Abstract
Background
Pandemic isolation has increased the burden of informal caregiving. Work patterns have changed compared to pre-pandemic levels, with many countries promoting remote work due to government lockdown policies. The importance of support measures necessary to balance caregiving and work has been reaffirmed, as indicated by an increase in psychological distress. This study primarily investigates the factors associated with the increase in caregiving frequency among working family caregivers during the COVID-19 pandemic.
Methods
This cross-sectional study involved a sample of 403 individuals who provided complete data from a fully structured web-based survey of workers. The participants were both employed and providing family care during the April–May 2020 coronavirus pandemic in Japan. Consequently, a binomial logistic regression analysis was conducted.
Results
The effect of work type on the frequency of caregiving during the pandemic was studied. It was 0.97 times odds (p = 0.008) for those who cohabited and used “measures to avoid transfer,” 5.27 times (p = 0.018) higher odds for those who utilized “reduced working hours compared to those who did not,” and 2.46 times (p < 0.001) higher odds for each hour increase in time spent at home during pandemic. Those who lived separately and used “long-term leave” were 17.07 times higher odds compared to those who did not (p = 0.005), used “short-term leave” were 9.60 times higher odds compared to those who did not (p = 0.007), while those who used “reduced working hours” were 7.13 times higher caregiving frequency compared to those who did not (p = 0.013). Regardless of whether they live together or apart, caregivers who believe “It is a filial duty to provide care to parents by oneself as a child” were found to provide care more frequently when taking “Short-term leave” compared to when they did not take it.
Conclusion
Main factors influencing caregiving frequency during the COVID-19 pandemic included the acquisition of “reduced working hours” under flexible work arrangements (FWA), the caregiver's attitude toward caregiving, and living with the cared-for person. Contrary to previous studies, this research demonstrated that the use of FWA did not necessarily alleviate the caregiver's burden, and differences were observed in the increase or decrease in caregiving frequency based on the caregiver's attitudes and living arrangements.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-026-27149-y.
Keywords: Caregiver, COVID-19, Informal care, Pandemic, Work arrangement
Introduction
Currently, the world is aging at a significantly faster rate than in the past. The World Health Organization (WHO) has recognized the need for countries to develop effective health and social systems to address this challenging demographic shift. The aging population poses economic and social consequences in many countries, including labor shortages and increased health and welfare burdens. Japan, Italy, and Germany are among the three countries with the highest aging rates [1], and with this aging population, informal care is becoming increasingly important. Informal care is defined as “private care provided by family, relatives, friends, and neighbors” [2]. Recent reports indicate that informal care has a more balanced gender distribution, although women still play a predominant role, with daughters being the primary sources of informal care in all three countries [3]. Governments in these countries emphasize a strategy favoring informal care and community-based long-term care services over formal care [4, 5].
The prevalence of informal care is closely related to caregivers' work arrangements, particularly work interference with family (WIF) and flexible work arrangements (FWA). Flexible Work arrangement is a system established by the government or companies to facilitate a balance between family and work. For example, studies of older adults who combined paid work and informal care have found that they experienced WIF and that FWAs moderated the relationship between work status preferences and WIF [6].
Regarding the relationship between FWAs and caregiving frequency, those who spend more time caring are more likely to work part-time or in shifts than non-caregivers [7]. Additionally, a study found that the number of weekly caregiving hours is associated with work-related factors such as contract hours and family-work conflict [8]. Research on the impact of informal caregiving on caregivers' subjective well-being has emphasized how caregiving responsibilities can lead to a loss of employment opportunities and a reduction in working hours [9]. This highlights the economic challenges and the potential strain on the subjective well-being of informal working caregivers in Japan due to balancing work and caregiving responsibilities. However, the research on the type of support needed for working family caregivers to remain in the workforce is limited, highlighting the need to find ways to balance work and informal care without leaving the workforce or reducing work hours. Thus, clarifying the association between work and caregiving among informal caregivers may provide important insights for formulating policies regarding work-caregiving balance in an aging society.
There is also a growing need to understand this relationship within the framework of role conflict theories. Role conflict arises when expectations associated with different social roles are incompatible, such that participation in one role becomes more difficult because of participation in another [10]. For working family caregivers, this tension is particularly relevant because employment and caregiving both require time, energy, and psychological commitment. When caregiving demands increase, work responsibilities may become harder to maintain; conversely, when work demands intensify or become less flexible, the time available for caregiving may be reduced. Previous reports have highly valued options such as flexible work scheduling and the ability to work from home for working family caregivers to continue providing informal care [11]. Evidence suggests that older adults who take up caregiving have work preferences that do not match their actual work status [12], and those in involuntary part-time work (preferring full-time work) are more likely to take on caregiving responsibilities. When preferences do not align with actual status or work hours, individuals may experience work status incongruence [13].
Prior studies have demonstrated that a fit between desired and actual work hours reduces work-life conflict [14]. FWAs have been proposed to help individuals manage work and family care commitments [15] and have been a focus of government policy in recent years. However, such arrangements do not necessarily reduce caregiving demands. In some situations, they may instead make it easier for caregivers to reallocate time from paid work to caregiving, particularly when caregiving is already being prioritized within the household.
The COVID-19 pandemic made these dynamics especially salient. During the pandemic, many caregivers experienced simultaneous changes in work conditions, increased time spent at home, and reduced access to formal care services [16, 17]. In Europe, the frequency of informal care for parents increased in most countries during the first wave of the pandemic in the spring/summer of 2020 [16]. This trend was mainly due to the pandemic-led downsizing of care networks and the closure, reduction, or reallocation of services provided by professionals, the pandemic accelerated large-scale transformations in work arrangements, such as the expansion of remote work, flexible schedules, and telecommuting, especially in developed countries [3, 4]. This created a unique context in which work–care role conflict may have been amplified, not only because caregiving needs became more difficult to delegate, but also because the boundaries between paid work and informal care were often disrupted. These changes presented both opportunities and challenges for working caregivers: while FWAs can theoretically alleviate work-family conflict [5, 6], they may also blur the boundaries between work and home, potentially increasing caregiving demands during work hours [7].Examining caregiving frequency during this period therefore provides an opportunity to better understand how employment arrangements were associated with caregivers’ responses to competing role demands.
Thus, the pandemic represents a unique opportunity to empirically examine how changes in employment conditions—such as telework and reduced working hours—interact with caregiving responsibilities.
Although the relationship between caregiving and work has been studied, research on this relationship during a pandemic is limited. In light of these gaps, this study aimed to examine how employment arrangements were associated with changes in “Caregiving frequency (Increase or stable in daily caregiving time)” among working family caregivers during the COVID-19 pandemic in Japan. The primary outcome variable was an increase in caregiving frequency during the COVID-19 pandemic, defined as a self-reported increase in daily caregiving frequency compared with the pre-pandemic period. Using logistic regression analysis, we investigated whether specific employment arrangements, including various types of flexible work arrangements—as well as living arrangements and caregiving attitudes, were independently associated with increased caregiving frequency.
We hypothesized that employment arrangements limiting temporal flexibility or increasing work–care conflict would be associated with a higher likelihood of increased caregiving frequency, whereas work arrangements designed to support work–care balance would be associated with a lower likelihood of caregiving frequency increase during the pandemic.
Method
Study design and subjects
This study was a cross-sectional investigation conducted in Japan from April to May 2020 using an anonymous questionnaire. The data were derived from a survey commissioned by the Non-Profit Organization (NPO) Tonarino Kaigo (Kanagawa, Japan), and the study was conducted by the marketing company Macromill Inc. (Japan). The objective was to derive practical and policy implications for supporting the reconciliation of work and caregiving by empirically examining the relationship between changes in work arrangements during the pandemic and the frequency of caregiving. Participants who met the selection criteria (Supplementary Material 1) were pre-registered with Macromill Inc. and resided in Japan. They were randomly contacted via email, and responses were collected. Participants were divided into four groups based on (1) caregiving experience (current or previous) and (2) whether they had ever left their job due to caregiving responsibilities. The survey closed after 400 responses were collected from each group. To reduce bias in the online survey, participants were recruited to reflect regional distribution. A total of 436 individuals were selected, fulfilling two conditions: (1) they were employed at the time of data collection (including those on leave or in new jobs), and (2) they were informal caregivers. Ultimately, 403 participants were included. The flow chart (Fig. 1) illustrates the sample selection procedure.
