Early-Life Parent–Child Relationships and Adult Children’s Support of Unpartnered Parents in Later Life

Early-Life Parent–Child Relationships and Adult Children’s Support of Unpartnered Parents in... Abstract Objectives The proportion of older adults who are unpartnered has increased significantly over the past 25 years. Unpartnered older adults often rely on their adult children for support. Most previous studies have focused on proximal factors associated with adult children’s support of their parents, while few have examined distal factors, such as parent–child relationships formed during childhood. This study fills the gap by investigating the direct and indirect associations between early-life parent–child relationships and adult children’s upward transfers to unpartnered parents. Method Data came from two supplements to the Panel Study of Income Dynamics, in which respondents were asked about their relationships with mothers and fathers before age 17 and their transfers of time and money to parents in 2013. Path models were estimated for unpartnered mother–adult child dyads and father–adult child dyads separately. Results For adult children of unpartnered mothers, psychological closeness has a direct, positive association with time transfer, while physical violence has an indirect association with time transfer through adult children’s marital status. For adult children of unpartnered fathers, psychological closeness has neither a direct nor an indirect association with time or money transfer, but physical violence has a direct, negative association with time transfer. Discussion Early-life parent–child relationships play a pivotal role in influencing adult children’s caregiving behavior, both directly and indirectly. Our findings suggest that by improving their relationships with children early in life, parents may be able to increase the amount of time transfer that they receive in late life. Money transfer, Physical violence, Psychological closeness, Time transfer The share of individuals over age 50 in the United States increased from 26% in 1990 to 35% in 2015 (Authors’ calculations based on U.S. Census data). The significant growth of this population is largely attributable to the aging of the Baby Boom generation. Compared with prior cohorts, Boomers are more likely to be unpartnered (Lin & Brown, 2012). Whereas partnered older adults often pool resources together and provide mutual care, unpartnered older adults do not reap these benefits of partnership. Consequently, unpartnered older adults have lower average income and more often live in poverty and rely on public assistance than the partnered (Lin & Brown, 2012). Unpartnered older adults also tend to report more health conditions, worse self-rated health, and more depressive symptoms than their partnered counterparts (Hughes & Waite, 2009). Adult children play a major role in providing care to unpartnered older adults in times of need (Wolff et al., 2017). Prior research on caregiving has generally focused on how adult children’s current resources and constraints, such as marital status, income, and health status, are associated with their support of parents (Couch, Daly, & Wolf, 1999; Laditka & Laditka, 2001; Sarkisian & Gerstel, 2008). While laying an important foundation for understanding adult children’s caregiving behavior, these studies do not answer the question of why adult children differ in their current resources and constraints. Complementing the caregiving literature, the child development literature has suggested that parent–child relationships formed during childhood (referred to as early-life parent–child relationships hereafter) have lasting impacts on adult development. Children’s psychological closeness with and the physical violence they experience by their parents are significantly related to children’s chance of successfully establishing intimate relationships, achieving high socioeconomic status, and maintaining good health during adulthood (Colman & Widom, 2004; Covey, Menard, & Franzese, 2013; Springer, 2009). This line of research, however, has not examined whether children’s successful launch into adulthood could later affect their abilities to help parents when needs arise. To the best of our knowledge, only three studies (Kong & Moorman, 2016; Whitbeck, Hoyt, & Huck, 1994; Whitbeck, Simons, & Conger, 1991) have investigated how early-life parent–child relationships are related to adult children’s caregiving behavior later in life. Whitbeck et al. (1994) and Kong and Moorman (2016) found a direct association of parental rejection and childhood abuse with adult children’s assistance to parents, though Whitbeck et al. (1991) did not. These studies also considered that early-life parent–child relationships may have an indirect association with adult children’s support of their parents in later life, but none of them has examined the mediating roles of adult children’s marital status, income, and health status, even though researchers have viewed these as vital determinants of caregiving behavior (Couch et al., 1999; Laditka & Laditka, 2001; Sarkisian & Gerstel, 2008). The study aims to extend the caregiving literature by including early-life parent–child relationships as a predicator and examining the possible mediating roles of adult children’s current resources and constraints. Using newly available data from the Panel Study of Income Dynamics, we address two research questions: First, do early-life parent–child relationships have an enduring effect on adult children’s time and money transfers to their unpartnered parents in later life? Second, is some of the enduring effect mediated by adult children’s marital status, income, and health status? By clarifying the links among early-life parent–child relationships; adult children’s marital status, income, and health status; and adult children’s caregiving behavior, this study can help us better understand how adult children’s caregiving behavior gradually takes shape over the life course. Life Course Perspective The life course perspective suggests that human development is a life-long process in which the patterns of later-life adaptation are influenced by earlier life experiences (Elder, Johnson, & Crosnoe, 2003). Research on caregiving in later life has extensively examined how adult children care for parents with health problems or economic hardship (McGarry & Schoeni, 1995; Pezzin, Pollak, & Schone, 2015). These studies, however, tend to center on proximal factors, such as adult children’s current resources and constraints, and ignore distal factors, such as early-life parent–child relationships. Research on child development, in contrast, has shown how early-life parent–child relationships set up a chain of events resulting in life conditions that promote or disturb adult development (Currie & Widom, 2010; Lee, Tsenkova, & Carr, 2014; Mullen, Martin, Anderson, Romans, & Herbison, 1996). Nevertheless, this line of research has not examined the impact of adult development on caregiving behavior in later life. In this study, we adopt the life course perspective to connect these two lines of literature. We bring early-life parent–child relationships to the caregiving literature by considering not only that early-life parent–child relationships can continue to influence adult children’s support of their parents in later life, but also that early-life parent–child relationships can influence adult children’s current resources and constraints, which in turn affect the magnitudes of time and money transfers to parents. By integrating proximal and distal predictors, we provide a more comprehensive understanding of adult children’s caregiving behavior than prior studies do. Early-Life Parent–Child Relationships and Later-Life Upward Transfers Parent–child interactions in later life are shaped by the relationships they formed earlier in the life course (Rossi & Rossi, 1990; Silverstein, Conroy, Wang, Giarrusso, & Bengtson, 2002). Psychological closeness with and physical violence by parents during childhood are two important dimensions of early-life parent–child relationships that may later affect adult children’s provision of support to their parents. Prior studies have shown that adult children are motivated to help parents when they have formed secure attachments to them, whereas adult children who have developed avoidant or fearful attachments are relatively less likely to help their parents when parents’ needs arise (Carpenter, 2001; Cicirelli, 1993). Felt psychological closeness to parents while growing up is likely to enable individuals to form secure attachment to parents, whereas the experience of parental physical violence during childhood may facilitate the development of avoidant or fearful attachments to parents. Not only can early-life parent–child relationships continue to influence parent–child interactions in later life, but early-life parent–child relationships also can influence whether children successfully launch into adulthood, which in turn affects the resources that adult children have to help their parents. Examining the mediating roles of adult children’s marital status, income, and health status would provide a more comprehensive understanding of adult children’s caregiving behavior. Marital Status Close relationships with parents during childhood foster a sense of trust and security that enables adult children to form and maintain healthy intimate relationships (Belt & Abidin, 1996). Conversely, adults with histories of parental physical violence learn to use aggression or withdraw communication in intimate relationships, often leading to marital dissolution (Colman & Widom, 2004; Mullen et al., 1996). At the same time, couples who remain married may be distracted from investing in relationships with extended kin, because marriage requires the couples’ time and energy to maintain their household. Relative to their unmarried counterparts, married adult children tend to give less instrumental support and financial assistance to their parents (Pezzin et al., 2015; Sarkisian & Gerstel, 2008). Income Researchers have found that parent–child closeness is positively associated with children’s academic performance in high school (Turley, Desmond, & Bruch, 2010). By comparison, children who experienced physical violence by their parents are less likely to do well in school, which in turn influences their socioeconomic status attainment and results in lower incomes (Covey, Menard, & Franzese, 2013; Currie & Widom, 2010). As adult children’s hourly wages increase, time becomes more valuable, and thus adult children with higher incomes may spend less time providing instrumental support to parents. Nonetheless, higher income relaxes the family’s budget constraints, permitting greater financial assistance to parents. Several studies have shown that adult children’ incomes are positively related to the provision of money transfers to parents, but they are inversely or unrelated to time transfers (Couch et al., 1999; McGarry & Schoeni, 1995). Health Status Childhood is a critical period during which individuals develop psychosocial resources that they can use to conquer adverse circumstances in adult life and that can have lasting impacts on health. Emotional support received from parents during childhood helps children develop a sense of self-worth, learn effective ways to exert personal control, and establish supportive social relationships. These psychosocial resources subsequently help people handle adversities and maintain mental and physical health during adulthood (Shaw, Krause, Chatters, Connell, & Ingersoll-Dayton, 2004). By contrast, children who are exposed to parental physical violence are more likely to develop depression and poor coping strategies, such as stress-induced eating, cigarette smoking, and heavy drinking, than those who are not, which in turn could worsen their health during adulthood (Lee et al., 2014; Springer, 2009). Caregiving often places enormous mental and physical demands on caregivers. Therefore, adults who are healthier are more likely to become caregivers of their parents and remain in the caregiver role (Fredman, Lyons, Cauley, Hochberg, & Applebaum, 2015; Laditka & Laditka, 2001). Nevertheless, not all studies have found that adult children’s health is related to their provision of instrumental support to parents (e.g., Fingerman, VanderDrift, Dotterer, Birditt, & Zarit, 2011). In sum, the child development literature has established the association of early-life parent–child relationships with adult children’s marital status, income, and health status; at the same time, the caregiving literature has shown the association of adult children’s marital status, income, and health status with upward transfers of time and money to their parents. Together, these two lines of literature suggest that early-life parent–child relationships are linked to later-life upward transfer through adult children’s marital status, income, and health status. Variation by Parent’s Gender Mothers and fathers differ in both their relationships with children and their needs for support in later life. Because mothers are more likely than fathers to express love and affection toward their children, adult children tend to recall more positive early-life relationships with mothers than with fathers (Schafer, Morton, & Ferraro, 2014). Although fathers use corporal punishment on children less often than mothers (Straus & Stewart, 1999), when they do, they tend to use harsher forms of physical discipline (Bender et al., 2007; Nobes, Smith, Upton, & Heverin, 1999). On average, mothers spend about twice as much time with their children as fathers do (Parker & Wang, 2013). Therefore, mother–child relationships may have greater influences on shaping children’s intimate relationships, academic performance, and psychosocial resources than do father–child relationships. In addition, unpartnered women experience greater economic and health vulnerabilities compared to unpartnered men (Dupre, Beck, & Meadows, 2009; Lin, Brown, & Hammersmith, 2017). Given that women and men differ in their relationships with children and their needs for support, it is important to examine unpartnered mothers and fathers separately. The Present Study This study fills the gap in the literature by examining the long-term effect of early-life parent–child relationships on later-life upward transfers, as well as the mediating roles of adult children’s marital status, income, and health status. Based on the findings from the child development literature and the caregiving literature, we anticipated that adult children’s psychological closeness to parents while growing up would be directly related to their greater transfers of time and money to parents, but that parental physical violence would be directly associated with fewer upward transfers. In addition, we expected some of these associations to be mediated by adult children’s marital status, income, and health status. Specifically, we conjectured psychological closeness to be positively but physical violence to be inversely related to adult children’s marital status, income, and health status. Adult children’s resources and constraints, in turn, are hypothesized to have different associations with upward transfers. We anticipated: (a) being married to be negatively related to the provision of time and money transfers, (b) higher incomes to be positively associated with money transfer but negatively related to time transfer, and (c) health conditions to be inversely correlated with transfers of time and money. We also expected that psychological closeness has a greater influence on upward transfers to mothers, whereas physical violence has a greater influence on upward transfers to fathers. Because most mothers are the primary caregivers shaping children’s future trajectories, we conjectured that the indirect association, mediated by adult children’s marital status, income, and health status, would be stronger for mothers than for fathers. In this study, we also considered other factors that might be associated with upward transfers, including parents’ resources and needs, adult children’s demographic characteristics, and family composition. Parents with fewer resources and greater needs, such as those who are older and who have poorer health or lower income, need more help from adult children (Henretta, Soldo, & Van Voorhis, 2011; Pezzin et al., 2015). As for adult children’s characteristics, extant studies have demonstrated that daughters, minorities, and adult children with more resources are more likely than their respective counterparts to provide help to their parents (Henretta et al., 2011; Pezzin et al., 2015; Wong, Kitayama, & Soldo, 1999). Parent–child proximity also has been shown to be positively