Sensorineural hearing loss among type 2 diabetic patients and its association with peripheral neuropathy: a cross-sectional study from a lower middle-income country

Abstract

Introduction

Despite potential links between diabetes and sensorineural hearing loss (SNHL), routine hearing assessments for diabetic patients are not standard practice. Our study aimed to investigate the prevalence of SNHL and its association with diabetes-related factors among patients with type 2 diabetes mellitus (T2DM).

Research design and methods

This cross-sectional study was conducted at the Diabetes Clinic, Jinnah Postgraduate Medical Centre, Karachi, Pakistan, from May to September 2021. A total of 396 patients fulfilling the inclusion criteria participated after informed consent. Data collection involved a sociodemographic profile, Michigan Neuropathy Screening Instrument examination followed by pure-tone audiometry and laboratory tests including haemoglobin A1C (HbA1c). HL was defined using better ear four-frequency pure-tone average of ≥26 dB HL and graded as per WHO criteria. Statistical analyses were performed using SPSS. χ², independent t-test and multinomial logistic regression analyses were applied. P<0.05 at 95% CI was considered significant.

Results

Our study revealed a high prevalence of SNHL among patients with T2DM. Mild HL was seen in 55.8%, while 18.7% suffered from moderate HL. Common audiological symptoms included difficulty understanding speech in noisy surroundings (44.2%), balance problems (42.9%), sentence repetition (35.9%), tinnitus (32.3%) and differentiating consonants (31.1%). Hearing impairment predominantly affected low (0.25–0.5 kHz) and high (4–8 kHz) frequencies with a significant difference at 4 kHz among both sexes (t (394)=2.8, p=0.004). Peripheral neuropathy was significantly associated with SNHL on multinomial logistic regression after adjusting with age, sex, body mass index and the presence of any comorbidities. Diabetes duration, HbA1c or family history of diabetes was found unrelated to SNHL severity.

Conclusions

The study highlights the substantial prevalence of SNHL among patients with T2DM and emphasises the importance of targeted audiological care as part of a holistic approach to diabetes management. Addressing HL early may significantly improve communication and overall quality of life.

Strengths and limitations of this study.

• Objectively measures peripheral neuropathy and sensorineural hearing loss among patients with type 2 diabetes mellitus.

• Potential confounders are controlled by patient exclusion (age related, noise/drug induced, etc) and during analyses through statistical methods.

• Non-availability of control group and observational study design unable to link causative association.

• Patient-reported symptoms introduce recall bias.

• Non-probability convenience sampling limits the generalisability of the findings.

Introduction

Diabetes mellitus (DM) stands as a persistent metabolic ailment marked by elevated blood sugar levels.¹ As of the year 2021, the International Diabetes Federation estimates that the global count of individuals grappling with diabetes has reached an estimated 537 million, with projections indicating a further rise.² It is noteworthy to highlight that surveys have revealed a significant shift in diabetes prevalence within the UK, doubling from 2.39% in 2000 to 5.32% in 2013.³ Furthermore, the prevalence of prediabetes has surged from 11.6% in 2003 to 35.3% in 2011.⁴ Forecasts indicate a substantial 67% upswing in diabetes prevalence within lower middle-income countries (LMIC) from 2010 to 2030.⁵

Complications arising from diabetes constitute a principal contributor to untimely morbidity and mortality on a global scale.^{6 7} These complications span both macrovascular and microvascular domains with diabetic nephropathy, neuropathy, visual loss and cardiovascular deaths being well-known contributors to healthcare budgets. Ears are no exception; phenomena such as the thickening of the basal membrane of the cochlea’s stria vascularis capillaries on the lateral wall, alongside other microvascular and neuropathic changes, may lead to sensorineural hearing loss (SNHL).^8–10

The presence of auditory deficiencies among patients with diabetes might be correlated with coexisting conditions like compromised immunity, which heightens susceptibility to infections of the outer, middle or inner ear. Persistent infections like malignant otitis externa, which impact the external ear, are frequently observed among patients with diabetes and often coincide with structural damage.¹¹ This phenomenon may be elucidated by the acceleration of hearing apparatus degeneration spurred by hypertension.¹² Investigations have displayed an expedited hearing disability in individuals under age 60 years.¹³

