How to cite this article: Zamora-Vega O, Gómez-Díaz RA, Delgado-Solís M, Vázquez-Estupiñán F, Vargas-Aguayo AM, Wacher-Rodarte NH. [Association between depression and hearing loss in patients with type 2 diabetes]. Rev Med Inst Mex Seguro Soc. 2016;54 Suppl 2:S140-7.
Received: November 2nd 2015
Judged: May 2nd 2016
Omar Zamora-Vega,a Rita Angélica Gómez-Díaz,a Margarita Delgado-Solís,b Felipe Vázquez-Estupiñán,c Alejandro Martín Vargas-Aguayo,d Niels H. Wacher-Rodartea
aUnidad de Investigación en Epidemiología Clínica
bEspecialista en Audiometría y Foniatría
cServicios de Atención Psiquiátrica, Secretaría de Salud
dEx Jefe del Servicio de Otorrinolaringología
a,b,dHospital de Especialidades, Centro Médico Nacional Siglo XXI, Instituto Mexicano del Seguro Social, Ciudad de México, México
Communication with: Rita Angélica Gómez Díaz
Telephone: 55 5627 6900, extension 21481
Background: The association between diabetes, hearing loss and depression is unknown, and needs to be understood clearly and precisely. Our objective was to estimate the association between depression and hearing loss in patients with type 2 diabetes.
Methods: Cross-sectional analytical study. Patients of 40 years or older with type 2 diabetes were studied. Anthropometric measurements and biochemical parameters were recorded. The Beck Depression Inventory and a pure tone audiometry were applied. Groups were compared with chi squared test and logistic regression for confounders.
Results: 150 patients were included (76 % women). Average age was 56 ± 9.3 years, with 12.4 ± 6.5 years of progression of diabetes, weight 67.4 ± 11.6 kg; 31 % were obese; 25.5 % hypertensive (126.3 ± 19.3/79.4 ± 19.7 mm Hg) and 80.7 % had poor metabolic control (HbA1c ≥ 7 %). Of all the patients, 45.3 % presented hearing loss and 32.4 % depression. Diagnosis of depression in patients with hearing loss remained significant after adjusting for confounders (odds ratio [OR] = 2.3; 95 % confidence interval [CI] = 1.051-5.333, p = 0.037). Women had greater risk of depression, difference that remained significant (OR = 3.2; 95 % CI = 1.268-8.584, p = 0.014) after adjustment.
Conclusions: Almost half of the patients with diabetes presented hearing loss and more than three times the risk of depression. Subjects with depression and diabetes presented more hearing loss (> 20 dB) than those without diabetes and/or without depression.
Keywords: Type 2 diabetes mellitus; Hearing loss; Depression
Patients with diabetes are more prone to atherosclerotic disease than the general population. Many of the complications of the elderly appear earlier in patients with diabetes. 1 According to the World Health Organization (WHO), for 2025 it is estimated that 25% of adults aged 65 to 75 years and between 70 and 80% of those over 75 will suffer from sensorineural hearing loss.2 Diabetes can cause hearing loss by microcirculation compromise, neuropathic factors, and chronic hyperglycemia, with deleterious effects on the organ of Corti, 3-5 causing varying degrees of sensorineural hearing loss. 6-8 The most common form of hearing loss in patients with type 2 diabetes is progressive, sensorineural, bilateral, and symmetrical, particularly affecting the high frequencies, but with more severe losses especially in elderly patients. 9 The evaluation of hearing loss by audiometry is a rapid, accessible, and non-invasive method, taking into account the type, degree, and configuration of hearing loss.
On the other hand, a strong relationship is found between hearing loss and depression. 10 This is the mood characterized by loss of interest and the ability to enjoy things, decreased activity levels, and tiredness following minimal exertion, for two or more consecutive weeks. 11 To establish a diagnosis of depression one must have at least two of the following symptoms: decreased attention and concentration; loss of self-confidence and feelings of inferiority; ideas of guilt and of being useless; bleak future outlook; thoughts or acts of suicide and self-harm; sleep disturbances; and loss of appetite. Certain questionnaires are used to identify and quantify the intensity of depressive symptoms or to evaluate treatment. Among them is the Beck questionnaire. 12
The association of diabetes and depression causes poor adherence to treatment, poor metabolic control, increased presence of complications, and increased mortality. If left untreated, these conditions contribute to deteriorating health. The association between diabetes and depression is bidirectional.13 Depression and hearing loss are diseases that lead to life with disabilities. Furthermore, the comorbidity of depression and diabetes increases health costs if compared with diabetes only. 14
The presence of diabetes increases the incidence of depression around 20%, while depression is associated with a 60% increase in the risk of new cases of type 2 diabetes. 15
According to the above, diabetes, like other comorbidities, such as hypertension, cardiovascular disease, sleep disorders, cancer, etc., has been strongly associated with hearing loss and depression. 16
As suggested by Petrak et al., multidisciplinary collaboration between medical and psychological specialties is needed to improve the treatment of patients with diabetes, hearing loss, and depression. 17
Therefore, the aim of our study was to estimate the strength of association between depression and hearing loss in patients with type 2 diabetes.
