Suicide and suicidal ideation was a significant public health concern before the COVID-19 pandemic, but the pandemic has made it even more urgent.Reference John, Pirkis, Gunnell, Appleby and Morrissey1 O'Connor et al reported that suicidal ideation increased among young people during lockdown (March to April 2020) in the UK.Reference O'Connor, Wetherall, Cleare, McClelland, Melson and Niedzwiedz2 The results identified those aged 18–29 and 30–59 years at higher risk compared with those aged ≥60 years, suggesting that the pandemic is affecting the mental health of the working-age population more than the population aged ≥60 years. Another longitudinal survey conducted in the USA reported that suicidal ideation was more prevalent among employed than unemployed respondents.Reference Czeisler, Lane and Petrosky3 Suicidal ideation among the employed population should be an area of concern.
On 7 April 2020, the Japanese Government declared a state of emergency, which lasted until 25 May 2020. The Japanese Government and local governments did not impose a lockdown, but called on citizens to voluntarily refrain from non-essential and non-urgent outings and avoid the ‘three C's’ (confined spaces, crowded places and close contact). Although Japan's emergency declaration was much less restrictive than the lockdowns in Europe or the USA (comprising request-based measures for close or shortened business hours and stay-at-home requests instead of orders), poor mental health in the general public was reported.Reference Yamamoto, Uchiumi, Suzuki, Yoshimoto and Murillo-Rodriguez4 However, the effects on suicidal ideation and its risk factors, even after the conclusion of the state of emergency, have not been examined yet among the Japanese working population.
We conducted a longitudinal analysis on suicidal ideation by using the Employee Cohort Study in the COVID-19 pandemic in Japan (E-COCO-J),Reference Sasaki, Kuroda, Tsuno and Kawakami5 and investigated its risk factors.
Members of a pre-existing cohort were invited to take part. The cohort consisted of full-time employees recruited from community-dwelling people all over Japan by an internet research company in February 2019, stratified by gender and 10-year age groups (N = 4120). Through an invitation email from the internet research company, we further invited these respondents to participate in a baseline online survey of the E-COCO-J cohort. After completing an online baseline survey during 19–22 March 2020 (time point 1, n = 1448), the respondents, excluding the unemployed (n = 27), were invited to complete a survey during 22–26 May 2020 (time point 2) and 7–12 August 2020 (time point 3). The state of emergency was declared from 7 April to 25 May 2020. The survey time point is shown in Supplementary Fig. 1 available at https://doi.org/10.1192/bjo.2021.1035, along with the daily number of COVID-19 cases in Japan. At the peak of the first wave of the outbreak, Japan reported approximately 700 positive cases in a single day. Time points 1 and 2 are before and after the first wave of the COVID-19 pandemic, respectively, and time point 3 coincides with the peak of the second wave. Supplementary Fig. 2 shows the flow chart of participant recruitment. Suicidal ideation and loneliness in the past 30 days were measured by the items ‘I feel like I want to die’ and ‘I feel lonely’, respectively, at time points 2 and 3. The response option was scored with a four-point Likert scale: 1 (almost never), 2 (sometimes), 3 (often) and 4 (almost always). The variables were categorised into yes (2–4) and no (1). The McNemar test was used to assess differences in proportion of loneliness or suicide ideation for the total sample, and separately for groups classified based on gender, age, education, occupation type and pre-existing mental health conditions, between time points 2 and 3. Potentially relevant factors associated with suicidal ideation at time point 3 were analysed with the multiple logistic regression model (forced entry method), adjusted for gender, age, education, occupation type (measured in 2019), suicidal ideation and loneliness at time point 2, and pre-existing mental health conditions at time point 1 (current or past treatment for depression, anxiety or mood instability). This study was approved by the Research Ethics Committee of the University of Tokyo (approval number 10856-(2)(3)(4)(5)). Online informed consent was obtained from all participants, with full disclosure and explanation of this study's purpose and procedures. We explained that their participation was voluntary, and they could withdraw from the study for any reason simply by not completing the questionnaire.
The analytic sample consisted of currently employed respondents (N = 875) who responded to baseline and two follow-up surveys at time points 2 and 3. Participant characteristics are shown in Table 1. A comparison of respondents in the analytic sample (N = 875) and those who responded to time point 1 but dropped out at time point 2 or 3, or were excluded owing to unemployment (n = 573), showed that the analytic sample was more likely to be older (>40 years), male, employed in a managerial/non-manual occupation, have higher educational attainment and have no pre-existing mental health conditions (Supplementary Table 1).
Suicidal ideation increased between time points 2 and 3 in the total sample (P = 0.008), as did loneliness (P = 0.002) (Table 2). Suicidal ideation increased significantly among women, younger adults (aged <39 years), those with higher educational attainment and those without pre-existing mental health conditions (P = 0.028, P = 0.048, P = 0.003 and P = 0.044, respectively). The adjusted model 1 showed that younger age (adjusted odds ratio 1.54, 95% CI 1.07–2.22, P = 0.021), loneliness at time point 2 (adjusted odds ratio 1.52, 95% CI 1.02–2.26, P = 0.041) and suicidal ideation at time point 2 (adjusted odds ratio 15.17, 95% CI 9.95–23.13, P < 0.001) were significantly associated with suicidal ideation at time point 3 (Table 3). The fully adjusted model 2 showed that younger age (adjusted odds ratio 1.57, 95% CI 1.09–2.28, P = 0.017), pre-existing mental health conditions at time point 1 (adjusted odds ratio 2.17, 95% CI 1.28–3.67, P = 0.004) and suicidal ideation at time point 2 (adjusted odds ratio 15.40, 95% CI 10.06–23.58, P < 0.001) were significantly associated with suicidal ideation at time point 3 (Table 3).
a. High educational attainment was indicated as an undergraduate degree and beyond.
