Nicotine & Tobacco Research, 2018, 1–8

doi:10.1093/ntr/nty026
Original investigation
Received July 12, 2017; Editorial Decision January 15, 2018; Accepted February 14, 2018
Advance Access publication March 14, 2018

Original investigation

Smoking, Smoking Cessation, and the Risk of

Hearing Loss: Japan Epidemiology Collaboration
on Occupational Health Study
Huanhuan Hu PhD,1 Naoko Sasaki MD,2 Takayuki Ogasawara MD,2
Satsue Nagahama MD,3 Shamima Akter PhD,1 Keisuke Kuwahara PhD,1,4
Takeshi Kochi MD,5 Masafumi Eguchi MD,5 Ikuko Kashino PhD,1
Taizo Murakami MD,6 Makiko Shimizu MD,6 Akihiko Uehara MD,7
Makoto Yamamoto MD,8 Tohru Nakagawa MD,9 Toru Honda MD,9
Shuichiro Yamamoto MD,9 Ai Hori PhD,10 Chihiro Nishiura MD,11
Hiroko Okazaki MD,12 Teppei Imai MD,13 Akiko Nishihara MD,13
Toshiaki Miyamoto MD,14 Kentaro Tomita MD,15 Isamu Kabe MD,5
Tetsuya Mizoue PhD,1 Naoki Kunugita MD,16 Seitaro Dohi MD12
for the Japan Epidemiology Collaboration on Occupational Health Study
Group
1
Department of Epidemiology and Prevention, National Center for Global Health and Medicine, Tokyo, Japan;
2
Mitsubishi Fuso Truck and Bus Corporation, Kanagawa, Japan; 3All Japan Labour Welfare Foundation, Tokyo,
Japan; 4Teikyo University Graduate School of Public Health, Tokyo, Japan; 5Furukawa Electric Co, Ltd, Tokyo, Japan;
6
Mizue Medical Clinic, Keihin Occupational Health Center, Kanagawa, Japan; 7Seijinkai Shizunai Hospital, Hokkaidō,
Japan; 8Yamaha Corporation, Shizuoka, Japan; 9Hitachi Ltd, Ibaraki, Japan; 10Department of Global Public Health,
University of Tsukuba, Ibaraki, Japan; 11Tokyo Gas Co, Ltd, Tokyo, Japan; 12Mitsui Chemicals, Inc, Tokyo, Japan; 13Azbil
Corporation, Tokyo, Japan; 14 Nippon Steel & Sumitomo Metal Corporation Kimitsu Works, Chiba, Japan; 15 Mitsubishi
Plastics, Inc, Tokyo, Japan; 16Department of Environmental Health, National Institute of Public Health, Saitama, Japan

Corresponding Author: Huanhuan Hu, Department of Epidemiology and Prevention, National Center for Global Health and
Medicine, Toyama 1-21-1, Shinjuku-ku, Tokyo 162–8655, Japan; E-mail: hu.huanhuan@yahoo.com

Abstract
Introduction: We aimed to determine the prospective association of smoking status, smoking
intensity, and smoking cessation with the risk of hearing loss in a large Japanese cohort.
Methods: The cohort study included 50 195 employees, who were aged 20–64 years and free of hear-
ing loss at baseline. Participants were followed up for a maximum of 8 years. Pure-tone audiometric
testing was performed annually to identify hearing loss at 1 and 4 kHz. Cox proportional hazards
regression models were used to investigate the association between smoking and hearing loss.
Results: During follow-up, 3532 individuals developed high-frequency hearing loss, and 1575
developed low-frequency hearing loss. The hazard ratio (HR) associated with current smokers was
1.6 (95% confidence interval [CI] = 1.5 to 1.7) and 1.2 (95% CI = 1.1 to 1.4) for high- and low-frequency
hearing loss, respectively, as compared with never smokers. The risk of high- and low-frequency
hearing loss increased with the number of cigarettes smoked per day (both p for trend <.001).
The HR associated with former smokers was 1.2 (95% CI = 1.1 to 1.3) and 0.9 (95% CI = 0.8 to 1.1)

© The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. 1
For permissions, please e-mail: journals.permissions@oup.com.

Downloaded from https://academic.oup.com/ntr/advance-article-abstract/doi/10.1093/ntr/nty026/4925604

by University of California, Santa Barbara user
on 14 March 2018
 2 Nicotine & Tobacco Research, 2018, Vol. 00, No. 00

for high- and low-frequency hearing loss, respectively. The analysis by quitting years showed a
decline in risk of hearing loss after quitting smoking, even among those who quitted less than
5 years before baseline.
Conclusions: Smoking is associated with increased risk of hearing loss, especially at the high
frequency, in a dose–response manner. The excess risk of hearing loss associated with smoking
disappears in a relatively short period after quitting.
Implications: The prospective association between smoking and hearing loss has not been well
studied. To the best of our knowledge, our study is the largest to date investigating the association
between smoking and incident hearing loss. Our results indicate that smoking is associated with
increased risk of hearing loss in a dose–response manner. Quitting smoking virtually eliminates
the excess risk of hearing loss, even among quitters with short duration of cessation. These results
suggest that smoking may be a causal factor for hearing loss, although further research would be
required to confirm this. If so, this would emphasize the need for tobacco control to prevent or
delay the development of hearing loss.

