The Effect of Comorbidities on Cochlear Implantation Outcomes in Adults under 60

Abstract

Introduction

Prior studies have demonstrated that comorbid conditions can negatively impact cochlear implantation (CI) outcomes in elderly patients, but few have examined how comorbidities affect younger adult CI recipients. This study examines the relationship between comorbidities and CI outcomes in adults under 60 years old.

Methods

We reviewed all CI recipients between 20 and 60 years old from 2015 to 2019 at a tertiary academic medical center. Patient data were collected including comorbidities, demographics, etiology, and length of deafness (LOD). Patients’ Charlson Comorbidity Index (CCI) was calculated. The primary outcome was speech perception scores at 1 year on the consonant-nucleus-consonant (CNC) word test.

Results

There were 118 patients who underwent CI (20–29 years [15%], 30–39 years [22%], 40–49 years [21%], 50–60 years [42%]), averaging 1.8 comorbidities. Mean LOD was 19.7 years, and most etiologies were unknown (53.4%). 34% had no comorbidities, and the most frequent comorbidities were hypertension (14%), asthma (10%), anxiety (8%), acoustic neuroma (8%), and arthritis (7%). Comorbidity frequency was similar across ages, but cardiovascular comorbidities varied by patient decade (50–60 years: 41% vs. 20–49 years: 12–22%, p = 0.004). Compared to studies on elderly CI outcomes, our cohort had fewer comorbidities with reduced cardiac events and neurological conditions. We did not find differences in 1-year CNC scores or complications based on the number of comorbidities or any single comorbidity. However, there was a difference in individual improvement in CNC word scores by age group (p = 0.024). Patients’ CCI did not correlate to post-op scores.

Conclusion

Subjects showed improved speech understanding post-CI. The number and type of comorbidities were not meaningful predictors of 1-year speech perception scores, suggesting adult CI users under 60 years with comorbidities can expect comparable outcomes to those without comorbidities.

Plain Language Summary

This study investigated how comorbidities affect outcomes in younger adults (under 60 years old) who receive cochlear implants (CIs). A retrospective review was performed of CI recipients aged 20–60 years old between 2015 and 2019. We collected the following data: patient demographics, hearing loss duration, underlying causes of hearing loss, and comorbidities. The primary outcome was speech perception improvement, measured by consonant-nucleus-consonant word recognition tests one year after implantation. We included 118 patients, and common comorbidities were high blood pressure, asthma, anxiety, and arthritis. Comorbidity rates were similar across age groups, although cardiovascular conditions were more common in patients aged 50–60. Compared to older CI recipients from other studies, our cohort had fewer cardiac and neurological conditions. Results showed that speech perception improved significantly after CI, regardless of the number or type of comorbidities. No strong relationship was found between speech outcomes and number of comorbidities or any single comorbidity. However, age groups differed slightly in individual improvement. Overall, these findings suggest that younger adults with comorbidities can expect similar success with CIs as those without comorbid conditions.

Introduction

Prior studies have demonstrated that comorbid conditions are associated with hearing loss [1] and can negatively impact cochlear implantation (CI) outcomes in elderly patients [2–5]. Specifically, a patient’s number of comorbidities and conditions related to cardiovascular and neurological disease have been found to play a more significant role in speech comprehension and postoperative complications [2, 6, 7]. However, few studies have explored the impact of comorbidities on younger adult patients under 60 years old, who comprise almost 30% of adult CI candidates in the USA [8].

It is prudent to examine this younger adult cohort because the same comorbidity can have a different effect on CI outcomes depending on a patient’s age [1, 9], and younger adults typically have a different comorbidity profile and burden compared to geriatric patients [7, 9]. Furthermore, younger adults often experience different environmental pressures such as being employed and highly social [7, 10, 11]. Because hearing loss is associated with social isolation and early retirement, understanding factors that affect speech outcomes in this age group is particularly important [10, 11].

