Increased life expectancy and recent social changes in Europe are related to an increase in the number of grandparents providing supplementary care to their grandchildren (Triadó et al., 2014). These are grandparents who provide care in addition to the parental care received by their grandchildren (Arpino et al., 2018). The Survey of Health, Ageing and Retirement in Europe (SHARE, a survey of European people aged 50 and over dwelling in the community) showed that approximately 52% of grandparents were supplementary providers of care to children under 16 years of age (Arpino et al., 2018). However, there are significant differences in care intensity across Europe. Southern European countries, such as Greece (24.8%), Italy (20.3%) and Spain (15.2%), have more parents with at least one child looked after regularly by a grandparent (more than 15 h per week) than Northern European countries, such as Denmark (3.6%), Sweden (3.6%) and the Netherlands (6.9%). The higher number of grandparents acting as regular caregivers in Mediterranean countries may be explained by the lack of part-time job opportunities, parental leave benefits for working parents, as well as insufficient formal childcare resources (Di Gessa et al., 2015).

Providing care for a grandchild occasionally (as a voluntary option and leisure activity) may not be as demanding as assuming the obligation of caregiving regularly. Data drawn from the SHARE survey (waves 1 and 2) shows 32.3 average monthly hours of non-intensive childcare in Europe (1.07 h per week) (Brunello & Rocco, 2016) while regular caregivers reach 30 h per week in Europe (Di Gessa et al., 2015). This means that many regular caregivers are providing care for their grandchildren for an amount of time similar to holding a full-time job, and many of them are performing tasks that are usually part of parents’ responsibilities. A study conducted in Italy showed that it was more likely to find regular care provided to grandchildren if the grandparents were women, were married, had an adequate economic status and good health levels (Zamberletti et al., 2018). Also, Di Gessa et al. (2015) found that in Mediterranean countries the probability of engaging in regular caregiving was higher in younger grandparents, more commonly women that were not working and that had fewer grandchildren (Di Gessa et al., 2015). Considering the high levels of care that some grandparents are providing, the effect of caregiving for grandchildren on grandparents’ HRQoL may be related to its intensity.

Caregiving for grandchildren regularly may require grandparents to learn new educational methodologies that can be challenging and emotionally demanding (Noriega et al., 2017). Regular caregivers not only participate in playful activities but also day-to-day tasks with their grandchildren (Viguer et al., 2010). The supervision of grandchildren, with whom grandparents take part in day-to-day tasks, by setting norms and limits is crucial to help grandchildren internalize a set of values that guide their behaviors (Noriega et al., 2017). Setting norms and limits is especially important for primary school aged children because they are in a period in which they enroll in primary school and must cope with new academic and social demands (Erikson, 2000). While grandparents were brought up in an educational system in which children did not argue with authority (Noriega et al., 2017), current primary school aged grandchildren often express their disagreement and can show difficulties associated with behavioral problems (Triadó et al., 2014). However, these challenges can also have positive consequences. For example, a recent study of grandparents acting as supplementary caregivers in Spain found that their personal growth increased when they updated their socializing styles and moved to a more authoritative one when socializing primary school aged children (Noriega et al., 2020).

Grandparents’ HRQoL

HRQoL is a multidimensional concept defined as the subjective assessment of the influence of health status and the level of physical, psychological and social functioning on the possibility of achieving goals in life (Shumaker & Naughtozn, 1995). It is a crucial concept for older adults because it analyzes different dimensions of quality of life related to physical and mental health and its relationship with well-being (Kane, 2003; Karimi & Brazier, 2016). HRQoL assessment is increasingly used in public health research with older adults and plays an important role in the development of health services (Huong et al., 2017).