Introduction
Hearing loss is an important public health and societal concern. It
adversely affects interpersonal communication, social interaction,
psychosocial well-being, and quality of life1 and is ranked as the
fourth leading cause of years lived with disability globally.2 Many
people with hearing loss receive inadequate treatment or rehabilita-
tion, even in high-income countries,3,4 and the current production
of hearing aids meets less than 10% of the global need.5 The World
Health Organization estimated that more than 5% of the world’s
population has disabling hearing loss.6 In Japan, one study showed
that about 10% of workers aged 40–79 years had hearing loss at
1 or 4 kHz.7 Thus, identifying modifiable risk factors related to
hearing loss is important for reducing the burden of hearing loss.
Experimental studies suggested that smoking may affect the audi-
tory system by direct ototoxicity of nicotine,8,9 cochlear ischemia due
to increased levels of carboxyhemoglobin,10 and smoking-mediated
increased blood viscosity.11
A meta-analysis of five cross-sectional studies showed that cur-
rent smokers were more likely to have hearing loss.12 However,
evidence from prospective studies conflicts. Of eight cohort studies
on the associations between smoking and hearing loss,13–20 only
three reported a significant positive association.18–20 Of five cohort
studies that assessed the dose–response association between
smoking intensity and hearing loss,14,17–20 only two found a dose–
response relationship.19,20 Cohort studies based on self-reported
hearing loss found that past smoking was also associated with
hearing loss,17,20 whereas cohort studies with audiometric assess-
ments did not detect risk elevation among past smokers.18,19 All
these cohort studies had some limitations, including relatively
small sample sizes (500–2000 participants),13–16,18,19 short follow-
up period (≤5 years),15,16,19,20 or self-reported hearing loss.17,20 To
understand better the association between smoking and hearing
loss, large-scale studies using audiometry over long periods of
follow-up are needed.
In Japan, employees are required by law to receive an annual
health examination, including a hearing test. This provides a valu-
able opportunity to identify potentially modifiable risk factors asso-
ciated with hearing loss. The purpose of the present study was to
determine the prospective association of smoking status, smoking
intensity, and smoking cessation with the risk of hearing loss in a
large working population-based cohort study with up to 8 years of
follow-up.
Methods
Study Design
The Japan Epidemiology Collaboration on Occupational Health
(J-ECOH) Study is an ongoing, multicenter epidemiologic study
among workers from 12 companies covering various industries (elec-
tric machinery and apparatus manufacturing; steel, chemical, gas,
and nonferrous metal manufacturing; automobile and instrument
manufacturing; plastic product manufacturing; and health care). In
Japan, workers are obliged to undergo a health checkup at least once
a year under the Industrial Safety and Health Act; nearly all workers
attend their health checkup each year. Participants in the J-ECOH
Study received anthropometric measurements, physical examination
including hearing test, and laboratory examination (blood glucose,
blood lipids, etc) during the health checkup. Checkup participants
also completed a questionnaire about medical history, health-related
lifestyle, and work environment. The details of the J-ECOH Study
have been described previously.21,22 As of May 2016, 11 participating
companies (12 worksites) provided health checkup data obtained
between January 2008 and December 2015 or between April 2008
and March 2016.
Prior to the collection of data, the conduct of the J-ECOH Study
was announced in each company by using posters that explained
the purpose and procedure of the study. Participants did not provide
their verbal or written informed consent to join the study but were
allowed to refuse their participation. This procedure conforms to the
Japanese Ethical Guidelines for Epidemiological Research, where the
procedure of obtaining consent may be simplified for observational
studies using existing data. The study protocol including consent
procedure was approved by the Ethics Committee of the National
Center for Global Health and Medicine, Japan.
In this study, the data from the earliest health checkup (mostly
carried out in 2008) were regarded as the baseline data; however, if
the 2008 dataset contained a large amount of missing data, the data
from the 2009 or 2010 (one company each) health checkup were
used as the baseline. The outcome of the present prospective analysis
was ascertained using data from the health checkups after the base-
line and through March 2016.
Study Participants
As shown in Supplementary Figure S1, of 92 823 participants aged
20–64 years, we excluded participants from one company for
which data on hearing tests were not available, and participants

Downloaded from https://academic.oup.com/ntr/advance-article-abstract/doi/10.1093/ntr/nty026/4925604

