1. Introduction

Unpaid informal care contributes to an estimated 70–95% of the world’s care needs, and significantly impacts the lives of those who provide it, imposing both positive and negative effects [1–3]. Moreover, unpaid care is deeply gendered, with women accounting for an estimated 80% of informal carers globally [4]. The COVID-19 pandemic highlighted not only the importance and scale of unpaid care provision but catapulted it to being an essential frontline component of the pandemic response across the globe, with demands increasing exponentially as public services were restricted [5, 6]. Ultimately the pandemic resulted in greater unpaid caring responsibilities and increased caregiving intensity for carers worldwide, the effects of which are only just starting to be understood. This review addresses a gap in the existing literature by synthesising and evaluating the quantitative evidence examining the association between informal unpaid caregiving and mental health (compared to non-caring), during the COVID-19 pandemic.

Mental health (MH) disorders are a considerable public health concern worldwide. Even prior to the pandemic, MH disorders (depression and anxiety in particular) were persistently among the top causes of the global disease burden [7]. In creating additional and unprecedented hazards to mental health, the pandemic triggered relatively higher rates of stress, anxiety, depression, and psychological distress across the globe [8, 9]. All major emergencies and disasters generate high levels of fear, uncertainty, and stress, and stress is an important pathway through which our social and physical environments affect our MH [10, 11]. Extensive literature also acknowledges the stressful nature of caregiving [12], and the caregiver stress process through which caregiving is theorised to affect MH [13]. This is supported by several (pre-COVID) reviews reporting a negative association between unpaid caregiving and MH for carers of all ages [14–20]. Compared to non-carers, caregivers were likely disproportionately affected over the course of the pandemic due to the additional (and interwoven) emotional, health-based, and financial stressors they experienced [21]. Yet, whilst much initial attention and focus were directed toward the paid care workforce, unpaid caregivers remained largely hidden in comparison [22]. Fortunately, increasing attention and recognition of the experiences of informal caregivers has emerged, and the emerging literature base reflects this.

The pandemic was a globally unifying experience and, whilst different countries experienced differing length and severity of public health measures, strict lockdowns were commonplace, as were strict quarantine measures and reduced service access. Such measures resulted in additional stressors for caregivers. Not only did reduced access to support services (e.g., home health assistance, allied health care, respite care) increase caring intensity/responsibilities, lockdowns severely disrupted protective social support networks and connections [21, 23]. Moreover, the danger of a life-threatening virus only intensified caregiver worries regarding care recipient health [24], often prompting further isolation due to fear of contracting and spreading COVID-19 to their loved one [21, 23]. Additionally, many caregivers also had to negotiate new or intensified economic stressors or job insecurities because of the pandemic [21]. In contrast, some positive effects have been postulated, including deeper connections with care recipients [25], spending more time with loved ones [26], and having a greater sense of purpose [27]. Therefore, the possible MH effects for caregivers (compared to non-carers) in the context of the pandemic are potentially mixed.

Despite our proximity to the pandemic, several reviews have already been published in this space. A rapid systematic review examined the MH impact of the pandemic on caregivers of adult family members [28]. From the eight included studies (screening ceasing in September 2021), the review reported two main themes: family caregivers COVID-19 related stress, and (mal) adaptive strategies to the “new” normality [28]. Meanwhile, a 2021 systematic review examined the MH of caregivers of dementia patients specifically, reporting worsening mental health and increased burden because of the COVID-19 pandemic and its associated public health measures [29]. Lastly, another systematic review examined the psychological impact of the pandemic (lockdown and quarantine measures) on caregivers of children and adolescents (<18yrs), reporting in pooled estimates that 52.3% and 27.4% of caregivers developed anxiety and depression, respectively, while being in isolation with children [30].