Fig. 1.
Flowchart of participant selection. Participants were selected from a nationwide web-based survey targeting residents of Japan who had current or past experience providing family care. Of the 1,600 participants initially screened, 436 met the eligibility criteria of being employed at the time of the survey and currently providing family care. After excluding 33 participants with missing data, 403 participants were included in the final analysis
The certification for long-term care needs measures the level of care services required for the elderly across five areas: direct daily living assistance, indirect daily living assistance, behavioral and psychological symptoms of dementia (BPSD)-related activities, functional training, and medical-related activities. It is classified as “Requiring Support” (levels 1 and 2) or “Requiring Long-Term Care” (levels 1–5, with 5 being the highest). The certification committee for Long-term care needs to evaluate cases based on investigations of mental and physical conditions and physicians’ letters of opinion [18].
Measures
For each person being cared for, data were collected on whether family members lived with the individual, gender, age, whether the person had dementia, certification for long-term care need, (“Requiring Support” (levels 1 and 2) or “Requiring Long-Term Care” (levels 1–5, with 5 being the highest)) and the number of days the person had lived with the caregiver (Questionnaire developed for this study: Supplementary Material 2).
Caregivers were asked about their gender, age, marital status, household income, number of employees in their organization, occupation (International Standard Classification of Occupations (ISCO) [19]), intention to leave employment due to caregiving, and attitudes toward informal care. Details are provided in Supplementary Material 1.
For COVID-19-related variables, we investigated whether caregivers used or had used FWA, whether there was an increase or decrease in the time spent at home per day during the pandemic compared to the pre-pandemic state, and whether there was an increase or decrease in daily informal care time compared to pre-pandemic levels.
Main Outcome- Caregiving Frequency: Participants were asked to report changes in their average daily informal caregiving hours compared with the pre–COVID-19 period. Response options included: reduced by more than 3 h, reduced by 2–3 h, reduced by 1–2 h, reduced by less than 1 h, no change, increased by less than 1 h, increased by 1–2 h, increased by 2–3 h, and increased by more than 3 h. In the present study, this variable was operationalized as the primary outcome variable. Participants were classified into two groups: those reporting no increase or a decrease in caregiving frequency (stable group = 0) and those reporting an increase in caregiving frequency (increased group = 1) during the COVID-19 pandemic.
Flexible Work Arrangements: Participants were asked whether they had used FWA before or after the pandemic. The response options included: long-term leave, short-term leave, reduced working hours, flextime, remote work, limitations on overtime work, and measures to avoid transfer.
The question on “attitudes toward informal care” included three options regarding the use of public long-term care services and the provision of care by family members: “Care for those who need care should be provided by family members, not by public care service providers,” “It is a filial duty to provide care to parents by oneself as a child,” and “If one has problems with care, one should not try to solve them by oneself, but rely on public services and others outside the family”.
Statistical analysis
Caregivers were divided into two groups: Individuals whose daily informal caregiving hours remained the same or decreased during the COVID-19 pandemic (the “stable group”) and those whose daily informal caregiving hours increased during the COVID-19 pandemic (the “increased group”). This dichotomized variable (increase vs. no increase in caregiving frequency) was treated as the primary outcome variable in all subsequent analyses.
The analysis was divided into whole subjects, cohabitation, and separate residences. The following data were analyzed: cohabitation or separate residence, gender, age of the caregiver and care recipient, marital status, occupation, presence of dementia, caregiver's intention to leave work during the COVID-19 pandemic, use of seven types of FWA (long-term leave, short-term leave, Flexible working hours, reduced working hours, remote work, limitations on overtime work, and measures to avoid transfer), and caregiver’s thoughts about informal care.
The analytical procedure consisted of two sequential steps. First, prior to multivariable modeling, bivariate correlation analyses were conducted between the primary outcome variable (increased caregiving frequency during the COVID-19 pandemic) and candidate explanatory variables. Variables that demonstrated statistically significant associations at p < 0.05 were subsequently entered into the multivariable logistic regression models. Binary logistic regression was then performed to estimate adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for factors associated with increased caregiving frequency during the COVID-19 pandemic. Because this was a cross-sectional study, these findings represent statistical associations and suggestive relationships rather than causal effects.
Subsequently, with caregiving hours as the dependent variable (stable group = 0/increased group = 1), we compared Model 1 with basic characteristics (age of the caregiver, gender of the care recipient, marital status of the caregiver, household income, occupation); Model 2 with Model 1 plus the intention to leave work during the COVID-19 pandemic and seven types of FWA; and Model 3 with Model 1 and Model 2 plus daily hours at home during the COVID-19 pandemic as the independent variable. Additionally, three thoughts about informal caregiving were included in the independent variables of Model 3 with caregiving frequency (stable group/increased group) as the dependent variable: Model 1: “Care for those who need care should be provided by family members, not by public care service providers”; Model 2: “It is a filial duty to provide care to parents by oneself as a child”; Model 3: “If one has problems with care, one should not try to solve them by oneself, but rely on public services and others outside the family.” A binomial logistic regression analysis was conducted. The sample size (n = 403) provided an events-per-variable ratio greater than 10, ensuring sufficient statistical power and stable estimation in the logistic regression models.
Ethical considerations
This study was approved by the Ethics Review Committee of Kyoto University Graduate School of Medicine and the Ethics Committee of Kyoto University Hospital, Kyoto, Japan (approval number: R2786-1). The study was conducted in accordance with the tenets of the Declaration of Helsinki and the “Ethical Guidelines for Medical and Health Sciences Research Involving Human Subjects” of the Ministry of Health, Labor and Welfare, Japan. Consent was obtained by having subjects read the terms and conditions regarding ethical considerations (free participation, possibility of withdrawal, protection of personal information, no disadvantage for nonparticipation, etc.) at the time of data collection and by answering a questionnaire. No personally identifiable information was collected during the data collection process.
We followed the STROBE guidelines for reporting cross-sectional studies (STROBE 2007 v4), and a completed STROBE checklist is provided as a separate file.
Results
Basic characteristics and factors related to informal care (Table 1)
Table 1.
