Grandparenting and psychological well-being: how important is grandparent role centrality?

Abstract

The purpose of the study was to examine the association between the centrality of the grandparent’s role and psychological well-being in the population of European grandparents, aged 50 and older. The study was a secondary analysis of data gathered in the first wave of the Survey of Health, Ageing and Retirement in Europe. The sample included grandparents who looked after their grandchildren without the presence of the parents (n = 3,888). Respondents’ scores on the Center for Epidemiological Studies Depression Scale were regressed on the frequency of contact with grandchildren, beliefs and attitudes about grandparenting and grandparent-focused role occupancy, and on a composite measure of grandparent role centrality, controlling for age, gender, education, household income, physical status, and country. The findings negated the hypothesized positive association between grandparent role centrality and psychological well-being. Only one of the three indices of interest—grandparent-focused role occupancy—was related to the psychological well-being outcome when background and health variables were taken into account. The study refutes the central claim of role theory according to which salient roles are more beneficial to the psychological well-being of the individual than are other roles, especially in old age. It also questions the theoretical framework of grandparent role meaning that is commonly cited in the literature.

Significant changes in the socio-demographic profile of the aging population raise questions about the role of grandparenting in the discourse about successful aging. In today’s world, people can expect to spend several decades in the grandparent role during some of which they are still both employed and/or otherwise active in many areas of their lives. Various studies have examined the associations between selected aspects of grandparenthood and the psychological well-being of grandparents. The results of most of these studies tend to indicate that grandparents and grandchildren engage in frequent interaction and that the relationships between them are characterized by both satisfaction and by emotional closeness (Drew and Silverstein 2007).

The grandparent role is directly linked to the distribution of responsibility between the family and the welfare system in caring for the young generation. It is surprising, therefore, that relatively few studies specifically address the changing role of the grandparent in the context of modern society. It is even more curious that only a small number of studies focus on contemporary noncustodial grandparenting, even though it is a far more common occurrence. The current analysis attempts to fill this gap in the existing body of knowledge regarding normative aspects of noncustodial grandparenthood. In particular, it examines the centrality of the grandparent’s role in the life of grandparents and its association with their psychological well-being. For this purpose, this study employed data from the first wave of the Survey of Health, Aging and Retirement in Europe (SHARE). The internationally comparable and multidisciplinary database provided by SHARE allows the consideration of multidimensional aspects of the grandparent role within a single analysis.

Theoretical background

The main theoretical perspectives that inform the inquiry into contemporary grandparenting are role theory and identity theory. A role is a structural position that an individual holds within a social group, whereas identity is a subjective self evaluation of this role. Studies that have been done on role fulfillment have found that individuals invest their social roles with commitment, self-esteem, identity meaning, and salience (Burnette 1999; Reitzes and Mutran 2002; Reitzes and Mutran 2004; Thiele and Whelan 2008). Moreover, people do not merely perform a social role but actively participate in the process of defining it by subjectively infusing it with personal significance. By adding meaning and value to a certain role, an individual channels his or her personal preferences and priorities. A role, thus, becomes integrated into one’s self concept.

Symbolic interactionism posits that the construction of one’s self-concept is a continuous process that occurs throughout the life course during which some roles become more central to the individual’s self-concept than others. This is because of the fact that people organize their roles and identities into a hierarchy of relative importance to their self-concept. More salient roles are given more physical and material resources as well as personal investments of time, energy, and commitment (Reitzes and Mutran 2002). In respect to the study of grandparenthood, much of the research relies on the conceptual framework developed by Kivnick (1983) in which the centrality of the grandparent role is the first among five dimensions used to explore the self concept and the role meaning of grandparents.

There is empirical evidence that salient roles have greater impact on the mental health of individuals than less central ones. Drew and Silverstein (2004) investigated psychological well-being of family members, intergenerational role identities, and investment in family roles. They found evidence that both role hierarchy and finding meaning in role enactment were linked to positive well-being. Martire et al. (2000) found that greater centrality of four major social roles, such as mother, wife, employee, and parent, was associated with better psychological well-being.