To the best of our knowledge, there is a lack of studies considering supplementary caregivers’ HRQoL (a concept that goes beyond just health), while the ones focused on health status have shown mixed results. Data from the US Nurses’ Health Study indicated a higher risk of coronary disease in grandmothers who provided regular care to their grandchildren for more than 20 h per week (Lee et al., 2003). Another study conducted in Spain indicated poorer perceived health as the number of years providing regular care (more than 12 h per week) increased (Triadó et al., 2014). These results may be associated with Role Strain Theory (Goode, 1960), in which high levels of care engagement can compete with other role responsibilities, increasing grandparents’ stress levels. If the demands exceed grandparents’ abilities to cope with stress associated with supplementary caregiving, this may have negative consequences for their health (Di Gessa et al., 2016). In contrast, other studies have found a positive relationship between the provision of supplementary care and grandparents’ health in different European countries participating in the SHARE survey (Danielsbacka et al., 2019) and in China (Zhou et al., 2017). In these studies, grandparents’ health did not deteriorate because of the amount of care provided. However, perceiving caregiving as a stressor had a negative impact on grandparents’ health (Hughes et al., 2007; Triadó et al., 2014). This means that the effects on supplementary caregivers’ health appear to be mediated by the context and circumstances of caregiving (e.g., the demands of caregiving for grandchildren and available resources) instead of by the intensity of the care (Hughes et al., 2007; Triadó et al., 2014). However, these authors also state that it is more likely to find difficulties associated with caregiving when it is more intense (e.g., grandparents who care for grandchildren daily may have to cope more often with grandchildren’s behavioral problems and set more limits).

Conceptual Framework

There is limited literature that examines supplementary caregivers’ HRQoL and the positive factors that mediate the relationship between care provided by grandparents and their HRQoL. Moreover, most of the studies have focused on the adverse effects of grandparents’ caregiving. There is a need to explore the positive effects of grandparents’ caregiving and associated factors promoting positive caregiving. The Resiliency Model of Family Stress, Adjustment and Adaptation (McCubbin, 1993) considers psychological resources and coping strategies as protective factors that may minimize the impact of caregiving difficulties. This model has been supported by Musil et al. (2006) in grandmothers (primary caregivers living in multigenerational homes and non-caregivers). Specifically, they found that coping resources, subjective social support and resourcefulness mediated the effects of grandmothers’ family life events. In our study, social support and character strengths were used as protective factors. Detecting risk factors that may have negative consequences on grandparents’ mental and physical health is essential, but so is detecting protective factors that may minimize the impact of the difficulties associated with caregiving. This focus on the positive aspects may contribute to developing more effective family intervention programs and to reducing grandparents’ vulnerability by promoting protective factors and building resilience.

Protective Factors

Amount and satisfaction of social support

Social relationships play a significant role in promoting health and helping the individual cope with adverse situations. The integration in the community through the development of social links is especially important for older adults (Schönfeld et al., 2017). According to Role Enhancement Theory (Moen et al., 1995), the development of multiple roles can increase social integration which, in turn, is related to greater emotional gratification, sense of identity, available resources and social recognition. Specifically, caregiving for grandchildren may provide non-custodial grandmothers with opportunities to increase their social networks and further their personal growth (Moore & Rosenthal, 2014), as well as keeping them more active and re-energized (Fuller-Thomson et al., 2014). Zhou et al. (2017) favored the mediating role played by social support between health and care provided by grandparents. Most of the previous research has focused on grandparents as primary caregivers (grandparents who are the main caregivers of their grandchildren and develop a parental role) and has considered their social support in terms of the number of their social and interpersonal relationships. There is evidence to suggest that the appraisal of the available social support is more important than the number of interpersonal contacts and has a greater impact on physical and psychological heath (Antonucci & Israel, 1986; Vandervoort, 1999). The reason for this is that different individuals have varying needs with regards to relationships and attach different meanings to these contacts (Saranson et al., 1987). For this reason, it is essential that research includes both the number of contacts and the satisfaction with them.

Character strengths

Character strengths are understood as the natural ability by which thoughts, feelings and behaviors allow the individual to achieve their goals (Peterson et al., 2006). Research has supported a relationship between character strengths and physical and mental health outcomes (life satisfaction, happiness, wellness, positive affect, quality of life and sense of purpose) (Buschor et al., 2013; Proctor et al., 2011). To our knowledge, there is only one study that has found a positive association between HRQoL and character strengths, but it was carried out on university students (Proctor et al., 2011). This study determined that character strengths were associated with increased HRQoL among students. Research examining the link between character strengths and social relationships is also limited. Older adults’ character strengths might increase by sharing their personal experience with other people (Margelisch, 2017) while strengths could contribute to creating or maintaining social relationships that are crucial at all life stages (Martínez-Martí & Ruch, 2014).

Therefore, the present study aims to relate character strengths to grandparents’ perception of the social support they received and their HRQoL, and also to further examine the protective role of character strengths and social support for HRQoL. Also, this study examines whether character strengths exhibit an interdependent association with HRQoL beyond the influence of health and social support. Despite the lack of research that examines the possible relationships between these two variables, it is expected that character strengths buffer the stress associated with caregiving.