This is an analytical cross-sectional design. Patients attending their scheduled appointments at the Unidad de Investigación en Epidemiología Clínica of the Hospital de Especialidades of the Centro Médico Nacional Siglo XXI of the Instituto Mexicano del Seguro Social Institute over 12 months (from January to December 2014) were invited to participate.
Patients of both sexes with type 2 diabetes and non-diabetic subjects, 40 years of age or older, who agreed to sign the informed consent were included. The study did not include patients with previous otologic surgeries or illnesses or those under antipsychotic/antidepressant treatment; nor those with a history of acute otologic disease in the last three months, or those who work in noisy environments without the use of hearing protection equipment, or subjects with a history of mental illness or prior diagnosis of depression.
Subjects who did not answer the questionnaires were excluded, as well as those who had no audiometry test or with incomplete data.
After informed consent, an interview and complete clinical examination were performed, recording age, weight, height, sex, and blood pressure. The body mass index (BMI) was calculated according to the Quetelet index (weight/height 2). A peripheral venous blood sample was taken to determine the fraction of A1c glycated hemoglobin (HbA1C), fasting glucose, creatinine, and lipid profile. The Beck questionnaire was applied to diagnose depression; patients were also taken the Otolaryngology department for audiometry.
The cutoff point for depression with the Beck questionnaire was ≥ 12 points according to the total score.
Regarding the statistical analysis, quantitative results are reported as mean and standard deviation; qualitative data were expressed as a percentage. The analysis used Chi-squared or Fisher's exact test according to the type and distribution of the variable. The variables (age, gender, socioeconomic status, HbA1c, and time of disease progression) were categorized and multiple logistic regression analysis was done. A p-value < 0.05 was considered statistically significant.
150 subjects were included. 76% were women (n = 114). The average age and average weight of the total population were 56 ± 9.3 years and 60.6 ± 10.0 kg, respectively. 75 patients had diabetes and had 12.4 ± 6.5 years of disease progression. The average BMI calculated in subjects with diabetes was 28.8 ± 4.4 kg/m2, versus 26.91 ± 3.6 kg/m2 for subjects without diabetes (p = 0.005) (Table I).
|Table I Anthropometric characteristics of study subjects|
( = 75)
( = 75)
(n = 150)
|Mean ± SD||Mean ± SD||Mean ± SD|
|Age (in years)||56 ± 9.3||57 ± 9.4||57 ± 9.3||0.597|
|Weight (in kg)||69.24 ± 12.8||65.73 ± 10.04||67.4 ± 11.6||0.064|
|Height (in m)||1.54 ± 0.09||1.56 ± 0.07||1.55 ± 0.08||0.343|
|BMI (in kg/m2)||28.8 ± 4.4||26.91 ± 3.6||27.87 ± 4.20||0.005|
|SBP (in mm Hg)||125||118-130||120||110-130||120||113-130||0.068|
|DBP (in mm Hg)||80||71-85||80||70-85||80||70-85||0.307|
|SD = standard deviation; BMI = body mass index; Med = medium; IR = interquartile range; SBP = systolic blood pressure; DBP = diastolic blood pressure|
As expected, a significant difference (p < 0.001) was observed when comparing glucose values and HbA1c in subjects without diabetes (86.7 mg/dL and 5.7%), with subjects with diabetes (177.1 mg/dL and 7.9%), respectively. Also, lower mean HDL cholesterol values were noted in subjects with diabetes (46.1 ± 12.5 mg/dL) than in those without diabetes (50.5 ± 14.4 mg/dL) (p = 0.031). The same was observed with triglyceride values (164 mg/dL) versus (128.5 mg/dL) (p = 0.001) Table II.