The literature also notes that the salience of a particular role is measured on several dimensions that usually include at least one behavioral aspect of participation in a role and an affective–cognitive assessment (Peterson 1999). Some studies include behavioral and affective dimensions only (Perrone and Civiletto 2004), whereas others emphasize role definition and the variety of role enactment styles (Silverstein and Marenco 2001). The behavioral dimension refers to the amount of time spent in the role. A number of studies on grandparenthood agree that activities performed with the grandchildren do indeed represent a key behavioral dimension of the grandparent role (Peterson 1999; Hayslip et al. 2003; Reitzes and Mutran 2004; Thiele and Whelan 2008). This dimension—often referred to in the literature as “frequency of contact”—may vary from visiting and socializing occasionally to providing full-day child care as a form of active support to the parents. Studies that focused on the contact between grandparents and their grandchildren in the English-speaking culture found that 60% of American and 42% of Australian grandparents report contact on at least a weekly basis (Thiele and Whelan 2008).

A number of studies report a significant association between frequency of contact with grandchildren and life satisfaction among grandparents as well as positive assessment of the grandparent role (Peterson 1999). Reitzes and Mutran (2004) found that the centrality of the grandparent role was positively related to frequency of contact for grandfathers but not for grandmothers. Nevertheless, it was equally positively related to grandparents’ role satisfaction for both genders.

As noted earlier, another relevant dimension of grandparenthood involves the attitudinal aspect. The literature notes that beliefs and attitudes that individuals hold grant instruction and guidelines in role enactment. Jendrek (1994) investigated grandparents who provided both formal and informal care for their grandchildren and found that grandparents are ready to assume responsibility and even to become the sole caretaker for underage children because they feel that they are obligated to do so. It is the perception of their familial duty that prompts them to take this demanding role despite the detrimental consequences that it may have to their well-being. Another study that investigated grandparents’ perceptions and financial obligations to grandchildren and step grandchildren found that grandparents view the relationship with the young generation as continuous and firmly embedded in multigenerational family ties (Ganong and Coleman 1998). It was this perception that directed their decision-making process about the financial transfers to the grandchildren. These beliefs were found to be firm and stable over time and were not affected by strenuous life events, such as parents’ divorce.

An additional study that refers to the same dimension of grandparenthood focuses on the population of grandparents who provide full-time care for their grandchildren without the presence of the parents (Hank and Buber 2009). The study was based on the findings from the 2004 SHARE and investigated cross-national variations of grandparents’ child care in ten European countries. In addition to questions about frequency of provided child care, the respondents were asked how much they agree with a statement that it is a grandparent’s duty to help grandchildren’s parents in looking after young grandchildren. The findings of the study reveal high compliance with the statement in Mediterranean countries as well as in Germany and France, and much weaker support in Denmark and Netherlands, even among grandparents who frequently cared for their grandchildren.

In contrast, the probability of providing some kind of child care was highest among Danish, Dutch, French, and Swedish grandparents and lowest among the Mediterranean grandparents. Yet, the frequency of child care among grandparents who supplied this form of family support was high among Mediterranean countries and low among grandparents from Nordic countries and France. The investigators present a few explanations for this phenomenon. First is the North/South divide in the interpretation of what it means to look after young grandchildren. Second is the high prevalence of multigenerational co-residence as a norm of transferring resources between generations in Southern Europe, as opposed to Northern Europe where family support rarely involves co-residence and, thus, is less intense. Finally, variations in public child care provision and in maternal labor force participation between the North and the South are also seen as partly responsible for the differences in grandparents’ support.

Another approach to the study of role salience is to consider it in terms of role hierarchy. Studies on the topic tend to place particular emphasis on multiple role occupancy in older age. The assumption is that roles tend to shrink in scope during the later stages of life, leaving an older person with only tenuous roles and deprived of social identity (Adelmann 1994). On the other hand, there is evidence that despite role loss and various changes that are typically associated with later life, many older people continue to be productively involved in multiple roles. This contrariety may be why the relationship between multiple role occupancy and psychological well-being is somewhat ambiguous.

