Respiratory Medicine 123 (2017) 87e93

Contents lists available at ScienceDirect

Respiratory Medicine
journal homepage: www.elsevier.com/locate/rmed

Chronic bronchitis in relation to hospitalization and mortality over

three decades
Margit K. Pelkonen a, *, Irma-Leena K. Notkola b, Tiina K. Laatikainen c, Pekka Jousilahti c
a
Department of Pulmonary Diseases, Kuopio University Hospital, Kuopio, Finland
b
University of Eastern Finland, Kuopio, Finland
c
National Institute for Health and Welfare, Helsinki, Finland

a r t i c l e i n f o a b s t r a c t

Article history: Background: The study examines the predictive value of chronic bronchitis for all cause and cause-
Received 22 September 2016 specific hospitalizations and for mortality during the last three decades.
Received in revised form Methods: The study population consists of altogether 47 896 men and women aged 25e74 years who
9 December 2016
participated in the National FINRISK Study between 1982 and 2007. The study protocol included a
Accepted 26 December 2016
Available online 27 December 2016
standardized questionnaire on the symptoms of chronic bronchitis, smoking habits and other risk factors
and clinical measurements at the study site. Data on hospitalizations were obtained from the National
Hospital Discharge Registry, and data on the underlying causes of deaths from the National Causes of
Keywords:
Chronic bronchitis
Death register. The study cohorts were followed up until the end of 2011.
COPD Results: In study subjects with symptoms of chronic bronchitis the mean annual days of hospitalization
Smoking were almost two-fold higher than in study subjects without chronic bronchitis. The increase was seen in
Hospitalization all age -groups and both in 5-year periods for each cohort and during the whole 30-year follow-up. More
Mortality specifically, hospitalizations were increased for respiratory diseases and cancer. Chronic bronchitis
increased hospitalizations more in smokers and ex-smokers than in never smokers. Furthermore, chronic
bronchitis was associated with increased all-cause mortality (hazard ratio (HR) 1.23) and mortality from
respiratory causes, cardiovascular diseases and cancer. Smokers and ex-smokers with chronic bronchitis
had an increased risk to die (HRs 2.89 and 1.69, respectively) compared with never-smokers without
chronic bronchitis.
Conclusion: Symptoms of chronic bronchitis can help to identify individuals who are at risk for increased
hospitalizations and mortality.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction bronchitis and hospitalizations [6,8,11,12]. In these studies the du-

The major risk factor for chronic bronchitis, characterized by
chronic mucus production from airways, is cigarette smoking [1,2].
Exposure to airborne particles may also contribute to the devel-
opment of chronic bronchitis [3,4]. Chronic bronchitis itself has
been associated with an accelerated decline in lung function and
the development of airway obstruction [5,6], an increased risk for
respiratory infection [7], increased medical costs [8] and increased
respiratory, cardiovascular and all-cause mortality [5,9,10]. There
are only a few studies examining the relation between chronic
* Corresponding author. Center for Medicine and Clinical Research, Division of
Respiratory Medicine, Kuopio University Hospital, Box 1777, 70211 Kuopio, Finland.
E-mail address: Margit.Pelkonen@kuh.fi (M.K. Pelkonen).
rations of the follow-ups have varied from three months to one
year, and only the relations between chronic bronchitis and all-
cause hospitalizations or COPD hospitalizations have been stud-
ied [6,8,11,12].
In the present study, we first examined how the symptoms of
chronic bronchitis in 1982 were related to hospitalizations during
30 years of follow-up. Then we studied how the symptoms of
chronic bronchitis were associated with hospitalizations during a
5-year period after each examination. We also investigated if there
were any trends in the association between chronic bronchitis and
hospitalization by the year of examination. The relation between
chronic bronchitis and 30-year mortality was also studied to clarify
the whole process from chronic bronchitis to increased morbidity
and mortality.

http://dx.doi.org/10.1016/j.rmed.2016.12.018
0954-6111/© 2017 Elsevier Ltd. All rights reserved.
88 M.K. Pelkonen et al. / Respiratory Medicine 123 (2017) 87e93

