Introduction

Mothers and children are usually embedded in complex extended family support networks that interact with individual, familial and contextual characteristics to shape family functioning and maternal and child well-being. Very often these social support networks include the involvement of non-parental adults, that is aunts, uncles, grandparents or adult family friends, who help parents raise children. This support may be particularly prevalent and influential among economically disadvantaged and Latino families, given needs stemming from economic hardship (e.g., Ceballo and McLoyd 2002) and/or immigration, and cultural preferences endorsing collective child-rearing (e.g., Baca Zinn and Pok 2002; Sarkisian et al. 2006). Moreover, social support involving care for children may be a key buffer in high-risk ecological contexts that threaten parental and child well-being (e.g., Ceballo and McLoyd 2002; Appleyard et al. 2007). However, limited research considers the impact of parenting support provided by non-parental adults on maternal functioning. This omission could lead to the development of short-sighted programs to promote family well-being. The present study examines the extent to which parenting support, specifically engagement in several child rearing tasks, provided by non-parental adults is linked to maternal parenting efficacy and depressive symptoms among a community sample of mothers and children.

This study is grounded in a bioecological framework (Bronfenbrenner and Morris 2006). Much research on parenting focuses on parent–child dyads, yet these relationships are embedded in multiple relationship systems. We conceptualize social support as a proximal process in which mothers and children are engaged, especially when it involves sustained direct interactions with children across time. Moreover, the nature of social support and its influence on maternal functioning likely varies systematically as a function of other contextual and individual characteristics. That is, the transactional nature of social support suggests that characteristics of the stressor, the individual, the environment and the types of social support available shape the processes through which social support is related to individual outcomes (e.g., Antonucci and Jackson 1990). More specifically, this study applies a Process Person Context model (Bronfenbrenner and Morris 2006) to examine the associations between characteristics of social support networks and maternal wellbeing, and how these associations may vary according to contextual (race/ethnicity) and family (maternal marital status) characteristics.

Parenting Efficacy and Depressive Symptoms

Parenting efficacy, the extent to which a parent feels confident and effective in her abilities as a parent to shape her child’s development, is a key marker of maternal functioning. Parenting efficacy is positively linked to maternal well-being (Jones and Prinz 2005) and to children’s social adjustment (Scaramella et al. 2008). More specifically, parenting efficacy may mediate associations between maternal (e.g., depression), child (e.g., difficult temperament), and family (e.g., low-income) risks and mothers’ competent parenting behaviors (Teti et al. 1996). Similarly, elevated maternal depressive symptoms influence the use of negative parenting practices, and thus indirectly impact children (e.g., Goodman et al. 2011; Lovejoy et al. 2000). For example, maternal depressive symptoms are consistently linked to less responsive and engaged parenting, and more intrusive and harsh parenting (e.g., Lovejoy et al. 2000; Riley et al. 2008). Parenting efficacy and depressive symptoms may be particularly vulnerable dimensions of functioning among low-income and single mothers coping with multiple stressors (Jackson 2000; Scaramella et al. 2008). Therefore, identifying factors that may increase parenting efficacy and decrease depressive symptoms represents an important research goal.

The Protective Role of Social Support

Social support, which is generally conceptualized as consisting of emotional, informational and instrumental dimensions, is consistently linked to child wellbeing, either directly via interactions with the child, or indirectly via influences on parenting (Cochran and Niegro 1995). Perceived availability of social support protects mothers’ well-being in the presence of multiple sources of stress (e.g., Hashima and Amato 1994; Osofsky and Thompson 2000). Higher perceived levels of social support are associated with increased parenting efficacy (e.g., Cochran and Niegro 1995; Marshall et al. 2001; Suzuki et al. 2009) and reduced depressive symptoms (e.g., Bost et al. 2002; Cairney et al. 2003; Orthner et al. 2004) among mothers. For example, Marshall et al. (2001) reported that maternal perception of social-emotional support from friends and family was positively related to parenting efficacy among mothers of elementary school age children. Similarly, Jackson (2000) found that among low-income mothers, perceptions of available social-emotional support were linked to fewer self-reported depressive symptoms. Perceptions of social-emotional support in particular may both validate mothers’ feelings of effective parenting, and provide an outlet to support mothers when they encounter the inevitable frustrations stemming from parenting and other life stressors that contribute to depressive symptoms and undermine parenting efficacy. Relatedly, provision of social support, especially instrumental support (i.e., tangible goods and services) may reduce the number of stressors that parents encounter (Cairney et al. 2003; Hashima and Amato 1994). However, most research linking social support to parental wellbeing focuses on social-emotional or broad-based instrumental support with little focus on how individuals may support mothers by engaging directly in caregiving tasks (Cochran and Niegro 1995; Marshall et al. 2001).