Basic characteristics and factors related to informal care
| Whole (n = 403) | Cohabitation ( n = 202) | Living apart ( n = 201) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Caregiving frequency n(%) | Caregiving frequency n(%) | Caregiving frequency n(%) | ||||||||
| Stable | Increased | p value | Stable | Increased | p value | Stable | Increased | p value | ||
| Cohabitation (Yes) | 135(33.4) | 67(16.6) | < 0.001 | |||||||
| Caregiver's sex (Female) | 130(42.9) | 41(41.0) | 0.738 | 44(21.8) | 29(14.4) | 0.136 | 86(42.8) | 12(6.0) | 0.119 | |
| Caregiver | Under 40 years Average 78.0 ± 12.9SD years | 41(10.2) | 33(8.2) | 13(6.4) | 25(12.4) | 28(13.9) | 8(4.0) | |||
| 40–49 years | 72(17.9) | 21(5.2) | 30(14.9) | 14(6.9) | 42(20.9) | 7(3.5) | ||||
| 50–59 years | 121(30.0) | 29(7.2) | 51(25.2) | 16(7.9) | 70(34.8) | 13(6.5) | ||||
| Over 60 years | 69(17.1) | 17(4.2) | < 0.001 | 41(20.3) | 12(5.9) | < 0.001 | 28(13.9) | 5(2.5) | 0.772 | |
| Marital status (Yes) | 187(46.4) | 65(16.1) | 0.556 | 67(33.2) | 40(19.8) | 0.177 | 120(59.7) | 25(12.4) | 0.612 | |
| Occupation | Civil servant | 12(3.0) | 7(1.7) | 6(3.0) | 3(1.5) | 6(3.0) | 4(2.0) | |||
| Employer/board member | 8(2.0) | 3(0.7) | 5(2.5) | 2(1.0) | 3(1.5) | 1(0.5) | ||||
| Company employee (Administrative) | 50(12.4) | 27(6.7) | 22(10.9) | 24(11.9) | 28(13.9) | 3(1.5) | ||||
| Company employee (Technical) | 41(10.2) | 18(4.5) | 16(7.9) | 11(5.5) | 25(12.4) | 7(3.5) | ||||
| Company employee (Others such as service and sales workers) | 58(14.4) | 6(1.5) | 21(10.4) | 3(1.5) | 37(18.4) | 3(1.5) | ||||
| Self-employed/Free enterprise | 42(10.4) | 9(2.2) | 25(12.4) | 4(2.0) | 17(8.5) | 5(2.5) | ||||
| Non-regular employment | 77(19.1) | 19(4.7) | 33(16.3) | 13(6.4) | 44(21.9) | 6(3.0) | ||||
| Other | 15(3.7) | 11(2.7) | 0.003 | 7(3.5) | 7(3.5) | 0.005 | 8(4.0) | 4(2.0) | 0.095 | |
| Care recipient's sex (Female) | 204(50.6) | 56(13.9) | 0.04 | 91(45.0) | 38(18.8) | 0.136 | 113(56.2) | 18(9.0) | 0.161 | |
| Care recipient | Under 70 years | 32(7.9) | 12(3.0) | 16(7.9) | 10(5.0) | 16(8.0) | 2(1.0) | |||
| 70–79 years | 63(15.6) | 26(6.5) | 34(16.8) | 20(9.9) | 29(14.4) | 6(3.0) | ||||
| 80–89 years | 148(36.7) | 38(9.4) | 56(27.7) | 21(10.4) | 92(45.8) | 17(8.5) | ||||
| Over 90 years | 60(14.9) | 24(6.0) | 0.306 | 29(14.4) | 16(7.9) | 0.569 | 31(15.4) | 8(4.0) | 0.822 | |
| Dementia (Yes) | 180(44.7) | 61(15.1) | 0.778 | 84(41.6) | 39(19.3) | 0.582 | 96(47.8) | 22(10.9) | 0.31 | |
| Intention to leave | I don't think I should leave my job at all | 137(34.0) | 23(5.7) | 61(30.2) | 13(6.4) | 76(37.8) | 10(5.0) | |||
| I don't think I should leave my job | 122(30.3) | 44(10.9) | 54(26.7) | 28(13.9) | 68(33.8) | 16(8.0) | ||||
| I should leave my job | 32(7.9) | 22(5.5) | 14(6.9) | 16(7.9) | 18(9.0) | 6(3.0) | ||||
| I really should leave my job | 12(3.0) | 11(2.7) | < 0.001 | 6(3.0) | 10(5.0) | < 0.001 | 6(3.0) | 1(0.5) | 0.366 | |
| FWAs | Long-term leave | 12(3.0) | 24(6.0) | < 0.001 | 8(4.0) | 17(8.4) | < 0.001 | 4(2.0) | 7(3.5) | < 0.001 |
| Short-term leave | 23(5.7) | 34(8.4) | < 0.001 | 8(4.0) | 23(11.4) | < 0.001 | 15(7.5) | 11(5.5) | < 0.001 | |
| Reduced working hours | 28(6.9) | 36(8.9) | < 0.001 | 15(7.4) | 25(12.4) | < 0.001 | 13(6.5) | 11(5.5) | < 0.001 | |
| Flexible working hours | 26(6.5) | 25(6.2) | < 0.001 | 11(5.4) | 17(8.4) | < 0.001 | 15(7.5) | 8(4.0) | 0.012 | |
| Remote work | 15(3.7) | 21(5.2) | < 0.001 | 7(3.5) | 14(6.9) | < 0.001 | 8(4.0) | 7(3.5) | 0.001 | |
| Limitations on overtime work | 11(2.7) | 14(3.5) | < 0.001 | 6(4.4) | 10(5.0) | 0.009 | 5(2.5) | 4(2.0) | 0.02 | |
| Measure to avoid transfer | 15(3.7) | 12(3.0) | 0.014 | 7(3.5) | 6(3.0) | 0.304 | 8(4.0) | 6(3.0) | 0.006 | |
| Thought 1* | 65(16.1) | 37(9.2) | 0.002 | 36(17.8) | 29(14.4) | 0.017 | 29(14.4) | 8(4.0) | 0.344 | |
| Thought 2† | 177(43.9) | 68(16.9) | 0.089 | 86(42.6) | 48(23.8) | 0.261 | 91(45.3) | 20(10.0) | 0.496 | |
| Thought 3§ | 281(69.7) | 85(21.1) | 0.020 | 123(60.9) | 57(28.2) | 0.195 | 158(78.6) | 28(13.9) | 0.066 | |
Dependent variable: Increase or decrease in daily caregiving frequency during the COVID-19 pandemic (Stable group: Increase or decrease in daily caregiving frequency since the pandemic ≤ 0, Increased group: Increase or decrease in daily caregiving frequency since the pandemic > 0)
* Care for those who need care should be provided by family members, not by public care service providers
† It is a filial duty to provide care to parents by oneself as a child
§ If one has problems with care, one should not try to solve them by oneself but rely on public services and others outside the family
In total, 403 working informal caregivers were included in the study. Of these, 28 participants whose certification for long-term care needs were unknown, and five participants whose caregivers' dementia status was unknown, were excluded.
The mean age of the caregivers was 47.3 ± 10.2 years, and that of the care recipients was 78.0 ± 12.9 years. Among them, 303 (75.2%) were in the stable group with no increase in caregiving frequency due to the COVID-19 pandemic, while 100 (24.8%) were in the increased caregiving frequency group. To identify factors that increase caregiving frequecy, we first extracted the characteristics of the stable group.
There was a significant tendency in the stable group for caregivers to be older than 40 years (p < 0.001) and for care recipients to be female (p = 0.040), compared to the increased group. Significant differences were also found between the two groups regarding their intention to leave their jobs and attitudes toward caregiving. A significantly higher proportion of caregivers in the stable group answered, “I don't feel I should leave my job at all” and “I don't feel I should leave my job” (34.0%, 30.3%, p < 0.001) during the COVID-19 pandemic compared to the increased group. Regarding informal care, significantly more caregivers in the stable group (69.7%, p = 0.020) thought that “If one has problems with care, one should not try to solve them by oneself, but rely on public services and others outside the family” (Thought 3), compared to the increased group. Conversely, more caregivers in the increased group took long-term leave (p < 0.001), short-term leave (p < 0.001), reduced working hours (p < 0.001), flextime (p < 0.001), remote work (p < 0.001), and limited overtime (p < 0.001).
Next, to investigate whether living with the care recipient was associated with an increase in caregiving frequency, we divided caregivers into two categories: those who live with the care recipient and those who do not. Among caregivers living with the care recipient, 135 (66.8%) were in the stable group and 67 (33.2%) in the increased group. Significantly more caregivers in the stable group were 40 years or older (p < 0.001) and indicated, “I don't feel I should leave my job at all” or “I don't feel I should leave my job” (30.2%, 26.7%, p < 0.001) regarding their intention to leave their jobs during the COVID-19 pandemic, compared to the increased group. Conversely, significantly more caregivers in the increased group (17.8%, p = 0.017) believed that “Care for those who need care should be provided by family members, not by public care service providers” (Thought 1) compared to the stable group. Additionally, the increased group was significantly more likely to adopt the following FWA: long-term leave (p < 0.001), short-term leave (p < 0.001), reduced working hours (p < 0.001), flextime (p < 0.001), remote work (p < 0.001), and limitations on overtime work (p < 0.009).