2. Methods
2.1. Subjects
Cross-sectional risk factor surveys have been conducted every
fifth year in Finland by the National Public Health Institute since
1972 [13e15]. In each survey, an independent population sample
was drawn from the population register in five geographical areas:
the provinces of North Karelia and Kuopio in eastern Finland since
1972, the Turku-Loimaa region in south-western Finland since
1982, the capital area including the cities of Helsinki and Vantaa
since 1992 and the Oulu province in north-western Finland since
1997 and Lapland since 2002. In each study year, a random sample
stratified by sex and 10-year age group was taken from each study
area so that at least 250 people (at least 200 in 2007) were included
in each subgroup [13e15]. The sampling and study methods com-
plied with the protocol of the World Health Organisation Multi-
national MONItoring of trends and determinants in CArdiovascular
disease (MONICA) project [16], and since 2002, the later recom-
mendations of the European Health Risk Monitoring project
(EHRM) [17].
The study population consisted of altogether 47896 men and
women who participated in the cross-sectional surveys between
1982 and 2007 (Table 1). Subjects without data on smoking were
excluded. Also those who had quit smoking less than one month
before the examination were excluded (because of their short
abstinence they could not be recorded as ex-smokers). In addition,
those subjects who had no data on the symptoms of chronic
bronchitis or asthma (altogether n ¼ 1540) and subjects reporting
asthma diagnosed by a doctor (altogether n ¼ 1727) were also
excluded [13].
2.2. Measurement of chronic bronchitis, smoking status and other
variables
At each study year, the participants completed a self-
administered questionnaire [13]. Definition of chronic bronchitis
was based on a positive response to the same standard question ‘Do
you bring up phlegm on most days or nights for at least three
months each year’ [5,13,18]. Smoking, marital status, education, a
history of diagnosed hypertension during the preceding 12 months
and a history of diagnosed myocardial infarction were asked with
standardized questions in a self-administered questionnaire. Body
mass index (BMI) (kg/m [2]) was calculated from height and weight
measured at the examination.
Smoking status was classified into the three categories: never-,
ex- and current smokers (Table 1). Smokers had smoked regularly
at least one year (cigarettes, cigars or pipe) and had smoked during
the preceding month. Ex-smokers had stopped smoking at least
one month before the survey. The educational background was
classified into four categories: an elementary school, a vocational
school, an upper secondary school/college and an academic degree.
Marital status was classified into four categories: married/cohabi-
tation without marriage, unmarried, divorced and widowed. The
area of residence was classified into four categories: North Karelia
and Kuopio, south-western Finland (Turku-Loimaa area), Oulu and
Lapland, and the capital area.
2.3. Assessment of hospitalization and mortality
The annual days of hospitalization were calculated by summing
up the days of hospitalization and then dividing the sum by the
time of the follow-up in years (or by the time until death if the
subject died before the end of the follow-up). The hospitalization
data was derived from the National Hospital Discharge Registry,
which has the healthcare records of inpatients provided by all
hospitals and municipal health centers in Finland. The reports
include e.g. admission and discharge dates and discharge diagnoses
coded by the doctor responsible for the discharge. After participa-
tion in the FINRISK Study data on hospitalizations of the study
subjects from that point on was linked to the survey data.
Between 1982 and 2012 three different revisions of the Inter-
national Classification of Diseases (ICD) occurred in Finland (i.e. ICD-
8 until 1986, ICD-9 between 1987 and 1995 and ICD-10 since 1996).
In the present study, the concordance table was used for bridging
the three revisions of ICD [19], and the first-listed (main) discharge
diagnoses were classified into the following four major categories
(Table 2): respiratory causes (the corresponding ICD-8, ICD-9 and
ICD-10 codes are listed in at Table 2), cardiovascular diseases,
cancer and other causes. COPD (¼ chronic obstructive pulmonary
disease), ischemic heart disease and lung cancer were also rean-
alysed separately.
There were altogether 31883 subjects with hospitalizations
(total of 199887 hospitalizations) between 1982 and the end of
2011 (Table 1). The data on the length of the visit was missing in
3238 hospital visits, and the data on the main discharge diagnosis
was missing in 53 hospitalizations.
Data on the underlying causes of death were obtained from the
National Causes of Death register. There were altogether 7257
deaths between 1982 and the end of 2011. The underlying causes of

Table 1
Description of the study population by the year of examination.

n Year of examination

1982 1987 1992 1997 2002 2007 Total

Invited 11,395 7932 7927 11,500 13,498 12,000 64,252

Examined (men/women) 4615/4732 3109/3370 2849/3202 4253/4193 4482/5098 3740/4253 23048/28848
Non-responders (men/women) 1212/836 853/600 1116/760 1747/1307 2267/1651 2260/1747 9455/6901

Subjects with hospitalizations between examination and 2011 8453 5660 4690 5973 4910 2197 31,883
Subjects with hospitalizations during 5 years after examination 3950 2830 2479 3522 3398 2197 18,376
Number of deaths during 1982e2011 3181 1484 817 1161 472 142 7257

Smoking status
never smokers 3962 2804 2596 3752 4294 3634 21,042
ex-smokers 2289 1675 1622 2317 2490 2425 12,818
current smokers 2791 1805 1811 2117 2725 1855 13,104
excludeda 305 195 22 260 71 79 932