The literature on multiple role occupancy reports two main perspectives: role enhancement versus role strain. According to the role enhancement theory, people with multiple roles enjoy higher levels of psychological well-being because of an increase in the feeling of personal power, resources, and emotional gratification (Adelmann 1994; Ahrens and Ryff 2006; Nordenmark 2004; Reid and Hardy 1999). The alternative perspective, however, emphasizes the cost of many, sometimes contradictory, demands placed on the individual that result in role conflict and role overload, thus reducing the level of psychological well-being (Perrone and Civiletto 2004; Rozario et al. 2004).

A study based on a large national sample of American women investigated functional role involvement of grandmothers who occupied multiple social roles and varied in their characteristics. Its findings indicated that caring for grandchildren on a regular basis (but not being a primary caregiver) did not interfere with the maintenance of other social roles, such as employee, volunteer, member of a community organization, and so on. It was especially true for young grandmothers, who were married, healthy, had higher education, and were active in the labor market. These women were found to be highly involved in multiple roles with no indication for compromised levels of psychological well-being (Baydar and Brooks-Gunn 1998).

To summarize, role salience is a key component in the self concept of the individual. It is especially true in the later stages of life when the number of social roles available to a person decreases and when role content can become tenuous. The literature notes that because family is an important social institution that not only nourishes and assists the individual but also confirms one’s personal identity, intergenerational relationships in later life have special importance (Drew and Silverstein 2004). The current analysis attempts to examine the nature of grandparent role centrality and its relationship to psychological well-being among noncustodial grandparents. Based on the literature, it posits that a multidimensional approach to investigating the role salience of grandparents is appropriate. It thus offers a multidimensional model of grandparent role centrality in which “centrality” is constructed from three separate dimensions of grandparenthood that have not yet been combined into one single analysis: (a) the behavioral aspect as expressed by frequency of contact with grandchildren, (b) the cognitive aspect as expressed by beliefs and attitudes about grandparenting, and (c) grandparent-focused role occupancy as expressed by the number of social roles in addition to grandparenthood that a respondent holds (the greater the number and frequency of roles, the less the extent of grandparent-focused role occupancy). The main hypothesis of this analysis is that the centrality of the grandparent role is positively associated with the psychological well-being of the grandparent, meaning that those who have placed the role on a higher level within their role hierarchy will tend to report higher levels of well-being.

Method

The analysis is based on data from the first wave of the SHARE survey from 2004 (release 2.0.1). The first-wave data include persons aged 50 years and older and their spouses regardless of age from 11 European countries: Denmark, Sweden, Austria, Germany, France, Switzerland, Netherlands, Belgium, Spain, Italy, and Greece. The data were collected through face-to-face computer-assisted personal interviews that were complemented by a self-completion questionnaire. The questionnaire was designed such that there were three types of respondents: financial, household, and family. Financial respondents answered the questions on assets and financial transfers. Household respondents responded to questions on household features, including consumption. Family respondents provided information regarding children and social support, including questions related to grandparenting. In single-person households, the respondent answered all questions, that is, he/she was the financial, household, and family respondent. In multiperson households, survey questions were asked differentially. Usually, the first-responding respondent in the household was chosen to be the family respondent as well.

Sample

Sampling design in each SHARE country varied, insofar as institutional conditions and regulations regarding sampling were different across the countries (Klevmarken et al. 2005). The respective sampling frames were selected in relation to the available infrastructures in each county. All samples were drawn through probability sampling. The sampling procedures varied from simple random sampling to multistage designs. The most common design was two-stage sampling, with geographical areas (generally municipalities) used as primary sampling units, and households or individuals as secondary sampling units. Simple random sampling was applied in Denmark and Sweden, and three-stage sampling was employed in Austria, Belgium, Greece, and Italy.

