Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Fam Soc. Author manuscript; available in PMC 2011 June 2.
Published in final edited form as:
Fam Soc. 2010 January 1; 91(4): 385–393.
doi:  10.1606/1044-3894.4030
PMCID: PMC3106443

Patterns and Predictors of Support Group Use by Custodial Grandmothers and Grandchildren

Gregory C. Smith, Professor, Julian Montoro Rodriguez, Professor, and Patrick A Palmieri, Director


Despite increasing popularity of support groups for custodial grandfamilies, the few published studies to date raise questions regarding actual support group usage and disregard predictors of such use. This study examined patterns of self-reported use, need, and unmet need of both grandparent and grandchild support groups. In addition, the Andersen Social Behavioral Model was used to identify predisposing, enabling, and need factors that predicted use in a national sample of 733 grandfamilies recruited by either convenience or population-based strategies. Reported use and need for support groups were lower in the population-based sample, and predictors varied by sampling strategy. The findings are discussed in terms of how they inform service providers and researchers regarding factors that influence support group use by custodial grandfamilies.

In response to rising numbers of custodial grandparents (CGP) who are raising custodial grandchildren (CGC) due to problems in the parent generation (Simmons & Dye, 2003), support groups have become the predominant intervention for assisting this population (Patrick & Hayslip, 2003). Support groups offer family caregivers education and support and typically encompass seven themes: information about the care receiver's situation; group members as a mutual support system; emotional impacts of caregiving; self-care; problematic relationships; developing and using support systems outside the group; and learning new skills (Roberto & Qualls, 2003).

Support groups are the most used source of education and support by CGP, with many regarding them as their most valued resource (Roe, 2000; Minkler, Driver, Roe, & Bedeian, 1993). In fact, three fourths of programs and services for CGP examined nationwide in the 1990s by the Brookdale Grandparent Information Project consisted solely or primarily of support groups (Minkler et al., 1993). By 2000, over 400 CGP support groups across the U.S. were registered with the American Association of Retired Persons (AARP) Grandparent Information Center (Vacha-Haase, Ness, Dannison, & Smith, 2000).

Because CGC experience an array of emotional and behavioral difficulties stemming from the traumatic circumstances surrounding the need for grandparental care (Smith & Palmieri, 2007), support groups for these children have likewise been advocated by human service professionals and CGP alike (Dannison & Smith, 2003; Landry-Meyer, 1999; Smith, 2003; Smith, Savage-Stevens, & Fabian, 2002). Child support groups are composed of children who share a common status or predicament that produces stress with the goal of minimizing this stress through mutual support and sharing of coping strategies, information, and confidence (O'Rourke & Worzbyt, 1996). Although there are numerous reports on the effectiveness of such groups for children and adolescents who face diverse stressors (Gladding, 1995), their specific benefit to CGC has yet to be explored.

Despite the increasing popularity of support groups for the growing population of custodial grandfamilies, recent studies raise questions regarding how many CGP and CGC actually attend such groups. A focus group study sponsored by AARP revealed that few grandparents demonstrated enthusiasm over their own participation in support groups, even though most felt that other CGP would have a lot to gain from them (Robinson & Kensinger, 1996). Emick and Hayslip (1999) found that 75% of CGP raising CGC without behavioral or emotional problems, and over 50% of those raising CGC with such problems, had never gone to any sort of intervention to help cope with their caregiving responsibilities. Gerard, Landry-Meyer and Roe (2006) similarly reported that 61 % of CGP in their sample had never used support groups. Even less is known about the use of support groups by CGC. Some investigators, however, have found CGP to express considerable interest in support group for their grandchildren (Kopera-Frye, Wiscott, & Begovic, 2003; Smith et al., 2003). Furthermore, there are no published studies on what variables, if any, are associated with the actual use of support groups by these families.

The present study includes three aims. First, we examine perceived need, use, and unmet need for support groups by CGP and their CGC as reported by a national sample of 733 custodial grandmothers. This aim is important because the limited number of past studies on the use of support groups by custodial grandfamilies have involved small samples from particular geographic locales which restrict their generalizability.

A second aim is to explore predictors of service use by way of the prominent Andersen Social Behavioral Model (SBM, Andersen, 1995) to identify variables that predict the use of both CGP and CGC support groups among a large national sample. Roberto, Dolbin-MacNab, and Finney (2008) proclaimed that “the SBM provides a useful framework for practitioners when implementing strategies designed to influence grandparents’ health care practices and behaviors” (p.85). Findings from the present study will inform human service professionals regarding factors that either facilitate or impede the use of support groups by these families. Although originally developed to examine medical care use, the SBM has been broadened to other services used by family caregivers (Montoro-Rodriguez, Kosloski & Montgomery, 2003; Monahan, Greene, & Cleman, 1992). According to the SBM, factors that independently affect decisions about service use fall into the three categories of predisposing, enabling, and need factors.

