The Impact of Cumulative Stress on Asthma Outcomes Among Urban Adolescents

Abstract

Introduction:

Many families living in low-income, urban areas experience a number of stressors (e.g., poverty, neighborhood stress, family functioning) that place adolescents at risk for worse asthma outcomes. Adolescents may face additional challenges (e.g., peer pressure, school stress) that add to their overall stress and influence their disease care and health outcomes. The current study examined the impact of a cumulative risk model of stressors including poverty, neighborhood stress, school stress, peer pressure, and caregiver-adolescent conflict on asthma outcomes (e.g., emergency department [ED] visits, asthma control, quality of life [QOL]) among urban adolescents (13–17 years).

Methods:

Data were collected from 61 urban families of adolescents with asthma (54.1% female; 93.4% African American) in the United States. Caregivers and adolescents completed questionnaires assessing stressors and asthma outcomes separately during a research session.

Results:

Cumulative risk was significantly associated with worse adolescent QOL and asthma control, and more ED visits. The cumulative risk index was also a more robust predictor of QOL and asthma control than any one individual predictor. Poverty, neighborhood stress, and school stress emerged as individual predictors of ED visits. Further, adolescents with well-controlled asthma had significantly lower neighborhood and school-related stress scores.

Conclusions:

Findings suggest that beyond the risk conferred by individual risk factors, an accumulation of stress can have an especially negative impact on asthma outcomes for urban adolescents. Future intervention work aimed at improving asthma outcomes should consider incorporating strategies for minimizing overlapping sources of stress in adolescents’ daily lives.

Although overall pediatric asthma rates are plateauing, rates among adolescents and children living in poverty continue to rise (Akinbami, Simon, & Rossen, 2016). Further, racial disparities in pediatric asthma are well documented, with prevalence, emergency department (ED) visits, hospitalizations, and death rates from asthma higher among minority children (Akinbami, Moorman, Simon, & Schoendorf, 2014). For youth with asthma, adolescence is a period in which many begin to take over management of their illness. Additionally, adolescents from low-income urban areas are more likely to be exposed to chronic stress (e.g., poverty, community violence, family instability), which is a key mechanism in disease progression and management (Yonas, Lange, & Celedon, 2012). Despite focused efforts to reduce asthma morbidities, adolescents still experience frequent asthma attacks and asthma-related ED use remains high (Center of Disease Control [CDC], 2006). Thus, existing research suggests that factors outside routine medical care, such as chronic stress, impact daily asthma management and related outcomes.

Adolescents with Asthma

Adolescents with asthma face a unique set of stressors that may place them at higher risk for asthma morbidity (Bitsko, Everhart, & Rubin, 2014) and are an understudied group in pediatric asthma research. Rates of psychopathology increase in adolescence along with the struggle to balance family, peer, and academic demands; moreover, adolescence is a critical developmental period in which children increase autonomy and become active participants in daily asthma management care (Bitsko et al., 2014). Theoretically, stress (e.g., peer-related stress; Sadof & Kaslovsky, 2011) undermines health in part because it fosters non-compliance of asthma management. Thus, understanding the role of cumulative stress on disease outcomes in this critical period of development is an important first step toward informing tailored inventions, with long-term implications for the development of healthy disease management routines.

Additionally, stressful life events in adolescence have been related to newly diagnosed cases of asthma in young adulthood (Oren, Gerald, Stern, Martinez, & Wright, 2017), suggesting long-term effects of stress experienced in adolescence on asthma outcomes across later developmental periods. In adolescence, individuals make gains in cognitive development, abstract reasoning, and information processing; therefore, adolescents may be better able to understand limitations due to their disease and experience challenges to their growing independence (Bitsko et al., 2014). Thus, cumulative stress may have a particularly salient impact on asthma outcomes among adolescents.

Multiple Sources of Stress

The current study used the ecobiodevelopmental and toxic stress frameworks as guiding principles in assessing multiple stressors that adolescents with asthma living in urban areas are at risk for experiencing, such as poverty and family/community violence (American Academy of Pediatrics, 2014; Johnson, Riley, Granger, & Riis, 2013; Shonkoff et al., 2012). Identifying early sources of stress may have both immediate and long-term ramifications for health outcomes among youth. Specifically, asthma-related health outcomes are of importance given that asthma is one of the most common childhood chronic conditions and pediatric asthma disparities persist (Akinbami et al., 2014). This study considered a cumulative risk model comprised of poverty, neighborhood stress, school-related stress, peer pressure, and caregiver-adolescent conflict.