by University of California, Santa Barbara user
on 14 March 2018
 Nicotine & Tobacco Research, 2018, Vol. 00, No. 00
from another two companies for which there was no information
on past smoking. Of participants in the remaining eight companies
(n = 80 323), we excluded those who had hearing loss (defined as
hearing thresholds at 1 kHz >30 dB or at 4 kHz >40 dB in at least
one ear) or lacked data on the hearing test at baseline. Of the remain-
ing 59 333 participants, we excluded participants with missing data
on smoking, with self-reported cancer, and with missing data nec-
essary for the diagnosis of hypertension, diabetes, or dyslipidemia
that were used in the main analysis. We further excluded partici-
pants who did not attend any subsequent health checkups or who
attended but did not receive a hearing test. Finally, 50 195 partici-
pants remained for our first set of analyses to assess the association
between the risk of hearing loss among former and current smokers
and that for never smokers. In the second set of analyses, to assess
the association between smoking intensity among current smokers
(number of cigarettes smoked per day) and the risk of hearing loss,
data were available for seven companies (n = 39 112). In the third
set of analyses, which focused on duration of smoking cessation and
the risk of hearing loss, data were available for two major companies
(n = 34 912).
Assessment of Smoking Status
Smoking status (never, former, or current) was identified by a self-
administered questionnaire at baseline. Of seven companies where
the number of cigarettes smoked per day was further asked, four
companies collected data on daily cigarette consumption in a con-
tinuous scale, and the other three companies used a categorical
question (≤10, 11–20, or >20 cigarettes/day). Of two companies
where duration of smoking cessation was available, participants in
one company self-reported years since quitting; in the other com-
pany, participants self-reported age at quitting. Years since quitting
were thus calculated as age at baseline visit minus the recalled age
at quitting.
Assessment and Definition of Hearing Loss
As part of an annual health checkup, pure-tone audiometric test-
ing was performed by trained audiometric technicians using a pure
tone audiometer. The types of audiometers used in each worksite are
provided in Supplementary Table S1. Participants received a hearing
test for each ear at frequencies of 1 and 4 kHz to check whether
their hearing thresholds exceed the thresholds recommended by the
Japanese government guideline (>30 dB at 1 kHz and >40 dB at
4 kHz). http://labor.tank.jp/anei/sisinetc/souon-guideline.html (in
Japanese, accessed May 2017).
Given that smoking appears to be more strongly related vato
high-frequency hearing loss than low-frequency hearing loss,19,23–25
we treated high- and low-frequency hearing loss separately. We
defined low-frequency hearing loss as a hearing threshold of more
than 30 dB at 1 kHz and high-frequency hearing loss as a hearing
threshold of more than 40 dB at 4 kHz in at least one ear (unilateral
hearing loss). In addition, we alternatively defined hearing loss in
both ears (bilateral hearing loss).
Assessment of Covariates
Covariates included age, sex, worksite, body mass index (BMI), his-
tory of cardiovascular disease (CVD), hypertension, diabetes, and
dyslipidemia. The body height and weight were measured using a
scale while the participant wore light clothes and no shoes. BMI was
calculated as the weight in kilograms divided by the squared height
Downloaded from https://academic.oup.com/ntr/advance-article-abstract/doi/10.1093/ntr/nty026/4925604
by University of California, Santa Barbara user
on 14 March 2018
3
in meters. History of CVD was identified on the basis of self-report
of stroke and ischemic heart disease (angina and myocardial infarc-
tion). Blood pressure (BP) was measured with the participants in a
sitting position using an automatic BP monitor. Plasma glucose was
measured with either the enzymatic or glucose oxidase peroxida-
tive electrode method. Glycated hemoglobin (HbA1c) was measured
using a latex agglutination immunoassay, high-performance liquid
chromatography, or the enzymatic method. Triglyceride, low-density
lipoprotein cholesterol, and high-density lipoprotein cholesterol
levels were measured with the enzymatic method. All the labora-
tories involved in the health checkups of the participating compa-
nies had received satisfactory scores (rank A or a score of >95 out
of 100) from external quality-control agencies. Hypertension was
defined as systolic BP ≥140 mmHg, diastolic BP ≥90 mmHg, or
receiving medical treatment for hypertension.26 Diabetes was defined
as HbA1c ≥6.5%, fasting plasma glucose ≥126 mg/dl, random
plasma glucose ≥200 mg/dl, or currently under medical treatment for
diabetes, according to the American Diabetes Association criteria for
the diagnosis of diabetes.27 Dyslipidemia was defined as triglyceride
level ≥150 mg/dl, low-density lipoprotein cholesterol level ≥140 mg/
dl, high-density lipoprotein cholesterol level <40 mg/dl, or receiv-
ing medical treatment for dyslipidemia, based on the criteria for the
Japan Atherosclerosis Society.28
Because questions on lifestyle (except for smoking) and work
environment were markedly different across the participating com-
panies, these factors were adjusted for a sensitivity analysis using
data in one company (n = 31 440). Averaged daily alcohol consump-
tion was calculated as drinking frequency multiplied by ethanol con-
sumption per drinking day. Weekly minutes of leisure time physical
activity were calculated as frequency of physical activity multiplied
by duration of the activity engaged per day and summed across up
to 20 activities, as described elsewhere.29 Participants were asked to
select up to five characteristics of worksite environments among a
list of 14 items (hot, cold, noisy, etc). Occupational noise exposure
was identified if “noisy” was selected by the participant.
Statistical Analyses
Characteristics of the study participants were described as means
for continuous variables and percentages for categorical variables.
Pearson chi-square test for categorical variables or analysis of vari-
ance for continuous variables was used to examine differences in
baseline characteristics among never smokers, former smokers, and
current smokers.
Person-time was calculated from the date of the baseline exami-
nation to the date of the first diagnosis of hearing loss in a subse-
quent examination or to the date of the last examination, whichever
occurred first. We used the Cox proportional hazards model to esti-
mate hazard ratios (HRs) and 95% confidence intervals (CIs) for inci-
dent hearing loss by categories of smoking status, considering never
smoker as the reference category. The first model was adjusted for
baseline age (years, continuous), sex, and worksite. The second model
(main analysis) was additionally adjusted for BMI (kg/m2, continu-
ous), CVD (yes or no), hypertension (yes or no), diabetes (yes or no),
and dyslipidemia (yes or no). We further adjusted for occupational
noise exposure (yes or no), alcohol intake (<23 or ≥23 g ethanol/
day), and leisure time physical activity (<150 or ≥150 min/week) in a
sensitivity analysis using data from one major company. Interaction
of smoking and occupational noise exposure on hearing loss was
examined in the sensitivity analysis. Population attributable fraction
(PAF) was calculated using the formula: PAF = pd [(HR − 1)/HR],
 4 Nicotine & Tobacco Research, 2018, Vol. 00, No. 00

