Hearing Impairment and School Engagement Outcomes in US Children

Key Points

Question

How does school engagement (participation in extracurricular activities and educational performance) compare among US children with parent-reported deafness or hearing problems vs those without deafness or hearing problems as well as within subgroups of those with deafness or hearing problems?

Findings

In this cross-sectional study of 155 178 school-aged children (weighted, 49 340 700 children) in the US, children with deafness or hearing problems were less likely to participate in 3 of 5 extracurricular activities and more likely to perform worse in 5 of 5 educational performance variables compared with children without deafness or hearing problems. Moreover, although subgroup analyses of children with deafness or hearing problems with and without hearing loss sequelae (speech and/or language disorders, behavioral and/or conduct problems, or neuropsychiatric and/or learning disorders) revealed significant differences, the presence of sequelae only partially mediated these associations.

Meaning

These findings support the need for improvements in direct intervention for deafness and hearing problems in the educational setting.

Abstract

Importance

Ensuring appropriate school engagement for deaf or hard of hearing (DHH) children in the US is important for improving their long-term outcomes as they grow into adults.

Objective

To examine the associations between hearing loss (HL), its sequelae (speech and/or language disorders, behavioral and/or conduct problems, and neuropsychiatric and/or learning disorders), and various school engagement measures among school-aged children in the US.

Design, Setting, and Participants

This cross-sectional study combined data from the nationally representative 2016-2021 National Survey of Children’s Health. A total of 155 178 randomly selected children (weighted, approximately 49 340 700 children) aged 6 to 17 years with a parent or caregiver who responded to an address-based survey by mail or online were included. All analyses were weighted to account for the probability of selection and nonresponse and to reflect population-based estimates representative of all noninstitutionalized school-aged US children and adolescents residing in housing units.

Main Outcomes and Measures

Diverse school engagement measures, including extracurricular participation in sports, clubs, paid work, volunteer work, and organized lessons, as well as educational performance variables, including missed school days, not caring about doing well in school, not doing required homework, grade repetition, and parent or guardian contacted by school.

Results

Of the estimated 49 340 700 children aged 6 to 17 years (41.1% aged 6-10 years; 51.1% male; 54.7% without underrepresented minority status), an estimated 1.4% (95% CI, 1.2%-1.5%) were reported by their parent or guardian to have deafness or hearing problems. Having DHH status was associated with significantly worse outcomes on 8 of 10 school engagement measures (eg, participation in sports: adjusted odds ratio [AOR], 0.75 [95% CI, 0.60-0.93]; missed school days: AOR, 2.98 [95% CI, 2.21-4.00]), even after adjustment for age, sex, underrepresented minority status, highest educational level of parent or guardian, federal poverty level of the household, and primary language in the household. Moreover, although subgroup analyses of DHH children with and without HL sequelae revealed significant differences (speech and/or language disorder: AOR, 5.83 [95% CI, 4.31-7.89]; behavioral and/or conduct problem: AOR, 2.75 [95% CI, 2.10-3.60]; neuropsychiatric and/or learning disorder: AOR, 3.06 [95% CI, 2.39-3.91]), HL sequelae only partially mediated the associations between these disparities.

Conclusions and Relevance

In this cross-sectional study, DHH status itself may have been the primary factor directly associated with school engagement disparities. These findings suggest the need for greater emphasis on educational accommodations and support for hearing status itself, independent of the presence or absence of HL sequelae.

This cross-sectional study uses data from the 2016-2021 National Survey of Children’s Health to examine the associations between hearing loss, its sequelae, and school engagement measures among school-aged children in the US.