Whilst the above evidence provides an important starting point to understanding how the global pandemic has impacted the MH of unpaid caregivers, no reviews to date have examined the effect of caregiving compared to non-caregiving within this context. This is important because the pandemic exerted a significant impact on the MH of the population. Population studies have demonstrated that MH among adults [31] and children [32] declined, with effects more pronounced in some subgroups. Without a comparator, it is not possible to distinguish the effects of the pandemic from the effects of caring. In addition, compared to extant reviews, our review criteria encompass the breadth of what constitutes unpaid caregiving, rather than being restricted to care for older adults, or care for children, or for specific care recipients based on disease or condition. Our review addresses both gaps and extends to include all published work in this space up until March 2023. Lastly, understanding the impact of the pandemic on carers is not merely an exercise in assuaging curiosity, but it is of significant practical importance. Not only are informal care demands predicted to continue to increase, but COVID-19 will not be humanity’s last experience of a pandemic. As such, gleaning what we can from this pandemic is vital in informing responses to future pandemics.

The main aims of this review are:

1. To summarise the quantitative evidence examining the association between informal unpaid caregiving and MH (compared to non-caring) in the context of the COVID-19 pandemic.
2. To assess the quality of the existing evidence.

2. Methods

3. Results

3.1. Study characteristics

A total of 5,816 studies were identified by the search, from which 2,743 duplicates were removed, resulting in a total of 3,073 records being screened by title and abstract. From these, 145 records were assessed for eligibility through full-text screening. As per Fig 1 (PRISMA flowchart), 125 of these were excluded due to not meeting the eligibility criteria, resulting in a total of 20 eligible studies (totalling 113,151 participants) for inclusion in the systematic review. S1 Table documents the studies excluded at full text review with reasons for exclusion, the most common reason being the absence of a non-caring comparator.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Study selection diagram (PRISMA flowchart).

https://doi.org/10.1371/journal.pone.0297097.g001

Of the included studies, six employed a longitudinal design [41–46], whilst fourteen were cross-sectional [47–60]. Geographically, there were five studies from the United Kingdom [41, 42, 45, 46, 56], three from the US [44, 49, 55], three from Italy [48, 50, 60], two from Japan [57, 58], and one each from Australia [52], Austria [59], Canada [43], China [54], Hong Kong [53], Latin America [51], and New Zealand [47]. Whilst the majority of studies were sampled from the general population [41, 43–45, 49–52, 54–57], some studies were restricted to older adults [46, 47, 58], one was restricted to working age adults (20-64yrs) [59], another to young adults (18-29yrs) [60], one to undergraduate students [48], another to grandparents [42], and one study examined adult women only [53]. The studies varied in how they defined and analysed informal caregiving. The majority operationalised unpaid caregiving as a binary variable (care provision versus no care provision), whilst seven studies categorised caregiving in a variety of ways. Two studies categorised carers as new carers or continued/existing carers [41, 46], one study had a very broad definition of caregiving [46], whilst the other restricted their caring exposure to out-of-home caregiving [41]. One study divided carers into short-term and long-term (>12 months) caregivers [44], whilst another examined childcare and eldercare separately as well as together (sandwich care) [54]. Two studies inferred caregiving based on family/household composition. One study classified family caregivers if any child or adult in the household had a disability (presumed care) vs having a disability themselves, with the comparator group being no household members with disability (including self); the presumed non-caring comparator [51]. Another study (examining young adult carers) presumed caregiving based on who (if anyone) in their household had a serious physical or mental health condition (i.e., parent or other ill family member, rather than being explicitly self-reported [60]. The sixth study focussed on care among grandparents, examining changes in grandparental care because of the pandemic. Those doing mostly the same (or increased) grandparental care were classified as caregivers, and non-carers were those grandparents who were caring for grandchildren pre-pandemic and then had completely stopped caregiving since the pandemic began [42].