Predisposing factors include characteristics of caregivers that influence their inclination to use support groups. Race, gender, and age of both the CGP and CGC are hypothesized to influence use, as attitudes and behaviors regarding service use are known to vary significantly across the groups defined by these demographic variables (Monahan et al., 1992). We further hypothesize that CGP who are unemployed will be more likely to use support groups because they have more free time to attend groups for themselves or to transport a CGC to a group (Goodman, Potts, & Pasztor, 2007). Our model also hypothesizes that the predisposing variable of time in the CGP role will affect service use because grandmothers new to this role may be more stressed and tend to view support groups as a resource for coping with this transition (Ross & Aday, 2006).

Enabling variables are those concerning the caregiver's ability to access support group services. We hypothesize that greater use of CGP and CGC support groups will be associated with living in non-rural settings where support groups and transportation to them are more prevalent (Cohen & Pyle, 2000), higher income to cover related costs such as transportation and day care (Kaminski & Murrell, 2008), low risk neighborhoods where access is safe, CGP's higher educational attainment (Roberto & Qualls, 2003), and greater overall social support (i.e., being married and perceiving greater expressive support) which enables access to needed services (Thompson et al., 2007). We also hypothesize that the legal status of CGP in relation to their grandchildren as a key enabling variable given that having legal arrangements in place facilitates overall service use within this population (Gerard et al., 2006; Goodman et al., 2007).

Need factors involve objective and subjective characteristics of the entire caregiving situation that indicate a vulnerability for which support groups provide an appropriate intervention (Monahan et al., 1992). The need variables hypothesized to be associated with greater use in our model are number of CGC having medical or psychiatric diagnoses, CGP's self-reported change in physical health since caregiving began, as well as CGP's perceived role captivity and depression. Our hypotheses are consistent with the view that greater service use is related to strain associated with caring for children with behavioral and physical disorders (Brannan & Helfinger, 2005) and to a generally vulnerable family environment (Thompson et al., 2007). We further hypothesize that the same predisposing, enabling, and need factors will meaningfully predict both CGP and their CGC support group use given that (a) children typically receive services as a result of decisions made by adults in their lives, instead of by their own efforts (Thompson, et al., 2007); and (b) the family factors included in our model are believed to play a significant role in determining if supportive services are used by custodial grandfamilies (Roberto et al., 2008).

A third aim is to investigate if the patterns and predictors of support group use vary across families recruited by two different sampling strategies: convenience (n = 346 [47%]) and population-based (n = 386 [53%]). Because CGP sampled via both techniques were required to meet identical eligibility criteria, we are able to examine if differences in the patterns and predictors of support group use by both CGP and CGC are attributable to sampling strategy. This is an important aim because different types of sampling result in different types of selection that, in turn, may influence measurement parameters, including means, variability, and relationships among variables (Hultsch et al, 2002).

In family caregiving research, there is a particular concern that participants recruited via convenience are biased toward greater levels of psychological distress and service use (Pruchno et al., 2008). Given that all prior studies on support group use by custodial families have involved small convenience samples, the present study will help determine if samples derived from convenience versus population-based strategies represent different populations. Any differences found may be attributable in part to the effects of sampling technique, whereas similarities found across sampling techniques may be held with greater confidence (Karney et al., 1995).



Participants were 733 custodial grandmothers (M age = 56 years, SD = 8.1) providing full-time care to a grandchild in total absence of the child's parents for at least three months (M = 6.4 years, SD = 4.0, range = 3 months to 16 years). The sample was restricted to custodial grandmothers because they generally experience higher levels of strain and family vulnerability than co-resident grandparents due to greater child-care responsibilities, less available social support, and greater exposure to adverse life-style changes (Bower & Myers, 1999). Grandmothers were recruited across the 48 contiguous states for an NIMH-funded study of stress and coping among custodial grandparents through a combination of convenience (e.g., social service agencies; Internet, radio, and newspaper ads) and population-based methods. Population-based sampling involved recruitment letters sent to randomly generated lists purchased from Survey Sampling, Inc., of the approximately 38 million households containing children under age 18.

Quota sampling was used to obtain a sample of custodial grandmothers that was half Black and half White. Blacks were over sampled because the probability of custodial care is much higher among Blacks than Whites (Bryson & Casper, 1999). Grandmothers’ self reported race was chosen as the analytical variable because all measures of constructs in the proposed model were obtained directly from them. Thus, any potential racial differences observed in the present study may be ascribed to grandmothers. Other racial and ethnic populations were excluded because they comprise a much smaller number of the overall CGP population than do non-Hispanic Whites and Blacks (Bryson & Casper, 1999). Otherwise, the sample was diverse regarding marital status, education, residential locale, work status, and income. Key demographic information regarding the sample and descriptive data for all predictors is shown in Table 1.