Neighborhood Stress and Poverty.

The neighborhood has been described as the broadest level of an individual’s social network, which includes influences from socioeconomic status (SES) and exposure to violence, both of which have been associated with asthma morbidity (e.g., ED visits, symptoms; Bellin et al., 2014; Chen & Schreier, 2008). Child and parent reports of neighborhood stress and perceived safety have been associated with pediatric asthma outcomes. For instance, more neighborhood stress and less perceived safety have been associated with higher rates of functional limitations and uncontrolled asthma, more night-time symptoms, and increased need for quick-relief medication (Koinis-Mitchell, Kopel, Salcedo, McCue, & McQuaid, 2014; Kopel et al., 2015). Children living in high poverty neighborhoods and crowded conditions have higher hospital readmission rates than children living in less disadvantaged neighborhoods (Liu & Pearlman, 2009). Together these findings suggest that the physical environment and potential resources afforded in higher SES communities can impact health outcomes among children and adolescents with asthma. Further, a lack of resources and safety may serve as stressful environments that could lead to a worsening of outcomes.

School and Relationship Stressors.

Other prominent stressors identified by adolescents include school and relationships (LaRue & Herrman, 2008; Rew, Tyler, Fredland, & Hannah, 2012). Regardless of age, gender, or grade level, adolescents consistently report school (e.g., performance, worry about grades, graduating/admission into college) as a main source of stress (LaRue & Herrman, 2008; Rew et al., 2012; Stuart, 2006). Adolescents also described relationships with parents and peers as common sources of worry/concern, including strain in the parent-adolescent relationship, pressure from parents, and distractions from peers/romantic partners (LaRue & Herrman, 2008; Rew et al., 2012; Stuart, 2006). Social relationships have been identified as a barrier to asthma care/management in adolescents through embarrassment, reluctance to take medication around peers, and low management self-efficacy (Rhee, Belyea, Ciurzynski, & Brasch, 2009). Prior research has linked harsh/conflicted parent-child interactions to asthma symptom intensity, as well as the downregulation of anti-inflammatory responses (i.e., anti-inflammatory gene expression), and stimulation of cytokines that lead to inflammation (Ehrlich, Miller, & Chen, 2015; Tobin et al., 2015a; Tobin et al., 2015b). In addition to the physical environment and availability of resources, personal relationships and the pressure to perform in school may add to the overall stress impacting health outcomes among adolescents with asthma.

Cumulative Risk

Cumulative risk models offer a parsimonious method for identifying families most in need of intervention, as exposure to multiple, simultaneous risk factors has been associated with worse developmental consequences (Evans, Li, & Whipple, 2013). Cumulative risk models have been used in asthma research to examine the impact of multiple, overlapping risk factors on adverse outcomes rather than the individual impact of each factor alone (Koinis-Mitchell et al., 2007; Evans & English, 2002). Such models offer a more accurate depiction of how families experience risk in that stressful events and processes do not occur in isolation (Everhart, Fiese, & Smyth, 2008). Additionally, cumulative risk models might confer more predictive power (i.e., explain more variance in outcomes) than any one factor (Sameroff, Seifer, Baldwin, & Baldwin, 1993). Prior work has also suggested ways to improve upon these models, including the incorporation of a strong theoretical reasoning for the selection of factors in models (Evans et al., 2013). Moreover, prior studies have dichotomized risk factors, which were often arbitrarily based on sample-specific cut-off points and reduced the variability of factors (Sameroff et al., 1993). To overcome this challenge while maximizing variability, this study prioritized using standardized continuous variables.

Several studies have examined cumulative risk within pediatric asthma populations and found that accumulated stress was: 1) more predictive of asthma morbidity than poverty or asthma severity alone, and 2) associated with greater functional impairment, more ED visits and hospitalizations, and poor health-related QOL (Josie, Greenly, & Drotar, 2007; Koinis-Mitchell et al., 2012; Koinis-Mitchell et al., 2007). Although these studies exemplify the predictive power of cumulative risk models, prior literature has relied heavily on parent-report, specifically parent-perceived stress, and has dichotomized factors before calculating the risk index. Further, the prior studies included a wide age-range of children and adolescents with children as young as 5-to-7 and ranging in age up to 13-to-17. In the current study, adolescents self-reported on all stressors (except poverty) to examine how adolescents’ own perceptions of stress affect their health outcomes. Additionally, this study included both adolescent-reported outcomes (asthma-related quality of life [QOL] and asthma control), and caregiver-reported outcomes (asthma-related ED visits within the last year). Moreover, following the example of Koinis-Mitchell and colleagues (2007), we also examined the predictive value of the cumulative risk model against each individual stressor in the model. Doing so enabled us to determine whether the impact of our cumulative risk model on asthma outcomes was driven by the effects of particular stressors.