As can be observed, there is limited literature that examines what may mediate the relationship between care provided by grandparents and HRQoL. The Resiliency Model of Family Stress, Adjustment and Adaptation (McCubbin, 1993) offers a positive viewpoint that considers protective factors which, in turn, buffer the stress perceived when dealing with family demands. Social support has been shown to be a protective variable that significantly influences primary caregivers’ health. Nevertheless, social support’s influence on supplementary caregivers’ HRQoL has not been studied in detail yet, nor has the role played by grandparents’ character strengths. The study of protective factors is of great interest to be able to detect grandparents in situations of vulnerability and develop preventive and intervention programs that promote protective factors that, in turn, may mitigate the possible difficulties associated with caregiving. Not only may grandparents benefit from these programs but also grandchildren, since it is more likely that grandparents with higher levels of health provide better quality care.

Purpose and Research Questions

Based on the Resiliency Model of Family Stress, Adjustment and Adaptation of McCubbin (1993), this study addresses the following research questions: (1) Are there significant differences in socio-demographic characteristics, social support, character strengths and HRQoL (mental and physical) between grandparents who act as caregivers regularly and those who do so occasionally? (2) Are there positive relationships between social support (number and appraisal), HRQoL and character strengths? and (3) Do social support and character strengths mediate the relationship between the amount of caregiving grandparents provide for their grandchildren and grandparents’ HRQoL (in both regular and occasional supplementary care providers)?

Considering these research questions, we hypothesized that (1) we will not find differences among regular and occasional caregivers; (2) social support, character strengths and HRQoL will show positive associations; and (3) the effects of the amount of care on grandparents’ mental and physical HRQoL will be mediated by grandparents’ psychological resources (social support and character strengths) (see Fig. 1).

Fig. 1
figure 1

Hypothesized model

Full size image

Method

Participants

Three hundred grandparents living in Spain completed a self-report survey. Participants were supplementary caregivers of grandchildren aged between 6 and 12 years old.

Measures

Socio-demographic data

The following variables were included: age, gender (0 = men; 1 = women), marital status (0 = single; 1 = married or living with a partner; 2 = divorced/separated; 3 = widowed), level of education (0 = secondary school incomplete; 1 = secondary school complete; 2 = post-secondary school; 3 = university complete), job status (0 = working full-time; 1 = working part-time; 2 = unemployed; 3 = housewife/househusband; 4 = retired) and number of grandchildren participants were looking after. We also asked participants to choose one of their grandchildren aged 6–12 they were providing additional care for and to respond to the following variables: lineage (0 = maternal grandparent; 1 = paternal grandparent), grandchildren’s gender (0 = boy; 1 = girl) and age.

Character strengths

We applied the Psychological Resources Inventory (IRP-77; Martínez, 2007). This scale evaluates character strengths through 77 items. We used the total score of grandparents’ character strengths. The scoring ranged from 0 to 3, where 0 corresponds to completely disagree and 3 to completely agree. Some examples of items are the following: I tend to analyze situations from different viewpoints, I think things will go well in the future and I think challenges are opportunities, rather than difficulties. Scores under 140 represent low levels of character strengths, between 141 and 170 are characteristic of medium levels and scores between 171 and 231 are representative of high levels. This instrument has shown a high internal consistency in our sample of grandparents, with a Cronbach’s alpha equivalent to 0.94 for the total scale.

Social support

We administered the SSQ3 version (López et al., 2004) of the Short Form Revised of the Social Support Questionnaire of Saranson et al. (1987). It comprises two subscales: number of interpersonal contacts (3 items, ranging from 0 to 27) and satisfaction with social support (3 items, ranging from 3 to 18). To assess the number of interpersonal contacts, participants were asked to indicate (1) the number of people they trust when they need help, (2) the number of people that help them feel relaxed when they are tense and (3) the number of people that can comfort them when they feel upset. To assess social support satisfaction, they were required to indicate their degree of satisfaction with the people they indicated in the situations scoring in a Likert scale ranging from 1 (very unsatisfied) to 6 (very satisfied). This instrument showed a high internal consistency in this sample, achieving a Cronbach’s alpha of 0.87 in both subscales.