associated with instrumental support (Pillemer & Suitor, 2014). In addition, holding parents’ needs constant, the more siblings adult children have, the less likely they are to provide time and money transfers to their parents (Henretta et al., 2011; Pezzin et al., 2015). We also expected these characteristics to be correlated with early-life parent–child relationships and adult children’s marital status, income, and health status (Belt & Abidin, 1996; Covey et al., 2013; Shaw et al., 2004). Method Data came from two supplements to the Panel Study of Income Dynamics (PSID): the 2013 Family Roster and Transfer Module (R&T) and the 2014 Childhood Retrospective Circumstances Study (CRCS). The PSID began in 1968, with an original sample of 18,230 individuals living in 4,802 households. The original sample and their descendants were followed annually until 1997 and every other year thereafter. In 2013, the R&T was administered to respondents who had participated in the PSID core survey to help identify the structure of extended families and the flow of intergenerational assistance within and between households in contemporary American society (Schoeni, Bianchi, Hotz, Seltzer, & Wiemers, 2015). The R&T asked respondents for information about their parents, as well as the time and money transfers they provided to their parents who were younger than 80 years old. In 2014, the CRCS was conducted to collect information on the 2013 PSID participants’ childhood circumstances (McGonagle & Freedman, 2015) by asking them to identify the woman and the man who spent the most time raising them before age 17 and their relationships with these parent figures. Combining the R&T and the CRCS with the PSID core survey offered a unique opportunity to examine how early-life parent–child relationships are related to adult children’s support of their unpartnered parents in later life and whether this association is mediated by adult children’s marital status, income, and health status. In 2013, 9,063 households containing 13,697 individuals were interviewed in the core survey (including the R&T), with a response rate of 91% (PSID Main Interview User Manual, 2015). Of them, 12,985 were subsequently selected for the CRCS (excluding proxies and Spanish interviews), and 8,072 completed the self-administered survey, with a response rate of 62% (McGonagle & Freedman, 2015). Altogether, 7,851 individuals completed the R&T, the CRCS, and the PSID core survey. Because of the focus of this study, we restricted our analytic sample to respondents whose parents were unpartnered in 2013 (when transfer information was requested), yielding 3,207 parent–adult child dyads. After further excluding dyads in which parents were younger than age 50 (n = 428) and parents who were not the parent figures when the respondent was growing up (n = 384), the final sample comprised 1,837 mother–adult child dyads and 558 father–adult child dyads. Measures Upward transfers were measured by whether respondents gave time and money transfers to their parents during the year prior to the R&T interview. Time transfer included hours spent on giving help with errands, rides, chores, or hands-on care. Money transfer referred to the amount of money, loans, or gifts of $100 or more given. Because the distributions of these two measures were skewed, log transformation was used in the multivariate analysis. Psychological closeness was appraised by four questions asked about mother figures and father figures, respectively, in the CRCS. Respondents were asked before age 17: (a) how much respondents could confide in their mother/father figure about things that were bothering them, (b) how much their mother/father figure understood respondents’ problems and worries, (c) how emotionally close respondents were with their mother/father figure, and (d) how much love and affection mother/father figure gave respondents. Response categories ranged from 1 (not at all), 2 (a little or not very), 3 (some or somewhat), to 4 (a lot or very). These four questions were adapted from the National Survey of Midlife Development in the U.S. Study (e.g., Shaw et al., 2004) and were highly correlated (α = 0.90). Physical violence was gauged using four questions asked about mother figures and father figures, respectively, in the CRCS. Respondents were asked how often before age 17 their mother/father figure (a) pushed, grabbed, or shoved the respondents; (b) slap or hit the respondents; (c) threw something at the respondents; and (d) physically harmed the respondents in any other way. Response categories included 1 (never), 2 (not very often), 3 (sometimes), and 4 (often). These questions were adapted from the Conflict Tactics Scale (Straus, Gelles, & Steinmetz, 2006), with a reliability of 0.84. Parent’s characteristics included age, marital status, educational attainment, employment status, and health status at the time of the respondents’ interviews. Age was measured in years. Marital status included being widowed (coded 1) or divorced or separated (coded 0). Educational attainment consisted of 9 categories: no education (= 0), 0–5th grade (= 1), 6–8th grade (= 2), 9–11th grade (= 3), 12th grade (= 4), 12th grade plus nonacademic training (= 5), some college (= 6), college degree (= 7), and advanced degree (= 8). Employment status comprised working (coded 0) or not (coded 1). Health status ranged from 1 (poor) to 4 (excellent). Two family characteristics during the respondents’ childhood were also considered: The number of time periods when the family was struggling financially (before age 5, ages 6–12, and ages 13–17, ranging from 0 to 3) and whether parents had alcohol or drug problems while the respondents were growing up (0 = No, 1 = Yes). Adult child’s characteristics encompassed marital status, family income, number of health conditions, gender, age, race, educational attainment, employment status, presence of any child at home, and parent–child proximity. Marital status was a dichotomous variable (0 = unmarried, 1 = married or cohabiting). Family income was measured in 2013 dollars. Because the distribution of family income was skewed, log transformation was used in the multivariate analysis. The number of health conditions was measured by the number of chronic conditions (e.g., heart attack, hypertension, diabetes, etc.) that had been diagnosed by a doctor, ranging from 0 (no conditions) to 11. Daughter was coded 1 and son was 0. Age and education were measured in years. Race consisted of whites (reference category) versus nonwhites. Employment status was captured by working (coded 0) or not (coded 1). The presence of any offspring in the respondent’s household was a dichotomous variable (1 = yes; 0 = no). Parent–child proximity was measured in miles, using information from the PSID restricted file, where parent–child coresidence was coded 0. Because the distribution of parent–child proximity was skewed, log transformation was used in the multivariate analysis. Finally, family composition, a continuous measure, was assessed by the number of the respondent’s siblings. Analytic Strategy Two analyses were conducted. First, we identified the characteristics of unpartnered parents and their adult children, using the mean or percentage (as appropriate), and examined whether these characteristics differed significantly between mother–adult child dyads and father–adult child dyads. Next, we estimated a path model, as illustrated in Figure 1, to test our hypotheses regarding the associations of early-life parent–child relationships, adult children’s resources, and upward time and money transfers. In this path model, psychological closeness and physical violence are viewed as two dimensions of early-life parent–child relationships that have direct associations with upward transfer of time or money, as well as indirect associations through adult children’s marital status, income, and health conditions. Figure 1. View largeDownload slide Path model of how early-life parent–child relationships are associated with upward transfers to unpartnered parents through adult children's marital status, income, and health conditions. Figure 1. View largeDownload slide Path model of how early-life parent–child relationships are associated with upward transfers to unpartnered parents through adult children's marital status, income, and health conditions. We treated psychological closeness and physical violence as two latent factors to control for possible measure errors in the items and let these two factors be correlated. We also allowed the residuals of adult children’s marital status, income, and health conditions to be related, as those who are married tend to have higher incomes and better health than those who are not. Moreover, prior studies have shown that adult children who provide parents with time transfers are also likely to give money transfers (Couch et al., 1999). Thus, we let the residuals of these two types of transfers to be correlated. Because most respondents did not provide transfers to their parents during the year prior to the interview, we used tobit regressions for the hour and dollar measures in the path models. In addition, we included parents’ and adult children’s characteristics and family composition in the models and allowed them to be correlated with early-life parent–child relationships, adult children’s resources, and upward time and money transfers for possible confounding associations. This path analysis was conducted for mothers and fathers separately. We first estimated the path model specified in Figure 1. The model fits the data well for both mothers and fathers, as their comparative fit indexes (CFI) and Tucker–Lewis indexes (TLI) exceed 0.95 and the root mean square errors of approximations (RMSEAs) are smaller than 0.06 (Hu & Bentler, 1999). We further adopted the model-trimming approach (Kline, 2011) to find the most parsimonious model that uses the fewest parameters but fits data equally well as the model illustrated in Figure 1. Specifically, we gradually constrained nonsignificant coefficients to 0, one at a time, and compared the more-restricted model with the less-restricted one. We repeated the procedure until further pruning any path would significantly worsen the model’s goodness of fit. The path models were estimated using the WLSMV estimator provided in the statistical package Mplus version 7.4 (Muthén & Muthén, 2015). Information was missing in some respondents’ reports, ranging from 0.001% (number of siblings) to 11% (parent’s education). We used a multiple imputation procedure to handle missing cases, such that the missing value for a single variable was imputed as a function of other covariates in the analysis (Acock, 2005). To preserve the randomness of imputed variables, the study results were based on 10 random, multiple-imputed replicates. All estimates were weighted to adjust for unequal probabilities of selection, nonresponse, and sample attrition (PSID Main Interview User Manual, 2015). Results Characteristics of Unpartnered Parents and Their Adult Children The prevalence and magnitude of time and money transfers given to unpartnered mothers and fathers by their adult children differed somewhat, as shown in Table 1. For time transfer, adult children of unpartnered mothers were more likely to help and spent more hours helping their mothers with errands, rides, chores, or hands-on care as compared with adult children of unpartnered fathers with respect to their fathers. Nearly one half of adult children gave time transfers to their mothers (48%), whereas slightly more than one third of adult children helped their fathers (36%). Among adult children who helped, they spent an average of 6 hr (= 300.33/52) and 4 hr (= 202.60/52) each week assisting their mothers and fathers, respectively. By contrast, fewer adult children provided financial support to their parents. Only 15% of adult children gave money to their mothers, compared to 10% to fathers. When adult children helped, they tended to give slightly more money to their fathers than mothers ($1,647 vs $1,378), but the differences in the prevalence and magnitude of financial assistance to mothers versus fathers did not reach statistical significance. Table 1. Weighted Descriptive Statistics for All Variables by Unpartnered Parent’s Gender   Mothers  Fathers  Difference    Mean or %  SD  Mean or %  SD  Upward time transfer   Any help  48.30    35.94    **   If helps, number of hours last year  300.33  617.24  202.60  345.65  *  Upward money transfer   Any help  15.10    9.98       If helps, dollar amounts last year  1,378.36  4,278.27  1,647.17  2,795.27    Psychological closenessa   Confide  2.74  1.03  2.20  0.99  ***   Understand  2.93  0.96  2.36  0.97  ***   Closeness  3.18  0.85  2.70  0.98  ***   Love  3.37  0.83  2.94  0.91  ***  Physical violenceb   Push, grab, or shove  1.39  0.72  1.41  0.73     Slap or hit  1.62  0.82  1.60  0.82     Throw things at  1.23  0.58  1.18  0.54     Physically harm  1.18  0.52  1.24  0.62    Parent’s characteristics   Age  73.03  11.89  70.54  12.52  **   Widowed  69.01    46.74    ***   Education  4.64  1.62  5.06  1.63  ***   Working  22.08    31.05    **   Health  2.90  1.14  2.96  1.16     Financial struggles before child age 17  1.12  1.26  1.08  1.21     Alcohol/drug problem before child age 17  7.70    19.96    ***  Adult child’s characteristics   Married  61.00    52.65    *   Family income (median, in $1,000)  70.60    65.67       Number of health conditions  1.00  1.31  0.79  1.05  *   Daughter  54.89    57.39       Age  46.90  11.78  42.18  12.24  ***   White  79.23    77.98       Education  13.99  2.19  14.39  2.15  **   Working  71.74    77.87    *   Any child lives in household  35.37    40.50       Distance from parent (in miles)  212.28  489.25  255.17  590.06    Family composition   Number of siblings  2.68  1.97  2.46  1.89     Unweighted N  1,837  558      Mothers  Fathers  Difference    Mean or %  SD  Mean or %  SD  Upward time transfer   Any help  48.30    35.94    **   If helps, number of hours last year  300.33  617.24  202.60  345.65  *  Upward money transfer   Any help  15.10    9.98       If helps, dollar amounts last year  1,378.36  4,278.27  1,647.17  2,795.27    Psychological closenessa   Confide  2.74  1.03  2.20  0.99  ***   Understand  2.93  0.96  2.36  0.97  ***   Closeness  3.18  0.85  2.70  0.98  ***   Love  3.37  0.83  2.94  0.91  ***  Physical violenceb   Push, grab, or shove  1.39  0.72  1.41  0.73     Slap or hit  1.62  0.82  1.60  0.82     Throw things at  1.23  0.58  1.18  0.54     Physically harm  1.18  0.52  1.24  0.62    Parent’s characteristics   Age  73.03  11.89  70.54  12.52  **   Widowed  69.01    46.74    ***   Education  4.64  1.62  5.06  1.63  ***   Working  22.08    31.05    **   Health  2.90  1.14  2.96  1.16     Financial struggles before child age 17  1.12  1.26  1.08  1.21     Alcohol/drug problem before child age 17  7.70    19.96    ***  Adult child’s characteristics   Married  61.00    52.65    *   Family income (median, in $1,000)  70.60    65.67       Number of health conditions  1.00  1.31  0.79  1.05  *   Daughter  54.89    57.39       Age  46.90  11.78  42.18  12.24  ***   White  79.23    77.98       Education  13.99  2.19  14.39  2.15  **   Working  71.74    77.87    *   Any child lives in household  35.37    40.50       Distance from parent (in miles)  212.28  489.25  255.17  590.06    Family composition   Number of siblings  2.68  1.97  2.46  1.89     Unweighted N  1,837  558    Note: *p < .05, **p < .01, ***p < .001. aFour-point scale ranging from not at all (=1), a little/not very (=2), some/somewhat (=3) to a lot/very (=4). bFour-point scale ranging from never (=1), not very often (=2), sometimes (=3), to often (=4). View Large Table 1. Weighted Descriptive Statistics for All Variables by Unpartnered Parent’s Gender   Mothers  Fathers  Difference    Mean or %  SD  Mean or %  SD  Upward time transfer   Any help  48.30    35.94    **   If helps, number of hours last year  300.33  617.24  202.60  345.65  *  Upward money transfer   Any help  15.10    9.98       If helps, dollar amounts last year  1,378.36  4,278.27  1,647.17  2,795.27    Psychological closenessa   Confide  2.74  1.03  2.20  0.99  ***   Understand  2.93  0.96  2.36  0.97  ***   Closeness  3.18  0.85  2.70  0.98  ***   Love  3.37  0.83  2.94  0.91  ***  Physical violenceb   Push, grab, or shove  1.39  0.72  1.41  0.73     Slap or hit  1.62  0.82  1.60  0.82     Throw things at  1.23  0.58  1.18  0.54     Physically harm  1.18  0.52  1.24  0.62    Parent’s characteristics   Age  73.03  11.89  70.54  12.52  **   Widowed  69.01    46.74    ***   Education  4.64  1.62  5.06  1.63  ***   Working  22.08    31.05    **   Health  2.90  1.14  2.96  1.16     Financial struggles before child age 17  1.12  1.26  1.08  