WHO approximates that over 360 million individuals, equivalent to about 5% of the global populace, endure incapacitating hearing loss (HL),¹⁴ a figure on the rise due to the ageing of populations.¹⁵ Given the irreparable nature of SNHL, identifying preventable origins of such impairment assumes paramount importance in clinical and public health realms.¹⁶

Many morbidity and mortality indices have been devised and validated for diabetes-related complications. However, none of them seems to be inclusive of the audiological aspect of diabetes-related inner ear pathology.¹⁷

Our study, conducted at the Diabetes Clinic, Jinnah Postgraduate Medical Centre (JPMC), Karachi, aimed to assess the frequency of SNHL and its associations with diabetes-related factors among individuals aged 18–60 years. The study involved the Michigan Neuropathy Screening Instrument (MNSI) which comprised neurological evaluation, monofilament testing, vibration perception tests and ankle reflex assessments. The study’s outcomes are intended to provide insights into hearing disabilities among patients with T2DM and enable us to inform potential interventions, especially in the context of a public-sector hospital in a LMIC.

Methodology

Patient and public involvement

How was the development of the research question and outcome informed by their priorities, experience and preferences? How did you involve patients in the design of this study? Were patients involved in the recruitment or conduct of the study?

No patient involved

How were (or will) patients/public be involved in choosing the methods and agreeing on plans for dissemination of the study results to participants and linked communities?

Participating patients consented to disseminate the results of the study in an anonymised form through scientific journals, meetings and public awareness seminars.

This cross-sectional study was conducted at Diabetes Clinic, Ward 7, JPMC, Karachi, Pakistan. Data were collected from May 2021 to September 2021 through non-probability convenience sampling. Jinnah Sindh Medical University issued the institutional review approval (Ref: JSMU/IRB/2021/–441). A study from Southern Punjab¹⁸ reported the prevalence of HL among diabetic patients to be 46.1%. Using this as the anticipated frequency at a 95% CI and p<0.05, a sample size of 382 was calculated through the OpenEpi calculator. After adding 10% as non-response, a sample size of 425 was obtained. Patients with T2DM diagnosed as per American Diabetes Association criteria, aged between 18 and 60 years, were eligible for the study. Exclusion criteria comprised of patients having upper respiratory tract infection, history of ototoxic drugs (eg, antituberculosis drugs), recent ear surgery, occupational noise exposure, family history of otosclerosis, use of hearing aid and unilateral HL (defined by WHO-proposed hearing impairment grade as <20 dB HL in the better ear and ≥35 dB HL in worse ear based on pure-tone audiometry (PTA)). Subsequently, patients demonstrating discharging ear, perforated tympanic membrane and impacted earwax on otoscopic examination were also excluded from the study.

After written informed consent, data were collected in anonymised form using a structured proforma. Sociodemographic history, clinical, neurological, audiological and laboratory parameters, followed by quality-of-life aspects, were recorded. Demographic and clinical evaluation included age, sex, body mass index (BMI), waist circumference, blood pressure, diabetes-related history and hearing symptoms. Blood samples for HbA1c, fasting blood sugar, complete blood count, Urea/ Creatinine/ Electrolytes, Liver Function Tests, total cholesterol (TC) and triglycerides were obtained.

All patients underwent neurological evaluation using MNSI. Neuropathy was classified with a cumulative score of ≥7.

Monofilament testing was done using a 10 g monofilament pressed perpendicularly against the surface until buckled for a brief period (<2 s). Tested areas were the dorsum of big toes and the surfaces below the first, second and fifth metatarsal heads. 8 of 10 correct responses with eyes closed were graded normal, and 1–7 indicated reduced sensations, while no correct answer was regarded as absent (scores of 0, 0.5 and 1 were assigned respectively).

The vibration perception test was done bilaterally with a tuning fork of 128 Hz kept over the dorsum of the big toe, between the nail bed and the distal interphalangeal joint. The test was considered ‘normal’ if the examiner felt a vibration in the forefinger for <10 s after the patient stopped feeling vibration at their big toes. Examination finding was classified as ‘decreased’ when vibrations felt >10 s by the examiner. ‘Absent’ vibration perception was labelled when the patient had no sensation (scores of 0, 0.5 and 1 were assigned respectively).

