key: cord-0821270-9zuq9k21
authors: Czeisler, Mark É; Drane, Alexandra; Winnay, Sarah S; Capodilupo, Emily R; Czeisler, Charles A; Rajaratnam, Shantha MW; Howard, Mark E
title: Mental Health, Substance Use, and Suicidal Ideation Among Unpaid Caregivers in the United States During the COVID-19 Pandemic: Relationships to Age, Race/Ethnicity, Employment, and Caregiver Intensity
date: 2021-09-03
journal: J Affect Disord
DOI: 10.1016/j.jad.2021.08.130
sha: 6641b2268ab1641a21ed09b2868627360c53d107
doc_id: 821270
cord_uid: 9zuq9k21

BACKGROUND: : Unpaid caregivers of adults play critical roles in health care systems by providing care to older adults and those with chronic conditions. The COVID-19 pandemic has heightened caregiving needs, forcing some into caregiving roles and disrupting others. We sought to estimate the prevalence of and identify factors associated with adverse mental health symptoms, substance use, and suicidal ideation amongst unpaid caregivers of adults versus non-caregivers. METHODS: : During June 24-30, 2020, surveys were administered to U.S. adults. Quota sampling and survey weighting were implemented to improve sample representativeness of age, gender, and race/ethnicity. RESULTS: : Of 9,896 eligible invited adults, 5,412 (54.7%) completed surveys and 5,011 (92.6%) met screening criteria and were analyzed, including 1,362 (27.2%) caregivers. Caregivers had higher adverse mental health symptom prevalences than non-caregivers, including suicidal ideation (33.4% vs 3.7%, p<0.0001). Symptoms were more common among caregivers who were young vs older adults (e.g., aged 18-24 vs ≥65 years, aPR 2.75, 95% CI 1.95-3.88, p<0.0001) and with moderate and high vs low Caregiver Intensity Index scores (2.31, 1.65-3.23; 2.81, 2.00-3.94; both p<0.0001). LIMITATIONS: : Self-report data may be subject to recall, response, and social desirability biases; unpaid caregivers were self-identified; child caregiving roles were not assessed; and internet-based survey samples might not fully represent the U.S. population. CONCLUSIONS: : Caregivers experienced disproportionately high levels of adverse mental health symptoms. Younger caregivers and those with higher caregiving intensity were disproportionately affected. Increased visibility of and access to mental health care resources are urgently needed to address mental health challenges of caregiving.

During June 2020, caregivers reported significantly a higher prevalence of adverse mental and behavioral health symptoms than non-caregivers, including symptoms of an anxiety disorder, depressive disorder, or COVID-19-related trauma-and stressor-related disorders (TSRDs), having started or increased substance use to cope with the pandemic, and suicidal ideation (Czeisler et al., 2020a). A study of 1,459 pediatric and adult brain tumor patients and 530 caregivers in 33 countries found that caregivers were significantly more anxious than patients, and 42.8% of caregivers felt that their caregiver burden has significantly increased during the pandemic (Voisin et al., 2020).

Caregivers represent a significant demographic in the U.S. In 2020, the pre-pandemic estimated prevalence of caregivers was 19.2% of adults aged ≥18 years, or 47.9 million Americans (The National Alliance for Caregiving and Public Policy Institute, 2020). This estimate represents an increase in caregivers of more than eight million compared with 2015 (The National Alliance for Caregiving and Public Policy Institute, 2015). People may have taken up unplanned caregiving roles during the pandemic due to mobility restrictions related to community mitigation activities designed to reduce potential exposure to SARS-CoV-2 for older adults. Moreover, some caregivers who had been providing care before the pandemic may have faced barriers and disruptions to their routines and livelihood. Both scenarios require caregivers to make sacrifices to care for others during a time when their own lives may have been disrupted.

Addressing the needs of the disproportionately affected population of caregivers is critically important for the health and well-being of caregivers and the persons for whom they provide care. To effectively address these needs during the COVID-19 pandemic and afterwards, studies are needed to determine the prevalence and characteristics of caregivers, and to identify stressors that may be targets for support systems and prevention and intervention efforts. This study had 6 three specific aims: (1) to estimate the prevalence of U.S. caregivers during the pandemic and compare the demographic characteristics of this population with non-caregivers, (2) to evaluate demographic characteristics associated with adverse mental and behavioral health symptoms separately among caregivers and non-caregivers, and (3) to analyze caregiving characteristics associated with adverse mental and behavioral health symptoms among caregivers.