The entire 11 country sample encompassed 22,377 respondents (http://www.share-project.org/t3/share/fileadmin/pdf_documentation/Tabelle_Sample_Release2.0.1_Wave1.pdf). The overall response rate was about 62%. This was similar to the response rate reported in the 1994 European Community Household Panel (62%) and the 1996 European Labor Force Survey (63%), and higher than that which was obtained in the relevant countries in the European Social Survey in 2004 (55%) (De Luca and Peracchi 2005).

The sample for the current analysis was limited to family respondents who answered both the computer-assisted personal interviews and the self-completion “drop-off” questionnaire (the response rate for the drop-off was about 85%). This group numbered some 13,873 individuals. From this group, 4,460 respondents who were not grandparents were excluded, thus leaving 8,579 grandparents.

This study sample included only grandparents aged 50 years and older who provided a positive answer to the following question: “During the last 12 months, have you regularly or occasionally looked after your grandchild/grandchildren without the presence of the parents?” In SHARE, only persons who responded positively to this probe were asked two follow-up questions on frequency of contact and hours of care, variables that were necessary for the current analysis. Respondents with missing data on both the follow-up questions (9.8%) were also excluded from the current inquiry. Thus, the final study sample numbered 3,888 individuals.

Variables

The main independent variable of the analysis is the centrality of the grandparent role. This variable displays the combined effect of three separate dimensions of grandparenthood, each represented by a separate variable: frequency of contact with grandchildren, beliefs and attitudes about grandparenting, and grandparent-focused role occupancy.

“Frequency of contact” was constructed from two questions: “On average, how often did you look after the grandchildren in the past 12 months?” (from “Almost daily” to “Less often than once a month”) and “About how many hours did you look after your grandchildren on a typical day/week/month?” The number of hours given was multiplied by 30 if the contact was daily, by 4 if it was weekly, and left untouched if it was monthly. The number of yearly hours was divided by 12. Respondents who reported frequency of contact, but failed to report the number of hours per contact, were assigned mean values for that category of frequency.

Beliefs and attitudes were measured by three statements about grandparenting: (1) Grandparents’ duty is to be there for grandchildren in cases of difficulty (such as divorce of parents or illness), (2) Grandparents’ duty is to contribute toward the economic security of grandchildren and their families, and (3) Grandparents’ duty is to help grandchildren’s parents in looking after young grandchildren. The respondent was asked to indicate the degree of agreement with each of these statements on a scale that ranged from 5 (strongly agree) to 1 (strongly disagree). Internal reliability between the three items, as measured by Cronbach’s alpha, was sufficient to relate to the summary scores as a scale (α = 0.76). Thus, the higher the summary score, the stronger the agreement with the grandparents’ duty.

The grandparent-focused role occupancy variable was measured by combining two related items. First, the respondent was asked to indicate how many other activities from a given list of seven he or she had done in the last month. These included, for example, doing volunteer or charity work, going to a sport, social or other kind of club, and taking part in a religious organization. Second, the respondent had to indicate the frequency of the engagement in each of the indicated activities: daily (3), weekly (2), or less often (1). The product of activities and frequency yielded a composite score of “multiple role occupancy” in which the highest potential score for a person who engaged in all seven activities on a daily basis was 21. In the current sample, the scores ranged from 0 to 14. To measure grandparent-focused role occupancy, the scores were reverse coded such that the higher the score, the lower the extent of other activity. Stated differently, respondents with the highest score on the grandparent-focused role occupancy variable reported having engaged in no additional activity beyond that of grandparenting.

The main variable of grandparent role centrality was calculated in two stages. First, the three measurements of grandparenthood—frequency of contact, the summary score for beliefs on grandparenting, and the grandparent-focused role occupancy measure—were standardized by calculating a logarithm, a procedure commonly used to overcome differences in scales between variables (Zeltzer-Zubida and Spalter 2008). In the second stage of the calculation, the standardized measurements were summed to reflect the cumulative effect of the centrality of the grandparent role. Thus, the higher the summary score, the more central the grandparent role in the life of the respondent.