Among caregivers who lived apart from their care recipients, 168 (83.6%) were in the stable group, and 33 (16.4%) were in the increased group. The stable group was significantly more likely than the increased group to acquire the following FWAs: short-term leave (p < 0.001), reduced working hours (p < 0.001), flextime (p = 0.012), remote work (p = 0.001), limitations on overtime work (p = 0.020), and measures to avoid transfer (p = 0.006). In contrast, significantly more caregivers in the increased group took long-term leave (p < 0.001). Factors significant for cohabitation, such as age, caregiver's intention to leave work, and attitudes toward caregiving, were not significant for those living separately.
The results indicate that factors influencing the increase or decrease in caregiving frequency during the COVID-19 pandemic differ depending on whether the caregiver lives with or apart from the care recipient. Therefore, we considered whether living with the care recipient was related to caregiving frequency during the COVID-19 pandemic. Consequently, we conducted a binomial logistic regression analysis separately for cohabitation and separate residence.
We assessed the impact of acquiring FWAs, the intention to leave work, the length of time spent at home, and attitudes toward informal care on the time spent providing care in the case of cohabitation (Tables 2 and 4)
Table 2.
The impact of the acquisition of FWA, the intention to leave, the length of time spent at home, and attitudes toward informal care on the time spent providing care (In the case of cohabitation)
| Dependent variable: Increase or decrease in daily caregiving frequency in COVID-19pandemic | Cohabitation | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Model 1* | Model 2† | Model 3§ | ||||||||
| Odds Ratio | 95%CI | p value | Odds Ratio | 95%CI | p value | Odds Ratio | 95%CI | p value | ||
| Caregiver's age | 0.77 | 0.66/0.92 | 0.003 | 0.81 | 0.67/0.99 | 0.037 | 0.80 | 0.63/1.02 | 0.070 | |
| Family income | 1.13 | 0.94/1.36 | 0.178 | 1.08 | 0.86/1.34 | 0.512 | 1.02 | 0.79/1.33 | 0.863 | |
| Occupation | Civil servant | 1(Reference) | 1(Reference) | 1(Reference) | ||||||
| Employer/Board member | 0.61 | 0.06/6.07 | 0.671 | 0.57 | 0.04/7.31 | 0.663 | 0.59 | 0.04/8.70 | 0.700 | |
| Company employees (Administrative) | 1.85 | 0.36/9.38 | 0.458 | 0.73 | 0.10/5.20 | 0.755 | 0.17 | 0.02/1.55 | 0.117 | |
| Company employees (Technical) | 0.84 | 0.15/4.81 | 0.845 | 0.23 | 0.03/2.07 | 0.190 | 0.11 | 0.01/1.34 | 0.084 | |
| Company employees (Others such as service and sales workers) | 0.21 | 0.03/1.50 | 0.120 | 0.09 | 0.01/0.92 | 0.042 | 0.03 | 0.00/0.40 | 0.008 | |
| Self-employed/Free enterprise | 0.42 | 0.07/2.63 | 0.352 | 0.21 | 0.02/1.84 | 0.158 | 0.10 | 0.01/1.02 | 0.052 | |
| Non-regular employment | 0.68 | 0.13/3.69 | 0.655 | 0.34 | 0.04/2.52 | 0.289 | 0.09 | 0.01/0.87 | 0.037 | |
| Other | 1.43 | 0.20/10.06 | 0.722 | 0.96 | 0.11/8.76 | 0.973 | 0.47 | 0.04/6.13 | 0.563 | |
| Care recipient's sex (Male vs Female) | 0.85 | 0.41/1.78 | 0.666 | 0.86 | 0.35/2.08 | 0.736 | 0.70 | 0.24/2.01 | 0.508 | |
| Intention to leave | I don't feel I should leave my job at all | 1(Reference) | 1(Reference) | |||||||
| I don't feel I should leave my job | 2.27 | 0.88/5.87 | 0.089 | 3.00 | 0.96/9.38 | 0.058 | ||||
| I feel I should leave my job | 2.05 | 0.52/8.04 | 0.304 | 1.52 | 0.31/7.59 | 0.608 | ||||
| I feel I really should leave my job | 6.13 | 1.23/30.48 | 0.027 | 5.78 | 0.99/33.72 | 0.051 | ||||
| FWA | Long-term leave (No vs Yes) | 1.21 | 0.29/5.03 | 0.789 | 3.12 | 0.60/16.26 | 0.177 | |||
| Short-term leave (No vs Yes) | 3.20 | 0.95/10.70 | 0.059 | 3.16 | 0.82/12.27 | 0.096 | ||||
| Reduced working hours (No vs Yes) | 2.90 | 1.02/8.26 | 0.047 | 5.27 | 1.33/20.85 | 0.018 | ||||
| Flexible working hours (No vs Yes) | 3.35 | 0.85/13.23 | 0.084 | 3.53 | 0.80/15.53 | 0.095 | ||||
| Remote work (No vs Yes) | 1.75 | 0.40/7.70 | 0.457 | 2.23 | 0.32/15.42 | 0.417 | ||||
| Limitations on overtime work (No vs Yes) | 0.65 | 0.12/3.52 | 0.619 | 0.21 | 0.03/1.47 | 0.115 | ||||
| Measure to avoid transfer (No vs Yes) | 0.12 | 0.01/1.21 | 0.072 | 0.03 | 0.00/0.39 | 0.008 | ||||
| Length of time spent at home/1 day | 2.46 | 1.75/3.47 | < 0.001 | |||||||
95%CI 95% Confidence Interval
Dependent variable: Increase or decrease in daily caregiving frequency in COVID-19pandemic (Stable group: Increase or decrease in daily caregiving frequency in the COVID-19 pandemic ≤ 0, Increased group: Increase or decrease in daily caregiving frequency in the COVID-19 pandemic > 0)
*Basic characteristics
†Model1 + Intention to leave(I don't think I should leave my job at all/I don't think I should leave my job/I should leave my job/I really should leave my job dummy variable)and FWA(Long-term leave/Short-term leave/Reduced working hours/Flexible working hours/Remote work/Limitations on overtime work/Measure to avoid transfer; Dummy variable)
§ Model 2 + Length of time spent at home/1 day(in COVID-19 pandemic)
Table 4.