1.21     Alcohol/drug problem before child age 17  7.70    19.96    ***  Adult child’s characteristics   Married  61.00    52.65    *   Family income (median, in $1,000)  70.60    65.67       Number of health conditions  1.00  1.31  0.79  1.05  *   Daughter  54.89    57.39       Age  46.90  11.78  42.18  12.24  ***   White  79.23    77.98       Education  13.99  2.19  14.39  2.15  **   Working  71.74    77.87    *   Any child lives in household  35.37    40.50       Distance from parent (in miles)  212.28  489.25  255.17  590.06    Family composition   Number of siblings  2.68  1.97  2.46  1.89     Unweighted N  1,837  558      Mothers  Fathers  Difference    Mean or %  SD  Mean or %  SD  Upward time transfer   Any help  48.30    35.94    **   If helps, number of hours last year  300.33  617.24  202.60  345.65  *  Upward money transfer   Any help  15.10    9.98       If helps, dollar amounts last year  1,378.36  4,278.27  1,647.17  2,795.27    Psychological closenessa   Confide  2.74  1.03  2.20  0.99  ***   Understand  2.93  0.96  2.36  0.97  ***   Closeness  3.18  0.85  2.70  0.98  ***   Love  3.37  0.83  2.94  0.91  ***  Physical violenceb   Push, grab, or shove  1.39  0.72  1.41  0.73     Slap or hit  1.62  0.82  1.60  0.82     Throw things at  1.23  0.58  1.18  0.54     Physically harm  1.18  0.52  1.24  0.62    Parent’s characteristics   Age  73.03  11.89  70.54  12.52  **   Widowed  69.01    46.74    ***   Education  4.64  1.62  5.06  1.63  ***   Working  22.08    31.05    **   Health  2.90  1.14  2.96  1.16     Financial struggles before child age 17  1.12  1.26  1.08  1.21     Alcohol/drug problem before child age 17  7.70    19.96    ***  Adult child’s characteristics   Married  61.00    52.65    *   Family income (median, in $1,000)  70.60    65.67       Number of health conditions  1.00  1.31  0.79  1.05  *   Daughter  54.89    57.39       Age  46.90  11.78  42.18  12.24  ***   White  79.23    77.98       Education  13.99  2.19  14.39  2.15  **   Working  71.74    77.87    *   Any child lives in household  35.37    40.50       Distance from parent (in miles)  212.28  489.25  255.17  590.06    Family composition   Number of siblings  2.68  1.97  2.46  1.89     Unweighted N  1,837  558    Note: *p < .05, **p < .01, ***p < .001. aFour-point scale ranging from not at all (=1), a little/not very (=2), some/somewhat (=3) to a lot/very (=4). bFour-point scale ranging from never (=1), not very often (=2), sometimes (=3), to often (=4). View Large Early-life parent–child relationships comprised two latent constructs, psychological closeness and physical violence, and each construct was gauged by four items with responses ranging from 1 (not at all/never), 2 (a little/not very/not very often), 3 (some/somewhat/sometimes) to 4 (a lot/very/often). Adult children recalled moderate levels of psychological closeness to their parents while growing up, with overall averages ranging from 2.20 to 3.37, but adult children of unpartnered mothers felt psychologically closer to their mothers than adult children of unpartnered fathers did to their fathers. Specifically, compared with adult children of unpartnered fathers with respect to their fathers, adult children of unpartnered mothers responded that they could confide in their mothers more (2.74 vs 2.20), that their mothers understood their problems and worries more (2.93 vs 2.36), that they were emotionally closer to their mothers (3.18 vs 2.70), and that their mothers gave them more love and affection (3.37 vs 2.94). Different from psychological closeness, the levels of physical violence were on the lower end of the scale (ranging from 1.18 to 1.62), and there was no gender difference. Mothers and fathers were equally likely to push, grab, or shove (1.39 vs 1.41); slap or hit (1.62 vs 1.60); throw something at (1.23 vs 1.18); or physically harm (1.18 vs 1.24) the respondents prior to age 17. Unpartnered mothers and fathers in the sample differed in several aspects. Unpartnered mothers tended to be older than unpartnered fathers (73 vs 71 years old). Two thirds of the mothers were widows, whereas less than one half of fathers were widowers. Fathers had more education than mothers, with a high-school degree, on average. Fathers were more likely than mothers to be working at the time of the respondents’ interviews (31% vs 22%). In addition, more fathers than mothers experienced alcohol or drug problems while respondents were growing up (20% vs 8%). Nonetheless, the differences between mothers’ and fathers’ current health and their financial struggles during the respondents’ childhood did not reach statistical significance. Adult children of unpartnered mothers and those of unpartnered fathers were equally likely to be white (more than three quarters), women (more than one half), and having offspring living in the household (more than one third). Moreover, they had a median family income of $65,000, lived about 200 miles away from their parents, and had two siblings. Despite these similarities, adult children of unpartnered mothers were older than those of unpartnered fathers (47 vs 42 years old), which may explain why adult children of unpartnered mothers were more likely to be married (61% vs 53%), had more health conditions (1 vs 0.79), were less educated (13.99 vs 14.39), and were less likely to be working (72% vs 78%) than adult children of unpartnered fathers. Early-Life Mother–Child Relationships and Support of Unpartnered Mothers We applied the path model illustrated in Figure 1 to the unpartnered mother–adult child sample and took the model-trimming approach to obtain the most parsimonious model. The final model fits the data well (Hu & Bentler, 1999), with the model fit statistics RMSEA, CFI, and TLI equal to 0.01, 0.99, and 0.99, respectively. To simplify the presentation, we show only the coefficients pertinent to early-life mother–child relationships; adult children’s marital status, income, and health conditions; and time and money transfers in Figure 2. All coefficients involved confounding variables are presented in Supplementary Table 1. Figure 2. View largeDownload slide Path model of how early-life mother–child relationships are associated with upward transfers to unpartnered mothers through adult children's marital status, income, and health conditions. Figure 2. View largeDownload slide Path model of how early-life mother–child relationships are associated with upward transfers to unpartnered mothers through adult children's marital status, income, and health conditions. We found that psychological closeness and physical violence had different patterns of association with adult children’s transfers of time and money to their unpartnered mothers. Specifically, the factor loadings of the four items measuring psychological closeness ranged from 0.82 to 0.99, indicating that the latent construct, psychological closeness, represented a large amount of the commonality among these items. Moreover, adult children who felt psychologically close to their mothers while growing up spent more hours helping their mothers in later life (0.71, p < .001). Such a direct association, however, was not found between psychological closeness and money transfer. Concerning indirect associations, we found that psychological closeness was related to income only (−0.05, p < .05), but income was unrelated to either time or money transfer, suggesting no indirect association between psychological closeness and transfer of time or money through adult children’s marital status, income, or health conditions. With regard to physical violence, the factor loadings of its four indicators varied from 0.62 to 0.96, signifying that the latent construct, physical violence, well represented these items. We did not find a direct association between physical violence and transfer of time or money. Nonetheless, physical violence was significantly related to marital status (−0.18, p < .01), income (−0.09, p < .01), and health conditions (0.27, p < .001). Whereas income and health conditions were not associated with transfer of time or money, marital status was significantly related to time transfer (−0.66, p < .001), indicating that physical violence had an indirect association with time transfer via its influence on adult children’s marital status (0.12, p < .03). As for the confounding variables, mothers’ and adult children’s characteristics and family composition were also related to time and money transfers, as shown in Supplementary Table 1. Adult children spent more hours helping mothers when the mothers had more education or poorer health; when the families experienced longer periods of financial struggles while the adult children were growing up; and when the adult children were daughters, lived closer, or had fewer siblings. Mothers’ characteristics and family composition were unrelated to adult children’s money transfer, but nonwhite children and children with more education gave more financial support to their mothers than their respective counterparts. As expected, more mothers’ characteristics (e.g., financial struggles and the use of alcohol and drugs) were associated with early-life mother–child relationships than were adult children’s characteristics, whereas more adult children’s characteristics (e.g., age, race, employment status, parenthood, and proximity) were related to their marital status, income, and health conditions than were mothers’ characteristics. Early-Life Father–Child Relationships and Support of Unpartnered Fathers We also applied the path model illustrated in Figure 1 to the unpartnered father–adult child sample and took the model-trimming approach to obtain the most parsimonious model. The final model also fits the data well, with the model fit statistics RMSEA, CFI, and TLI equal to 0.01, 0.96, and 0.95, respectively. We show only the coefficients pertinent to early-life father–child relationships; adult children’s marital status, income, and health conditions; and time and money transfers to fathers in Figure 3. All coefficients involved confounding variables are presented in Supplementary Table 2. Figure 3. View largeDownload slide Path model of how early-life father–child relationships are associated with upward transfers to unpartnered fathers through adult children's marital status, income, and health conditions. Figure 3. View largeDownload slide Path model of how early-life father–child relationships are associated with upward transfers to unpartnered fathers through adult children's marital status, income, and health conditions. As shown in Figure 3, both latent constructs, psychological closeness and physical violence, well represent their respective measures, as the factor loadings range from 0.85 to 0.95 for the former and from 0.52 to 0.94 for the latter. The path model also reveals that psychological closeness did not have a direct association with either time or money transfer. In addition, psychological closeness was unrelated to marital status, income, or health conditions, indicating that psychological closeness did not indirectly relate to time or money transfer through its associations with adult children’s resources and constraints. Like psychological closeness, physical violence was unrelated to marital status, income, or health conditions, so physical violence did not have an indirect association with time or money transfer. Nonetheless, adult children who more often experienced physical violence by fathers before age 17 spent fewer hours helping the fathers in later life (−2.24, p < .001). Physical violence, however, did not have a direct association with money transfer. Only two covariates were associated with adult children’s time transfer to fathers, as shown in Supplementary Table 2. Adult children spent more hours helping their fathers when the fathers had poorer health or when adult children lived closer to the fathers. None of the covariates was significantly related to adult children’s money transfer to fathers. Like the findings for mothers, more fathers’ characteristics (e.g., financial struggles and the use of alcohol and drugs) were associated with early-life father–child relationships than were adult children’s characteristics, whereas more adult children’s characteristics (e.g., race, employment status, and parenthood) were related to their marital status, income, and health conditions than were fathers’ characteristics. Discussion How to care for unpartnered older adults is a rising social concern, as the number of unpartnered older adults, who face greater economic and health vulnerabilities than the partnered (Hughes & Waite, 2009; Lin & Brown, 2012), continues to grow. Caregiving literature has extensively examined how adult children’s current resources and constraints influence their abilities to help parents (Couch et al., 1999; Laditka & Laditka, 2001; Sarkisian & Gerstel, 2008), but rarely considered relationships that adult children formed with their parents during childhood. Drawing on the life course perspective (Elder et al., 2003), we bring the child development literature to caregiving research to provide new insights into adult children’s caregiving behavior. We found that mothers’ use of physical violence on their children was inversely associated with adult children’s marital status, which, in turn, was negatively related to time transfer to the mothers. This finding suggests that adult children’s marital status, which scholars have long viewed as a key determinant of caregiving is actually a mediator of the association of early-life mother–child relationships and adult children’s support of unpartnered mothers in later life, an insight that would have been overlooked if early-life mother–child relationships had not been considered in the study. After taking into account various possible mediating and confounding variables, we found that early-life parent–child relationships still have unique and independent associations with adult children’s caregiving behavior. Specifically, adult children who were psychologically closer to their mothers while growing up spent more hours helping them in later life than did adult children who were less attached. In addition, adult children who more often experienced physical violence by their fathers while growing up spent fewer hours helping them in later life than did those who experienced physical violence less often. These findings indicate that early-life parent–child relationships have an influence on adult children’s time transfer to their parents in later life that caregiving researchers should not ignore. We also found that psychological closeness and physical violence have different patterns of associations with adult children’s time transfer to unpartnered mothers and fathers, respectively. Such differences are likely attributable to the different roles that mothers and fathers play in childrearing. Mothers are usually the primary caregivers—spending more time caring for children than do fathers (Parker & Wang, 2013). The patterns of parent–child interactions in later life reflect the psychological bonds formed early in the life course (Rossi & Rossi, 1990), which may explain the direct association of psychological closeness with time transfer to mothers, but not fathers. Although mothers also discipline children more often than fathers, the negative consequences may be buffered by maternal warmth (McKee et al., 2007). This may explain why physical violence was not found to be directly associated with time transfer to mothers. Nevertheless, children could still imitate aggressive behavior from mothers who exercised physical violence and use it in their own intimate relationships (Colman & Widom, 2004; Mullen et al., 1996). Along this line, we observed an indirect association between mothers’ use of physical violence and time transfer through adult children’s marital status. Fathers less often discipline children (Straus & Stewart, 1999), but when they do, they tend to discipline children more harshly than mothers (Bender et al., 2007; Nobes et al., 1999). The use of harsh discipline during childhood may have a scarring effect on subsequent father–child interactions, which may explain why we observed a direct association between physical violence and time transfer to fathers. There are several null findings. Early-life parent–child relationships have neither direct nor indirect associations with adult children’s provision of financial assistance to their mothers or fathers. This likely occurs because both mothers and fathers in the sample were relatively well-off, with a median family income of $65,000 (shown in Table 1). When parents have less financial need, there may not be enough variance in the sample to identify possible covariates. We also did not find any pathways from early-life parent–child relationships to upward transfers