Social support from extended family members who engage in direct childrearing tasks, referred to in the anthropological literature as “alloparenting,” has been an important influence on children’s development throughout human evolutionary history (Hrdy 2009). Examining parenting support provided by non-parental adults is particularly applicable to economically disadvantaged and Latino families, the sample for the present study, for several reasons. First, economically disadvantaged parents are likely to rely extensively on assistance from others within their social networks due to the inability to purchase services (Henly et al. 2005; Orthner et al. 2004). This reliance is particularly likely to include child care and direct support for children, especially among single mothers (Osofsky and Thompson 2000; Jones et al. 2007). Second, Mexican–American parents often report family-based social support given needs tied to immigration and economic disadvantage, and cultural preferences endorsing raising children as a collective process (e.g., Baca Zinn and Pok 2002; Roosa et al. 2002). In fact, single mothers in Mexico are significantly more likely to reside with and depend on extended kin, especially for child care, in comparison to Mexican married or American single mothers (Shin 2013). Therefore, parenting support may be a key buffer in high-risk ecological contexts that threaten parent and child wellbeing, particularly when alloparents engage in regular and sustained interactions with children (Appleyard et al. 2007; Jones et al. 2007). Thus, provision of parenting support by non-parental adults may be an important, but under-measured, family resource.

Social support is complex and multidimensional, and thus it requires multiple indicators. Although the source of support (e.g., formal versus informal and friend versus family) has been the subject of extensive research, the function of support has received less attention, especially among Latino populations. The development of new family support programs and the identification of eligible families for existing programs requires precise, culturally valid and sensitive measures of family risks and strengths. One important step in this direction is the consideration of perceived support related directly to parenting within a framework that also accounts for global social-emotional support. Moreover, limited research has considered how children’s perceptions of the support they receive from other adults may be related to parental functioning (Cochran and Niegro 1995), even though children provide a unique perspective that may be over-looked by parents, especially those who are burdened by other stressors. For example, if mothers are experiencing high levels of depression, they may be less likely to report or notice the assistance provided by others, but these individuals may be particularly key resources for children exposed to maternal depression. In a rare exception, Marshall et al. (2001) measured child reported availability of supportive adults, that is the presence of someone other than a parent who made them feel better when they talked about their feelings. Availability of support was unrelated to parenting behaviors or child well-being, however, this measure did not account for children’s perceptions of the availability of more instrumental parenting support (i.e., engagement in direct caregiving tasks) that is the focus of the present study. This tangible assistance provided by non-parental adults may be especially helpful to mothers. In another study, Sterrett et al. (2009) found that child reports of support from caregivers other than mothers in single-mother families interacted with mothers’ positive parenting such that the combination of positive parenting and positive youth-nonparental adult relationships was linked to reduced internalizing and externalizing symptoms. However, this study did not examine how children’s perceptions of non-parental adult involvement were linked to maternal functioning, which could be an alternative pathway by which non-parental adults impact children.