Current Study

The current study developed a cumulative risk index comprised of stressors salient to adolescents in urban environments (i.e., poverty, neighborhood stress, school-related stress, peer pressure, caregiver-adolescent conflict) and tested whether the index was associated with each asthma outcome. This study built upon the prior body of work by: 1) grounding the cumulative risk model in a firm theoretical framework (i.e., ecobiodevelopmental and toxic stress frameworks), 2) choosing developmentally appropriate risk factors for adolescents (e.g., school-related stress, peer pressure, caregiver-adolescent conflict), and 3) maintaining the continuous nature of risk exposures when possible. We hypothesized that higher levels of cumulative risk would predict worse asthma control and QOL, as well as more ED visits. Additionally, we hypothesized that the cumulative risk index would be more strongly associated with asthma outcomes than any one stressor in the model. Finally, we hypothesized that adolescents with poorly controlled asthma would report higher levels of individual stressors as compared to adolescents with well-controlled asthma.

Methods

Participants

Sixty-one adolescent and caregiver dyads from an urban center completed the current study. Adolescents were 13 to 17 years old (M=14.77, SD=1.40), primarily African American/Black (93.4%), and most caregivers were biological mothers (90.2%). See Table 1 for full demographic information.

Table 1.

Caregiver and Adolescent Demographics

Demographic Items	Total Sample N=61
Caregiver
Age, M years (SD)	41.93 (8.39)
Race/Ethnicity, n (%)
African American/Black	56 (91.8)
Latino/a	2 (3.2)
Caucasian/White	1 (1.6)
Mixed/Multi-Racial	1 (1.6)
Other	1 (1.6)
Relation to child, n (%)
Biological Mother	55 (90.2)
Grandmother	5 (8.2)
Biological Father	1 (1.6)
Relationship Status, n (%)
Single/Never Married	35 (57.4)
Married	13 (21.3)
Divorced	8 (13.1)
Separated	3 (4.9)
Widowed	2 (3.3)
Education, n (%)
Less than a high school education	17 (27.9)
High school degree	18 (29.5)
Some college	21 (34.4)
College degree or higher	5 (8.2)
Income (Past Month), n (%)
Less than $1,000	24 (39.3)
$1,000 – $1,999	17 (27.9)
$2,000 – $3,999	17 (27.9)
Greater than $4,000	3 (4.9)
Household size, M (SD)	4.48 (1.95)
Adolescent
Age, M years (SD)	14.77 (1.40)
Sex (Female), n (%)	33 (54.1)
Race/Ethnicity, n (%)
African American/Black	57 (93.4)
Latino/a	2 (3.2)
Caucasian/White	1 (1.6)
Mixed/Multi-Racial	1 (1.6)

Procedures

After Institutional Review Board approval was obtained, potential participants were identified by electronic health record review of a major hospital system serving the area. Flyers were also distributed at pulmonology/pediatric clinics and nearby community/neighborhood centers. Researchers assessed participant eligibility over the phone. Adolescent inclusion criteria included 13-to-17 years of age with persistent asthma according to NHLBI guidelines (2013); persistent asthma was determined from caregiver report of criteria (e.g., daytime symptoms, quick-relief medication use). Dyads needed to live in an urban area as identified by zip code and speak English. Exclusion criteria included the adolescent having an additional pulmonary disease, developmental delay, or severe psychiatric or other medical illness.

Eligible families were invited to complete a research session. Depending on family preference, this session took place in a research office or the family’s home. Consent and assent procedures were completed prior to questionnaire administration and both were required in order for a dyad to participate in the study. Trained research assistants administered questionnaires to caregivers and adolescents. Dyads were separated to the extent that the setting would allow (e.g., opposite sides of the room). Caregivers and adolescents were each compensated for their time.

Measures

Demographic Information.

Caregivers reported their household size, marital status, and relationship to the adolescent as well as the adolescent’s race/ethnicity, age, and grade. SES information, also reported by caregivers, included their education, occupation, and total household income (past year and month).

Poverty.