through adult children’s income and health conditions. The lack of mediating effects through income is likely because we controlled for family financial struggles while the respondents were growing up; such financial struggles were strongly associated with early-life parent–child relationships and time transfer, at least for mother–adult child dyads (shown in Supplementary Table 1). In addition, the lack of mediating effects through adult children’s health may be attributable to the fact that most adult children in the sample were relatively young and had few health conditions. This study has some limitations. First, like other studies examining early-life parent–child relationships (Kong & Moorman, 2016; Whitbeck et al., 1991; Whitbeck et al., 1994), we also relied on respondents’ retrospective reports, which could have been influenced by their current support of parents. Yet, prior studies have established good reliability and validity for respondents’ recall of childhood family circumstances (Hardt & Rutter, 2004). Second, this study focused on three types of resources and constraints that could mediate the associations between early-life parent–child relationships and upward transfers. There may be other factors that could mediate such associations, such as parents’ expectations for caregiving (Pillemer & Suitor, 2014). Future studies should continue to explore possible meditating factors. Third, we found fewer significant paths for the father sample than for the mother sample, likely due to the former’s relatively smaller sample size. Further studies using larger samples of both mothers and fathers are needed. Finally, we have controlled more confounding variables than previous studies on a similar topic did (Kong & Moorman, 2016; Whitbeck et al., 1991; Whitbeck et al., 1994). Nevertheless, other confounding variables that were not available in current study may be worth considering in the future, including parental divorce during childhood and the geographic location of siblings. Despite these limitations, this study makes important theoretical and practical contributions to the existing caregiving literature. Regarding theoretical contributions, this study shows that early-life parent–child relationships play a pivotal role in influencing adult children’s caregiving behavior, both directly and indirectly. Moreover, the pathways through which early-life parent–child relationships are associated with adult children’s caregiving behavior differ for unpartnered mothers and fathers, underscoring the importance of developing respective theoretical frameworks for mothers and fathers. As for practical contributions, practitioners generally suggest that providing caregivers with respite services and psychosocial support can enhance caregiving to older adults, because they reduce caregiver burden, help caregivers cope with stress, and maintain caregivers’ well-being (Lopez-Hartmann, Wens, Verhoeven, & Remmen, 2012). Interestingly, practitioners seldom discuss what older adults can do to increase the likelihood of receiving support from their adult children. This study demonstrates that parents may be able to increase support from their adult children by fostering psychological closeness and avoiding the use of physical violence when their children are growing up. Some parents may have difficulty creating and maintaining warm and violence-free relationships with their children, and social programs should be designed for and provided to these parents. As the number of unpartnered older adults continues to grow, many will need help from their adult children. It is important for researchers and practitioners to recognize that adult children’s decisions on whether to help their parents rest not only on their current resources and constraints, but also on the relationships that adult children had with their parents while growing up. Supplementary Material Supplementary data is available at The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences online. Funding This research was supported in part by the Center for Family and Demographic Research, Bowling Green State University, which has core funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (P2CHD050959). Conflict of Interest I.-F. Lin serves on Journal Editorial Board. Acknowledgments The authors thank the editor and two anonymous reviewers for their helpful comments. Any opinions expressed here are solely those of the authors and not of the funding agency or center. Author Contributions: I.-F. Lin and H.-S. Wu conceived the study, planned the analyses, and wrote the article. I.-F. Lin prepared the data file and performed the analyses. References Acock, A. C. ( 2005). Working with missing values. Journal of Marriage and Family , 67, 1012– 1028. doi: 10.1111/j.1741- 3737.2005.00191.x Google Scholar CrossRef Search ADS   Belt, W., & Abidin, R. R. ( 1996). The relation of childhood abuse and early parenting experiences to current marital quality in a nonclinical sample. Child Abuse & Neglect , 20, 1019– 1030. doi: 10.1016/0145-2134(96)00092-0 Google Scholar CrossRef Search ADS PubMed  Bender, H. L., Allen, J. P., McElhaney, K. B., Antonishak, J., Moore, C. M., Kelly, H. O., & Davis, S. M. ( 2007). Use of harsh physical discipline and developmental outcomes in adolescence. Development and Psychopathology , 19, 227– 242. doi: 10.1017/S0954579407070125 Google Scholar CrossRef Search ADS PubMed  Carpenter, B. D. ( 2001). Attachment bonds between adult daughters and their older mothers: associations with contemporary caregiving. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 56, P257– P266. doi: 10.1093/geronb/56.5.P257 Google Scholar CrossRef Search ADS   Cicirelli, V. G. ( 1993). Attachment and obligation as daughters’ motives for caregiving behavior and subsequent effect on subjective burden. Psychology and Aging , 8, 144– 155. doi: 10.1037/0882-7974.8.2.144 Google Scholar CrossRef Search ADS PubMed  Colman, R. A., & Widom, C. S. ( 2004). Childhood abuse and neglect and adult intimate relationships: A prospective study. Child Abuse & Neglect , 28, 1133– 1151. doi: 10.1016/j.chiabu.2004.02.005 Google Scholar CrossRef Search ADS PubMed  Couch, K. A., Daly, M. C., & Wolf, D. A. ( 1999). Time? Money? Both? The allocation of resources to older parents. Demography , 36, 219– 232. doi: 10.2307/2648110 Google Scholar CrossRef Search ADS PubMed  Covey, H. C., Menard, S., & Franzese, R. J. ( 2013). Effects of adolescent physical abuse, exposure to neighborhood violence, and witnessing parental violence on adult socioeconomic status. Child Maltreatment , 18, 85– 97. doi: 10.1177/1077559513477914 Google Scholar CrossRef Search ADS PubMed  Currie, J., & Widom, C. S. ( 2010). Long-term consequences of child abuse and neglect on adult economic well-being. Child Maltreatment , 15, 111– 120. doi: 10.1177/10775 59509355316 Google Scholar CrossRef Search ADS PubMed  Dupre, M. E., Beck, A. N., & Meadows, S. O. ( 2009). Marital trajectories and mortality among US adults. American Journal of Epidemiology , 170, 546– 555. doi: 10.1093/aje/kwp194 Google Scholar CrossRef Search ADS PubMed  Elder, G. H., Johnson, M. K., & Crosnoe, R. ( 2003). The emergence and development of life course theory. In J. T. Mortimer & M. J. Shanahan (Eds.), Handbook of the life course  (pp. 3– 19). New York, NY: Kluwer Academic. doi: 10.1007/978-0-306-48247-2_1 Google Scholar CrossRef Search ADS   Fingerman, K. L., VanderDrift, L. E., Dotterer, A. M., Birditt, K. S., & Zarit, S. H. ( 2011). Support to aging parents and grown children in black and white families. The Gerontologist , 51, 441– 452. doi: 10.1093/geront/gnq114 Google Scholar CrossRef Search ADS PubMed  Fredman, L., Lyons, J. G., Cauley, J. A., Hochberg, M., & Applebaum, K. M. ( 2015). The relationship between caregiving and mortality after accounting for time-varying caregiver status and addressing the healthy caregiver hypothesis. The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences , 70, 1163– 1168. doi: 10.1093/gerona/glv009 Google Scholar CrossRef Search ADS   Hardt, J., & Rutter, M. ( 2004). Validity of adult retrospective reports of adverse childhood experiences: Review of the evidence. Journal of Child Psychology and Psychiatry, and Allied Disciplines , 45, 260– 273. doi: 10.1111/j.1469-7610.2004.00218.x Google Scholar CrossRef Search ADS PubMed  Henretta, J. C., Soldo, B. J., & Van Voorhis, M. F. ( 2011). Why do families differ? Children’s care for an unmarried mother. Journal of Marriage and the Family , 73, 383– 395. doi: 10.1111/j.1741-3737.2010.00813.x Google Scholar CrossRef Search ADS PubMed  Hu, L.-T., & Bentler, P. M. ( 1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling , 6, 1– 55. doi: 10.1080/10705519909540118 Google Scholar CrossRef Search ADS   Hughes, M. E., & Waite, L. J. ( 2009). Marital biography and health at mid-life. Journal of Health and Social Behavior , 50, 344– 358. doi: 10.1177/002214650905000307 Google Scholar CrossRef Search ADS PubMed  Kline, R. B. ( 2011). Principles and practice of structural equation modeling  ( 3rd ed.). New York, NY: The Guilford Press. Kong, J., & Moorman, S. M. ( 2016). History of childhood abuse and intergenerational support to mothers in adulthood. Journal of Marriage and Family , 78, 926– 938. doi: 10.1111/jomf.12285 Google Scholar CrossRef Search ADS   Laditka, J. N., & Laditka, S. B. ( 2001). Adult children helping older parents: Variations in likelihood and hours by gender, race, and family role. Research on Aging , 23, 429– 456. doi: 10.1177/0164027501234003 Google Scholar CrossRef Search ADS   Lee, C., Tsenkova, V., & Carr, D. ( 2014). Childhood trauma and metabolic syndrome in men and women. Social Science & Medicine (1982) , 105, 122– 130. doi: 10.1016/j.socscimed.2014.01.017 Google Scholar CrossRef Search ADS PubMed  Lin, I.-F., & Brown, S. L. ( 2012). Unmarried Boomers confront old age: A national portrait. The Gerontologist , 52, 153– 165. doi: 10.1093/geront/gnr141 Google Scholar CrossRef Search ADS PubMed  Lin, I.-F., Brown, S. L., & Hammersmith, A. M. ( 2017). Marital biography, social security receipt, and poverty. Research on Aging , 39, 86– 110. doi: 10.1177/0164027516656139 Google Scholar CrossRef Search ADS PubMed  Lopez-Hartmann, M., Wens, J., Verhoeven, V., & Remmen, R. ( 2012). The effect of caregiver support interventions for informal caregivers of community-dwelling frail elderly: A systematic review. International Journal of Integrated Care , 12, e133. doi: 10.5334/ijic.845 Google Scholar PubMed  McGarry, K., & Schoeni, R. F. ( 1995). Transfer behavior in the health and retirement study: Measurement and the redistribution of resources within the family. The Journal of Human Resources , 30, S184– S226. doi: 10.2307/146283 Google Scholar CrossRef Search ADS   McGonagle, K., & Freedman, V. A. ( 2015). The panel study of income dynamics’ childhood retrospective circumstances study (PSID-CRCS) user guide: Final release 1 . Institute for Social Research, University of Michigan. Retrieved from https://psidonline.isr.umich.edu/CRCS/2014UserGuide.pdf. McKee, L., Roland, E., Coffelt, N., Olson, A. L., Forehand, R., Massari, C., Jones, D., Gaffney, C. A., & Zens, M. S. ( 2007). Harsh discipline and child problem behaviors: The roles of positive parenting and gender. Journal of Family Violence , 22, 187– 196. doi: 10.1007/s10896-007-9070-6 Google Scholar CrossRef Search ADS   Mullen, P. E., Martin, J. L., Anderson, J. C., Romans, S. E., & Herbison, G. P. ( 1996). The long-term impact of the physical, emotional, and sexual abuse of children: A community study. Child Abuse & Neglect , 20, 7– 21. doi: 10.1016/0145-2134(95)00112-3 Google Scholar CrossRef Search ADS PubMed  Muthén, L. K., & Muthén, B. O. ( 2015). Mplus user’s guide  (7th ed.). Los Angeles, CA: Muthén & Muthén. Nobes, G., Smith, M., Upton, P., & Heverin, A. ( 1999). Physical punishment by mothers and fathers in British homes. Journal of Interpersonal Violence , 14, 887– 902. doi: 10.1177/088626099014008006 Google Scholar CrossRef Search ADS   Parker, K., & Wang, W. ( 2013). Modern parenthood: Roles of moms and dads converge as they balance work and family . Washington DC: Pew Research Center. Retrieved from http://www.pewsocialtrends.org/2013/03/14/modern-parenthood-roles-of-moms-and-dads-converge-as-they-balance-work-and-family/. Pezzin, L. E., Pollak, R. A., & Schone, B. S. ( 2015). Bargaining power, parental caregiving, and intergenerational coresidence. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 70, 969– 980. doi: 10.1093/geronb/gbu079 Google Scholar CrossRef Search ADS   Pillemer, K., & Suitor, J. J. ( 2014). Who provides care? A prospective study of caregiving among adult siblings. The Gerontologist , 54, 589– 598. doi: 10.1093/geront/gnt066 Google Scholar CrossRef Search ADS PubMed  PSID Main Interview User Manual( 2015). Institute for Social Research , University of Michigan. Retrieved from http://psidonline.isr.umich.edu/data/Documentation/UserGuide2013.pdf. Rossi, A. S., & Rossi, P. H. ( 1990). Of human bonding: Parent-child relations across the life course . New York, NY: Aldine de Gruyter. Sarkisian, N., & Gerstel, N. ( 2008). Till marriage do us part: Adult children’s relationships with their parents. Journal of Marriage and Family , 70, 360– 376. doi: 10.1111/j.1741-3737.2008.00487.x Google Scholar CrossRef Search ADS   Schafer, M. H., Morton, P. M., & Ferraro, K. F. ( 2014). Child maltreatment and adult health in a national sample: Heterogeneous relational contexts, divergent effects? Child Abuse & Neglect , 38, 395– 406. doi: 10.1016/j.chiabu.2013.08.003 Google Scholar CrossRef Search ADS PubMed  Schoeni, R. F., Bianchi, S. M., Hotz, V. J., Seltzer, J. A., & Wiemers, E. E. ( 2015). Intergenerational transfers and rosters of the extended family: A new substudy of the panel study of income dynamics. Longitudinal and Life Course Studies , 6, 319– 330. doi: 10.14301/llcs.v6i3.332 Google Scholar CrossRef Search ADS PubMed  Shaw, B. A., Krause, N., Chatters, L. M., Connell, C. M., & Ingersoll-Dayton, B. ( 2004). Emotional support from parents early in life, aging, and health. Psychology and Aging , 19, 4– 12. doi: 10.1037/0882-7974.19.1.4 Google Scholar CrossRef Search ADS PubMed  Silverstein, M., Conroy, S. J., Wang, H., Giarrusso, R., & Bengtson, V. L. ( 2002). Reciprocity in parent-child relations over the adult life course. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 57, S3– S13. doi: 10.1093/geronb/57.1.S3 Google Scholar CrossRef Search ADS   Springer, K. W. ( 2009). Childhood physical abuse and midlife physical health: Testing a multi-pathway life course model. Social Science & Medicine , 69, 138– 146. doi: 10.1016/j.socscimed.2009.04.011 Google Scholar CrossRef Search ADS   Straus, M. A., Gelles, R. J., & Steinmetz, S. K. ( 2006). Behind closed doors: Violence in the American family . New Brunswick, NJ: Transaction Publishers. Straus, M. A., & Stewart, J. H. ( 1999). Corporal punishment by American parents: National data on prevalence, chronicity, severity, and duration, in relation to child and family characteristics. Clinical Child and Family Psychology Review , 2, 55– 70. doi: 10.1023/A:1021891529770 Google Scholar CrossRef Search ADS PubMed  Turley, R. N. L., Desmond, M., & Bruch, S. K. ( 2010). Unanticipated educational consequences of a positive parent-child relationship. Journal of Marriage and Family , 72, 1377– 1390. doi: 10.1111/ j.1741-3737.2010.00771.x Google Scholar CrossRef Search ADS   Whitbeck, L., Hoyt, D. R., & Huck, S. M. ( 1994). Early family relationships, intergenerational solidarity, and support provided to parents by their adult children. Journal of Gerontology , 49, S85– S94. doi: 10.1093/geronj/49.2.S85 Google Scholar CrossRef Search ADS PubMed  Whitbeck, L. B., Simons, R. L., & Conger, R. D. ( 1991). The effects of early family relationships on contemporary relationships and assistance patterns between adult children and their parents. Journal of Gerontology , 46, S330– S337. doi: 10.1093/geronj/46.6.s330 Google Scholar CrossRef Search ADS PubMed  Wolff, J. L., Mulcahy, J., Huang, J., Roth, D. L., Covinsky, K., & Kasper, J. D. ( 2017). Family caregivers of older adults, 1999–2015: Trends in characteristics, circumstances, and role-related appraisal. The Gerontologist . doi: 10.1093/geront/gnx093 Wong, R., Kitayama, K. E., & Soldo, B. J. ( 1999). Ethnic differences in time transfers from adult children to elderly parents: Unobserved heterogeneity across families? Research on Aging , 21, 144– 175. doi: 10.1177/0164027599212002 Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journals of Gerontology Series B: Psychological Sciences and Social Sciences Oxford University Press

Early-Life Parent–Child Relationships and Adult Children’s Support of Unpartnered Parents in Later Life

Loading next page...