Ankle reflex assessment was done using a clinical hammer with the patient sitting relaxed and feet in a neutral position. The Achilles tendon was percussed directly in a slight dorsiflexed foot. If the reflex was obtained, this was counted as normal. If elicited with the Jendrassik manoeuvre, it was recorded as reduced and considered absent when not obtained, even after the Jendrassik manoeuvre (scores of 0, 0.5 and 1 were assigned, respectively).

Audiological examination was conducted by a professional audiologist in a calm room, through a portable audiometer. Testing frequencies for air conduction were 0.25 kHz, 0.5 kHz (low tone), 1 kHz, 2 kHz (middle tone), 4 kHz and 8 kHz (high tone). For bone conduction, testing frequencies were 0.5 kHz, 1 kHz, 2 kHz and 4 kHz. An air-bone gap of ≥15 dB HL at any of these frequencies was set as a threshold for conductive HL. We categorized HL as per WHO criteria: mild, 26–40 dB HL; moderate, 41–60 dB HL; severe, 61–80 dB HL; and profound, ≥81 dB HL. Better ear four-frequency (0.5 kHz, 1 kHz, 2 kHz and 4 kHz) pure-tone average (4 f PTA) was calculated for both ears, and the ear with lower 4 f PTA was termed the ‘better ear’ and used to categorise the HL. Disabling HL was >40 dB HL in the better ear (grades 2, 3 and 4).

We used SPSS V.23 (IBM Corp) for data entry and statistical analyses. Categorical data (eg, sex, family history of diabetes, complications, audiological symptoms, etc) were expressed as numbers and percentages. Mean, median, mode, SD, SE of mean, IQR, percentages and proportions were used for quantitative variables. Univariate analyses were conducted to determine significant variables among patients with T2DM with normal and impaired hearing. χ² test was applied for categorical variables. The independent t-test was applied to compare mean hearing thresholds with normal and abnormal MNSI scores. A multinomial logistic regression model was constructed to find significant associations for SNHL for diabetes-related factors adjusted with age, sex, BMI and the presence of any comorbidities (history of ischaemic heart disease and hypertension). All analyses were conducted at 95% CI, and p<0.05 was deemed statistically significant.

Results

Out of 436 participants, 396 patients with T2DM fulfilled the inclusion criteria. Patients were excluded because of unilateral HL (2), conductive or mixed HL (4), ototoxic drugs history (6) and incomplete forms (19). The majority of patients were females (305/396). Table 1 specifies the sociodemographic, clinical and metabolic parameters of the study population.

Table 1.

Sociodemographic, clinical and metabolic parameters of patients with T2DM (n=396)

Variables	N (%)/M (±SD.D)
Females (N (%))	305 (77%)
Age (years)	48.61 (±9.32)
Duration of diabetes (years)	10.29 (±6.13)
Family history of diabetes	145 (36.6%)
Mode of treatment
Oral hypoglycaemic agents	216 (54.5%)
Insulin	89 (22.5%)
Both	91 (23%)
Comorbidities
History of IHD	75 (18.9%)
HTN*	278 (70.2%)
Substance abuse
Pan eater	28 (7.1%)
Gutka** (betel quid)	12 (3%)
Chalia (areca nut)	31 (7.8%)
Naswar†	20 (5.1%)
Smoking	16 (4%)
SBP (mm Hg)	125.27 (±17.4)
DBP (mm Hg)	78.23 (±10.78)
BMI (kg/m2)	27.57 (±5.31)
WC (cm)	101 (101)
HbA1c (%) (mean±SD.D)	9.4 (±2.25)
FBS (mg/dL)	231.17 (220)
TC (mg/dL)	188.2 (185)
TG (mg/dL)	184 (184)
HDL-C (mg/dL)	43.6 (42)
LDL-C (mg/dL)	111.8 (110)
Right ear 4fPTA (dB HL)	34.93 (±10.31)
Left ear 4 f PTA (dB HL)	34.57 (±9.8)
MNSI history score (mean±SD.D)	5.81 (±2.8)
Normal (<7)	217 (54.8%)
Abnormal (>7)	179 (45.2%)
MNSI exam score (mean±SD.D)	2.3 (±1.9)
Monofilament (normal)	245 (61.8%)
Vibration perception (normal)	225 (56.8%)
Ankle reflex (normal)	313 (78%)

Values are expressed as frequencies, M (mean, mode and median), IQR or SD.