Variations by Family Characteristics

The influence of parenting support on maternal parenting efficacy and depressive symptoms may vary as a function of family characteristics. We consider two of these potential moderators in the present study. First, parenting support may be most protective in single-mother families, as reliance on alloparental support may in part compensate for father absence (Hrdy 2009; Jones et al. 2007; Jackson 2000). In fact, given consistent positive associations between spousal support and maternal parenting efficacy (e.g., Cochran and Niegro 1995), support provided by kin and friends may be particularly facilitative of parenting efficacy among single mothers (Osofsky and Thompson 2000). Single mothers also may be at greater risk for depression than married mothers (e.g., Cairney et al. 2003), and thus social support may alleviate some of the stressors contributing to depressive symptoms.

Further, although social support may be protective among all parents, it may be especially important to examine among parents of Mexican origin. Research on Latino adults in general consistently indicates the presence and reliance on extensive family support networks (e.g., Roosa et al. 2002; Marshall et al. 2001), and Mexican origin and other Latino parents are likely to report extensive social support from family (Haxton and Harknett 2009; Niska 1999; Shin 2013). However, there has been surprisingly little research identifying the implications of family-based or other social support for parental well-being among Mexican–Americans (Barnett et al. 2013). Moreover, Mexican–American cultural values such as familism and sociocentrism favor raising children as a collective or group process (Baca Zinn and Pok 2002). Thus, the focus on the dyadic parent–child relationship or the nuclear family fails to capture the diversity of interactions that shape child development. In fact, family support has been linked to better self-rated mental health among Latino adults (Aranda et al. 2001; Rivera et al. 2008; Rodriguez et al. 2007) and Mexican origin mothers of toddlers (Barnett et al. 2013). In the present study, the 53 % of the mothers who self- identified as Latino/Hispanic were mostly of Mexican origin due to the study location in Arizona. The vast majority (98 %) of mothers who did not identify as Latino/Hispanic were White.

The Present Study

The present study draws data from a community sample of mothers and 10–12 year-old children. We examine the extent to which mothers’ perceived global social-emotional support and children’s perceived parenting support by non-parental adults are associated with mother-reported parenting efficacy and depressive symptoms. We expect that social-emotional and parenting support will be uniquely and positively associated with parenting efficacy, and uniquely and negatively associated with depressive symptoms. However, we expect these associations to be qualified by two moderating effects such that parenting support will be particularly protective, that is positively related to parenting efficacy and negatively related to depressive symptoms, among (a) single-mother families, and (b) Latino/Hispanic families.

Method

Participants

The analyses include 59 families comprised of mother and child dyads drawn from fourth, fifth and sixth grade classrooms in two elementary schools in a low-income district in a medium sized southwestern city. Mean child age was 10.08 years (SD = 1.02), and mean mother age was 36.47 years (SD = 5.56). The sample is predominantly economically disadvantaged. Mean mother-reported annual per capita household income was $6,100 (SD = $4,706), with 73 % of mothers earning at least a high school diploma or equivalent. Marital status included 42 % single, 15 % cohabiting with a romantic partner, and 42 % married. Approximately 53 % of mothers self-identified as Latino/Hispanic, and 51 % of mothers reported that they were born in the United States. Mothers were given the option of participating in English or Spanish; 22 (37 %) mothers chose to be interviewed in Spanish. All children were interviewed in English.

Procedure

All study measures and procedures were approved by the by the Institutional Review Board of the affiliated researchers.

Recruitment

Participants were recruited through two methods. First, letters requesting permission to contact parents for more information about the study were sent home with every 4th–6th grader in the two targeted elementary schools. This information was provided in English and Spanish. Parents had the option of indicating that they were not interested in participating. All children who returned a permission form were given a coupon for a free item at a local business. Second, research team members attended school-wide events to meet directly with interested parents.

Interviews

Mothers and children were interviewed separately. All interviews took place in private settings with trained research assistants. In order to account for differences in literacy skills among participants, interviewers read the interview questions to the participants while the participants followed along with their own copies. All maternal questionnaires were translated into Mexican Spanish unless a Spanish version previously validated with Mexican–Americans was available. The translator has many years of experience translating research documents between English and Spanish in the United States and Mexico. Two native Spanish speakers who are bilingual reviewed the English and Spanish versions and resolved any inconsistencies. Mothers were compensated $25 for their participation in a 60–90 min in-person interview. All maternal interviews were conducted by female interviewers. The present analyses only consider data from select measures.