Amount of care provided for the grandchild

We asked participants to indicate the average daily hours of care provided for their grandchildren. This question was adopted from the SHARE survey (Arpino et al., 2018). This variable was used to categorize “occasional” and “regular” caregivers and was also included in the SEM model tested.

Health-related quality of life

We used the Spanish version of the Short-Form Health Survey (SF-12; Vilagut et al., 2008). This scale consists of 12 items that evaluate to what extent one’s current physical and mental health can be a limitation to perform everyday tasks. The items are scored heterogeneously. Item 1 asks on a Likert-type scale from 1 (excellent) to 5 (poor) about the perception of participants’ general health; items 2 and 3 ask through a Likert-type scale from 1 (yes, it limits me a lot) to 3 (no, it doesn’t limit me at all) if one’s current health limits the individual’s ability to carry out activities; items 4–7 assess whether during the last 4 weeks the person has had any difficulties in their daily activities due to their physical (items 4 and 5) or emotional health (items 6–7) in a dichotomous way (1 = yes; 2 = no); item 8 evaluates the extent to which pain has hindered their usual work with a Likert-type scale from 1 (not at all) to 5 (a lot); items 9–11 ask the degree to which the person has felt a series of emotions (calm, energy and sadness respectively) from 1 (always) to 5 (never); item 12 asks with what frequency their physical health or emotional problems have hindered social activities, scoring from 1 (not at all) to 5 (a lot). These 12 questions are grouped into two subscales: physical and mental health. Total scores range from 0 to 100, where higher scores mean higher levels of HRQoL. This instrument has shown an adequate internal consistency in our sample for the total score (Cronbach’s alpha = 0.89), as did mental and physical health (0.67 to 0.85 respectively).

Procedure

The Innovation and Ethics Research Committee approved the study. We recruited participants through primary schools, older adults’ associations and federations from different backgrounds in Spain using snowball sampling. The researchers asked participants to recommend others who fitted the criteria (e.g., face-to-face interactions, telephone calls and emails). When grandparents expressed interest in taking part in the study, we asked them to sign the written informed consent agreement. Subsequently, we gave them a self-applied questionnaire. We integrated into one questionnaire all aspects of the study (demographics, social support, character strengths and HRQoL). It took approximately 1 h to complete the questionnaire. To avoid potential biases in the responses, we explained the objective of the study, guaranteed anonymity and gave several instructions at the beginning of the questionnaire in which we invited grandparents to answer honestly and stressed the importance of not leaving blank responses.

Data analyses

We first grouped participants by the amount of care they were providing for their grandchildren using Wellard’s (2011) cutoff criteria: Regular care providers (more than 10 h of care per week) and occasional care providers (fewer than 10 h per week). Custodial grandparents were excluded from this study because they assume a parenting role (a role very different compared to supplementary caregiving). To analyze differences between regular and occasional care providers, chi-square statistics were performed in gender, marital status, socio-economic status, educational level, job status and lineage. We conducted t-tests to analyze mean differences in age, character strengths, social support (amount and satisfaction) and perceived health between grandparents acting as caregivers regularly and those acting thus occasionally. Then, path analysis using the Maximum Likehood method (ML) was developed to test the effects on grandparents’ HRQoL (mental and physical). The variables included in the model were the following: socio-demographic variables (grandparent’s gender and age, chosen grandchild’s gender and age and number of grandchildren and lineage), care for grandchildren (average daily hours of care provided for their grandchildren), protective factors (number of interpersonal contacts, satisfaction with social support and character strengths) and HRQoL (mental and physical health). The model tested can be observed in Fig. 1. Five indexes were used to analyze the overall fit of the model: chi-square statistic (χ²), the ratio of chi-square to degree of freedom (χ²/gl), the goodness-of-fit index (GFI), the comparative fit index (CFI) and the Root Mean Square Error of Approximation (RMSEA). An indication of a good model fit occurs when scores in GFI and CFI are over 0.90, and in RMSEA are below 0.06 (Hu & Bentler, 1999). We used bootstrapping to analyze direct and indirect effects.

Finally, a multiple-group path analysis was conducted to assess factor invariance by the grandparent group (regular versus occasional caregivers). The dataset was split into occasional and regular caregivers and then the model was tested with each set of data separately. According to Byrne (2016), chi-square differences (∆χ2) are the traditional method that analyzes invariance and is the most widely accepted when conducting multi-group comparisons. When ∆χ2 is statistically non-significant, it suggests that all specified equality constraints are plausible. Since ∆χ2 value is sensitive to sample size, Cheung and Rensvold (2002) also recommend conducting CFI differences (ΔCFI = 0.001) and suggest that its difference value should not exceed 0.01.

