The COVID‐19 pandemic impact on wellbeing and mental health in people with psychotic and bipolar disorders

Abstract Introduction The COVID‐19 pandemic affects people globally, but it may affect people with psychotic and bipolar disorders disproportionally. Our aims were to investigate the pandemic impact on perceived wellbeing and mental health in this population, including which pandemic‐related factors have had an impact. Methods People with psychotic and bipolar disorders (N = 520; female = 81%; psychotic disorders n = 75/bipolar disorder n = 445) completed an online survey about wellbeing and mental health in the early phase of the COVID‐19 pandemic (June 5–July 5, 2020). Results Many participants experienced deteriorated wellbeing and mental health after the pandemic outbreak, especially in life satisfaction, meaning in life, positive feelings, depression, anxiety, and self‐harm/suicidal ideation. Experienced recovery from mental health difficulties was significantly lower after compared to before the outbreak. Participants with psychotic disorders had significantly poorer wellbeing and mental health than participants with bipolar disorders, although they experienced significantly more worsening only of psychotic symptoms. Nearly half the participants reported coping with the situation; however, most factors potentially important to wellbeing and mental health changed adversely, including sufficiency and quality of treatment. More loneliness, low coping, insufficient mental health treatment during the COVID‐19 pandemic, pandemic worry, more insomnia symptoms, and increased alcohol use predicted poor wellbeing and poor mental health. Conclusions During a pandemic, it is particularly important that mental health services strive to offer the best possible treatment under the current conditions and target loneliness, coping strategies, pandemic worry, insomnia, and increased alcohol use to uphold wellbeing and reduce mental health difficulties. For some, teletherapy is an agreeable substitute for traditional therapy.


INTRODUCTION
The Coronavirus disease-19 (COVID-19) developed fast into a pandemic with increasing death rates worldwide. Most governments have initiated public measures to prevent the spread of COVID-19, including wearing face masks, physical/social distancing, and self-isolation (WHO, 2020b). The pandemic has an unprecedented impact on people's lives by causing worry about contracting the virus and challenges living with social precautions. This may have negative consequences for peoples' mental health (Holmes et al., 2020;Kumar & Nayar, 2021).
Indeed, studies report that distress, disturbed sleep, anxiety, and depression have been common reactions in the general population during the current and previous virus outbreaks (e.g., the Severe Acute  (Alkhamees et al., 2020;Rajkumar, 2020b;Salari et al., 2020;Torales et al., 2020;Vindegaard & Benros, 2020). The pandemic implications for people's wellbeing, that is, positive feelings, life satisfaction, meaning in life, and social connectedness (Chan et al., 2018) are more uncertain. Alongside potential negative effects, people may experience a renewed sense of shared purpose in the joint combat of the virus (Brown et al., 2020), and feelings of happiness, freedom, and an increased sense of calm have been reported resulting from a slower pace of life (Simblett et al., 2021).
Although the pandemic affects people globally, there are concerns that it impacts people with severe mental disorders disproportionately.
Severe mental disorders include psychotic disorders (PsD; characterized by presence of psychotic symptoms) and bipolar disorders (BD; characterized by the presence of (hypo)manic and depressive symptoms). Psychotic disorders and bipolar disorders overlap in terms of symptomatology, with around 50% having secondary affective symptoms and psychotic symptoms, respectively (Romm et al., 2010;Simonsen et al., 2011). Thus, they can be regarded dimensionally rather than categorically different. People with PsD and BD may be more vulnerable during the pandemic due to higher sensitivity to stress, smaller social networks, high prevalence of substance use, sensitivity to circadian rhythm disruption, and dependency of social, community, and mental health services, which are all factors potentially affected by the pandemic (Brown et al., 2020;Holmes et al., 2020;Kozloff et al., 2020;Rajkumar, 2020a). Distress, loneliness, sleep problems, change in substance use, practical/financial problems, and lack of treatment may reduce wellbeing and affect mental health by exacerbation of anxiety, depression, suicidality, mania, or psychosis.
Based on PsD and BD overlapping in terms of symptomatology, it is of interest to investigate whether or how they potentially differ in response to the life-changing situation brought about by the pandemic.
Some studies have investigated the impact of the COVID-19 outbreak on this population, with reports of higher levels of distress, alcohol use, sleep disruption, anxiety, depression, and poorer coping strategies compared to healthy controls (González-Blanco et al., 2020;Solé et al., 2020;Van Rheenen et al., 2020), with 30% showing symptom relapse and 5% reporting increased suicidality after the outbreak (Muruganandam et al., 2020). However, Pinkham et al. (2020) found no change in affective or psychotic symptoms and an increase in wellbeing in people with schizophrenia spectrum and affective disorders.
Thus, the findings are somewhat inconsistent, with little knowledge about pandemic impact on wellbeing and psychotic symptoms specifically. Increased knowledge about pandemic effects may suggest measures that could mitigate negative consequences during the ongoing and future pandemics.
The primary aims of this study were thus to investigate the COVID-19 impact on the experience of both wellbeing and mental health difficulties in people with PsD and BD, including which pandemic-related factors have had an impact.