Children participated in a 45 min interview for which they were compensated with books of their choice. Children reported on a variety of constructs, including parenting behaviors and non-parental parenting support. Children also completed two computer-based inhibitory control tasks. Only data from the children’s reports of support from non-parental adults are used in the present study.

Measures

Demographics

Mothers reported marital/relationship status, race/ethnicity, and total annual household income and size. In order to compare single mothers to partnered mothers, those who reported currently being married and those who reported living with a romantic partner were combined to form one group.

Depressive Symptoms

Mothers completed the Center for Epidemiologic Studies Depression Scale (Radloff 1977), a commonly used index. Participants rated how frequently they experienced 20 depressive symptoms during the last week on a 4-point scale ranging from 1 (none of the time) to 4 (most of the time). The items were averaged to create a measure of total depressive symptoms (α = 77), with higher scores reflecting higher levels of depressive symptoms.

Parenting Efficacy

Mothers responded to six items designed specifically for low-income mothers with low educational attainment (Scaramella et al. 2008). Items tapped how mothers felt they could change their child’s behavior, handle problems their child may have, and correct their child’s behavior. Items were rated on a 5-point Likert scale, ranging from 1 (never) to 5 (always). Items were summed (α = 0.73) such that higher scores indicate higher perceptions of parenting efficacy.

Global Social-Emotional Support

Mothers responded to the Multidimensional Scale of Perceived Social Support (Zimet et al. 1988), which has been validated with Mexican–Americans (Knight et al. 2010). Mothers were asked to rate the perceived availability of social support on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Sample items include, “I get the emotional help and support I need from my family,” and “There is a special person in my life who cares about my feelings”. Items are summed (α = 0.91) so that higher scores indicate higher levels of perceived social support.

Child-Reported Parenting Support

Children responded to a novel measure of children’s perceptions of support from non-parental adults related to specific childrearing tasks, the Social Support for Parents and Children Scales (Barnett 2010). Children were asked to identify the two adults, other than their mother and father, who are most important to them. Children most frequently listed their maternal grandmothers, followed by aunts, uncles and paternal grandmothers. They then rated on a 3-point scale ranging from 1 (never) to 3 (all the time) how often those individuals performed nine tasks such as “helps with homework,” “attends important events,” “does fun things with you,” “takes care of you when you’re sick,” “prepares meals for you,” and “teaches you new things.” Items were summed for each individual (α = 0.82 and 0.83), and then averaged across the two individuals with higher scores reflecting higher average parenting support.

Analytical Strategy

In addition to examination of bivariate relations among variables of interest, a series of hierarchical regression analyses were computed. Specifically, two sets of identical models were computed. The dependent variable for one model was parenting efficacy, and the dependent variable for the other model was depressive symptoms. In order to better compare the results from the two sets of models, we included maternal depressive symptoms as an independent variable in the models predicting parenting efficacy, and likewise parenting efficacy as an independent variable in the models predicting depressive symptoms. First, the main effects of race/ethnicity, marital/partnered status, depressive symptoms/parenting efficacy, and social-emotional support on each outcome were assessed. Second, child reported parenting support was added to the models. Finally, interaction terms were included in the models to test whether race/ethnicity and marital/partnered status moderated the relation between child-reported parenting support and maternal self-reported parenting efficacy and depressive symptoms. All statistically significant interaction effects were evaluated by following standard pick-a-point procedures (Preacher et al. 2006).