Introduction

Hearing loss (HL) is one of the most common congenital abnormalities in the US, with 1.7 per 1000 infants deaf or hard of hearing (DHH) at birth.¹ Due to progressive and late-onset HL, this prevalence increases with age, with 2.4% to as many as 14.9% of school-aged children reported to have HL.^2,3 A majority of school-aged children with HL are educated in mainstream classes with typically hearing peers but require hearing-assistive technologies (eg, hearing aids or cochlear implants and frequency modulation systems) and supportive educational services (eg, 504 plans or individualized education plans [IEPs]) to optimize their access to sound and language⁴ and to reduce the risk of potential sequelae that occur when intervention is not provided for HL, such as speech and/or language disorders (hereinafter, speech/language disorders), psychosocial difficulties, behavioral and/or conduct problems (hereinafter, behavioral/conduct problems), or neuropsychiatric and/or learning disorders (hereinafter, neuropsychiatric/learning disorders).^{5,6,7,8,9,10,11} Ultimately, these sequelae of inadequately managed HL are thought to underlie challenges in academic achievement and lower rates of literacy and high school graduation among DHH individuals.^6,10 Because HL is a recognized disability, all DHH children are eligible for educational accommodations in the form of a 504 plan (a federal civil rights protection based on Section 504 of the Rehabilitation Act of 1973),¹² yet eligibility for the most comprehensive set of supportive services, including special education services through an IEP, which is based on the Individuals with Disabilities Education Act of 1990,¹³ is contingent on identification of developmental and educational sequelae associated with DHH status.¹⁴ However, the qualifying criteria for these educational sequelae are intentionally broad and subject to wide spectrums of interpretation. Accordingly, depending on an IEP eligibility team’s subjective explanation of DHH status vs HL sequelae as the primary factor associated with educational performance impacts, there is potential for provisional creation of additional IEP eligibility based on having these HL sequelae. This situation can ultimately result in placing substantial emphasis on supporting HL sequelae and DHH children most affected by these sequelae (rather than all DHH children) as well as unequal IEP access for DHH children depending on the absence or presence of HL sequelae.

Few studies have examined the associations between HL, its sequelae (speech/language disorders, behavioral/conduct problems, and neuropsychiatric/learning disorders), and more specific, granular downstream consequences at school. For example, a systematic review and meta-analysis⁵ of studies evaluating quality of life in children with HL found that children with HL have difficulties in situations affecting interactions with others, participation in social and school activities, and emotional well-being. However, the studies included in this meta-analysis⁵ were restricted to broad school engagement measures and institution- or geography-specific cohorts with relatively narrow age and grade groupings that were not nationally representative. Two nationally representative studies^8,11 have found that US children with a current hearing problem had an odds ratio of 1.9 for grade repetition,¹¹ and US adolescents with HL were 31% less likely to participate in sports.⁸ Yet, to our knowledge, no study has been conducted among school-aged children with HL to assess a multivariable (10 total) comprehensive landscape of school engagement (participation in extracurricular activities and educational performance), nor has any study evaluated this set of variables using parents’ perceptions of HL and more than 2 years of data from a nationally representative data set. Furthermore, the association between HL, its sequelae, and these school engagement outcomes among school-aged children in the US has yet to be elucidated.

The purpose of this study was to assess the odds of educational performance (ie, missed school days, grade repetition, doing required homework, caring about doing well in school, and parent or guardian contacted by school) and engagement in extracurricular activity (ie, sports, clubs, paid work, volunteer work, and organized lessons) outcomes by parent-reported hearing status (ie, with or without deafness or hearing problems). Our secondary aim was to assess the odds of these outcomes for subgroups of DHH children with and without HL sequelae (ie, speech/language disorders, behavioral/conduct problems, and neuropsychiatric/learning disorders) as well as to evaluate the extent to which these sequelae mediate the associations found between HL and school engagement. To address these aims, we used the National Survey of Children’s Health (NSCH), a nationally representative data set on the health and well-being of children aged 0 to 17 years in the US, to explore the extent to which contemporary data continue to reveal the implications of HL for school-aged children at the national level and hypothesize about the potential mechanisms underlying these associations.¹⁵