Setting and procedures
The Norwegian Government announced a nationwide lockdown on March 12, 2020 in order to reduce the spread of COVID-19. Protecting healthcare professionals was highlighted, with the intention to uphold healthcare services (Government.no, 2020a). Physical distancing (2 m indoors/1 m in public spaces) and good hand hygiene were emphasized, and number of people allowed to meet was restricted. The general recommendation was to "avoid contact with other people." People who tested positive for COVID-19 were to be isolated, and those who had been exposed to the virus had to quarantine. Day-care centers, educational institutions, fitness centers, swimming pools, and establishments providing hair and body care were closed. Restaurants, bars, etc. were closed except where visitors could keep a distance of at least 1 m. Cultural events, sporting events, organized sporting activities, and visits to holiday properties (e.g., summerhouses) were prohibited. Use of public transportation and all non-essential travels were advised against.
Healthcare workers were prohibited from traveling abroad. Access to public healthcare facilities was restricted, and visits were prohibited.
One-to-one health services (e.g., private mental health therapy, physiotherapy, etc.) that could not uphold physical distancing were closed.

Participants
People with PsD and BD were invited by OUS to respond to how they experienced the COVID-19 pandemic. The first item in the survey was to check off for the diagnostic group that the participants identified with. We chose to use the term "psychotic disorders" rather than the diagnostic equivalents (e.g., F20-F29 in the ICD-10 [WHO, 1992]), because this is a more commonly used term in Norway. There were no other inclusion or exclusion criteria.

Measurements
In the survey, participants were informed that the questions concerned the situation during the COVID-19 pandemic, meaning after March 12, 2020. Questions about demographics covered age, gender, place of birth, immigration background, education, and marital status.
Wellbeing was measured using a Norwegian guideline list for wellbeing measurement (Nes et al., 2018), which is based on the OECD Guidelines on Measuring Subjective Well-being (OECD, 2013) and Diener's Flourishing Scale (Diener et al., 2009

Statistical analysis
Statistical analyses were performed with IBM SPSS Statistics 26. Analyses were two-tailed with a pre-set significance level of .05. Diagnostic group differences were analyzed with t-test, Mann-Whitney Relevant independent variables were entered into the regression anal-yses for each of the two dependent variables, taking out those that did not have a significant contribution one by one.

Demographics
Five hundred and twenty participants completed the survey (BD, n = 445; PsD, n = 75). Demographic data are presented in Table 1. The only significant diagnostic group differences were BD participants having higher education and being married or cohabitant more often than the PsD participants.
The participants' experience of recovery (i.e., improvement) from mental health difficulties was significantly reduced from before to after the COVID-19 outbreak (see Figure 1). Furthermore, the median recovery experience score was reduced from 6 to 3 (0-10 scale) from before to after the outbreak (z = −9.7, p < .001). PsD participants had lower median recovery scores post outbreak than BD participants (1 vs. 3; U = 13,793.0, p = .015).