*Chewing tobacco preparation made of crushed areca nut, tobacco, catechu, paraffin wax, slaked lime and flavourings (wikipedia.org).

†Moist, powdered or paste tobacco dip (wikipedia.org).

BMI, body mass index; DBP, diastolic blood pressure; FBS, fasting blood sugar; 4 f PTA, four-frequency pure-tone average; HbA1c, haemoglobin A1C; HDL-C, high-density lipoprotein cholesterol; HTN, hypertension; IHD, ischaemic heart disease; LDL-C, low-density lipoprotein cholesterol; MNSI, Michigan Neuropathy Screening Instrument; SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; WC, waist circumference.

Using better ear 4 f PTA, a WHO hearing impairment grade was assigned. One-fourth (25.3%, n=100) had normal hearing thresholds. More than half (55.8%, n=221) of the patients had mild hearing impairment, whereas 18.7% (n=74) of patients suffered from moderate HL. Only one patient had profound HL which was grouped with moderate HL for further analyses.

The most common audiological symptom was difficulty understanding speech in noisy surroundings (44.2%, 175), followed by balance problems (42.9%, 170), need for repetition of sentences (35.9%, 142), tinnitus (32.3%, 128), differentiating consonants (31.1%, 123), difficulty understanding speech in quiet surroundings (23%, 91) and falls (19.4%, 77).

4 f PTA was used to compare the mean thresholds between the patients having these symptoms. The independent t-test was applied to find statistically significant differences in mean thresholds. Patients who reported speech understanding difficulties, proprioception defects and decreased efficiency in their routine tasks had significantly higher mean hearing thresholds than those who did not (p=0.001). Nonetheless, the presence of tinnitus did not affect average thresholds.

Low-tone (0.25–0.5 kHz) and high-tone (4–8 kHz) frequency losses were more prevalent. Middle tones (1–2 kHz) were comparatively less affected in both ears. We also noted that 4 kHz was more affected in males (40.80 dB HL) than in females (36.08 dB HL) bilaterally (t (394)=2.8, p=0.004). Figure 1A,B demonstrates mean pure-tone thresholds (dB HL) at individual frequencies (Hz) among males and females.

Figure 1.

(A)Right ear mean pure-tone thresholds (dB HL) among males and females. (B)Left ear mean pure-tone thresholds (dB HL) among males and females.

The independent t-test for continuous variables and χ² test for categorical variables were performed at a 95% CI between mean pure-tone normal and impaired hearing thresholds. This yielded age, high TC and abnormal MNSI scores significantly associated with SNHL (table 2).

Table 2.

Sociodemographic, clinical and metabolic parameters of patients with T2DM in comparison with normal and impaired hearing (n=396)

	Normal hearing (n=100)	Impaired hearing (n=296)	P value
Males (%)	23 (25.3)	68 (74.7)	0.31
Females (%)	77 (25.2)	228 (74.8)	0.31
Age (years)	44.9 (±9.96)	49.85 (±8.77)	<0.001
Duration of diabetes (years)	9.36 (±6.04)	10.61 (±6.17)	0.07
SBP (mm Hg)	123.38 (1.69)	125.9 (1.02)	0.21
DBP (mm Hg)	79.11 (1.17)	77.9 (.61)	0.34
BMI (kg/m2)	28.06 (±5.84)	27.4 (±5.1)	0.28
WC (cm)	102.8 (1.3)	100.47 (.72)	0.11
HbA1c (%)	9.29 (±2.3)	9.45 (±2.23)	0.52
FBS (mg/dL)	224.69 (10.4)	233.35 (5.29)	0.43
TC (mg/dL)	179.3 (5.04)	191.23 (2.94)	0.04
TG (mg/dL)	196.35 (11.32)	220 (7.32)	0.09
HDL-C (mg/dL)	44.4 (2.24)	43.48 (.78)	0.59
LDL-C (mg/dL)	108.6 (10.06)	112.37 (3.9)	0.72
Right ear 4 f PTA (dB HL)	23.04 (.43)	38.95 (.49)	<0.001
Left ear 4 f PTA (dB HL)	22.85 (.38)	38.59 (.45)	<0.001
MNSI history	5.32 (±2.8)	5.97 (±2.8)	0.04
MNSI exam	1.35 (.15)	2.63 (.15)	<0.001