Methods

Data Source

The NSCH, a mail and online survey administered by the Maternal and Child Health Bureau of the Health Resources & Services Administration of the US Department of Health and Human Services, served as the data source for this study.¹⁵ The NSCH is an annual survey that uses sampling weights to generate population-based estimates representative of noninstitutionalized US children and adolescents aged 0 to 17 years; children with special health care needs and children younger than 5 years were oversampled to ensure adequate representation.¹⁵ The survey’s content focuses on a child’s physical and emotional health and factors associated with a child’s well-being, such as health care access, use, and quality; family and neighborhood characteristics; and school and after-school experiences. Per University of California San Francisco institutional review board guidelines, since all data were unidentifiable, deidentified, and publicly available, the study did not qualify as human participant research and did not require institutional review board review

We combined data from the 2016-2021 NSCH and, due to relevant survey item age restrictions, limited the study population to school-aged children and adolescents aged 6 to 17 years. All values in this study did not reflect the total survey sample size of 155 178 responses, but sample weights were used to generate estimates for a total population of approximately 49 340 700 school-aged US children and adolescents. These estimates were rounded to the nearest thousand or million as appropriate.

From the NSCH, we extracted and coded demographic variables, including age in years, sex (male or female), race and ethnicity (underrepresented minority [URM] status or non-URM status), highest educational level of parent or guardian (less than college or college and higher), federal poverty level (FPL) of the household (<200% or ≥200%), and primary language in the household (English or non-English). Respondents were classified as being from a URM group based on the National Institutes of Health working definition of URM groups (ie, those who self-report as Black, Hispanic, Native American or Alaska Native, Native Hawaiian or Other Pacific Islander, or ≥2 races when ≥1 races from the aforementioned categories were defined as being from a URM group). We focused on survey questions regarding hearing health status (termed deafness or hearing problems), HL sequelae, and school engagement outcomes (extracurricular activities and educational performance). The survey items for all study variables are shown in eTable 1 in Supplement 1.

Statistical Analysis

All categorical variables were dichotomized (variable coding is shown in eTable 1 in Supplement 1). After incorporating clustering, stratification, and weights to account for survey design, descriptive statistics were used to assess the demographic characteristics of the population. Associations between demographic variables and DHH status and each combined DHH and sequelae subgroup were also assessed using Pearson χ² tests and t tests. Univariate statistics were computed for the association between DHH status and subgroups, demographic variables, and all outcome variables. Multivariate logistic regression models were then used to examine the associations of DHH status and DHH subgroups with outcome variables, adjusting for covariates specified a priori, including age, sex, URM status, highest educational level of parent or guardian, FPL of the household, and primary language in the household. These factors were chosen because they have previously been found to be associated with access to obtaining a hearing diagnosis or intervention^16,17,18,19; they were also used to characterize sociodemographic differences between DHH children vs non-DHH children.

Mediation analyses were subsequently performed to evaluate whether HL sequelae (speech/language disorder, behavioral/conduct problem, and neuropsychiatric/learning disorder) helped to explain any associations between DHH status and each outcome measure. The independent variable was DHH status, the HL sequelae were mediators, and all school engagement variables were outcome variables. Each mediation model also included all demographic covariates to adjust for potential confounding variables, and a sensitivity analysis was performed by excluding demographic covariates in the mediation models. With statistical significance set at 2-tailed P = .05, mediation results were interpreted as representing full mediation (ie, direct and indirect effect estimates were significant, and total effect estimates were not significant), partial mediation (ie, all effect estimates were significant), and no mediation (ie, indirect effect estimates were not significant). Odds ratios (ORs) with 95% CIs and mediation effect estimates were computed with statistical significance set at 2-tailed P = .05. Stata software, version 17 (StataCorp LLC), was used for all statistical analysis.

Results

The 6-year combined NSCH data set identified an estimated 666 300 children (95% CI, 589 200-743 500 children) aged 6 to 17 years nationally as having deafness or problems with hearing, representing approximately 1.4% (95% CI, 1.2%-1.5%) of the total school-aged study population. Of those, 151 100 DHH children (95% CI, 114 600-187 500) were reported to have a current speech/language disorder, 132 100 (95% CI, 106 900-157 400) were reported to have a current behavioral/conduct problem, and 241 900 (95% CI, 202 100-281 600) were reported to have a current neuropsychiatric/learning disorder (eTable 2 in Supplement 1).