To assess mental and behavioral health among adults aged ≥18 years with residence in the U.S. who had provided unpaid care for adults during the COVID-19 pandemic, we conducted a cross-sectional analysis of an Internet-based survey study conducted during June 24-30, 2020 for Participants included both first-time respondents and respondents who had completed a related survey during April 2-8, May 5-12, 2020, or both intervals. Demographic quota sampling was used to recruit respondents based on population estimates for age, gender, race, and ethnicity based on the 2010 U.S. Census. Potential respondents likely to qualify based on demographic characteristics listed in their Qualtrics panelist profile were targeted during recruitment; demographic questions were then included in the survey to determine their eligibility. Potential respondents received invitations and could opt to participate by activating a survey link directing them to the participant information and consent page preceding the survey. Ineligible respondents who did not meet inclusion criteria (e.g., age <18 years, not a U.S. resident) or exceeded set quotas (i.e., maximum demographic characteristic quota already met) were not empaneled in the survey.

The survey instruments included individual questions, validated questionnaires, and COVID-19-specific questionnaires used to assess respondent attitudes, behaviors, and beliefs related to COVID-19 and its mitigation, along with mental and behavioral health consequences of the COVID-19 pandemic.

Demographic variables included gender, categorized age, combined race/ethnicity, disability status, marital status, household occupancy, 2019 household income, U.S. Census region, urban/rural classification using self-reported ZIP codes, employment status, and, among employed respondents, self-identified essential worker status and weekly paid work hours.

Caregiving variables included the method by which caregivers provided care (in-person in-home only; in-person out-of-home only; virtually only; and both in-person and virtually), the person for whom they were providing care, weekly unpaid caregiving hours, caregiver experience in months, and caregiving intensity assessed using the 12-or 14-item ARCHANGELS Caregiver Intensity Index (CII; see Supplement (p 1) for additional details), which is composed of three subscales: Caregiver Load based on four items (situation stability, impact on expenses, family strife, and preparedness), Caregiver Impacts based on four items (emotional state, work, personal time, and stress), and Caregiver Buffers based on six items (support, insurance knowledge, selfefficacy, financial knowledge, sense of purpose, and employer support). Caregivers who were also employed completed all 14 items, while those who were not employed completed all items 8 except for the work and employer support items. The sum of items in each subscale is normalized from 0-100, and the normalized sum of the three subscales is used to categorize total CII scores as Low (0-25), Moderate (26-55), or High (≥56). Symptoms of COVID-19TSRDs were assessed via the six-item Impact of Event Scale (IES-6) to screen for overlapping symptoms of posttraumatic stress disorder (PTSD), acute stress disorder (ASD), and adjustment disorders (ADs) (Hosey et al., 2019). Respondents also reported whether they had started or increased substance use, (e.g., alcohol, drugs) to cope with stress or emotions related to COVID-19, or if they had seriously considered trying to kill themselves (suicidal ideation) in the prior 30 days. See Supplement (pp 1-2) for additional details.

All surveys underwent Qualtrics, LLC standard data quality screening procedures, and a secondary cleaning conducted by the investigators; see Supplement (p 2). Respondents who failed an attention or speed check, along with any responses that failed data quality screening procedures, were excluded from the analysis.

All statistical analyses were conducted using Python (version 3 

Summary statistics (counts and percentages) were used to estimate the prevalence of adverse mental and behavioral health symptoms among caregivers and non-caregivers, overall and by demographic characteristics. Univariable Rao-Scott adjusted Pearson chi-squared tests were used to test for differences in observed and expected frequencies among groups by characteristic with a Bonferroni adjustment and evaluated at a significance level of α = 0.05. Additionally, to identify whether specific demographic characteristics were independently associated with adverse mental and behavioral health symptoms within these populations, multivariable Poisson regressions with robust standard errors were used to estimate adjusted prevalence ratios (aPRs) and 95% confidence intervals (95% CIs) for adverse mental and behavioral health symptoms among caregivers, evaluated at a significance level of α = 0.05.

Summary statistics (counts and percentages) were used to estimate the prevalence of adverse mental and behavioral health symptoms among caregivers, overall and by caregiving characteristics. Univariable Rao-Scott adjusted Pearson chi-squared tests were used to test for differences in observed and expected frequencies among caregivers by characteristic with a Bonferroni adjustment and evaluated at a significance level of α = 0.05. To identify whether specific caregiving characteristics were independently associated with adverse mental and behavioral health symptoms, multivariable Poisson regressions with robust standard errors were used to estimate aPRs and 95% CIs for adverse mental and behavioral health symptoms, evaluated at a significance level of α = 0.05.