3.3.1
Concerns and coping F I G U R E 1 Participants experience of recovery † (i.e., improvement) from their mental health difficulties before and after the COVID-19 outbreak ‡ (N = 520). † Experience of recovery from mental health difficulties was measured on a 0 ("not at all")-10 ("to a great extent") scale; low level ≤ 5, high level ≥ 6. ‡ Before the COVID-19 outbreak = before March 12 2020; after the COVID-19 outbreak = June 5-July 5, 2020. Change from before to after the COVID-19 outbreak was analyzed with McNemars' test (χ 2 = 81.5, p < .001). 53% (n = 275); 47% (n = 245) / 26% (n = 133); 74% (n = 387) 3.3.2 Housing, daily activity, and economy Table 3 shows that in the BD group there was a significant increase in living with family/cohabitant from before to after the outbreak. There was also a significant reduction in studying and full-time work and increase in temporarily lay-offs and "other" activities. One-in-three reported worsened personal economy after the outbreak (see Table 4). Table 4 shows that the majority of participants experienced worsened social life post outbreak, feeling more isolated, outside the community and lonely. Family conflicts had worsened for some.

TA B L E 3
Changes in housing and daily activity from before to after the COVID-19 outbreak a in the total sample and across diagnostic groups

Substance use
Users of alcohol and illicit drugs reported increased use post outbreak (Table 4).

Medication use
Some participants taking antipsychotics used more post outbreak, with comparatively fewer reporting increased use of mood stabilizers and antidepressants (Table 4). Half of the participants using anxiolytics reported increased use post outbreak. Table 4 also shows data on sleep. Participants reported insomnia symptoms half of the nights and for many this had increased post outbreak.

Sleep
Troubling nightmares were less frequent, but one-in-four reported an increase. There were no pre-post changes in sleep duration (p = .212).

TA B L E 4
Pandemic-related factors: Changes in factors potentially important to wellbeing and mental health difficulties from before to after the COVID-19 outbreak a in the total sample and across diagnostic groups

3.3.7
Mental health services Table 5 shows that in the total sample and among BP participants significantly fewer participants were in treatment for their mental disorder after compared to before the outbreak, with fewer receiving treatment from general practitioners.
Regarding community and charity mental health support, n = 86 participants had support pre-outbreak. The group difference (PsD  Table 6 presents results from the binary logistic regression analyses with poor wellbeing as dependent variable and age, being single, worry about pandemic consequences, low coping, loneliness, insufficient treatment, poorer economy, increased alcohol use, and insomnia symptoms entered as independent variables. Insufficient treatment, more loneliness, and low coping significantly and independently predicted poor wellbeing in both the first and the final model. Table 6 also presents results from the binary logistic regression analyses with poor mental health as dependent variable and age, worry about pandemic consequences, low coping, loneliness, insufficient treatment, poorer economy, increased alcohol use, and insomnia symptoms entered as independent variables. Insufficient treatment, more loneliness, low coping, worry about pandemic consequences, more nights with insomnia symptoms, and increased alcohol use significantly and independently predicted poor mental health in the final model. All variables apart from age, poor economy and increased alcohol use were significant in the first model. Increased alcohol use became significant in the model when age and personal economy were removed. TA B L E 6 Logistic regression analyses with poor wellbeing a (no/yes, n = 233/287) and poor mental health b (no/yes, n = 178/342) as dependent variables a Poor wellbeing was defined as a score of ≤5 (on a 0-10 scale) and in addition responding that the item had become "worse" after the COVID-19 outbreak (i.e., after March 12, 2020) on either one of the following items: life satisfaction, meaning in life, social support, or feeling happy. b Poor mental health was defined as a score of ≥5 (on a 0-10 scale) and in addition responding that the item had become "worse" after the COVID-19 outbreak (i.e., after March 12, 2020) on either one of the following items: anxious, worried, depressed mood, little interest/pleasure, self-harm/suicidal ideation, ideas of persecution, or hallucinations. c Model chi-square = 171.884 df = 3, p = .000. The model as a whole explained between 28.1% (Cox and Snell R 2 ) and 37.7% (Nagelkerke R 2 ) of the variance and correctly identified 75.0% of the cases. d Loneliness past 2 weeks was scored on a scale from 0 ("not at all")-10 ("to a great extent"). e Model chi-square = 180.081, df = 6, p = .000. The model as a whole explained between 29.3% (Cox & Snell R 2 ) and 40.5% (Nagelkerke R 2 ) of the variance and correctly identified 78.5% of the cases.