Values are expressed as frequencies, mean, median, SD or SE of mean.

BMI, body mass index; DBP, diastolic blood pressure; FBS, fasting blood sugar; HbA1c, haemoglobin A1C; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; MNSI, Michigan Neuropathy Screening Instrument; 4 f PTA, four-frequency pure-tone average; SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; WC, waist circumference.

To control the confounding effect of age and duration of diabetes, we constructed a multinomial logistic regression model to determine the odds of SNHL with clinical and metabolic parameters. The models were adjusted for age, sex, BMI and the presence of any comorbidities (table 3). The severity of HL increased significantly with increasing MNSI scores (OR: 0.23 (0.11–0.48), p<0.001), illustrated in figure 2A,B. We also found that the duration of diabetes, sex, HbA1c and family history of diabetes were unrelated to HL severity (p>0.05).

Table 3.

Relationship of characteristics of T2DM with a degree of HL (WHO better ear) (n=396)

Factors	Unadjusted OR (95% CI) Mild HL (n=221; 55.8%)	Adjusted OR (95% CI) Mild HL (n=221; 55.8%)	Unadjusted OR (95% CI) Moderate–severe HL (n=75; 18.9%)	Adjusted OR (95% CI) Moderate–severe HL (n=75; 18.9%)
Family history of T2DM
No (n=251)	1.39 (0.82–2.37)	1.46 (0.84–2.54)	1.78 (0.92–3.44)	1.63 (0.80–3.28)
Yes (n=145)	Reference	Reference	Reference	Reference
P value	0.21	0.17	0.08	0.17
Duration of T2DM
<10 years (n=203)	0.88 (0.53–1.46)	0.95 (0.56–1.62)	0.93 (0.49–1.76)	1.33 (0.67–2.66)
>10 years (n=193)	Reference	Reference	Reference	Reference
P value	0.64	0.86	0.83	0.41
HBA1c
<6.5% (n=41)	0.74 (0.32–1.68)	0.78 (0.33–1.83)	1.05 (0.38–2.87)	0.88 (0.30–2.61)
6.5–7.5% (n=52)	0.88 (0.41–1.87)	0.82 (0.37–1.82)	1.97 (0.82–4.7)	1.58 (0.62–4.03)
>7.5% (n=302)	Reference	Reference	Reference	Reference
P value	0.47	0.57	0.91	0.82
Vision problem
No (n=123)	0.53 (0.31–1.78)	0.68 (0.39–1.12)	0.41 (0.20–0.83)	0.51 (0.23–1.11)
Yes (n=273)	Reference	Reference	Reference	Reference
P value	0.01	0.18	0.01	0.09
MNSI exam score
Normal (n=241)	0.28 (0.16–0.51)	0.30 (0.16–0.55)	0.20 (0.10–0.40)	0.23 (0.11–0.48)
Abnormal (n=154)	Reference	Reference	Reference	Reference
P value	<0.001	<0.001	<0.001	<0.001

Normal hearing thresholds are kept as a reference. Variables are adjusted for age, sex, BMI and the presence of any comorbidities (history of IHD and HTN).

HBA1c, haemoglobin A1C; HL, hearing loss; MNSI, Michigan Neuropathy Screening Instrument; T2DM, type 2 diabetes mellitus.

Figure 2.

(A)Right ear mean pure-tone thresholds (dB HL) for normal and abnormal MNSI examination. (B)Left ear mean pure-tone thresholds (dB HL) for normal and abnormal MNSI examination. MNSI, Michigan Neuropathy Screening Instrument.

