Occupation- and industry-specific cancer mortality among Japanese women from 1980 to 2015

Many Japanese women quit their jobs for childrearing. However, with the initiatives of the government focusing on flexible work environments and provision of paternal leave, the proportion of women continuing their employment is increasing. As Japan’s population rapidly ages the occupational and industrial structure, including the work environment, is changing. As one of the world’s largest industrial economies, Japan has many industries with a high risk of occupational exposure to environmental carcinogens, which is an important avoidable cause of cancer mortality [1]. Approximately 115,000 fatal and non-fatal accidents directly attributable to industry occur yearly in Japan, with one-third of the fatal accidents occurring in the construction sector [2]. According to the Global Burden of Disease Report 2019, occupational cancer causes 538,000 deaths annually [3]. Lung cancer, mesothelioma, leukaemia, ovarian and laryngeal cancer are some of the most common cancers related to occupational exposures among Japanese women [3]. Mortality disparities associated with environmental factors in different occupations and industries pose a significant occupational health challenge. Past studies have shown that occupation-specific mortality in Japan does not follow a clear socioeconomic gradient like its western counterparts [4] and white-collar workers such as professors and managers are known to have higher gastric, lung and cancer mortality rates than blue-collar workers [5‐8].

The financial crisis in the late 1990s has led to an increase in this mortality inequality [5, 9]. The prevalence of cancer mortality in Japan has been rising in recent years. Smoking cigarettes, a family history of cancer, environmental factors such as dust, chemical exposure, and diet impose risk for lung cancer [10, 11] while age, sex, radiation, socio-economic status and alcohol consumption is associated with gastric cancer [12‐14]. Modifiable risk factors like obesity, diabetes, smoking and alcohol consumption and exposure to pesticides, radiation and sedentary employment is associated with pancreatic cancer [15]. Alcohol consumption, smoking, and a variety of lifestyle and dietary factors increase the risk of colorectal cancer death, including lack of exercise, obesity, eating red or processed meat, and overconsumption of refined carbohydrates [16, 17]. There were 150,838 cancer deaths among Japanese women in 2015 [16], 2–5% of which were attributable to occupational exposure to carcinogens [18‐21]. Lung, gastric, pancreatic and colorectal cancer were among the leading causes of cancer mortality among Japanese women aged 25–64 years and constitute approximately 50% of all cancer deaths. If detected at stage I, the 5-year relative survival rate of gastric and colorectal cancer in Japan is more than 95% followed by lung cancer at 81%, while that of pancreatic cancer is below 43% [16]. However, this rate decreases to 5%, 9%, 1.7% and 19% for lung, gastric, pancreatic and colorectal cancer, respectively if detected at stage IV. These figures make it even more important to detect these cancers at an early stage. Workers from specific occupations and industries have an elevated risk of cancer morbidity and mortality [22]. The Japanese government has set a priority to increase the cancer screening rates to greater than 50% [23, 24]. Nevertheless, the cancer screening rate in Japan remains the lowest among its OECD counterparts [25].

Around 70% of the female Japanese workforce is aged 25–64 years, and this proportion is increasing, with women also increasingly entering once male-dominated sectors in industry and transportation [26, 27]. These women are exposed to potentially hazardous carcinogens in the workplace. It is essential to analyse the risk among women due to the gender specific differences in workplace exposure, responses, carcinogenic potency and susceptibility. It is generally possible to characterize women’s risks based on the results of studies conducted on men, but this is not always the case [28], and past studies of occupational risk in Japan have often focused only on men. Recent studies have focused on occupation as a means of identifying women at high risk as a proxy measurement of health inequality [29]. However, most studies to date have explored leading causes of cancer by occupation and industry among Japanese men only [7]. A study by Okui explored trends in cancer mortality among manual and non-manual female occupational workers [8]. However, there is still insufficient evidence concerning the relative risk among Japanese women of developing lung, pancreatic, gastric and colorectal cancer in relation to occupational and industrial classification. This study will help in highlighting the association between socio-economic differences and cancer mortality. This study aims to examine the difference in mortality risk for key cancers among women and explore the effect of the mid-1990s economic crisis separately for occupational and industrial categories.

A repeated cross-sectional study from 1980 to 2015 was conducted with five-yearly mortality data from the Japanese Population Census and the National Vital Statistics conducted in the same year. The government records the occupation and industry at the time of death, as reported by a family member of the deceased, and records the underlying cause of death from death certificates. We examined the four leading cancers – lung, gastric, pancreatic and colorectal cancer – in Japanese women aged between 25 and 64 years. For 1980, 1985 and 1990 we used the ninth revision of the International Classification of Disease, and the tenth revision for 1995, 2000, 2010 and 2015. Occupations were classified as: managers (administrative and managerial); professional; clerk; sales; service; security; agriculture; and manufacturing. The industry was classified as: agriculture and forestry; fishing; mining (mining and quarrying of stone and gravel); construction; manufacturing; electricity (electricity/gas/heat supply/water industry); transportation/communication (transport and postal activities/information and communications); wholesale/retail (wholesale and retail trade); finance (finance and insurance); real estate (real estate and goods rental and leasing business); service; and public services. Occupations and industries were classified according to the International Standard Occupational and Industrial Classification [30‐33].

The Japanese dataset provided aggregated data by age category and does not provide individual-level data. Thus, we could not adjust for additional risk factors for cancer such as smoking and personal or family history of cancer. Many women may change their job after their cancer diagnosis, especially non-office workers. Thus, the occupation obtained from the certificate that provides the occupation or industry at the time of death may not reflect the nature of employment at the time of cancer diagnosis, which may affect the results. Additionally, the occupation and industry classification of the deceased was reported by the family of the deceased, which could have led to misclassification. The use of broad occupational and industrial categories in the current analysis might pose limitations in interpreting the result. Analysing the data with specific occupational and industrial categories would be helpful in better understanding the cancer mortality situation. The occupational and industrial classification has been revised over the years and may affect the mortality trend. Similarly, the transition of disease coding from ICD-9 to ICD-10 may also have affected the mortality trends. As women become an increasingly large proportion of the workforce and move into higher-risk manufacturing and industrial sectors, it is important to consider the economic cost of workplace-related mortality among women as well as men. This was not possible with the available data in the present study but should be considered in future research.