 
/lp/ou_press/early-life-parent-child-relationships-and-adult-children-s-support-of-0BamF30FqI
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
1079-5014
eISSN
1758-5368
D.O.I.
10.1093/geronb/gby020
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives The proportion of older adults who are unpartnered has increased significantly over the past 25 years. Unpartnered older adults often rely on their adult children for support. Most previous studies have focused on proximal factors associated with adult children’s support of their parents, while few have examined distal factors, such as parent–child relationships formed during childhood. This study fills the gap by investigating the direct and indirect associations between early-life parent–child relationships and adult children’s upward transfers to unpartnered parents. Method Data came from two supplements to the Panel Study of Income Dynamics, in which respondents were asked about their relationships with mothers and fathers before age 17 and their transfers of time and money to parents in 2013. Path models were estimated for unpartnered mother–adult child dyads and father–adult child dyads separately. Results For adult children of unpartnered mothers, psychological closeness has a direct, positive association with time transfer, while physical violence has an indirect association with time transfer through adult children’s marital status. For adult children of unpartnered fathers, psychological closeness has neither a direct nor an indirect association with time or money transfer, but physical violence has a direct, negative association with time transfer. Discussion Early-life parent–child relationships play a pivotal role in influencing adult children’s caregiving behavior, both directly and indirectly. Our findings suggest that by improving their relationships with children early in life, parents may be able to increase the amount of time transfer that they receive in late life. Money transfer, Physical violence, Psychological closeness, Time transfer The share of individuals over age 50 in the United States increased from 26% in 1990 to 35% in 2015 (Authors’ calculations based on U.S. Census data). The significant growth of this population is largely attributable to the aging of the Baby Boom generation. Compared with prior cohorts, Boomers are more likely to be unpartnered (Lin & Brown, 2012). Whereas partnered older adults often pool resources together and provide mutual care, unpartnered older adults do not reap these benefits of partnership. Consequently, unpartnered older adults have lower average income and more often live in poverty and rely on public assistance than the partnered (Lin & Brown, 2012). Unpartnered older adults also tend to report more health conditions, worse self-rated health, and more depressive symptoms than their partnered counterparts (Hughes & Waite, 2009). Adult children play a major role in providing care to unpartnered older adults in times of need (Wolff et al., 2017). Prior research on caregiving has generally focused on how adult children’s current resources and constraints, such as marital status, income, and health status, are associated with their support of parents (Couch, Daly, & Wolf, 1999; Laditka & Laditka, 2001; Sarkisian & Gerstel, 2008). While laying an important foundation for understanding adult children’s caregiving behavior, these studies do not answer the question of why adult children differ in their current resources and constraints. Complementing the caregiving literature, the child development literature has suggested that parent–child relationships formed during childhood (referred to as early-life parent–child relationships hereafter) have lasting impacts on adult development. Children’s psychological closeness with and the physical violence they experience by their parents are significantly related to children’s chance of successfully establishing intimate relationships, achieving high socioeconomic status, and maintaining good health during adulthood (Colman & Widom, 2004; Covey, Menard, & Franzese, 2013; Springer, 2009). This line of research, however, has not examined whether children’s successful launch into adulthood could later affect their abilities to help parents when needs arise. To the best of our knowledge, only three studies (Kong & Moorman, 2016; Whitbeck, Hoyt, & Huck, 1994; Whitbeck, Simons, & Conger, 1991) have investigated how early-life parent–child relationships are related to adult children’s caregiving behavior later in life. Whitbeck et al. (1994) and Kong and Moorman (2016) found a direct association of parental rejection and childhood abuse with adult children’s assistance to parents, though Whitbeck et al. (1991) did not. These studies also considered that early-life parent–child relationships may have an indirect association with adult children’s support of their parents in later life, but none of them has examined the mediating roles of adult children’s marital status, income, and health status, even though researchers have viewed these as vital determinants of caregiving behavior (Couch et al., 1999; Laditka & Laditka, 2001; Sarkisian & Gerstel, 2008). The study aims to extend the caregiving literature by including early-life parent–child relationships as a predicator and examining the possible mediating roles of adult children’s current resources and constraints. Using newly available data from the Panel Study of Income Dynamics, we address two research questions: First, do early-life parent–child relationships have an enduring effect on adult children’s time and money transfers to their unpartnered parents in later life? Second, is some of the enduring effect mediated by adult children’s marital status, income, and health status? By clarifying the links among early-life parent–child relationships; adult children’s marital status, income, and health status; and adult children’s caregiving behavior, this study can help us better understand how adult children’s caregiving behavior gradually takes shape over the life course. Life Course Perspective The life course perspective suggests that human development is a life-long process in which the patterns of later-life adaptation are influenced by earlier life experiences (Elder, Johnson, & Crosnoe, 2003). Research on caregiving in later life has extensively examined how adult children care for parents with health problems or economic hardship (McGarry & Schoeni, 1995; Pezzin, Pollak, & Schone, 2015). These studies, however, tend to center on proximal factors, such as adult children’s current resources and constraints, and ignore distal factors, such as early-life parent–child relationships. Research on child development, in contrast, has shown how early-life parent–child relationships set up a chain of events resulting in life conditions that promote or disturb adult development (Currie & Widom, 2010; Lee, Tsenkova, & Carr, 2014; Mullen, Martin, Anderson, Romans, & Herbison, 1996). Nevertheless, this line of research has not examined the impact of adult development on caregiving behavior in later life. In this study, we adopt the life course perspective to connect these two lines of literature. We bring early-life parent–child relationships to the caregiving literature by considering not only that early-life parent–child relationships can continue to influence adult children’s support of their parents in later life, but also that early-life parent–child relationships can influence adult children’s current resources and constraints, which in turn affect the magnitudes of time and money transfers to parents. By integrating proximal and distal predictors, we provide a more comprehensive understanding of adult children’s caregiving behavior than prior studies do. Early-Life Parent–Child Relationships and Later-Life Upward Transfers Parent–child interactions in later life are shaped by the relationships they formed earlier in the life course (Rossi & Rossi, 1990; Silverstein, Conroy, Wang, Giarrusso, & Bengtson, 2002). Psychological closeness with and physical violence by parents during childhood are two important dimensions of early-life parent–child relationships that may later affect adult children’s provision of support to their parents. Prior studies have shown that adult children are motivated to help parents when they have formed secure attachments to them, whereas adult children who have developed avoidant or fearful attachments are relatively less likely to help their parents when parents’ needs arise (Carpenter, 2001; Cicirelli, 1993). Felt psychological closeness to parents while growing up is likely to enable individuals to form secure attachment to parents, whereas the experience of parental physical violence during childhood may facilitate the development of avoidant or fearful attachments to parents. Not only can early-life parent–child relationships continue to influence parent–child interactions in later life, but early-life parent–child relationships also can influence whether children successfully launch into adulthood, which in turn affects the resources that adult children have to help their parents. Examining the mediating roles of adult children’s marital status, income, and health status would provide a more comprehensive understanding of adult children’s caregiving behavior. Marital Status Close relationships with parents during childhood foster a sense of trust and security that enables adult children to form and maintain healthy intimate relationships (Belt & Abidin, 1996). Conversely, adults with histories of parental physical violence learn to use aggression or withdraw communication in intimate relationships, often leading to marital dissolution (Colman & Widom, 2004; Mullen et al., 1996). At the same time, couples who remain married may be distracted from investing in relationships with extended kin, because marriage requires the couples’ time and energy to maintain their household. Relative to their unmarried counterparts, married adult children tend to give less instrumental support and financial assistance to their parents (Pezzin et al., 2015; Sarkisian & Gerstel, 2008). Income Researchers have found that parent–child closeness is positively associated with children’s academic performance in high school (Turley, Desmond, & Bruch, 2010). By comparison, children who experienced physical violence by their parents are less likely to do well in school, which in turn influences their socioeconomic status attainment and results in lower incomes (Covey, Menard, & Franzese, 2013; Currie & Widom, 2010). As adult children’s hourly wages increase, time becomes more valuable, and thus adult children with higher incomes may spend less time providing instrumental support to parents. Nonetheless, higher income relaxes the family’s budget constraints, permitting greater financial assistance to parents. Several studies have shown that adult children’ incomes are positively related to the provision of money transfers to parents, but they are inversely or unrelated to time transfers (Couch et al., 1999; McGarry & Schoeni, 1995). Health Status Childhood is a critical period during which individuals develop psychosocial resources that they can use to conquer adverse circumstances in adult life and that can have lasting impacts on health. Emotional support received from parents during childhood helps children develop a sense of self-worth, learn effective ways to exert personal control, and establish supportive social relationships. These psychosocial resources subsequently help people handle adversities and maintain mental and physical health during adulthood (Shaw, Krause, Chatters, Connell, & Ingersoll-Dayton, 2004). By contrast, children who are exposed to parental physical violence are more likely to develop depression and poor coping strategies, such as stress-induced eating, cigarette smoking, and heavy drinking, than those who are not, which in turn could worsen their health during adulthood (Lee et al., 2014; Springer, 2009). Caregiving often places enormous mental and physical demands on caregivers. Therefore, adults who are healthier are more likely to become caregivers of their parents and remain in the caregiver role (Fredman, Lyons, Cauley, Hochberg, & Applebaum, 2015; Laditka & Laditka, 2001). Nevertheless, not all studies have found that adult children’s health is related to their provision of instrumental support to parents (e.g., Fingerman, VanderDrift, Dotterer, Birditt, & Zarit, 2011). In sum, the child development literature has established the association of early-life parent–child relationships with adult children’s marital status, income, and health status; at the same time, the caregiving literature has shown the association of adult children’s marital status, income, and health status with upward transfers of time and money to their parents. Together, these two lines of literature suggest that early-life parent–child relationships are linked to later-life upward transfer through adult children’s marital status, income, and health status. Variation by Parent’s Gender Mothers and fathers differ in both their relationships with children and their needs for support in later life. Because mothers are more likely than fathers to express love and affection toward their children, adult children tend to recall more positive early-life relationships with mothers than with fathers (Schafer, Morton, & Ferraro, 2014). Although fathers use corporal punishment on children less often than mothers (Straus & Stewart, 1999), when they do, they tend to use harsher forms of physical discipline (Bender et al., 2007; Nobes, Smith, Upton, & Heverin, 1999). On average, mothers spend about twice as much time with their children as fathers do (Parker & Wang, 2013). Therefore, mother–child relationships may have greater influences on shaping children’s intimate relationships, academic performance, and psychosocial resources than do father–child relationships. In addition, unpartnered women experience greater economic and health vulnerabilities compared to unpartnered men (Dupre, Beck, & Meadows, 2009; Lin, Brown, & Hammersmith, 2017). Given that women and men differ in their relationships with children and their needs for support, it is important to examine unpartnered mothers and fathers separately. The Present Study This study fills the gap in the literature by examining the long-term effect of early-life parent–child relationships on later-life upward transfers, as well as the mediating roles of adult children’s marital status, income, and health status. Based on the findings from the child development literature and the caregiving literature, we anticipated that adult children’s psychological closeness to parents while growing up would be directly related to their greater transfers of time and money to parents, but that parental physical violence would be directly associated with fewer upward transfers. In addition, we expected some of these associations to be mediated by adult children’s marital status, income, and health status. Specifically, we conjectured psychological closeness to be positively but physical violence to be inversely related to adult children’s marital status, income, and health status. Adult children’s resources and constraints, in turn, are hypothesized to have different associations with upward transfers. We anticipated: (a) being married to be negatively related to the provision of time and money transfers, (b) higher incomes to be positively associated with money transfer but negatively related to time transfer, and (c) health conditions to be inversely correlated with transfers of time and money. We also expected that psychological closeness has a greater influence on upward transfers to mothers, whereas physical violence has a greater influence on upward transfers to fathers. Because most mothers are the primary caregivers shaping children’s future trajectories, we conjectured that the indirect association, mediated by adult children’s marital status, income, and health status, would be stronger for mothers than for fathers. In this study, we also considered other factors that might be associated with upward transfers, including parents’ resources and needs, adult children’s demographic characteristics, and family composition. Parents with fewer resources and greater needs, such as those who are older and who have poorer health or lower income, need more help from adult children (Henretta, Soldo, & Van Voorhis, 2011; Pezzin et al., 2015). As for adult children’s characteristics, extant studies have demonstrated that daughters, minorities, and adult children with more resources are more likely than their respective counterparts to provide help to their parents (Henretta et al., 2011; Pezzin et al., 2015; Wong, Kitayama, & Soldo, 1999). Parent–child