The demographic characteristics of the total NSCH sample, DHH children, and non-DHH children as well as the differences in demographic characteristics between DHH and non-DHH children are shown in Table 1. Among 155 178 randomly selected DHH children (weighted, approximately 49 340 700 children) included in the analysis, 41.1% were aged 6 to 10 years, 25.5% were aged 11 to 13 years, and 33.5% were aged 14 to 17 years; 51.1% were male and 48.9% were female. With regard to race and ethnicity, 54.7% of children had non-URM status and 45.3% had URM status. Compared with non-DHH children, there was higher representation from children aged 6 to 13 years, female children, and English-speaking households among DHH children. However, there were no significant differences based on sex, URM status, or primary language in the household between DHH and non-DHH children. Parental educational level and FPL were associated with DHH status, with lower parental educational level and higher FPL significantly more common among DHH vs non-DHH children (eg, less than college educational level: 66.0% vs 51.9%; P < .001; FPL <200%: 54.9% vs 41.1%; P < .001).

Table 1. Participant Characteristics.

Characteristic	Total population, weighted %	Weighted % (95% CI)		P valuea
		DHH children	Non-DHH children
Sample size, No.
Unweighted	155 178	1984	152 276	NA
Weighted	49 340 700	666 300 (589 200-743 500)	48 674 300 (48 170 900-49 177 700)
Age, y
6-10	41.1	42.5 (36.7-48.5)	41.1 (40.5-41.7)	.23
11-13	25.5	29.5 (24.6-35.0)	25.4 (24.9-25.9)
14-17	33.5	28.0 (23.6-32.9)	33.5 (33.0-34.1)
Sex
Female	48.9	52.7 (47.0-58.3)	48.9 (48.3-49.5)	.19
Male	51.1	47.3 (41.7-53.0)	51.1 (50.5-51.7)
Race and ethnicity
URM status	45.3	49.6 (43.8-55.4)	54.8 (54.2-55.4)	.14
Non-URM status	54.7	50.4 (44.6-56.2)	45.2 (44.6-45.8)
Highest educational level of parent or guardian
Less than college	52.1	66.0 (61.3-70.4)	51.9 (51.3-52.5)	<.001
College and higher	47.9	34.0 (29.6-38.7)	48.1 (47.5-48.7)
FPL of household, %
<200	41.3	54.9 (49.3-60.4)	41.1 (40.5-41.7)	<.001
≥200	58.7	45.1 (39.6-50.7)	58.9 (58.3-59.5)
Primary language in household
Non-English	14.6	12.5 (8.7-17.7)	14.7 (14.1-15.2)	.38
English	85.4	87.5 (82.3-91.3)	85.3 (84.8-86.0)

Abbreviations: DHH, deaf or hard of hearing; FPL, federal poverty level; NA, not applicable; URM, underrepresented minority.

^a P values were calculated using χ² tests. Statistical significance was set at 2-tailed α = .05.

The ORs, adjusted for demographic covariates, of each school engagement outcome among DHH children are shown in Figure 1A. When compared with non-DHH children, DHH children were less likely to participate in 3 of the 5 extracurricular activities (sports: adjusted OR [AOR], 0.75 [95% CI, 0.60-0.93]; paid work: AOR, 0.71 [95% CI, 0.52-0.95]), volunteer work: AOR, 0.73 [95% CI, 0.60-0.88]) and more likely to have worse outcomes on all 5 educational performance variables (grade repetition: AOR, 1.82 [95% CI, 1.39-2.37]; missed school days: AOR, 2.98 [95% CI, 2.21-4.00]; parent or guardian contacted by school: AOR, 1.95 [95% CI, 1.57-2.43]; did not care about doing well in school: AOR, 2.06 [95% CI, 1.54-2.74]; did not do required homework: AOR, 2.36 [95% CI, 1.78-3.12]). All χ² test comparisons of percentages of each outcome between DHH and non-DHH children were statistically significant; for example, 44.2% of DHH children participated in sports compared with 55.3% of non-DHH children (P = .01), and 12.0% of DHH children repeated school grades compared with 6.3% of non-DHH children (P < .001) (Figure 1B; eTable 2 in Supplement 1).

