication adherence monitoring devices. Approximately 3 weeks later, participants returned for a follow-up appointment and the data from their medication adherence monitoring devices were downloaded. Starting in Mid-March 2020 during the COVID-19 lockdown, participants received prepaid envelopes to mail their adherence monitoring devices back for data download since all study sessions switched to a virtual format. Days that the devices were in transit were not counted for adherence calculations. All data used in these analyses were collected prior to participants being randomized into the behavioral intervention to avoid any influence of the intervention.

Measures

Caregivers completed a demographics questionnaire to provide information on their age, sex, and education level. Caregivers also reported their total monthly family income and household composition, to categorize families as living above or below the federal poverty line. Self-report of child ethnicity was measured using a questionnaire that was designed for guiding children in labelling their racial and ethnic identities(16).

Conners-3 ADHD Index Abbreviated (Conners 3-AI):

The Conners-3AI child version was used to measure pediatric ADHD symptoms. The Conners-3AI is a validated, 10-item self-report questionnaire that distinguishes between children with and without ADHD(17). Each item in the questionnaire provides a statement corresponding to an ADHD symptom, and the child is prompted to signify how true that statement has been about them in the past month. Examples of items include “I have trouble keeping myself organized”, “I have too much energy to stay still”, and “I have trouble following instructions”(17) to assess ADHD symptoms such as disorganization, hyperactivity, and difficulties focusing on tasks, respectively. The Conners-3AI raw scores are recoded and converted to T-scores (range 0–90) based on child age, to determine if it is clinically significant for a diagnosis of ADHD. The Conners-3AI has strong psychometric properties including excellent discriminative validity, predictive validity, construct validity, internal consistency, and reliability(17).

Asthma Controller Medication Adherence:

ICS controller medication adherence data were collected via Doser (Meditrack, Wellesley, MA) and Smartinhaler (Hailie, Auckland, New Zealand) devices fitted to the child’s asthma medications. Dosers are devices that are attached to the top of a metered dose inhaler (MDI) medication to track its use for 30 days. Dosers have been found to be dependable for tracking inhaled asthma medication use in clinical research, and more accurate as compared to other means such as self-report and pharmacy records(18). For the MDI medications that Dosers did not fit onto (i.e., Budesonide/Formoterol), Smartinhalers were utilized to collect medication adherence data. Smartinhalers are electronic monitors that reliably track how often an inhaled asthma medication is used, with storage for over 6,000 data points(19). To monitor the use of Leukotriene Receptor Antagonist (LTRA) medications, the Medication Events Monitoring System (MEMS-6) TrackCaps (AARDEX, Switzerland) were utilized. TrackCaps are widely used electronic medication monitors that attach to pill bottles and record each date and time the bottle is opened and closed(20).

Data were cleaned to remove any extra puffs or bottle-openings that would artificially inflate adherence. Following data cleaning and reduction procedures(21), medication adherence was calculated as a percentage by dividing the number of total doses taken by the number of prescribed doses.

QR Medication Use:

The use of QR medications was also tracked via Dosers that were fitted to participants’ MDIs and collected data for a 30-day period. Overall QR medication use was defined as the frequency of medication administration within the observed period, calculated by dividing the number of days of QR medication use by the number of days of data available (typically 30) and then converting it to a percent value. The data were non-normally distributed, so it was dichotomously split into use ≤ twice a week and use > twice a week based on the National Heart, Lung, and Blood Institute guidelines for asthma control(22). Thereby if the frequency of QR medication use over the recorded 30-day period was greater than 28.6% it was coded as QR use indicating poorly controlled asthma (high-level use), and any frequency lower than that threshold was coded as QR use indicating well-controlled asthma (low-level use)(22).