proximity also has been shown to be positively associated with instrumental support (Pillemer & Suitor, 2014). In addition, holding parents’ needs constant, the more siblings adult children have, the less likely they are to provide time and money transfers to their parents (Henretta et al., 2011; Pezzin et al., 2015). We also expected these characteristics to be correlated with early-life parent–child relationships and adult children’s marital status, income, and health status (Belt & Abidin, 1996; Covey et al., 2013; Shaw et al., 2004). Method Data came from two supplements to the Panel Study of Income Dynamics (PSID): the 2013 Family Roster and Transfer Module (R&T) and the 2014 Childhood Retrospective Circumstances Study (CRCS). The PSID began in 1968, with an original sample of 18,230 individuals living in 4,802 households. The original sample and their descendants were followed annually until 1997 and every other year thereafter. In 2013, the R&T was administered to respondents who had participated in the PSID core survey to help identify the structure of extended families and the flow of intergenerational assistance within and between households in contemporary American society (Schoeni, Bianchi, Hotz, Seltzer, & Wiemers, 2015). The R&T asked respondents for information about their parents, as well as the time and money transfers they provided to their parents who were younger than 80 years old. In 2014, the CRCS was conducted to collect information on the 2013 PSID participants’ childhood circumstances (McGonagle & Freedman, 2015) by asking them to identify the woman and the man who spent the most time raising them before age 17 and their relationships with these parent figures. Combining the R&T and the CRCS with the PSID core survey offered a unique opportunity to examine how early-life parent–child relationships are related to adult children’s support of their unpartnered parents in later life and whether this association is mediated by adult children’s marital status, income, and health status. In 2013, 9,063 households containing 13,697 individuals were interviewed in the core survey (including the R&T), with a response rate of 91% (PSID Main Interview User Manual, 2015). Of them, 12,985 were subsequently selected for the CRCS (excluding proxies and Spanish interviews), and 8,072 completed the self-administered survey, with a response rate of 62% (McGonagle & Freedman, 2015). Altogether, 7,851 individuals completed the R&T, the CRCS, and the PSID core survey. Because of the focus of this study, we restricted our analytic sample to respondents whose parents were unpartnered in 2013 (when transfer information was requested), yielding 3,207 parent–adult child dyads. After further excluding dyads in which parents were younger than age 50 (n = 428) and parents who were not the parent figures when the respondent was growing up (n = 384), the final sample comprised 1,837 mother–adult child dyads and 558 father–adult child dyads. Measures Upward transfers were measured by whether respondents gave time and money transfers to their parents during the year prior to the R&T interview. Time transfer included hours spent on giving help with errands, rides, chores, or hands-on care. Money transfer referred to the amount of money, loans, or gifts of $100 or more given. Because the distributions of these two measures were skewed, log transformation was used in the multivariate analysis. Psychological closeness was appraised by four questions asked about mother figures and father figures, respectively, in the CRCS. Respondents were asked before age 17: (a) how much respondents could confide in their mother/father figure about things that were bothering them, (b) how much their mother/father figure understood respondents’ problems and worries, (c) how emotionally close respondents were with their mother/father figure, and (d) how much love and affection mother/father figure gave respondents. Response categories ranged from 1 (not at all), 2 (a little or not very), 3 (some or somewhat), to 4 (a lot or very). These four questions were adapted from the National Survey of Midlife Development in the U.S. Study (e.g., Shaw et al., 2004) and were highly correlated (α = 0.90). Physical violence was gauged using four questions asked about mother figures and father figures, respectively, in the CRCS. Respondents were asked how often before age 17 their mother/father figure (a) pushed, grabbed, or shoved the respondents; (b) slap or hit the respondents; (c) threw something at the respondents; and (d) physically harmed the respondents in any other way. Response categories included 1 (never), 2 (not very often), 3 (sometimes), and 4 (often). These questions were adapted from the Conflict Tactics Scale (Straus, Gelles, & Steinmetz, 2006), with a reliability of 0.84. Parent’s characteristics included age, marital status, educational attainment, employment status, and health status at the time of the respondents’ interviews. Age was measured in years. Marital status included being widowed (coded 1) or divorced or separated (coded 0). Educational attainment consisted of 9 categories: no education (= 0), 0–5th grade (= 1), 6–8th grade (= 2), 9–11th grade (= 3), 12th grade (= 4), 12th grade plus nonacademic training (= 5), some college (= 6), college degree (= 7), and advanced degree (= 8). Employment status comprised working (coded 0) or not (coded 1). Health status ranged from 1 (poor) to 4 (excellent). Two family characteristics during the respondents’ childhood were also considered: The number of time periods when the family was struggling financially (before age 5, ages 6–12, and ages 13–17, ranging from 0 to 3) and whether parents had alcohol or drug problems while the respondents were growing up (0 = No, 1 = Yes). Adult child’s characteristics encompassed marital status, family income, number of health conditions, gender, age, race, educational attainment, employment status, presence of any child at home, and parent–child proximity. Marital status was a dichotomous variable (0 = unmarried, 1 = married or cohabiting). Family income was measured in 2013 dollars. Because the distribution of family income was skewed, log transformation was used in the multivariate analysis. The number of health conditions was measured by the number of chronic conditions (e.g., heart attack, hypertension, diabetes, etc.) that had been diagnosed by a doctor, ranging from 0 (no conditions) to 11. Daughter was coded 1 and son was 0. Age and education were measured in years. Race consisted of whites (reference category) versus nonwhites. Employment status was captured by working (coded 0) or not (coded 1). The presence of any offspring in the respondent’s household was a dichotomous variable (1 = yes; 0 = no). Parent–child proximity was measured in miles, using information from the PSID restricted file, where parent–child coresidence was coded 0. Because the distribution of parent–child proximity was skewed, log transformation was used in the multivariate analysis. Finally, family composition, a continuous measure, was assessed by the number of the respondent’s siblings. Analytic Strategy Two analyses were conducted. First, we identified the characteristics of unpartnered parents and their adult children, using the mean or percentage (as appropriate), and examined whether these characteristics differed significantly between mother–adult child dyads and father–adult child dyads. Next, we estimated a path model, as illustrated in Figure 1, to test our hypotheses regarding the associations of early-life parent–child relationships, adult children’s resources, and upward time and money transfers. In this path model, psychological closeness and physical violence are viewed as two dimensions of early-life parent–child relationships that have direct associations with upward transfer of time or money, as well as indirect associations through adult children’s marital status, income, and health conditions. Figure 1. View largeDownload slide Path model of how early-life parent–child relationships are associated with upward transfers to unpartnered parents through adult children's marital status, income, and health conditions. Figure 1. View largeDownload slide Path model of how early-life parent–child relationships are associated with upward transfers to unpartnered parents through adult children's marital status, income, and health conditions. We treated psychological closeness and physical violence as two latent factors to control for possible measure errors in the items and let these two factors be correlated. We also allowed the residuals of adult children’s marital status, income, and health conditions to be related, as those who are married tend to have higher incomes and better health than those who are not. Moreover, prior studies have shown that adult children who provide parents with time transfers are also likely to give money transfers (Couch et al., 1999). Thus, we let the residuals of these two types of transfers to be correlated. Because most respondents did not provide transfers to their parents during the year prior to the interview, we used tobit regressions for the hour and dollar measures in the path models. In addition, we included parents’ and adult children’s characteristics and family composition in the models and allowed them to be correlated with early-life parent–child relationships, adult children’s resources, and upward time and money transfers for possible confounding associations. This path analysis was conducted for mothers and fathers separately. We first estimated the path model specified in Figure 1. The model fits the data well for both mothers and fathers, as their comparative fit indexes (CFI) and Tucker–Lewis indexes (TLI) exceed 0.95 and the root mean square errors of approximations (RMSEAs) are smaller than 0.06 (Hu & Bentler, 1999). We further adopted the model-trimming approach (Kline, 2011) to find the most parsimonious model that uses the fewest parameters but fits data equally well as the model illustrated in Figure 1. Specifically, we gradually constrained nonsignificant coefficients to 0, one at a time, and compared the more-restricted model with the less-restricted one. We repeated the procedure until further pruning any path would significantly worsen the model’s goodness of fit. The path models were estimated using the WLSMV estimator provided in the statistical package Mplus version 7.4 (Muthén & Muthén, 2015). Information was missing in some respondents’ reports, ranging from 0.001% (number of siblings) to 11% (parent’s education). We used a multiple imputation procedure to handle missing cases, such that the missing value for a single variable was imputed as a function of other covariates in the analysis (Acock, 2005). To preserve the randomness of imputed variables, the study results were based on 10 random, multiple-imputed replicates. All estimates were weighted to adjust for unequal probabilities of selection, nonresponse, and sample attrition (PSID Main Interview User Manual, 2015). Results Characteristics of Unpartnered Parents and Their Adult Children The prevalence and magnitude of time and money transfers given to unpartnered mothers and fathers by their adult children differed somewhat, as shown in Table 1. For time transfer, adult children of unpartnered mothers were more likely to help and spent more hours helping their mothers with errands, rides, chores, or hands-on care as compared with adult children of unpartnered fathers with respect to their fathers. Nearly one half of adult children gave time transfers to their mothers (48%), whereas slightly more than one third of adult children helped their fathers (36%). Among adult children who helped, they spent an average of 6 hr (= 300.33/52) and 4 hr (= 202.60/52) each week assisting their mothers and fathers, respectively. By contrast, fewer adult children provided financial support to their parents. Only 15% of adult children gave money to their mothers, compared to 10% to fathers. When adult children helped, they tended to give slightly more money to their fathers than mothers ($1,647 vs $1,378), but the differences in the prevalence and magnitude of financial assistance to mothers versus fathers did not reach statistical significance. Table 1. Weighted Descriptive Statistics for All Variables by Unpartnered Parent’s Gender   Mothers  Fathers  Difference    Mean or %  SD  Mean or %  SD  Upward time transfer   Any help  48.30    35.94    **   If helps, number of hours last year  300.33  617.24  202.60  345.65  *  Upward money transfer   Any help  15.10    9.98       If helps, dollar amounts last year  1,378.36  4,278.27  1,647.17  2,795.27    Psychological closenessa   Confide  2.74  1.03  2.20  0.99  ***   Understand  2.93  0.96  2.36  0.97  ***   Closeness  3.18  0.85  2.70  0.98  ***   Love  3.37  0.83  2.94  0.91  ***  Physical violenceb   Push, grab, or shove  1.39  0.72  1.41  0.73     Slap or hit  1.62  0.82  1.60  0.82     Throw things at  1.23  0.58  1.18  0.54     Physically harm  1.18  0.52  1.24  0.62    Parent’s characteristics   Age  73.03  11.89  70.54  12.52  **   Widowed  69.01    46.74    ***   Education  4.64  1.62  5.06  1.63  ***   Working  22.08    31.05    **   Health  2.90  1.14  2.96  1.16     Financial struggles before child age 17  1.12  1.26  1.08  1.21     Alcohol/drug problem before child age 17  7.70    19.96    ***  Adult child’s characteristics   Married  61.00    52.65    *   Family income (median, in $1,000)  70.60    65.67       Number of health conditions  1.00  1.31  0.79  1.05  *   Daughter  54.89    57.39       Age  46.90  11.78  42.18  12.24  ***   White  79.23    77.98       Education  13.99  2.19  14.39  2.15  **   Working  71.74    77.87    *   Any child lives in household  35.37    40.50       Distance from parent (in miles)  212.28  489.25  255.17  590.06    Family composition   Number of siblings  2.68  1.97  2.46  1.89     Unweighted N  1,837  558      Mothers  Fathers  Difference    Mean or %  SD  Mean or %  SD  Upward time transfer   Any help  48.30    35.94    **   If helps, number of hours last year  300.33  617.24  202.60  345.65  *  Upward money transfer   Any help  15.10    9.98       If helps, dollar amounts last year  1,378.36  4,278.27  1,647.17  2,795.27    Psychological closenessa   Confide  2.74  1.03  2.20  0.99  ***   Understand  2.93  0.96  2.36  0.97  ***   Closeness  3.18  0.85  2.70  0.98  ***   Love  3.37  0.83  2.94  0.91  ***  Physical violenceb   Push, grab, or shove  1.39  0.72  1.41  0.73     Slap or hit  1.62  0.82  1.60  0.82     Throw things at  1.23  0.58  1.18  0.54     Physically harm  1.18  0.52  1.24  0.62    Parent’s characteristics   Age  73.03  11.89  70.54  12.52  **   Widowed  69.01    46.74    ***   Education  4.64  1.62  5.06  1.63  ***   Working  22.08    31.05    **   Health  2.90  1.14  2.96  1.16     Financial struggles before child age 17  1.12  1.26  1.08  1.21     Alcohol/drug problem before child age 17  7.70    19.96    ***  Adult child’s characteristics   Married  61.00    52.65    *   Family income (median, in $1,000)  70.60    65.67       Number of health conditions  1.00  1.31  0.79  1.05  *   Daughter  54.89    57.39       Age  46.90  11.78  42.18  12.24  ***   White  79.23    77.98       Education  13.99  2.19  14.39  2.15  **   Working  71.74    77.87    *   Any child lives in household  35.37    40.50       Distance from parent (in miles)  212.28  489.25  255.17  590.06    Family composition   Number of siblings  2.68  1.97  2.46  1.89     Unweighted N  1,837  558    Note: *p < .05, **p < .01, ***p < .001. aFour-point scale ranging from not at all (=1), a little/not very (=2), some/somewhat (=3) to a lot/very (=4). bFour-point scale ranging from never (=1), not very often (=2), sometimes (=3), to often (=4). View Large Table 1. Weighted Descriptive Statistics for All Variables by Unpartnered Parent’s Gender   Mothers  Fathers  Difference    Mean or %  SD  Mean or %  SD  Upward time transfer   Any help  48.30    35.94    **   If helps, number of hours last year  300.33  617.24  202.60  345.65  *  Upward money transfer   Any help  15.10    9.98       If helps, dollar amounts last year  1,378.36  4,278.27  1,647.17  2,795.27    Psychological closenessa   Confide  2.74  1.03  2.20  0.99  ***   Understand  2.93  0.96  2.36  0.97  ***   Closeness  3.18  0.85  2.70  0.98  ***   Love  3.37  0.83  2.94  0.91  ***  Physical violenceb   Push, grab, or shove  1.39  0.72  1.41  0.73     Slap or hit  1.62  0.82  1.60  0.82     Throw things at  1.23  0.58  1.18  0.54     Physically harm  1.18  0.52  1.24  0.62    Parent’s characteristics   Age  73.03  11.89  70.54  12.52  **   Widowed  69.01    46.74    ***   Education  4.64  1.62  5.06  1.63  ***   Working  22.08    31.05    **   Health  2.90  1.14  2.96  1.16     Financial struggles