We conducted a population-based nested case-control study within the long-term resident cohort in Amagasaki City. The long-term resident cohort, who had been continuously living in Amagasaki City before 1975 and at least until the beginning of 2002, were followed from 1 January 2002 to 31 December 2015 for death or outmigration status using the Basic Resident Records in our previous study[13]. We observed 143,929 residents aged 40 years or more at baseline. The cause of death, mesothelioma, was identified as code C45 according to the International Classification of Disease, 10th Revision (ICD-10), using the National Vital Statistics. The Amagasaki City government provided a data set of malignant mesothelioma deaths that included both pleural and peritoneal mesothelioma between 2002 and 2015 with completed interviews regarding occupational and non-occupational histories related to asbestos exposures, which was obtained from the surveillance based on support by the Ministry of Environment.
The cumulated number of mesothelioma deaths was 379, including 13 peritoneal and 366 pleural mesothelioma deaths between 2002 and 2015 in the entire city, which were target subjects for the interview survey by the Amagasaki City government (Fig. 1). Of these, 110 death cases were excluded because no bereaved families existed or they had moved out of the city. The remaining 269 death cases were contacted for interview and 102 refused or did not respond, resulting in 167 death cases with completed interviews. Of these, 133 death cases met criteria of the Amagasaki City long-term resident cohort and were used as the nested cases. Case interviews of the bereaved families, with informed consent, were conducted by well-trained nurse interviewers employed by the government between 2002 and 2015.
Control candidates were randomly selected from the long-term resident cohort populations matched by sex and age (birth date ± 12 months) and alive on the date of death of corresponding case. Furthermore, three or more matched controls with completed interviews were required per case. To obtain three matched controls per case, 40 candidates per case were randomly extracted from the Amagasaki City long-term resident cohort. Of 5,320 candidates, 1,129 candidates were excluded because they had no family members or had moved out of the cohort. The municipal government sent requests to 4,191 subjects with identified addresses in Amagasaki City to obtain consent for their cooperation. After excluding 1,471 subjects who did not respond, 1,575 subjects who refused, and 70 subjects who responded as under consideration, we sent the individual questionnaires and consent forms for interview in a closed document to 1,075 subjects who had given their permission to be contacted. The interviews were then conducted in the order of consent, by home visit or telephone, and were terminated when at least three controls per case had been secured. Therefore, excluding 54 subjects for whom interviews were canceled before sending questionnaires, we sent a total of 1,021 questionnaires and obtained 403 (39.5%) completed interviews with consents (Fig. 1).
A total of 133 cases and 403 matched controls in the Amagasaki City long-term resident cohort were confirmed as eligible data sets (Fig. 1). The number of case-control pairs was as follows: 9 sets for 1:1, 22 sets for 1:2, 74 sets for 1:3, 19 sets for 1:4, 5 sets for 1:5, 1 set for 1:6, and 3 sets for 1:7. Since no controls were still living for some cases especially elderly cases, there were a few sets in which the number of controls was less than three.
The interviews for cases were conducted with bereaved families between 2002 and 2015. Since most of those families were going through the process of compensation for mesothelioma related asbestos exposures due to the AC plant, they had sufficient information regarding their cases. On the other hand, most families of the controls did not have as much detailed information as did the bereaved families of mesothelioma death cases. For these reasons, the interviews for controls were conducted with the subjects themselves and their families, using the same questionnaire as for the cases, and by well-trained nurses with expertise in asbestos-related health hazards, who had been originally employed by Amagasaki City government for municipal surveillance and subsequently employed by the study group between 2015 and 2017.
The questionnaire consisted of more than 40 items, similar to those used for the Ministry of Environment’s survey regarding mesothelioma deaths due to environmental asbestos exposure. The questionnaire, which was performed by either a structured visit or a telephone interview, consisted of the following sections on the relationship of respondent with the index subject (cases and controls themselves): demographic characteristics including smoking history; status of occupational compensation certification; lifetime occupational history (date of entry/exit, employer’s name, address, occupational category, job description, and handling of asbestos products); residential history (date of entry/exit, and address) during the period 1957–1975; crocidolite, which is considered to be the most toxic type of asbestos and is associated with the highest mesothelioma risk [4], was definitely allowed to be used in the plant [11, 15, 16]; and other possible environmental asbestos exposures such as domestic (cohabitants who washed clothes at home or brought work materials home, handled asbestos products, or worked at places with walls or ceilings sprayed with asbestos) or household (any exposure to asbestos-containing materials in the home) exposure. Environmental (non-occupational) exposure is generally divided into three sources according to the exposure pathway: neighborhood, domestic, or household [3]. In our study, neighborhood exposure that resulted from living near an emission point of airborne asbestos fibers was estimated from individual residential histories. Domestic exposure was estimated in relation to cohabitants occupationally exposed to asbestos in the same house, and household exposure was estimated as it related to any exposure to asbestos-containing materials in the home, such as fire-proof sheets and asbestos-sprayed walls and ceilings. Because individuals were considered to have multiple exposure sources, we intended to evaluate four types of exposure independently, according to relevant rating procedures: occupational, domestic, household, and neighborhood.
