Respiratory Morbidities and Lung Function Abnormalities in Survivors of Bhopal Gas Disaster: A Cross-Sectional Study

Background Respiratory morbidities remained major morbidities among the survivors of the Bhopal gas disaster. We hypothesized that underlying lung function abnormalities, especially small airway dysfunctions (SAD), were responsible for persistent respiratory morbidity. We conducted a cross-sectional study between 2018 and 2020 in the severely exposed cohort of the Bhopal gas disaster. A standardized questionnaire was used to collect respiratory symptoms. The forced oscillation technique (FOT) and spirometry were used for assessing lung functions. Univariate and multivariate logistic regression analyses were used to examine the association. What is the bottom line? This study was the rst to investigate small airway dysfunction in the gas-exposed survivors. We observed a high prevalence of abnormalities in both spirometry and forced oscillation techniques parameters. The small airway dysfunction was independent of the abnormality in spirometry. Respiratory morbidity showed a signicant association with abnormalities in lung function. This study emphasized that spirometry is insucient to detect complex lung function abnormalities in participants exposed to toxic gases. The small airway function must be assessed in such a situation.


Introduction
Bhopal gas disaster is one of the worst chemical disasters in the world. On 3rd December 1984, toxic gases were leaked at Bhopal (capital of state Madhya Pradesh, India) from a nearby pesticide plant. Several thousand individuals were exposed to these toxic gases, and many died. After the disaster, the Municipal Corporation of Bhopal categorized the Bhopal city (wards) into gas-affected and non-affected areas based on the severity of symptoms in the residents of those areas. 1 The gas-affected areas were further strati ed into mild, moderate, and severely affected areas based on the mortality rate between 3rd and 6th December 1984.
In 1985, the Indian Council of Medical Research (ICMR) started the long-term population-based epidemiological study to assess the long-term health effects of toxic gas exposure. Cohorts for that study were assembled by randomly recruiting individuals from both gas-affected and non-affected areas. The strati cation of mild, moderate, and severely exposed cohorts was based on the recruited subjects' residential addresses. Formation of cohorts and morbidity pro le of these cohorts over the last thirty years had been published. [1][2][3] Respiratory morbidities continued to be one of the commonest morbidities in the exposed cohorts. We hypothesized that lung function abnormalities, especially small airway dysfunction (SAD), are responsible for persistent respiratory symptoms.
The aim of our study was (i) to characterize the respiratory symptoms, (ii) to assess lung function by forced oscillation technique (FOT) and spirometry, and (iii) to assess the relationships of respiratory symptoms with lung function abnormalities in survivors of the Bhopal gas disaster.

Study population
This community-based cross-sectional study was conducted from February 2018 to February 2020. The Institutional Ethics Committee of the National Institute for Research in Environmental Health approved the study protocol (NIREH/BPL/1EC-prj.19/2017-18/1458). The initial plan was to recruit from all cohorts, but permission to recruit only from the severely exposed cohort was granted. The cohort population of the ICMR study was variable on each survey since inception due to the high attrition rates. [1][2][3] The number of individuals of the severe cohort available before the initiation of this study was 4170. Therefore, we decided to recruit 4000 severely exposed. The exclusion criteria were those who declined to take part and unavailable on two attempts. An earlier study showed that individuals with age <20 years at the time of the disaster had lesser respiratory morbidity than those with age ≥ 20 years. 3 Therefore, we strati ed our study participants into two age groups: age ≤ 55 years (i.e., age was ≤ 20 years at the time of the disaster) and age more than 55 years (i.e., age > 20 years at the time of the disaster) to verify the previous observation.