before child age 17  1.12  1.26  1.08  1.21     Alcohol/drug problem before child age 17  7.70    19.96    ***  Adult child’s characteristics   Married  61.00    52.65    *   Family income (median, in $1,000)  70.60    65.67       Number of health conditions  1.00  1.31  0.79  1.05  *   Daughter  54.89    57.39       Age  46.90  11.78  42.18  12.24  ***   White  79.23    77.98       Education  13.99  2.19  14.39  2.15  **   Working  71.74    77.87    *   Any child lives in household  35.37    40.50       Distance from parent (in miles)  212.28  489.25  255.17  590.06    Family composition   Number of siblings  2.68  1.97  2.46  1.89     Unweighted N  1,837  558      Mothers  Fathers  Difference    Mean or %  SD  Mean or %  SD  Upward time transfer   Any help  48.30    35.94    **   If helps, number of hours last year  300.33  617.24  202.60  345.65  *  Upward money transfer   Any help  15.10    9.98       If helps, dollar amounts last year  1,378.36  4,278.27  1,647.17  2,795.27    Psychological closenessa   Confide  2.74  1.03  2.20  0.99  ***   Understand  2.93  0.96  2.36  0.97  ***   Closeness  3.18  0.85  2.70  0.98  ***   Love  3.37  0.83  2.94  0.91  ***  Physical violenceb   Push, grab, or shove  1.39  0.72  1.41  0.73     Slap or hit  1.62  0.82  1.60  0.82     Throw things at  1.23  0.58  1.18  0.54     Physically harm  1.18  0.52  1.24  0.62    Parent’s characteristics   Age  73.03  11.89  70.54  12.52  **   Widowed  69.01    46.74    ***   Education  4.64  1.62  5.06  1.63  ***   Working  22.08    31.05    **   Health  2.90  1.14  2.96  1.16     Financial struggles before child age 17  1.12  1.26  1.08  1.21     Alcohol/drug problem before child age 17  7.70    19.96    ***  Adult child’s characteristics   Married  61.00    52.65    *   Family income (median, in $1,000)  70.60    65.67       Number of health conditions  1.00  1.31  0.79  1.05  *   Daughter  54.89    57.39       Age  46.90  11.78  42.18  12.24  ***   White  79.23    77.98       Education  13.99  2.19  14.39  2.15  **   Working  71.74    77.87    *   Any child lives in household  35.37    40.50       Distance from parent (in miles)  212.28  489.25  255.17  590.06    Family composition   Number of siblings  2.68  1.97  2.46  1.89     Unweighted N  1,837  558    Note: *p < .05, **p < .01, ***p < .001. aFour-point scale ranging from not at all (=1), a little/not very (=2), some/somewhat (=3) to a lot/very (=4). bFour-point scale ranging from never (=1), not very often (=2), sometimes (=3), to often (=4). View Large Early-life parent–child relationships comprised two latent constructs, psychological closeness and physical violence, and each construct was gauged by four items with responses ranging from 1 (not at all/never), 2 (a little/not very/not very often), 3 (some/somewhat/sometimes) to 4 (a lot/very/often). Adult children recalled moderate levels of psychological closeness to their parents while growing up, with overall averages ranging from 2.20 to 3.37, but adult children of unpartnered mothers felt psychologically closer to their mothers than adult children of unpartnered fathers did to their fathers. Specifically, compared with adult children of unpartnered fathers with respect to their fathers, adult children of unpartnered mothers responded that they could confide in their mothers more (2.74 vs 2.20), that their mothers understood their problems and worries more (2.93 vs 2.36), that they were emotionally closer to their mothers (3.18 vs 2.70), and that their mothers gave them more love and affection (3.37 vs 2.94). Different from psychological closeness, the levels of physical violence were on the lower end of the scale (ranging from 1.18 to 1.62), and there was no gender difference. Mothers and fathers were equally likely to push, grab, or shove (1.39 vs 1.41); slap or hit (1.62 vs 1.60); throw something at (1.23 vs 1.18); or physically harm (1.18 vs 1.24) the respondents prior to age 17. Unpartnered mothers and fathers in the sample differed in several aspects. Unpartnered mothers tended to be older than unpartnered fathers (73 vs 71 years old). Two thirds of the mothers were widows, whereas less than one half of fathers were widowers. Fathers had more education than mothers, with a high-school degree, on average. Fathers were more likely than mothers to be working at the time of the respondents’ interviews (31% vs 22%). In addition, more fathers than mothers experienced alcohol or drug problems while respondents were growing up (20% vs 8%). Nonetheless, the differences between mothers’ and fathers’ current health and their financial struggles during the respondents’ childhood did not reach statistical significance. Adult children of unpartnered mothers and those of unpartnered fathers were equally likely to be white (more than three quarters), women (more than one half), and having offspring living in the household (more than one third). Moreover, they had a median family income of $65,000, lived about 200 miles away from their parents, and had two siblings. Despite these similarities, adult children of unpartnered mothers were older than those of unpartnered fathers (47 vs 42 years old), which may explain why adult children of unpartnered mothers were more likely to be married (61% vs 53%), had more health conditions (1 vs 0.79), were less educated (13.99 vs 14.39), and were less likely to be working (72% vs 78%) than adult children of unpartnered fathers. Early-Life Mother–Child Relationships and Support of Unpartnered Mothers We applied the path model illustrated in Figure 1 to the unpartnered mother–adult child sample and took the model-trimming approach to obtain the most parsimonious model. The final model fits the data well (Hu & Bentler, 1999), with the model fit statistics RMSEA, CFI, and TLI equal to 0.01, 0.99, and 0.99, respectively. To simplify the presentation, we show only the coefficients pertinent to early-life mother–child relationships; adult children’s marital status, income, and health conditions; and time and money transfers in Figure 2. All coefficients involved confounding variables are presented in Supplementary Table 1. Figure 2. View largeDownload slide Path model of how early-life mother–child relationships are associated with upward transfers to unpartnered mothers through adult children's marital status, income, and health conditions. Figure 2. View largeDownload slide Path model of how early-life mother–child relationships are associated with upward transfers to unpartnered mothers through adult children's marital status, income, and health conditions. We found that psychological closeness and physical violence had different patterns of association with adult children’s transfers of time and money to their unpartnered mothers. Specifically, the factor loadings of the four items measuring psychological closeness ranged from 0.82 to 0.99, indicating that the latent construct, psychological closeness, represented a large amount of the commonality among these items. Moreover, adult children who felt psychologically close to their mothers while growing up spent more hours helping their mothers in later life (0.71, p < .001). Such a direct association, however, was not found between psychological closeness and money transfer. Concerning indirect associations, we found that psychological closeness was related to income only (−0.05, p < .05), but income was unrelated to either time or money transfer, suggesting no indirect association between psychological closeness and transfer of time or money through adult children’s marital status, income, or health conditions. With regard to physical violence, the factor loadings of its four indicators varied from 0.62 to 0.96, signifying that the latent construct, physical violence, well represented these items. We did not find a direct association between physical violence and transfer of time or money. Nonetheless, physical violence was significantly related to marital status (−0.18, p < .01), income (−0.09, p < .01), and health conditions (0.27, p < .001). Whereas income and health conditions were not associated with transfer of time or money, marital status was significantly related to time transfer (−0.66, p < .001), indicating that physical violence had an indirect association with time transfer via its influence on adult children’s marital status (0.12, p < .03). As for the confounding variables, mothers’ and adult children’s characteristics and family composition were also related to time and money transfers, as shown in Supplementary Table 1. Adult children spent more hours helping mothers when the mothers had more education or poorer health; when the families experienced longer periods of financial struggles while the adult children were growing up; and when the adult children were daughters, lived closer, or had fewer siblings. Mothers’ characteristics and family composition were unrelated to adult children’s money transfer, but nonwhite children and children with more education gave more financial support to their mothers than their respective counterparts. As expected, more mothers’ characteristics (e.g., financial struggles and the use of alcohol and drugs) were associated with early-life mother–child relationships than were adult children’s characteristics, whereas more adult children’s characteristics (e.g., age, race, employment status, parenthood, and proximity) were related to their marital status, income, and health conditions than were mothers’ characteristics. Early-Life Father–Child Relationships and Support of Unpartnered Fathers We also applied the path model illustrated in Figure 1 to the unpartnered father–adult child sample and took the model-trimming approach to obtain the most parsimonious model. The final model also fits the data well, with the model fit statistics RMSEA, CFI, and TLI equal to 0.01, 0.96, and 0.95, respectively. We show only the coefficients pertinent to early-life father–child relationships; adult children’s marital status, income, and health conditions; and time and money transfers to fathers in Figure 3. All coefficients involved confounding variables are presented in Supplementary Table 2. Figure 3. View largeDownload slide Path model of how early-life father–child relationships are associated with upward transfers to unpartnered fathers through adult children's marital status, income, and health conditions. Figure 3. View largeDownload slide Path model of how early-life father–child relationships are associated with upward transfers to unpartnered fathers through adult children's marital status, income, and health conditions. As shown in Figure 3, both latent constructs, psychological closeness and physical violence, well represent their respective measures, as the factor loadings range from 0.85 to 0.95 for the former and from 0.52 to 0.94 for the latter. The path model also reveals that psychological closeness did not have a direct association with either time or money transfer. In addition, psychological closeness was unrelated to marital status, income, or health conditions, indicating that psychological closeness did not indirectly relate to time or money transfer through its associations with adult children’s resources and constraints. Like psychological closeness, physical violence was unrelated to marital status, income, or health conditions, so physical violence did not have an indirect association with time or money transfer. Nonetheless, adult children who more often experienced physical violence by fathers before age 17 spent fewer hours helping the fathers in later life (−2.24, p < .001). Physical violence, however, did not have a direct association with money transfer. Only two covariates were associated with adult children’s time transfer to fathers, as shown in Supplementary Table 2. Adult children spent more hours helping their fathers when the fathers had poorer health or when adult children lived closer to the fathers. None of the covariates was significantly related to adult children’s money transfer to fathers. Like the findings for mothers, more fathers’ characteristics (e.g., financial struggles and the use of alcohol and drugs) were associated with early-life father–child relationships than were adult children’s characteristics, whereas more adult children’s characteristics (e.g., race, employment status, and parenthood) were related to their marital status, income, and health conditions than were fathers’ characteristics. Discussion How to care for unpartnered older adults is a rising social concern, as the number of unpartnered older adults, who face greater economic and health vulnerabilities than the partnered (Hughes & Waite, 2009; Lin & Brown, 2012), continues to grow. Caregiving literature has extensively examined how adult children’s current resources and constraints influence their abilities to help parents (Couch et al., 1999; Laditka & Laditka, 2001; Sarkisian & Gerstel, 2008), but rarely considered relationships that adult children formed with their parents during childhood. Drawing on the life course perspective (Elder et al., 2003), we bring the child development literature to caregiving research to provide new insights into adult children’s caregiving behavior. We found that mothers’ use of physical violence on their children was inversely associated with adult children’s marital status, which, in turn, was negatively related to time transfer to the mothers. This finding suggests that adult children’s marital status, which scholars have long viewed as a key determinant of caregiving is actually a mediator of the association of early-life mother–child relationships and adult children’s support of unpartnered mothers in later life, an insight that would have been overlooked if early-life mother–child relationships had not been considered in the study. After taking into account various possible mediating and confounding variables, we found that early-life parent–child relationships still have unique and independent associations with adult children’s caregiving behavior. Specifically, adult children who were psychologically closer to their mothers while growing up spent more hours helping them in later life than did adult children who were less attached. In addition, adult children who more often experienced physical violence by their fathers while growing up spent fewer hours helping them in later life than did those who experienced physical violence less often. These findings indicate that early-life parent–child relationships have an influence on adult children’s time transfer to their parents in later life that caregiving researchers should not ignore. We also found that psychological closeness and physical violence have different patterns of associations with adult children’s time transfer to unpartnered mothers and fathers, respectively. Such differences are likely attributable to the different roles that mothers and fathers play in childrearing. Mothers are usually the primary caregivers—spending more time caring for children than do fathers (Parker & Wang, 2013). The patterns of parent–child interactions in later life reflect the psychological bonds formed early in the life course (Rossi & Rossi, 1990), which may explain the direct association of psychological closeness with time transfer to mothers, but not fathers. Although mothers also discipline children more often than fathers, the negative consequences may be buffered by maternal warmth (McKee et al., 2007). This may explain why physical violence was not found to be directly associated with time transfer to mothers. Nevertheless, children could still imitate aggressive behavior from mothers who exercised physical violence and use it in their own intimate relationships (Colman & Widom, 2004; Mullen et al., 1996). Along this line, we observed an indirect association between mothers’ use of physical violence and time transfer through adult children’s marital status. Fathers less often discipline children (Straus & Stewart, 1999), but when they do, they tend to discipline children more harshly than mothers (Bender et al., 2007; Nobes et al., 1999). The use of harsh discipline during childhood may have a scarring effect on subsequent father–child interactions, which may explain why we observed a direct association between physical violence and time transfer to fathers. There are several null findings. Early-life parent–child relationships have neither direct nor indirect associations with adult children’s provision of financial assistance to their mothers or fathers. This likely occurs because both mothers and fathers in the sample were relatively well-off, with a median family income of $65,000 (shown in Table 1). When parents have less financial need, there may not be enough variance in the sample to identify possible covariates. We also did not find any pathways from early-life