* P < 0.05.
a. Adjusted for gender, age, education, occupation type, and loneliness and suicidal ideation at time point 2.
b. Adjusted for gender, age, education, occupation type, and loneliness and suicidal ideation at time point 2, and pre-existing mental health condition at time point 1.
c. High educational attainment was indicated as an undergraduate degree and beyond.
* P < 0.05.
Suicidal ideation significantly increased between time point 2 (May 2020) and time point 3 (August 2020) in the total analytic sample of Japanese employees (N = 875). Factors significantly associated with suicidal ideation were younger age (aged <39 years), suicidal ideation at time point 2 and pre-existing mental health conditions. Loneliness at time point 2 also showed significant association, without adjusting for pre-existing mental health conditions.
Echoing the study findings by O'Connor et al,Reference O'Connor, Wetherall, Cleare, McClelland, Melson and Niedzwiedz2 the younger population in both Japan and the UK may be at greater risk of suicidal ideation during the pandemic. In fact, since July 2020, the number of suicides in Japan each month has increased among young people and women compared with that of the same month in 2019.6 In more detail, monthly suicide rates in Japan declined by 14% during the first 5 months of the pandemic (February to June 2020), but monthly suicide rates increased by 16% during the second wave (July to October 2020), with a larger increase among women (37%) and children and adolescents (49%).Reference Tanaka and Okamoto7 In contrast, there was no significant association with suicidal ideation in gender among full-time employees in this study, whereas suicidal ideation longitudinally increased among women. The literatureReference Matilla-Santander, Ahonen, Albin, Albin, Baron and Bolíbar8,Reference Ruffolo, Price, Schoultz, Leung, Bonsaksen and Thygesen9 reported that women who were employed part-time or unemployed may be more affected by COVID-19, but they were not part of our sample. Future monitoring may be worthwhile for the female population, who are more likely to be affected by COVID-19.
Pre-existing mental health conditions and pre-existing suicidal ideation at time point 2 increased the risk of suicidal ideation at time point 3. Pre-existing psychiatric disorders are suggested as one of the risk factors for suicide during the COVID-19 pandemic.Reference Raj, Ghosh, Singh, Verma and Arya10 A Chinese cross-sectional study reported that psychiatric patients showed higher suicidal ideation and poorer mental health than healthy controls during early lockdown in the COVID-19 pandemic.Reference Hao, Tan and Jiang11 Employees with pre-existing mental health conditions may be at risk.
Loneliness may be one of the key factors in identifying the risk of, and preventing, suicide.Reference Stickley and Koyanagi12 As almost all groups showed an increase in loneliness between time points 2 and 3 (Table 2), less social interaction (i.e. social disconnectedness) may universally be taken into consideration in the prolonged pandemic.Reference Raj, Ghosh, Singh, Verma and Arya10
This study had several limitations. We did not measure these variables at time point 1 (baseline), so we could not compare these with early stages of the COVID-19 pandemic. This study has limited generalisability because it was a survey of full-time employees, leading to an underestimation of the risk of suicidal ideation and loneliness by excluding people who were unemployed or who had a precarious job. The generalisability may also be limited by the online recruitment procedure and sampling bias resulting from differences in drop-out rates related to gender, age, occupation and pre-existing mental health conditions. The suicidal ideation of this study was more passive than in previous study the previous study by O'Connor et al,Reference O'Connor, Wetherall, Cleare, McClelland, Melson and Niedzwiedz2 which collected data on the frequency of active suicidal ideation, attempts and self-harm attempts (‘How many times in the past week have you thought of taking your life?’). Such a difference may cause overestimation of the frequency of suicidal ideation in this study.
Future studies are needed for practical public health measures to reduce the risk of suicide in the COVID-19 pandemic, investigating both individual risk (e.g. illness) and social risk (e.g. finance, isolation).Reference Gunnell, Appleby, Arensman, Hawton, John and Kapur13
In summary, the results suggest that suicidal ideation of employees has increased from May to August 2020 during the ongoing COVID-19 outbreaks, parallel to the observed increased suicide cases in Japan. Being young, experiencing loneliness, having a pre-existing mental health condition and prior suicidal ideation were suggested to be a risk factor of suicidal ideation. National and community support is needed to target people who are likely to be left behind.Reference United Nations14,Reference Sher15
Supplementary material is available online at https://doi.org/10.1192/bjo.2021.1035.
The data that support the findings of this study are available from the corresponding author, N.K., upon reasonable request.
N.K. was in charge of this study, supervising the process and providing his expert opinion. N.S. and N.K. organised the study design and analysed the data. Collaborators R.K., K.T. and K.I. ensured that questions related to any part of this work's accuracy or integrity were appropriately investigated and resolved. All authors conducted the survey. N.S. and N.K. wrote the first draft of the manuscript, and all other authors critically revised it. All authors approved the final version of the manuscript.
This work was supported by internal funds of the Department of Mental Health, Graduate School of Medicine, The University of Tokyo. The sponsors had no role in the design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review and approval of the manuscript; or the decision to submit the manuscript for publication.
Declaration of interest
N.K. reports receiving grants from Fujitsu Ltd and SBAtWork Corporation; personal fees from the Occupational Health Foundation, Japan Dental Association, Sekisui Chemicals, Junpukai Health Care Center and Osaka Chamber of Commerce and Industry; and non-financial support from Japan Productivity Center, outside the submitted work. All other authors claim no conflicts of interest.