As an exploratory analysis of individual CII items to determine the relative strength of correlations between caregiver perceptions and adverse mental and behavioral health symptoms, non-parametric Spearman correlations were calculated between each CII item and mental and behavioral health measures.

The Monash University Human Research Ethics Committee reviewed and approved the study protocol (ID #24036). All participants provided informed electronic consent prior to study commencement. Investigators received anonymized responses. (Table 1 ). There was not a significant difference in caregiver status by gender or 2019 household income, though compared with non-caregivers, caregivers were significantly more commonly of young age (e.g., 18-24 years = 26.6% vs 8.0%, respectively, group p<0.0001) and

either Black or Hispanic race/ethnicity (Black = 18.8% vs 9.7%; Hispanic = 29.0% vs 11.6%, group p<0.0001). White respondents accounted for 44.5% of caregivers and 70.8% of noncaregivers. Caregivers also more commonly reported living with a disability than not (37.9% vs 17.0%, p<0.0001), and, among employed caregivers, essential than nonessential worker status (73.7% vs 47.8%, p<0.0001) ( Table 1) .

Adverse mental and behavioral health symptoms were more prevalent among caregivers than non-caregivers (symptoms of anxiety or depressive disorder = 57.6% vs 21.5%, respectively; symptoms of a COVID-19-related TSRD = 49.0% vs 17.9%; having started or increased substance use to cope with the pandemic = 35.0% vs 6.3%; suicidal ideation = 33.4% vs 3.7%; one or more of these symptoms = 69.6% vs 31.0%; all p<0.0001) ( Tables 2,3) .

Among caregivers, adverse mental and behavioral health symptoms were most prevalent among those aged 18-24 years (e.g., one or more symptom, vs those aged ≥65 years; 88.5% vs 18.8%, group p<0.0001), and were more prevalent among Black and Hispanic caregivers than White caregivers (80.2% and 89.4%, respectively, vs 53.4%, group p<0.0001) and among those with than those without disabilities (85.8% vs 59.8%, p<0.0001) ( Table 2 ). There were also differences by employment status, as caregivers who were employed (76.1%) or students (79.0%) had higher prevalences of adverse mental and behavioral health symptoms than those who were retired (29.9%) or unemployed (59.3%) (group p<0.0001). Among employed caregivers, adverse mental and behavioral health symptoms were more common among essential than nonessential workers (81.6% vs 60.6%, p<0.0001), and were most prevalent among those who worked >60 hours in the previous week and decreased with weekly work hours (e.g., vs those who worked ≤20 hours; 96.9% vs 59.3%, group p<0.0001). Overall, demographic characteristics associated with adverse mental and behavioral health symptoms among caregivers were also observed among non-caregivers (Table 3) .

Adjusted prevalence ratios for select demographic variables associated with significantly different prevalences of symptoms of anxiety or depressive disorder, suicidal ideation, and one or more adverse mental or behavioral health symptom, are shown in Figure S2 . Specifically, adjusted prevalence ratios for adverse mental health symptoms were higher among young caregivers aged 18-24 years vs caregivers aged 45-64 years (e.g., anxiety or depressive disorder 

Among caregivers, mental and behavioral health also differed by caregiving characteristics ( for more hours and those who had been caregiving for fewer than 12 months had higher prevalences of adverse mental health symptoms, which may reflect stressors from being forced into a caregiving role, starting as a caregiver during the pandemic, or survival bias (Czeisler et al., 2021d), whereby those who were still providing care after 12 months were more resilient to stressors associated with the role.

The findings in this report reveal that unpaid caregiving for adults is common, has likely increased during the COVID-19 pandemic, and is represented broadly across demographics.

Further, the report underscores the significant impact associated with caregiving on mental and behavioral health and highlights the compounding impact of intersectionality with those who identify in multiple groups having elevated experiences of adverse mental and behavioral health. Czeisler et al., 2020b, 2021b) , which may be related to a combination of their own perceived risk of SARS-CoV-2 infection and to their perceived risk and grief about potentially infecting the person for whom they are caring.