|Table II Biochemical characteristics of study participants|
(n = 75)
(n = 75)
|Mean ± SD||Mean ± SD|
|Glucose (in mg/dL)||177.1 ± 72.9||86.7 ± 9.9||< 0.001|
|Total cholesterol (in mg/dL)||191.3 ± 46.1||192.7 ± 40.6||0.658|
|HDL cholesterol (in mg/dL)||46.1 ± 12.5||50.5 ± 14.4||0.031|
|LDL cholesterol (in mg/dL)||126.4 ± 37.5||135.0 ± 33.4||0.701|
|HbA1c (%)||7.9||6.58-10.1||5.7||5.5-6.0||< 0.001|
|Creatinine (in mg/dL)||0.73||0.66-0.88||0 81||0.70-0.89||0.130|
|Triglycerides (in mg/dL)||164.0||123.2-209.5||128.5||102.5-159.2||< 0.001|
|SD = standard deviation; HDL cholesterol = high-density cholesterol; LDL cholesterol = low-density cholesterol; HbA1c = A1c glycosylated hemoglobin; IR = interquartile range|
34 patients with diabetes presented hearing loss (45.3%), i.e., that condition was 17.3% more frequent than in subjects without diabetes (28%). The hearing loss was moderate in 6.6% vs. 8% of patients with and without diabetes (p = 0.103), respectively.
21.3% of patients with diabetes (n = 16) presented depression (Beck questionnaire ≥ 12 points), while only 14.7% (n = 11) of those without diabetes had it (p = 0.288). 9.3% of patients with diabetes (n = 7) and 10.6% without diabetes (n = 8) had a score ≥ 17 (0.415). When comparing patients with and without diabetes, 37.3% (n = 28) versus 16% (n = 12) (p = 0.007) presented obesity, respectively- Table III.
|Table III Social and special characteristics of patients|
(n = 75)
(n = 75)
|Diagnosis of hearing loss|
|Degree of hearing loss|
|No hearing loss||41||54.7||54||72||0.103|
|Moderate hearing loss or more||5||6.6||6||8|
The risk of depression in subjects who had hearing loss and diabetes was 3.44 (95% confidence interval [CI] = 1.05-11.2). In subjects with diabetes with low socioeconomic status, the odds ratio (OR) was 3.7 (95% CI = 1.19-12.03).
When the risk was calculated adjusted for sex, age, obesity, years of diabetes diagnosis, and degree of metabolic control, the risk in patients with diabetes and hearing loss increased with an OR of 5.4; also for those who had low socioeconomic status and diabetes, OR increased to 5.91- Table IV.
|Table IV Odds ratio describing the association between depression and hearing loss in patients with and without diabetes|
|Variable||No diabetes||No diabetes||Diabetes||Diabetes|
|Raw OR||95% CI||Adjusted OR||95% CI||Raw OR||95% CI||Adjusted OR||95% CI|
|Low socio-economic statusb||2.6||0 72-9.6||3.03||0.75-12.2||3.7||1.19-12.03||5.91||1.45-24.1|
|Age ≥ 60 yearsd||1.01||0.26-3.8||0.96||0.22-4.1||1 92||0.61-5.9||1.72||0.40-7.3|
|Evolution ≥ 14 yearsf||---||---||---||---||1.4||0.43-4.5||1.22||0.28-5.3|
|≥ 7% HbA1cg||---||---||---||---||1.5||0.43-5.3||0.75||0.16-3.4|
|Compared with the variable tono hearing loss, b medium socioeconomic status, cman, d< 60 years old, enot obese, f< 14 years evolution, gHbA1c < 7%|
Logistic regression analysis was made using depression according to the Beck questionnaire (≥ 12 points) as a dependent variable. After adjusting for age, gender, BMI, time of evolution, HbA1c, HDL-C, LDL-C, and triglycerides, the independent association between hearing loss and depression was confirmed (beta = 0.748, OR 2.113, 95% CI = 1.003-4.449; p = 0.049). Gender had a borderline association (beta 0.703 with OR 2.20 [p = 0.082, 95% CI .914-4.464]) Table IV.
|Table V Logistic regression analysis for the association between hearing loss and related variables|
|Beta *||Exp-beta||95% CI||p|
|* The beta coefficient of depression includes Beck questionnaire scoring with a value ≥ 12 and is adjusted for each variable (age, gender, body mass index, time of evolution, HbA1c, HDL-C, LDL-C and triglycerides)
Exp-beta = beta exponent; CI = confidence interval; BMI = body mass index;
HDL-C = high-density cholesterol; LDL-C = low-density cholesterol; HbA1c = A1c glycosylated hemoglobin
Overt hearing loss was observed in subjects with depression (Beck questionnaire ≥ 12 points), with a predominance in the left ear in subjects with diabetes, compared with those without diabetes (Annex 1).