Results

Correlational and Descriptive Analyses

As shown in Table 1, only one independent variable was correlated with the dependent variables. Mother-reported social-emotional support was inversely correlated with depressive symptoms (r = −0.41, p < 0.01). Next, we computed T-tests to determine whether there were mean differences among the independent and dependent variables in the analyses according to mothers’ self-identified race/ethnicity (Latino/Hispanic versus non-Latino/Hispanic) and marital status (single versus married/partnered). There were no mean differences in the independent and dependent variables based on maternal self-reported race/ethnicity. However, partnered mothers (mean = 3.31; SD = 0.68) on average reported more social-emotional support than single mothers (mean = 2.89; SD = 0.64; t [58] = −2.47, p < 0.05).

Table 1 Bi-variate correlations and descriptives of dependent and independent variables

Hierarchical Multiple Regression Models Predicting Parenting Efficacy

Next, hierarchical multiple regression equations were computed to evaluate the hypotheses regarding the links between social support and parenting efficacy. As shown in Table 2, including marital status, race/ethnicity, maternal depressive symptoms and mother-reported social-emotional support in the first step of the model failed to account for a statistically significant portion of the observed variance in parenting efficacy, or to produce a statistically significant predictor, though the effect of depressive symptoms was moderate in size (β = −0.30, p < 0.05). The inclusion of child-reported parenting support in the next step also failed to produce a statistically significant model. Next, the interaction terms representing the products of the standardized values of child-reported parenting support and each of the hypothesized moderators were added to the model. This addition resulted in a statistically significant overall model fit, accounting for 21 % of the observed variance in parenting efficacy. In this final model, two interactions terms, child-reported parenting support X marital status and child reported parenting support X race/ethnicity emerged as statistically significant. These interactions were probed following identical procedures. First, we plotted the simple slopes of the lines defining the relationship between child-reported parenting support and parenting efficacy for single mothers and for married/partnered mothers. As depicted in Fig. 1, for single mothers only (b = −4.06, p < 0.01), higher levels of levels of parenting support predicted lower levels of parenting efficacy. Second, we plotted the simple slopes of the lines defining the relationship between child-reported parenting support and parenting efficacy for Latino/Hispanic and non-Latino/Hispanic mothers. As shown in Fig. 2, there was a negative association between parenting support and parenting efficacy among Latino/Hispanic mothers only (b = −2.29, p < 0.05).

Table 2 Hierarchical regression model predicting maternal parenting efficacy
Fig. 1
figure 1

Mother reported parenting support negatively predicts parenting efficacy for single mothers

Fig. 2
figure 2

Mother reported parenting support negatively predicts parenting efficacy for latino/hispanic mothers

Hierarchical Multiple Regression Models Predicting Maternal Depressive Symptoms

Table 3 presents the results of the hierarchical regression models computed to predict maternal depressive symptoms. The first model estimated the influences of marital/partnered status, race/ethnicity, parenting efficacy, and social-emotional support on maternal depressive symptoms. This model accounted for 12 % of the observed variance in depressive symptoms. Mother-reported global social-emotional support was inversely related to depressive symptoms, and attained a moderate effect size (β = −0.39, p < 0.05). Next, child reported parenting support was added to the model, but this addition failed to account for a statistically significant increase in the amount of explained variance. Third, the interaction terms representing the products of the standardized values of child reported parenting support and the two hypothesized moderators were added to the model. The addition of these variables accounted for a significant increase (8 %) in the proportion of the overall variance accounted for in depressive symptoms (F for change in R 2 = 3.21, p < 0.05). Above and beyond the inverse association between mother-reported global social-emotional support and depressive symptoms, there was a statistically significant interaction between race/ethnicity and child-reported parenting support. Following procedures similar to those outlined above, as shown in Fig. 3, we found that for Latino/Hispanic families only, children’s perceived receipt of greater parenting support was associated with more maternal self-reported depressive symptoms (b = 0.26, p < 0.05).