Characterization of chronic respiratory symptoms
We developed a questionnaire based on the INSEARCH (Indian Study on Epidemiology of Asthma, respiratory symptoms and Chronic Bronchitis) questionnaire to characterize respiratory symptoms and smoking habits. 4 The death of family members (living with them) within one year from the disaster, whether admitted to a hospital within one week after the exposure, the distance of residence from the pesticide plant at the time of exposure, and treatment received for other co-morbidities (e.g., diabetes, hypertension, coronary artery disease, chronic kidney diseases, etc.) were also collected. A trained technician administered the questionnaire during home visits. The principal investigator re-examined the responses at the time of lung function testing. The complaint of breathlessness was de ned as an a rmative response to at least one of the questions on breathlessness in different situations, i.e., a feeling of morning breathlessness, breathlessness on exertion, breathlessness without exertion, and breathlessness at night. The cough complaint was de ned as an a rmative response to either having cough in the morning or at night. The diagnosis of chronic bronchitis was based on the cough with expectoration for ⩾3 months with an a rmative response to one of the two questions, i.e., usually rst cough in the morning and bring out phlegm from the chest rst time in the morning.
Those who consented to lung function testing were bought to our respiratory clinic.
Forced Oscillatory Technique (FOT) The respiratory system resistance (Rrs) and reactance (Xrs) were measured by the Resmon Pro Full device (Restech Srl, Milan, Italy) at 5, 11, and 19 Hz sinusoidal signals. The equipment was calibrated daily before the procedure with reference impedance. The age in completed years, gender, standing height to the nearest centimeter without shoes, and weight rounded off to the nearest kilogram were recorded. FOT was performed in an upright sitting position, putting the nose clip, supporting the cheeks, and according to the European Respiratory Society's recommendation. 5 Each individual performed at least three FOT, and each test was continued until 15 accepted breaths were recorded. The mean of three acceptable tests was retained for the analysis. The FOT parameters included in this analysis were whole-breath respiratory system resistance at 5 Hz (R5), whole-breath respiratory system reactance at 5 Hz (X5), and frequencydependence of resistance calculated as the difference of whole-breath resistance at 5 and 19 Hz (R [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] ). We published the predicted values with the upper limit of normal (ULN) and lower limit of normal (LLN) of each FOT parameter for healthy Indian adults. 6 Until now, the standards to describe the abnormality of FOT parameters have not been developed. 7 Therefore, we considered R5 and R [5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] were abnormal if the recorded values were higher than the ULN; and X5 was abnormal if the recorded values were lower than the LLN. An abnormal FOT was de ned as abnormal R5 or R 5-19, or X5, or all three parameters were abnormal.

Spirometry:
After FOT, spirometry was performed according to the ATS-ERS recommendation using PowerCube Diffusion+ (GANSHORN Medizin Electronic, Germany). 8 Spirometry was performed in a sitting position and wearing a nose clip. Each participant performed at least three acceptable maximal forced expiratory maneuvers. The highest value of forced vital capacity (FVC) and forced expiratory volume in one second (FEV 1 ) were recorded. The North Indian regression equations were used to estimate the predicted values of spirometry parameters, and the lower limit of normal (LLN) was set at the 5 th percentile. 9 If both FEV 1 /FVC ≥0.70 and FVC ≥ LLN, spirometry was categorized as normal spirometry. If FEV 1 /FVC <0.70, spirometry was categorized as an obstructive pattern according to the GOLD (Global Initiative for Chronic Obstructive Lung Disease) guideline. 10 If FEV 1 /FVC ≥0.70 and FVC < LLN, spirometry was categorized as a restrictive pattern. The severities of obstruction and restriction were based on FEV 1 % of predicted according to the ATS-ERS recommendation, i.e., mild (≥70); moderate (≥60 to <70); moderately severe (≥50 to <60); severe (≥35 to <50); and very severe (<35). 8 If the mid-expiratory ow rate (MMEF) was less than LLN; it was categorized as low MMEF.

ANALYSIS
Categorical variables were summarized in counts and percentages. The proportions were presented with 95% con dence intervals (CI) and were compared by Pearson's chi-square test. The parametric variables were presented as mean± standard deviation, and the Student's t-test was used to compare the two groups. Logistic regression analysis was performed to assess the association between respiratory symptoms and abnormalities in FOT and spirometry. The odds ratios (OR) with 95% CI were used to assess the associations. The signi cance (p-value) was set at the 0.05 level. Statistical analysis was carried out by IBM ® SPSS ® Statistics Version 20.