Chronic bronchitisb 1284 785 855 1085 1175 879 6063

a
Subjects without data on smoking or when they had last smoked as well as ex-smokers who had quit smoking less than one month ago were excluded.
b
In subjects having data on the smoking status.
 M.K. Pelkonen et al. / Respiratory Medicine 123 (2017) 87e93
Table 2
Bridging of the three revisions of the International Classification of Diseases (ICD).
Cause of hospitalization/death ICD-8
Respiratory 460e519
COPD 490e493, 518
Cardiovascular 390e438, 450 (excluding 435)
Ischemic heart disease 410e414
Cancer 140e199, 230e239
Lung cancer 162
Other causes Rest (001e796), E800eE999, N800eN999, Y00eY89
death were coded into the same four major categories as the hos-
pitalization diagnoses (respiratory, cardiovascular, cancer, and
other causes). There were altogether six subjects with a missing
cause of death.
2.4. Statistical methods
There were altogether 40756 subjects with all the variables
needed for the multivariate analyses. At first, a non-parametric test
was used to calculate if the distribution of the annual days of
hospitalization was different in subjects without and with chronic
bronchitis (Mann- Whitney U Test). Analyses were done separately
for subjects with examination in 1982 (the outcome was possible
hospitalization between 1982 and 2011) (n ¼ 8430) and for subjects
with examination in 1982, 1987, 1992, 1997, 2002 or 2007 (the
outcome was possible hospitalization in the following five years
after each exam) (n ¼ 40756). Analysis of covariance (ANCOVA) was
used for multivariate analyses. Our data did not follow a normal
distribution. Because of the large study population, however, the
ANCOVA analysis gives a robust estimate on the significance of the
differences. The ANCOVA analyses were adjusted for the following
variables: sex, the year of examination, the 10-year age-group,
smoking status (a three-category variable), education and marital
status and the area of residence (four-category variables), the
presence of hypertension and a history of diagnosed myocardial
infarction as binary variables. In smokers, an additional adjustment
for pack-years (¼ years of smoking  the daily number of smoked
cigarettes/20) was done. When studying the interaction effect be-
tween chronic bronchitis and the study year the interaction term
chronic bronchitis * the study year was added to the ANCOVA an-
alyses. Similarly, when studying the interaction effect between
chronic bronchitis and the smoking category the interaction term
chronic bronchitis * smoking category was added to the analyses.
The c2 test was used to compare the number of the four major
classifications of diseases by the presence of chronic bronchitis.
Cox proportional time-dependent hazards regression models
were used to analyse the relation between the presence of chronic
bronchitis and all cause and cause-specific mortality. Among 40756
subjects altogether 1278 subjects had a duplicate case in the FIN-
RISK Study between 1982 and 2007. Among these subjects with
duplicate measurements the first values were used in the mortality
analyses. The mortality analyses (n ¼ 39472) were adjusted for the
same co-variables as the ANCOVA analyses (sex, the year of ex-
amination, the 10-year age-group, smoking status, education,
marital status, the area of residence, the presence of hypertension
and a history of diagnosed myocardial infarction).
Statistical analyses were performed by SPSS 19.0 for Windows.
The study was approved in 2001 by the ethics committee at the
National Institute for Health and Welfare in Helsinki.
3. Results
In subjects with the examination in 1982, the means for the
annual days of hospitalization (- between 1982 and 2011-) were
89
ICD-9 ICD-10
460e519 J00eJ99
490e494, 496 J40eJ47
390e438 (excluding 435) I00eI69
410e414 I20eI25
140e199, 235e239 C00eC80, C97, D37eD48
162 C33eC34
Rest (001e999), E800eE999, V01eV82 Rest (A00eZ99)
almost two-fold higher in those with chronic bronchitis than in
those without it, and the means were greater in subjects with
chronic bronchitis in all age groups (Table 3). The 50th and 90th
percentiles for these means (all age groups together) were 0.76 and
8.21 in subjects without chronic bronchitis and 1.64 and 13.4,
respectively, in subjects with chronic bronchitis (not shown in
tables).
The means for the annual days of hospitalization for the 5 years
after each examination were also significantly higher in those with
chronic bronchitis than in those without it for all age groups
(Table 3). The 50th and 90th percentiles for these means (all age
groups together) were 0.00 and 2.40 in subjects without chronic
bronchitis and 0.00 and 4.40, respectively, in subjects with chronic
bronchitis (not shown).
In multivariate analyses (Table 4), in subjects with chronic
bronchitis the cause-specific hospitalization (-between 1982 and
2011-) was increased for COPD and lung cancer. For the 5 years after
each exam, in subjects with chronic bronchitis the cause-specific
hospitalization was increased for respiratory causes (- in general
and COPD-) and for cancer, (- and more specifically lung cancer-)
(Table 4). In subjects with the exam in 1982 (not shown in tables),
altogether 36%, 37%, 22% and 5%, respectively, of subjects with
chronic bronchitis were hospitalized because of one, two, three and
all four major classifications of diseases (the corresponding per-
centages in subjects without chronic bronchitis were 49%, 36%, 13%
and 2%) (p < 0.001). For the 5 years after each exam (not shown),
there was a similar trend so that subjects with chronic bronchitis
were significantly (p < 0.001) more often hospitalized for several
major causes than subjects without chronic bronchitis (the corre-
sponding percentages, respectively, in subjects with (and without)
chronic bronchitis were 79%, 18%, 3% and 0.3% (87%, 12%, 1% and
0.1%)).
In the multivariate analyses, there was a decreasing trend in the
annual days of hospitalization during the follow-up (p < 0.001) (not
shown in tables). There was an interaction (p ¼ 0.027) between
chronic bronchitis and the study year so that chronic bronchitis
increased all-cause hospitalization more at the beginning than at
the end of the follow-up (Table 5). The interaction between chronic
bronchitis and smoking on 5-year all-cause hospitalization was also
significant (p ¼ 0.001) so that hospitalization due to chronic
bronchitis was increased in smokers and ex-smokers (Table 5). In
smokers, after additional adjusting for pack-years all-cause hospi-
talization was still increased (p < 0.001) in those with chronic
bronchitis (not shown).
Subjects with chronic bronchitis had an increased risk to die and
cause-specific mortality was increased from respiratory causes
(and COPD), cardiovascular diseases, cancer (and lung cancer) and
other causes (Table 6). There was an interaction (p ¼ 0.021) be-
tween gender and chronic bronchitis on all-cause mortality so that
especially men with chronic bronchitis had an increased mortality
risk. The interaction between chronic bronchitis and smoking was
also significant (p ¼ 0.019) (Fig. 1) so that it was the smokers and
ex-smokers with chronic bronchitis who were at increased risk of
death and not the non-smokers.
90 M.K. Pelkonen et al. / Respiratory Medicine 123 (2017) 87e93

