Respiratory symptoms and respiratory deaths: A multi-cohort study with 45 years observation time.

This study determined the association between respiratory symptoms and death from respiratory causes over a period of 45 years. In four cohorts of random samples of Norwegian populations with 103,881 participants, 43,731 persons had died per 31 December 2016. In total, 5,949 (14%) had died from respiratory diseases; 2,442 (41%) from lung cancer, 1,717 (29%) chronic obstructive pulmonary disease (COPD), 1,348 (23%) pneumonia, 119 (2%) asthma, 147 (2%) interstitial lung disease and 176 (3%) other pulmonary diseases. Compared with persons without respiratory symptoms the multivariable adjusted hazard ratio (HR) for lung cancer deaths increased with score of breathlessness on effort and cough and phlegm, being 2.6 (95% CI 2.1-3.2) for breathlessness score 3 and 2.1 (95% CI 1.7-2.5) for cough and phlegm score 5. The HR of COPD death was 6.4 (95% CI 5.4-7.7) for breathlessness score 3 and 3.0 (2.4-3.6) for cough and phlegm score 5. Attacks of breathlessness and wheeze score 2 had a HR of 1.6 (1.4-1.9) for COPD death. The risk of pneumonia deaths increased also with higher breathlessness on effort score, but not with higher cough and phlegm score, except for score 2 with HR 1.5 (1.2-1.8). In this study with >2.4 million person-years at risk, a positive association was observed between scores of respiratory symptoms and deaths due to COPD and lung cancer. Respiratory symptoms are thus important risk factors, which should be followed thoroughly by health care practitioners for the benefit of public health.

Introduction

Respiratory diseases, such as pneumonia, lung cancer and chronic obstructive pulmonary disease (COPD), remain a leading cause of disability and death worldwide [1-3]. Respiratory symptoms are important indicators of these diseases, and recording of respiratory symptoms is a cheap and easy screening method.

In the 1970s, at least one respiratory symptom was reported by 40% of the adult inhabitants of the Oslo city: cough in the morning, attacks of breathlessness, breathless when climbing stairs or wheezing [4]. The symptom burden remains high also in recent times, with approximately 20% reporting wheeze and >25% cough [5].

Respiratory symptoms are associated with all-cause and cardiovascular mortality [6-9]. Less is known about the association between respiratory symptoms and respiratory causes of death. The limited number of respiratory deaths in previous studies [10-12], and the non-random nature of some of the cohorts [10], limit generalisability of findings. In a recent general population study with <15 years of follow-up time, individuals with chronic respiratory symptoms and normal spirometry had increased mortality due to respiratory diseases [8, 13].

Large cohorts are needed to observe a sufficient number of deaths due to asthma and interstitial pneumonia, which are rare causes of death in the general population. We therefore extended a previous analysis [12] to investigate the long-term relationship between respiratory symptoms and mortality for several respiratory diseases by combining four general population cohorts [6].

This study reports on the association of self-reported subsets of respiratory symptoms such as breathlessness on effort, cough and phlegm, and attacks of breathlessness/wheeze with the mortality from lung cancer, pneumonia/influenza, COPD, asthma and interstitial pneumonias over 45 years.

Materials and methods

Study population

This multicohort study used harmonized individual-level data from four cross-sectional surveys in the City of Oslo 1972 and 1998–1999, Hordaland County in 1985, 1988–1989 (including Sauda municipality in Rogaland county) and 1998–1999 [14]. Oslo, the capital of Norway, had 477,476 inhabitants in 1972 and 499,693 persons in 1998. Hordaland county is a combined rural and urban (Bergen) population with 399,702 persons in 1985, 405,063 in 1988 and 428,823 in 1998. Sauda is a rural municipality with 5,416 inhabitants in 1988.

The target populations were born 1902–1973. The sample frames were updated lists from the Norwegian National Population Registry. Invitees were drawn at random for the 1972 (Oslo 72), 1985 (Hordaland 85) and 1998–1999 (Oslo/Hordaland 98–99) surveys. The 1988–1990 survey (Støvlunge 88–90) invited all men born 1914–1958, plus a 10% sample of the general population of City of Bergen examined in a previous cohort in 1965–1970 [14, 15], leading to about 6% women in this cohort. We excluded those included in one cohort from later cohorts. The eligible population sample comprised 158,702 unique persons. We have previously presented the recruitment and pooling of these cohorts [6].

The analyses included respondents who provided information on smoking, education, occupational exposure to dust/gas and respiratory symptoms, in total 103,881 persons; 65% of the sample of unique persons (Table 1).

Table 1

Flowchart of randomly sampled individuals in the study of respiratory symptoms and all cause deaths in Norway.

	Oslo cohort 1972	Hordaland county cohort 1985	Hordaland county and Sauda municipality Rogaland county 1988–1990	Hordaland and Oslo counties cohort 1998–1999	Total
Sample drawn from target populations	19998	4992	112235	25000	162225
Present in previous samples, excluded	1*	2	1893	1627	3523
Eligible, unique persons	19997	4990	110342	23373	158702
Eligible after missing times excluded	19892	4982	108812	23210	156896
Respondents (response to at least 1 of 17 questionnaire items)	17690	4461	77003	15870	115024
Respondents to smoking habits	17680	4404	76675	15623	114380
Respondents to smoking habits and education	17377	4347	75406	14994	112124
Respondents to smoking habits, education and occupational exposure	16445	4307	71958	14765	107475
Respondents to all items on respiratory symptoms (in analysis)	16084	4137	69168	14492	103881

*removed duplicate record.

The study was approved by the Committee on Medical Research Ethics (reference 2017/1679), The Norwegian Data Inspectorate (07/00414) and The Norwegian Directorate of Health (07/948).