Effects of attitudes toward informal care on FWA, intentions to leave job, and caregiving frequency
| Model 1 | Model 2 | Model 3 | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Cohabitation | ¶AOR | 95%CI | p value | ¶AOR | 95%CI | p value | ¶AOR | 95%CI | p value | |
| Intention to leave | I don't feel I should leave my job at all | 1(Reference) | 1(Reference) | 1(Reference) | ||||||
| I don't feel I should leave my job | 2.28 | 0.88/5.89 | 0.089 | 2.24 | 0.87/5.79 | 0.094 | 2.00 | 0.76/5.28 | 0.161 | |
| I feel I should leave my job | 2.05 | 0.52/8.05 | 0.303 | 2.02 | 0.51/7.99 | 0.318 | 1.93 | 0.49/7.66 | 0.351 | |
| I feel I really should leave my job | 6.17 | 1.23/30.90 | 0.027 | 6.29 | 1.26/31.38 | 0.025 | 6.73 | 1.34/33.76 | 0.020 | |
| FWA | Long-term leave (No/Yes) | 1.21 | 0.29/5.03 | 0.792 | 1.30 | 0.31/5.36 | 0.720 | 1.14 | 0.27/4.71 | 0.861 |
| Short-term leave (No/Yes) | 3.18 | 0.94/10.72 | 0.062 | 3.35 | 1.01/11.18 | 0.049 | 3.12 | 0.91/10.66 | 0.070 | |
| Reduced working hours (No/Yes) | 2.92 | 1.00/8.52 | 0.049 | 2.74 | 0.96/7.80 | 0.059 | 2.56 | 0.88/7.47 | 0.084 | |
| Flexible working hours (No/Yes) | 3.38 | 0.85/13.49 | 0.085 | 3.05 | 0.77/12.08 | 0.113 | 3.56 | 0.89/14.23 | 0.072 | |
| Remote work (No/Yes) | 1.76 | 0.40/7.71 | 0.456 | 1.87 | 0.42/8.36 | 0.411 | 1.73 | 0.39/7.58 | 0.468 | |
| Limitations on overtime work (No/Yes) | 0.65 | 0.12/3.54 | 0.621 | 0.65 | 0.12/3.50 | 0.616 | 0.66 | 0.12/3.57 | 0.633 | |
| Measure to avoid transfer (No/Yes) | 0.12 | 0.01/1.21 | 0.072 | 0.12 | 0.01/1.27 | 0.079 | 0.13 | 0.01/1.33 | 0.086 | |
| Thought 1* | 0.98 | 0.55/1.73 | 0.934 | |||||||
| Thought2† | 1.40 | 0.80/2.48 | 0.241 | |||||||
| Thought3§ | 0.67 | 0.35/1.28 | 0.225 | |||||||
| Model 1 | Model 2 | Model 3 | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Living Apart | ¶AOR | 95%CI | p value | ¶AOR | 95%CI | p value | ¶AOR | 95%CI | p value | |
| Intention to leave | I don't feel I should leave my job at all | 1(Reference) | 1(Reference) | 1(Reference) | ||||||
| I don't feel I should leave my job | 5.47 | 1.22/24.61 | 0.027 | 6.43 | 1.34/30.97 | 0.020 | 4.84 | 1.06/22.16 | 0.042 | |
| I feel I should leave my job | 20.88 | 2.69/161.85 | 0.004 | 40.01 | 4.45/359.37 | 0.001 | 19.58 | 2.79/137.57 | 0.003 | |
| I feel I really should leave my job | 0.71 | 0.01/61.38 | 0.880 | 1.55 | 0.02/112.77 | 0.841 | 0.78 | 0.01/65.76 | 0.912 | |
| FWA | Long-term leave (No/Yes) | 16.82 | 2.32/121.99 | 0.005 | 13.05 | 1.83/93.22 | 0.010 | 14.17 | 1.83/109.80 | 0.011 |
| Short-term leave (No/Yes) | 11.34 | 2.23/57.81 | 0.003 | 15.47 | 2.86/83.75 | 0.001 | 11.10 | 2.26/54.54 | 0.003 | |
| Reduced working hours (No/Yes) | 6.46 | 1.06/39.53 | 0.043 | 6.13 | 0.97/38.89 | 0.054 | 6.00 | 0.97/37.07 | 0.054 | |
| Flexible working hours (No/Yes) | 0.29 | 0.04/2.19 | 0.229 | 0.16 | 0.02/1.43 | 0.102 | 0.29 | 0.04/2.07 | 0.216 | |
| Remote work (No/Yes) | 1.96 | 0.27/14.05 | 0.503 | 3.11 | 0.43/22.55 | 0.261 | 2.03 | 0.28/14.54 | 0.480 | |
| Limitations on overtime work (No/Yes) | 0.03 | 0.001/0.82 | 0.037 | 0.05 | 0.002/1.35 | 0.076 | 0.04 | 0.002/1.00 | 0.050 | |
| Measure to avoid transfer (No/Yes) | 6.31 | 0.74/53.82 | 0.092 | 7.10 | 0.82/61.35 | 0.075 | 6.77 | 0.81/56.35 | 0.077 | |
| Thought 1* | 0.99 | 0.42/2.30 | 0.980 | 0.45 | ||||||
| Thought 2† | 0.17/1.16 | 0.099 | ||||||||
| Thought 3§ | 0.78 | 0.34/1.76 | 0.544 |
AOR Adjusted Odds Ratio
95% CI 95% Confidence Interval
Dependent variable: increase or decrease in daily caregiving frequency during the COVID-19 pandemic (Stable group: increase or decrease in daily caregiving frequency during the pandemic ≤ 0, Increased group: increase or decrease in daily caregiving frequency during the pandemic > 0)
¶ Caregiver's sex, age (under 40 years, 40–49, 50–59, over 60 years dummy variable), marital status, occupation (Civil Servant, Employer/Board member, Company employee (Administrative), Company employee (Technical), Company employee (Others such as service and sales workers), Self-employed/Free enterprise, Non-regular employment, other dummy variable) were adjusted
*Care for those who need care should be provided by family members, not by public care service providers
† It is a filial duty to provide care to parents by oneself as a child
§ If one has problems with care, one should not try to solve them by oneself but rely on public services and others outside the family
Here, we describe the results of the binomial logistic regression analysis of the factors influencing the length of caregiving frequency in the case of cohabitation. Model 1 presents the basic characteristics, while Model 2 investigates the impact of the use of FWA and caregivers’ intention to leave work on the increase or decrease in time spent providing care during the COVID-19 pandemic. Model 3 further investigates the impact of “time spent at home” on caregiving frequency, in addition to the factors analyzed in Model 2.
For the items significantly associated with caregiving frequency, the odds ratio for caregiving frequency relative to caregivers’ age in Model 1 was OR = 0.77 (95% CI: 0.66/0.92, p = 0.003). In Model 2, those who answered, “I feel I really should leave my job” had 6.13 times higher odds (95% CI: 1.23/30.48, p = 0.027) compared to those who responded, “I don't feel I should leave my job at all.” In Model 3, caregivers who received the “measure to avoid transfer” had lower odds than those who did not receive it (OR = 0.03, 95% CI: 0.00/0.39, p = 0.008), while caregivers who received “reduced working hours” had significantly higher odds compared to those who did not (OR = 5.27, 95% CI: 1.33/20.85, p = 0.018). Odds were 2.46 times higher for an increase of one hour per day spent at home during the COVID-19 pandemic (95% CI: 1.75/3.47, p < 0.001).
Moreover, we examined the effects of caregivers’ attitudes toward informal caregiving in Model 2 of Table 2 (the relationship between FWAs, caregivers' intention to leave work, and the increase or decrease in daily caregiving frequency during the COVID-19 pandemic).
Table 4 added “Caregiver’s attitudes toward caregiving (Thoughts 1–3)” to Table 2, model 2, and none of these thoughts indicated a significant association with caregiving frequency, and the effects of FWAs and intention to leave maintained the same trend as in Table 2. Among caregivers who believe “Care for those who need care should be provided by family members, not by public care service providers,” those who took “shortened work hours” had 2.92 (95% CI: 1.00/8.52, p = 0.049) times more caregiving frequency than those who did not take it. Additionally, among caregivers who believe “It is a filial duty to provide care to parents by oneself as a child,” those who took short-term leave spent 3.35 times (95% CI: 1.01/11.18, p = 0.049) more caregiving frequency than those who did not take it.
The impact of acquiring FWAs, the intention to leave work, the length of time spent at home, and attitudes toward informal care on the time spent providing care in the case of living separately (Tables 3 and 4)
Table 3.