Table 3
Means (standard deviation ¼ SD) for the annual days of hospitalization by sex and age at the examination and p-values for means from non-parametric tests.b

Age at the examination Men

Examination 1982, hospitalization 1982e2011 Examination 1982e2007d, 5-year hospitalization thereafter

a c
n Chronic bronchitis p-value na Chronic bronchitis p-valuec
No Yes No Yes

Mean (SDb) Mean (SDb) Mean (SDb) Mean (SDb)

25e34 950/135 1.07 (5.79) 1.59 (4.23) <0.001 3603/484 0.49 (3.13) 0.86 (4.69) <0.001
35e44 960/171 1.82 (5.29) 2.98 (8.79) <0.001 3965/623 0.65 (3.56) 1.39 (7.02) <0.001
45e54 944/133 4.19 (11.8) 6.18 (11.8) <0.001 3902/754 1.18 (6.59) 2.20 (7.14) <0.001
55e64 686/226 8.09 (14.9) 11.2 (18.9) 0.001 3756/922 2.12 (8.56) 4.46 (15.7) <0.001
65e74 e e e e 1333/329 3.82 (15.3) 5.24 (11.6) <0.001
All 3436/728 3.44 (10.1) 6.14 (13.5) <0.001 16 559/3112 1.33 (7.21) 2.82 (10.7) <0.001

Age at the examination Women

Examination 1982, hospitalization 1982e2011 Examination 1982e2007d, 5-year hospitalization thereafter
na Chronic bronchitis p-valuec na Chronic bronchitis p-valuec
No Yes No Yes
Mean (SDb) Mean (SDb) Mean (SDb) Mean (SDb)

25e34 917/99 1.11 (2.90) 1.25 (1.78) 0.009 4265/407 1.06 (3.33) 1.88 (12.1) 0.029
35e44 970/122 1.47 (4.56) 5.76 (27.0) 0.005 4487/561 0.78 (3.26) 1.16 (5.00) 0.005
45e54 840/196 3.81 (14.0) 4.76 (9.71) 0.035 4467/637 1.08 (6.15) 1.32 (6.51) 0.002
55e64 920/161 10.12 (21.2) 11.15 (16.9) <0.001 4233/703 1.63 (8.42) 2.07 (6.58) <0.001
65e74 e e e e 1093/232 3.20 (13.8) 3.65 (9.61) 0.019
All 3751/515 4.09 (13.4) 6.32 (17.3) <0.001 18 545/2540 1.26 (6.48) 1.79 (7.78) <0.001

Altogether n ¼ 8430 (7187 and 1243, respectively, without and with chronic bronchitis) in 1982.
Altogether n ¼ 40 756 (35 104 and 5652, respectively, without and with chronic bronchitis) in 1982, 1987, 1992, 1997, 2002 and 2007.
a
N in analyses without/with chronic bronchitis.
b
SD ¼ standard deviation.
c
Mann-Whitney U test.
d
Examination in 1982/1987/1992/1997/2002/2007.

Table 4
Means (standard deviation ¼ SD) and multivariate adjusted p-values for the annual days of cause-specific hospitalization by the presence of chronic bronchitis.

Cause of hospitalization Examination in 1982, hospitalization 1982e2011

n of subjects with hospitalizations 1982-2011a Chronic bronchitis p-valuec

No Yes

mean (SDb) mean (SDb)

Respiratory 1420/396 0.33 (4.57) 0.72 (5.18) 0.069

COPD 296/141 0.05 (0.87) 0.24 (2.03) <0.001
Cardiovascular 2336/531 0.84 (6.11) 1.17 (7.07) 0.784
Ischemic heart disease 1226/285 0.23 (2.32) 0.26 (1.19) 0.138
Cancer 1293/288 0.49 (4.60) 0.88 (6.03) 0.193
Lung cancer 101/62 0.04 (0.73) 0.33 (4.32) <0.001
Other 6090/1110 2.13 (6.92) 3.44 (9.07) <0.001
All causes 6413/1168 3.78 (11.9) 6.21 (15.2) <0.001

Examination in 1982/1987/1992/1997/2002/2007, 5-year hospitalization thereafter

n of subjects with 5-year hospitalizationsa Chronic bronchitis p-valuec
No Yes
mean (SDb) mean (SDb)

Respiratory 932/390 0.05 (0.85) 0.18 (1.86) <0.001

COPD 143/116 0.01 (0.33) 0.07 (1.34) <0.001
Cardiovascular 1673/466 0.20 (3.48) 0.34 (2.99) 0.562
Ischaemic heart disease 761/238 0.06 (0.71) 0.11 (0.83) 0.088
Cancer 834/202 0.16 (2.90) 0.36 (4.33) 0.003
Lung cancer 33/36 0.02 (0.83) 0.10 (2.61) <0.001
Other 11 875/2332 0.88 (4.75) 1.48 (7.31) <0.001
All causes 13 176/2675 1.29 (6.83) 2.36 (9.52) <0.001

Altogether n ¼ 8430 (7187 and 1243, respectively, without and with chronic bronchitis) in 1982.
Altogether n ¼ 40 756 (35 104 and 5652, respectively, without and with chronic bronchitis) in 1982, 1987, 1992, 1997, 2002 and 2007.
a
N of subjects without/with chronic bronchitis and with hospitalizations (and with data needed for the covariance analyses).
b
SD ¼ standard deviation.
c
From ANCOVA analyses for each cause of hospitalization separately, adjusted for the following variables: sex, age group, smoking status, body mass index, the year of
examination, education, marital status, a history of myocardial infarction, the presence of hypertension and the area of residence.