parent–child relationships to upward transfers through adult children’s income and health conditions. The lack of mediating effects through income is likely because we controlled for family financial struggles while the respondents were growing up; such financial struggles were strongly associated with early-life parent–child relationships and time transfer, at least for mother–adult child dyads (shown in Supplementary Table 1). In addition, the lack of mediating effects through adult children’s health may be attributable to the fact that most adult children in the sample were relatively young and had few health conditions. This study has some limitations. First, like other studies examining early-life parent–child relationships (Kong & Moorman, 2016; Whitbeck et al., 1991; Whitbeck et al., 1994), we also relied on respondents’ retrospective reports, which could have been influenced by their current support of parents. Yet, prior studies have established good reliability and validity for respondents’ recall of childhood family circumstances (Hardt & Rutter, 2004). Second, this study focused on three types of resources and constraints that could mediate the associations between early-life parent–child relationships and upward transfers. There may be other factors that could mediate such associations, such as parents’ expectations for caregiving (Pillemer & Suitor, 2014). Future studies should continue to explore possible meditating factors. Third, we found fewer significant paths for the father sample than for the mother sample, likely due to the former’s relatively smaller sample size. Further studies using larger samples of both mothers and fathers are needed. Finally, we have controlled more confounding variables than previous studies on a similar topic did (Kong & Moorman, 2016; Whitbeck et al., 1991; Whitbeck et al., 1994). Nevertheless, other confounding variables that were not available in current study may be worth considering in the future, including parental divorce during childhood and the geographic location of siblings. Despite these limitations, this study makes important theoretical and practical contributions to the existing caregiving literature. Regarding theoretical contributions, this study shows that early-life parent–child relationships play a pivotal role in influencing adult children’s caregiving behavior, both directly and indirectly. Moreover, the pathways through which early-life parent–child relationships are associated with adult children’s caregiving behavior differ for unpartnered mothers and fathers, underscoring the importance of developing respective theoretical frameworks for mothers and fathers. As for practical contributions, practitioners generally suggest that providing caregivers with respite services and psychosocial support can enhance caregiving to older adults, because they reduce caregiver burden, help caregivers cope with stress, and maintain caregivers’ well-being (Lopez-Hartmann, Wens, Verhoeven, & Remmen, 2012). Interestingly, practitioners seldom discuss what older adults can do to increase the likelihood of receiving support from their adult children. This study demonstrates that parents may be able to increase support from their adult children by fostering psychological closeness and avoiding the use of physical violence when their children are growing up. Some parents may have difficulty creating and maintaining warm and violence-free relationships with their children, and social programs should be designed for and provided to these parents. As the number of unpartnered older adults continues to grow, many will need help from their adult children. It is important for researchers and practitioners to recognize that adult children’s decisions on whether to help their parents rest not only on their current resources and constraints, but also on the relationships that adult children had with their parents while growing up. Supplementary Material Supplementary data is available at The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences online. Funding This research was supported in part by the Center for Family and Demographic Research, Bowling Green State University, which has core funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (P2CHD050959). Conflict of Interest I.-F. Lin serves on Journal Editorial Board. Acknowledgments The authors thank the editor and two anonymous reviewers for their helpful comments. Any opinions expressed here are solely those of the authors and not of the funding agency or center. Author Contributions: I.-F. Lin and H.-S. Wu conceived the study, planned the analyses, and wrote the article. I.-F. Lin prepared the data file and performed the analyses. References Acock, A. C. ( 2005). Working with missing values. Journal of Marriage and Family , 67, 1012– 1028. doi: 10.1111/j.1741- 3737.2005.00191.x Google Scholar CrossRef Search ADS   Belt, W., & Abidin, R. R. ( 1996). The relation of childhood abuse and early parenting experiences to current marital quality in a nonclinical sample. Child Abuse & Neglect , 20, 1019– 1030. doi: 10.1016/0145-2134(96)00092-0 Google Scholar CrossRef Search ADS PubMed  Bender, H. L., Allen, J. P., McElhaney, K. B., Antonishak, J., Moore, C. M., Kelly, H. O., & Davis, S. M. ( 2007). Use of harsh physical discipline and developmental outcomes in adolescence. Development and Psychopathology , 19, 227– 242. doi: 10.1017/S0954579407070125 Google Scholar CrossRef Search ADS PubMed  Carpenter, B. D. ( 2001). Attachment bonds between adult daughters and their older mothers: associations with contemporary caregiving. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 56, P257– P266. doi: 10.1093/geronb/56.5.P257 Google Scholar CrossRef Search ADS   Cicirelli, V. G. ( 1993). Attachment and obligation as daughters’ motives for caregiving behavior and subsequent effect on subjective burden. Psychology and Aging , 8, 144– 155. doi: 10.1037/0882-7974.8.2.144 Google Scholar CrossRef Search ADS PubMed  Colman, R. A., & Widom, C. S. ( 2004). Childhood abuse and neglect and adult intimate relationships: A prospective study. Child Abuse & Neglect , 28, 1133– 1151. doi: 10.1016/j.chiabu.2004.02.005 Google Scholar CrossRef Search ADS PubMed  Couch, K. A., Daly, M. C., & Wolf, D. A. ( 1999). Time? Money? Both? The allocation of resources to older parents. Demography , 36, 219– 232. doi: 10.2307/2648110 Google Scholar CrossRef Search ADS PubMed  Covey, H. C., Menard, S., & Franzese, R. J. ( 2013). Effects of adolescent physical abuse, exposure to neighborhood violence, and witnessing parental violence on adult socioeconomic status. Child Maltreatment , 18, 85– 97. doi: 10.1177/1077559513477914 Google Scholar CrossRef Search ADS PubMed  Currie, J., & Widom, C. S. ( 2010). Long-term consequences of child abuse and neglect on adult economic well-being. Child Maltreatment , 15, 111– 120. doi: 10.1177/10775 59509355316 Google Scholar CrossRef Search ADS PubMed  Dupre, M. E., Beck, A. N., & Meadows, S. O. ( 2009). Marital trajectories and mortality among US adults. American Journal of Epidemiology , 170, 546– 555. doi: 10.1093/aje/kwp194 Google Scholar CrossRef Search ADS PubMed  Elder, G. H., Johnson, M. K., & Crosnoe, R. ( 2003). The emergence and development of life course theory. In J. T. Mortimer & M. J. Shanahan (Eds.), Handbook of the life course  (pp. 3– 19). New York, NY: Kluwer Academic. doi: 10.1007/978-0-306-48247-2_1 Google Scholar CrossRef Search ADS   Fingerman, K. L., VanderDrift, L. E., Dotterer, A. M., Birditt, K. S., & Zarit, S. H. ( 2011). Support to aging parents and grown children in black and white families. The Gerontologist , 51, 441– 452. doi: 10.1093/geront/gnq114 Google Scholar CrossRef Search ADS PubMed  Fredman, L., Lyons, J. G., Cauley, J. A., Hochberg, M., & Applebaum, K. M. ( 2015). The relationship between caregiving and mortality after accounting for time-varying caregiver status and addressing the healthy caregiver hypothesis. The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences , 70, 1163– 1168. doi: 10.1093/gerona/glv009 Google Scholar CrossRef Search ADS   Hardt, J., & Rutter, M. ( 2004). Validity of adult retrospective reports of adverse childhood experiences: Review of the evidence. Journal of Child Psychology and Psychiatry, and Allied Disciplines , 45, 260– 273. doi: 10.1111/j.1469-7610.2004.00218.x Google Scholar CrossRef Search ADS PubMed  Henretta, J. C., Soldo, B. J., & Van Voorhis, M. F. ( 2011). Why do families differ? Children’s care for an unmarried mother. Journal of Marriage and the Family , 73, 383– 395. doi: 10.1111/j.1741-3737.2010.00813.x Google Scholar CrossRef Search ADS PubMed  Hu, L.-T., & Bentler, P. M. ( 1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling , 6, 1– 55. doi: 10.1080/10705519909540118 Google Scholar CrossRef Search ADS   Hughes, M. E., & Waite, L. J. ( 2009). Marital biography and health at mid-life. Journal of Health and Social Behavior , 50, 344– 358. doi: 10.1177/002214650905000307 Google Scholar CrossRef Search ADS PubMed  Kline, R. B. ( 2011). Principles and practice of structural equation modeling  ( 3rd ed.). New York, NY: The Guilford Press. Kong, J., & Moorman, S. M. ( 2016). History of childhood abuse and intergenerational support to mothers in adulthood. Journal of Marriage and Family , 78, 926– 938. doi: 10.1111/jomf.12285 Google Scholar CrossRef Search ADS   Laditka, J. N., & Laditka, S. B. ( 2001). Adult children helping older parents: Variations in likelihood and hours by gender, race, and family role. Research on Aging , 23, 429– 456. doi: 10.1177/0164027501234003 Google Scholar CrossRef Search ADS   Lee, C., Tsenkova, V., & Carr, D. ( 2014). Childhood trauma and metabolic syndrome in men and women. Social Science & Medicine (1982) , 105, 122– 130. doi: 10.1016/j.socscimed.2014.01.017 Google Scholar CrossRef Search ADS PubMed  Lin, I.-F., & Brown, S. L. ( 2012). Unmarried Boomers confront old age: A national portrait. The Gerontologist , 52, 153– 165. doi: 10.1093/geront/gnr141 Google Scholar CrossRef Search ADS PubMed  Lin, I.-F., Brown, S. L., & Hammersmith, A. M. ( 2017). Marital biography, social security receipt, and poverty. Research on Aging , 39, 86– 110. doi: 10.1177/0164027516656139 Google Scholar CrossRef Search ADS PubMed  Lopez-Hartmann, M., Wens, J., Verhoeven, V., & Remmen, R. ( 2012). The effect of caregiver support interventions for informal caregivers of community-dwelling frail elderly: A systematic review. International Journal of Integrated Care , 12, e133. doi: 10.5334/ijic.845 Google Scholar PubMed  McGarry, K., & Schoeni, R. F. ( 1995). Transfer behavior in the health and retirement study: Measurement and the redistribution of resources within the family. The Journal of Human Resources , 30, S184– S226. doi: 10.2307/146283 Google Scholar CrossRef Search ADS   McGonagle, K., & Freedman, V. A. ( 2015). The panel study of income dynamics’ childhood retrospective circumstances study (PSID-CRCS) user guide: Final release 1 . Institute for Social Research, University of Michigan. Retrieved from https://psidonline.isr.umich.edu/CRCS/2014UserGuide.pdf. McKee, L., Roland, E., Coffelt, N., Olson, A. L., Forehand, R., Massari, C., Jones, D., Gaffney, C. A., & Zens, M. S. ( 2007). Harsh discipline and child problem behaviors: The roles of positive parenting and gender. Journal of Family Violence , 22, 187– 196. doi: 10.1007/s10896-007-9070-6 Google Scholar CrossRef Search ADS   Mullen, P. E., Martin, J. L., Anderson, J. C., Romans, S. E., & Herbison, G. P. ( 1996). The long-term impact of the physical, emotional, and sexual abuse of children: A community study. Child Abuse & Neglect , 20, 7– 21. doi: 10.1016/0145-2134(95)00112-3 Google Scholar CrossRef Search ADS PubMed  Muthén, L. K., & Muthén, B. O. ( 2015). Mplus user’s guide  (7th ed.). Los Angeles, CA: Muthén & Muthén. Nobes, G., Smith, M., Upton, P., & Heverin, A. ( 1999). Physical punishment by mothers and fathers in British homes. Journal of Interpersonal Violence , 14, 887– 902. doi: 10.1177/088626099014008006 Google Scholar CrossRef Search ADS   Parker, K., & Wang, W. ( 2013). Modern parenthood: Roles of moms and dads converge as they balance work and family . Washington DC: Pew Research Center. Retrieved from http://www.pewsocialtrends.org/2013/03/14/modern-parenthood-roles-of-moms-and-dads-converge-as-they-balance-work-and-family/. Pezzin, L. E., Pollak, R. A., & Schone, B. S. ( 2015). Bargaining power, parental caregiving, and intergenerational coresidence. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 70, 969– 980. doi: 10.1093/geronb/gbu079 Google Scholar CrossRef Search ADS   Pillemer, K., & Suitor, J. J. ( 2014). Who provides care? A prospective study of caregiving among adult siblings. The Gerontologist , 54, 589– 598. doi: 10.1093/geront/gnt066 Google Scholar CrossRef Search ADS PubMed  PSID Main Interview User Manual( 2015). Institute for Social Research , University of Michigan. Retrieved from http://psidonline.isr.umich.edu/data/Documentation/UserGuide2013.pdf. Rossi, A. S., & Rossi, P. H. ( 1990). Of human bonding: Parent-child relations across the life course . New York, NY: Aldine de Gruyter. Sarkisian, N., & Gerstel, N. ( 2008). Till marriage do us part: Adult children’s relationships with their parents. Journal of Marriage and Family , 70, 360– 376. doi: 10.1111/j.1741-3737.2008.00487.x Google Scholar CrossRef Search ADS   Schafer, M. H., Morton, P. M., & Ferraro, K. F. ( 2014). Child maltreatment and adult health in a national sample: Heterogeneous relational contexts, divergent effects? Child Abuse & Neglect , 38, 395– 406. doi: 10.1016/j.chiabu.2013.08.003 Google Scholar CrossRef Search ADS PubMed  Schoeni, R. F., Bianchi, S. M., Hotz, V. J., Seltzer, J. A., & Wiemers, E. E. ( 2015). Intergenerational transfers and rosters of the extended family: A new substudy of the panel study of income dynamics. Longitudinal and Life Course Studies , 6, 319– 330. doi: 10.14301/llcs.v6i3.332 Google Scholar CrossRef Search ADS PubMed  Shaw, B. A., Krause, N., Chatters, L. M., Connell, C. M., & Ingersoll-Dayton, B. ( 2004). Emotional support from parents early in life, aging, and health. Psychology and Aging , 19, 4– 12. doi: 10.1037/0882-7974.19.1.4 Google Scholar CrossRef Search ADS PubMed  Silverstein, M., Conroy, S. J., Wang, H., Giarrusso, R., & Bengtson, V. L. ( 2002). Reciprocity in parent-child relations over the adult life course. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 57, S3– S13. doi: 10.1093/geronb/57.1.S3 Google Scholar CrossRef Search ADS   Springer, K. W. ( 2009). Childhood physical abuse and midlife physical health: Testing a multi-pathway life course model. Social Science & Medicine , 69, 138– 146. doi: 10.1016/j.socscimed.2009.04.011 Google Scholar CrossRef Search ADS   Straus, M. A., Gelles, R. J., & Steinmetz, S. K. ( 2006). Behind closed doors: Violence in the American family . New Brunswick, NJ: Transaction Publishers. Straus, M. A., & Stewart, J. H. ( 1999). Corporal punishment by American parents: National data on prevalence, chronicity, severity, and duration, in relation to child and family characteristics. Clinical Child and Family Psychology Review , 2, 55– 70. doi: 10.1023/A:1021891529770 Google Scholar CrossRef Search ADS PubMed  Turley, R. N. L., Desmond, M., & Bruch, S. K. ( 2010). Unanticipated educational consequences of a positive parent-child relationship. Journal of Marriage and Family , 72, 1377– 1390. doi: 10.1111/ j.1741-3737.2010.00771.x Google Scholar CrossRef Search ADS   Whitbeck, L., Hoyt, D. R., & Huck, S. M. ( 1994). Early family relationships, intergenerational solidarity, and support provided to parents by their adult children. Journal of Gerontology , 49, S85– S94. doi: 10.1093/geronj/49.2.S85 Google Scholar CrossRef Search ADS PubMed  Whitbeck, L. B., Simons, R. L., & Conger, R. D. ( 1991). The effects of early family relationships on contemporary relationships and assistance patterns between adult children and their parents. Journal of Gerontology , 46, S330– S337. doi: 10.1093/geronj/46.6.s330 Google Scholar CrossRef Search ADS PubMed  Wolff, J. L., Mulcahy, J., Huang, J., Roth, D. L., Covinsky, K., & Kasper, J. D. ( 2017). Family caregivers of older adults, 1999–2015: Trends in characteristics, circumstances, and role-related appraisal. The Gerontologist . doi: 10.1093/geront/gnx093 Wong, R., Kitayama, K. E., & Soldo, B. J. ( 1999). Ethnic differences in time transfers from adult children to elderly parents: Unobserved heterogeneity across families? Research on Aging , 21, 144– 175. doi: 10.1177/0164027599212002 Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

The Journals of Gerontology Series B: Psychological Sciences and Social SciencesOxford University Press

Published: Feb 8, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off