An income-to-needs ratio was calculated based on household income and the poverty threshold for a family of that size. Poverty thresholds were retrieved for all years of the study and matched based on participant visit date. If the income-to-needs ratio was less than 1 the family was considered below the poverty line (U.S. Department of Health and Human Services, 2017). Caregivers self-reported their family’s income bracket (using provided income ranges), which allowed us to determine whether or not the family was living below the poverty line, but not the exact ratio; therefore, a dichotomous variable was used in the cumulative risk index.

Neighborhood Stress Index

(Attar, Guerra, & Tolan, 1994). In the 23-item stress index, adolescents reflected on the past year when determining whether an event had occurred. If the adolescent responded “yes”, they were then asked to rate how stressful the event was on a 1 “not at all stressful/upsetting” to 3 “very stressful/upsetting” scale. The index assesses three domains of stress including life transitions, circumscribed events, and exposure to violence. Higher scores indicate more perceived neighborhood-related stress. The total scale had good internal consistency (α=0.82).

Adolescent Stress Questionnaire

(ASQ; Byrne, Davenport, & Mazanov, 2007). The 58-item ASQ assesses multiple domains of stress that adolescents may experience. The current study utilized the school performance and peer pressure subscales. The school performance subscale assessed stress related to studying/keeping up with schoolwork, concentration, understanding of material, and teacher expectations. The peer pressure subscale assessed stress related to fitting in, physical appearance, feeling judged, and disagreements with peers. Adolescents rated how stressful they found each event/circumstance on a 1 “not at all stressful (or is irrelevant to me)” to 5 “very stressful” scale. The school performance subscale had excellent internal consistency (α=0.90) and the peer pressure subscale had adequate internal consistency (α=0.76).

Conflict Behavior Questionnaire

(CBQ; Robin & Foster, 1989). The CBQ is comprised of two separate forms, one for the parent/caregiver and one for the adolescent, which assess the parent-child relationship. Each of the 20 items was replied to in a true/false format reflecting the parent-child relationship over the past two weeks. Caregivers completed a form about their relationship with the child enrolled in the study; similarly, adolescents completed a form about the caregiver participating in the study with them. The internal consistency of the caregiver-reported CBQ was 0.92; internal consistency of the adolescent-reported CBQ of the primary caregiver was 0.88.

Asthma Assessment Form

(AAF; Rosier et al., 1994). The AAF evaluates asthma severity, missed days of school, doctor visits, ED visits, and hospital stays due to asthma over the past 4 weeks and 12 months. Caregivers completed this form and number of ED visits over the last 12 months was used as an outcome in analyses.

Asthma Control Test

(ACT; Nathan et al., 2004). Asthma control was assessed via adolescent self-report with the 5-item ACT designed to be completed by children 12 years or older. Items are rated on a 5-point scale assessing symptoms over the past 4 weeks. Lower scores on the ACT indicate worse asthma control; the scale had adequate internal consistency (α=0.79).

Pediatric Asthma Quality of Life Questionnaire

(PAQLQ; Juniper et al., 1996). The 23-item PAQLQ assesses physical, emotional, and social impairment due to asthma. Adolescents rated responses on a 7-point scale from 1 “extremely bothered/all of the time” to 7 “not at all bothered/none of the time”. An overall score was determined from the mean of all responses, with lower scores indicating poorer QOL; internal consistency was excellent (α=0.95).

Data Analysis

Analyses were conducted using IBM SPSS Statistics, Version 24.0 software (IBM Corp., Armonk, NY, USA). Adolescent race/ethnicity, sex, and age were examined as potential covariates and variable skew and kurtosis were assessed. A cumulative risk score was calculated for each family. For poverty, 1 was assigned to families that were under the poverty threshold and 0 for families above the poverty threshold. Continuous variables, including neighborhood stress, school-related stress, peer pressure, and caregiver-adolescent conflict, were standardized before risk indices were calculated. In accord with prior research, scores from each risk variable were summed to create the cumulative risk index.

Hierarchical multiple regression analyses were conducted to assess the association between the cumulative risk index and each asthma outcome. Relevant covariate(s) were entered in the first step of the model and the risk index was entered in the second step. Hierarchical regressions were also conducted to assess whether the cumulative risk index predicted each asthma outcome with each individual risk factor in the model. Covariate(s) were entered in the first step, the single stressor in the second step, and the cumulative risk index was entered in the third step of each model. Independent t-tests were conducted to determine differences in continuous risk factors across level of asthma control. A chi-square analysis was used to assess the association between poverty and asthma control. ACT scores of 19 and below were classified as “poorly controlled”; scores above 19 were classified as “well-controlled” (Nathan et al., 2004).