4. Discussion bronchitis than in those without it. More specifically, hospitaliza-

tion in subjects with chronic bronchitis was increased from respi-
In the present study, the mean annual days of hospitalization ratory causes and cancer. Especially smokers and ex-smokers with
were approximately two-fold higher in subjects with chronic chronic bronchitis had increased hospitalizations and mortality.
 M.K. Pelkonen et al. / Respiratory Medicine 123 (2017) 87e93 91

Table 5 1.47-fold increased risk of lung cancer [23]. In the NHANES study,
Means (standard deviation ¼ SD) and multivariate adjusted p-values for the annual subjects aged 45 þ with COPD were more likely than subjects
days of 5 year hospitalization by the year of examination, smoking status and the
presence of chronic bronchitis.
without COPD to have cancer [24]. In the present study, in subjects
with chronic bronchitis the cause-specific hospitalization for lung
n of subjects Chronic bronchitis p-valuec cancer was also clearly increased. In addition, in the present study,
with hospitalizations
in 5 years after exama
No Yes there was an additive effect so that subjects with chronic bronchitis
mean (SDb) mean (SDb) more often had hospitalizations because of several major causes of
diseases than those without chronic bronchitis.
Study year
1982 2846/615 1.54 (6.46) 3.05 (12.14) <0.001 In the present study, there was a decreasing trend in the annual
1987 1873/397 1.52 (7.65) 2.80 (11.62) 0.011 days of hospitalization during the follow-up, which is in line with
1992 1953/410 1.09 (4.32) 2.02 (7.80) <0.001 an earlier result showing a decrease in average length of stay at
1997 2452/517 1.38 (8.00) 2.34 (8.32) 0.085 hospital after 1970 [25]. In addition, in the present study chronic
2002 2447/481 1.08 (6.18) 2.07 (7.51) 0.005
2007 1605/255 1.10 (7.77) 1.52 (7.97) 0.723
bronchitis was more strongly associated with hospitalizations at
Smoking status the beginning of the follow-up than at the end. This may at least be
Never 6173/870 1.27 (7.47) 1.79 (6.52) 0.211 partly explained by decreasing male smoking between 1982 and
Ex-smoker 3812/591 1.30 (6.10) 2.64 (9.12) <0.001 2007 [13e15]. Furthermore, in the present study chronic bronchitis
Smoker 3191/1214 1.31 (6.37) 2.66 (11.45) <0.001
was associated with hospitalization and mortality especially in
Altogether n ¼ 40 756 (35 104 and 5652, respectively, without and with chronic smokers and ex-smokers, and chronic bronchitis was not of great
bronchitis) in 1982, 1987, 1992, 1997, 2002 and 2007. importance with regards to mortality and hospitalizations among
a
N of subjects without/with chronic bronchitis and with hospitalizations (and
with data needed for covariance analyses).
never-smokers.
b
SD ¼ standard deviation. Previously, the reduction in life expectancy due to severe COPD
c
From ANCOVA analyses for examination year separately, adjusted for the has been approximately two years in never-smokers compared
following variables: sex, age group, smoking status, body mass index, the year of with a nine-year loss in smokers with severe COPD [26]. On the
examination, education, marital status, a history of myocardial infarction, the
other hand, in a longitudinal study never smokers with chronic
presence of hypertension and the area of residence.
bronchitis also had increased mortality when the presence of
chronic bronchitis and the decline in pulmonary function were
In an earlier study, the risk of all-cause hospitalizations (in the evaluated in repeated examinations [5]. Earlier, chronic bronchitis
past three months) increased from 6% in subjects without any has been a risk factor for mortality with HRs of 1.2e1.3[5,10]. In
respiratory conditions to 15e18% in subjects with asthmatic bron- subjects with chronic bronchitis, cause-specific mortality is
chitis [11]. In another earlier study (the COPDgene Study) [20], in increased from COPD, respiratory diseases, lung cancer, cardiovas-
subjects with COPD the presence of chronic bronchitis led to more cular disease and coronary heart disease [5,10,27,28]. According to
severe exacerbations of COPD requiring hospitalizations. During an our earlier results from the FINRISK Study male smokers with
8e10-year follow-up, in subjects with chronic bronchitis the FEV1 chronic bronchitis had smoked more than their female counter-
-adjusted relative risk of COPD hospitalization was two to three- parts [13] which may partly explain the interaction between gender
fold higher than in subjects without chronic bronchitis [6]. and chronic bronchitis on mortality.
Among subjects with COPD aged 40years, those with chronic The strength of this study was a large study population, a long
bronchitis have had lower lung function, worse general health follow-up and comprehensive hospitalization and mortality data.
status and more exacerbations of chronic obstructive pulmonary In the present study, the participation rates decreased from older
disease than subjects without chronic bronchitis [21]. In the pre- cohorts to younger cohorts probably leaving out more of those with
sent study hospitalizations from respiratory causes were more than higher morbidity than healthier individuals. However, according to
three-fold higher in subjects with chronic bronchitis than in sub- an earlier questionnaire study non-responders, evaluated by a
jects without chronic bronchitis, and the risk for COPD hospitali- telephone interview, have not reported more respiratory symptoms
zation was more than 4-fold higher. The causal relationship may go than responders [29]. On the other hand, in another postal ques-
as follows: smoking or other airborne irritating particles [3,22] lead tionnaire study elderly smokers tended to under-report the
to chronic bronchitis which, in turn, leads to increased hospitali- symptoms of chronic bronchitis [30]. In our study, three different
zations, partly through the development of COPD. revisions of the International Classification of Diseases were used.
In a pooled analysis, a history of chronic bronchitis conferred a Thus, there may have been changes in reporting of the causes of