Healthcare Utilization:

Caregiver report questions were utilized to measure the child’s healthcare utilization (HCU) associated with asthma over the past year. HCU was assessed in 3 domains: hospitalizations, emergency room (ER) visits, and Urgent Care visits in the last 12 months. Due to the highly skewed distribution of these data, hospitalizations and ER visits were dichotomously coded as either absent (no visits) or present (1 or more visits) for the past year. Urgent Care visits were dichotomously coded as less than two visits and two or more visits per year.

Asthma Control:

The Asthma Control Test (ACT) was used to measure asthma control (23) for pediatric participants ages 12–17 years old. The Childhood Asthma Control Test (C-ACT)(24) was used for participants ages 10–11 years old. Both measures have strong reliability and proven validity in categorizing children and adolescents as having well-controlled or poorly controlled asthma(23, 24). For Spanish-speaking study participants, the Spanish ACT or Spanish C-ACT was used, both of which also have strong reliability(25) and predictive validity(26). These questionnaires assess the frequency and intensity of the child’s asthma symptoms, their use of their quick-relief medications, and how asthma has impacted their daily life over the past month. The C-ACT is completed collaboratively by the child and their caregiver, with the child completing the first 4 questions and their caregiver completing the last 3 questions. The ACT consists of 5 questions that are completed independently by the adolescent (23). The total score of the C-ACT or ACT measure was computed, with a score from 0–18 indicating the child had poor asthma control and a score from 19–27 indicating the child had well-controlled asthma (24). Based on this clinical cutoff, the total scores were then dichotomized to indicate the child either had self-reported well-controlled or poorly controlled asthma.

Pulmonary Function:

Spirometry testing was used to assess pulmonary function as an objective measure of asthma control. Pulmonary function testing was conducted using a handheld spirometer (nSpire, Longmont, CO). All spirometry testing procedures and equipment met or exceeded the American Thoracic Society standards, including having at least three forced expiratory maneuvers for acceptable results to be achieved (27),(28). The current analyses used percent predicted Forced Expiratory Volume in one second (FEV₁) and the FEV₁/Forced Vital Capacity ratio (FEV₁ /FVC) as measures of pulmonary function. FEV₁ is often used as a strong barometer of lung function, and the FEV₁ /FVC ratio is a key indicator for assessing asthma control (28). The percent predicted FEV₁ and FEV₁ /FVC ratio were computed using the National Health and Nutrition Examination Survey III equation (29) for predicted values based on the child’s sex, age, height, and race.

Analyses

Descriptive statistics were run to characterize the sample and examine distributional characteristics of the data. Structural equation models (SEM) were run to examine the direct and indirect effects of ADHD, asthma severity, and controller medication adherence on the endogenous variables of asthma control, pulmonary function (FEV₁ and FEV₁/FVC), acute healthcare utilization (ER, urgent care visits and hospitalizations), and quick relief medication use. SEM allows for missing data on the endogenous variable using full maximum likelihood methods which assume that the data are missing at random. A latent variable for controller medication adherence was comprised of the observed variables for the objectively measured ICS and LTRA medication use. Exogenous variables for the direct effects on the latent controller medication adherence variable included child’s age, sex, and race/ethnicity. Figure 1 illustrates the SEM model for ADHD on controller medication adherence and the endogenous variables. Separate models were run for each endogenous variable.

Figure 1.

Structural Equation Model for Asthma Control, Pulmonary Function, Quick Relief Medication Use, and Acute Healthcare Utilization

The following direct effects were specified for each model: child’s age, sex, ethnicity, asthma severity, and ADHD symptoms on controller medication adherence; and asthma severity, ADHD symptoms, and controller medication adherence on each endogenous variable (i.e., asthma control, FEV₁ % predicted, FEV₁/FVC, ER visits, urgent care visits, hospitalizations, and quick relief medication use). The following indirect effects were specified: ADHD symptoms through controller medication adherence on each endogenous variable listed above and asthma severity through controller medication adherence on each endogenous variable. Standardized coefficients and standard errors were examined and model fit indices computed using maximum likelihood estimation. AIC and BIC model selection were used to distinguish among a set of possible models examining the relationships described above. The best-fitting model was selected based on the lowest AIC and BIC. Following conventions outlined by Kline (30), several model fit indices are reported: χ²/df ratio <2, comparative fit index (CFI), Tucker–Lewis Index (TLI) ≥.90, Root Mean Square Error of Approximation ≤.05, and standardized root mean square residual (SRMR) <.08. SAS version 9.4 of the SAS System for Windows was used for the descriptive analyses and MPlus version 8.8 was used for the SEM analyses.

