The COVID-19 pandemic is a public health threat of international concern, intensifying peoples’ psychological risk and vulnerability by strengthening mental health stressors such as fear, panic and uncertainty. The unexpected fear of COVID-19 has been reported to be associated with suicide occurrences, similar to prior pandemics.

Identifying the factors associated with fear of COVID-19 could help us to develop better mental health strategy and practice to improve the situation here in Bangladesh. This was the first attempt to present a Geographic Information System (GIS)-based distribution of fear of COVID-19 across the country's administrative districts in a nationwide sample.

Data for a total of 10 067 individuals were collected by an online survey during the first wave of the pandemic (1 to 10 April 2020); data for 10 052 participants were finally analysed after excluding 15 transgender individuals. The survey questionnaire included items concerning sociodemographic, behavioural and health-related variables, COVID-19-related issues, and the Bangla Fear of COVID-19 Scale.

The mean fear of COVID-19 scores was 21.30 ± 6.01 (out of a possible 35) in the present sample. Female gender, highly educated, non-smoker, non-alcohol consumer, having chronic diseases, using social media, and using social media and not using newspapers as COVID-19 information sources were associated with a higher level of fear of COVID-19. Higher levels of fear of COVID-19 were found in districts of Magura, Panchagarh, Tangail, Sunamganj and Munshiganj; by contrast, Kushtia, Pirojpur, Chapainawabganj, Jhalokathi and Naogaon districts had lower fear of COVID-19. Based on the GIS-distribution, fear of COVID-19 was significantly associated with the district as well as in respect to its gender-based and education-level-based associations. However, fear of COVID-19 and COVID-19 cases were heterogeneously distributed across the districts; that is, no consistent association of higher COVID-19 cases with higher fear of COVID-19 was found.

This study being exploratory in nature may help to facilitate further studies, as well as directing governmental initiatives for reducing fear of COVID-19 in at-risk individuals. Providing adequate resources and mental health services in the administrative regions identified as highly vulnerable to fear of COVID-19 is recommended.

. Self-isolation regime is an effective measure to contain the pandemic (Alfano, Ercolano, 2020), but the psychological factors predicting compliance with stay-at-home recommendations (CSHR) are understudied. We hypothesized that 1) defensive optimism and constructive optimism will have opposite effects on CSHR, 2) the effect of defensive optimism will be mediated through a decrease of anxiety (fear of infection).

. The purpose of this study was to assess the direct and indirect (through the fear of infection) effects of defensive optimism (belief that coronavirus problem is exaggerated) and constructive optimism (belief that people’s efforts help to prevent infection and spread of the virus) on CSHR, controlling for dispositional optimism.

. A longitudinal study (from 10/4/2020 till 2/6/2020) was conducted on a sample of 306 university students (89% women, MA=21.20, SD=4.54) using a single-item measure of CSHR, LOT-R (Scheier et al., 1994), the scales of defensive and constructive optimism (Gordeeva, Sychev, 2020), and anxiety in a pandemic situation questionnaire (Tkhostov, Rasskazova, 2020).

. During seven-week interval CSHR has decreased dramatically (Cohen’s d=0.66, p<0.001) while the other variables remained stable. Using SEM we have showed that CSHR at the end of study (T2) is predicted by the CSHR (T1) and through it by the defensive optimism (negative effect, p<0.05) and constructive optimism (positive effect, p<0.001). Negative effect of defensive optimism on CSHR is also mediated by the fear of infection (T2), reducing it. Dispositional optimism is associated only with constructive optimism.

. Defensive and constructive optimism/ pessimism are essential in explaining health-related behavior.

No significant relationships.