Missing data were under 1.61%. We used the Expectation-Maximization algorithm to control missing data following Tabachnick and Fidell’s (2014) recommendations when using path analysis. SPSS Statistics 25 and AMOS 22 were used.

Results

There were 107 regular and 193 occasional care providers. Grandparents’ mean age was 71.55 (SD = 7.32), ranging between 51 and 98, and 56% were women. The average number of grandchildren per grandparent was 3.53 (SD = 4.4). 22% of grandparents were providing daily care, 51% weekly, 20.3% between one and three times per month and only 6.7% between one and twelve times per year. All participants were Caucasian (Table 1 shows socio-demographic data of each group).

Table 1 Comparison of Grandparents’ Demographic Characteristics by Caregiver Group
Full size table

We found significant differences between grandparent groups in age, lineage and gender. Regular caregivers showed a higher rate of women (63.6%), maternal grandparents (72%) and a lower average age (M = 69.46%; SD = 7.02) compared with occasional caregivers (51.8; 54.9%; M = 71.28; SD = 7.00).

We also found a significant difference in perceived physical health (t = −2.23; p = 0.03), showing regular caregivers achieving higher scores (M = 48.24; SD = 9.26) compared with occasional ones (M = 45.52; SD = 10.47). In contrast, there were no significant differences between grandparents’ groups in character strengths, social support and perceived mental health (see Table 2).

Table 2 Means and standard deviations for study variables by grandparent group
Full size table

Based on the Resiliency Model of Family Stress, Adjustment and Adaptation (McCubbin, 1993), we hypothesized that the effects of the amount of care on grandparents’ mental and physical health would be mediated by grandparents’ protective resources (social support and character strengths) (Fig. 1). This model showed a poor data fit (χ² = 82.331; χ²/gl = 26; p = 0.001; GFI = 0.953; CFI = 0.828; RMSEA = 0.085). There were also several insignificant paths as can be observed in Table 3.

Table 3 Structural equation model results of the hypothesized model
Full size table

To increase the degrees of freedom we deleted some insignificant paths. We used the modification fit indexes information to conduct modifications in the model. At each step, we eliminated the parameter that produced the best improvement to the fit. This procedure lasted until data reached a good data fit (Jöreskog, 1993). This new model showed an excellent model fit (χ² = 57.518; χ²/gl = 30; p = 0.002; GFI = 0.966; CFI = 0.091; RMSEA = 0.055) (see Fig. 2) (Table 4).

Fig. 2
figure 2

Final structural equation model

Full size image
Table 4 Final structural equation model results
Full size table

According to the model hypothesized, we found that the influence of the amount of care on grandparents’ mental health was mediated positively by grandparents’ character strengths and social support. However, the effect of care on physical health was only mediated by character strengths, but not by social support (see Table 5). We also expected a direct influence of grandparents’ caregiving on social support and character strengths. However, we only found a direct effect of grandparents’ caregiving on character strengths while the effect of caregiving on social support was mediated by character strengths.

Table 5 Direct and indirect effects using bootstraping
Full size table

As can be observed, character strengths were associated directly and indirectly with physical and mental health (HRQoL) and social support (number of contacts and satisfaction). The number of interpersonal social relationships was not related to HRQoL. However, the higher the number of contacts grandparents had, the greater their satisfaction with social support. Grandparents showed higher levels of mental health when social support satisfaction was better. Regarding socio-demographic characteristics, younger grandparents and maternal grandparents gave more care. Younger grandparents also showed higher levels of character strengths and physical health. Men showed higher levels of HRQoL (mental and physical) than women. Finally, grandparents with more grandchildren showed a higher level of social support but did not exhibit more satisfaction.

To analyze whether the model works both for regular and occasional supplementary care providers we developed a multi-group analysis. The chi-square differences in this study were not significant (p = 0.19). We also calculated CFI differences following Cheung and Rensvold’s (2002) suggestions. As a result, CFI differences (ΔCFI = 0.001) were lower than the criteria established by these authors (0.01). That is, the model did not differ by grandparent groups and hence the model proposed in Fig. 2 works for both occasional and regular caregivers.