Results

Preliminary Analyses

Descriptive information for study measures are presented in Table 2. Approximately 63.9% of adolescents had poorly controlled asthma (i.e., ACT scores ≤ 19), 72.0% of adolescents had been to the ED at least once in the last year, and 45.9% reported two or more asthma-related ED visits. Additionally, 77.0% of families lived below the poverty line. Cumulative risk scores ranged from −4.31 to 9.59 (M=0.77, SD=2.98) with higher scores indicating greater risk. In this sample, 47.5% of adolescents had cumulative risk scores higher than the mean, with the top quartile having scores greater than three times the mean (> 2.92).

Table 2.

Descriptives of Measures

Measure	Mean (SD)	Score Range
Child QOL	5.28 (1.30)	2.17 – 7.00
Asthma Control (ACT)	17.82 (4.95)	7–25
ED Visits (in the last 12 months)	1.56 (1.43)	0 – 6
Below Poverty Line	(47) 77.0%
Neighborhood Stress	12.67 (9.16)	0–35
School Performance (ASQ)	18.05 (8.45)	7–34
Peer Pressure (ASQ)	10.79 (4.76)	7–32
Caregiver-Adolescent Conflict (CBQ-Adolescent Report)	4.20 (4.37)	0–18
Cumulative Risk Index	0.77 (2.98)	−4.31 – 9.59

Adolescent age was significantly correlated with ED visits (r=.37, p=.003), with older adolescents having been seen at the ED more frequently in the last year. Both QOL (F(1,59)=9.09, p=.004) and asthma control (F(1,59)=8.54, p=.005) differed by adolescent sex. Male adolescents reported better QOL (M=5.80, SD=1.28) than female adolescents (M=4.85, SD=1.18). Additionally, male adolescents had higher levels of asthma control (M=19.71, SD=4.72) than female adolescents (M=16.21, SD=4.61). None of the outcome variables differed by adolescent race/ethnicity. Adolescent sex was controlled for in models predicting QOL and asthma control, and adolescent age was controlled for in models predicting asthma-related ED visits. The ASQ peer pressure scores were slightly skewed (2.14) and kurtotic (6.18); therefore, a natural log transformation was computed.

Cumulative Risk Analyses

First, in multiple regression analyses, the cumulative risk index was examined as a correlate of each asthma outcome (QOL, asthma control, ED visits; see Table 3). Cumulative risk was significantly associated with QOL after controlling for adolescent sex, such that higher risk scores were related to lower QOL scores. Higher cumulative risk scores were also associated with worse asthma control, controlling for adolescent sex. Finally, higher cumulative risk scores were associated with more asthma-related ED visits in the last 12 months after controlling for child age. Additionally, analyses were re-run splitting the cohort into two based on adolescent age; younger adolescents (13–14 years) and older adolescents (15–17 years). All findings remained significant except that the cumulative risk index was no longer associated with ED visits among older adolescents (R²=.01, t(1,32)=0.43, p=.67).

Table 3.

Estimates of the Cumulative Risk Index (Poverty, Neighborhood Stress, School Stress, Peer Pressure, Caregiver-Adolescent Conflict) as a Predictor of Adolescent Asthma Outcomes

Outcome	∆R2	∆F	b	SE	95% CI	t	df
Quality of Life (QOL)	.36	42.02***	−.28	.04	−0.36, −0.19	−6.48***	58
Asthma Control	.27	26.45***	−.91	.18	−.1.26, −0.56	−5.14***	58
Emergency Department (ED) Visits	.06	4.44*	.12	.06	0.01, 0.24	2.11*	58

Note.

^***p < .001

^**p < .01

^*p < .05.

In the QOL and asthma control models, child sex was included as a covariate in the first step. In the ED visits model child age was included as a covariate in the first step.

Cumulative Risk Indices Compared to Individual Stressors

Adolescent QOL and asthma control were correlated with all stressors included in the model, whereas number of ED visits was correlated with only poverty, neighborhood stress, and school-related stress (Table 4). Hierarchical regression analyses were conducted to determine whether cumulative risk indices were a stronger predictor of the three outcomes than any one individual stressor. The cumulative risk index was a more robust predictor of QOL and asthma control than poverty, neighborhood stress, school-related stress, peer pressure, and caregiver-adolescent conflict, controlling for adolescent sex (see Table 5 for model estimates). However, when examining asthma-related ED visits, the cumulative risk index was only a stronger predictor than peer pressure.

