The COVID-19 pandemic has presented an unprecedented source of stress for people around the world (
1–
7). Contributing factors have included fear of contracting COVID-19 and of decreased access to food and essential services (
8). The social isolation necessitated by infection control procedures also has contributed to distress (
9).
Persons with serious mental illness may be particularly vulnerable to the effects of the pandemic. Compared with the general population, such individuals may have more limited resources, smaller social networks, and higher utilization of health care and social services that may be disrupted by the pandemic (
10–
12). Persons with serious mental illness also have relatively high rates of tobacco smoking and co-occurring medical conditions, such as diabetes, which are risk factors for COVID-19 (
13–
15). Recent studies also indicate that persons with schizophrenia have had excess COVID-19–related morbidity and mortality (
16,
17). In addition, worry about the pandemic may exacerbate preexisting anxiety in this population, and enforced isolation may aggravate feelings of loneliness and depression (
18).
The experiences of persons with serious mental illness during the COVID-19 pandemic have not been extensively characterized because the pandemic began to emerge in the United States in early 2020, and few studies on this topic have been published (
5,
19–
26). Most of these studies found that individuals with serious mental illness, or psychiatric disorders more broadly, experienced more acute distress during the pandemic than did persons in the general population (
19,
20,
24–
26). Other studies found that persons with serious mental illness did not experience any worsening of psychiatric symptoms compared with before the pandemic or did not experience a higher level of mood symptoms compared with persons without psychiatric disorders (
21–
23). Most of these studies were performed during the early stages of the pandemic.
The purpose of this study was to characterize the experiences associated later in the pandemic of a cohort of individuals diagnosed as having schizophrenia, bipolar disorder, or major depressive disorder as well as persons without a psychiatric disorder. We explored whether participants believed they had contracted COVID-19, their mental distress and suicidal thoughts, the impact of pandemic-related worries, their substance use, and their access to medical care. The responses of persons with serious mental illness were compared with responses of those without a psychiatric disorder and were correlated with current and previously collected clinical and demographic data.
Methods
Persons in the study had enrolled in a research study at the Stanley Research Program at Sheppard Pratt between July 2000 and February 2020 and had given permission to be contacted again. Inclusion and exclusion criteria have been described previously (
27–
30). All participants were previously assessed with a structured clinical interview to confirm a psychiatric diagnosis or the absence of one. Most individuals were assessed with a cognitive battery, the Repeatable Battery for the Assessment of Neuropsychological Status (
31). Participants with a psychiatric disorder were assessed with a measure of psychiatric symptom severity, the Positive and Negative Syndrome Scale (
32), and a subset who enrolled after December 2016 were assessed with the Columbia Suicide Severity Rating Scale (C-SSRS) (
33).
Persons were contacted for the current study by phone, e-mail, or U.S. postal mail and were invited to participate. Written informed consent was obtained after the study was explained. Participants were individually interviewed by videoconferencing or telephone. Interview questions focused on whether participants, persons in their household, or other persons close to them had contracted COVID-19; mental distress and suicidal thoughts; responses to the COVID-19 pandemic, including its overall impact and the degree of worry related to different aspects of the pandemic; current use of tobacco and alcohol and changes in use since the start of the pandemic; and changes in access to medical care since the start of the pandemic and recent missed health care visits and medications. Items about contracting COVID-19, tobacco smoking, and health care access were modified from items in the Johns Hopkins University COVID-19 Community Response Survey (
34); items about the impact of and concerns related to the pandemic were from a Pew Research Center survey (
35) and the C-SSRS. The study was approved by the Sheppard Pratt Institutional Review Board.