Table 1
SBM Predictor Variables by Sampling Strategy

If custodial grandmothers were caring for multiple CGC, then a target grandchild was selected using the most recent birthday technique. The target grandchildren were 391 girls and 342 boys (M age = 9.8 years, SD = 3.7, Range = 4 to 17 years). Most grandmothers (65.8%) provided care to a target grandchild who was born to a daughter. Multiple reasons for providing full-time care were reported by the majority of respondents. Most of these reasons concerned crisis or tragedy within the parent generation (e.g., substance abuse: 55.4%; incarceration: 42.6%). Grandmothers were excluded if they were caring for CGC due to the death of their own offspring because our focus was on families where CGC were being cared for due to problems faced currently by the child's birth parent.


As part of a larger telephone survey conducted by trained interviewers at a public research university in Northeast Ohio, custodial grandmothers were read a list of 27 services that provide support to custodial grandfamilies. This list was developed with input from professionals at national organizations providing advocacy and services for CGP-headed households. After each service was read to them, respondents were asked to answer “yes” or “no” if they had “used” and “needed” that service within the past year. The two services examined in this paper were read to respondents as “support groups for grandparents” and “support groups for grandchildren”. All variables included in our model of service use predictors are shown in Table 1, and only those for which measurement is not obvious from Table 1 are described here.

Predisposing Variables

Grandmothers’ work status was measured dichotomously in terms of whether or not the respondents were employed (full- or part-time, inside or outside the home) when the interview was conducted. Time as a CGP was measured by asking how long (in years) grandmothers had been caring for any CGC full-time without a biological parent's involvement.

Enabling Variables

Income was measured by asking grandmothers “What is your approximate total yearly household income after taxes?” Responses ranged from 1 (under $10,000) to 10 (more than $125,000). Expressive support was measured by the Expressive Support Scale (Pearlin, Mullan, Semple, & Skaff, 1990). Items (e.g., “You have friends that you can talk to when you are feeling down or discouraged”) were rated from 1 (strongly disagree) to 5 (strongly agree), with potential scores ranging from 8 – 40 (α = .89). Service satisfaction was measured with one item developed for this study: “The mental health system has been supportive of my needs as a caregiving grandparent” rated from 1 (strongly disagree) to 4 (strongly agree). Grandmothers’ health was assessed by one item that was self-rated from 1 (poor) to 5 (excellent). Neighborhood risk was scored dichotomously in terms of whether respondents expressed agreement (0) or disagreement (1) with the following statement: “My neighborhood provides a safe, clean, and comfortable environment for raising my grandchild”.

Need Variables

Total medical and psychiatric diagnoses were measured by summing the number of specific diagnoses within their respective category reported by grandmothers as having been made by health care professionals for any CGC in their care. Grandmothers’ health change was measured by one item asking the extent to which their physical health had changed since caregiving began: 1 (I have been sick a great deal), 2 (I haven't felt as good), 3 (I haven't noticed any change), and 4 (I have been healthier). Role captivity was assessed by adapting three items (e.g., “How often have you felt trapped by caring for your grandchild”?) that were originally developed by Pearlin et al (1990) for use with family caregivers of Alzheimer's victims. Items were rated from 1 (never) to 5 (very often), with potential scores ranging from 3 - 15 (alpha = .75). Grandmothers’ depression was assessed by the Center for Epidemiologic Studies Depression Scale (CES-D, Radloff, 1977), a 20-item measure of depressive symptoms (α = .90). For each item, participants endorsed the response that best described how often they had felt a particular way in the past week, from 0 (rarely or none of the time - less than 1 day) to 3 (most or all of the time - 5 to 7 days).

Analytic Plan

We obtained descriptive results for all variables for the overall sample as well as for the convenience and probability-based samples separately. Differences between sampling approaches were assessed with chi-square tests and independent sample t-tests for categorical and continuous variables, respectively. Logistic regression was used to identify the predisposing, enabling, and need factors that predicted CGP and CGC support group use. Analyses were conducted separately by sample type, as well as for the combined sample with sampling strategy included as a control variable. Statistically significant odds ratios greater than 1.00 reflect higher likelihood of use; those below 1.00 reflect lower likelihood. An alpha level of .05 was used for all statistical tests. The analyses were conducted with SPSS software.