Figure 1. School Engagement Challenges Among Deaf or Hard of Hearing Children.

Error bars represent 95% CIs. AOR indicates adjusted odds ratio; and DHH, deaf or hard of hearing.

Compared with non-DHH children, DHH children were also more likely to have speech/language disorders (AOR, 5.83; 95% CI, 4.31-7.89), behavioral/conduct problems (AOR, 2.75; 95% CI, 2.10-3.60), and neuropsychiatric/learning disorders (AOR, 3.06; 95% CI, 2.39-3.91). This finding was consistent with the theoretical model proposing that HL sequelae can be associated with being DHH and may mediate the association between DHH status and school engagement outcomes. However, an alternate hypothesis is that DHH status confers risk of worse school engagement outcomes independent of these HL sequelae. To first assess this hypothesis, we performed subgroup analyses to compare the school engagement outcomes between subgroups of DHH children who did or did not report each of these HL sequelae. These subgroup analyses found that almost all school engagement outcomes were different between cohorts of DHH children who differed based on the presence or absence of speech/language disorders, behavioral/conduct problems, or neuropsychiatric/learning disorders (Figure 2; eTable 2 in Supplement 1). For example, χ² analysis revealed that DHH children with speech/language disorders had differences in percentages for all 10 outcomes, DHH children with behavioral/conduct problems had differences in percentages for 9 outcomes (with the exception of paid work), and DHH children with neuropsychiatric/learning disorders had differences in percentages for 7 outcomes (with the exception of the educational performance outcomes of paid work, volunteer work, and organized lessons).

Figure 2. School Engagement Challenges Among Subgroups of Deaf or Hard of Hearing Children.

AOR indicates adjusted odds ratio; and DHH, deaf or hard of hearing.

To further evaluate the hypothesis that these HL sequelae may mediate the association between DHH status and the school engagement outcomes, we performed mediation analyses (Figure 3A). Because DHH status was associated with each of the school engagement outcomes, DHH status was associated with each of the 3 HL sequelae, and HL sequelae were associated with each of the school engagement outcomes (data not shown), the criteria to perform mediation analysis were met. Mediation models considered each of the HL sequelae as potential mediators of the association between DHH status and school engagement outcomes. We found that the HL sequelae only partially mediated all of the observed associations with school engagement outcomes (Figure 3B; Table 2). Across each combination of potential mediator and outcome variable, the percentage mediated, or the percentage of the total association of DHH status with each outcome attributed to the mediator, was lower than or equal to that from DHH status directly. For example, 28.2% of the total effect estimate of DHH status for sports participation was mediated by having a speech/language disorder compared with 71.8% being considered a direct consequence of DHH status, and the ratio of the indirect effect estimate to direct effect estimate was 39.3%. Similarly, 75.7% of the total effect estimate for grade repetition was attributed to being a direct consequence of DHH status, while only 33.3% was mediated by having a behavioral/conduct problem, and the magnitude of the indirect effect estimate was only 32.1% of the direct effect estimate. The sensitivity analysis that excluded the demographic variables for potential confounders remained essentially unchanged (eTable 3A-C in Supplement 1).

Figure 3. Mediation Analyses.

The proportion of total effect estimate mediated was calculated by dividing the indirect effect estimate by the total effect estimate, controlled for age, sex, underrepresented minority status, highest educational level of parent or guardian, federal poverty level of the household, and primary language in the household.

Table 2. Mediation Analyses of Association of Deafness or Hearing Problems With School Engagement Outcomes.