Discussion

We found that 74.7% of diabetic patients were experiencing some level of HL. This figure closely mirrors but is greater than 66.2%, as documented by Al-Rubeaan et al.¹⁹ Moreover, it is also greater than the values identified by Bamanie and Al-Noury in Saudi Arabia, Mozaffari et al in Iran, Uchida et al in Japan and Cheng et al in the USA which reported percentages of 52.3%, 30.0%, 39.1% and 46.0%, respectively.^20–23 Among these, mild hearing impairment (55.8%) emerged as the prevailing category, followed by moderate HL (38.2%). Al-Rubeaan et al reported figures of 47.7% and 38.2%, respectively. These results underscore a notable prevalence of HL within the broader diabetic population, emphasising the importance of regular audiological care in comprehensive diabetes management approaches.

The connection between the severity of peripheral neuropathy, indicated by MNSI scores, and the extent of HL was noticed. This observation implies a potential shared underlying mechanism between diabetic neuropathy and vestibule-cochlear damage. Diabetic neuropathy, a frequent complication of diabetes, involves harm to peripheral nerves, including those connected to the auditory system.^{24 25} The combined impact of microvascular alterations, metabolic disruptions and persistent inflammation linked to diabetes could be influential in the development of both peripheral neuropathy and SNHL.^{26 27} Ren et al’s study similarly revealed a parallel association with MNSI (OR=1.38). Additional measures performed in their study, like Semmes-Weinstein monofilament, OR=1.24, and vibration perception threshold, OR=1.19,²⁵ highlight a connection between diabetic neuropathy and HL.²⁵

Our participants’ reported audiological symptoms shed more light on their hearing challenges. Predominantly, comprehending speech in noisy surroundings emerged as the most frequent symptom (44.2%). This highlights the communication hurdles that individuals with diabetes encounter. In addition, sentence repetition (35.9%) and tinnitus (32.3%) were commonly noted symptoms. These observations signify the intricate nature of hearing impairment in diabetes, affecting various dimensions of people’s daily lives, including social interactions, work productivity and overall life quality. Balance problems reported by 42.9% of patients are particularly worrisome, potentially impacting safety and the risk of falls.²⁸ Agrawal et al noted that vestibular dysfunction, independently, increased the odds of falling more than twofold among patients with diabetes.²⁹ The requirement for sentence repetition might signify difficulties in auditory processing and comprehension, potentially leading to breakdowns in communication.³⁰ Tinnitus, characterised by the perception of ringing or buzzing sounds in the ears, can also significantly impede one’s quality of life and well-being.³¹ Mousavi et al noted 26.4% (vs 32.3% in our study) prevalence of tinnitus among diabetic patients in their 2021 study.³²

Analysing audiometric data unveiled distinct frequency loss patterns. Low frequencies (0.25–0.5 kHz) and high frequencies (4–8 kHz) exhibited greater impairment, whereas middle frequencies (1–2 kHz) were relatively less affected. This HL pattern aligns with the heightened vulnerability of the basal and apical cochlear regions in diabetic patients.³³ A meta-analysis in 2014 also identified greater hearing impairment at 6–8 kHz (high tone).³⁴ Further studies showed that low frequencies tend to be affected.^{35 36} Comprehending these alterations is crucial for devising targeted interventions and optimising hearing rehabilitation approaches tailored for individuals with diabetes. HL in low frequencies can disrupt the perception of vowel sounds, pivotal for speech comprehension and effective communication.³⁷ This can result in conversational misunderstandings and diminished speech clarity. Conversely, high-frequency loss impacts the discernment of consonant sounds, essential for distinguishing words with similar vowel patterns.³⁸ This challenge can impede accurate speech comprehension, particularly when consonants convey word meanings. Moreover, high-frequency loss influences the perception of environmental sounds like sirens, alarms and bird songs, pivotal for situational awareness and safety.³⁹ Diabetic individuals with high-frequency HL might overlook these vital auditory cues, heightening the risk of accidents or overlooking crucial warning signals.