Questionnaire

We used a questionnaire, which was a modification of one approved by the British Medical Research Council`s (MRC) Committee on Research into Chronic Bronchitis in 1966. The validity of the Norwegian respiratory questionnaire has been evaluated [16], and compared with the original MRC questionnaire [17].

The questionnaire included 11 questions about respiratory symptoms (S1 Table), covering current cough, phlegm, wheezing, periods of cough and/or phlegm and breathlessness, which were aggregated into three symptom groups: breathlessness scored 0 to 4; cough and/or phlegm symptoms scored 0 to 5, and attacks of breathlessness and wheezing scored 0 to 2 [6]. A higher score represents more severe symptoms.

The questionnaire also included questions on smoking history [18] and occupational exposure to air pollution. Smoking was categorized as current smokers (daily at the time of the study), ex-smokers, or never-smokers. Tobacco consumption was estimated from number of cigarettes per day (1 cigarette = 1 g), and grouped as <10 g, 10–19 g and ≥20 g. Occupational exposure to air pollution was defined by responding “yes” to “Have you been exposed to particles, gases or damp at your working place?”

No self-reported cardiopulmonary disease was defined as negative replies to 10 questions on treatments by physician or hospitals of asthma, bronchitis, emphysema, pleuritis, lung tuberculosis, other pulmonary disease, myocardial infarction, angina pectoris, hearth failure or other hearth diseases.

Follow-up and census data

Date of death, emigration and cause of death until 31 December 2016 were obtained from the National Cause of Death Registry. All inhabitants of Norway have a unique personal identification number that allows complete follow-up until death or emigration. In total, 156,896 persons were initially observed. A total of 103,881 respondents with known smoking status and complete responses were followed; median follow-up was 27.4 years, maximum 45.2 years. They represented 2,449,538 person-years at risk.

Highest attained education was extracted from the national census for each decade and grouped according to the maximum length of education using three levels: compulsory education (7–10 years), medium level (11–13 years) and university level (≥14 years).

The overall inclusion rate in the analyses was 65% of the target population and varied for the baseline examination from 81% of the target population of the cohort of 1972 to 62% in the cohort of 1998–1999. The distribution of participants of the various groups was almost identical in the target population and the population of respondents with regard to age, sex and education (S2 Table).

Classification of causes of death

Classification of respiratory causes as underlying causes of death was done using the European Shortlist for Causes of Death, 2012 [19, 20], with two modifications: 1) Because there were few cases with influenza, this was combined with pneumonia; 2) Interstitial pneumonias were extracted from “Other diseases of the respiratory system” and analyzed as a separate entity (Table 2).

Table 2

Classification of underlying causes of death.

Respiratory causes	EU 2012 classification	ICD-8	ICD-9	ICD-10
Lung cancer (trachea, bronchus, lung)	15	162.0–162.9	162.0–162.9	C33, C34.0–C.34.9
COPD	56	491–492, 518	491–492, 494, 496	J42, J43.9, J44.0-J44.9, J47
Pneumonia incl. influenza	52, 53	480–480, 470–474	481,482, 485–486, 470, 471, 487	J10.1, J11.0-J11.1, J13–J14, J15.2–J15.9, J18.0–J18.9
Asthma	55	493	493, 493.9	J45.0, J45.9, J46
Interstitial pneumonias*	57 (subset)	515, 516.3, 516.9	515, 516,3, 516.9	J84.0, J84.1, J84.9
Other respiratory	57 (excl. interstitial pneumonias*)	460–478, 495, 500–519 (excl.*)	460–478, 495, 500–519 (excl. *)	J00-J06, J20–J39, J60–J99 (excl. *)

Statistical analysis

Descriptive statistics are presented in frequency tables as number (%). For start dates for observations in the cohorts, we used 5 October 1972 in the Oslo county 1972 cohort, the 15th of the actual starting month in the Hordaland county and Sauda municipality cohort 1988–1990, and the actual start date in the other two cohorts. Other missing start dates for participants were imputed using the median start date in the same cohort: 31 December 1989 in the Hordaland county and Sauda municipality cohort 1988–1990 (n = 100) and 5 October 1998 in the Oslo and Hordaland counties cohort 1998–1999 (n = 4).

The cohort members were followed until death or censored at the date of emigration or end of follow-up on 31 December 2016, whichever came first. For some people that emigrated (n = 231), we did not have a date of emigration, but only an interval. These cases were censored at the mid-point of the interval [6]. We did not impute missing values for other variables.

We pooled the four cohorts and analyzed the association of symptom scores of breathlessness on effort, cough or/and phlegm and attacks of breathlessness or/and wheeze with cause-specific mortality. We used Cox proportional hazards analysis, with age as the dependent variable [21]. We also repeated the analysis replacing the respiratory symptom scores with a variable with no respiratory symptom = 0, any respiratory symptom = 1. These analyses were prepared using shared frailty for study cohort, i.e. incorporating cluster-specific random effects to account for within-cluster homogeneity in outcomes [22].

All analyses were multivariable, adjusting for sex, education (<10, 11–13, ≥14 years), smoking habits (never, ex-, current-smoker) and occupational exposure (dust/fume vs. none). The results are presented as hazard ratios (HR) of death with 95% confidence intervals (CI).

Finally, we conducted analyses in strata of the pooled sample (men, women, never smokers, those without a history of cardiopulmonary disease) according to death from lung cancer, COPD and pneumonia using the same approach and the same covariates. For other causes of death, we did not conduct stratified analyses because of few events.

The proportional hazards assumption was checked graphically using log-log plots and was considered acceptable. We chose a significance level of 0.05 using two-sided tests. Stata version 16.1 (StataCorp, College Station, TX, USA) was used for all statistical analyses.