DISCUSSION
The main findings in this study are that the majority of participants with BD and particularly PsD experienced low levels of wellbeing and high levels of mental health difficulties in the early phase of the COVID-19 pandemic, with around half reporting that they had experienced worsening post outbreak. Participants' experience of being in recovery from mental health difficulties was significantly lower after compared to before the outbreak. Amongst pandemic-related factors, low coping with the situation, loneliness, and insufficient treatment had a negative impact on both wellbeing and mental health difficulties, while worrying about pandemic consequences, increased alcohol use, and insomnia symptoms only affected mental health adversely. Others have reported that depression and anxiety symptoms are prevalent reactions to the COVID-19 pandemic in the general population (Rajkumar, 2020b;Salari et al., 2020), and even higher levels of depression and anxiety in people with PsD and BD compared to healthy controls (González-Blanco et al., 2020;Van Rheenen et al., 2020). The worsening of self-harm/suicidal ideation in 33% of participants is highly concerning, as pandemic disruption to mental health services may reduce prevention of suicides (Sher, 2020). Whether our results would be similar in people without a pre-existing mental disorder is unknown, but a recent study finds no increase in self-reported mental disorders or suicidal ideation from before the pandemic compared to the early phase of the pandemic (March 12 to May 31, 2020) in the general Norwegian population (Knudsen et al., 2021). Furthermore, this study presents data from the Norwegian Cause of Death Registry, which showed no increase in suicide deaths from March to May 2014-2018 compared to March to May 2020. Also worth noting are findings that levels of anxiety in people seeking help for anxiety and depression increased in the first 4 weeks of the pandemic, but then declined the subsequent weeks . Whether our target group experienced the same fluctuations in symptom levels is unknown.

Wellbeing and mental health difficulties
In line with expected seasonal increase in mania during spring and summer (Wang & Chen, 2013), 20-50% of the total sample reported increased symptoms of mania; however, only 6% experienced an increase in ideas of grandiosity. As approximately 50% in both groups experienced more irritable mood post outbreak, this item probably captured irritability beyond mania. A prospective study of people with BD found no significant increase of depressive symptoms, but an increase in symptoms of (hypo)mania in the first wave of the pandemic, with a decrease thereafter (Koenders et al., 2021). with PsD is in line with one study (Muruganandam et al., 2020), yet inconsistent with another (Pinkham et al., 2020).