Though adults under 60 years old undergoing CI overall demonstrate excellent outcomes [12], the level of speech perception does vary, and it can be difficult to predict who may be a poorer performer [13]. While prior studies have established the effect of age, duration of deafness, and hearing loss etiology on CI outcomes, there is limited evidence on the role of comorbid conditions, especially for younger adults [14, 15]. Therefore, we aimed to examine the relationship between comorbidities and CI outcomes in adults aged 20–60 years old.

Methods

We performed a retrospective review of the electronic medical records of all CI recipients aged 20–60 years old at a single academic medical center from 2015 to 2019. This study was granted an exemption from requiring written informed consent from subjects and was approved by the New York University Institutional Review Board (i23-00952). Patients with single-sided deafness and those who underwent revision or sequential bilateral surgery were included. Patients’ comorbidities, demographic characteristics, hearing loss etiology, duration of deafness, hearing loss progression, preoperative hearing aid use, and preoperative aided consonant-nucleus-consonant (CNC) word score were collected. Only CNC scores were used because these scores equate to sentences in noise (HINT) scores, and very few patients had both pre-op and post-op HINT scores. We calculated each subject’s Charlson Comorbidity Index (CCI) score [16]. The CCI is a commonly used, validated metric for quantifying comorbidities in patients based on age, history of MI, congestive heart failure, peripheral vascular disease, stroke, dementia, chronic obstructive pulmonary disease, connective tissue disease, peptic ulcer disease, liver disease, diabetes, hemiplegia, chronic kidney disease, solid tumor, leukemia, lymphoma, and AIDS [16]. Patient comorbidities were also characterized according to the following: cardiac, otologic, neurologic, autoimmune, gastrointestinal, pulmonary, renal, and cancer (online suppl. Table 2; for all online suppl. material, see https://doi.org/10.1159/000546569). The primary outcome of the study was improvement in speech perception scores from preoperatively to 1-year post-activation on the CNC word test. Secondary outcomes included complications (intraoperatively, 30 days post-op, and 6 months post-op) and need for revision.

Statistical analyses included χ², ANOVA with Tukey correction for multiple comparison of means, and Pearson’s correlation coefficient. A generalized linear model was developed for change in 1-year speech perception scores using variables significant on one-way univariate analysis. A p value <0.05 was considered statistically significant.

Results

Patient Demographics and Comorbidities

118 patients underwent CI (n = 18 (15%) 20–29 years, n = 26 (22%) 30–39 years, n = 25 (21%) 40–49 years, n = 49 (42%) 50–60 years), and 60% were female. The average duration of hearing loss was 19.7 years, and 78% of patients had progressive hearing loss, hearing which worsened as subjectively noted by the patient or objectively by audiogram. The most common hearing loss etiologies were unknown (53.4%) and neurofibromatosis type 2 (7.6%) (Table 1). 71.2% of patients utilized a hearing aid preoperatively in the to-be-implanted ear for an average of 16.8 years.

Table 1.