Strengths of this study include recruitment of a large sample of unpaid caregivers from a demographically diverse sample of U.S. adults and utilization of validated screening instruments for mental health. Limitations of this study follow. First, unpaid caregivers of adults were selfidentified, and whether they were caregivers of children or adolescents was not assessed; future research could continue to assess mental health among multigenerational caregivers. Second, a diagnostic evaluation for anxiety disorder or depressive disorder was not conducted; however, clinically validated screening instruments were used to assess symptoms. Third, substance use was self-reported; therefore, responses might be subject to recall, response, and social desirability biases. Fourth, the novel nature of the ARCHANGELS Caregiver Intensity Index and the specific use within this research precludes exact comparisons with normative data on caregiving intensity before the pandemic. Finally, Internet-based survey samples might not be fully representative of the 2020 U.S. population and may therefore have limited generalizability.

However, standardized and supplementary data quality screening procedures were applied, and the prevalence of symptoms of anxiety disorder and depressive disorder were largely consistent with findings from the Household Pulse Survey during June 2020 (Centers for Disease Control and Prevention, 2020).

The COVID-19 pandemic both introduced new challenges (e.g., barriers to in-person care provision, COVID-19 concerns) and exacerbated longstanding challenges (e.g., financial and time strains) associated with caregiving. Therefore, prevention efforts and cultural changes may be required both during and beyond the pandemic to properly address the factors associated with caregiving that contribute to elevated experiences of adverse mental health. This is of increasing importance to the economy, as even before the pandemic, a 2015 study estimated the value of unpaid caregiver labor to be USD$470 billion (Beltrán-Sánchez et al., 2015) .

Given the high prevalence of employed caregivers and its compounding mental health impact, reducing the stigma that can be associated with caregiver status and establishing visible and al., 2021c) . Similarly, a study of 350 caregivers of people with COVID-19 in Iran reported prevalence estimates of anxiety, depression, and stress between 75% and 80%, with higher levels among those who were younger, not exercising, or employed as in a health-related occupation (Jafari-Oori et al., 2021) . Finally, given that approximately 20% of bereaved of depression and complicated grief (Schulz et al., 2008) , caregivers might benefit from preparation for and support during this experience.

Further characterization of caregivers and assessment of mental health, substance use, and suicidal ideation will be required to determine the extent to which increased prevalence of caregiving and elevated adverse mental and behavioral health symptoms progress over the course of the pandemic and beyond. Investment in support systems that reflect the diverse caregiving population and improves their ability to provide care will improve societal health and well-being during this critical health crisis and beyond. * Bonferroni-corrected Rao-Scott adjusted Pearson chi-squared test was used to test for differences in observed and expected frequencies among groups. Significance was assessed at p<0.05. † "Other" race includes American Indian or Alaska Native, Native Hawaiian or Pacific Islander, or Other. ‡ Persons who had a disability were defined as such based on a qualifying response to either one of two questions: "Are you limited in any way in any activities because of physical, mental, or emotional condition?" and "Do you have any health conditions that require you to use special equipment, such as a cane, wheelchair, special bed, or special telephone?" https://www.cdc.gov/brfss/questionnaires/pdf-ques/2015-brfss-questionnaire-12-29-14.pdf. * See Table 3 for the adverse mental and behavioral health symptoms among those who were not unpaid caregivers of adults, by select respondent demographics. † p<0.05 for Bonferroni-corrected Rao-Scott adjusted Pearson chi-squared test between caregivers and noncaregivers. ‡ p<0.05 for Bonferroni-corrected Rao-Scott adjusted Pearson chi-squared test between demographics among caregivers. § "Other" race includes American Indian or Alaska Native, Native Hawaiian or Pacific Islander, or Other. ** Persons who had a disability were defined as such based on a qualifying response to either one of two questions: "Are you limited in any way in any activities because of physical, mental, or emotional condition?" and "Do you have any health conditions that require you to use special equipment, such as a cane, wheelchair, special bed, or special telephone?" https://www.cdc.gov/brfss/questionnaires/pdf-ques/2015-brfss-questionnaire-12-29-14.pdf. * See Table 2 for the adverse mental and behavioral health symptoms among those who were unpaid caregivers of adults, by select respondent demographics. † p<0.05 for Bonferroni-corrected Rao-Scott adjusted Pearson chi-squared test between caregivers and noncaregivers. ‡ p<0.05 for Bonferroni-corrected Rao-Scott adjusted Pearson chi-squared test between demographics among caregivers. § "Other" race includes American Indian or Alaska Native, Native Hawaiian or Pacific Islander, or Other. ** Persons who had a disability were defined as such based on a qualifying response to either one of two questions: "Are you limited in any way in any activities because of physical, mental, or emotional condition?" and "Do you have any health conditions that require you to use special equipment, such as a cane, wheelchair, special bed, or special telephone?" https://www.cdc.gov/brfss/questionnaires/pdf-ques/2015-brfss-questionnaire-12-29-14.pdf. 