Discussion

Since many grandparents are providing supplementary childcare in Western countries, it is important to consider how caregiving may be associated with their HRQoL. The main contribution of our study was finding that character strengths and social support act as protective factors that mediated between the participants’ amount of care provided for their grandchildren and their HRQoL based on the Resiliency Model of Family Stress, Adjustment and Adaptation of McCubbin (1993). Moreover, these results were applicable for both regular and occasional grandparents.

When comparing grandparents by care intensity, regular caregivers showed higher scores in physical HRQoL. In line with the Role Enhancement Theory, Fuller-Thomson et al. (2014) stated that looking after a grandchild can re-energize regular caregivers and help them remain active. The higher scores in physical HRQoL in regular caregivers could also be related to the fact that the grandparents in this group were significantly younger compared to the group of grandparents providing occasional care. Literature has extensively supported the higher engagement in caregiving of younger grandparents, who generally show better levels of health and energy (Danielsbacka et al., 2019; Di Gessa et al., 2015; Ku et al., 2013). As our study is not a longitudinal one, we cannot establish causal relationships.

There was a higher percentage of women in the regular caregivers’ group. Other studies have found higher levels of care and greater role satisfaction in grandmothers because they are expected to provide care to a greater extent than grandfathers (Arpino et al., 2018). This difference may be related to traditional gender roles, in which women are expected to perform a caregiver role (Sharma et al., 2016; Troll 1983). Also, there was a higher percentage of maternal grandparents in the regular caregivers’ group compared to the occasional one. These results could be related to the matrilineal advantage in Mediterranean countries, according to which the ties with the maternal lineage tend to be strengthened (Roberto & Stroes, 1992).

Younger grandparents showed higher levels of physical health and character strengths. Being older is considered a risk factor for a lower HRQoL (Etxeberria et al., 2019; Tajvar et al., 2008). Danielsbacka et al. (2019) suggested that a grandparent’s life stage must be considered when analyzing the relationship between childcare and health over time, after finding that providing more frequent childcare significantly reduced limitations in activities of daily living (physical health) among the youngest grandparents but increased limitations in activities of daily living among the oldest grandparents. Regarding character strengths, Baumann et al. (2020) found positive associations of most character strengths with age. However, younger people showed a more positive impact when using character strengths. Since there is a lack of studies conducted in older adults on this topic, further research should address these issues.

In line with previous research, men showed higher levels of HRQoL (mental and physical) than women (Gouveia et al., 2018; Orfila et al., 2006; Zamberletti et al., 2018). Finally, grandparents with more grandchildren showed more social interactions, but did not exhibit more satisfaction. Rico et al. (2001) state that what increases grandparents’ satisfaction is not associated with the number of grandchildren, but with having close relationships with them.

When analyzing the relationships between the outcome variables, the model tested based on the Resiliency Model of Family Stress, Adjustment and Adaptation (McCubbin, 1993) showed an excellent data fit to the model hypothesized for both regular and occasional caregivers. According to our hypothesis, character strengths mediated the effect of care on both physical and mental HRQoL. They also showed a direct effect on HRQoL. To our knowledge, this is the first study that analyzes character strengths in grandparents. Studies developed with other samples found similar results. Proctor et al. (2011) also found positive effects of character strengths’ use on university students’ perceived health, while Peterson et al. (2006) found that participants who recovered from serious illness showed a higher use of character strengths. These results highlight character strengths’ protective role as a mediator between the amount of care provided and HRQoL. In this regard, theoretical models have also supported the importance of cognitive resources as stress mediators. For example, Musil et al. (2009) found that resourcefulness was a mediator between stress and mental health for grandmothers (caregivers, non-caregivers and those living in multigenerational families).

Character strengths showed a direct effect on social support. This result supports Martínez-Martí and Ruch’s (2014) statement in which character strengths can contribute to creating or maintaining significant social relationships. In contrast, and contrary to our expectations, social support did not show any direct or indirect effect on character strengths. This result is in line with Snyder and Lopez (2009). These authors consider character strengths as internal variables, similar to personal traits that are part of the person’s background. Social support is an external variable that may enhance the effect of character strengths on health variables by increasing the available resources that help to express the character strengths in behaviors or actions.