Table 3 Hierarchical regression model predicting maternal depressive symptoms
Fig. 3
figure 3

Child reported parenting support positively predicts maternal depressive symptoms for latino/hispanic mothers

Discussion

The goal of the present study was to use an ecologically grounded approach to examine links between children’s perceptions of direct parenting support and mothers’ self-reported parenting efficacy and depressive symptoms, above and beyond the links between mother-reported social-emotional support and these aspects of maternal functioning. More specifically, we considered the extent to which the associations between child-reported parenting support provided by non-parental adults varied according to race/ethnicity and maternal marital status. We hypothesized that parenting support represented an important, under-studied resource that would be positively associated with parenting efficacy, and negatively associated with depressive symptoms, especially among those families most at risk due to single motherhood, and among Latino/Hispanic families who may hold cultural values endorsing this kind of support. However, for the most part the findings indicate that higher levels of parenting support may in fact be markers for family risk rather than promoters of maternal well-being. First, we discuss the findings regarding the main effects of social support on maternal well-being. Second, we consider the findings regarding moderating influences of family characteristics. Next, we highlight study strengths and limitations. We conclude by discussing the implications of the findings.

Social Support, Parenting Efficacy, and Maternal Depressive Symptoms

The present findings underscore the multidimensional and complex nature of social support. For example, mothers’ perceptions of available social-emotional support were unrelated to their children’s reports of parenting support. Importantly, this could be due to the perspective of different reporters, or it could be because direct support to children in the form of engagement in specific child rearing tasks is distinct from the social-emotional support that mothers may perceive as being directed at them. Mixed evidence emerged regarding the direct links between social support and the two indices of maternal well-being. First, the findings suggest that consistent with existing research, mothers’ perceptions of the availability of global social-emotional support were negatively associated with depressive symptoms (Cairney et al. 2003; Orthner et al. 2004). Therefore, social-emotional support appears to be an important resource for mothers. However, there were no direct associations between child-reported parenting support and maternal self-reported functioning. Perhaps, as the findings described below regarding moderation by family characteristics suggest, the meaning of this support differs across groups, and thus there are no observable effects when the sample is combined. It is worth noting that although not included in the present analyses, mothers were also asked to identify the adults (other than a spouse) who are most helpful in caring for their children. Mothers and children generally identified the same individuals, suggesting that mothers are aware of who is caring for their children, and that these individuals are integral members of whole family support networks.

The hypotheses regarding the differential associations between parenting support and maternal well-being as a function of family characteristics were partially supported, however in the opposite direction than expected. First, when Latino/Hispanic children reported more parenting support, their mothers reported feeling less efficacious in their role as a parent. Second, child-reported parenting support and mother-reported parenting efficacy were also inversely related among single mothers. Post hoc analyses failed to reveal a statistically significant three-way interaction between race/ethnicity, marital status and parenting support in the prediction of parenting efficacy, although sample size limited the power to detect such an effect. Importantly, there were no mean differences in the amount of parenting support reported by children by race/ethnicity or maternal marital status. Finally, among Latino/Hispanic families only, child reports of parenting support were associated with increased maternal depressive symptoms.

Taken together, these findings suggest that the mothers who are the most in need, that is, those who feel the least efficacious and the most depressed, receive the most parenting support from others. These mothers may either solicit this assistance, or others in their support networks may recognize the need, and thus step into help mothers and children. In other words, high levels of parenting support may be indicative of network or family-based crisis management due to maternal dysfunction. These effects may be most apparent among single mothers, at least in terms of parenting efficacy, because without spousal support, single mothers may rely more on extended family members for support (Cairney et al. 2003; Hrdy 2009). Second, given Latino, especially Mexican–American, cultural values such as familism that endorse the involvement of extended family members in raising children (Niska 1999; Shin 2013), the extended networks of Latino/Hispanic mothers may be particularly responsive to mother and child needs, and thus will be most likely to assist when mothers feel ineffective as parents or depressed. The protective function of parenting support for children, perhaps especially when maternal functioning is impaired, should be explored in future research.