The impact of the acquisition of FWA, the intention to leave work, the length of time spent at home, and attitudes toward informal care on the time spent providing care (In the case of living apart)
| Dependent variable: increase or decrease in daily caregiving frequency in COVID-19pandemic | Living Apart | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Model 1* | Model 2† | Model 3§ | ||||||||
| Odds Ratio | 95%CI | p value | Odds Ratio | 95%CI | p value | Odds Ratio | 95%CI | p value | ||
| Caregiver's age | 0.89 | 0.72/1.12 | 0.327 | 1.08 | 0.77/1.50 | 0.668 | 1.10 | 0.78/1.56 | 0.569 | |
| Family income | 1.00 | 0.79/1.28 | 0.983 | 0.78 | 0.56/1.10 | 0.153 | 0.77 | 0.55/1.09 | 0.136 | |
| Occupation | Civil servant | 1(Reference) | 1(Reference) | 1(Reference) | 0.111 | |||||
| Employer/Board member | 0.70 | 0.05/9.83 | 0.791 | 1.63 | 0.06/42.55 | 0.771 | 2.28 | 0.08/63.07 | 0.628 | |
| Company employees (Administrative) | 0.20 | 0.03/1.19 | 0.077 | 0.14 | 0.01/1.49 | 0.104 | 0.16 | 0.02/1.61 | 0.120 | |
| Company employees (Technical) | 0.36 | 0.07/1.76 | 0.207 | 0.27 | 0.04/2.04 | 0.204 | 0.30 | 0.04/2.22 | 0.237 | |
| Company employees (Others such as service and sales workers) | 0.04 | 0.004/0.47 | 0.010 | 0.004 | 0.0001/0.20 | 0.006 | 0.00 | 0.0001/0.21 | 0.007 | |
| Self-employed/Free enterprise | 0.53 | 0.10/2.83 | 0.460 | 0.70 | 0.08/5.91 | 0.742 | 0.66 | 0.08/5.40 | 0.695 | |
| Non-regular employment | 0.13 | 0.02/0.74 | 0.021 | 0.11 | 0.01/1.05 | 0.055 | 0.11 | 0.01/0.98 | 0.048 | |
| Other | 0.70 | 0.11/4.60 | 0.710 | 0.70 | 0.05/10.17 | 0.796 | 0.83 | 0.06/11.60 | 0.892 | |
| Care recipient's sex (Male vs Female) | 0.53 | 0.21/1.32 | 0.173 | 0.66 | 0.18/2.40 | 0.530 | 0.67 | 0.18/2.44 | 0.539 | |
| Intention to leave | I don't feel I should leave my job at all | 1(Reference) | 1(Reference) | |||||||
| I don't feel I should leave my job | 5.45 | 1.24/23.88 | 0.024 | 5.74 | 1.29/25.52 | 0.022 | ||||
| I feel I should leave my job | 20.70 | 2.98/144.00 | 0.002 | 20.42 | 2.94/141.73 | 0.002 | ||||
| I feel I really should leave my job | 0.70 | 0.01/54.50 | 0.872 | 0.57 | 0.01/61.93 | 0.816 | ||||
| FWA | Long-term leave (No vs Yes) | 16.79 | 2.33/121.14 | 0.005 | 17.07 | 2.35/124.12 | 0.005 | |||
| Short-term leave (No vs Yes) | 11.30 | 2.28/56.09 | 0.003 | 9.60 | 1.85/49.84 | 0.007 | ||||
| Reduced working hours (No vs Yes) | 6.47 | 1.06/39.50 | 0.043 | 7.13 | 1.12/45.36 | 0.038 | ||||
| Flexible working hours (No vs Yes) | 0.29 | 0.04/2.08 | 0.218 | 0.29 | 0.04/2.17 | 0.228 | ||||
| Remote work (No vs Yes) | 1.96 | 0.28/13.87 | 0.502 | 1.56 | 0.19/12.46 | 0.676 | ||||
| Limitations on overtime work (No vs Yes) | 0.03 | 0.001/0.81 | 0.037 | 0.04 | 0.002/1.19 | 0.063 | ||||
| Measure to avoid transfer (No vs Yes) | 6.35 | 0.79/51.18 | 0.083 | 6.67 | 0.76/58.50 | 0.087 | ||||
| Length of time spent at home/1 day | 1.19 | 0.83/1.17 | 0.350 | |||||||
95%CI 95% Confidence Interval
Dependent variable: Increase or decrease in daily caregiving frequency in COVID-19pandemic (Stable group: Increase or decrease in daily caregiving frequency in the COVID-19 pandemic ≤ 0, Increased group: Increase or decrease in daily caregiving frequency in the COVID-19 pandemic > 0)
*Basic characteristics
†Model1 + Intention to leave(I don't think I should leave my job at all/I don't think I should leave my job/I should leave my job/I really should leave my job dummy variable)and FWA(Long-term leave/Short-term leave/Reduced working hours/Flexible working hours/Remote work/Limitations on overtime work/Measure to avoid transfer; Dummy variable)
§ Model 2 + Length of time spent at home/1 day(in COVID-19 pandemic)
The same study was conducted as in the case of cohabitation.
In Model 2, concerning their intention to leave jobs, caregivers who answered “I feel I should leave my job” had 20.70 times higher odds (95% CI: 2.98/144.00, p = 0.002), and those who answered “I don't feel I should leave my job” had 5.45 times higher odds compared to those who responded, “I don't feel I should leave my job at all” (95% CI: 1.24/23.88, p = 0.024) regarding informal caregiving frequency. Caregivers who took “short-term leave” had 11.30 times higher odds (95% CI: 2.28/56.09, p = 0.003), those who took “long-term leave” had 16.79 times higher odds (95% CI: 2.33/121.14, p = 0.005), and those who opted for “reduced working hours” had 6.47 times higher odds (95% CI: 1.06/39.50, p = 0.043) compared to those who did not take these leaves, respectively. Caregivers who chose the “limitations on overtime work” option had significantly lower odds than those who did not (OR = 0.03, 95% CI: 0.001/0.81, p = 0.037). The effect of the COVID-19 pandemic on time spent at home was examined in Model 3, where the relationship between “limitations on overtime work” and caregiving frequency became insignificant.
Moreover, we examined the effects of caregivers’ attitudes on informal caregiving in Table 4. Similar to the result of cohabitation, none of these factors showed a significant association with caregiving frequency. Also, the effects of FWAs and intention to leave maintained the same trend as in Table 3. Caregivers who believe “It is a filial duty to provide care to parents by oneself as a child” and caregivers who believe “If one has problems with care, one should not try to solve them by oneself but rely on public services and others outside the family” showed no increase in caregiving frequency even when taking “Reduced working hours” (p = 0.054, p = 0.054).
Furthermore, caregivers holding either of these two beliefs who took “long-term leave” had care frequencies 13.05(95% CI: 1.83/93.22, p = 0.010) times and 14.17 times (95% CI: 1.83/109.80, p = 0.011) higher, respectively, compared to those who did not take it. However, compared to the results in Table 3 (which did not consider “Thought”), this showed a suppression of the increase in caregiving frequency.
Conversely, caregivers who believed “It is a filial duty to provide care to parents by oneself as a child” and took “short-term leave” had a 15.47 times higher caregiving frequency compared to those who did not take leave. This increase in caregiving frequency was confirmed when compared to the results in Table 3, which did not consider “caregivers' thoughts on informal caregiving.”
Discussion
The findings of this study suggest that changes in caregiving frequency among working family caregivers during the COVID-19 pandemic were shaped not by a single factor, but by the interaction between work arrangements, living circumstances, and caregivers’ views on family responsibility. In particular, the results indicate that flexible work arrangements did not operate uniformly as protective measures.
Despite utilizing “reduced working hours” through FWA, many caregivers reported an increase in caregiving hours. Arksey [20] and Bernard et al. [21] found that employed caregivers often did not benefit from reduced working hours or adjustments to their work schedules. Spann et al. suggest that FWA, such as working from home, may lead to more work interruptions, as caregivers can become distracted or feel guilty for not giving more focused attention to the care recipient [22].
These previous findings are consistent with the interpretation of our results within the framework of role conflict theory. Role conflict arises when the demands of different social roles become difficult to reconcile because they compete for the same limited time, energy, and psychological resources [10]. This study suggests that reduced working hours may not necessarily be associated with a decrease in caregiving frequency; rather, increased flexibility in working hours may have led to an increase in caregiving frequency.
Additionally, this study found that caregivers who lived with their care recipients increased their caregiving frequency more than those who lived apart. Specifically, caregiving frequency increased by 2.55 times for each additional hour per day that the caregiver spent with the recipient. These results suggest that increased time spent with the care recipient correlates with increased caregiving frequency during the pandemic.