Results

Among the 103,881 individuals (response rate 64%), 78% were men, and mean baseline age was 46.8 years (range 15–92 years). Altogether 43,731 (42%) had died as of 31 December 2016 and had a specified cause of death. In total 5,949 persons (14% of all deaths) had a pulmonary death; lung cancer 2,442 (41%), COPD 1,717 (28%), and pneumonia 1348 (23%) including 40 cases of influenza deaths (Table 2). In addition, 147 (2.5%) died from interstitial pneumonia, 119 from asthma (2.0%) and 176 due to other pulmonary causes (3.0%).

The prevalence of baseline symptom scores and the distribution of symptoms according to the principal causes of death is shown in S3 Table.

Respiratory mortality

The overall crude mortality rate (MR) of respiratory diseases was 243 per 100,000 person-years. In total, 6.1% of men (4971/81510) and 4.4% of women (978/22371) died from respiratory causes. Among individuals with pulmonary deaths, 87% were ever smokers (Table 3).

Table 3

Crude number of deaths (%) according to respiratory causes in the pooled cohort*.

	All pulmonary	Lung cancer	COPD	Pneumonia	Asthma	Interstitial pneumonia	Other pulmonary
Age, years
15–29	90 (2)	57 (2)	21 (1)	6 (1)	2 (1)	2 (1)	2 (1)
30–44	593 (10)	376 (15)	140 (8)	44 (3)	6 (5)	12 (8)	15 (9)
45–59	2052 (34)	1007 (41)	601 (35)	295 (22)	39 (33)	50 (34)	60 (34)
≥60	3214 (54)	1002 (41)	955 (56)	1003 (74)	72 (61)	83 (56)	99 (56)
Sex
Male	4971 (84)	2138 (88)	1454 (85)	1018 (76)	87 (73)	126 (86)	148 (84)
Female	978 (16)	304 (12)	263 (15)	330 (24)	32 (27)	21 (14)	28 (16)
Highest attained education
Compulsory education (<11 years)	2656 (45)	1022 (42)	815 (47)	642 (48)	60 (50)	52 (35)	65 (37)
Medium level (11–13 years)	2861 (48)	1237 (51)	797 (46)	600 (45)	56 (47)	75 (51)	96 (55)
University level (>13 years)	432 (7)	183 (7)	105 (6)	106 (8)	3 (3)	20 (14)	15 (9)
Smoking status
Never	781 (13)	109 (4)	119 (7)	442 (33)	16 (13)	41 (28)	54 (31)
Previous	1378 (23)	434 (18)	366 (21)	424 (31)	34 (29)	69 (47)	51 (29)
Current	3790 (64)	1899 (78)	1232 (72)	482 (36)	69 (58)	37 (25)	71 (40)
No. of cigarettes per day
0–9	995 (21)	324 (15)	299 (21)	291 (37)	30 (35)	25 (26)	26 (24)
10–19	2283 (49)	1092 (51)	715 (49)	337 (43)	39 (46)	46 (48)	54 (49)
≥20	1395 (30)	732 (34)	442 (30)	150 (19)	16 (19)	25 (26)	30 (27)
Occupational exposure gas/dust
Yes	2921 (49)	1290 (53)	926 (54)	487 (36)	56 (47)	81 (55)	81 (46)
No	3028 (51)	1152 (47)	791 (46)	861 (64)	63 (53)	66 (45)	95 (54)
Cohort study
Oslo 72	1262 (21)	427 (17)	333 (19)	400 (30)	46 (39)	23 (16)	33 (19)
Hordaland 85	154 (3)	62 (3)	47 (3)	36 (3)	3 (3)	2 81)	4 (2)
Støvlunge 88–90	4313 (72)	1854 (76)	1263 (74)	887 (66)	66 (55)	115 (78)	128 (73)
Oslo/Hordaland 98–99	220 (4)	99 (4)	74 (4)	25 (2)	4 (3)	7 (5)	11 (6)
Total	5949 (100)	2442 (100)	1717 (100)	1348 (100)	119 (100)	147 (100)	176 (100)

*Some numbers do not add to a total of 100 percent because of rounding

The hazard of death due to pulmonary disease was higher in men compared with women, in those with only compulsory education compared with university education and in smokers versus never smokers (Table 4). Scores on breathlessness, cough and phlegm and attacks of breathlessness and ever wheezing were strongly associated with increased hazard of respiratory mortality. However, breathlessness showed a stronger association with pulmonary mortality than cough and phlegm did.

Table 4

Hazard ratios (HR) for death with 95% confidence intervals and p-values according to pulmonary cause of death, multivariable proportional hazards regression analysis (n = 103,881).