Table 6
Number of deaths in subjects without and with chronic bronchitis and multivariate adjusted hazard ratios (95% confidence interval ¼ CI)a for all cause and cause-specific
mortality 1982e2011.

Cause of death n of deaths Hazard ratio (95% CI)b p-value

No chronic bronchitis Chronic bronchitis

Respiratory 162 75 1.64 (1.24e2.17) 0.002

COPD 62 47 2.12 (1.44e3.14) <0.001
Cardiovascular 1758 514 1.22 (1.10e1.35) <0.001
Ischemic heart disease 1162 333 1.13 (0.99e1.28) 0.064
Cancer 1207 345 1.31 (1.15e1.48) <0.001
Lung cancer 209 121 1.77 (1.40e2.23) <0.001
Other causes 1367 336 1.16 (1.02e1.31) 0.022
All cause 4494 1270 1.24 (1.16e1.32) <0.001
a
From Cox proportional hazards regression model, adjusted for sex, age group, smoking status, body mass index, the year of examination, education, marital status, a
history of myocardial infarction, the presence of hypertension and the area of residence. Subjects without chronic bronchitis were the reference group. Altogether n ¼ 39 472
(34 012 and 5460, respectively, without and with chronic bronchitis).
b
95% CI ¼ 95% confidence interval.
92 M.K. Pelkonen et al. / Respiratory Medicine 123 (2017) 87e93

Fig. 1. Multivariate adjusted hazard ratios (HR) for all-cause mortality during 1982e2011 by smoking status and the presence of chronic bronchitis from Cox proportional hazards
regression model. Adjusted for age group, the year of examination, education, marital status, body mass index, a history of myocardial infarction, the presence of hypertension and
the area of residence. Never smokers without chronic bronchitis were the reference group. Altogether n ¼ 39,472 (34,012 and 5460, respectively, without and with chronic
bronchitis). I-bars indicate confidence intervals.

