Study Design
This study used data from the Japan COVID-19 and Society Internet Survey (JACSIS). JACSIS is a nation-wide online cohort study in Japan, conducted since August 2020 to the present. Participants were community dwelling people aged 16–79 years. The baseline sample of JACSIS was collected in 2020 (n = 28,000). A follow-up survey was conducted in 2022 for the participants in 2020 or 2021, and new participants were invited to join the 2022 survey, creating a total of N = 32,000 people participated the 2022 survey. In this study, a cross-sectional design was applied using the JACSIS 2022 data.
Recruitment
The study utilized email messages to request survey participation from a research panel consisting of individuals aged 15 to 79 years who were registered with Rakuten Insight, Inc. The company’s registered panel consisted of over 2.2 million individuals with diverse sociodemographic backgrounds, representing the national population across all 47 prefectures of Japan. Potential participants were selected using a simple random sampling method based on sex, age, and prefecture category in accordance with the official Japanese demographic composition as of October 1, 2019. Those who agreed to participate in the survey were given access to a designated website. Participants were allowed to skip any questions or discontinue the survey at any point. Priority was given to participants who had completed the survey in 2020 or 2021 during the 2022 data collection, but new participants were also invited until the final sample size was reached (N = 32,000).
Management of data quality
To ensure the validity of the data, respondents who exhibited discrepancies or provided artificial/unnatural responses were removed from the study. Specifically, three question items were used to identify such responses: “Please choose the second from the bottom,” “choosing positive in all of a set of questions for using drugs,” and “choosing positive in all of a set of questions for having chronic diseases.” Respondents who were found to have provided such responses (n = 3320) were excluded from the study.
Participants
The study sample included community dwelling people in Japan over 18 years old and without missing data. Following these criteria, participants aged under 18 (n = 13) and participants with missing data of area deprivation index (n = 700) were excluded.
Measurement variables
Adverse Childhood Experiences (ACEs)
The Adverse Childhood Experiences Japanese version (ACE-J) is a questionnaire that assesses individuals’ exposure to ACEs in Japan (31). The items were limited to one for each category of adversities, except parental loss. Parental loss was assessed as either parental death and divorce or separation. Following the expanded concept of ACEs, childhood poverty, bullying, hospitalization due to chronic disease, and exposure to life-threatening natural disasters were included. The ACE-J consisted of 15 items in total. Participants were asked whether they had following experiences under 18 years old. The response options were “Yes” or “No.” One item related to emotional neglect was reversed question; “I felt I was loved by my parents.” After reversing the score of this item, the summed score of the ACEs were used as the number of ACEs experienced.
Living area characteristics
The Area Deprivation Index (ADI) and Densely Inhabited District (DID) were used to assess living area characteristics. Both indices were created at a zip code level in Japan, with a total of 113,107 zip codes included in the analysis. Each zip code included approximately 1,100 residents.
Area Deprived Index (ADI). In this study, the Area Deprivation Index (ADI) was utilized as a measure of neighborhood deprivation. The data used to construct the ADI was obtained from the 2010 Population Census of Japan, and the specific methodology used to calculate the ADI has been previously described (17). The ADI is a composite indicator that considers several poverty-related census variables, including the unemployment rate, proportion of elderly couple households, elderly single-occupier households, single mother households, rented houses, sales and service workers, agricultural workers, and blue-collar workers. A higher ADI score is indicative of greater neighborhood deprivation. In this study, the ADI was divided into four categories by using quartile to facilitate analysis.
Densely Inhabited District (DID). The urbanization level of the study areas was measured using the Densely Inhabited District (DID) data obtained from the 2015 Population Census of Japan. The DID score is an indicator of the level of population density in a given area, with a higher score indicating a greater level of urbanization. To facilitate analysis, the DID score was dichotomized into high and low categories. The data source for this information is publicly available (37). DIDs that lack a zip code centroid were deemed non-DID, indicating a rural area, and were allocated to the urbanization level with the lowest category. DIDs that possess a zip code centroid, on the other hand, were classified as DID, denoting an urbanized area, and were arranged into tertiles based on population density. These tertiles correspond to the second, third, and highest categories of urbanization level.
Sociodemographic characteristics
Age, sex, educational attainment (less than high school, vocational/college, undergraduate, graduate over), and marital status (married, single/divorced), household income (< 3, 3–5, 5–8, 8–10, over 10 million yen, no response/unknown), and working status (paid work, no paid work, students) were measured as demographic variables.
Statistical analysis
The associations of ACEs with ADI and DID were examined using one-way ANOVA for the total number of ACEs and the chi-square test for the prevalence of ACE 4+. Considering neighborhood deprivation as a reflection of people’s disadvantaged sociodemographic status, the crude model was considered the primary analysis. To explore the effects of individual-level factors, the authors employed multivariable logistic regression to analyze the associations between ADI and ACE 4 + in Japan, controlling for age, sex, marital status, and education, as an additional analysis. As no relevant area-based information was available, multi-level analysis was not conducted in the current study.