Social support mediated the effect of the amount of care on grandparents’ mental HRQoL, in line with the results found by Musil et al. (2009) in primary caregivers and non-caregivers’ grandmothers. These results highlight the importance of having high levels of social support. According to Musil et al. (2009), social support may provide positive social experiences, increase role rewarding and improve the ability to cope with stressful events. Satisfaction with social support also showed a direct effect on mental HRQoL, while the number of social contacts did not. Recent research has not found a significant relationship between total social contact and HRQoL, concluding that the quality of social support has a higher impact on HRQoL in older adults than quantity does (Lu et al., 2020). Contrary to our expectations, social support did not have any effect on physical HRQoL. Other types of social support that have shown a relationship with physical health, such as instrumental social support, were not analyzed. Future research should analyze the relationship between social support and health in greater depth, considering the different dimensions of social support and using standard criteria.

We must mention some limitations. The first limitation is recruitment of volunteers, yielding a sample of convenience. It may be that only grandparents who were engaged in positive relationships with grandchildren and experiencing resilience volunteered. Second, this is a cross-sectional study that does not allow us to make causal inferences and establish predictions. Longitudinal studies are needed to ensure the stability of these results. Third, future studies must consider contextual factors. Southern European countries promote caregiving for children within the family, while Nordic countries show higher institutional child-care services. The impact of care on grandparents’ health may differ depending on socio-demographic factors and motives for caregiving. Finally, data was self-reported, and this may have increased participants’ social desirability. Future studies should include other reliable measures, such as participants’ objective health or instrumental support provided by healthcare professionals.

Implications for Practice

These results support the assertion that the amount of care by itself does not negatively affect supplementary caregivers’ HRQoL. Character strengths mediated between grandparents’ amount of caregiving and their physical and mental HRQoL. Social support only mediated between the amount of caregiving and mental HRQoL. These results highlight the importance of considering protective factors (i.e., social support and character strengths) when developing policies directed to grandparents who provide supplementary care to their grandchildren.

Our findings also extend our understanding of the variables that professionals could consider while developing intervention programs for grandparents who provide supplementary care. For example, intervention programs that promote the use of character strengths associated with active coping strategies may help grandparents to cope with possible care difficulties. Also, support group interventions, in which grandparents may exchange reflections upon their experiences with other grandparents, can help enhance grandparents’ social support. Regarding this, some researchers have found that support group interventions focused on custodial grandparents reduced distancing as a coping strategy (Burnette, 1998) and their sense of social isolation (Kelley et al., 2001), while increased social support (Kelley et al., 2001, 2007) and encouragement (Kelley et al., 2001). However, these interventions have been developed with custodial grandparents who perform a parental role with their grandchildren. Since there is a lack of interventions methodologically and rigorously focused on grandparents providing supplementary care to their grandchildren, future studies should test these hypotheses.

Finally, the development of policies and programs could contribute to the improvement of supplementary caregivers’ role rewards and the strengthening of family ties. Some examples could be the following: helping grandchildren’s parents balance work-family balance; recognizing the role played by supplementary caregivers in society; increasing subsidized formal child-care services; maximizing more flexible services adapted to families’ needs; increasing parental leave benefits and part-time opportunities; developing institutional campaigns that aim to change the negative stereotypes associated with older adults; or promoting intergenerational programs. Child-care activities are a way to achieve active aging and contribute to current families and society. However, grandparents should be able to make clear to their adult children how much care they are able to or want to provide for their grandchildren. Working on grandparent-adult-child communication is needed to set a limit on the demands placed on grandparents, state child-care conditions and, in turn, reduce grandparents’ burdens and misunderstandings among generations.

Conclusion

There is limited literature that analyzes what may be mediating the relationship between care provided by grandparents and HRQoL and most of these studies are based on a negative perspective. While research must consider the negative aspects of caregiving, this study applies the Resiliency Model of Family Stress, Adjustment and Adaptation (McCubbin, 1993) to offer a positive orientation. Our results showed the protective role of character strengths and social support as mediators among the amount of care provided for grandchildren and grandparents’ HRQoL. This focus on positive aspects may contribute to the development of future studies focused on developing effective family intervention programs that prevent grandparents’ vulnerability by promoting protective factors and building resilience. There is a need to develop future studies tailored to grandparents providing supplementary care for their grandchildren to know what specific character strengths protect their mental and physical health. In addition, grandchildren’s and adult children’s perceptions should be analyzed to develop intergenerational programs that consider the viewpoints of the three generations. Finally, it is critical to develop intervention programs methodologically and rigorously tested in longitudinal randomized controlled trials.