Pandemic-related factors and their impact on wellbeing and mental health difficulties
The participants were obviously concerned, with high levels of worrying about pandemic consequences, keeping updated, and following recommendations. Worrying about consequences predicted poor mental health, in line with the general population where such worries have been linked to anxiety (Academy of Medical Sciences, 2020).
Despite most participants being concerned, as many as 45% reported coping with the situation "a lot." Nevertheless, low coping with the situation predicted both poor wellbeing and poor mental health, suggesting that coping strategies are essential. Almost all participants reported keeping updated via relevant channels, indicating that people with severe mental disorders do keep in touch with world events, contrary to accounts of this population being inadequately informed about the pandemic (Hölzle et al., 2020;Muruganandam et al., 2020).
Almost all participants reported that they were following recommendations about hand washing and social distancing, in line with a current study where only 13% of people with severe mental disorders were observed to have problems following protective measures (Mork et al., in press). Studies from previous pandemics of people with PsD found markedly inadequate adherence to protective measures (Brown et al., 2020). The inconsistencies between studies concerning keeping updated and following recommendations may reflect different methodology and different epidemics/pandemics. However, it may also reflect that people with severe mental disorders are heterogeneous, also in the face of a pandemic, and that they require different interventions.
The majority experienced deterioration in their social life post outbreak, presumably in line with the general population. Several COVID-19 campaigns have been designed around "we-are-all-in-thistogether" (Nilsen & Skarpenes, 2020;Society of Editors, 2020). This message did clearly not have the full-intended effect on the majority of our sample, who were feeling more outside of the community. The worsening of social isolation and loneliness is not in keeping with the general population, where levels of loneliness remained stable or even decreased during early months of the pandemic (Luchetti et al., 2020;NIPH, 2020). Our findings are in line with previous concerns that social restrictions may impact more severely on people with severe mental disorders (Brown et al., 2020). We found that loneliness predicted both poor wellbeing and poor mental health, consistent with previous findings (Beutel et al., 2017;Eglit et al., 2018). However, over half of the participants experienced social support, with few experiencing a reduction post outbreak. This can be regarded as a resource for interventions aimed to counteract pandemic consequences.
Users of illicit drugs and alcohol reported increased use post outbreak, in line with other studies (Pinkham et al., 2020;Van Rheenen et al., 2020). Increased substance use may reflect maladaptive coping with pandemic distress, anxiety and depression, and/or result from a more monotone/boring life or fewer regulating social constraints.
Increased alcohol consumption requires attention, as increased alcohol use predicted poor mental health, as anticipated (Rajkumar, 2020a).
The most prominent finding concerning medication use is that half the participants using anxiolytics reported increased use, probably reflecting pandemic concerns and increase in anxiety symptoms.
Whether the increased use of medications was according to prescription or self-medication is unknown. However, reduced access to general practitioners and mental health services and a reluctance to seek help during the lockdown may have caused participants to selfmedicate.
Insomnia symptoms were prevalent, and for many participants sleep problems had worsened post outbreak. Other studies report more post outbreak sleep disruptions in people with severe mental disorders compared to healthy controls (Solé et al., 2020;Van Rheenen et al., 2020). The worsening is concerning due to adverse outcomes associated with poor sleep (Laskemoen et al., 2019). In line with this, we found that high levels of insomnia symptoms predicted poor mental health.
Significantly fewer BD participants received treatment post outbreak, suggesting that they were not prioritized and/or more reluctant to seek help. In fact, almost half of the total sample had refrained from contacting mental health services for issues they normally would. Bearing in mind the poorer mental health of participants with PsD compared to participants with BD, we regard it as positive that participants with PsD did not report reduction in being in treatment. The group of participants in inpatient treatment before and after the outbreak was small, and the reduction in inpatient treatment in the total sample from before to after the outbreak was no longer significant after Bonferroni correction. A small group of participants received treatment from ambulatory teams, and an interesting finding was that ambulatory care was unchanged post outbreak. Ambulatory teams in Norway provide care for people with severe mental health difficulties, including poor functioning, thus maintaining ambulatory care for this group may have been prioritized after the outbreak. Still, more than one-third of the total sample reported insufficient treatment for their mental disorders and reduced treatment quality post outbreak. Our findings that experienced recovery from mental health difficulties was halved, and that insufficient treatment predicted both poor wellbeing and poor mental health suggests that participants needed the same or more treatment in these troubled times. In line with this, studies have found increased mental help seeking and service demand in the early weeks of the pandemic, and later when COVID-19 transmission was high (Staples et al., 2021;Titov et al., 2020). However, a WHO survey found that 93% of responding countries reported pandemic disruptions to mental health services, despite goals to ensure continue of care (WHO, 2020a), indicating how uncertainty about timing and impact of a pandemic poses challenges for planners and service providers. The Norwegian Directorate of Health (2021) reports that number of adult patients treated in public mental health outpatient clinics remained the same in 2019 and 2020, mainly due to a 50% increase in treatment delivered via phone or videoconference in 2020, while number of admissions to mental health inpatient wards was reduced by 7% from 2019 to 2020. Of note is that these numbers do not reflect number of treatment sessions and may not reflect mental health services in the early phase of the pandemic, when they were trying to adjust to the situation. Some of our participants had teletherapy, most prevalently via phone or video conference. For some participants this was an agreeable solution, while others did not want to continue with this mode of treatment delivery, indicating that teletherapy on its own is not adequate for everyone. In many countries, the pandemic also led to increased capacity for mental health helplines (WHOa, 2020). However, few participants in our study had called mental helplines. People with PsD and BD may find helplines to be inadequate to accommodate their specialized needs; continuity in care and an established therapeutic alliance may be especially important for this population.