Data were analyzed with logistic regression for dichotomous variables and with ordered logistic regression for outcomes measured on an ordinal scale. Regression analyses were performed with diagnostic group as the dependent variable and age, race, and gender as covariates. Odds ratios (ORs) were computed for each psychiatric diagnostic group relative to the group without a psychiatric diagnosis. Log rank was used to assess overall differences among groups, and the Wald test was used to compare individual ORs. Additional covariates, including previously collected demographic characteristics and measures of symptom severity and cognitive functioning were added to the regression analysis for the item assessing feeling overwhelmed by COVID-19 to determine their effect on the overall model. Mixed-effects models were used to compare previous and current responses to C-SSRS suicide variables, with age at study entry, age at time of survey, gender, diagnostic group, race, and time between measures as fixed-effect covariates and with variations within individuals as random effects. A value of α≤0.0125 was considered statistically significant for analyses with four groups to partially adjust for multiple comparisons, and a p value of 0.0125–0.05 was considered indicative of a trend. Analyses focused on single diagnostic groups were not adjusted. Data were analyzed in Stata, version 16.0.
Results
Research staff attempted to contact 790 persons, of whom 280 were located and communicated with study staff. Of these individuals, 195 (70%) signed consent forms and completed the study. Each interview lasted approximately 30 minutes. A total of 84 (43%) interviews were conducted over videoconferencing (i.e., Zoom), and 111 (57%) were done by telephone. The individuals who participated did not differ significantly in terms of age at time of initial enrollment, gender, or diagnostic group from those whom research staff attempted to contact but who did not participate. We found significant differences in terms of race (White persons were more likely than people of other races to complete the survey) and year of initial enrollment (participants had been initially enrolled more recently than those who did not participate).
The interviews were conducted between July 15, 2020, and January 22, 2021. During this period, Maryland, where most participants resided, had emerged from a lockdown and underwent phases of reopening and restrictions. Telehealth services were largely available. A COVID-19 vaccine was not available to any participants when they were interviewed.
As shown in
Table 1, the 195 participants were divided among those diagnosed as having schizophrenia (N=50), bipolar disorder (N=51), and major depressive disorder (N=29) and those with no psychiatric disorder (N=65). The mean±SD age of participants was 41.0±13.3 years (range 19–72); 83 (43%) participants were male, 126 (65%) were White, 46 (24%) were Black, and 23 (12%) were of “other” race. The mean follow-up interval since initial enrollment was 6.14±4.0 years.
Infection With COVID-19
Seventeen (9%) participants reported that they had received a positive COVID-19 test or had had symptoms consistent with COVID-19, and 10 (5%) reported that a person in their household had contracted COVID-19. These percentages did not differ significantly among diagnostic groups. Eighty (41%) participants indicated that a close friend or family member outside of their household had contracted COVID-19. Significantly fewer persons with schizophrenia reported this outcome than participants without a psychiatric disorder; fewer persons with depression also reported this outcome, but this finding was not statistically significant (
Table 2).
Mental Distress
Participants were asked about the frequency of symptoms of mental distress over the past 7 days. As shown in
Figure 1, most of the groups with a psychiatric disorder reported more frequently feeling nervous, depressed, lonely, and hyperaroused, compared with those without a psychiatric disorder, and some groups also reported feeling less hopeful and having more difficulty sleeping.
Participants were interviewed about items from the C-SSRS pertaining to suicidal thoughts and behavior. Forty-seven persons (24% of the sample) responded that they had passive suicidal ideas, and 27 (14%) indicated that they had experienced thoughts of killing themselves since the start of the pandemic. These responses included one (2%) and no individuals with no psychiatric disorder, eight (16%) and six (12%) with schizophrenia, 26 (51%) and 18 (35%) with bipolar disorder, and 12 (41%) and three (10%) with depression, respectively.
These responses to the C-SSRS were compared with participants’ responses at the time of their earlier study participation for the 42 participants with a psychiatric disorder (11 with schizophrenia, 22 with bipolar disorder, and nine with depression) who had been previously assessed with this instrument. There was a significant decrease in response to the item measuring nonspecific suicidal ideas (OR=0.18, 95% confidence interval [CI]=0.04–0.73, p=0.017, as measured by mixed-effects models using age, gender, race, and time between assessments as fixed-effect covariates).