Psychological well-being was measured on an adapted abbreviated version of the Center for Epidemiological Studies Depression Scale (CES-D) (Radloff 1977), an instrument widely used in such research. Of the 20 original items in the scale, 11 were kept and 3 others were added, as in the baseline questionnaire of the Health and Retirement Study in the United States. The 11 CES-D items were: felt depressed, everything was an effort, sleep was restless, happy, lonely, people were unfriendly, enjoyed life, felt sad, felt people disliked me, could not get going, and appetite was poor. The three additional items were: had a lot of energy, felt tired, and felt really rested when I woke up in the morning. Respondents were asked to indicate how frequently they experienced each symptom during the previous week on a scale of 1–4. A higher summary score on this measure indicated greater depression (α = 0.82). In this analysis, a log transformation of the CES-D measure was used to overcome the potential skewedness that is commonly associated with this measure.

Background characteristics were tapped through six variables: (a) age; (b) gender (men = 1, women = 2); (c) economic status, as measured by gross household income; (d) level of education, classified using the International Standard Classification of Educational Degrees 1997 of the United Nations Educational Scientific and Cultural Organization; (e) country of residence; and (f) health status, as measured on five question regarding the respondent’s ability to perform five actions: pulling a large object, stooping, lifting, reaching, and picking up a small coin. Similar to the Physical Activity Scale used in the Yale Health and Aging Project (Cornoni-Huntley et al. 1985), this last scale tests self-reported physical limitations. Difficulty in performance on each item was coded as 1; no difficulty was coded as 0. The resultant summary scale score was an average of the five items and ranged from 0 to 5, the higher the score, the greater the functional health limitation.

Analysis

The analysis progressed in three stages. First was a univariate description of all the study variables. Second, bivariate analyses tested associations between the study variables and the well-being outcome measure—depression. ANOVA was employed for the country of residence variable and Pearson’s correlations were used for all the others. In the final stage, multivariate hierarchal regression was performed to test the unique associations between the grandparenthood variables and the well-being outcome. The regression was done using three main models and two supplementary ones. In the first model, the background variables were entered. In the second model, to test the added value of grandparenthood to well-being, the three variables representing the separate dimensions of grandparenthood were added. In the third model, the combined centrality variable replaced the three separate dimensions of grandparenthood in the regression. Additional models (4 and 5) entered interaction terms.

In all the multivariate analyses, country was controlled to neutralize possibly differing national effects on the psychological well-being outcome measure. It was not the aim of the current inquiry to explain country differences in relation to depressive symptoms, to the extent to which they may exist, but rather to clarify the unique relationship between certain aspects of grandparenting and psychological well-being. Thus, procedures appropriate for such cross-cultural analysis were not employed here.

Results

Overall, 58% of the respondents in the sample were female and 42% were male. The average gross annual household income was 43,358 Euros (SD = 45,893). Thirty-one percent reported having elementary school education or less, 22% had partial high school education, 29% had a high school diploma, and 18% had post secondary and higher education. More than 63% of the respondents had no physical limitations, and 21% reported having only one limitation. At the time of the survey, 23% were still actively employed.

Table 1 presents the univariate analysis of the grandparenthood variables. The table reveals that although frequency of contact with the grandchildren could reach up to 720 h per month (that is, 24 h per day times 30 days per month), the reported monthly average was about 57 h. The median shows, moreover, that some 50% of the grandparents took care of their grandchildren 20 h a month or less, corresponding roughly to about 2½ full days. In respect to the beliefs and attitudes that respondents held about grandparenthood, the table shows that there was a fair amount of agreement with all three statements about the duties of grandparents. The median score of 4 out of 5 reflects that at least 50% of the respondents expressed net agreement (a score of 3 meant neither agree nor disagree, and lower scores were net disagreement). Moreover, the mean for the entire sample was almost identical in the rate of agreement.

Table 1.