in which pd is the proportion of cases exposed to the risk factor and
HR is the adjusted HR in the second model.30
Multivariable-adjusted HRs for incident hearing loss by cat-
egories of smoking intensity who smoked 1 to 10, 11 to 20, and
>20 cigarettes/day, compared with persons who never smoked,
were estimated. Tests for trends were performed by treating the cat-
egories of smoking intensity as ordinal numbers and modeling this
variable as a continuous variable. To assess the association between
smoking cessation and the risk of hearing loss, we classified past
smokers into three groups according to years since quitting: <5, 5
to 9, and ≥10 years. We decided not to further divide those who
quitted ≥10 years before baseline into subgroups after consider-
ing the distribution of quitting years (14% with 10–14 years, 8%
with 15–19 years, and 18% with ≥20 years) and confirming similar
associations with hearing loss among these subgroups. All statistical
analyses were performed using SAS version 9.3 (SAS Institute, Cary,
NC). A two-sided p < .05 was considered statistically significant.
Results
Table 1 shows baseline characteristics of the study population by
smoking status. Former smokers were older and had higher BMI,
BP, and blood glucose. Correspondingly, former smokers had higher
prevalence of hypertension, diabetes, and CVD than never smokers
and current smokers. Current smokers were more likely to report
occupational noise than their counterparts.
Smoking Status and Risk of Hearing Loss
Table 2 shows adjusted HRs for incident unilateral hearing loss
according to baseline smoking status. During a maximum of 8 years
of follow-up, 3532 (7.0%) individuals developed unilateral high-
frequency hearing loss, and 1575 (3.1%) developed unilateral low-
frequency hearing loss, corresponding to an incident rate of 12.2
and 5.3 per 1000 person-years, respectively. After adjusting for age,
sex, worksite, BMI, CVD, hypertension, diabetes, and dyslipidemia,
the HRs for high-frequency hearing loss were 1.2 (95% CI = 1.1
to 1.3) in former smokers and 1.6 (95% CI = 1.5 to 1.7) in current
smokers compared with never smokers. For low-frequency hearing
loss, the adjusted HRs were 0.9 (95% CI = 0.8 to 1.1) in former
smokers and 1.2 (95% CI = 1.1 to 1.4) in current smokers. Similar
associations were observed after further adjustment for occupational
noise exposure, alcohol intake, and leisure time physical activity in
the sensitivity analysis. The interaction between smoking and occu-
pational noise exposure was not statistically significant. The PAF of
high- and low-frequency hearing loss onset due to current smoking
were 17.5% and 6.6%, respectively.
Smoking Intensity and Risk of Hearing Loss
Figure 1 shows adjusted HRs for incident unilateral hearing loss
according to baseline smoking intensity. Compared with never
smokers, the adjusted HRs of high-frequency hearing loss among
persons who smoked 1 to 10, 11 to 20, and more than 20 cigarettes/
day were 1.4 (95% CI = 1.2 to 1.6), 1.6 (95% CI = 1.5 to 1.8), and
1.7 (95% CI = 1.5 to 2.0), respectively (Supplementary Table S2).
Regarding low-frequency hearing loss, the adjusted HRs were 1.1
(95% CI = 0.9 to 1.4), 1.2 (95% CI = 1.0 to 1.4), and 1.4 (95%
CI = 1.1 to 1.8) for persons who smoked 1 to 10, 11 to 20, and
more than 20 cigarettes/day, respectively. The risk of both high- and
low-frequency hearing loss increased significantly as the number of
cigarettes smoked per day increased (both p for trend <.001).
Smoking Cessation and Risk of Hearing Loss
Figure 2 shows a rapid decline in risk of hearing loss among people
who quit <5 years before baseline compared with current smokers.
For high-frequency hearing loss, the HRs of hearing loss decreased
from 1.6 (95% CI = 1.4 to 1.7) for current smokers to 1.2 (95%

Table 1. Characteristics of Study Population According to Smoking Status at Baseline, the Japan Epidemiology Collaboration on
Occupational Health Study, 2008–2015

Never smoker (n = 21 374) Past smoker (n = 9828) Current smoker (n = 18 993)

% Mean (SD) % Mean (SD) % Mean (SD) p

Men 70.1 95.8 95.1 <.001

Age (years) 43.3 (9.4) 46.3 (8.7) 42.8 (9.4) <.001
BMI (kg/m2) 23.0 (3.5) 23.8 (3.1) 23.6 (3.5) <.001
SBP (mmHg) 120.4 (15.2) 123.5 (15.1) 120.1 (14.6) <.001
DBP (mmHg) 75.5 (10.9) 78.6 (10.5) 75.2 (10.3) <.001
Blood glucose (mg/dl) 97.2 (17.1) 102.0 (19.9) 99.5 (20.3) <.001
HbA1c (%) 5.5 (0.6) 5.6 (0.7) 5.6 (0.7) <.001
TG 105.3 (73.5) 127.9 (89.1) 138.9 (110.6) <.001
HDL 60.8 (15.3) 58.5 (14.9) 54.2 (13.8) <.001
LDL 118.5 (29.7) 121.0 (29.2) 118.8 (31.1) <.001
Hypertension 17.1 25.3 16.1 <.001
Diabetes 4.8 8.0 7.4 <.001
Dyslipidemia 37.5 47.2 49.0 <.001
CVD 0.9 2.6 0.7 <.001
Leisure time physical activity 12.8 16.8 13.0 <.001
(≥150 min/week)a
Alcohol drinker (>23 g ethanol/day,)a 13.8 34.0 31.7 <.001
Occupational noise exposurea 18.7 21.2 24.8 <.001

BMI = body mass index; CVD = cardiovascular disease; DBP = diastolic blood pressure; HbA1c = glycated hemoglobin; HDL = high-density lipoprotein-
cholesterol; LDL = low-density lipoprotein-cholesterol; SBP = systolic blood pressure; TG = triglyceride.
a
Data were available for 31 440 people in one company.