Results

A total of 302 parent-child dyads participated in the current study. Descriptive and clinical characteristics of participants are included in Table 1. The majority of the pediatric participants were male (53.8%), with an average age of 12.8 years. The self-identified ethnicities of the pediatric participants included Puerto Rican (32.8%), Black (33.1%), Dominican (13.8%), and other Latinx (20.3%). The average caregiver age was 42.08 years, with the majority of caregivers self-identifying as either Puerto Rican (33.9%) or African American (23.7%). Most of the caregivers reported having a household income below the federal poverty threshold (72%). The mean T-score on the Conners-3AI measure of ADHD symptoms was in the high average range (Mean T = 62.8, SD = 1.45) (Table 2). In terms of medication use, the mean ICS adherence rate in the study sample was 37.9% and the mean LTRA adherence rate was 47.5%. The majority of the study participants (73.5%) were classified as having low level use of their QR medications (less than two days per week on average) (Table 2). For acute healthcare utilization over the past year, ER visits (57.4%) and urgent care visits (49.2%) related to asthma were fairly common amongst the study population, while a smaller portion of participants reported asthma-related hospitalizations (19.7%) (Table 2).

Table 1.

Sociodemographic Characteristics of Study Sample

Variable	M(SD) or N(%) or Median [IQR]
Age	12.8 [4.0]
Sex
Male	162 (53.8%)
Female	139 (46.2 %)
Race
White	93 (39.1%)
Black	108 (46.0%)
Mixed	35 (14.9%)
Ethnicity
Puerto Rican	95 (32.8%)
African American	96 (33.1%)
Dominican	40 (13.8%)
Other	59 (20.3%)
Caregiver Age	42.08 [10.81]
Poverty
< Poverty Threshold	170 (72.0%)
> Poverty Threshold	66 (28.0%)
Caregiver Education
Some HS	65 (28.1%)
HS Diploma	52 (22.4%)
Some College	62 (26.7%)
College/Grad	53 (22.8%)

Note. N= 302; Other= Self-described ethnic background from European, Asian, Middle Eastern, African, Caribbean, North, South, or Central American, and/or mixed descent not fitting into the provided categories. Poverty calculated via caregiver report of monthly income in $1,000 increments and number of individuals living in household to categorize families as living below or above the federal poverty line. Some HS= Some High School, HS Diploma= High School Diploma, College/Grad= 4-year college degree and/or graduate degree.

Table 2.

Clinical Characteristics of Study Sample

Variable	M(SD) or N(%) or Median [IQR]
Conners-3AI T-score	62.4 (15.92)
ER Visits
None	127 (42.6%)
1 or More	144 (57.4%)
Sick Visits
Less than 2	153 (50.8%)
2 or More	148 (49.2%)
Hospitalizations
None	187 (80.3%)
1 or More	46 (19.7%)
Quick-Relief Use
Low Level	155 (73.5%)
High Level	56 (26.5%)
ICS Adherence %	37.9 (28.79)
LTRA Adherence %	47.47 (32.1)
Asthma Control
Poorly controlled	120 (50.6%)
Well controlled	117 (49.4%)
% Predicted FEV1	93.71 (11.7)
FEV1/FVC	82.5 (11.12)

Note. N= 302 Quick-Relief Use Low Level= Less than two days a week, High Level= Two days a week or more. ICS= Inhaled Corticosteroids, LTRA= Leukotriene Receptor Antagonist. Asthma control rating based on results of ACT or C-ACT (based on age). FEV₁= Forced Expiratory Volume in one Second, FEV₁/FVC= ratio of FEV₁ and Forced Vital Capacity.