The descriptive comparisons on study variables by sampling strategy shown in Table 1 reveal more similarities than differences between custodial grandmothers sampled via convenience versus a population-based approach. Among the differences observed, convenience custodial grandmothers reported significantly more unemployment, legal arrangements, medical and psychiatric diagnoses among their CGC, role captivity, and depressive symptoms than did those in the population-based sample. These findings suggest an overall higher level of vulnerability within the convenience sample. Convenience custodial grandmothers also reported greater satisfaction with the mental health service system.

Patterns of Support Group Use and Need

Table 2 summarizes, separately by recruitment strategy, the patterns of use and need for CGP and CGC support groups. Unmet need was deemed to occur if a custodial grandmother reported a need for the service but did not report using it within the past year.

Table 2
Patterns of use, need, and unmet need for support groups by sampling strategy.

Grandparent support groups

The unmet need for CGP support groups was less than 25% and was not significantly different between the convenience and population-based samples. A higher percentage of convenience custodial grandmothers, however, reported both use (X2(1) = 132.15, p < .001) and need (X2(1) = 73.43, p < .001) for CGP support groups than did those in the population-based sample.

Grandchildren support groups

The findings regarding CGC support groups are similar to those regarding CGP groups. The unmet need for CGC support groups was less than 27% and not significantly different between convenience and population-based samples. A higher percentage of both use (X2(1) = 76.69, p < .001) and need (X2(1) = 34.25, p < .001) for CGC support groups was reported in the convenience sample.

Overlapping use of CGP and CGC support groups

In light of the similar pattern of findings observed for use of CGP and CGC support groups, we conducted additional analyses to examine the extent to which the use of both types of groups overlapped. With the convenience and population-based samples combined, 60.6% of grandmothers who were using CGP support groups also reported use of CGC groups. This pattern was similar in the convenience sample, where 64.0% of CGP support group users also reported support group use by grandchildren. In contrast, only 47.5% of the population-based sample using CGP groups also used CGC groups.

Predictors of Service Use

Grandparent support groups

Shown in Table 3 are the results of the logistic regressions for predicting use of CGP support groups. For all three regression analyses, model fit was acceptable (Convenience X2(8) = 7.51, p = .48; Population X2(8) = 10.13, p = .26; Combined X2(8) = 6.27, p = .62). Pseudo R2 values were: Convenience = .20; Population = .25; Combined = .36.

Table 3
Predictors for use of grandparent support groups.

For both samples, use was associated with greater role captivity. Although not significant, there was also a trend across both groups for lower income to be related to use.

Unique to convenience grandmothers, a lower likelihood of use was related significantly to being employed and residing in a small city (versus a rural setting). A higher likelihood of use was associated significantly with being in the CGP role for a longer period (though there was a similar non-significant trend for population grandmothers) and with more CGC having medical diagnoses. There was a non significant trend for more expressive support to be associated with less use.

For population-based grandmothers only, target grandchild in the 8 to 10 age groups (compared to age 15 to 17), custody and adoption (compared to having no legal arrangement), and better grandmothers’ health were associated significantly with higher likelihood of use. There were also trends for use to be associated with target grandchildren in the 4 to 7 age group and with foster parent status. Custodial grandmothers’ perception of improved health since becoming a caregiver was associated significantly with a lower likelihood of use.

With the two samples combined, foster parent status (compared to no legal arrangement), longer time in the CGP role, better health among grandmothers, more CGC with medical diagnoses, and higher perceived role captivity were associated significantly with higher likelihood of CGP support group use, whereas being employed, residing in a small city (compared to rural setting), higher income, and more expressive support, were associated significantly with a lower likelihood of use. There were trends for legal custody and higher service satisfaction being associated with higher usage and for improved health since beginning the CGP role being associated with lower usage. These results were obtained after controlling for sampling strategy, which also was significant such that probability sampling was associated with lower likelihood of using support groups.

Grandchildren support groups

Shown in Table 4 are the results of the logistic regressions for predicting use of CGC support groups. For all three regression analyses, model fit was acceptable Convenience X2(8) = 9.34, p = .31; Population X2(8) = 3.42, p = .91; Combined X2(8) = 11.02, p = .20). Pseudo R2 values were: Convenience = .16; Population = .33; Combined = .26.

Table 4
Predictors for use of grandchildren support groups.