Clinical implications
The clinical implications of our findings include that low wellbeing and increased anxiety and depression may be normalized as common reactions during a pandemic. New knowledge from this study is that mental health services should be particularly aware of potential increase in self-harm/suicidal ideation and psychotic symptoms in people with PsD and BD. Moreover, mental health workers should probe actively for signs of loneliness, insomnia, and increased alcohol use and offer adequate treatment. Poor coping should be targeted, for example, by introducing strategies suggested by other people with mental disorders, including cognitive coping strategies (Simblett et al., 2021) and behaviors such as staying connected, keeping busy, physical activity, staying calm, managing media intake, and maintaining routine (Academy of Medical Sciences, 2020). Peer-support could be beneficial in this context and is currently an underused resource, which should be promoted. Experiences of insufficient treatment suggests that mental health services were not adequately prepared for the COVID-19 pandemic; however, services should seek to deliver the best possible care under current conditions. This population's potential vulnerability and reluctance to seek help implies a need for active outreach via available channels. Development of customized teletherapy is needed to fit individual needs. Meanwhile, mental health services should strive to uphold physical sessions when called for, requiring supply and training in infectious control equipment. When conditions prevent optimal mental health services, all available resources should be considered, including family members who already provide important support (Eckardt, 2020). Empowering family members requires establishing contact pre-crises and supporting them during strenuous times.

Strengths and limitations
Our sample is relatively large, potentially from across the country, with a wide age range. The majority of respondents being female is also seen in other online surveys (Academy of Medical Sciences, 2020; Solé et al., 2020). Diagnosis was based on self-report; thus, the reliability of the diagnoses and specific diagnostics are uncertain. Moreover, comorbid disorders were not recorded. The sample is skewed towards people with BD. This may at least partly be explained by the user organization for bipolar disorder in Norway sharing the survey to its members, while there is not an equivalent user organization for people with psychotic disorders in Norway. The representativity of our sample may be biased regarding severity of illness and service satisfaction.

CONCLUSIONS
Our findings indicate that the early stages of the COVID-19 pandemic had serious consequences for wellbeing and mental health difficulties in many people with PsD and BD. Adverse change in treatment sufficiency, loneliness, insomnia symptoms, alcohol drinking, pandemic worry, and low coping was related to the deterioration. Our findings suggest a need to increase general disaster preparedness in mental health services to ensure provision of sufficient care under suboptimal conditions. Future research should investigate what service users think the health services could do to improve during a pandemic.

ACKNOWLEDGMENTS
We thank Sindre Hembre Kruse, Karoline Fløystad-Thorsen, Kristine Gjermundsen, TIPS consultants in the South-Eastern Health Region of Norway, and TIPS Stavanger for help with designing and distributing the survey.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1002/brb3.2559