Participants with increased ADHD symptoms had significantly reduced adherence to asthma controller medications in the models for asthma control, pulmonary function, and QR medication use (Table 3). Also, higher levels of asthma severity were associated with better controller medication adherence in the models examining the above endogenous variables. No direct effects were found for ADHD symptoms on self-reported asthma control, FEV₁% predicted, FEV₁/FVC, or quick-relief medication use. Higher levels of asthma severity were associated with worse asthma control, lower FEV₁% predicted and FEV₁/FVC, and increased quick-relief medication use (β = 0.34, p < .0001). Greater controller medication adherence was associated with higher FEV₁/FVC, but no associations were found with asthma control, FEV₁% predicted, or quick relief medication use. Adherence to controller medications did not mediate any of the relationships between ADHD symptoms or asthma severity and asthma control, pulmonary function and QR medication use.

Table 3.

Structural Equation Model for Asthma Control, Pulmonary Function, and Quick-Relief Medication Use

Variable	Asthma Control		FEV1 % Predicted		FEV1/FVC		QRM Use
Model R 2	0.25	.005	.21	.008	.18	.02	.28	.003
Model Fit	X2/df=1.82 RMSEA=.06 CFI=.89 TLI=.78 SRMR=.07		X2/df=0.77 RMSEA=0 CFI=1.00 TLI=1.00 SRMR=.04		X2/df=2.12 RMSEA=.07 CFI=.86 TLI=.72 SRMR=.06		X2/df=1.11 RMSEA=.02 CFI=.98 TLI=.96 SRMR=.05
Controller Medication Adherence Direct Effects:	β (SE)	p-Value	β (SE)	p-Value	β (SE)	p-Value	β (SE)	p-Value
Child Age	−0.22 (.09)	.02	−0.21 (.08)	.01	−0.21 (.08)	.008	−0.22 (.10)	.02
Child Sex	−0.07 (.09)	.46	−0.13 (.09)	.13	−0.17 (.08)	.03	−0.10 (.09)	.29
Child Ethnicity	−0.29 (.09)	.001	−0.28 (.09)	.002	−0.21 (.09)	.01	−0.33 (.09)	<.0001
Asthma Severity	0.20 (.09)	.03	0.19 (.09)	.04	0.20 (.09)	.02	0.20 (.09)	.03
Conners T-score	−0.27 (.08)	.001	−0.22 (.09)	.01	−0.18 (.08)	.03	−0.28 (.09)	.001
Model R 2	.24	.005	.06	.08	.09	.03	.15	.02
Dependent Variable Direct Effects:
Controller Medication Adherence	0.08 (.10)	.44	0.14 (.09)	.11	0.21 (.08)	.01	0.11 (.14)	.43
Asthma Severity	−0.49 (.07)	<.0001	−0.23 (.07)	.001	−0.27 (.07)	<.0001	0.34 (.09)	<.0001
Conners T-score	0.03 (.06)	.67	−0.02 (.07)	.72	0.09 (.07)	.17	0.03 (.07)	.68
Indirect effect:
Conners T-score Adherence Dependent Variable	−0.04 (.03)	.16	−0.03 (.02)	.17	−0.04 (.02)	.08	−0.03 (.04)	.45
Asthma Severity Adherence Dependent Variable	0.03 (.03)	.22	0.03 (.02)	.20	0.04 (.03)	.11	0.02 (.03)	.45

Table 4 reveals that children with greater ADHD symptoms had significantly reduced adherence to asthma controller medications in the models for acute healthcare utilization (i.e., ER and urgent care visits and hospitalizations). Similar to the results above, children with more severe asthma reported increased controller medication adherence in the healthcare utilization models. There were no direct effects observed for ADHD symptoms and asthma severity on each of the healthcare utilization variables. There was a direct effect of controller medication adherence on ER visits and urgent care visits but not hospitalizations. Controller medication adherence mediated the relationship between ADHD symptoms and ER visits. No other indirect effects were significant in the healthcare utilization models.