Income was the only variable that was a statistically significant predictor across both sampling strategies, with higher income being associated with lower likelihood of using CGC support groups. For convenience custodial grandmothers only, being older, living in a safe neighborhood, and greater perceived role captivity were associated with higher likelihood of use. For population-based custodial grandmothers only, target grandchild between ages 8 to 14 (compared to age 15 to 17) and foster parent status (compared to no legal arrangement) were associated significantly with higher likelihood of use, whereas improved custodial grandmothers’ health since taking on the caregiver role was associated with lower likelihood of use. There was also a non-significant tendency for higher use to be associated with target grandchildren being in the 4 to 7 age group and with greater satisfaction with the mental health system. With the two samples combined, target grandchildren of age 11 to 14 (compared to age 15 to 17); foster parent status (compared to no legal arrangement), higher service satisfaction, and greater role captivity were associated significantly with higher likelihood of use, whereas higher income was associated with lower likelihood of use. There was also a non-significant trend for higher use to be associated with target grandchildren being in the 4 to 7 age group, increased time in the caregiver role, more medical diagnoses among grandchildren, and higher role captivity. These results were obtained after controlling for sampling strategy, which also was significant such that probability sampling was associated with lower likelihood of using support groups.


The central aims of the present study were to (1) examine the perceived need, use, and unmet need for support groups by CGP and CGC; (2) identify variables that best predict the use of CGP and CGC support groups using the conceptual framework of the SBM; and (3) investigate if the patterns and predictors of service use vary across custodial grandfamilies recruited by convenience versus population-based strategies.

Patterns of need, use, and unmet need

Use and need for both CGP and CGC support groups varied by sampling strategy. Among convenience custodial grandmothers, users exceeded non-users (58.5% to 41.5%) for CGP support groups but not for CGC groups (41.5% to 58.5%). Among population-based custodial grandmothers there were a much lower percentage of grandmothers reporting use of support groups by both themselves (17.1%) and their grandchildren (12.4%). Whereas the majority of respondents in the convenience sample perceived a need for both CGP (71.3%) and CGC (61.2%) support groups, most of the population-based respondents did not perceive such a need for either type of support group (60.1% and 60.4%, respectively). However, perceived unmet need for support groups did not differ across convenience and population-based samples (22.1% for CGP support groups and 25.8% for CGC support groups).

These findings, coupled with the overall higher vulnerability observed among the convenience sample regarding the comparisons involving study independent variables (Table 1), reinforce concerns within the caregiving literature that convenience samples are biased toward higher distress and service use. This is not surprising in view of the fact that our convenience sampling plan drew heavily upon extant support groups and settings likely to sponsor or refer families to support groups (e.g., churches, social service agencies, legal services). Thus, we believe that our finding of relatively lower use within the population-based sample provides a more accurate picture of support group use across the overall population of custodial grandfamilies.

It is interesting that the unmet need for both CGP and CGC support groups was quite similar (range = 20.4% to 26.3%) among the convenience and population-based samples despite the different patterns of use and perceived need found between the two samples. The observed level of unmet need is particularly striking in regards to the population-based custodial grandmothers who reported low use of support groups for either themselves (17.1%) or grandchildren (12.4%). These findings may be due to the fact that many CGP are hesitant to seek or accept help from community organizations or government agencies due to skepticism about the value of formal assistance and reservations about depending on outside help (Gerard et al, 2006). Our population-based findings are also consistent with the AARP focus group study where few grandparents expressed enthusiasm about participating in support groups, even though most felt that other CGP would gain from them (Robinson & Kensinger, 1996).

Predictors of Support Group Use

Like our findings concerning use and need, those regarding predictors of support group use reveal more differences than similarities between the two sample types. A key variable emerging from our analyses was income, with lower income related to higher use of both CGP and CGC support groups regardless of sample type. One likely explanation for this finding is that CGP with less financial means may turn to support groups as an affordable alternative to more expensive services (e.g. family or individual counseling). It is also possible that poorer CGP have more contact with other services that refer them to support groups. In any case, our findings are consistent with past studies where CGP from support groups reported financial strain as their greatest problem (Kopera-Frye et al., 2003; Smith & Monahan, 2007). Thus, custodial grandmothers may have considerable unmet financial needs that should be covered in support groups and followed up on (Smith & Monahan, 2007).

Variables not associated with the use of either CGP or CGC support groups at statistically significant levels were target grandchild gender, grandmothers’ race, education, marital status, CGC psychiatric diagnoses, and custodial grandmothers’ depression. That these variables were not significant in either the convenience, population-based, or combined samples lends confidence to the conclusion that they are not meaningful predictors of support group use.

The only variable that was significantly related to use of CGP support groups in both samples was greater role captivity. This variable was also related to the use of CGC support groups in the convenience sample only. The salience of this variable is not surprising given that prior investigators have found increased caregiver burden to be associated with service use among custodial grandfamilies in particular (Goodman et al., 2007) and with the use of behavioral health services by children in general (Brannon et al., 2005). In fact, caregiver burden has been found elsewhere to overshadow children's symptoms in predicting mental health service use (Angold et al., 1998).