Downloaded from https://academic.oup.com/ntr/advance-article-abstract/doi/10.1093/ntr/nty026/4925604

by University of California, Santa Barbara user
on 14 March 2018
 Nicotine & Tobacco Research, 2018, Vol. 00, No. 00 5

Table 2. Adjusted Hazard Ratios (HRs) for Incident Hearing Loss (Unilateral) According to Baseline Smoking Status at Baseline, the Japan
Epidemiology Collaboration on Occupational Health Study, 2008–2015

Model 1 Model 2 Model 3a

Incidence (/1000
n Case person years) HR 95% CI HR 95% CI HR 95% CI PAF (%)b

High-frequency hearing loss

Never smoker 21 374 1015 8.2 1.0 Referent 1.0 Referent 1.0 Referent
Former smoker 9828 866 15.7 1.2 1.1, 1.3 1.2 1.1, 1.3 1.2 1.0, 1.3 4.1
Current smoker 18 993 1651 15.0 1.6 1.5, 1.7 1.6 1.5, 1.7 1.6 1.4, 1.7 17.5
Low-frequency hearing loss
Never smoker 21 374 642 5.2 1.0 Referent 1.0 Referent 1.0 Referent
Former smoker 9828 308 5.4 0.9 0.8, 1.1 0.9 0.8, 1.1 0.9 0.7, 1.0
Current smoker 18 993 625 5.5 1.2 1.1, 1.4 1.2 1.1, 1.4 1.1 1.0, 1.3 6.6

CI = confidence interval; PAF = population-attributable fraction.

Model 1 adjusted for age (years), sex, and worksite.
Model 2 (main analysis) adjusted for all factors in model 1 plus BMI (kg/m2), CVD, hypertension, diabetes, and dyslipidemia.
Model 3 (sensitivity analysis) adjusted for all factors in model 2 plus alcohol intake (<23 or ≥23 g ethanol/day), leisure time physical activity (<150 or ≥150 min/
week), and occupational noise exposure (yes or no).
a
Data available for 31 440 subjects in one company.
b
We calculated PAF (%) using multivariable-adjusted hazard ratio in model 2 (main analysis).

Figure 1. Adjusted hazard ratios for incident hearing loss (unilateral) according to baseline smoking intensity, the Japan Epidemiology Collaboration on
Occupational Health Study, 2008–2015. Estimates were adjusted for age (years), sex, worksite, body mass index (kg/m2), cardiovascular disease, hypertension,
diabetes, and dyslipidemia.

Downloaded from https://academic.oup.com/ntr/advance-article-abstract/doi/10.1093/ntr/nty026/4925604

by University of California, Santa Barbara user
on 14 March 2018
 6 Nicotine & Tobacco Research, 2018, Vol. 00, No. 00

Figure 2. Adjusted hazard ratios for incident hearing loss (unilateral) by years since quitting smoking before baseline, the Japan Epidemiology Collaboration on
Occupational Health Study, 2008–2015. Estimates were adjusted for age (years), sex, worksite, body mass index (kg/m2), cardiovascular disease, hypertension,
diabetes, and dyslipidemia.

CI = 1.0 to 1.4), 1.1 (95% CI = 0.9 to 1.3), and 1.1 (95% CI = 1.0
to 1.3) for people who quit <5, 5 to 9, and ≥10 years before, respec-
tively, compared with never smokers (see Supplementary Table S3).
For low frequency hearing loss, the HRs of hearing loss decreased
from 1.2 (95% CI = 1.0 to 1.3) for current smokers to 0.9 (95%
CI = 0.7 to 1.3), 0.8 (95% CI = 0.5 to 1.1), and 0.9 (95% CI = 0.7
to 1.1) for people who quit <5, 5 to 9, and ≥10 years before, respec-
tively, compared with never smokers.
Bilateral Hearing Loss
Similar association patterns were observed when we alternatively
defined hearing loss in both ears, although the associations between
smoking and bilateral low-frequency hearing loss were not statisti-
cally significant (Supplementary Table S4, and Figures S2 and S3).
Discussion
Based on a large-scale cohort study among a Japanese working
population, we demonstrated that smoking is associated with an
increased risk of hearing loss in a dose–response manner. The magni-
tude of the association between smoking and high-frequency hearing
loss was stronger than that of low-frequency hearing loss. The risk
of hearing loss associated with smoking appears to decrease within
5 years after quitting smoking. To our knowledge, the present study
is the largest cohort study examining the association between smok-
ing and hearing loss.
In this study, current smokers had a significantly increased risk of
hearing loss, even after adjustment for alcohol consumption, leisure
time physical activity, and occupational noise exposure in sensitiv-
ity analysis, suggesting an independent association of smoking with
hearing loss. Our findings are in line with the finding of a meta-
analysis based on five cross-sectional studies12 and those from recent
cross-sectional studies23,24,31,32 published after that meta-analysis.
Our results are also consistent with some18–20 but not all13–17 cohort
studies. The heterogeneity of results between cohort studies might
be because of the differences in the procedure of hearing assessment,
follow-up period, covariate adjustment, study population, and sam-
ple size. With a large sample size, long follow-up period, and objec-
tive assessment of hearing loss, our study provides strong evidence
that smoking is an independent risk factor for hearing loss.
We observed that the risk of hearing loss increased with cigarette
consumption. In agreement with our findings, cross-sectional studies
have reported that the odds ratios for hearing loss increased with the
number of cigarettes smoked per day25 and pack-years.31,33 Of three
cohort studies reporting a significant association between smoking and
hearing loss, one did not find a dose–response relationship between