Of the thirteen studies that examined caregiving (vs non care) as binary exposure, nuance existed in how the exposure was measured/operationalised. Some studies defined caregiving as broadly providing unpaid care or assistance to anyone with a disability, chronic illness or who was elderly [43, 47, 59]. Other studies nominated family member caregiving specifically [49, 57, 58], whilst four studies only considered caregiving for someone residing in the same household [45, 48, 50, 53]. In addition to the grandparental care study [42], two studies specifically examined childcare, one classifying caregivers based solely on whether they were the primary carer for a child [55], and another examining two caregiving exposures separately [52], one being childcare and the other caregiving for dependant relatives.

Every included study utilised a validated measure of MH (in line with the review criteria), all of which were self-reported via survey-based instruments. Seven studies examined depression and anxiety as two separate MH outcomes [43, 47–49, 52, 53, 60], whilst twelve studies examined only depression/physiological distress [41, 42, 44–46, 50, 54–59], and one study examined only anxiety [51]. A descriptive summary and main findings of included studies (including effect estimates and confidence parameters where reported) can be found in Table 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Descriptive summary of included studies.

https://doi.org/10.1371/journal.pone.0297097.t001

3.3. Data synthesis

To complement the narrative synthesis, Table 2 presents an effect direction plot [39] providing visual synthesis of the effect direction of all included studies in the review. Moreover, to aid interpretation and enable more nuanced data synthesis, three Albatross plots [40] (Fig 2A–2C) present different subgroup estimates based on exposures and outcomes of included studies. These were:

a) Adult caregiving and depression

Eleven studies [44, 45, 47–50, 52, 56–58, 60] (comprising 14 estimates) presented sufficient data to calculate a standardised mean difference (SMD) or Cohen’s d, allowing comparison and construction of an albatross plot for the association between adult caregiving and depression (Fig 2A). All studies show a positive effect size; nine estimates falling between SMD contours of 0.15 and 0.05, three estimates between 0.25 and 0.15, and one below 0.05. This indicates a moderate positive association between caregiving and poorer mental health, with no definitive gender subgroup differences.

b) Adult caregiving and anxiety

Six studies [47–49, 51, 52, 60] (comprising 8 estimates, two of which were unadjusted) [47, 49] presented sufficient data to construct an albatross plot for adult caregiving and anxiety (Fig 2B). All studies showed a positive effect size; five estimates falling between SMD contours of 0.15 and 0.05, and three null results with SMD<0.05. This indicates a small positive association between adult caregiving and anxiety.

c) Childcare and depression

Whilst few studies in the review examined childcare [42, 52, 54, 55], we constructed an albatross plot (Fig 2C) for those with available estimates [42, 52, 55] to allow visual representation of the opposing effect direction reported for this small sub-group. Notably, whilst one contributing study to this negative result had the lowest risk of bias (10 stars) [42], another had the highest risk (3 stars) [55]. Therefore, whilst this plot indicates a small negative association between childcare and depression, it must be interpreted with caution given the minimal contributing data, and presumably biased data for one estimate.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2.

a. Albatross plot—Adult caregiving and depression. b. Albatross plot—Adult caregiving and anxiety. c. Albatross plot—Childcare and depression.

https://doi.org/10.1371/journal.pone.0297097.g002

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Effect direction plot/table.

https://doi.org/10.1371/journal.pone.0297097.t002

4. Discussion

To our knowledge, this is the first systematic review to synthesise and assess the quality of the quantitative evidence examining the association between unpaid caregiving and MH during the pandemic. By imposing a requirement for a non-caring comparator, our review accounted for the effect of the pandemic to interrogate the relationship between informal caregiving and MH during this period. Our results indicate that, in the pandemic setting, unpaid caregiving was detrimental to MH, particularly depression and psychological distress. Fifteen out of the twenty included studies reported poorer MH in caregivers compared to non-carers (either overall, or in at least one category when caregiving was interrogated as a categorical variable). Three studies reported no association. An overall positive association was reported in two studies examining childcare. The overall quality of the evidence was moderate.