That the likelihood of using CGP support groups was higher among grandmothers who had been caregivers for a longer time (although not quite reaching significance in the population-based sample) suggests that it may take these families a while to either identify the availability of support groups or feel a need to use them. Practitioners should be aware that CGP who are new to this role may need help with identifying support groups that would facilitate their transition. Our clinical work with other family caregiver populations has indicated that many caregivers do not recognize the need for supportive services until they become fully immersed in caregiving.

In light of our belief that the population-based sample is generally more representative of the overall CGP population than the convenience sample, it is worth noting the other significant predictors found among the population-based custodial grandmothers. Age of the target grandchild, type of legal arrangements, and custodial grandmothers’ changes in health were significant predictors of both CGP and CGC support group use in the population-based sample. Grandmother health was a significant predictor of CGP support group use only.

That use of CGC groups was more likely to occur when the target grandchild was younger than age 15 seems in line with Mayer's (2002) observation that late adolescent CGC are more successful than younger children at making sense out of what has happened to them. In turn, they may have less need for support groups as an aid to their adjustment. Use of both CGP and CGC support groups was also related to formal legal arrangements such as foster parent status, custody, and adoption. This may be due to the fact that CGP with ties to the formal legal system have greater access and referrals to support services and support groups are sometimes mandated for continuation of guardianship or foster parenthood (Baird, 2003; Cohen & Pyle, 2000).

Use of both custodial CGP and CGC support groups was less likely to occur when custodial grandmothers reported improved health after taking on the caregiver role, whereas better health in general among grandmothers was related to increased likelihood of using CGP support groups. Although grandmothers in better health may have the physical ability and energy to transport either themselves or a grandchild to settings where support groups are conducted, those who regard their health as improving since the onset of caregiving may perceive less overall need for support groups.

Implications for Future Research & Practice

Our findings regarding comparisons between the convenience and population-based samples imply that data on support group use obtained solely from convenience samples may be misleading with respect to both (a) exaggerating patterns of use and need, and (b) yielding different predictors of service use than those occurring in population-based samples. Practitioners, researchers, and policy makers should interpret any findings regarding service use obtained via convenience samples only cautiously until they are either replicated across multiple convenience samples or validated against a population-based sample.

Despite other differences found by sample type, the level of unmet need for CGP and CGC support group ranged from 20% to 26% for both the convenience and population-based samples, meaning that approximately one quarter of families who perceive a need for these services were not using them. Researchers and practitioners must take future steps to identify what factors account for this lack of use despite self-reported need. One possible reason for this unmet need is suggested by our finding that those CGP without formal legal arrangements in place for their CGC were less likely to use both CGP and CGC support groups. If CGP s are not already involved with the service system, whether necessitated by their own situations or by the intervention of child protective services, then they are less likely to receive information pertaining to educational opportunities, support groups, and other services (Baird, 2003; Goodman et al., 2007). It is essential that other outlets such as pediatricians, schools, day care centers, senior centers, and religious organizations be used to advertise the availability of support groups to custodial grandfamilies in view of past findings that many grandparents seem to be unaware of their existence (Ross & Aday, 2006).

The high amount of overlapping use of CGP and CGC support groups observed within a given family in our sample is also instructive to clinicians, and seems congruent with the view that children use services due to decisions made by adult caregivers instead of by their own efforts (Thompson, et al., 2007). This overlapping use also raises the critical question of whether or not using one type of group leads to use of the other. One possibility, for example, is that CGPs who experience high burden (role captivity) enter support groups for themselves and consequently become aware of SCG support groups. In addition, because lack of child care is an issue that is likely to prevent the use of CGP support groups (Kopera-Frye, Wiscott & Begovic, 2003; Qualls & Roberto, 2003), the overlapping use observed here may reflect the previously found desire of CGP to have CGC groups occur at the same time and place as GCP support groups (Smith et al., 2002). Practitioners who serve either CGP or CGC should make a point of enquiring about what services all family members might benefit from.

Practitioners should also bear in mind our findings that use of both CGP and CGC support groups were more likely to occur when CGC were of younger ages. One possibility is that caring for younger children is perceived to be more difficult than caring for older children, which may then prompt CGPs living with younger CGC to seek out support groups. As stated above, it could be that late adolescent CGC have less need for support groups as an aid to their adjustment because they are more successful than younger CGC at making sense out of what has happened to them Mayer, 2002). Or it might also be the case that suitable support groups for teenage CGC are less accessible than support groups for younger children. For example, CGP in our prior studies have reported that it is quite difficult for older grandchildren to find time in their daily schedules to attend support groups and strongly believed that support groups for CGC should be age appropriate in terms of content and procedures (Smith et al. 2002)/

Our findings also suggest that perceived burden in the caregiving role is a far more salient predictor of support group use than either depressive symptoms reported by CGP or the presence of psychiatric diagnoses among CGC. This is a somewhat reassuring finding to clinicians in that it appears to reflect an accurate perception on the part of CGPs that support groups can be effective at reducing stress and are not meant to serve as substitutes for treatment of psychological disorders (Monahan et al., 1992). It may also be true that support group leaders are appropriately referring CGP and CGC with psychological problems to other sources of assistance.