Annex 1 Comparison of average audiometry values by ear by subgroup
Our results demonstrate the independent association between depression and hearing loss in people over 40 years with type 2 diabetes. Both comorbidities are common and may go unnoticed if they are not looked for intentionally. Depression and hearing loss diminish the quality of life and are barriers to adherence in the management of chronic diseases. The strength of the association justifies the search for both conditions when either is detected. In addition, they should be considered in the evaluation of therapeutic barriers that can be resolved in the management of diabetes.
Nearly half of patients with diabetes had hearing loss and three times the risk of depression. In subjects with diabetes and depression, the presence of hearing loss had greater predominance in the left ear than those without diabetes or without depression.
Several studies have correlated diabetes with hearing loss, but there is still controversy about this, and its association with depression has only been studied separately. 18-21 The present study evaluated the association between depression and hearing loss in patients with type 2 diabetes.
Diabetes affects a growing number of Mexicans over 40 years. 22 Its presence is associated with loss of years of quality life and premature disability. The effectiveness of the available therapy in the health sector is lower than that reported in other countries. Identifying the conditions that limit the quality of life and efficacy of treatment should be part of the systematic evaluation of people with diabetes. These include depression and hearing loss, so their deliberate search should be part of the initial evaluation of patients with diabetes.
In our study, the group of patients with diabetes showed hearing loss in almost all hearing frequencies, compared to those without diabetes. The prevalence of hearing loss was 45.3%; the value is unexpected since patients were asymptomatic. The severity of hearing loss was moderate in 6.6%, enough to have a negative impact on quality of life. These results are consistent with the studies of Hernandez et al.5 and Bainbridge et al. 9
The frequency of hearing loss in women was found in three-quarters of people with diabetes, close to that reported by Imarai et al., who found 64% hearing loss in females.23
There is three times the risk of depression in subjects with hearing loss and diabetes. The American Diabetes Association categorizes depression and hearing loss as common comorbid conditions that may complicate the management of diabetes, and therefore recommends that they be part of the initial evaluation of patients with this disease.24 A study of 93 diabetic patients in Spain found a higher prevalence than ours (42.8% mild unilateral hearing loss and 57.1% bilateral moderate, predominantly in the right ear in 59.1% and 40.9% in the left). Unlike that reported by Sambola, 25 our patients with depression had overt bilateral hearing loss, predominantly observed in the left ear.
Our data confirm the high prevalence of depression in people with diabetes (21.3%). Multiple reports show that depression is one of the most common barriers to adherence in people with diabetes. 14-17 Our data confirm that cases with depression have higher levels of HbA1c and glucose levels than the rest of the participants in the study who had diabetes.
Our data also show that there is an independent association between depression and hearing loss in people with diabetes. Of the confounding variables analyzed, only age had a borderline role, probably because the patients are younger and have less time of evolution of the disease. This is opposite to what was reported by Michikawa et al., who support a causal role between hyperglycemia and age-related hearing loss, which turned out to be associated with the degree of control of HbA1c.26
However, the risk of depression for those who have diabetes and low socioeconomic status is increased almost six times. Furthermore, the severity of hearing loss was significantly higher when it coexisted with depression.
We recognize the limitations of this study, which has to do with the small size of our sample, plus the fact that causality cannot be established because the design is transversal. It is necessary to make an intentional search for patients with depression and hearing loss to begin management early and thereby slow related complications.
Nearly half of patients with diabetes had hearing loss and more than three times the risk of depression. Subjects with depression and diabetes showed greater hearing loss (> 20 dB) than those without diabetes or without depression. Hearing loss and depression are conditions that must be looked for intentionally in people with diabetes.
To Master of Science Rafael Mondragón González for his support in the laboratory, to Ricardo Saldaña Espinoza for his support with data management, and to Drs Aline Berenice Herrera Rangel and Leticia del Carmen Cruz Hernandez, for their role in the inclusion of patients in the protocol.
Conflict of interest statement: The authors have completed and submitted the form translated into Spanish for the declaration of potential conflicts of interest of the International Committee of Medical Journal Editors, and none were reported in relation to this article.