Univariate analysis of grandparenthood dimensions

Grandparenthood dimensions	%	Range	Mean	SD	Median
Frequency of contact (hours) with grandchildren		0–720a	56.8	105.0	20
Attitudes toward grandparenting		1–5	3.9	0.8	4
Social roles and activities		1–14	3.0	2.1	3
Providing help to family, friends, or neighbors	34
Gone to a sport, social, or other kind of club	23
Done voluntary or charity work	16
Taken part in a religious organization	15
Attended educational or training course	8
Cared for a sick or disabled adult	7
Taken part in political/community organization	5
The centrality of the grandparent’s role		0–10.2	4.9	2.0	5.1

^aThe maximum, 720 h, reflects 24 h per day in a 30 month day

Initial analysis of the grandparent-focused role occupancy variable (not shown in the table) revealed that 38% of the respondents were not engaged in any social activities outside the family circle. The remaining 62% who were engaged in any other activity did so less frequently than once a week. Among them, Table 1 shows that the most prevalent activity was providing help to family, friends, or neighbors (34%), followed by attending a sport, social, or other kind of club (23%). Other activities had lower rates of participation. The combined activity score, which takes into account both range and frequency of activity, showed a relatively limited degree of involvement on the part of the respondents. This was because both the mean and the median reflected scores of 3 on a scale of 1–14.

Table 1 also shows the results of the composite measure of centrality of the grandparent’s role in the current study sample. As may be seen, the median (and to a slightly lesser degree, the mean) fell about midway in the score range. This suggests that grandparenthood centrality existed to a moderate degree among respondents. Additional analysis (not shown in the table) revealed that less than 10% of the respondents could be said to have reported a high degree of centrality of the grandparent role in their lives, given their relatively high scores on this measure.

The bivariate associations between the dimensions of grandparenthood, background variables, and psychological well-being are presented in Table 2. As may be seen, all the study variables were related to well-being, but not necessarily in the direction hypothesized. Older respondents, women, and respondents who had more physical limitations displayed more depressive symptoms. Those with higher levels of education and higher household income displayed fewer depressive symptoms. The association between well-being and the country of residence variable was also found to be significant. Accordingly, this variable was entered into the multivariate stage of analysis to neutralize possible country-based differences.

Table 2.

Bivariate associations between background and grandparenting variables and depressive symptoms among grandparents

Variable	Depressive symptoms
	R	F
Age	0.06**
Gender	0.15**
Education	−0.04**
Income	−0.13**
Physical limitation	0.30**
Country of residence		9.2**
Frequency of contact	0.06**
Attitudes toward grandparenting	0.07**
Grandparent-focused role occupancy	0.10**
Centrality of grandparent’s role	0. 05*

* P < 0.05; ** P < 0.01

As for the grandparenthood variables, respondents who had more frequent contact with their grandchildren actually displayed more depressive symptoms, as did those who expressed higher compliance with the statements about grandparents’ duties and those who had higher role salience. Specifically, grandparent-focused role occupancy was positively associated with the number of symptoms, meaning that grandparents who were less engaged in other social roles beyond the role of grandparent were more depressed. Finally, the summary centrality measure was positively (but weakly) related to the depression outcome measure. That is, the more central the role of grandparent, the greater the number of depressive symptoms.

The multivariate analysis is presented in Table 3. Model 1 shows that the number of depressive symptoms was positively related to female gender and to physical limitation, and negatively related to household income. Country of residence showed differing associations by most of the countries, with Austria serving as the reference category. Model 2 reveals that after controlling for background and physical limitation, the variables frequency of contact with grandchildren and attitudes toward grandparenting were not related to grandparents’ well-being. Grandparent-focused role occupancy retained its earlier positive bivariate association with depressive symptoms, albeit to a lesser degree. However, the effect size for the addition of the variables in model 2 was minimal (Cohen’s f ² = 0.002).

Table 3.