Past, Present, and Future of Healthy Life Expectancy. Cold Spring Harb Perspect Med 5, a025957

The Pathophysiology of Perceived Social Isolation: Effects on Health and Mortality

How COVID-19 Affects the Brain

Mental Health -Household Pulse Survey -COVID-19

Early public adherence with and support for stay-at-home COVID-19 mitigation strategies despite adverse life impact: a transnational cross-sectional survey study in the United States and Australia

Delay or avoidance of routine, urgent and emergency medical care due to concerns about COVID-19 in a region with low COVID-19 prevalence

Mental Health, Substance Use, and Suicidal Ideation During the COVID-19 Pandemic -United States

Delay or Avoidance of Medical Care Because of COVID-19-Related Concerns -United States

Mental Health Among Parents of Children Aged <18 Years and Unpaid Caregivers of Adults During the COVID-19 Pandemic -United States

Uncovering survivorship bias in longitudinal mental health surveys during the COVID-19 pandemic

Mental health, substance use, and suicidal ideation during a prolonged COVID-19-related lockdown in a region with low SARS-CoV-2 prevalence

The association between subjective caregiver burden and depressive symptoms in carers of older relatives: A systematic review and meta-analysis

Positive Caregiving Characteristics as a Mediator of Caregiving Burden and Satisfaction With Life in Caregivers of Older Adults

Health impact of objective burden, subjective burden and positive aspects of caregiving: an observational study among caregivers in Switzerland

Telehealth Benefits and Barriers

Covid-19: adverse mental health outcomes for healthcare workers

The effectiveness of telemental health applications: a review

The unfolding COVID-19 pandemic: A probability-based, nationally representative study of mental health in the United States

Self-labeling as having a mental or physical illness: the effects of stigma and implications for help-seeking

Screening for posttraumatic stress disorder in ARDS survivors: validation of the Impact of Event Scale-6 (IES-6)

Psychiatric distress in family caregivers of patients with COVID -19

The hard work of balancing employment and caregiving: what can canadian employers do to help?

Measuring depression outcome with a brief self-report instrument: sensitivity to change of the Patient Health Questionnaire (PHQ-9

A 4-item measure of depression and anxiety: validation and standardization of the Patient Health Questionnaire-4 (PHQ-4) in the general population

Assistive technologies for ageing populations in six low-income and middle-income countries: a systematic review

The value of maintaining social connections for mental health in older people

Self-identifying as a caregiver: Exploring the positioning process

Pandemic fear" and COVID-19: mental health burden and strategies

Association between recognition and help-seeking preferences and stigma towards people with mental illness

Mental health before and during the COVID-19 pandemic: a longitudinal probability sample survey of the UK population

Mental health responses to the COVID-19 pandemic: a latent class trajectory analysis using longitudinal UK data

Disparities in use of telehealth at the onset of the COVID-19 public health emergency

Differences between caregivers and noncaregivers in psychological health and physical health: a meta-analysis

Employee Sleep Enhancement and Fatigue Reduction Programs: Analysis of the 2017 CDC Workplace Health in America Poll

Disparities In Telehealth Use Among California Patients With Limited English Proficiency

Stigma as a barrier to recognizing personal mental illness and seeking help: a prospective study among untreated persons with mental illness

Bereavement after caregiving

Prevalence of and Risk Factors Associated With Mental Health Symptoms Among the General Population in China During the Coronavirus Disease

Mental Health Disorders Related to COVID-19-Related Deaths

Effects of COVID-19 on College Students' Mental Health in the United States: Interview Survey Study

2021. 6-month neurological and psychiatric outcomes in 236 379 survivors of COVID-19: a retrospective cohort study using electronic health records

Public Policy Institute, 2020. Caregiving in the US 2020 | The National Alliance for Caregiving [WWW Document

Caregiving in the US 2015 | The National Alliance for Caregiving

Younger people are more vulnerable to stress, anxiety and depression during COVID-19 pandemic: A global crosssectional survey

Brain tumors and COVID-19: the patient and caregiver experience

Investigating Mental Health of US College Students During the COVID-19 Pandemic: Cross-Sectional Survey Study

The Effect of Long Working Hours and Overtime on Occupational Health: A Meta-Analysis of Evidence from