	All pulmonary		Lung cancer		COPD		Pneumonia incl. influenza		Asthma		Interstitial pneumonia		Other pulmonary
	HR	95%CI	HR	95%CI	HR	95%CI	HR	95%CI	HR	95%CI	HR	95%CI	HR	95%CI
Sex
Male	1		1		1		1		1		1		1
Female	0.69***	[0.64,0.75]	0.60***	[0.52,0.69]	0.61***	[0.52,0.72]	0.88	[0.75,1.03]	0.8	[0.48,1.31]	0.60*	[0.36,0.99]	0.55*	[0.34,0.89]
Highest attained education
Compulsory education (<11 years)	1		1		1		1		1		1		1
Medium level (11–13 years)	0.48***	[0.46,0.51]	0.55***	[0.50,0.60]	0.47***	[0.43,0.52]	0.38***	[0.34,0.42]	0.54**	[0.37,0.79]	0.59**	[0.41,0.84]	0.58***	[0.42,0.80]
University level (>13 years)	0.20***	[0.18,0.22]	0.25***	[0.21,0.29]	0.19***	[0.16,0.24]	0.13***	[0.11,0.16]	0.09***	[0.03,0.28]	0.36***	[0.21,0.62]	0.19***	[0.11,0.34]
Smoking
Never	1		1		1		1		1		1		1
Previous	1.88***	[1.72,2.06]	4.16***	[3.37,5.14]	2.92***	[2.37,3.60]	1.24**	[1.08,1.42]	2.18*	[1.18,4.03]	1.92**	[1.28,2.87]	1.05	[0.71,1.56]
Current	2.76***	[2.55,2.99]	10.57***	[8.69,12.86]	4.63***	[3.82,5.62]	0.76***	[0.66,0.87]	1.63	[0.92,2.88]	0.59*	[0.37,0.94]	0.9	[0.62,1.32]
Occupational exposure to air pollution
Yes	1		1		1		1		1		1		1
None	1.33***	[1.26,1.41]	1.21***	[1.11,1.32]	1.38***	[1.24,1.53]	1.64***	[1.45,1.85]	1.63*	[1.08,2.44]	0.80	[0.56,1.15]	1.22	[0.88,1.68]
Breathless on effort, score
0	1		1		1		1		1		1		1
1	1.44***	[1.33,1.55]	1.23***	[1.09,1.39]	1.96***	[1.71,2.24]	1.27*	[1.06,1.52]	1.21	[0.65,2.24]	1.6	[0.95,2.70]	2.22***	[1.45,3.40]
2	2.21***	[2.03,2.40]	1.84***	[1.61,2.10]	3.47***	[3.01,3.99]	1.66***	[1.34,2.06]	2.78***	[1.61,4.80]	1.56	[0.79,3.09]	1.03	[0.47,2.26]
3	3.86***	[3.43,4.33]	2.63***	[2.14,3.23]	6.44***	[5.37,7.73]	2.99***	[2.20,4.07]	6.34***	[3.51,11.44]	4.84***	[2.29,10.25]	4.61***	[2.14,9.93]
4	3.64***	[3.08,4.32]	1.69**	[1.17,2.45]	6.70***	[5.27,8.51]	3.05***	[1.88,4.95]	5.33***	[2.51,11.31]	3.64*	[1.04,12.81]	8.72***	[3.38,22.51]
Cough and phlegm, score
0	1		1		1		1		1		1		1
1	1.27***	[1.19,1.37]	1.24***	[1.11,1.38]	1.67***	[1.46,1.91]	1.08	[0.93,1.26]	0.96	[0.53,1.74]	1.25	[0.80,1.93]	1.07	[0.71,1.61]
2	1.64***	[1.51,1.79]	1.64***	[1.44,1.87]	1.97***	[1.67,2.32]	1.50***	[1.22,1.84]	1.03	[0.50,2.09]	1.32	[0.69,2.56]	1.78*	[1.08,2.94]
3	1.74***	[1.56,1.93]	1.70***	[1.45,1.99]	2.34***	[1.95,2.80]	1.37*	[1.04,1.81]	2.02*	[1.06,3.87]	0.64	[0.20,2.07]	0.28	[0.07,1.16]
4	2.09***	[1.87,2.34]	1.72***	[1.43,2.07]	3.10***	[2.58,3.73]	1.36	[0.97,1.90]	3.34***	[1.81,6.15]	2.34*	[1.06,5.19]	0.94	[0.36,2.43]
5	2.16***	[1.90,2.44]	2.06***	[1.70,2.51]	2.96***	[2.43,3.62]	1.43	[0.97,2.11]	1.47	[0.69,3.16]	1.82	[0.67,4.94]	1.26	[0.49,3.21]
Attacks of breathlessness and wheeze, score
0	1		1		1		1		1		1		1
1	1.06	[0.99,1.14]	1.05	[0.95,1.16]	1.33***	[1.18,1.50]	0.81**	[0.69,0.95]	2.94***	[1.74,4.98]	0.71	[0.43,1.18]	0.60*	[0.37,0.97]
2	1.16**	[1.06,1.27]	0.98	[0.85,1.14]	1.64***	[1.41,1.90]	0.64***	[0.49,0.83]	6.48***	[3.65,11.52]	1.05	[0.55,1.99]	0.96	[0.53,1.75]

* p<0.05

** p<0.01

*** p<0.001

The pattern of associations was similar in subgroups, except that associations for phlegm and cough were weaker and non-significant in women and never-smokers, as well as for attacks of breathlessness/wheeze in men and never-smokers (Table 5).

Table 5

Hazard ratios for death with 95% confidence intervals and p-values, all pulmonary causes of death, in subgroups of the total population.

Multivariable proportional hazards regression analysis.