This pattern is important because it suggests that cohabitation and increased time spent at home were not merely background characteristics, but contextual conditions that likely intensified work–care role overlap during the pandemic. COVID-19 disrupted formal care services and narrowed opportunities to rely on support outside the household [16]. Under such circumstances, caregivers who lived with care recipients and spent more time at home were likely exposed more continuously to care needs, making it more difficult to preserve boundaries between work and caregiving. In this sense, increased caregiving frequency may reflect not only more available time, but also a greater immediacy of care demands within daily life.
The factors influencing caregiving frequency varied depending on whether the caregiver and care recipient cohabited.
Whether caregivers lived with or apart from the care recipient, utilization of the “shortened work hours” option within the flexible work arrangement was significantly associated with increased caregiving frequency. For those living apart, “long-term leave” and “short-term leave” were additionally associated with increased caregiving frequency. Rather than viewing these findings as isolated associations, a broader pattern emerges: employment arrangements that increased caregivers’ time or role salience within the home were more likely to be associated with increased caregiving frequency, whereas arrangements that preserved work continuity or reduced role disruption were more likely to be associated with stable caregiving frequency. Although previous studies have rarely examined the effects of specific types of flexible work arrangements on caregiving frequency, our results suggest that different forms of flexibility should not be treated as equivalent. Their practical meaning depends on whether they help caregivers manage competing role demands or instead shift the balance toward informal care.
Regardless of whether they live together or apart, caregivers who believe “It is a filial duty to provide care to parents by oneself as a child” were found to provide care more frequently when taking “Short-term leave” compared to when they did not take it. On the other hand, it was inferred that the use of other FWAs was more strongly related to changes in time spent at home during the pandemic than to caregivers' attitudes toward caregiving. This finding suggests that caregivers’ attitudes did not simply coexist with employment conditions but shaped how newly available time was used. For caregivers with a strong sense of filial obligation, short-term leave may have created an immediate opportunity to prioritize caregiving over work. This interpretation is compatible with role conflict theory, which proposes that when competing role demands intensify, individuals allocate their limited resources according to both situational demands and role values [10]. In this study, short-term leave appears to have functioned not only as a work adjustment, but also as a trigger that activated pre-existing caregiving norms.
This study reveals several factors that increase or decrease caregiving frequency for working informal caregivers. One of these factors is the use of FWA, a system designed to balance caregiving and work. Such arrangements are generally believed to contribute to a decrease in caregiving frequency [6, 11]. However, we found no studies that examine which type of FWA use affects the frequency of care. Surprisingly, the results of this study suggest that the use of FWA does not necessarily lead to a reduction in caregiver frequency. Our findings therefore add an important nuance to the existing literature. Previous studies have generally evaluated flexible work arrangements in terms of work–family conflict or perceived balance [6, 12, 15], but fewer studies have examined whether different arrangements actually change caregiving behavior itself. The present results suggest that flexibility should not automatically be interpreted as reducing burden. In the context of a pandemic, flexibility may instead be absorbed into caregiving, especially when formal support is constrained and care needs must be met within the household.
In this study, caregiving frequency decreased when the measure to avoid transfer was used in cohabitation. Although we could not find similar reports from previous studies, this measure may have allowed caregivers to avoid situations requiring adaptation to new circumstances. Additionally, they may have been able to continue utilizing the support networks and resources they had been using in the community. Conversely, in the case of living separately, caregiving frequency decreased when “limitations on overtime work” were implemented. Although no similar previous studies were found, Nadash et al. reported that working in an organization with a flexible culture reduced psychological distress partly through work-family conflict [23], and limiting overtime may be related to a supportive workplace climate. These findings suggest that some work arrangements may be effective not because they increase flexibility in a general sense, but because they reduce instability or help preserve manageable role boundaries. Avoiding transfer may allow caregivers to maintain established support networks and familiar daily routines, while limiting overtime may prevent work from expanding into time that would otherwise be needed for coordination of care. This interpretation is also consistent with the argument that workplace culture and perceived support are central to how caregivers experience the compatibility between work and family responsibilities [23, 24]. Accordingly, what matters may not be the formal existence of a flexible arrangement alone, but whether it is embedded in a work environment that genuinely reduces role conflict. In practice, this means that schedule adjustments should be accompanied by realistic workload expectations, organizational understanding, and access to support outside the family.
During the COVID-19 pandemic, many forms of flexible work arrangements (FWA) beyond short-term leave—such as reduced working hours—changes in time spent at home had a greater impact on caregiving frequency than caregivers' personal values. When spending more time at home, caregivers were more likely to notice additional care needs and had an environment enabling immediate caregiving actions. Thus, increased time at home is presumed to have been a factor driving up caregiving frequency. This aligns with multiple international studies showing increased informal care hours among remote workers during the pandemic [17].
This study revealed that caregivers who believe “It is a filial duty to provide care to parents by oneself as a child,” regardless of whether they live together or apart, increased their caregiving frequency when taking short-term leave compared to those who did not take leave. This result can be interpreted within the framework of role conflict theories [10], which arise when balancing caregiving and work roles become difficult. For caregivers with a strong sense of responsibility that “I should be the one providing care,” short-term leave likely functioned as an “opportunity to prioritize caregiving actions,” resulting in increased caregiving frequency. This finding underscores the importance of the COVID-19 context itself. The pandemic altered not only work arrangements, but also the spatial and temporal organization of everyday life. Increased time at home may therefore have operated as a practical mechanism linking pandemic-related workplace change to greater caregiving involvement. In this respect, the pandemic provided a setting in which work and caregiving became more tightly interwoven, making role conflict more visible in daily behavior.
Overall, the underlying pattern observed in this study is that employment arrangements, cohabitation, and caregiving values interacted in ways that either intensified or buffered work–care role conflict during the pandemic. Flexible work arrangements were not inherently beneficial or harmful; rather, their meaning depended on whether they helped caregivers maintain workable boundaries between roles or instead increased their practical and psychological engagement in caregiving. This interpretation helps explain why flexibility did not uniformly reduce caregiving frequency and why the effects differ according to living arrangement and caregiving attitudes. This study provides insights like the above, though its cross-sectional design limits causal interpretation.
The strength of this study lies in our ability to collect data on working family caregivers during the COVID-19 pandemic from all regions of Japan. We clarified the influence of FWA acquisition status, caregivers' intention to leave their jobs, and their attitudes toward caregiving—factors that have rarely been reported—on the burden of caregiving. From a policy perspective, expanding FWA alone is insufficient. Comprehensive support systems, including boundary management assistance and psychological support, are required to enable caregivers to appropriately balance the demands of their work roles and caregiving roles. Healthcare professionals must consider the characteristics of individual caregivers and provide appropriate support.
Limitation
This study offers insights into factors associated with increased caregiving frequency, although its cross-sectional design limits causal interpretation. Although logistic regression models were used to estimate adjusted associations, these estimates do not establish temporal directionality or causality; due to the cross-sectional nature of the data, causation remains unclear. Therefore, longitudinal studies are needed to clarify the temporal direction and causal pathways between caregiving frequency and employment-related factors. Also, Data were collected by randomly soliciting individuals who met the criteria, with participation concluding upon reaching the target number. Consequently, the response rate could not be calculated due to the unknown denominator. Nevertheless, since the survey was conducted nationwide by a reputable marketing company, sampling bias is considered low. While living arrangements may influence caregiving frequency, causality cannot be assessed within this design. Future longitudinal research is necessary to clarify the relationship between living arrangements and caregiving frequency. These limitations highlight the need for further study and verification.
Conclusions
Factors influencing caregiving frequency during the COVID-19 pandemic included the caregiver's attitude toward caregiving, the acquisition of especially “reduced working hours” under FWA, and cohabitation with the cared-for person. Contrary to previous studies, this research indicated that the use of FWA did not necessarily alleviate the caregiver's burden. Differences in the increase or decrease in caregiving frequency were observed based on the caregiver's attitude and living arrangements. These findings highlight the importance of implementing support measures to balance caregiving and work.