Table 4.

Zero-Order Correlations

	1	2	3	4	5	6	7	8
Poverty (1)	1.00
Neighborhood Stress (2)	.36**	1.00
School Stress (3)	.27*	.44**	1.00
Peer Pressure (4)	.16	.13	.47**	1.00
Caregiver-Adolescent Conflict (5)	.17	.29*	.29*	.33**	1.00
QOL (6)	−.34**	−.51**	−.67**	−.38**	−.34**	1.00
Asthma Control (7)	−.36**	−.40**	−.59**	−.34**	−.32*	.80**	1.00
ED Visits (8)	.30*	.28*	.32*	.07	.12	−.26*	−.35**	1.00

Note.

^*p < .05

^**p < .01

Table 5.

Cumulative Risk Estimates from Regression Analyses Comparing the Cumulative Risk Index with Each Individual Stressor

Stressor being Compared to the Cumulative Risk Index	∆R2	∆F	b	SE	95% CI	t	df
Outcome: Quality of Life
Poverty	.28	32.72***	−.27	.05	−0.37, −0.18	−5.72***	57
Neighborhood Stress	.18	20.81***	−.26	.06	−0.37, −0.14	−4.45***	57
School Performance	.06	7.88**	−.18	.06	−0.30, −0.05	−2.81**	57
Peer Pressure	.29	33.40***	−.32	.06	−0.43, −0.21	−5.78***	57
Caregiver-Adolescent Conflict	.25	29.11***	−.32	.06	−0.45, −0.20	−5.40***	57
Outcome: Asthma Control
Poverty	.19	18.44***	−.84	.20	−1.23, −0.45	−4.29***	57
Neighborhood Stress	.17	26.93***	−.96	.23	−1.43, −0.49	−4.12***	57
School Performance	.05	5.15*	−.60	.26	−1.13, −0.07	−2.27*	57
Peer Pressure	.21	20.17***	−1.04	.23	−1.51, −0.58	−4.49***	57
Caregiver-Adolescent Conflict	.16	15.32***	−.99	.25	−1.49, −0.48	−3.91***	57
Outcome: Emergency Department (ED) Visits
Poverty	.03	2.04	.09	.06	−0.04, 0.22	1.43	57
Neighborhood Stress	.03	2.41	.12	.08	−0.04, 0.28	1.55	57
School Performance	.001	0.10	.03	.09	−0.15, 0.21	0.32	57
Peer Pressure	.08	5.89*	.19	.08	0.03, 0.34	2.43*	57
Caregiver-Adolescent Conflict	.05	3.21†	.14	.08	−0.02, 0.30	1.79†	57

Note.

^***p < .001

^**p < .01

^*p < .05

^†p < .10.

Differences in Stressors Across Level of Asthma Control

Approximately one third of the sample (36.1%) reported well-controlled asthma. Adolescents who had well-controlled asthma had significantly better QOL (M=6.30, SD=0.75) than those with poorly controlled asthma (M=4.71, SD=1.20; F(58.3)=9.17, t=−6.39, p <.001). Significantly lower levels of neighborhood stress were found among adolescents with well-controlled asthma (M=7.32, SD=8.29) compared to adolescents with poorly controlled asthma (M=15.69, SD=8.27; F(59)=0.14, t=3.79, p <.001). Additionally, school-related stress scores were significantly lower among those with well-controlled asthma (M=13.09, SD=6.58) versus those with poorly controlled asthma (M=20.84, SD=8.15; F(59)=2.94, t=3.81, p <.001). Peer-pressure scores were lower among adolescents with well-controlled asthma (M=9.05, SD=2.98) compared to those without (M=11.77, SD=5.31; F(59)=1.31, t=2.45, p=.02). Caregiver-adolescent conflict scores did not significantly differ across the two groups (F(59)=0.47, t=1.50, p=.14). As indicated by a chi-square test (χ²(1)=6.28, p=.01), about 87% (34 of 39) of adolescents with poorly controlled asthma lived below the poverty line and 41% (9 of 22) with well-controlled asthma lived below the poverty line.