	Men		Women		Never smokers		Without cardiopulmonary disease
	Hazard ratio	95%CI	Hazard ratio	95%CI	Hazard ratio	95%CI	Hazard ratio	95%CI
Highest attained education
Compulsory education (<11 years)	1		1		1		1
Medium level (11–13 years)	0.50***	[0.47,0.53]	0.42***	[0.37,0.48]	0.35***	[0.30,0.41]	0.56***	[0.51,0.61]
University level (>13 years)	0.21***	[0.19,0.24]	0.12***	[0.09,0.16]	0.13***	[0.10,0.17]	0.22***	[0.18,0.26]
Smoking
Never	1		1				1
Previous	2.32***	[2.08,2.58]	1.05	[0.83,1.32]			2.03***	[1.75,2.36]
Current	3.41***	[3.09,3.77]	1.67***	[1.44,1.93]			2.61***	[2.27,3.00]
Occupational exposure to air pollution
Yes	1		1		1		1
No	1.37***	[1.30,1.46]	1.00	[0.84,1.19]	1.58***	[1.33,1.88]	1.30***	[1.18,1.42]
Breathless on effort, score
0	1		1		1		1
1	1.52***	[1.40,1.66]	1.00	[0.82,1.23]	1.65***	[1.28,2.11]	1.47***	[1.31,1.66]
2	2.35***	[2.14,2.58]	1.54***	[1.25,1.90]	2.16***	[1.64,2.84]	2.39***	[2.12,2.69]
3	4.06***	[3.57,4.62]	2.79***	[2.12,3.67]	3.29***	[2.16,5.01]	4.05***	[3.50,4.67]
4	3.96***	[3.28,4.80]	2.45***	[1.68,3.56]	4.27***	[2.54,7.17]	3.61***	[2.96,4.41]
Cough and phlegm, score
0	1		1		1		1
1	1.31***	[1.21,1.42]	1.11	[0.93,1.32]	0.91	[0.73,1.12]	1.21**	[1.07,1.36]
2	1.74***	[1.58,1.92]	1.19	[0.95,1.50]	1.24	[0.88,1.75]	1.51***	[1.31,1.74]
3	1.87***	[1.67,2.09]	1.17	[0.88,1.54]	1.43	[0.96,2.13]	1.58***	[1.35,1.85]
4	2.14***	[1.89,2.42]	1.85***	[1.40,2.45]	1.3	[0.77,2.19]	2.12***	[1.82,2.48]
5	2.32***	[2.02,2.65]	1.51*	[1.10,2.07]	2.60***	[1.59,4.24]	1.96***	[1.66,2.31]
Attacks of breathlessness and wheeze, score
0	1		1		1		1
1	1.02	[0.95,1.10]	1.30**	[1.10,1.54]	1.15	[0.91,1.44]	1.12*	[1.01,1.24]
2	1.08	[0.98,1.20]	1.70***	[1.36,2.11]	1.29	[0.93,1.79]	1.06	[0.94,1.20]
Sex
Male					1		1
Female					1.22*	[1.01,1.47]	0.68***	[0.59,0.77]
N	81510		22371		34916		26723

* p<0.05

** p<0.01

*** p<0.001

Lung cancer

The crude MR of lung cancer was 100 per 100,000 person-years. Ex-smokers had fourfold and current smokers tenfold increased hazard of lung cancer death relative to never smokers. Breathlessness, cough and phlegm were also associated with an increased risk of lung cancer death, with an increased risk occurring from a score of 1 to 2 on both symptom scales (Table 4). Attacks of breathlessness and ever wheeze did not increase the risk for lung cancer deaths. Identical trends for risks of respiratory symptoms for lung cancer deaths were observed in stratified analyses of men, women, never smokers and in those without self-reported cardiopulmonary diseases (S4 Table).

Chronic obstructive pulmonary disease

The MR was 70 for COPD per 100,000 person-years. The pattern of adjusted HR for COPD with regard to sex and attained education was almost identical to the HR for lung cancer, but with lower HRs with regard to smoking habits (Table 4). The HR with increasing breathlessness on effort was higher for COPD than for other pulmonary causes of death.

The risk trends of respiratory symptoms for COPD deaths were similar for men, women, never smokers and for those without cardiopulmonary diagnosis (S5 Table).

Pneumonia

The MR of pneumonia was 55 per 100,000 person-years. The MR of pneumonia increased markedly for those aged ≥60 years at baseline compared with younger individuals (Table 3). There was no clear association between smoking habits and death due to pneumonia. There was a dose-response relationship between breathlessness score and pneumonia deaths, while this was not so obvious for cough and phlegm score (Table 4).

Associations between symptom burden and pneumonia deaths were not prominent in women. Attacks of breathlessness and wheeze score was negatively associated with deaths due to pneumonia in men (S6 Table).

Rare causes

Asthma was a rare cause of death with MR of 5 per 100,000 person-years, which was only 7% that of COPD deaths. The risk of asthma deaths showed the same pattern as COPD with regard to respiratory symptoms. However, higher attacks of breathless and wheeze score increased the risk of asthma mortality more than any other pulmonary causes of death (Table 4).

Interstitial pneumonia had a MR of 6 per 100,000 person-years. The risk pattern was similar for sex and education as for other pulmonary causes of deaths. The risk was lower in ex-smokers than in never smokers. An increased hazard of death from interstitial pneumonia was observed in those with breathlessness on effort (Table 4).

Occupational exposure to air pollution

There was a slightly increased crude rate of pulmonary death in those with occupational exposure to gas or dust. However, no increased risk was evident in multivariable models. It was rather the opposite trend with a higher risk of deaths in those without occupational exposure to dust and gases compared with those with occupational exposure, except for death from interstitial pneumonia (Table 4).

Any respiratory symptom

The presence of any respiratory symptom increased the HR (95%CI) for pulmonary death to 2.0 (1.9–2.1), lung cancer death to 1.7 (1.5–1.8), COPD death to 4.3 (3.8–4.8) and pneumonia to 1.4 (1.2–1.5) after adjustment for sex, education, smoking and occupational exposure (Fig 1).

Fig 1

Associations with cause-specific mortality.

Multivariable hazard ratios with 95% confidence intervals for any respiratory symptom (0 = no, 1 = yes), sex, education, smoking and occupational exposure to air pollution according to all pulmonary deaths, lung cancer, COPD or pneumonia. All subjects in pooled sample with available covariates (n = 107,136).

Discussion

This study has shown that respiratory symptoms were associated with respiratory deaths due to lung cancer, COPD, pneumonia and asthma. The scores of breathlessness, cough and phlegm had a dose-response relationship for specific causes of pulmonary deaths. These associations were stronger for deaths due to lung cancer and COPD than for pneumonia.

The burden of respiratory symptoms increased the risk of death due to lung disease, independent of subset of symptoms, although the subsets of respiratory symptoms were associated with different mortality risks. Breathlessness on effort was associated with a higher risk than cough and phlegm.

This study comprises a very large pooled cohort with a very long follow-up period, enabling us to present new data on cause-specific deaths. This is a strength of the present study and would not be possible or meaningful in smaller cohorts and explains why there is little previous information on the association between respiratory symptoms and the less common causes of respiratory deaths.