Multivariate analyses of background and grandparenting variables and depressive symptoms among grandparents

Variables	Model 1 β	Model 2 β	Model 3 β	Model 4 β	Model 5 β
Age	0.02	0.02	0.02	0.01	−0.02
Gender	0.10***	0.10***	0.10***	0.10***	0.10***
Education	−0.02	−0.02	−0.02	−0.02	−0.02
Income	−0.07***	−0.07***	−0.07***	−0.07***	−0.07***
Physical limitation	0.25***	0.25***	0.25***	0.25***	0.25***
Germany	0.04	0.04	0.03	0.03	0.03
Sweden	−0.09***	−0.09***	−0.09***	−0.09***	−0.09***
Netherlands	0.08***	0.08***	0.08***	0.08**	0.08**
Spain	0.08***	0.07***	0.08***	0.08***	0.08***
Italy	0.09***	0.09***	0.09***	0.08***	0.08***
France	0.03	0.03	0.04**	0.04**	0.04**
Denmark	−0.09***	−0.09***	−0.07***	−0.10***	−0.10***
Greece	0.09***	0.09***	0.09***	0.08***	0.08***
Switzerland	−0.01	0.00	0.00	0.00	−0.01
Belgium	0.10***	0.10***	0.10***	0.10***	0.10***
Frequency of contact		0.00
Attitudes toward grandparenting		0.01
Grandparent-focused role occupancy		0.04***
Centrality of grandparent’s role			0.00	−0.02	−0.05
Age × centrality				0.02
Gender × centrality					0.07
R²	0.173	0.175	0.175	0.175	0.175

** P < 0.01; *** P < 0.001

Most importantly, as may be seen in model 3, the composite measure for the centrality of the grandparent role was unrelated to the well-being outcome when other factors were taken into account. That is, the centrality of the grandparent role in the life of the respondent did not predict his or her psychological well-being. The strongest predictor was respondents’ physical status: the greater the extent of limitations, the more depressive symptoms. Both gender and income variables were the next strongest predictors with women and respondents with less income experiencing lower levels of well-being, respectively. Interestingly, age and level of education were not significantly related to the well-being outcome. Models 4 and 5 reveal, moreover, that the respective interactions of the demographic characteristics of age and gender with the composite measure of centrality had no association with the well-being outcome.

Discussion

This study investigated the association between the centrality of the grandparent’s role and psychological well-being in the population of European grandparents, aged 50 and older, from 11 countries. The findings negated the hypothesized positive association between grandparent role centrality and well-being. That is, only one of three indices of grandparenthood was significantly related to the well-being outcome addressed in this study—depressive symptoms—when background and health variables were taken into account as well. Moreover, the specific aspect that was related—grandparent-focused role occupancy—was positively associated with depressive symptoms; the greater the grandparent-focused role occupancy, the greater the number of depressive symptoms.

As recalled, role centrality in this analysis reflects the cumulative effect of three indices of grandparenthood covering a functional–behavioral aspect, a cognitive aspect, and a measure of role occupancy. Contrary to the literature, neither the behavioral aspect of grandparenthood nor the cognitive aspect was found here to be related to the well-being outcome when the measure of grandparent-focused role occupancy was taken into account as well. And, as mentioned, the combined measure of grandparent role centrality was not ultimately related to psychological well-being either.

One possible explanation for these results is that intergenerational contact in modern society diminishes as competing values come to replace those based on family solidarity. Gauthier (2002) and Silverstein and Marenco (2001) point out, in this context, that modern grandparents have many of the prevailing characteristics of the middle class. They are healthier, more affluent, and more educated than were grandparents of the previous generation. Moreover, they value personal independence and social involvement outside the family network and tend to rely on extra-familial sources, such as the workplace and social circles, for their sense of personal worth. Consequently, the rewards of the grandparent role tend to be more symbolic and less functional. It is possible, though, that people find it easier to formally comply with the support norm in respect to “proper” grandparenting, whereas their actual behavior reveals different child care patterns. This explains why the majority of grandparents in the current research agree with the statements about grandparents’ duty, while in practice a smaller proportion of them frequently interact with their grandchildren without the presence of the parents.