Prior to the COVID-19 pandemic, informal unpaid care provision was known to impact not only the lives but also the MH of those who provide it [17, 18]. Whilst the COVID-19 pandemic was unprecedented in the demand it placed on caregivers, paid and unpaid alike, it was unclear whether the MH of unpaid caregivers remained poorer compared to non-carers. This review has answered this question. In line with Carbone et al.’s review of dementia caregivers specifically [29], our review found that informal unpaid caregiving had a detrimental impact on mental health in the context of the pandemic. This is important, as whilst the WHO recently declared an end to COVID-19 as a global health emergency [61], response readiness for emerging variants and predicted future pandemics is vital [62]. Part of this preparedness should be to ensure that key learnings pertaining to the unique challenges faced by informal unpaid caregivers during the pandemic are prioritised, and aid in driving better policy in anticipation of future public health emergency events.

Our overall finding of poorer MH amongst caregivers during the pandemic is supported by other research aligned with this topic. This is predominantly via published work documenting caregivers’ experiences during COVID-19, which illustrate feasible pathways to the reported increased prevalence of depressive and anxiety disorders found by our review. For example, a US based survey of caregivers during 2020 showed that most carers reported an increase in stress (83%) and a greater feeling of loneliness (77%) during the pandemic [63]. This aligns with Dellafiore et al.’s earlier rapid systematic review reporting caregivers’ heightened stress as a key theme [28]. Moreover, a 2022 systematic review of the qualitative literature reported not only a rise in caregiving demands for carers, but a notable increase in caregiver negative emotions such as fear and uncertainty [64]. Importantly, factors such as knowledge of possible asymptomatic transmission of the virus and worries regarding no visitation rights if their loved care-recipient was hospitalised exacerbated these fears. Moreover, constant uncertainty was fueled by changeable and often abrupt and oscillating loss of health services as well as postponement/interruption of care recipient treatment, and, for some carers, economic insecurities was intensified. The same qualitative review also highlighted the decreased availability of social support, leading to concerns about carer wellbeing and ability to cope [64]. On top of this, in another study, carers described having to provide even greater emotional support than usual to their care recipients, subsequently reporting heightened mental exhaustion [65].

Therefore, in considering these recurring themes of caregiver stress, fear, uncertainty, and loneliness, coupled with decreased support, we reflect that the pandemic delivered somewhat of a perfect recipe for caregiver’s increased susceptibility to poorer MH. Whilst the mechanisms through which informal caregiving may impact MH in “normal times” are documented in detail elsewhere [15, 16, 18, 66], it is likely that the caregiving stress process model (SPM) [13, 67] also aptly describes many of the possible mechanisms at play during COVID-19. According to the SPM, it is the dual mix of objective circumstances and environment serving as primary stressors, as well as a carer’s subjective (secondary) experience of such stressors that contribute to MH effects [13, 67], both of which were heavily influenced by the pandemic. Another likely mechanism (well-known prior to the pandemic) is the family effect, whereby the carer wellbeing is contingent on the health and welfare of the care recipient [68]. Research prior to the pandemic examining the family effect showed informal carers suffer adverse MH effects as the health of their parents, spouses or loved ones deteriorated [14, 69]. We hypothesise that this family effect is highly relevant, with carers witnessing the fear, distress, and loneliness of their care recipients as the threat of the pandemic and the consequences of its associated public health measures unfolded. Lastly, whilst the employment status of carers varied within and between the studies included in this review, we also postulate that, for those working, the juggling of work commitments with care provision remained challenging, with time poverty pressures and reduced time for self-care (exercise, recreation, relaxation, sleep) possibly also contributing to poorer MH [24, 70, 71]. Ultimately however, we posit that the pandemic not only intensified many of the challenges associated with caregiving in “normal times” but, amid throwing the entire global population into turmoil, had serious and deeply troubling additional effects on unpaid carers, the extent of which may take many years to fully understand.