Alternatively, the findings linking increased parenting support to decreased maternal functioning may stem from the reciprocity that is inherent in social networks. While social support network members may provide crucial support to mothers, those same network members may also demand support from mothers, thus contributing to increased stress (Antonucci et al. 1998; Cochran and Niegro 1995; Taylor et al. 2012). Because single mothers may be particularly reliant on support networks, they may also be particularly burdened by the demands of support networks. Likewise, if there are cultural norms endorsing caring for extended family members even at the cost of your own well-being, then Latino/Hispanic mothers who receive the most support may also give the most support. Another interpretation is that the involvement of other adults in caring for children is perceived by mothers as interference, especially if this involvement is unwanted, and thus the support itself may make mothers feel less capable, leading to reduced parenting efficacy or increased depressive symptoms.

Study Strengths and Limitations

This study has several strengths, including measurement of children’s perceptions of parenting support provided by non-parental adults, an under-studied family process, among a predominantly economically disadvantaged sample composed of roughly equal numbers of Latino/Hispanic (predominantly Mexican origin) and non-Latino/Hispanic families. Broadly, the links between parenting support and maternal functioning and the consistent nomination by mothers and children of grandparents and other family members such as aunts and uncles underscores the value of considering extended family social support networks in models of family functioning, perhaps especially among low-income and ethnic minority families (Barnett 2008; Harknett and Knab 2007; Osofsky and Thompson 2000). These associations were above and beyond and in the opposite direction of the associations between mother-reported social-emotional support and depressive symptoms, thus suggesting a unique role for parenting support measured as direct engagement in caregiving activities. Importantly, we did not include other forms of support that extended family members may provide to mothers and children, such as financial support or parenting advice.

The study also has several limitations. First, the cross-sectional nature of the data precludes drawing conclusions regarding directionality or causality. This leads us to speculate as to whether higher levels of parenting support may be the cause or consequence of poor maternal functioning, as indexed by parenting efficacy and depressive symptoms. Although we have suggested that interactions between extended family support networks and maternal functioning are best conceptualized using a Process, Person, Context, Time model (Bronfenbrenner and Morris 2006), we have not, in fact, measured time. Second, we conducted analyses separately for parenting stress and depressive symptoms, and thus we cannot assume that the different patterns of predictors are in fact statistically significantly different from each other. Third, maternal report of race/ethnicity as Latino/Hispanic versus others is a crude approximation for ethnicity that may fail to capture culturally based variations in family functioning. Importantly, post hoc analyses that substituted language of interview (Spanish or English) resulted in a similar pattern of findings. Nonetheless, future work should consider more dynamic dimensions of sociocultural contexts, including acculturation and family values (e.g., Rodriguez et al. 2007; Schwartz et al. 2010). Finally, the generalizability of the findings beyond this sample may be limited.

Study Implications

The present findings have several implications for research and practice. In addition to the need for longitudinal studies that follow the same families over time to hint at causal processes, future work should also consider the direct and indirect implications of parenting support for children’s development. That is, while we focused on how children’s reports of parenting support are related to mothers’ wellbeing, there may be other implications for children. If non-parental adults are stepping in to care for children because mothers are struggling, then understanding if this support does in fact compensate for compromised maternal functioning by protecting or enhancing child development is a key question for researchers and practitioners. For example, Sterrett et al. (2009) demonstrated that youth reported non-parental involvement was associated with fewer behavior problems only when paired with positive maternal parenting among African American single mother families. Moreover, given that most children in the present study identified maternal grandmothers as playing key supportive roles, future research should explore the implications of intergenerational caregiving networks for the well-being of all family members. Turning to more practical implications, identifying variations in the function of social support, including parenting support, by specific family and maternal characteristics may aid in the identification of families most at-risk and in need of services, and the tailoring of services to benefit particular families. For example, these non-parental adults could be incorporated into existing family support and school-based programs. Further, given that the goal of many family intervention programs is to reduce the stressors encountered by parents and to bolster maternal functioning, it is important to consider how maternal perceptions of this support may influence its effectiveness. Finally, if extensive involvement of non-parental adults in raising children is an indicator of compromised maternal wellbeing, then it may be an important index for school personnel or support providers to measure in order to assess family risks and assets.