This study showed that the dimension of grandparent-focused role occupancy was the only dimension significantly associated with psychological well-being, and its association was negative (that is, it was positively related to depressive symptoms). This finding is congruent with the recent literature on multiple roles and activities in later life and provides a confirmation of the role enhancement perspective, which states that the more roles one holds the greater one’s well-being. In the same vein, studies have found that age identity is also affected by multiple roles. That is, people who are actively involved in a number of social roles usually believe that old age starts later than those with fewer roles available to them. They also view themselves as more competent and express a firmer belief that they will reach a very old age (Kaufman and Elder 2003).

Last, examination of interactions between the centrality variable and age and gender, respectively, showed no association. This means that the centrality of the grandparent’s role is unrelated to the grandparent’s psychological well-being regardless of age and gender. To summarize, the findings of this study showed that respondents who have only one social role available to them, that of a grandparent, scored lower on psychological well-being than the respondents who had at least one additional role. More than half of the grandparents in the analytical sample were actively involved in at least one social activity besides grandparenthood, and the more activities grandparents had, the better their psychological state.

The findings presented in this study were reinforced by additional analyses not shown in the article because of the lack of space. Parallel regressions were run for two additional well-being outcome measures—life satisfaction and the CASP Scale (Knesebeck et al. 2007). In both cases, similar results were obtained. Thus, we can confirm with confidence that the findings reported here are sufficiently robust.

Nevertheless, the present analysis has a few methodological limitations that should be noted. First, as is frequently the case in secondary analysis, the questions in SHARE were not designed specifically for the purpose of the current analysis and it was not possible to inquire further on additional topics that were not addressed. Thus, it was not possible to distinguish between custodial and noncustodial grandparents among the responding grandparents. In addition, the literature cites other characteristics that have been shown to be associated with grandparents’ psychological well-being and role enactment, such as grandchildren’s age, their number, and geographic proximity to the grandparents, and the parents’ health. These topics could not be addressed in this study and were left unexplored.

Another limitation was that to consider frequency of contact, information was available only for grandparents who took care of their grandchildren without the presence of the parents. Those that visited with their grandchildren only in the presence of the parents were excluded from the final sample. It is indeed possible that some of the latter group invest in their role as grandparent as well and have high role salience. Thus, further research needs to distinguish between grandparent caregiving and grandparenting per se.

A last limitation is that this analysis may have missed important country differences in contemporary grandparenting. Such differences can be explored through the use of multilevel analysis. Future studies should address the statistical dependencies in the data because respondents are, indeed, nested within countries.

Nevertheless, this study provides interesting results in relation to the question under consideration. Moreover, the multidimensional nature of grandparenthood was uniquely considered within a single analysis. This provides new insights into the interrelationships among the relevant grandparenting variables. In addition, the findings have two practical policy implications. First, the study results suggest that the scope and the content of the grandparent role should be considered by practitioners when assessing the quality of life of older individuals. Second, the results imply that caution must be used by welfare authorities when regarding grandparents as a safety net that can compensate for the shortcomings of the public welfare system. Modern grandparents may not constitute the unlimited resource that they were once believed to be.

In conclusion, this study found the centrality of the grandparent role to be largely unrelated to the psychological well-being of grandparents. It thus refutes the central claim of role theory according to which salient roles are more beneficial to the psychological well-being of the individual than are other roles, especially in old age. It also questions the theoretical framework of grandparent role meaning that is commonly cited in the literature.

The present findings illuminate the relationship between grandparent role salience and well-being in a different way. It seems that several grandparents today have shifted their focal point from family-oriented dimensions of self-concept to self-oriented ones. These dimensions, which can include parameters of health, income, and external activities, may be more related to well-being than is the enactment of family roles. Perhaps, because of the increase in longevity, “old age” has been postponed, and many modern grandparents in their 50s, 60s, and even those in their 70s maintain a characteristically younger life style. Future research on the issue of role salience and well-being among grandparents of different ages needs to take these issues into account.