Patient demographic and hearing characteristics

Characteristic	20–29, N = 181	30–39, N = 261	40–49, N = 251	50–60, N = 491	p value2
Age	25 (24, 27)	33 (32, 35)	45 (43, 47)	55 (52, 57)	<0.001
Sex					0.7
F	10 (56%)	15 (58%)	14 (56%)	31 (63%)
M	8 (44%)	11 (42%)	11 (44%)	18 (38%)
BMI	22.5 (21, 28)	24.3 (22, 25)	23.7 (22, 27)	25.8 (24, 30)	0.04
Surgery type					0.4
Bilateral	1 (5.0%)	2 (8%)	2 (8%)	2 (4%)
First	9 (50%)	16 (62%)	14 (56%)	32 (65%)
Revision	5 (28%)	1 (4%)	3 (12%)	2 (4%)
Sequential	2 (11%)	7 (27%)	6 (24%)	11 (22%)
Simultaneous bilateral	1 (6%)	0 (0%)	0 (0%)	2 (4%)
Single-sided deafness	1 (5.0%)	4 (15%)	6 (24%)	11 (22%)	0.37
Unknown etiology	7 (39%)	14 (54%)	11 (44%)	30 (61%)	0.3
NF2 neuroma hearing loss etiology	3 (17%)	2 (8%)	4 (16%)	4 (8%)	0.6
Length of deafness	22 (16, 24)	30 (9, 31)	10 (3, 26)	18 (5, 31)	0.4
Hearing aid use	15 (83%)	18 (69%)	18 (72%)	33 (67%)	0.6
Hearing aid implanted years	20 (8, 23)	16 (3, 29)	17 (2, 27)	11 (5, 27)	0.6
Number of comorbidities	2 (0, 4)	1 (0, 2)	1 (0, 2)	1 (0, 3)	0.2
Charlson Comorbidity Index (CCI)	0 (0, 0)	0 (0, 0)	0 (0, 0)	1 (1, 2)	<0.001
Cardiovascular	4 (22%)	2 (7.7%)	3 (12%)	20 (41%)	0.004
Otologic	5 (28%)	7 (27%)	6 (24%)	8 (16%)	0.6
Neurologic	4 (22%)	3 (12%)	5 (20%)	9 (18%)	0.8
Autoimmune	2 (11%)	1 (4%)	4 (16%)	12 (24%)	0.1
GI	2 (11%)	0 (0%)	3 (12%)	4 (8%)	0.4
Pulmonary	4 (22%)	5 (19%)	1 (4%)	6 (12%)	0.3
Renal	1 (5.0%)	1 (4%)	0 (0%)	4 (8%)	0.5
Cancer	1 (6%)	0 (0%)	3 (12%)	5 (10%)	0.3
Pre-op CNC words score	21 (0, 32)	4 (0, 12)	17 (0, 30)	0 (0, 12)	0.04

¹Median (IQR), n (%).

²Kruskal-Wallis rank-sum test, Pearson’s χ² test, Fisher’s exact test.

Of the entire cohort, 34% of patients had no comorbidities. The most frequent comorbidities were hypertension (14%), asthma (10%), anxiety (8%), acoustic neuroma (8%), and arthritis (7%) (online suppl. Table 1). Comorbidity frequency was similar across age groups, but the category of cardiovascular comorbidities varied by patient decade (50–60 years: 41% vs. 20–49 years: 12–22%, p = 0.004) (Fig. 1). Patients’ median CCI score was 1 (IQR: 0.1) (Table 1).

Fig. 1.

Comorbidity frequency by age group.

CI Outcomes and Complications

There were 52 patients (44.1%) who had 1-year post-op CNC word scores. Overall, patients demonstrated a mean improvement in CNC words scores of 53.4% (SD: 25.7). There were no differences in hearing outcomes between those undergoing first, bilateral, or revision CI, nor between those with or without single-sided deafness. For the entire patient cohort, there were no differences in 1-year CNC scores based on the number of comorbidities; however, patients aged 50–60 showed the greatest improvement, with an average CNC word score increase of 65% (Fig. 2). According to Tukey’s adjustment for multiple comparisons, this improvement was 25.9% higher than that of patients aged 40–49 (p = 0.046). Within each patient age group (20–29, 30–39, 40–49, 50–60), there were no differences in CNC word score improvement based on number of comorbidities (Table 2).

Fig. 2.

Change in CNC word scores by number of comorbidities (0, 1, 2, 3+) (a) and age group (b).

Table 2.