hospitalizations and deaths during the follow-up. However, these
changes in reporting were independent of the presence of chronic
bronchitis at each examination.
Another limitation was that there was only the baseline mea-
surement of smoking and the other adjusted variables. We know
that a large proportion of smokers have stopped smoking during
the follow up [1415]. However, for example misclassification of
smokers and non-smokers would only weaken the observed as-
sociation between chronic bronchitis and hospitalization and
mortality.
Unfortunately, in the present study pulmonary functions were
not measured and thus COPD classification could not be done.
However, productive cough in COPD has been previously associated
with the highest risk for exacerbations and a higher mortality risk
[12]. Furthermore, in a recent study current and former smokers
with respiratory symptoms had more exacerbations and evidence
of airway disease than corresponding asymptomatic smokers [31].
The means for the annual days of hospitalization were greater
for the 30-year period than for the 5-year periods, probably mostly
resulting from aging of the 1982 study cohort. Generally, the 50th
percentiles (¼medians) for the annual days of hospitalization were
clearly lower than the calculated means, meaning that half of the
subjects did not have many yearly hospitalization days during the
30-year follow-up. The 50th percentiles for the annual days of
hospitalization for the 5-year periods were extremely low, indi-
cating that at least half of the subjects did not have any yearly
hospitalization days then.
In developed countries smoking has been decreasing in men
and in more educated women [13e15,32,33]. Consequently, recent
trends in chronic bronchitis have been decreasing, and the preva-
lence of COPD has been stable [13,34]. The decreasing trends in
smoking and chronic bronchitis may in turn lead to decreasing
trends in hospitalizations.
Funding
The VTR Funding of the Kuopio University Hospital.
Conflict of interest statement
The material has not been published and is not under consid-
eration for publication elsewhere. No author has any conflict of
interest. We all have read the paper and approved the submission
as well as approved our names in this paper and thus sign consent
to publication.
References
[1] J.M. Sethi, C.L. Rochester, Smoking and chronic obstructive pulmonary disease,
Clin. Chest Med. 21 (1) (2000) 67e86.
[2] M.D. Eisner, N. Anthonisen, D. Coultas, N. Kuenzli, R. Perez-Padilla, D. Postma,
I. Romieu, E.K. Silverman, J.R. Balmes, Committee on Nonsmoking COPD,
Environmental and Occupational Health Assembly. An official American
 M.K. Pelkonen et al. / Respiratory Medicine 123 (2017) 87e93
Thoracic Society public policy statement: novel risk factors and the global
burden of chronic obstructive pulmonary disease, Am. J. Respir. Crit. Care Med.
182 (5) (2010) 693e718.
[3] J. Sunyer, D. Jarvis, T. Gotschi, R. Garcia-Esteban, B. Jacquemin, I. Aguilera,
U. Ackerman, R. de Marco, B. Forsberg, T. Gislason, J. Heinrich, D. Norb€ ack,
S. Villani, N. Künzli, Chronic bronchitis and urban air pollution in an inter-
national study, Occup. Envir Med. 63 (2006) 836e843.
[4] W.G. Kuschner, S. Hegde, M. Agrawal, Occupational history quality in patients
with newly documented, clinician-diagnosed chronic bronchitis, Chest 135 (2)
(2009) 378e383.
[5] M. Pelkonen, I.-L. Notkola, A. Nissinen, H. Tukiainen, H.O. Koskela, Thirty-year
cumulative incidence of chronic bronchitis and COPD in relation to 30-year
pulmonary function and 40-year mortality. A follow-up in middle-aged ru-
ral men, Chest 130 (4) (2006) 1129e1137.
[6] J. Vestbo, E. Prescott, P. Lange, Association of chronic mucus hypersecretion
with FEV1 decline and chronic obstructive pulmonary disease morbidity. The
Copenhagen City Heart Study Group, Am. J. Respir. Crit. Care Med. 153 (5)
(1996) 1530e1535.
[7] E. Prescott, P. Lange, J. Vestbo, Chronic mucus hypersecretion in COPD and
death from pulmonary infection, Eur. Respir. J. 8 (8) (1995) 1333e1338.
[8] M. Miravitlles, C. Murio, T. Guerrero, R. Gisbert, Costs of chronic bronchitis and
COPD. A 1-year follow-up study, Chest 123 (3) (2003) 784e791.
[9] S. Guerra, D.L. Sherill, C. Venker, C.M. Ceccato, M. Halonen, F.D. Martinez,
Chronic bronchitis before age 50 years predicts incident airflow limitation and
mortality risk, Thorax 64 (10) (2009) 894e900.
[10] P. Lange, J. Parner, E. Prescott, J. Vestbo, Chronic bronchitis in an elderly
population, Age Ageing 32 (6) (2003) 636e642.
[11] S. Accordini, A.G. Corsico, L. Calciano, R. Bono, I. Cerveri, A. Fois, P. Pirina,
R. Tassinari, G. Verlato, R. de Marco, The impact of asthma, chronic bronchitis
and allergic rhinitis on all-cause hospitalizations and limitations in daily ac-
tivities: a population-based observational study, BMC Pulm. Med. 15 (2015)
10.
[12] A. Lindberg, S. Sawalha, L. Hedman, L.G. Larsson, B. Lundb€ ack, E. Ro €nmark,
Subjects with COPD and productive cough have an increased risk for exac-
erbations and death, Respir. Med. 109 (1) (2015) 88e95.
[13] M.K. Pelkonen, I.L. Notkola, T.K. Laatikainen, H.O. Koskela, Twenty-five year
trends in prevalence of chronic bronchitis and the trends in relation to
smoking, Respir. Med. 108 (11) (2014) 1633e1640.
[14] E. Vartiainen, T. Laatikainen, M. Peltonen, A. Juolevi, S. M€ annisto€, J. Sudvall,