Collected occupational histories included “company name”, “address”, “employment period”, “job description”, and “the experience of direct asbestos handling”, which was obtained as dichotomous data (Yes or No), using a check list of asbestos-related products, and if any of these applied to the subject, we determined the subject to be a direct handler. Basically, interviews were supposed to be in an open-question manner, except when a dichotomous answer was required. The following questions related to occupational exposures were required answers as dichotomous data: direct handling, and working at a place with sprayed asbestos walls or ceilings.
In order to control confounding effects due to occupational exposure, when evaluating the effects of non-occupational exposure, we excluded “Definite” exposure (subjects who responded “Yes” to “the experience of direct asbestos handling through their lifetime”). Since evaluating the effects of occupational exposure was not the aim of this study, we did not try to quantitatively evaluate the cumulative dose of asbestos exposure for these subjects. As a result, we excluded 32 cases and 61 controls, for a remaining total of 101 cases and 342 controls. Occupational exposure was then categorized into three tiers: “Convincing”, “Possible”, and “None”, which were evaluated for probability of occupational exposure based on the individual’s occupational history obtained from interviews. Three raters, who had been judges for asbestos-related occupational compensations, assigned each obtained case and control to one of three occupational exposures probability categories, referring to publicly available materials provided by the Ministry of Health, Labor and Welfare, and to a list of industries that included 176 or more work places, in which many employees were certified as occupational compensations in Amagasaki City. To gain expert consensus, we took the following steps: 1) each expert assigned the probability of occupational exposure in two rounds, 2) a summary including demographics was provided to all the experts with the reasons of judgement, 3) the experts were encouraged to revise their earlier assignments in light of the other members’ comments, and 4) the final assignments were determined by a panel including three raters and all authors of the study.
We defined domestic exposure as exposure to asbestos fibers brought home by workers (who handled asbestos products or worked at places with walls or ceilings sprayed with asbestos) on their clothing or in their hair, or through living in the same house with occupationally-exposed individuals (who washed their clothes at home or brought home work materials), and coded it as either Yes or None in the dichotomous category.
We defined household exposure as exposure to asbestos-containing materials used in home structures (e.g., roofs, walls, insulation) or home improvement products, exposure to asbestos-sprayed walls at school, or exposure by playing at an asbestos factory, playing in the backyard of an asbestos factory, or playing with asbestos products. If any of the list of asbestos products applied to the subject, we coded it as Yes in the dichotomous category (None/Yes).
To evaluate for neighborhood asbestos exposure due to a large-scale AC plant, we used cumulative indices corresponding to the individuals’ residential histories during an exposure period from 1957 to 1975, which was established by consensus among experts as the period when a large amount of crocidolite, called “blue asbestos”, was used in the city and was considered the most toxic type of asbestos in relation to malignant mesothelioma. Based on this background, municipal surveillance, called the Epidemiological Analysis Survey of Asbestos Exposure, was conducted with the environmental exposure period set to 1957–1975. Focusing on airborne asbestos fibers derived from the AC plant, the relative concentration (unit, 1/m3) in each 100 m x 100 m grid was estimated, using a diffusion equation that considered meteorological conditions. The method assumed that an emission point of airborne asbestos was at the center of the premises of the plant and followed the diffusion equations as previously described in detail [11, 16]. Figure 2 depicts the distribution of geographically simulated relative asbestos concentrations across Amagasaki City, which were downscaled into a 10 m x 10 m grid by using spline interpolation. The contour lines show isolines of relative asbestos concentration, ranging from 1 to 107 (unit, 1/m3), and the color-coded concentric circles indicate crow-fly distances from the center of the plant premises.
By using residential histories based on the interviews and the Basic Resident Records in Amagasaki City, we obtained the geographic coordinate for each residential record. Geocoding with the Mapple Address Matching Tool (Shobunsya Inc.) was performed to convert the address of each residence into a geographic coordinate, a longitude and latitude pair, which represented a point on the city block called “Gaiku” (61.3% for cases and 79.9% for controls) or in the neighborhood called “Choh-aza” (38.7% for cases and 20.1% for controls). When the geocoding of old address names failed, we referred to detailed city maps to manually convert the old address into the current one. Based on the geographic coordinate, each residence was linked to the simulated relative asbestos concentration, as mentioned above.
We obtained residence-specific asbestos exposure calculated by the simulated relative concentration of airborne asbestos fibers multiplied by the duration at each residence (Table 3). Finally, we calculated the cumulative indices of neighborhood exposures by summing up the residence-specific exposures (unit, year/m3) for each individual subject, taking into account the residential histories between 1957 and 1975 in Amagasaki City.
Conditional logistic regression analysis was employed to compute odds ratios (OR) and 95% confidential intervals (95% CI) in relationship to the neighborhood asbestos exposure (quintiles of the cumulative dose of neighborhood asbestos exposure in all controls), using STATA software (version 14.2/MP; Stata Corp., College Station, TX, USA). Additionally, the other covariates for adjustment were assessed for three tiers of occupational exposure and dichotomous variables for both domestic and household exposures. We defined tests with P < 0.05 as statistically significant.