Study Limitations

Several limitations of this study limit the conclusiveness of these findings. Our measures of use are rudimentary and fail to distinguish between the diverse types of support groups available to custodial grandfamilies (e.g., professional versus non-professional, sponsorship, physical setting), and they do not address such issues as frequency of attendance and attrition. Likewise, our assessments of need and unmet need were self-reported rather than derived from objective determinations of need. Although based on the SBM, our model does not encompass all variables that predict support group use. Generalizability was restricted by not including racial groups other than Whites and Blacks in our sample. Even samples of custodial grandfamilies obtained via population-based strategies are not necessarily representative of the population because not everyone who is approached will agree to participate, and refusal may vary across different types of individuals. (Hultsch et al., 2002). Finally, the cross-sectional nature of our data precludes conclusions regarding the direction of causality among predictor variables. It is important for each these deficiencies to be addressed in future research.

Despite these limitations, ours is the first study to examine use of both custodial grandmothers and custodial grandchildren support groups within a large national sample and to identify key predictors of such use. Hopefully, these findings will lead to further research along these lines as well as inform practitioners on how to facilitate greater use of support groups by custodial grandparent families.


This research was supported by a grant from the National Institute of Mental Health awarded to the first author.


An earlier version of this paper was presented at the 59th Annual Scientific Meeting of the Gerontological Society of America, Dallas, TX. November (2006).

Contributor Information

Gregory C. Smith, College of Education, Health and Human Services Kent State University.

Julian Montoro Rodriguez, Department of Sociology California State University, San Bernadino.

Patrick A Palmieri, Center for the Treatment and Study of Traumatic Stress Summa Health System – (St. Thomas Hospital), Akron, OH.