Downloaded from https://academic.oup.com/ntr/advance-article-abstract/doi/10.1093/ntr/nty026/4925604

by University of California, Santa Barbara user
on 14 March 2018
 Nicotine & Tobacco Research, 2018, Vol. 00, No. 00
smoking and hearing loss, which is probably partly because of the
small number of current smokers (n = 238),18 whereas the other two
studies showed that HRs for hearing loss increased with the number
of cigarettes smoked per day.19,20 Specifically, the cohort study among
1554 male office workers in Japan showed that the HRs for high-
frequency hearing loss compared with never smokers were 1.8 (95%
CI = 0.9 to 3.6), 2.0 (95% CI = 1.0 to 4.1), and 2.2 (95% CI = 1.1 to
4.4) for people who smoked 1 to 20, 21 to 30, and ≥31 cigarettes/day,
respectively, and the corresponding HRs for low-frequency hearing
loss were 1.2 (95% CI = 0.7 to 2.3), 1.4 (95% CI = 0.7 to 2.6), and
1.8 (95% CI = 1.0 to 3.4), respectively.19 A cohort study in Denmark,
based on self-reported hearing loss among 4610 employees showed
that the HRs for hearing loss were 1.6 (95% CI = 1.1 to 2.3) for peo-
ple who smoked <15 g/day, and 1.8 (95% CI = 1.3 to 2.5) for those
who smoked ≥15 g/day, compared with never smokers.20 The dose–
response relationship observed in our study as well as in previous
cohort studies19,20 adds evidence to support the finding that smoking
is causally associated with hearing loss.
Consistent with previous cross-sectional23–25 and cohort19 studies,
we observed that the magnitude of the association between smok-
ing and high-frequency hearing loss was stronger than that of the
association between smoking and low-frequency hearing loss, sug-
gesting that smoking may have differential impact on high- and low-
frequency hearing loss. Some explanations for this finding are pos-
sible. First, the nicotinic acetylcholine receptor alpha9 and alpha10
subunits are expressed primarily on outer (high frequency) hair
cells, suggesting that tobacco may have a direct ototoxic effect on
outer hair cell function.34 If this is the case, a similar adverse effect
of electronic cigarettes on hearing as for tobacco smoking would be
expected. Future research should explore the association between
electronic cigarette use and hearing loss. Second, cigarette smok-
ing may damage the auditory system by cochlear ischemia due to
increased levels of carboxyhemoglobin.10 Some studies suggested that
the cochlea is more susceptible to ischemic change at its basal (high
frequency) end.35,36 Third, high-frequency hearing loss was associated
with high shear blood viscosity,37 which may be caused by smoking.
We observed that former smokers had a slightly increased risk of
high-frequency hearing loss (HR = 1.2, 95% CI = 1.1 to 1.3), but a
similar risk of low-frequency hearing loss (HR = 0.9, 95% CI = 0.8
to 1.1), than never smokers, suggesting a modest association between
past smoking and hearing loss. This finding is relatively consistent
with previous cohort studies.17–20 In the further analysis by quitting
years, we observed a rapid decline in the risk of hearing loss among
people who quit <5 years before baseline compared with current
smokers: HRs for high-frequency hearing loss decreased from 1.6
(95% CI = 1.4 to 1.7) for current smokers to 1.2 (95% CI = 1.0 to
1.4) for new quitters (<5 years) versus never smokers; HRs for low-
frequency hearing loss decreased from 1.2 (95% CI = 1.0 to 1.3) for
current smokers to 0.9 (95% CI = 0.7 to 1.3) for new quitters ver-
sus never smokers. A previous cohort study showed that compared
with never smokers, the HRs for hearing loss at 0.5–4 kHz were 1.4
(95% CI = 1.1 to 1.8) for current smokers, 1.2 (95% CI = 0.9 to
1.8) for people who stopped smoking <5 years before, and 1.1 (95%
CI = 0.9 to 1.4) for those who stopped ≥5 years before.18 Taken
together, quitting virtually eliminates the excess risk of hearing loss,
even among quitters with short duration of cessation.
Strengths and Limitations of this Study
This study has several strengths including the large population-based
cohort, long-term follow-up, use of audiometry to measure hear-
ing thresholds over time, and comprehensive covariate adjustment.
Downloaded from https://academic.oup.com/ntr/advance-article-abstract/doi/10.1093/ntr/nty026/4925604
by University of California, Santa Barbara user
on 14 March 2018
7
However, several limitations warrant mention. First, because of a small
number of participants who developed bilateral low-frequency hearing
loss, the present study lacked power to detect significant associations
between smoking and bilateral low-frequency hearing loss. However,
the association patterns of smoking with unilateral and bilateral hear-
ing loss were similar. Second, smoking habits were ascertained via a
questionnaire but not biologically verified (eg, exhaled carbon monox-
ide or salivary cotinine). Third, the existing yet limited studies showed
that other lifestyle factors (eg, physical activity, alcohol drinking, expo-
sure to loud music) were associated with hearing loss.31,38 The lack or
variation in questions on these data across participating companies