Dependent variable	Direct effect estimate (95% CI)	P value	Indirect effect estimate (95% CI)	P value	Total effect estimate (95% CI)	P value	Indirect/direct effect estimatea	Total effect estimate mediated, %b
DHH children with speech and/or language disorders
Extracurricular participation
Sports	−0.080 (−0.136 to −0.025)	.004	−0.032 (−0.041 to −0.022)	<.001	−0.112 (−0.167 to −0.057)	<.001	0.39	28.2
Clubs	−0.045 (−0.101 to 0.010)	.11	−0.030 (−0.040 to −0.021)	<.001	−0.075 (−0.131 to −0.020)	.008	0.67	40.1
Paid work	−0.067 (−0.125 to −0.009)	.02	−0.034 (−0.044 to −0.023)	<.001	−0.101 (−0.158 to −0.043)	.001	0.51	33.6
Volunteer work	−0.077 (−0.121 to −0.032)	.001	−0.024 (−0.032 to −0.016)	<.001	−0.101 (−0.144 to −0.057)	<.001	0.31	23.6
Organized lessons	−0.061 (−0.116 to −0.006)	.03	−0.020 (−0.028 to −0.013)	<.001	−0.082 (−0.136 to −0.027)	.003	0.33	25.0
School performance
Missed school days	0.068 (0.040 to 0.096)	<.001	0.010 (0.005 to 0.014)	<.001	0.077 (0.048 to 0.106)	<.001	0.14	12.3
Did not care about doing well in school	0.068 (0.019 to 0.117)	.006	0.038 (0.027 to 0.049)	<.001	0.106 (0.057 to 0.155)	<.001	0.56	35.9
Did not do required homework	0.095 (0.045 to 0.145)	<.001	0.034 (0.024 to 0.044)	<.001	0.129 (0.078 to 0.179)	<.001	0.36	26.4
Repeated ≥1 school grades	0.039 (0.013 to 0.065)	.003	0.019 (0.013 to 0.026)	<.001	0.058 (0.032 to 0.085)	<.001	0.50	33.3
Parent or guardian contacted by school	0.112 (0.058 to 0.165)	<.001	0.045 (0.032 to 0.058)	<.001	0.156 (0.103 to 0.210)	<.001	0.40	28.6
DHH children with behavioral and/or conduct problems
Extracurricular participation
Sports	−0.087 (−0.142 to −0.032)	.002	−0.020 (−0.027 to −0.012)	<.001	−0.107 (−0.161 to −0.052)	<.001	0.23	18.4
Clubs	−0.053 (−0.109 to 0.002)	.06	−0.017 (−0.023 to −0.011)	<.001	−0.070 (−0.126 to −0.014)	.01	0.32	24.0
Paid work	−0.078 (−0.135 to −0.022)	.007	−0.012 (−0.016 to −0.007)	<.001	−0.090 (−0.147 to −0.033)	.002	0.15	12.9
Volunteer work	−0.081 (−0.126 to −0.037)	<.001	−0.016 (−0.021 to −0.010)	<.001	−0.097 (−0.140 to −0.054)	<.001	0.19	16.1
Organized lessons	−0.060 (−0.114 to −0.006)	.03	−0.017 (−0.023 to −0.011)	<.001a	−0.077 (−0.131 to −0.022)	.006	0.28	21.9
School performance
Missed school days	0.065 (0.037 to 0.093)	<.001	0.009 (0.006 to 0.012)	<.001	0.074 (0.045 to 0.103)	<.001	0.14	12.1
Did not care about doing well in school	0.057 (0.019 to 0.096)	.004	0.045 (0.029 to 0.061)	<.001	0.102 (0.057 to 0.147)	<.001	0.78	44.0
Did not do required homework	0.080 (0.038 to 0.123)	<.001	0.044 (0.028 to 0.059)	<.001	0.124 (0.076 to 0.172)	<.001	0.55	35.3
Repeated ≥1 school grades	0.042 (0.016 to 0.068)	.002	0.013 (0.009 to 0.018)	<.001	0.055 (0.029 to 0.082)	<.001	0.32	24.3
Parent or guardian contacted by school	0.083 (0.030 to 0.135)	.002	0.062 (0.041 to 0.084)	<.001	0.145 (0.093 to 0.197)	<.001	0.75	43.0
DHH children with neuropsychiatric and/or learning disorders
Extracurricular participation
Sports	−0.086 (−0.141 to −0.03)	.002	−0.026 (−0.033 to −0.018)	<.001	−0.111 (−0.166 to −0.057)	<.001	0.30	23.1
Clubs	−0.052 (−0.108 to 0.003)	.07	−0.021 (−0.028 to −0.015)	<.001	−0.074 (−0.129 to −0.018)	.009	0.41	28.9
Paid work	−0.083 (−0.140 to −0.025)	.005	−0.011 (−0.015 to −0.006)	<.001	−0.094 (−0.150 to −0.037)	.001	0.13	11.7
Volunteer work	−0.082 (−0.127 to −0.038)	<.001	−0.017 (−0.022 to −0.011)	<.001	−0.099 (−0.142 to −0.056)	<.001	0.20	16.8
Organized lessons	−0.059 (−0.114 to −0.004)	.04	−0.021 (−0.027 to −0.015)	<.001	−0.080 (−0.134 to −0.025)	.004	0.35	26.1
School performance
Missed school days	0.065 (0.037 to 0.093)	<.001	0.012 (0.008 to 0.015)	<.001	0.077 (0.048 to 0.106)	<.001	0.18	15.1
Did not care about doing well in school	0.054 (0.009 to 0.099)	.02	0.054 (0.039 to 0.069)	<.001	0.108 (0.060 to 0.156)	<.001	1.00	50.0
Did not do required homework	0.078 (0.031 to 0.125)	.001	0.054 (0.039 to 0.069)	<.001	0.132 (0.082 to 0.182)	<.001	0.70	41.1
Repeated ≥1 school grades	0.033 (0.008 to 0.057)	.01	0.026 (0.018 to 0.033)	<.001	0.058 (0.032 to 0.084)	<.001	0.79	44.0
Parent or guardian contacted by school	0.088 (0.036 to 0.139)	.001	0.069 (0.050 to 0.088)	<.001	0.157 (0.104 to 0.210)	<.001	0.79	44.1