The study supports and extends the previous findings in one of the four cohorts with 30 years of follow up [12], which showed a strong association of respiratory symptoms with deaths due to obstructive lung disease and a weak association with pneumonia. We are aware of few other comparable studies with sub-classified respiratory death as outcomes. The Busselton study investigated the association between respiratory symptoms and respiratory deaths, except lung cancer which was grouped with cancers, with 20–26 years of follow-up. The study dichotomized the respiratory symptoms, whereas the present study had graded responses. The Busselton study included forced expiratory volume in 1 second (FEV1) and several cardiovascular risk factors in the analysis of about 4,300 subjects, in contrast to the present study with >100,000 subjects that included occupational exposure and education in the analysis, but no spirometry variable. The present study was far larger and had longer observation time than that study, which enabled the analysis of subgroups of respiratory deaths, such as lung cancer, COPD, pneumonia, and less common respiratory diseases as causes of death.

In a sample of the population of Copenhagen, Denmark, chronic respiratory symptoms were associated with deaths in individuals with normal spirometry and without known airway disease [8]. In another follow up study, dyspnoea was associated with all-cause mortality after adjustment for lung function, but specific pulmonary mortality was not analysed [23]. In a large primary care study, breathlessness was an early marker of chronic respiratory and cardiac disease, and it was associated with all-cause mortality, death from COPD as well as ischemic heart disease [24].

We did not find an association between symptoms of attacks of breathlessness/wheeze and death from lung cancer. In contrast, a recent study reported a possible association between asthma symptom control and the incidence of lung cancer [25].

Our findings are robust given that the pattern of associations between respiratory symptoms and cause-specific mortality was similar in strata of men and subjects without cardiopulmonary diseases, and the total pooled sample. In stratified analysis of women only and never-smokers, the effects were smaller or disappeared, in particular for cough and phlegm.

Strengths of the present study include a large population-based cohort study with randomly selected individuals with high response rate of the target population and with a long and complete follow-up. The outcome of respiratory-specific deaths is highly relevant in a clinical setting. Furthermore, the respiratory symptoms were self-reported and not biased by observers [4]. The items on respiratory symptoms were identical in all these sub-cohorts, and they were administered by postal questionnaires. All studies had the same primary investigator.

In the present population-based study, it is unlikely that we have missed individuals with substantial symptoms, but more likely that we have included more individuals with mild symptoms. Nevertheless, individuals with mild symptoms will only dilute an association and reduce the risk estimates. There was a clear dose-response relationship between respiratory symptoms and outcomes of death like lung cancer and COPD.

Some limitations of the study should be noted. In this large, pooled sample, we had access to only a limited number of common/harmonized covariates. It is possible that covariates such as cardiovascular risk factors, body mass index, genetic factors, more detailed history of allergy or infections may, or other variables may be confounders in the relationship between respiratory symptoms and the specific respiratory deaths. During such a long observation time, covariates may also change over time, however, this study only had available covariates at baseline.

Spirometry would have been a useful supplement to subjective perceptions of respiratory symptoms, but that was only available for a fraction of the participants in the pooled sample. However, spirometry is more resource-consuming than assessment of simple respiratory symptoms in clinical practice.

The largest sub-cohort included mostly only men, hence the pooled sample had relatively few women and may represent a bias, although we adjusted for sex in the analysis. In a previous report, however, we have shown a similar pattern of association between respiratory symptoms and all-cause mortality for men and women in stratified analyses [6].

As in any survey, there is a risk of non-participation bias; however, it is hard to delineate how this would impact the association between respiratory symptoms and the studied outcome. The response rate in the surveys in this study ranged 67–89%, which is higher than what would probably be achievable today, as survey response rates to epidemiological studies have decreased over the past decades [26]. However, response rates depend on the topic and length of the questionnaire, target population, and whether incentives are used [26-28].

Another potential limitation is that ICD codes are reported by medical doctors, and there will be misclassifications. A validity study of the European short list of respiratory death certificate and autopsy showed a very high agreement for lung cancer, intermediate for COPD but only fairly for pneumonia [29]. This type of misclassification is likely to be non-differential and would bias toward the null [30, 31], and cannot explain our positive results.

Pneumonia is a difficult outcome based on cause of death coding, and there might be different predisposing conditions. It is possible that deaths coded as pneumonia as underlying cause of death could reflect unrevealed chronic lung diseases, or be related to other covariates than those that were available in the pooled sample in these cohorts.

Occupational exposure to air pollution was not a significant predictor of deaths for lung cancer, COPD and pneumonia in the pooled cohorts in this study. On the contrary, we observed a reduced risk for lung cancer mortality as well as COPD mortality in those with occupational exposure to dust and gas. This can be due to the healthy worker effect, i.e. a selection of relatively healthier individuals into occupations with dusty exposures. Paradoxically, in 1985 we found that a positive answer to the same question on occupational exposure was associated with an increased risk of respiratory symptoms [32]. Further explanations for this reduced risk could be information biases with greater reduction in smoking, improved socioeconomic status or less obesity in the follow up of the occupational exposure subset of the population compared with the rest of the population, or possibly better health care or other health-promoting factors for those with occupational exposure.

Smoking might be a confounder of the association between respiratory symptoms and mortality outcomes. Although we controlled for smoking before inclusion in the study, the prevalence of smoking has changed during follow up. In a follow-up of men in a Norwegian work-force, 32% had quit smoking, and only 16% had started smoking over a period of 8 years [33]. In addition, the total proportion of daily smokers in Norway decreased from 42% to 9% during the follow up period [34]. As a considerable proportion of the persons who smoked and reported symptoms might have reduced their symptoms after quitting [35], our analysis probably underestimates rather than overestimates the relation between symptoms and cause-specific pulmonary mortality.