Discussion

The present study found that among adolescents with asthma, a cumulative risk model of stress (poverty, neighborhood stress, school-related stress, peer pressure, caregiver-adolescent conflict) was associated with asthma control, QOL, and asthma-related ED visits. Further, the cumulative risk index was a more robust predictor of QOL and asthma control than any one stressor; findings with ED visits were mixed. Importantly, adolescents with well-controlled asthma had lower scores on the neighborhood stress and school-related stress scales. This study extended prior research by focusing specifically on a cohort of adolescents 13-to-17 years of age and including adolescent perceptions of stress. Much of the previous work on cumulative risk among pediatric asthma populations has included a wide age range of children and adolescents in a single study and/or relied on parent reports of stressful experiences (Josie et al., 2007; Koinis-Mitchell et al., 2012; Koinis-Mitchell et al., 2007).

Consistent with previous research linking individual sources of stress (e.g., poverty, harsh parent-child interactions, neighborhood stress) with poorer health outcomes for children with asthma (Koinis-Mitchell et al., 2014; Kopel et al., 2005; Tobin et al., 2015a), we found that an accumulation of stress was associated with poorer QOL, asthma control, and more ED visits. Additionally, prior research has demonstrated an association between stressful life events among inner-city adolescents, including events related to school and family relationships, and asthma prevalence and morbidity (Turyk et al., 2017). Stress may be associated with worse asthma morbidity through several pathways: 1) stress is associated with the dysregulation of the immune system and HPA-axis, which may impact inflammatory responses related to asthma; 2) stress can negatively impact health behaviors which can lead to poorer asthma management; and 3) stress can directly affect the mental health of adolescents and caregivers, which can hinder asthma care (Shonkoff et al., 2012; Turyk et al., 2017; Wood, Miller, & Lehman, 2015). Current findings add to the growing body of literature linking stress and asthma morbidity among adolescents, and suggest that a cumulative risk index of stress may be important in asthma outcomes.

However, findings related to ED visits should be interpreted with caution given that the results only held in the younger adolescent group (13–14 years). This finding may suggest that among older adolescents, other factors or sources of stress may be more relevant to the frequency of ED visits. For example, prior research has found that compared to younger adolescents, older adolescents have worse asthma medication adherence (Mosnaim et al., 2014), which contributes to asthma-related ED visits. Moreover, increases in the desire for instant gratification may impact compliance to medications that do not provide an immediate effect (e.g., controller medication; Mosnaim et al., 2014). Additionally, the use of avoidance coping (as seen in adults) may aid in the explanation of higher rates of ED visits among older adolescents (Adams, Smith, Ruffin, 2000). Further research is needed to delineate factors driving differences in asthma-related healthcare utilization among younger versus older adolescents.

This study also examined the predictive nature of the cumulative risk index in comparison to each individual stressor included in the risk model. Consistent with previous research (e.g., Koinis-Mitchell et al., 2007), the cumulative risk index was a more robust predictor of QOL and asthma control than each individual stressor. However, the cumulative risk index was not a stronger predictor of ED visits (across the full adolescent age range) than poverty, neighborhood stress, or school-related stress, although, it was a stronger predictor when compared to peer pressure. This finding suggests that environmental factors (neighborhood and school) may be more salient predictors of asthma-related ED visits among adolescents. In previous research, factors related to neighborhood stress and poverty, including violence, crime, housing deterioration, living in lower SES areas, and perceived safety, have each been associated with worse asthma outcomes (e.g., poor asthma control, increased symptoms, ED visits; Chen & Schreier, 2008; Koinis-Mitchell et al., 2014; Kopel et al., 2015). Additionally, neighborhood stress is often related to access to care, such that distance from healthcare facilities and access to public transportation could lead to inconsistent asthma care, which can directly impact ED visits. School-related stress was also associated with more ED visits. Previous research has connected asthma to school absenteeism and difficulties in school (Moonie, Sterling, Figgs, & Castro, 2006). Findings from the current study suggest that poverty, neighborhood stress, and school-related stress may each be important predictors of asthma-related ED visits. Thus, unlike our findings for QOL and asthma control, factors related more to the environment (including school) and access to care, may be more relevant to ED use among adolescents (13 to 17 years of age).

When examining differences between adolescents with well-controlled asthma (36%) and adolescents with poorly controlled asthma, adolescents with well-controlled asthma had significantly lower levels of neighborhood and school-related stress. Although adolescents with well-controlled asthma had lower peer pressure scores as well, adolescents with poorly controlled asthma had scores twice as high on neighborhood stress and 1.62 times higher on school-related stress. Adolescents with well-controlled asthma were less likely to live in poverty (41%) than adolescents with poorly controlled asthma (87%). Caregiver-adolescent conflict scores did not significantly differ between groups.