Cochlear implantation intra- and postoperative outcomes

Characteristic	20–29, N = 181	30–39, N = 261	40–49, N = 251	50–60, N = 491	p value2
Anesthesia					0.6
General	18 (100%)	25 (96%)	25 (100%)	49 (100%)
Local	0 (0%)	1 (4%)	0 (0%)	0 (0%)
Intra-op complications	0 (0%)	0 (0%)	0 (0%)	0 (0%)	1
Complication within 30 days post-op	7 (39%)	5 (19%)	6 (24%)	5 (10%)	0.06
Complication within 6 months post-op	0 (0%)	0 (0%)	0 (0%)	0 (0%)	1
Complication at any time point	7 (39%)	5 (19%)	6 (24%)	5 (10%)	0.06
Δ 1-year post-op CNC words score	54 (12, 56)	58 (30, 66)	39 (27, 50)	64 (45, 79)	0.006
1-year post-op CNC words score	67 (57, 78)	62 (46, 74)	74 (46, 77)	78 (58, 84)	0.5
1-year post-op CNC phonemes score	82 (75, 90)	81 (63, 83)	87 (67, 87)	86 (77, 93)	0.5
1-year post-op AzBio Quiet score	81 (68, 91)	80 (57, 93)	89 (78, 94)	84 (73, 95)	0.8
1-year post-op AzBio Noise score	57 (52, 78)	69 (52, 85)	88 (67, 90)	72 (52, 82)	0.5
CI revision	1 (7%)	2 (9%)	1 (4%)	3 (6%)	0.9

¹Median (IQR), n (%).

²Kruskal-Wallis rank-sum test, Pearson’s χ² test, Fisher’s exact test.

Bold indicates statistically significant p values.

On multivariable analysis, age and autoimmune conditions were the only significant predictors, and they demonstrated positive and negative effects on CNC word scores, respectively (beta = 0.906, p < 0.001; beta = −16.83, p = 0.036) when also accounting for NF2 as an etiology/comorbidity and otologic comorbidities in general (Table 3). The multivariable analysis suggests that CNC scores improve with increasing age. When examining CNC improvement according to age decades, this positive relationship is mostly redemonstrated, except for a reduction in CNC score improvement in those aged 40–49 (20–29 years [54%], 30–39 years [58%], 40–49 years [39%], 50–60 years [64%]). Patients’ CCI score was not a significant predictor for CNC word scores when adjusted for age and length of deafness. Other correlations between specific comorbidities and CNC word scores can be seen in Figure 3.

Table 3.

Univariate and multivariate linear regression

Determinants	Δ 1-year post-op CNC words
	univariate		multivariate
	beta	p value	beta	p value
Age	0.825	0.002	0.906	0.000
BMI	0.019	0.979
Etiology of NF2 and/or neuroma	−31.958	0.007	−37.26	0.091
Length of deafness	0.135	0.607
Progression of deafness	12.090	0.501
Charlson Comorbidity Index	−0.974	0.67
Hearing aid use	11.529	0.155
Number of comorbidities	−2.340	0.075
Comorbidity categories
Cardiovascular	−3.323	0.662
Otologic	−20.304	0.012	−8.602	0.336
Neurologic	−7.793	0.379
Autoimmune	−17.756	0.024	−16.83	0.036
GI	−10.178	0.371
Pulmonary	−8.408	0.461
Renal	−1.414	0.922
Cancer	−15.449	0.135
Individual comorbidities
NF2	−29.006	0.037	30.93	0.212
Asthma	−18.958	0.126
GERD	7.074	0.625
Hypertension	9.943	0.26
Anxiety	18.144	0.104
Hyperlipidemia	0.885	0.938
Arthritis	−6.367	0.66
Migraine	−1.04	0.953
Hypothyroidism	−12.391	0.274
Allergies	−8.125	0.52
Depression	11.319	0.432

Bold indicates statistically significant p values.

Fig. 3.

Correlation heat map between hearing characteristics, comorbid conditions, and post-op speech perception outcomes.

There were no intra-op complications; all recorded complications (n = 23) occurred within 30 days post-op. 30-day complications included vertigo (n = 12), pain (n = 5), wound dehiscence (n = 1), facial paresis (n = 1), CSF leak (n = 1), dermatitis (n = 1), and hematoma (n = 1). There were no complications related to complete insertions of the electrode array. No specific comorbidity or category of comorbidities was associated with a post-op complication. There was no difference in likelihood of complication between age groups (p = 0.06) or according to the number of comorbidities (0, 1, 2, 3+). Patients’ CCI score did not correlate to any post-op complications.