Our study does not elucidate why having respiratory symptoms increases the risk for mortality from lung cancer, COPD and pneumonia, or why self-reports of such symptoms at any time have such long lasting effect. This study was an epidemiological study and not designed to investigate physiological mechanisms, therefore we can only speculate on such mechanisms. Furthermore, we do not have data that can confirm why for instance, dyspnea was a powerful predictor of COPD death, but a weaker predictor of lung cancer and pneumonia death.

It possible that the dyspnea domain captures both deconditioning by pulmonary, cardiovascular disease or other comorbidities, as well as anxiety-related conditions which may be associated with gradually developing COPD and death from COPD, and that this is less relevant for lung cancer or pneumonia, which have more rapid onset. In contrast, cough and phlegm might be more related to smoking habits and it is possible that there may be residual confounding even after adjustment for smoking. In conclusion, scores of respiratory symptoms were associated with long-term deaths due to COPD and lung cancer. Health personnel should be encouraged to ask patients about their symptom burden, as this may be a simple and cost-effective way to map increased mortality risk also among those without known illness. Awareness of even common respiratory symptoms is important in detecting deadly underlying diseases.

(DOCX)

Click here for additional data file.

Questions (Q) on respiratory symptoms and scores.

(PDF)

Click here for additional data file.

Descriptive statistics for participants at different stages according to response to questionnaire.

(PDF)

Click here for additional data file.

Prevalence of baseline symptoms and distribution of symptoms according to principal causes of respiratory death.

(PDF)

Click here for additional data file.

Hazard ratios (HR) with 95% confidence intervals and p-values for lung cancer death according to subgroup, multivariable proportional hazards regression analysis.

(PDF)

Click here for additional data file.

Hazard ratios (HR) with 95% confidence intervals and p-values for COPD death according to subgroup, multivariable proportional hazards regression analysis.

(PDF)

Click here for additional data file.

Hazard ratios (HR) with 95% confidence intervals and p-values for death from pneumonia according to subgroup, multivariable proportional hazards regression analysis.

(PDF)

Click here for additional data file.

12 Jul 2021

PONE-D-21-08054

Respiratory symptoms and respiratory deaths:

a multi-cohort study over 45 years

PLOS ONE

Dear Dr. Stavem,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

At this point we would request you to carefully go through the reviewers' comments and address them satisfactorily. There are several issues raised by the reviewers which you may address through approprite revisions or inclusions in your manuscript.

Please submit your revised manuscript by Aug 26 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Koustubh Panda, M. Tech., Ph.D

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Competing Interests section:

"I have read the journal's policy and the authors of this manuscript have the following competing interests: RN reports grants and personal fees from AstraZeneca, grants from GlaxoSmithKline, grants from Boehringer Ingelheim, outside the submitted work. All other authors have declared that no competing interests exist."

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

Reviewer #4: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: I Don't Know

Reviewer #4: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

Reviewer #4: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

Reviewer #4: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The study by Stavem et al. attempts to establish an association between respiratory symptoms and death from respiratory causes (including pneumonia, lung cancer, COPD, asthma, and other pulmonary diseases) in subjects over 45 years of age. However, the study has several shortcomings that seriously undermine the overall premises on which the conclusions have been drawn or the title has been framed for the reported study.

Major concerns:

The overall study design is not conforming to its title or the conclusions drawn. What is of particular concern is that -

(i) presence of possible or associated comorbidities in the subjects have not been considered which makes it questionable on whether the recorded deaths apparently from ‘respiratory causes’ can be attributed exclusively to respiratory symptoms

(ii) The study is significantly gender biased (84% male vs. 16% female) with a mean baseline age of 46.8 years which is extremely skewed for a cohort representing subjects above the age of 45 years. In fact, the basis for inclusion of 12% subjects in the age group of 15-44 years in a study aiming to examine respiratory symptoms related mortality in subjects above the age of 45 years is not only unclear but this contributes to the above aberrant effect on the age statistics as well as unnecessarily introduces complications in the data analysis because the factors governing respiratory diseases among such young group of subjects and the relatively aged group of 45 plus that the study tends to represent are extremely different.

(iii) Moreover, BMI, vitamin D deficiency, genetic factors, history of allergy and infections etc. can also play a determining role in the development of respiratory symptoms and associated mortality. Unfortunately, there is no consideration or mention of these factors in the study

(iv) The data used in the study have been collected between the year 1972-1999 and all subjects were born between 1902 to1973. There are thus grounds to suppose that the study methodologies and techniques used for assessment of the defined respiratory symptoms have not been consistent over such a long period

(v) The pathophysiological dynamics of various pulmonary diseases vary widely especially when parameters like cough and phlegm are used as ‘respiratory symptoms’. Thus, a more uniform and dependable approach like spirometric assessment of respiratory performance (eg. FEV1 and FVC) would have perhaps been more valuable for a study of this nature.

Minor concerns:

1. There have been several studies that have examined the relation between respiratory symptoms and mortality in specific age groups. The manuscript fails to critically portray these studies in sufficient detail and thereby enlighten how and why the present study is different from similar studies of the past (e.g. Busselton Health Study by Knuiman et.al.)

2. The authors express restrictions on data availability. The scope of such restrictions should have been defined in more detail as availability of all raw data related to studies of the present nature is somewhat essential.

In summary, the present study lacks sound design and critical supportive data to conclusively establish a relation between respiratory symptoms and mortality especially in subjects over the age of 45. Overall, the study fails to contribute any substantive knowledge or information that is new or valuable to our present understanding of the factors underlying respiratory diseases and related mortality.

Reviewer #2: In this cohort study, the authors evaluated the association of respiratory symptoms with mortality due to respiratory causes in Norway by leveraging population based health outcome data (National Cause of Death Registry) and cross sectional symptom surveys in 1972, 1985, 1988, and 1998. Although the authors did not make all data underlying their manuscript fully available, their rationale and a pathway for access to this data to confirm findings is reasonable. I’ve included my comments below

- Infection related pneumonia is not generally considered a chronic respiratory condition as mentioned in the beginning of the introduction. If you are referring to interstitial/histiopathologic pneumonia (such as UIP, EIP, etc)., please clarify this in the introduction.