Together, these findings suggest that adolescents with well-controlled asthma may either be more equipped to handle environment-related stressors (i.e., school, neighborhood) or may not perceive their experiences at school or in their neighborhoods as major sources of stress. In fact, low scores on the neighborhood and school-related stress assessments could indicate the absence of a particular stressor or that the adolescent did not find it to be stressful. It is possible that adolescents who have well-controlled asthma, which is often indicative of well-managed asthma, may also have more supports in place to handle stress from their everyday environment. For example, prior research has found that children and adolescents who have positive attitudes toward school, feel connected to family, use shift-and-persist strategies, and have higher levels of asthma-related self-efficacy exhibit less asthma morbidity in the face of stress (Chen et al., 2011; Koinis-Mitchell et al., 2012). Although more research is needed to better understand the association between stress and asthma control, future intervention work might consider arming adolescents with positive ways for coping with sources of stress not limited solely to their disease.

Limitations

This study had a limited sample size, which impacts the generalizability of findings. Findings are also limited to adolescents who met inclusion criteria. All participants were recruited from an urban area and findings may not generalize to adolescents living in other areas, such as rural settings, who may experience different stressors (e.g., different allergens, scarcity of specialty providers; Kopel, Phipatanakul, Gaffin, 2014). There may be more factors contributing to differences in asthma outcomes between younger and older adolescents that were beyond the scope of the current study. Additionally, other environmental exposures, such as air pollution, were not controlled for and may influence adolescents’ asthma outcomes. This study included specific stressors based on previous research and our theoretical background. Other sources of stress may be appropriate to include in future models, such as air pollution, household conditions, or environmental tobacco smoke exposure.

Data were self-report and subject to social desirability and time-recall biases. More objective assessments should be included in future research. For example, health records could be reviewed for number of ED visits. Moreover, although caregivers and adolescents completed questionnaires separately, dyads were sometimes in close proximity when the research sessions took place in family homes. Therefore, some reports may be an underestimation, such as caregiver-adolescent conflict behavior.

Future Directions and Clinical Implications

Findings suggest several directions for future research. First, although we theorize that the transition to self-management is a key mechanism by which stress affects asthma health outcomes during adolescence, we were not able to test this in the current study. A next step is to analyze this path using measures of asthma management behaviors specific to adolescents (i.e., motivation, parent-youth negotiation, parental monitoring). Second, results need to be replicated with larger groups of adolescents to determine if the same pattern of findings are seen among other samples. Third, future research could include additional stressors in cumulative risk indices, such as stress related to romantic partners or financial pressures, to gather more comprehensive assessments of adolescent stress. Finally, as previously mentioned, intervention work could target the accumulation of stress. Specifically, interventions could incorporate strategies for handling overlapping sources of stress and ways to minimize the cumulative effects. Our findings also suggest that interventions targeting singular outcomes may need to be tailored. For instance, in efforts to reduce ED visits, targeting issues related to access, such as transportation, may be important; on the other hand, to improve QOL, targeting overlapping connections among stress from home, school, and neighborhood environments may be imperative.

Together, previous research (Chen & Schreier, 2008; Tobin et al., 2015a; Koinis-Mitchell et al., 2012; Koinis-Mitchell et al., 2007) and findings from this study call for a focus on how the accumulation of stress among children/adolescents can complicate disease outcomes. Prior work has focused on the importance of including mental health screenings (i.e., depression and anxiety assessments) in the treatment of asthma among adolescents (Sadof & Kaslovsky, 2011; Towns & van Asperen, 2009); however, it may be equally important to assess sources of stress among adolescents and address their impact on disease outcomes. Doing so may prompt healthcare providers to provide families with referrals for resources (e.g., counselors, community centers) that may otherwise not be provided. Previous research has found that caregivers of inner-city children with asthma want and need asthma education and information on accessible community resources (Bellin et al., 2017). Psychoeducation regarding associations between multiple sources of stress and asthma outcomes could also be provided to families briefly during a clinic visit, in an educational handout, or even in stress-management programs. Prior research has found that community-based stress-management interventions for children with asthma have been effective in reducing stress and improving physical health, including pulmonary function (Long et al., 2011). Overall, findings from this study suggest that the accumulation of stress can have a negative impact on asthma outcomes among adolescents living in urban areas and more research is needed among this vulnerable population to further minimize pediatric asthma disparities.