Demographic characteristics
We could establish direct contact with 1695 severely exposed subjects during the study period. Out of 1695 individuals, 916 (55.7%) participated in this study.
The reasons for the low participation are mentioned in gure 1. The mean age of the study population was 55.2±12.3 years. The majority (98.9%) lived within a one-kilometer radius from the plant during exposure. 20.7% of individuals reported the death of at least one family member within one year. Two hundred twenty participants (24%) mentioned admission to a hospital within one week after the disaster. More than one year of tobacco smoking was reported by 18.4% of individuals, and most were current smokers (77.4%). Except for ve women, all smokers were men. Smokers in the ≤ 55-year age group were fewer than the older age group (13.9% vs. 22.7%, p<0.01).
The predominant comorbidity was systemic hypertension (37.0%; 95% CI: 33.9-40.2), followed by diabetes (14.4%; 95% CI: 12.3 -16.8). Out of 916, 558 participants (60.9%) took part in lung function testing. The demographic characteristics, co-morbidities, and respiratory symptoms of those who took part in lung function testing (n=558) and those who did not (n=358) are compared in Table 1. The demographic characteristics of both groups were comparable.
Those who participated in lung function assessment had fewer coronary artery diseases than those who did not (4.1% vs. 8.4%; p <0.01).

Forced oscillation technique
Acceptable FOT was performed by 471 of 558 individuals (84.4%) who participated in lung function assessments. The distribution of R5 and X5 of the study population is presented as the percentage predicted in gure 2. The R5 and X5 ≥150 % predicted was observed in 24.2% and 28.9% participants, respectively. The R5 >ULN was in 29.3%, R 5-19 >LLN was in 23.3%, and X5 <LLN was in 21.2% of participants (Table 3). We observe abnormal FOT in 37.7% of participants.
The distribution of abnormal FOT was not different between men and women (36.1% in men vs. 39.3% in women, P=0.27).
Three hundred eighty-six participants (69.2%) performed both acceptable FOT and spirometry. Table 5 illustrates the relationship between FOT parameters with spirometry outcomes of these individuals. The R5 > ULN, R 5-19 >ULN , and X5 <LLN were observed in 17.7%, 15.1%, and 8.1% of participants with normal spirometry. Individuals with obstructive spirometry had a higher risk of abnormal FOT parameters (adj OR: 3.93, 95% CI: 2.24-6.89). Both obstructive and restrictive spirometry had the highest risk of abnormal X5 among all FOT parameters.
Relationship of respiratory symptoms with lung function parameters