Of the individuals with suicidal thoughts in the current study, two persons with schizophrenia (4%), 12 (24%) with bipolar disorder, and two (7%) with depression reported that they had been thinking about how they might kill themselves; four individuals with bipolar disorder indicated that they had had suicidal intent, although not currently. Two participants with bipolar disorder had made a suicide attempt since the start of the pandemic. These results could not be statistically analyzed because of the small numbers.
Responses to the COVID-19 Pandemic
Participants were asked to indicate how much they agreed or disagreed with a series of 12 statements focused on pandemic-related worries. As shown in
Figure 2, individuals with schizophrenia indicated significantly more worry than did participants with no psychiatric disorder in response to five of the items (feeling overwhelmed by COVID-19, feeling worried about not having enough food because of COVID-19, feeling worried about medical bills if they get sick from COVID-19, experiencing difficulties concentrating, and having anxiety related to COVID-19), and they trended toward more worry about money and about getting COVID-19, but this finding did not meet the threshold for significance. Those with bipolar disorder also reported significantly more worry than the participants with no psychiatric disorder for two of these items: worry about having enough food and about money. The remaining items in this section showed a nonsignificant difference between any of the groups with a psychiatric disorder and those without a psychiatric disorder.
Participants were also queried about the extent to which their daily routine had been affected by the COVID-19 pandemic, the severity of their stress related to the pandemic, and the overall severity of the impact of the pandemic. Unlike the responses about worries, the responses to these items indicated no significant differences between any of the groups with a psychiatric disorder and those with no disorder (
Table 2).
We further examined the role of clinical and demographic measures associated with the response to feeling overwhelmed by COVID-19, which was the response that differed most among the groups. In addition to diagnosis, response to this item was significantly associated with lower age (OR=0.97, p=0.003) and female gender (OR=2.27, p=0.006). Regression models containing these covariates were not significantly altered by the addition of race, mother’s education, participant education and level of cognitive functioning, or symptom severity at time of their initial enrollment.
Tobacco and Alcohol Use
As shown in
Table 1, the groups differed in their current use of tobacco and alcohol. As shown in
Table 2, participants with schizophrenia reported significantly increased tobacco smoking, and participants with depression reported a trend toward increased alcohol use since the start of the pandemic, compared with participants with no psychiatric disorder. The other groups with a psychiatric disorder did not report a change in their use of these substances, which differed from that among participants with no psychiatric disorder.
Access to Care
All of the groups with a psychiatric disorder reported a higher rate of recent missed medical appointments than participants with no psychiatric disorder, as shown in
Table 2. Individuals with bipolar disorder and depression were more likely to report a higher rate of recent missed medications and showed trends for increased problems with access to medical care.
Discussion
This article describes the experiences of a cohort of 195 persons with serious mental illness or without a psychiatric disorder during the COVID-19 pandemic. Participants were queried about whether they believed they had contracted COVID-19, mental distress and suicidal thoughts, change in daily routine and worries related to the pandemic, tobacco and alcohol use, and access to care. All of the groups with a psychiatric disorder reported a higher frequency of symptoms of mental distress than participants with no disorder. Individuals with schizophrenia were the psychiatric group most likely to endorse COVID-19–related worries. The psychiatric groups reported more recent missed medical visits and medications than those with no psychiatric disorder. Of note, the proportion of individuals who reported having contracted COVID-19 did not differ among groups.
Our findings are consistent with several recent studies documenting that persons with serious mental illness reported a higher level of mental distress compared with persons without a psychiatric disorder during the pandemic (
19,
20,
22,
24–
26). This result is not surprising given the preexisting psychiatric disorders of these persons. On the other hand, we found that the groups with a serious mental illness did not report more overall COVID-19–related stress or changes in daily routines than did participants without a psychiatric disorder, perhaps reflecting the high level of disruption in day-to-day life experienced by all persons in the population affected by the pandemic (
2,
3). It is also possible that some persons with serious mental illness spend less time in structured daily activities, which may moderate the adverse effects of pandemic-related restrictions (
36).