did not allow us to control these factors in overall study subjects. In
a sensitivity analysis using data of one major company, however, we
confirmed that the association between smoking and hearing loss did
not materially change after adjusting for physical activity and alcohol
drinking. Fourth, data on occupational noise exposure was available
only in one of the participating companies. Although an analysis with
adjustment for noise exposure in that company showed little difference
in the association between smoking and hearing loss, the subjective
assessment of noise exposure should be acknowledged as a limitation.
Fifth, because information regarding otologic diseases, head injuries,
or drugs with potential impact on hearing acuity was not collected
in a uniform manner, we were unable to exclude cases of hearing loss
attributable to these above factors. Given that the hearing test was con-
ducted at only 1 and 4 kHz during health checkups, hearing loss at
other frequencies could not be identified. Last, the majority of study
participants were employees of large companies. Caution should thus
be exercised in generalizing the present finding to workers in smaller
companies, the self-employed, or the unemployed.
In conclusion, the present study indicates that smoking is associ-
ated with increased risk of hearing loss in a dose–response manner:
The association is more pronounced for high-frequency hearing loss
than for low-frequency hearing loss, and the excess risk of hearing
loss decreases after quitting. These results, obtained from a large-scale
prospective study, provide strong evidence to support that smoking is
a causal factor for hearing loss and emphasize the need for tobacco
control to prevent or delay the development of hearing loss.
Supplementary Material
Supplementary data can be found online at http://www.ntr.oxford-
journals.org.
Funding
This study was supported by the Industrial Health Foundation, Industrial
Disease Clinical Research Grants (140202-01), JSPS KAKENHI Grant
(16H05251), Grant of National Center for Global Health and Medicine (28-
Shi-1206). The funders had no role in study design; in the collection, analysis,
and interpretation of data; in the writing of the report; or in the decision to
submit the article for publication.
Declaration of Interests
None declared.
Acknowledgments
We thank Dr. Toshiteru Okubo (Chairperson of Industrial Health Foundation,
Director of Radiation Effects Research Foundation) for scientific advice on the
conduct of J-ECOH Study; Maki Konishi (National Center for Global Health
 8 Nicotine & Tobacco Research, 2018, Vol. 00, No. 00
and Medicine) for data management; and Rika Osawa (National Center for
Global Health and Medicine) for administrative support.
References
1. Ciorba A, Bianchini C, Pelucchi S, Pastore A. The impact of hearing loss on
the quality of life of elderly adults. Clin Interv Aging. 2012;7:159–163.
2. GBD 2015 Disease and Injury Incidence and Prevalence Collaborators.
Global, regional, and national incidence, prevalence, and years lived
with disability for 310 diseases and injuries, 1990–2015: a system-
atic analysis for the Global Burden of Disease Study 2015. Lancet.
2016;388(10053):1545–1602.
3. Nash SD, Cruickshanks KJ, Huang GH, et al. Unmet hearing health
care needs: the Beaver Dam offspring study. Am J Public Health.
2013;103(6):1134–1139.
4. Moon IJ, Baek SY, Cho YS. Hearing aid use and associated factors in South
Korea. Medicine (Baltimore). 2015;94(42):e1580.
5. World Health Organization. Millions of people in the world have hearing
loss that can be treated or prevented. http://www.who.int/pbd/deafness/
news/Millionslivewithhearingloss.pdf (accessed December 2017).
6. World Health Organization. Deafness and hearing loss. http://www.who.
int/mediacentre/factsheets/fs300/en/ (accessed May 2017).
7. Saito Y, Ikarashi1 Y, Miyamae A, et al. Hearing loss influenced by early
stage of hypertension, glucose intolerance and hyperlipidemia. Health
Eval Promot. 2014;41(2):268–273. (in Japanese).
8. Maffei G, Miani P. Experimental tobacco poisoning: resultant struc-
tural modifications of the cochlea and tuba acustica. Arch Otolaryngol.
1962;75(5):386–396.
9. Blanchet C, Eróstegui C, Sugasawa M, Dulon D. Acetylcholine-
induced potassium current of guinea pig outer hair cells: its depend-
ence on a calcium influx through nicotinic-like receptors. J Neurosci.
1996;16(8):2574–2584.
10. Fechter LD, Thorne PR, Nuttall AL. Effects of carbon monoxide on coch-
lear electrophysiology and blood flow. Hear Res. 1987;27(1):37–45.
11. Lowe GD, Drummond MM, Forbes CD, Barbenel JC. The effects of age
and cigarette-smoking on blood and plasma viscosity in men. Scott Med J.
1980;25(1):13–17.
12. Nomura K, Nakao M, Morimoto T. Effect of smoking on hearing loss:
quality assessment and meta-analysis. Prev Med. 2005;40(2):138–144.
13. Gates GA, Cobb JL, D’Agostino RB, Wolf PA. The relation of hearing in
the elderly to the presence of cardiovascular disease and cardiovascular
risk factors. Arch Otolaryngol Head Neck Surg. 1993;119(2):156–161.