Hearing impairment can additionally result in social isolation and withdrawal. Difficulties in following conversations and effectively engaging in group discussions may dissuade individuals from involvement in social events, resulting in emotions of loneliness and isolation. Given that social interactions are crucial for sustaining mental and emotional well-being, unaddressed HL can significantly affect an individual’s overall life of quality.⁴⁰

The study’s findings demonstrated a significant majority of diabetic patients in the sample were females (77.02%). This is consistent with data from other diabetes clinics operating in the morning hours in the country, even though there is a higher male prevalence of diabetes in the general population.⁴¹ A nationwide study in Pakistan reported a distribution of 51.17% males and 48.83% females.²⁶ A reason for this can be the exclusion of elderly males to prevent the overlap of age-related SNHL with diabetes-related SNHL. Additionally, a higher prevalence of metabolic syndrome components among females might contribute to increased diabetes, as noted by Alamgir et al.⁴²

Regarding the correlations between clinical and metabolic parameters and SNHL, the study pinpointed age, elevated TC and abnormal MNSI scores as significant factors. An Indian study indicated that patients with type 2 diabetes and dyslipidaemia exhibited a 65.7% prevalence of HL.⁴³ The role of advanced age as a HL risk has been long acknowledged,³³ and this study reaffirms its link to SNHL in diabetic patients. Another Indian study also noted a high rate of HL in the 16–50 age group, while the 5–15 age group had normal hearing.⁴⁴ The connection between dyslipidaemia, particularly high TC levels, and HL has been reported in several other research studies.⁴⁵ The precise mechanisms behind this relationship remain incompletely understood but potentially involve vascular and metabolic pathways contributing to cochlear damage.^{27 46}

Notably, our study did not find any significant associations between the duration of diabetes, sex, HbA1c levels and family history of diabetes with the severity of HL. HbA1c provides an estimate of time-limited glycaemic control. Similarly, the duration of diabetes does not describe the progression of diabetes. Despite their involvement in diabetic complications like neuropathy,^{17 19 47} these factors might impact SNHL differently across various populations, necessitating further investigation. A 2018 meta-analysis revealed an escalating HL trend with prolonged diabetes duration,⁴⁷ while another study⁴⁸ noted a J-shaped link between HbA1c and high-frequency hearing impairment among baseline participants with diabetes.

Given the high prevalence of SNHL among diabetic patients in our study, we conclude that audiological care and evaluation should be incorporated as part of a holistic approach to diabetes management. This may comprise of surveys, patient education, targeted clinical examination and timely specialist referral. The development of screening algorithms for SNHL among diabetic patients is another potential area of future research.

Detecting hearing issues early enables timely interventions, such as hearing aids or assistive devices, to alleviate hearing-related communication and life-quality consequences. Collaborative efforts among diabetes specialists, audiologists and otolaryngologists are vital for comprehensive diabetic care. Optimal glycaemic control, lipid regulation and cardiovascular risk management might curtail SNHL risk and severity. Furthermore, interventions addressing shared mechanisms behind diabetic neuropathy and SNHL, like neuroprotective agents or lifestyle changes, could potentially delay or prevent HL onset in diabetics.

Limitations

While this study provides valuable insights, it is crucial to acknowledge its limitations. First, due to its cross-sectional design, causal links between diabetes and HL cannot be firmly established. A longitudinal approach would offer more robust evidence regarding the temporal association between diabetes and the development of hearing impairment. Additionally, the use of self-reported audiological symptoms introduces the possibility of recall bias, potentially affecting the accuracy of symptom reporting. Furthermore, the absence of a non-diabetic control group hinders the comparison of HL prevalence between diabetic and non-diabetic individuals. The external validity of the findings is also constrained by the adoption of non-probability convenience sampling, which introduces selection bias. The study’s setting within a specific diabetes clinic in Karachi, a public-sector hospital, limits the generalisability of results to populations seeking care in private hospitals or specialised audiological clinics.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study involves human participants and was approved by Jinnah Sindh Medical University International Review Board (Approval Ref# JSMU/IRB/2021/-441). Participants gave informed consent to participate in the study before taking part.