In contrast to our overall findings, a small minority of studies included in this review reported a positive association between caregiving and MH during the pandemic. All three pertained to caregiving of children as their exposure (one being grandparental care [42], another for workers at Amazon who were primary carers of a child <18 [55], and one for childcare by women under 50years (the same study finding the opposite effect in older women and no effect in men)) [52]. It may be that caring for (presumably healthy) children during COVID provided much needed company when most households/persons were extremely isolated due to the pandemic. It is also possible that children provided distraction and a greater sense of purpose during this time. Nonetheless, given the small number of studies and the heterogeneity of the exposure across them, this review cannot draw robust conclusions about the overall effect of care for children during the pandemic on MH. Moreover, our findings do not align with the 2021 systematic review by Panda et al. who reported that caregivers of children and adolescents developed anxiety and depression while being in isolation with children, but they did not measure care per se [30]. Of note, another study in our review (reporting no association for childcare or eldercare as stand-alone exposures), did find sandwich care (both child and elder care) to negatively impact mental health [54]. This aligns with other research demonstrating that balancing caregiving responsibilities of older or disabled adults, whilst also having childcare responsibilities was identified as an increased risk factor (increased subjective burden and a lower carer-related quality of life scores) for caregivers during COVID-19 [65].

Overall, our review and the associated supporting research suggests that the pandemic was a particularly bleak time for informal unpaid caregivers. Our findings point not only to an ongoing need to alleviate the psychological risks of informal caregiving in a post-pandemic world, but also shine a light on the disproportionate MH penalty paid by caregivers because of the pandemic It is essential that these learnings ignite a concerted effort to instate policy and procedures that can be implemented in the face of future public health emergencies. Importantly, whilst we know that the pandemic accelerated the use (and acceptance) of digitally enabled psychosocial interventions [72], only limited work has been published regarding interventions initiated to support the caregiver MH during COVID to date [73]. A 2021 systematic review synthesised the available evidence and found only three studies (from China and Italy) that examined psychological support interventions for informal caregivers during the pandemic [74]. The review reported that all interventions were delivered in digital format, and that there was considerable heterogeneity between the type of treatment, digital service, and care recipients across the three studies [74]. Interventions included a cognitive and sensory–motor intervention designed to lower caregiver distress, a caregiver skills educational program, and telehealth psychological support services, with the review surmising interventions were effective in helping reduce caregiver distress and burnout while improving self-efficacy and wellbeing [74]. A 2023 systematic review of caregivers of people with dementia supported these findings, reporting that internet-based psychoeducational programs (comprising educational, psychological, and behavioural training) can improve some aspects of caregivers’ MH and emotional wellbeing [75]. Ultimately, whilst some possible policy actions and recommendations (with the aim of improving outcomes for informal caregivers) have been described in detail elsewhere [76, 77], there remains a dearth of evidence to inform suitable and targeted support for caregivers in the event of a future pandemic or similar widescale emergency [73], a gap that needs to be urgently addressed before the window of information and learnings from the recent pandemic recedes and closes.

5. Conclusion

This systematic review synthesised the quantitative evidence of the association between unpaid caregiving and MH during the COVID-19 pandemic. The findings demonstrate that the MH of informal carers, already poorer pre-COVID compared to non-caregivers, was disproportionally impacted because of the pandemic and its associated public health containment measures. Given that few studies examining childcare employed a non-caring comparator, further work is warranted to synthesise and assess the association between childcare and MH specifically. The overall quality of all included studies was moderate. Our review is timely as informal unpaid caregivers are increasingly being recognised as the forgotten or hidden care workers of the pandemic. Accordingly, this review provides important evidence highlighting the vulnerability of this group and should motivate actionable learnings and drive policies to ensure unpaid caregivers are better supported now, in the medium term, and crucially if, and when, another global public health emergency emerges.

Supporting information