Discussion
In this study, we characterized the respiratory symptoms in survivors of the Bhopal gas disaster and assessed their lung functions by both FOT and spirometry. The majority complained of breathlessness. Abnormalities in spirometry were observed in nearly half of those who performed acceptable spirometry. The abnormalities in FOT, suggestive of SAD, were independent of abnormalities in spirometry.
After any toxic inhalational disaster, population-based studies are vital in understanding the pathophysiological mechanisms of lung injury, long-term sequel, and developing appropriate medical management. Besides the Bhopal gas disaster, many individuals were exposed to the dust of building materials during the World Trade Center (WTC) attack. The gases of the Bhopal gas disaster were lethal compared to the dust of the WTC attack. Extensive research and long-term follow-up of the exposed population in the WTC attack helped us understand lung injury pathogenesis following toxic inhalation. However, long-term follow-up lung function in the Bhopal gas disaster survivors is lacking.
Aftermath any disaster, over or under-reporting of symptoms in the exposed population is anticipated due to various medical and non-medical causes. Most of our study population mentioned having at least one respiratory symptom. The morbidity data of the ongoing long-term epidemiological study of the ICMR are collected either from the head of the family or any adult family member available during the surveys. In contrast, the morbidity data of the current study were collected through face-to-face interaction. This might be a possible reason for observing higher respiratory morbidity than previous reports of long-term epidemiological study. [1][2][3] Lung function assessment is essential to evaluate the adverse effects of toxic inhalations. Immediately after the disaster and subsequently, several studies were conducted to estimate the lung function abnormalities in survivors of the Bhopal gas disaster. Spirometry was mostly used to assess lung functions. The type, severity, and distribution of abnormalities in spirometry were widely variable across those studies. 11 We also observed new abnormalities in spirometry and rapid decline in spirometry parameters in survivors with respiratory symptoms. 12 The last publication on lung function of the Bhopal gas disaster survivors was a retrospective analysis of spirometry reports of those who consulted a particular hospital for their respiratory problems. 12 That study reported an obstructive pattern as the predominant abnormality (50.8%), followed by a restrictive pattern (13.3%). The current study being a community-based might be the reason for the different distribution of abnormalities in spirometry. Cullinan et al. were the rst to report reduced MMEF, a subtle spirometry marker of SAD in the Bhopal Gas Disaster exposed population. 13 They observed that those were living near the pesticide plant had signi cantly reduced MMEF. However, they had not reported the prevalence of MMEF abnormalities. Nearly one-third of our population had MMEF < LLN. The individuals exposed to the WTC attack had persistent respiratory morbidity and lung function abnormalities over decades. 14,15 Therefore, lung function abnormalities in the exposed population of the Bhopal gas disaster, even over thirty years after the exposure, were not surprising. We observed that the lung function abnormalities were less in those exposed in their early life than older ones. This was possibly due to the wane of exposure effects as the children grown.
FEV 1 predominantly re ects the obstruction of medium to large-sized airways. A signi cant amount of small airway resistance must be built up to make FEV 1 abnormal. 16 Spirometry fails to detect complex heterogeneous pathology of airways, especially if subtle abnormalities are present in the small airways. FOT and Impulse Oscillometry (IOS) are two noninvasive techniques used to assess small airway function. Both these techniques measure respiratory system resistance (Rrs) and reactance (Xrs), the two components of respiratory impedance. Rrs represent impedance to air ow, primarily re ecting overall airway caliber. Whereas, Xrs represents impedance to volume changes and encompasses both the respiratory system's inertial and elastic properties. The evaluation of small airway function in participants exposed to WTC attack provided valuable information that was rarely apparent in the spirometry. 14, 15 Oppenheimer et al. highlighted IOS as a better way of identifying SAD in subjects exposed to the WTC attack. 17 The association of small airway dysfunctions in the exposed population of the WTC attack with persistent respiratory symptoms is well established. 14,19 Breathlessness in our study population had a signi cant association with the abnormalities of both FOT and spirometry. Like the WTC attack exposed population, we also found abnormalities in IOS parameters independent of spirometry abnormalities. Sriramachari histopathologically documented the involvement of small airways in lung autopsies of the exposed population of the Bhopal gas disaster. 18 He showed necrotizing bronchiolitis with denudation of the epithelium, in addition to severe tracheitis and bronchitis.
Our study is the rst to investigate and document SAD in survivors of the Bhopal gas disaster.
The restrictive spirometry of the WTC disaster was attributed to the involvement of small airways. 20 SAD closes small airways at tidal volume, leading to restrictive spirometry. We also observed the presence of SAD in participants with restrictive spirometry. Few individuals with restrictive spirometry in our study also undergone radiological evaluations; however, no evidence of brosis or scarring was found in their chest X-rays.
The major limitations of this study were cross-sectional design, low participation, recruitment was restricted to only severely exposed cohort, no comparator group, and nearly half of the recruited individuals participated in lung function assessment. Therefore, the observation of this study may not be generalized to the survivors of the Bhopal gas disaster. The causes of low participation were busy with livelihood, lack of interest, no nancial bene ts for participation, change of residence, local migration, etc. We were also unable to establish contact with the individuals who are usually at work during our study visits, and they were probably healthier. There was a potential risk of selection bias, as those with relatively fewer respiratory symptoms participated in lung function assessments. This study was conducted over thirty years after the exposure, and some participants were also exposed to occupational and ambient air pollution in the succeeding years. In the absence of longitudinal lung function assessments, we cannot attribute the observed lung function abnormalities to exposure only. Post-bronchodilator spirometry and FOT were not performed due to logistics issues. Lung volume estimation of participants with restrictive spirometry was also not performed to con rm reduced total lung capacity. The strength of this study was that participants were part of the original cohort of the long-term epidemiological study. Therefore, their exposure to the disaster was undoubted.
In conclusion, this study highlighted self-reported respiratory morbidity and abnormalities in lung functions. Those who were exposed in early childhood had lesser lung function abnormalities compared to older ones. Lung function abnormalities are complex, and small airway dysfunction was independent of spirometry. A comprehensive lung function assessment by both spirometry and FOT is indispensable in assessing the effects of inhalational exposures.

Declarations
Funding This project was supported by an Intramural Research Grant from the Indian Council of Medical Research, New Delhi. There was no role of the funding agency in the design, data collection, analysis, interpretation of data, and manuscript writing.
Grant/award number Not applicable

Con ict of Interest
The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this paper.
Running Title: Respiratory Morbidities and Lung Function abnormalities after the Bhopal Gas Disaster Contribution ship SD: Conceived the study, interpretation of data, and manuscript writing. NB: Data collection. YS: data interpretation. All authors read and approved the nal manuscript.