In contrast with responses to general questions about COVID-19–related stress, participants with a psychiatric disorder reported a higher level of specific pandemic-related worries than did those without a psychiatric disorder. The responses of individuals with schizophrenia, and to some extent the responses of those with bipolar disorder, were more likely to reflect feeling overwhelmed by the pandemic, difficulty concentrating because of the pandemic, and worry about food. Among persons with serious mental illness, individuals with schizophrenia may have been the most affected by the suspension of group activities and most prone to worries about how their daily needs would be met. Persons with schizophrenia may also have had the most limited access to smartphones and other digital technologies, which buffer the effects of COVID-19 restrictions (
37).
Suicidal behaviors and ideation have been identified as a problem related to the pandemic (
25,
38), but few studies have examined this issue systematically (
21,
26). A portion of persons in all of the groups with a serious mental illness reported passive suicidal ideas and/or nonspecific suicidal thoughts since the start of the pandemic. However, the subset of the group that had been questioned about suicidal behaviors prior to the pandemic did not display an increase in such behaviors during the pandemic. In fact, significantly fewer endorsed nonspecific suicidal thoughts than they had earlier. This may be because some participants were initially queried when receiving acute hospital-based care or because their COVID-19–related distress had not manifested as an increase in suicidal thoughts. However, notably, some individuals in the groups with a serious mental illness had active suicidal thoughts and behavior during the pandemic, and two individuals with bipolar disorder had attempted suicide. This finding underscores the importance of ongoing suicide prevention interventions.
Our study also found that women and younger persons were more likely to report feeling overwhelmed by COVID-19. Feeling overwhelmed was not independently associated with race, education level, cognitive functioning, or symptom severity at the time of enrollment. The association with age and gender is consistent with studies of the general population, which have shown that men and older persons are less affected by COVID-19–related stress (
39–
43). Possible explanations include that the experiences of older adults enable them to better manage challenges such as those related to the pandemic (
41) and that the daily activities of men may be less burdened by the pandemic (
43). While the variable of race was not statistically significant in our analyses, we cannot exclude the likelihood that race and the experience of race played a role in persons' responses to the pandemic.
The fact that persons with schizophrenia reported a disproportionate increase in tobacco smoking is of concern, given the already high prevalence of tobacco smoking in this population (
44,
45). Other studies have reported an increase in tobacco smoking during the pandemic (
46), but ours is among the first to study the issue among persons with mental illness (
26). That those with depression reported a trend for a disproportionate increase in alcohol use is also of concern, given the high comorbidity of depression and problematic alcohol use (
47) as well as reports of excess drinking during the pandemic (
24,
48).
Most of the groups with a psychiatric disorder were more likely to report recent missed medical appointments and medications compared with those without a psychiatric disorder. In part, this may be because psychiatric groups are likely to have a higher base rate of medical care and medication usage than nonpsychiatric participants. These results raise concerns about the long-term effects of COVID-19 in terms of health care access for individuals with serious mental illness and the need for heightened monitoring in this population.
Limitations of our study included the possibility that our participants may not be representative of persons in the diagnostic categories studied. Selection bias also may have occurred among those who responded to our study (e.g., those who were more stable economically may have been more easily located). In addition, the small sample size may limit confidence in interpreting the lack of statistically significant differences. Strengths of our study included that we conducted individual interviews rather than relying on online surveys. We had a high participation rate (70%) among persons we were able to locate. Our sample included three psychiatric diagnostic groups and a group with no psychiatric disorder, all of whom were extensively evaluated at the time of their original participation. Because the participants were part of an ongoing cohort, their current interview responses could be correlated with past measures. In addition, the study is among the first to report on the responses of persons with serious mental illness about the pandemic up to 10 months after the pandemic began. Further studies will need to determine the effects of the pandemic on persons with serious mental illness as the pandemic continues to evolve and vaccinations become more available.
Conclusions
Our study indicates the likelihood of long-term effects of COVID-19 among individuals with serious mental illness in terms of their mental health, substance use, and access to care. Additional studies should be performed to further define the specific areas that require heightened interventions. Psychiatric services should be prepared to proactively address the mental distress and worries associated with the pandemic.