P. Jousilahti, V. Salomaa, L. Valsta, P. Puska, Thirty-five-year trends in car-
diovascular risk factors in Finland, Int. J. Epidemiol. 39 (2) (2010) 504e518.
[15] E. Vartiainen, P. Jousilahti, G. Alfthan, J. Sundvall, P. Pietinen, P. Puska, Car-
diovascular risk factor changes in Finland, 1972-1997, Int. J. Epidemiol. 29 (1)
(2000) 49e56.
[16] WHO MONICA project Principal Investigators, The World Health Organization
MONICA Project (monitoring trends and determinants of cardiovascular dis-
ease): a major international collaboration, J. Clin. Epidemiol. 41 (2) (1988)
105e114.
[17] H. Tolonen, K. Kuulasmaa, T. Laatikainen, Wolf H and the European Health
Risk Monitoring Project. Recommendation for Indicators, International
Collaboration, Protocol and Manual of Operations for Chronic Disease Risk
Factor Surveys. October 2002, www-publications from the EHRM Project,
http://www.ktl.fi/ehrm.
[18] G.A. Rose, H. Blackburn, Cardiovascular Survey Methods, Monograph Series,
World Health Organization, Geneva, 1968, pp. 177e178.
[19] F. Janssen, A.E. Kunst, ICD coding changes and discontinuities in trends in
cause-specific mortality in six European countries, 1950-99, Bull. World
Health Organ. 82 (12) (2004) 904e913.
[20] V. Kim, M.K. Han, B.V. Gwendolyn, B.J. Make, J.D. Newell, J.E. Hokanson,
93
C.P. Hersh, D. Stinson, E.K. Silverman, G.J. Criner, The chronic bronchitic
phenotype of COPD. An analysis of the COPDGene Study, Chest 140 (2011)
626e633.
[21] M. Montes de Oca, R.J. Halbert, M.V. Lopez, R. Perez-Padilla, C. Talamo,
D. Moreno, A. Muino, R.B. Jardim, G. Valdivia, J. Pertuze, A.M.B. Menezes, The
chronic bronchitis phenotype in subjects with and without COPD: the PLAT-
INO study, Eur. Repir J. 40 (2012) 28e36.
[22] J. Regalado, R. Perez-Padilla, R. Sansores, S.I. Ramirez, M. Brauer, P. Pare ,
S. Vedal, The effect of biomass burning on respiratory symptoms and lung
function in rural Mexican women, Am. J. Respir. Crit. Care Med. 174 (2006)
901e905.
[23] D.R. Brenner, P. Boffetta, E.J. Duell, H. Bickebo €ller, A. Rosenberger,
V. McCormack, J.E. Muscat, P. Yang, H.-E. Wichmann, I. Brueske-Hohlfeld,
A.G. Schwartz, M.L. Cote, Anne Tjonneland, S. Friis, L. Le Marchand, Z.-
F. Zhang, H. Morgenstern, N. Szeszenia-Dabrowska, J. Lissowska, D. Zaridze,
P. Rudnai, E. Fabianova, L. Foretova, V. Janout, V. Bencko, M. Schejbalova,
P. Brennan, I.N. Mates, P. Lazarus, J.K. Field, O. Raji, J.R. McLaughlin, G. Liu,
J. Wiencke, M. Neri, D. Ugolini, A.S. Andrew, Q. Lan, W. Hu, I. Orlow, B.J. Park,
R.J. Hung, Previous lung diseases and lung cancer risk: a pooled analysis from
the International Lung Cancer Consortium, Am. J. Epidemiol. 176 (2012)
573e585.
[24] K. Schnell, C.O. Weiss, T. Lee, J.A. Krishan, J.L. Wolff, C. Boyd, The prevalence of
clinically-relevant comorbid conditions in patients with physician-diagnosed
COPD: a cross-sectional study using data from NHANES 1999-2008, BMC
Pulm. Med. 12 (2012) 12e26.
[25] V.R. Fuchs, Major trends in the U.S. Health economy since 1950, N. Engl. J.
Med. 366 (11) (2012) 973e977.
[26] R.M. Shavelle, D.R. Paculdo, S.J. Kush, D.M. Mannino, D.J. Strauss, Life expec-
tancy and years of life lost in chronic obstructive pulmonary disease: findings
from the NHANES III follow-up study, Int. J. Chron. Obstruct Dis. 4 (2009)
137e148.
[27] P. Lange, J. Nyboe, M. Appleyard, G. Jensen, P. Schnohr, Relation of ventilatory
impairment and of chronic mucus hypersecretion to mortality from
obstructive lung disease and from all causes, Thorax 45 (8) (1990) 579e585.
[28] P. Jousilahti, E. Vartiainen, J. Tuomilehto, P. Puska, Symptoms of chronic
bronchitis and the risk of coronary disease, Lancet 348 (9027) (1996)
567e572.
[29] €nmark, L. Ekerljung, J. Lo
E.P. Ro €tvall, K. Tore
n, E. Ro
€ nmark, B. Lundba €ck, Large
scale questionnaire survey on respiratory health in Sweden: effects of late-
and non-response, Respir. Med. 103 (12) (2009) 1807e1815.
[30] B. Lundba €ck, N. Stjernberg, L. Nystro € m, K. Lundb€ ack, M. Lindstro € m,
L. Rosenhall, An interview study to estimate prevalence of asthma and chronic
bronchitis, Eur. J. Epidemiol. 9 (1993) 124e133.
[31] P.G. Woodruff, R.G. Barr, E. Bleecker, S.A. Christenson, D. Couper, J.L. Curtis,
N.A. Gouskova, N.N. Hansel, E.A. Hoffman, R.E. Kanner, E. Kleerup, S.C. Lazarus,
F.J. Martinez, R. Paine III, S. Rennard, D.P. Tashkin, M.K. Han, Clinical signifi-
cance of symptoms in smokers with preserved pulmonary function, N. Engl. J.
Med. 374 (2016) 1811e1821.
[32] K. Giskes, A.E. Kunst, J. Benach, C. Borrell, G. Costa, E. Dahl, J.A.A. Dalstra,
B. Federico, U. Helmert, K. Judge, E. Lahelma, K. Moussa, P.O. Ostergren,
S. Platt, R. Prattala, N.K. Rasmussen, J.P. Mackenbach, Trends in smoking
behaviour between 1985 and 2000 in nine European countries by education,
J. Epidemiol. Community Health 59 (2005) 395e401.
[33] R. Peto, S. Darby, H. Deo, P. Silcocks, E. Whitley, R. Doll, Smoking, smoking
cessation, and lung cancer in the UK since 1950: combination of national
statistics with two case-control studies, BMJ 321 (7257) (2000) 323e328.
[34] T.M. Vasankari, O. Impivaara, M. Helio € vaara, S. Heistaro, K. Liippo, P. Puukka,
S. Saarelainen, M. Kanervisto, P. Jousilahti, No increase in the prevalence of
COPD in two decades, Eur. Respir. J. 36 (2010) 766e773.
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.