Supplementary Information
Acknowledgements
We would like to express our sincere gratitude to all those who contributed to the successful completion of this paper. Special thanks go to Ayae Kinoshita and Satoko Yoneyama for their invaluable guidance and expertise throughout the research process. We are also grateful to Jun Kawauchi and Daiki Takahashi, NPO Tonarino Kaigo for providing the resources.
Finally, we extend our thanks to the anonymous reviewers for their constructive feedback which helped enhance the quality of this publication.
Abbreviations
- COVID-19
Coronavirus Disease 2019
- FWA
Flexible Work Arrangements
- WIF
Work Interference with Family
Authors’ contributions
MH: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Writing – original draft, review & editing SY: Data curation, Formal analysis, Methodology, Supervision, Writing – review & editing AK: Funding acquisition, Methodology, Project administration, Supervision, Writing – review & editing All authors read and approved the final manuscript.
Funding
This work was supported by Management Expense Grant under Grant No. 021515, and Collaborative Research Projects at the Institute for the Future of Human Society for the Academic Year 2024 under Grant No. 022591.
Data availability
The data that support the findings of this study are restricted by legal or ethical reasons and are not publicly available. Data are, however, available from the authors upon reasonable request and with permission of relevant ethics committee or legal authority.
Declarations
Ethics approval and consent to participate
This study was approved by the Ethics Review Committee of Kyoto University Graduate School of Medicine and the Ethics Committee of Kyoto University Hospital, Kyoto, Japan (approval number: R2786-1). Informed consent was obtained from all individual participants included in the study.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Tooth L, Russell A, Lucke J, Byrne G, Lee C, Wilson A, et al. Impact of cognitive and physical impairment on carer burden and quality of life. Qual Life Res. 2008;17(2):267–73. 10.1007/s11136-007-9300-7. [DOI] [PubMed] [Google Scholar]
- 2.Zwar L, König HH, Hajek A. Psychosocial consequences of transitioning into informal caregiving in male and female caregivers: findings from a population-based panel study. Soc Sci Med. 2020;264:113281. 10.1016/j.socscimed.2020.113281. [DOI] [PubMed] [Google Scholar]
- 3.Gruber J, McGarry K, Hanzel C. Long-term Care Around the World. Cambridge, MA: National Bureau of Economic Research; 2023 Nov. p. w31882. Report No.: w31882. Available from: http://www.nber.org/papers/w31882.pdf.10.3386/w31882. Cited 2024 Mar 26.
- 4.Fu R, Iizuka T, Noguchi H. Long-term Care in Japan. Cambridge, MA: National Bureau of Economic Research; 2023 Nov. p. w31829. Report No.: w31829. Available from: http://www.nber.org/papers/w31829.pdf.10.3386/w31829. Cited 2024 Mar 28.
- 5.Stara V, Soraci L, Takano E, Kondo I, Möller J, Maranesi E, et al. Intrinsic capacity and active and healthy aging domains supported by personalized digital coaching: survey study among geriatricians in Europe and Japan on eHealth opportunities for older adults. J Med Internet Res. 2023;25:e41035. 10.2196/41035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Koreshi SY, Alpass F. Reconciling paid work and informal caregiving among older adults: implications for work interference with family. Innov Aging. 2023;7(9):igad115. 10.1093/geroni/igad115. (PubMed PMID: 38024332; PubMed Central PMCID: PMC10681359). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Harris EC, D’Angelo S, Syddall HE, Linaker C, Cooper C, Walker-Bone K. Relationships between informal caregiving, health and work in the Health and Employment After Fifty study, England. Eur J Public Health. 2020;30(4):799–806. 10.1093/eurpub/ckaa078. (PubMed PMID: 32494804; PubMed Central PMCID: PMC7445037). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Boumans NPG, Dorant E. The relationships of job and family demands and job and family resources with family caregivers’ strain. Scand J Caring Sci. 2021;35(2):567–76. 10.1111/scs.12873. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cheng M, Yang H, Yu Q. Impact of informal caregiving on caregivers’ subjective well-being in China: a longitudinal study. Arch Public Health. 2023;81(1):209. 10.1186/s13690-023-01220-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Greenhaus JH, Beutell NJ. Sources of conflict between work and family roles. Acad Manage Rev. 1985;10(1):76. 10.2307/258214. [Google Scholar]
- 11.Okabayashi H, Sugisawa H, Takanashi K, Nakatani Y, Sugihara Y, Hougham GW. A longitudinal study of coping and burnout among Japanese family caregivers of frail elders. Aging Ment Health. 2008;12(4):434–43. 10.1080/13607860802224318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Koreshi SY, Alpass F. Understanding the use of flexible work arrangements among older New Zealand caregivers. J Appl Gerontol. 2023;42(5):1045–55. 10.1177/07334648231152153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Holtom BC, Lee TW, Tidd ST. The relationship between work status congruence and work-related attitudes and behaviors. J Appl Psychol. 2002;87(5):903–15. 10.1037/0021-9010.87.5.903. [DOI] [PubMed] [Google Scholar]
- 14.Lu L. Working hours and personal preference among Taiwanese employees. Int J Workplace Health Manag. 2011;4(3):244–56. 10.1108/17538351111172608. [Google Scholar]
- 15.Shockley KM, Allen TD. When flexibility helps: another look at the availability of flexible work arrangements and work–family conflict. J Vocat Behav. 2007;71(3):479–93. 10.1016/j.jvb.2007.08.006. [Google Scholar]
- 16.Bergmann M, Wagner M. The impact of COVID-19 on informal caregiving and care receiving across Europe during the first phase of the pandemic. Front Public Health. 2021;9:673874. 10.3389/fpubh.2021.673874. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Raiber K, Verbakel E. Are the gender gaps in informal caregiving intensity and burden closing due to the COVID‐19 pandemic? Evidence from the Netherlands. Gend Work Organ. 2021;28(5):1926–36. 10.1111/gwao.12725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ministry of Health Labor and Welfare. 2010. Available from: Available at: https://www.mhlw.go.jp/english/wp/wp-hw4/10.html. Accessed 14 Aug 2025.
- 19.International Standard Classification of Occupations (ISCO). 2008. Available from: https://ilostat.ilo.org/methods/concepts-and-definitions/classification-occupation/#:~:text=The%20International%20Classification%20of%20Occupations,form%20that%20can%20be%20useful.
- 20.Arksey H. Combining informal care and work: supporting carers in the workplace: combining informal care and work. Health Soc Care Community. 2002;10(3):151–61. 10.1046/j.1365-2524.2002.00353.x. [DOI] [PubMed] [Google Scholar]
- 21.Bernard M, Phillips JE. Working carers of older adults: What helps and what hinders in juggling work and care? Community Work Fam. 2007;10(2):139–60. 10.1080/13668800701270075. [Google Scholar]
- 22.Spann A, Vicente J, Allard C, Hawley M, Spreeuwenberg M, Witte L. Challenges of combining work and unpaid care, and solutions: a scoping review. Health Soc Care Community. 2020;28(3):699–715. 10.1111/hsc.12912. [DOI] [PubMed] [Google Scholar]
- 23.Nadash P, Alberth AG, Tell EJ, Jansen T. Policies to sustain employment among family caregivers: the family caregiver perspective. J Appl Gerontol. 2023;42(1):3–11. 10.1177/07334648221125635. [DOI] [PubMed] [Google Scholar]
- 24.Plaisier I, Broese Van Groenou MI, Keuzenkamp S. Combining work and informal care: the importance of caring organisations. Hum Resour Manag J. 2015;25(2):267–80. 10.1111/1748-8583.12048. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are restricted by legal or ethical reasons and are not publicly available. Data are, however, available from the authors upon reasonable request and with permission of relevant ethics committee or legal authority.