Table 4.

Structural Equation Model for Acute Healthcare Utilization

Variable	ER Visits (Dichotomous)		Hospitalizations (Dichotomous)		Urgent Care Visits (Dichotomous)
Model R 2	.75	<.0001	.28	.003	.34	.001
Model Fit	X2/df=1.08 RMSEA=.03 CFI=.98 TLI=.94 SRMR=.09		X2/df=1.06 RMSEA=.02 CFI=.99 TLI=.97 SRMR=.09		X2/df=1.28 RMSEA=.03 CFI=.94 TLI=.89 SRMR=.07
Controller Medication Adherence Direct Effects:	β (SE)	p-Value	β (SE)	p-Value	β (SE)	p-Value
Child Age	−0.61 (.18)	.001	−0.24 (.10)	.01	−0.26 (.10)	.008
Child Sex	−0.21 (.13)	.12	−0.11 (.09)	.23	−0.10 (.09)	.30
Child Ethnicity	−0.47 (.15)	.002	−0.31 (.09)	.001	−0.37 (.10)	<.0001
Asthma Severity	0.45 (.21)	.03	0.21 (.09)	.02	0.22 (.10)	.03
Conners T-score	−0.59 (.20)	.003	−0.28 (.09)	.001	−0.30 (.09)	.001
Model R 2	.24	.05	.07	.24	.09	.19
Dependent Variable Direct Effects:
Controller Medication Adherence	0.46 (.15)	.002	0.19 (.13)	.15	0.28 (.13)	.03
Asthma Severity	0.05 (.14)	.70	0.14 (.10)	.17	0.07 (.09)	.45
Conners T-score	0.10 (.10)	.35	0.04 (.07)	.61	0.12 (.07)	.08
Indirect effect:
Conners T-score Adherence Dependent Variable	−0.27 (.11)	.02	−0.05 (.04)	.20	−0.08 (.05)	.08
Asthma Severity Adherence Dependent Variable	0.21 (.11)	.07	0.04 (.03)	.22	0.06 (.04)	.12

Discussion

The current study results support the hypothesis that ADHD symptoms are inversely associated with asthma controller medication adherence, as study participants with increased ADHD symptoms demonstrated reduced adherence to their ICS and/or LTRA medications. However, no associations were found between ADHD symptoms and the frequency of QR medication use, self-reported asthma control, pulmonary function, or acute asthma-related healthcare utilization. Asthma controller medication adherence was also not found to serve as a mediator in relation to ADHD and these asthma outcomes with the exception of ER visits. These findings have important clinical implications for providers. While there are substantial data indicating a strong relationship between the development of ADHD and asthma in children (31), there is minimal exploration in the literature on the influence of ADHD symptoms on asthma management. The current findings suggest that healthcare providers should closely support and monitor asthma controller medication adherence in pediatric patients with ADHD.

One possible interpretation of the results is that familial dynamics may play a role in asthma self-management strategies amongst children and adolescents with ADHD symptoms. Prior studies have indicated that pediatric ADHD symptoms are associated with difficulties in self-reported family asthma management, which partially mediates asthma symptoms and morbidity (32). It has also been demonstrated that caregivers of adolescents with ADHD regard them as less responsible for their asthma management than the caregivers of adolescents with asthma-only do (12). Accordingly, adolescents with ADHD may need increased support from caregivers and providers when transitioning to self-management of their asthma as they enter young adulthood (12).

