During the first two years following the onset of the COVID-19 pandemic, the incidence of suicide declined state-wide in California, while firearm suicide rates declined much more modestly. At the same time, the incidence of nonfatal self-harm presenting in UC hospitals increased. Differential variation by sociodemographic groups and geographic areas underlay these trends, suggesting differential exposure to or impact of pandemic-era risk and protective factors, and a need for tailored allocation of state resources and prevention efforts.
The overall decline in state-wide suicide rates during and following the onset of the pandemic parallels similar findings from other states,41–43 and the slight decline in firearm suicide rates aligns with recent CDC data released indicating firearm suicide rates remained level between 2019 and 2020.44 In California, the overall decline was driven by meaningful reductions in suicide among the groups most burdened by suicide – male, middle-aged, and white Californians. In contrast, female, young, Black, and Hispanic Californians experienced increases in suicide or firearm suicide.
As in prior research, we found that males in California consistently had a higher risk of suicide and females consistently had a higher risk of nonfatal self-harm.45 However, females experienced a slight increase in firearm suicide in 2020 compared to pre-pandemic, while males experienced a decrease. These results may reflect the fact that the recent firearm purchasing surge led to uniquely high firearm ownership among groups historically less likely to own firearms (e.g., women) and may indicate a potential shift toward more lethal methods among this group.24
Young people (ages 10–19) in California experienced an increase in fatal and nonfatal self-harm overall in 2020 and 2021 compared to years prior, a trend mirroring national findings.44 This is especially concerning given that suicide is the second leading cause of death for young people in California and nationally.46, 47 The magnitude of the increase among young people was not shared by any other age groups, most of whom experienced a decrease in suicide and self-harm. Preventative efforts, including lethal means safety and mental health supports, should be prioritized for adolescents and young adults—who were uniquely impacted by recent social isolation, uncertainty, stress, and fear—given their stage of life and the importance of socialization for healthy development.48, 49
White Californians experienced substantial declines in suicide, firearm suicide, and nonfatal self-harm during the pandemic. Given the size of the white population and the magnitude of suicide burden among this group, this decrease drove the overall decline observed in the aggregated data. By disaggregating the data, we discovered unique trends across distinct communities. For instance, unlike all other racial/ethnic groups, Black and Hispanic Californians experienced the largest relative increase in suicide/firearm suicide and non-fatal self-harm, respectively, following the onset of the pandemic. These findings are consistent with studies in Maryland and Connecticut documenting an increase in suicide mortality among Black residents and a decrease among white residents in the months following the onset of the pandemic compared to earlier time periods,43−42 and with national, pre-pandemic trends showing a greater increase in suicidal behavior among Black Americans, particularly youth, compared to white Americans, from 1991 to 2019.50
It is likely the racial/ethnic disparities we identified are related, in part, to the pandemic-driven amplification of the structural inequities that shape population health in the U.S.51 and the attrition of culturally-specific factors protective of suicide. The communities most burdened by the health, economic, and social crises of 2020 and 2021 already faced disproportionate threats to their health as a result of systemic racism8 and other systems of marginalization that concentrate greater risk factors associated with suicide (e.g., poverty, unemployment, and mass incarceration34, 35) and fewer protective factors (e.g., quality education, economic development, and culturally competent mental healthcare36–38). Further, Black and Latino Americans, who attend church at higher rates than white Americans,53 may have been disproportionately impacted by the restricted ability to gather for religious worship; and religiosity has been linked to reduction in suicide risk.32, 52 In addition, COVID-19 increased economic and labor market disparities along racial lines7, which have been connected to increased risk of suicide.54 Finally, perceived racial discrimination, which increased during the pandemic,52, 53 along with disparities in death from COVID-19 and police killings,7,55 has also been connected to suicide risk among racially/ethnically minoritized groups.56, 57
Another factor potentially contributing to the increase in suicide, particularly firearm suicide, among some groups may be the firearm purchasing surge of 2020 and 2021. There is an established connection between firearm access and risk of firearm suicide,25, 58–60 and surges in firearm purchasing, which California experienced at the onset of the pandemic, are associated with increases in firearm violence.61, 62 Further, a national study found that pandemic-era firearm purchasers were more likely to experience suicidality than non-owners and pre-pandemic purchasers.63 While we did not observe an increase in number of firearm suicides following the onset of the pandemic, the increase in proportion of suicides that involve a firearm could indicate a trend toward an increasing use of firearms for self-harm, even amidst an overall decreasing trend in death by suicide. Alternatively, the fact that nonfatal self-harm increased while overall suicide decreased may point to greater use of less-lethal (non-firearm) means of suicide, which could explain the increase in proportion of suicides that involve a firearm, even if firearm use for self-harm did not increase. Future studies with more granular data on method of suicide should explore this question. Either way, investment in firearm violence prevention strategies – including education on safe storage practices and promotion of extreme risk protection orders64 – may help reduce risk for firearm suicide.
For white populations and others who experienced a decline in suicide during the pandemic, a few potentially protective factors introduced during this period may have buffered or modified the expected association between the stressors of 2020 and 2021 and increased risk for suicide. For instance, a sense of shared experience may have offset the lack of social interaction by creating a feeling of collective purpose.65 In addition, people who lived with others during the stay-at-home order may have been alone less often or under higher levels of supervision or scrutiny within their home, which may have reduced self-harm. Reductions in in-person healthcare appointments at the start of the pandemic led to the widespread adoption of telehealth, which may have increased some individuals’ access to mental healthcare.66 Finally, COVID-19 relief payments may have offset financial strain for some.67 Each of these potentially protective factors are likely differential based on one’s access to remote work and other economic protections; non-Hispanic white and male Californians are relatively advantaged in both regards,68 which could help explain the decline in suicide and self-harm those groups experienced.
To our knowledge, this is the first study to assess incidence of firearm suicide and nonfatal self-harm across sociodemographic groups in California following the onset of the pandemic. There are, however, several limitations. First, due to data availability, we are only able to stratify deaths by suicide across one method (i.e., firearm) and we are unable to distinguish the methods used in nonfatal self-harm. As such, we cannot pinpoint the method of suicide or self-harm driving observed changes and are unable to compare fatal and nonfatal firearm self-harm trends (although the latter is rare given the lethality of firearm suicide attempts).69 Further, our inability to stratify death data across more than one domain restricts the nuance of our analyses. Future research should characterize risk across the intersection of multiple groups and compare changes in other methods of suicide. In addition, our nonfatal data may not generalize to the whole of California because only UC hospitals were captured in the sample, and all were located in urban areas. Rates for nonfatal data may over- or under-estimate the true burden across groups because of the inherent uncertainty in defining hospital catchment areas; however, assuming that there were no major changes in population size over our study period, relative rates across years should remain accurate. Finally, we are unable to identify the intent of nonfatal self-harm incidents and thus cannot distinguish between suicidal self-harm and non-suicidal self-injury, a prevalent condition with distinct etiology.70