Abbreviation: DHH, deaf or hard of hearing.

^a Indirect effect estimate divided by direct effect estimate.

^b The proportion of total effect estimate mediated was calculated by dividing the indirect effect estimate by the total effect estimate, adjusted for age, sex, underrepresented minority status, highest educational level of parent or guardian, federal poverty level of the household, and primary language in the household.

Discussion

This nationally representative cross-sectional study found that DHH children had worse school engagement compared with non-DHH children. This association has previously been hypothesized to be due to HL sequelae (eg, speech/language disorders or behavioral/conduct problems). However, while DHH subgroup analyses revealed associations with HL sequelae, all of the sequelae only partially mediated the association between DHH status and each outcome variable; that is, DHH status had a direct association with worse school engagement outcomes. Thus, we concluded that DHH status, not necessarily simply the sequelae of DHH status, was the primary factor associated with poor school engagement. These results suggest the need for reevaluation of the role DHH status alone plays in educational accommodations and supportive services. Greater attention to hearing-specific accommodations may need to be emphasized. Given this study’s findings, increased accessibility to direct support for hearing, such as hearing aids or cochlear implants, frequency modulation systems, seating changes, and teachers for the deaf, may improve DHH children’s school engagement. Future work should examine underlying factors that may explain the primacy of DHH status in these findings; these underlying factors may include the distinct academic stresses DHH children encounter and the additional burdens these children experience to accommodate their DHH status in the mainstream academic environment (eg, listening-related fatigue).