Furthermore, it is possible that the deficits associated with ADHD symptoms may also hinder other aspects of asthma self-management aside from the observed reduced medication adherence, such as environmental trigger avoidance, although this has not been examined in the literature. There is evidence from other disease models that the developmental challenges associated with ADHD may contribute to poor chronic disease self-management (33). In a study examining the associations between ADHD and diabetes self-management, it was noted that the participants with ADHD struggled with sustaining their attention, organizing themselves, and self-monitoring, all of which contributed to their difficulties in maintaining the consistent behaviors necessary to manage their chronic disease (33). Notably, these challenges were found to peak in adolescence due to the changing environmental, social, and hormonal factors during this developmental stage (33). It is likely that such factors may also contribute to poor self-management behaviors in adolescents with asthma, such as the observed association between increased ADHD symptoms and poor controller medication adherence.

Additionally, controller medication adherence was found to be positively associated with Urgent Care and ER Visits and to mediate the relationship between ADHD symptoms and ER Visits. Given that healthcare utilization reporting was retrospective, this may be explained by severe asthma exacerbations leading to subsequent increased controller medication adherence. Thus in reaction to exacerbations requiring acute healthcare utilization individuals may be increasingly likely to utilize preventative medications to avoid future flareups. However, the timing of these measures is a limitation of the current study so we are cautious in our interpretation of this finding.

While an association between pediatric ADHD symptoms and asthma controller medication adherence was established, no relationship was found between ADHD and QR medication use. Despite our hypothesis that ADHD symptoms, such as impulsivity, may promote increased QR medication use, it is certainly possible that children with ADHD are less engaged overall in their asthma management and thereby do not frequently use their controller or QR medications. This notion is indirectly supported by the evidence that children with attention deficits (as assessed by the Continuous Performance Task) are worse at perceiving their asthma symptoms accurately (34). Accordingly, it is possible that the absence of an association between ADHD symptoms and QR medication use in the current analyses may be attributed to a lack of recognition of asthma exacerbations requiring the use of QR medications.

The current analyses have several limitations that should be addressed in future investigations. Since cross-sectional data were utilized, the causal mechanism underlying the relationship between ADHD symptoms and asthma controller medication adherence that was identified could not be established. Also, many of the current study measures were self-report and retrospective which brings the potential for bias in responding (i.e., minimization or forgetfulness) and generates difficulty in making diagnostic determinations.

Future research may benefit from the use of additional diagnostic assessments such as the Structured Clinical Interview for the DSM-5 (SCID-5), behavioral assessments (i.e., the Continuous Performance Test), and the Parent and Teacher forms of the Conners-3AI to determine ADHD diagnoses more definitively. Additionally, prospective investigations should utilize objective measures of asthma outcomes such as healthcare utilization (i.e., a review of electronic medical records) as opposed to retrospective self-report. Future studies may also expand on the current design by examining these variables longitudinally and exploring how ADHD is associated with additional aspects of asthma management such as environmental avoidance of triggers.

In conclusion, these analyses demonstrated that ADHD symptoms in Black and Latinx children and adolescents were associated with reduced asthma controller medication adherence. This finding extends the existing literature on asthma and ADHD, by establishing that ADHD symptoms are associated with poor pediatric asthma management practices. Physicians should be aware of this relationship between pediatric ADHD and asthma so they can adjust their assessment and treatment of patients accordingly. Further understanding the relationship between these diseases is critical to improving pediatric asthma self-management and reducing the burden poorly controlled asthma places on minority populations.

Implications and Contributions.

This paper demonstrates that increased ADHD symptoms in a sample of minority adolescents were associated with reduced asthma controller medication adherence. This novel finding extends the field’s limited understanding of the implications of comorbid ADHD and asthma in youth. Providers should adjust their assessment and treatment of patients accordingly.

Abbreviations:

ADHD

Attention-Deficit Hyperactivity Disorder

US

United States

ICS

Inhaled Corticosteroids

QR

Quick-Relief

MDI

Metered Dose Inhaler

LTRA

Leukotriene Receptor Antagonist

HCU

Healthcare Utilization

ER

Emergency Room

ACT

Asthma Control Test

C-ACT

Childhood Asthma Control Test

FEV₁

Forced Expiratory Volume in One Second

FVC

Forced Vital Capacity