Discussion

Our study of 20- to 60-year-old CI recipients demonstrated a statistically significant improvement in 1-year CNC word scores. One-year speech perception scores did not differ according to number of comorbidities. Improvement in CNC word scores were different for patients between 40–49 and 50–60 years old, and on multivariable analysis, age and the category of autoimmune comorbidities were positive and negative significant predictors, respectively. Compared to previous studies of adult CI users which have demonstrated either positive partial correlations or negative correlations between length of deafness and speech perception scores, deafness duration did not significantly affect CNC word score improvement in our cohort [17–19]. We found no difference in post-op complications based on comorbid conditions or patient age.

Our study demonstrated that number of comorbidities does not significantly affect 1-year speech perception scores in younger adults. This was also true when examining within patient age groups of 20–29, 30–39, 40–49, and 50–60. In comparison, Giourgas et al. [20] demonstrated that elderly patients with no comorbidities had higher 12-month Freiburger Monosyllables Test scores than those with one or multiple comorbidities, and Hochmair-Desoyer/Schulz/Moser-Sentence test scores in noise were lower in those with multiple comorbidities versus one or no comorbid conditions. However, no statistical tests confirmed this relationship, and their results emphasized that neurological conditions were more impactful than multimorbidity. Though Lee et al. [2] also suggested that multimorbidity is correlated with worse performance, their study was restrained to limited CI performers with a mean age of 71.2 years at implantation, longer duration of hearing loss (mean 26.4 years), and higher proportion of cardiac comorbidities.

The effect of multimorbidity in younger adults may differ from literature on geriatric patients because older adults have a different comorbidity profile than younger adults with a greater overall number of comorbidities [5, 6] and a higher comorbidity burden in the areas of cardiac events [5, 6] and neurological insults [7, 20]. These differences in findings may be better explained by the types of comorbid conditions that are most prevalent in each age group rather than the number of comorbid conditions itself. For example, Wilkerson et al. [7] found the most common comorbidities in elderly patients to be hypertension, coronary artery disease, diabetes mellitus type 2, atrial fibrillation, chronic obstructive pulmonary disease, aortic valve stenosis, pulmonary fibrosis, and aortic and cerebral aneurysm [20]. The types of disease processes present in younger adults may tend to be more benign and thus have a lesser impact on outcomes [1]. Furthermore, the impact of a patient’s quantity of comorbid conditions might be mediated by their medication burden. Older adults are more sensitive to polypharmacy and thus medications for comorbid conditions may lead to increased general morbidity [2, 21]. Additionally, in our patient cohort, relatively few patients had significant multimorbidity with 35% of patients having no comorbidities and the average number of comorbidities being less than two. In a subject group with a greater overall number of comorbidities as well as a wider range of comorbidities, it is possible that we may have seen a greater effect from the number of comorbidities on speech outcomes. Because our study indicates no definitive comorbidity threshold that impacts post-CI speech outcomes in younger adults, patient counseling may instead shift to emphasis on recognizing notable comorbidities like severe cardiac or neurological conditions.

Nevertheless, no individual comorbidity (e.g., hypertension) was associated with worse speech perception scores in our patient cohort. On multivariable analysis, the autoimmune comorbidity category had a negative effect on speech perception. This contrasts with prior literature of limited performers and elderly adults which instead showed that cardiovascular [22], cancer, and neurological comorbidities are associated with inferior post-CI speech outcomes [20]. The negative effect of autoimmune conditions is likely erroneous given our small sample size as well as the underlying relationship between age and the development of arthritis, which was the most common autoimmune comorbidity in our patient cohort, which was not present in anyone under 40 years old. Furthermore, although our study captured many common comorbid conditions in middle-aged adults such as hypertension, asthma, anxiety, and hyperlipidemia, other conditions such as heart failure, cognitive impairment, and chronic obstructive pulmonary disease were not well represented. Similarly, for neurological comorbidities, our study included migraine (n = 8), seizures (n = 5), and headache (n = 3), without many ischemic conditions or conditions of aging including cognitive impairment often included in prior studies on elderly patients [7, 23]. Because our study did not find the known comorbidities in the elderly population to exist in our groups, we could not make a direct comparison between the two age groups. However, depression (n = 6) and anxiety (n = 10) have been previously identified as comorbidities that significantly impact hearing-related quality of life post-CI more so in middle-aged adults than older adults [9, 24]. Even though these psychiatric disorders may not be predictive of differential speech outcomes in our cohort, they may be important comorbidities to consider when managing a patient’s expectations for their CI outcomes and satisfaction [25, 26].