- What was the temporal association between symptoms and deaths due to pneumonia? Given the observation of increased breathlessness and death due to pneumonia, and the presumption that all deaths did not occur shortly after the survey, is it possible there was another intermediate factor that increased the risk of death due to pneumonia (such as undiagnosed chronic lung disease, bronchiectasis etc)?

- Please address the following limitations: Male biased cohort, limitations associated with survey data regarding selection bias.

Reviewer #3: First, a caveat. On the surface, the methods seem OK, but my expertise is dyspnea mechanisms and measurement, not epidemiologic stats. I am not equipped to make a serious critique of epidemiologic statistical methods in this study.

This analysis utilizes data collected in 4 previous cross sectional studies conducted between 1972 and 1990 in Norway utilizing a Norwegian adaptation of the MRC respiratory questionnaire in 100,000 people. The authors examine the relationship between respiratory symptoms and cause of death. Of the eligible subjects (born between 1902 and 1973), 44,000 had died by the end of 2016 when data collection was terminated. This is a pretty heterogeneous group – subjects would have been 43 to 114 years of age at the study termination, and there would have been a wide variation in the time between survey and death. The main finding of the paper is that subjects who reported respiratory symptoms during a population survey are more likely to die from respiratory disease than subjects who do not report respiratory symptoms. The sample was then subcategorized by respiratory disease (COPD, lung cancer, pneumonia …). The strongest association was between COPD as cause of death and reported breathlessness (increasing breathlessness score associated with increasing hazard ratio - a breathlessness score of 4 was associated with a 7-fold risk of COPD death). These findings are unsurprising to those familiar with the literature.

The information in this paper may be useful to some, but it would benefit from a less narrow view. I am in agreement with the authors’ conclusion that “Health personnel should be encouraged to ask patients about their symptom burden, as this may be a simple and cost-effective way to map increased mortality risk also among those without known illness. Awareness of even common respiratory symptoms is important in detecting deadly underlying diseases.” In fact, I have been supporting this idea for years. There is, however, substantial published information supporting this idea (much of which has not been cited by the authors). I’m not sure the present work, in its current form, adds much. Perhaps the authors could consider enhancing its usefulness.

1) The paper has a narrow focus on a particular epidemiologic question, and lacks broader context. There are several sources for broader information on dyspnea and cough. You could fit this together with your observations – why for instance, would dyspnea be such a powerful predictor of COPD death, but a much weaker predictor of lung cancer and pneumonia death? Why does it predict pneumonia death at all? Pneumonia is an acute disease. These people were not surveyed as part of a medical encounter, so they would be unlikely to have pneumonia symptoms at the time of survey. Is the connection indirect, through COPD? One might also consider the measurement circumstances and timing – what was the gap between survey and death? What was the age at death, etc. Also one might consider the physiologic mechanisms causing cough and breathlessness – how are these related to disease process, and does this inform the interpretation of the link between symptom and cause of death? The questionnaire used (a revision of the MRC) is quite old (consistent with the date of the data collection) – how does that limit the study? Do you think more modern instruments would improve sensitivity or specificity? Why? Or why not?

2) Can you argue the advantages to the clinician or to public health agencies of routinely assessing and documenting respiratory symptoms? It is not at all surprising or new that respiratory symptoms are associated with respiratory deaths. But are there advantages not elucidated in the paper? Do symptoms predict mortality better than other common indicators in your database? For example look at the key paper by Nishimura (Chest 2002) showing that dyspnea is a better predictor of COPD mortality than FEV1. Do symptoms predict mortality before other common indicators? (time from first report of symptoms to death may be within your database, but is not reported). Other important papers relating respiratory symptoms to mortality are missing from your references and discussion (eg, Abidov NEJM 2005, Santos PLOS One 2016).

3) Dyspnea usually indicates some problem in oxygen delivery and utilization, or problems with CO2 elimination – thus it is associated with many cardiovascular and metabolic diseases in addition to those classified as respiratory disease. Any consideration of dyspnea as a predictor of these other diseases is absent from your work, although these causes of deaths are probably in your database.

4) Information on prevalence of the various scores of the symptoms would be useful – both overall in the sample, and subdivided as in tables S3, S4, S5. By this I mean, for instance, how many people reported a score of 3 on dyspnea, etc.

5) A curious detail: Why do you suppose that a dyspnea score of 4 was associated with fewer lung cancer deaths than a score of 3?

Reviewer #4: The research study reported respiratory symptoms and death from respiratory causes. In summary, 14% of the examined population died from respiratory causes. Of these, 41% died from lung cancer, 29% from COPD, 23% from pneumonia, 2% from asthma, 2% from interstitial lung disease and 3% from other pulmonary diseases.

Minor revisions:

1- Line 191: State the proportion of men who died.

2- Title and Abstract: Clarify that 45 years refers to a span of time, not the participant's ages.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

Reviewer #4: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

28 Sep 2021

Rebuttal letter with comments to editor and reviewers has been uploaded as a file.

Submitted filename: Rebuttal_letter__PlosOne27sep2021_ks.docx

Click here for additional data file.

10 Nov 2021

Respiratory symptoms and respiratory deaths:

a multi-cohort study with 45 years observation time

PONE-D-21-08054R1

Dear Dr. Stavem,

We’re pleased to inform you that following a review of your revised submission your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Koustubh Panda, M. Tech., Ph.D

Academic Editor

PLOS ONE

12 Nov 2021

PONE-D-21-08054R1

Respiratory symptoms and respiratory deaths: a multi-cohort study with 45 years observation time

Dear Dr. Stavem:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Professor Koustubh Panda

Academic Editor

PLOS ONE