Moreover, given the current subjectivity of IEP eligibility interpretation, certain IEP teams may choose to focus on HL sequelae as the underlying reasons for differences in educational performance and assume that the absence of HL sequelae in a DHH child (eg, normal scores for speech and language testing or no documented cognitive or learning delays) means that their educational performance is not being affected. Consequently, the DHH child would be deemed ineligible for an IEP, and the available fallback option would be a 504 plan. Yet the accommodations provided by a 504 plan tend to be more generalized, whereas an IEP provides individualized trackable goals, support, and services as part of a specialized education program. Moreover, per the Individuals With Disabilities Education Act,¹³ children with disabilities are required to be allowed to participate in all services, including extracurricular and nonacademic activities, offered by a public school district to a child receiving general education. However, because the act does not define extracurricular and nonacademic activities,¹³ inclusion of accommodations in these settings is potentially vulnerable to variable interpretations and requires advocacy on the part of the IEP writer to describe required accommodations specific to noninstructional settings. Yet, given that 504 plans generally list accommodations, they inherently do not specify accommodations for noninstructional settings.

The results of this study suggest that DHH children, regardless of the presence of HL sequelae, are struggling with school engagement and require additional specialized support. The findings support an interpretation of IEP eligibility that also focuses on the upstream factors, namely DHH status itself, that impact educational performance, which can result in increased IEP accessibility and provision for all DHH children. Advocacy to secure more DHH-specific educational support may ultimately be necessary to change the landscape of school engagement for DHH children.

Limitations

This study has several limitations. First, the survey is intentionally nonmedical, includes a single item relevant to hearing health, and lacks specificity of the terms deafness or problems with hearing, which can be widely interpreted by respondents and combines children with diverse HL characteristics (eg, type and severity) and care needs. Given the absence of audiological data, we made the assumption that parental perception of HL is an indication the child has current HL, which is then associated with school engagement outcomes. Previous studies^{20,21,22,23,24} support this assumption and have found no association between parental perception and objective measures of HL; rather, these studies reported that parents often underestimated the extent of HL, especially for milder forms of HL. Moreover, given the inability to determine HL severity for this sample, we were unable to assess the association of HL severity with HL sequelae and school engagement outcomes. In addition, because all data are based on parental and caregiver reports, as with all surveys, there is potential for bias due to a lack of knowledge and required recall of the past year.

Due to the cross-sectional nature of our study, only associations between variables, not causal links, can be assessed. In any disparity analysis, it is also important to acknowledge deficiencies in inclusion. The NSCH is administered only in the English and Spanish languages and only via mail and online, which omits potential respondents with limited reading skills in these languages, with another primary language, or without reliable internet access; these limitations may also lead to omission of the most vulnerable children. However, despite these omissions, given that our school-aged HL prevalence estimates were substantially lower than in other published reports,^2,3 this data source may in fact underestimate the true burden on school engagement.

Although we controlled for 6 sociodemographic variables, due to sample size limitations and collinearity, we could not account for the entire spectrum of comorbidities (eg, cognitive, medical, and physical conditions) in our primary analysis. However, the results of the mediation models reinforce the major role of DHH status itself in our conclusions. Thus, overall, because of the content and focus of the NSCH, a major strength of this study is the identification of several specific school-related impacts experienced by children with HL, as revealed on a nationally representative level.

Conclusions

In this nationally representative cross-sectional study of children’s health, children with deafness or hearing problems were less likely to participate in 3 of 5 extracurricular activities and more likely to have worse performance in all 5 educational performance variables compared with children without deafness or hearing problems. Moreover, while subgroup analyses of DHH children with and without HL sequelae revealed differences, these sequelae only partially mediated the associations. Thus, these findings suggest that DHH status itself may be the primary factor associated with poor school engagement, which warrants greater emphasis on educational accommodations and support services based on hearing status itself, independent of the presence or absence of HL sequelae.

Supplement 1.

eTable 1. Survey Questions for Variables Included in This Study

eTable 2. Likelihood and Percentage of School Engagement Outcomes

eTable 3A. Unadjusted Mediation Analyses of Children With Deafness or Hearing Problems, Speech-Language Disorder, and School Engagement Outcomes

eTable 3B. Unadjusted Mediation Analyses of Children With Deafness or Hearing Problems, Behavioral/Conduct Problem, and School Engagement Outcomes

eTable 3C. Unadjusted Mediation Analyses of Children With Deafness or Hearing Problems, Neuropsychiatric/Learning Disorder, and School Engagement Outcomes

Supplement 2.

Data Sharing Statement