14. Brant LJ, Gordon-Salant S, Pearson JD, et al. Risk factors related to age-
associated hearing loss in the speech frequencies. J Am Acad Audiol.
1996;7(3):152–160.
15. Karlsmose B, Lauritzen T, Engberg M, Parving A. A five-year longitudi-
nal study of hearing in a Danish rural population aged 31–50 years. Br J
Audiol. 2000;34(1):47–55.
16. Gopinath B, Flood VM, McMahon CM, Burlutsky G, Smith W, Mitchell
P. The effects of smoking and alcohol consumption on age-related hearing
loss: the Blue Mountains Hearing Study. Ear Hear. 2010;31(2):277–282.
17. Shargorodsky J, Curhan SG, Eavey R, Curhan GC. A prospective study of
cardiovascular risk factors and incident hearing loss in men. Laryngoscope.
2010;120(9):1887–1891.
18. Cruickshanks KJ, Nondahl DM, Dalton DS, et al. Smoking, central adi-
posity, and poor glycemic control increase risk of hearing impairment.
J Am Geriatr Soc. 2015;63(5):918–924.
19. Nakanishi N, Okamoto M, Nakamura K, Suzuki K, Tatara K. Cigarette
smoking and risk for hearing impairment: a longitudinal study in Japanese
male office workers. J Occup Environ Med. 2000;42(11):1045–1049.
Downloaded from https://academic.oup.com/ntr/advance-article-abstract/doi/10.1093/ntr/nty026/4925604
by University of California, Santa Barbara user
on 14 March 2018
20. Burr H, Lund SP, Sperling BB, Kristensen TS, Poulsen OM. Smoking
and height as risk factors for prevalence and 5-year incidence of hear-
ing loss. A questionnaire-based follow-up study of employees in Denmark
aged 18–59 years exposed and unexposed to noise. Int J Audiol.
2005;44(9):531–539.
21. Hu H, Kurotani K, Sasaki N, et al.; Japan Epidemiology Collaboration on
Occupational Health Study Group. Optimal waist circumference cut-off
points and ability of different metabolic syndrome criteria for predicting
diabetes in Japanese men and women: Japan Epidemiology Collaboration
on Occupational Health Study. BMC Public Health. 2016;16:220.
22. Hu H, Nagahama S, Nanri A, et al. Duration and degree of weight change
and risk of incident diabetes: Japan Epidemiology Collaboration on
Occupational Health Study. Prev Med. 2017;96:118–123.
23. Fransen E, Topsakal V, Hendrickx JJ, et al. Occupational noise, smok-
ing, and a high body mass index are risk factors for age-related hear-
ing impairment and moderate alcohol consumption is protective: a
European population-based multicenter study. J Assoc Res Otolaryngol.
2008;9(3):264–276.
24. Sumit AF, Das A, Sharmin Z, et al. Cigarette smoking causes hearing impair-
ment among Bangladeshi population. PLoS One. 2015;10(3):e0118960.
25. Mizoue T, Miyamoto T, Shimizu T. Combined effect of smoking and occu-
pational exposure to noise on hearing loss in steel factory workers. Occup
Environ Med. 2003;60(1):56–59.
26. Shimamoto K, Ando K, Fujita T, et al.; Japanese Society of Hypertension
Committee for Guidelines for the Management of Hypertension. The
Japanese Society of Hypertension guidelines for the management of hyper-
tension (JSH 2014). Hypertens Res. 2014;37(4):253–390.
27. American Diabetes Association Standards of medical care in diabe-
tes—2015. Diabetes Care. 2015;38(suppl 1):S1–S94.
28. Teramoto T, Sasaki J, Ishibashi S, et al.; Japan Atherosclerosis Society.
Executive summary of the Japan Atherosclerosis Society (JAS) guidelines
for the diagnosis and prevention of atherosclerotic cardiovascular diseases
in Japan – 2012 version. J Atheroscler Thromb. 2013;20(6):517–523.
29. Honda T, Kuwahara K, Nakagawa T, Yamamoto S, Hayashi T, Mizoue T.
Leisure-time, occupational, and commuting physical activity and risk of
type 2 diabetes in Japanese workers: a cohort study. BMC Public Health.
2015;15:1004.
30. Rockhill B, Newman B, Weinberg C. Use and misuse of population attrib-
utable fractions. Am J Public Health. 1998;88(1):15–19.
31. Dawes P, Cruickshanks KJ, Moore DR, et al. Cigarette smoking, passive
smoking, alcohol consumption, and hearing loss. J Assoc Res Otolaryngol.
2014;15(4):663–674.
32. Chang J, Ryou N, Jun HJ, Hwang SY, Song JJ, Chae SW. Effect of ciga-
rette smoking and passive smoking on hearing impairment: data from a
population-based study. PLoS One. 2016;11(1):e0146608.
33. Virokannas H, Anttonen H. Dose-response relationship between smok-
ing and impairment of hearing acuity in workers exposed to noise. Scand
Audiol. 1995;24(4):211–216.
34. Lustig LR. Nicotinic acetylcholine receptor structure and function in
the efferent auditory system. Anat Rec A Discov Mol Cell Evol Biol.
2006;288(4):424–434.
35. Lin CD, Wei IH, Tsai MH, et al. Changes in guinea pig cochlea after tran-
sient cochlear ischemia. Neuroreport. 2010;21(15):968–975.
36. Surya Atmaja O, Sudarman K, Surono A. Preeclampsia as a risk factor
for damage of the cochlear outer hair cells function. Adv Otolaryngol
2016;2016:1798382.
37. Browning GG, Gatehouse S, Lowe GD. Blood viscosity as a factor in sen-
sorineural hearing impairment. Lancet. 1986;1(8473):121–123.
38. Curhan SG, Eavey R, Wang M, Stampfer MJ, Curhan GC. Body mass
index, waist circumference, physical activity, and risk of hearing loss in
women. Am J Med. 2013;126(12):1142.e1–1142.e8.