  • Andersen R. Revisiting the behavioral model and access to medical care: Does it matter? Journal of Health and Social Behavior. 1995;36:1–10. [PubMed]
  • Angold A, Messer SC, Stangl D, Farmer EM, Costelo EJ, Burns BJ. Perceived parental burden and service use for child and adolescent psychiatric disorders. American Journal of Public Health. 1998;88:75–80. [PubMed]
  • Baird AH. Through my eyes: Service needs of grandparents who raise their grandchildren, from the perspective of a custodial grandmother. In: Hayslip B, Patrick J, editors. Interventions for custodial grandparents. Springer; New York: 2003. pp. 59–65.
  • Bowers BF, Myers BJ. Grandmothers providing care for grandchildren: Consequences of various levels of caregiving. Family Relations. 1999;48:303–311.
  • Brannan AM, Heflinger CA. Child behavioral health service use and caregiver strain: Comparison of Managed Care and Fee-For-Service Medicaid Systems. Mental Health Services Research. 2005;7:197–211. [PubMed]
  • Bryson K, Casper LM. Current Population Reports. US Bureau of the Census; Washington, DC: May, 1999. Coresident grandparents and grandchildren. pp. 23–198.
  • Cohen CS, Pyle R. Support groups in the lives of grandmothers raising grandchildren. In: Cox CB, editor. To grandmother's house we go and stay: Perspectives on custodial grandparents. Springer; New York: 2000. pp. 235–267.
  • Dannison LL, Smith AB. Custodial grandparents’ community support program: Lessons learned. Children & Schools. 2003;25:87–98.
  • Emick ME, Hayslip B. Custodial grandparenting: Stresses, coping skills, and relationships with grandchildren. International Journal of Aging and Human Development. 1999;48:35–61. [PubMed]
  • Gerard JM, Landry-Meyer L, Roe JG. Grandparents raising grandchildren: The role of social support in coping with caregiving challenges. International of Aging and Human Development. 2006;62:359–383. [PubMed]
  • Gladding ST. Group work: A counseling specialty. 2nd ed Prentice Hall; Englewood Cliffs, NJ: 1995.
  • Goodman CC, Potts MK, Pasztor EM. Caregiving grandmothers with vs. without child welfare system involvement: Effects of expressed need, formal services, and informal social support on caregiver burden. Children and Youth Services Review. 2007;29:428–441.
  • Hultsch DF, MacDonald SWS, Hunter MA, Maitland SB, Dixon RA. Sampling and generalisability in developmental research: Comparison of random and convenience samples of older adults. International Journal of Behavioral Development. 2002;26:345–359.
  • Kaminski PL, Murrell AR. Counseling custodial grandchildren. In: Hayslip B, Kaminski P, editors. Parenting the custodial grandchild. Springer; New York: 2008. pp. 215–236.
  • Karney BR, Davila J, Cohan CL, Sullivan KT, Johnson MD, Bradbury TN. An empirical investigation of sampling strategies in marital research. Journal of Marriage & Family. 1995;57:909–920.
  • Kopera-Frye K, Wiscott RC, Begovic A. Lessons learned from custodial grandparents involved in a support group. In: Hayslip B, Patrick J, editors. Interventions for custodial grandparents. Springer; New York: 2003. pp. 243–256.
  • Landry-Meyer L. Research into action: Recommended intervention strategies for grandparent caregivers. Family Relations. 1999;48:381–389.
  • Mayer M. Grandparents raising grandchildren. School Psychology International. 2002;23:371–385.
  • Minkler M, Driver D, Roe KM, Bedeian K. Community interventions to support grandparent caregivers. The Gerontologist. 1993;33:807–8011. [PubMed]
  • Monahan DJ, Greene VL, Cleman PD. Caregiver support groups: Factors affecting use of services. Social Work. 1992;37:254–260. [PubMed]
  • Montoro-Rodríguez J, Kosloski K, Montgomery R. Evaluating a practice -oriented service model to increase the use of respite services among minorities and rural caregivers. The Gerontologist. 2003;43(6):916–924. [PubMed]
  • O'Rourke KO, Worzbt JC. Support groups for children. Accelerated Development; Bristol, PA: 1996.
  • Patrick JH, Hayslip B. Epilogue: The next stage in helping custodial grandparents. In: Hayslip B, Patrick J, editors. Interventions for custodial grandparents. Springer; New York: 2003. pp. 277–282.
  • Pearlin LI, Mullan J, Semple S, Skaff M. Caregiving and the stress process: An overview of concepts and their measures. The Gerontologist. 1990;30:583–594. [PubMed]
  • Pruchno RA, Brill JE, Shands Y, Gordon JR, Genderson MW, Rose M, Cartwright F. Convenience samples and caregiving research: How generalizable are the findings? Gerontologist. 2008;48:820–827. [PubMed]
  • Radloff LS. The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement. 1977;1:385–401.
  • Roberto KA, Qualls SH. Intervention strategies for grandparents raising grandchildren: Lessons learned from the caregiving literature. In: Hayslip B, Patrick J, editors. Interventions for custodial grandparents. Springer; New York: 2003. pp. 13–26.
  • Robinson M, Kensinger JL. Final report with African American and Hispanic grandparents about issues related to caring for their grandchildren. American Association of Retired Persons; Washington, DC: 1996.
  • Roe KM. Community interventions to support grandparent caregivers: Lessons learned from the field. In: Cox CB, editor. To grandmother's house we go and stay: Perspectives on custodial grandparents. Springer; New York: 2000. pp. 283–303.
  • Roberto KA, Dolbin-MacNab ML, Finney JW. Promoting health for grandmothers parenting young children. In: Hayslip B, Kaminski P, editors. Parenting the custodial grandchild. Springer; New York: 2008. pp. 75–89.
  • Ross MET, Aday LA. Grandparents who are raising their grandchildren. Journal of Family Issues. 2006;27:912–932.
  • Simmons T, Dye JL. Grandparents living with grandchildren: 2000. 2003. [February 16, 2009]. from
  • Smith CJ, Monahan DJ. KinNET: A demonstration project for a national support network for kinship care providers. Journal of Health & Social Policy. 2007;22:215–231. [PubMed]
  • Smith GC. How caregiving grandparents view support groups: An exploratory study. In: Hayslip B, Patrick J, editors. Interventions for custodial grandparents. Springer; New York: 2003. pp. 69–91.
  • Smith GC, Palmieri PA. Risk for psychological difficulties in children raised by custodial grandparents. Psychiatric Services. 2007;58:1303–1310. [PMC free article] [PubMed]
  • Smith GC, Savage-Stevens S, Fabian E. How caregiving grandparents view support groups for grandchildren in their care. Family Relations. 2002;51:274–281.
  • Thompson R, Lindsey MA, English DJ, Hawley KM, Lambert S, Browne DC. The Influence of family environment on mental health need and service use among vulnerable children. Child Welfare. 2007;86:57–74. [PubMed]
  • Vacha-Haase T, Ness CM, Dannison L, Smith A. Grandparents raising grandchildren: A psychoeducational group approach. Journal for Specialists in Group Work. 2000;25:67–78.