The Charlson Comorbidity Index (CCI), a weighted scoring system that predicts mortality in patients with specific comorbid conditions, served as a control for our results, although it was not associated with complications nor post-op speech outcomes [16]. This builds on a prior study which showed that CCI (beta <0.150) does not significantly affect 1-year post-CI quality of life measured by the Nijmegen Cochlear Implant Questionnaire, especially when compared to preoperative stress level (beta = −0.495, p < 0.001) [25]. In our patient cohort, most patients had a CCI score between 0 and 2, which perhaps was not a wide enough range to differentiate patients and see an effect on speech perception outcomes. CCI and frailty [27, 28] may not be a helpful tool in predicting CI outcomes for younger adults in part due to the overall better state of health and lower mortality risk, and future research into scoring systems with more utility in predicting outcomes for this age group would be valuable.

Lastly, our study did not demonstrate an association between the number or type of comorbidities with postoperative complications. Overall, there were no complications intra-op or within 6 months post-op, which is largely reflective of complication rates described in the literature [29]. Additionally, the most common 30-day complications included vertigo (n = 12) and pain (n = 5). Like prior literature, the number of comorbidities or specific comorbid conditions, such as hypertension and OSA, did not significantly affect post-op complication rates [7]. Clinically, comorbidities, such as MI or stroke, may play a greater role in assessing anesthesia risk rather than operative complications [29]. Our study reinforces that a patient’s comorbidities should not increase the likelihood of complications.

The study’s limitations include its sample size. Of the 118 patients included, only 44.1% of the patients had 1-year post-op CNC word scores. Another limitation of this study was accounting for the effect of hearing loss etiology. We found that an etiology of NF2 or acoustic neuroma was significant on univariate analysis for CNC word improvement score, unlike a prior study [30]. However, most patients in our cohort had an unknown underlying cause of hearing loss. Therefore, we could not fully explore the potential interaction between hearing loss etiology and CNC word score as well as its potential interaction with patient comorbidity burden. Of note, the difference in CNC word score between those aged 40–49 and 50–59 years old could be due to such unknown underlying etiologies of hearing loss or sampling error [31]. Furthermore, many common medical conditions [32] were not present in our cohort, and we were unable to classify comorbidities according to disease severity which may have provided a more nuanced understanding of their impact on CI outcomes. For instance, someone’s level of disability from a condition like migraines can vary widely. Like many existing studies on geriatric CI patients [7], our study is inherently limited in including those with comorbidities who were still “healthy enough” to undergo CI. However, this study still provides useful information as these healthier patients are the ones typically seeking CI. In conclusion, subjects showed improved speech understanding post-CI. No specific comorbidity nor number of comorbidities were clinically meaningful predictors of 1-year speech perception scores, suggesting adult non-elderly CI users with comorbidities can expect comparable outcomes to those without comorbidities.

Statement of Ethics

The study was approved and was granted an exemption from requiring written informed consent from subjects by the New York University Institutional Review Board, approval No. i23-00952.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This study was not supported by any sponsor or funder.

Author Contributions

J.A.S.: writing – original draft, data collection, and data analysis. C.S.: writing – original draft and editing and data collection. A.K.: writing – review and editing, data collection, and data analysis. E.R.S.: methodology, supervision, and data collection. S.B.W.: writing – review and editing, conceptualization, and supervision.

Funding Statement

This study was not supported by any sponsor or funder.

Data Availability Statement

The study’s data are not publicly available to protect the privacy of research participants but can be made available from the corresponding author (J.A.S.) upon reasonable request.