How Widowhood Shapes Adult Children’s Responses to Mothers’ Preferences for Care

Abstract

Objectives.

We examine whether recently widowed mothers who needed assistance for a chronic condition, serious illness, or injury were more likely to receive care from the children they preferred as caregivers than were mothers who were divorced or had been widowed for a longer period of time.

Method.

Data were analyzed from 130 widowed or divorced mothers aged 72–83 in the second wave of the Within-Family Differences Study, all of whom reported needing assistance for a chronic condition or a serious illness or injury within 2 years prior to T2.

Results.

The findings provided evidence that recent widowhood shapes patterns of caregiving. Analyses revealed that mothers who had been widowed within 4 years were substantially more likely to receive care from adult children whom they had identified several years earlier as preferred caregivers than were mothers who were divorced or had been widowed 4 or more years.

Discussion.

Research has shown that mothers are at an increased risk for declining psychological well-being when caregiving preferences are not met. Findings from this study suggest that mothers who are divorced or have been widowed for several years may be at greater risk for violation of their caregiving preferences, increasing their vulnerability to declines in psychological well-being.

Widowhood has interested family and gerontological scholars for several decades, providing a substantial body of research on the consequences of this major life event (Lee & DeMaris, 2007; Lopata, 1973; Pai & Ha, 2012; Utz, Reidy, Carr, Nesse, & Wortman, 2004). However, most researchers have studied widowhood primarily as an independent event, with relatively little attention to the ways in which its occurrence shapes other major life events that may occur at the same point in the life course. In particular, widowhood often coincides with the age at which women become vulnerable to steep declines in health, frequently requiring support from family members. Further, there is evidence that widowhood itself leads to health declines (Williams, 2004) and that individuals may receive lower quality formal care for serious health events following widowhood (Iwashyna & Christakis, 2003). Thus, new widows may be particularly vulnerable to health declines and in need of family assistance.

Because adult children are the primary source of support to older unmarried women, exploring how widowhood shapes offspring’s responses to their mothers’ caregiving needs can shed important light on the consequences of this common and highly stressful transition. Understanding these processes is particularly important in light of new evidence demonstrating the salience of the specific offspring in the family who provide care. In particular, recent research has shown that mothers typically have clear preferences regarding which children they prefer as their caregivers and that the violation of these preferences leads to lower psychological well-being (Suitor, Gilligan, & Pillemer, 2013). Recent widowhood is a transition that may affect whether mothers receive care from their preferred children and is thus worthy of exploration.

In this article, we address this issue using longitudinal data collected from 130 older mothers across 7 years, beginning at the point when the women were 65–75 years of age. We examine whether recently widowed mothers who required care for a chronic condition or a major illness or injury were more likely to receive care from the children they preferred as caregivers than were mothers who were divorced or had been widowed for a longer period of time. We focus on widowed and divorced mothers because these are the women who are most likely to rely on their offspring for support in the face of a major health event, as opposed to married mothers, who typically turn to their spouses first as their caregivers (Carr & Moorman, 2011).

Does Recent Widowhood Shape Patterns of Care?

Theory and research on status transitions have emphasized the ways in which the loss or acquisition of salient roles affect interpersonal relations. In the case of widowhood, the effects are typically positive on mothers’ relationships with their adult children, resulting in increased closeness and contact, at least for the first few years following death of a spouse (Fuller-Thomson, 2000; Khodyakov & Carr, 2009). Such increased contact and closeness are especially important at this time because widowhood is often followed by declines in health and greater dependence on adult children for support.

Adequately meeting mothers’ needs for care, however, may require more than simply having adult children available to provide assistance. As noted, recent research has shown that mothers typically have clear preferences regarding which children they prefer as their caregivers and that the violation of these preferences leads to lower psychological well-being (Suitor et al., 2013). Suitor and colleagues’ analyses revealed that mothers’ distress resulted from the inability to receive care from children whom they felt possessed highly desirable socioemotional characteristics. The study showed that in selecting their preferred caregivers, mothers valued socioemotional characteristics of the children and the parent–child dyads that ensured predictability and harmonious relations—a pattern consistent with principles of socioemotional selectivity theory (Carstensen, 1992; Carstensen, Fung, & Charles, 2003). Specifically, mothers preferred as their caregivers children whom they perceived to be the most empathetic and similar to themselves from among their offspring, consistent with other research (Pillemer & Suitor, 2006; Suitor & Pillemer, 2006).

Further, these patterns of preference remained remarkably stable across the 7-year period of the study as mothers moved from being independent to requiring assistance (Suitor, Gilligan, & Pillemer, in press; Suitor et al., 2013). Specifically, more than three quarters of the adult children did not change their status as preferred or not preferred future caregivers across the 7 years between T1 and T2 (Suitor et al., in press). In fact, even in cases in which mothers received extensive support from children, they had not identified as their preferred future caregivers, the mothers continued to prefer those offspring they had named at the first interview—those whom they identified throughout the study as the most empathetic and similar to them (Suitor et al., 2013). Thus, patterns of caregiving preference are highly stable, and adherence to these preferences has been shown to play a central role in optimal caregiving outcomes, making it important to identify the conditions under which mothers’ preferences are met or violated.

We propose that recent widowhood is a context in which mothers’ caregiving preferences are more likely to be met, relative to when mothers are divorced or have been widowed longer. First, adult children are likely to be more aware of their mothers’ caregiving preferences when recently widowed because of the increased contact with both mothers and siblings that directly follows this life event (Fuller-Thomson, 2000; Khodyakov & Carr, 2009). Such heightened contact provides greater opportunity for discussion of mothers’ future care, increasing the likelihood that the children would develop a clear understanding of their mothers’ preferences.

Second, the recent loss of their father is expected to lead adult children to be more attentive to their mothers’ caregiving preferences. In this argument, we draw on Carstensen’s theory of socioemotional selectivity (Carstensen, 1992; Carstensen et al., 2003), which posits that when individuals experience events that lead them to perceive that their time to achieve goals is limited, they tend to enhance their focus on socioemotional goals and reduce their focus on more instrumental goals. The theory was originally developed to explain changes in individuals’ goals as they entered the later years; however, tests of the theory have shown that such a shift in the salience of socioemotional and instrumental foci is also likely to occur under certain circumstances in earlier stages of the life course. In particular, individuals tend to increasingly emphasize socioemotional dimensions of their lives after experiencing “priming events” that shift their perspective regarding the time still available to them, regardless of their place in the life course (Fung & Carstensen, 2006).

Although this argument is generally used when considering individuals’ perspectives regarding the finitude of their own lives (Carstensen, 1992; Carstensen et al., 2003), we suggest that the same principles can be applied to adult children who are observing their parents’ experience of aging. In particular, we posit that the loss of one parent would heighten their children’s recognition that their time with their surviving parent is finite and create a sense of urgency to be responsive to parents needs and desires, especially when they experience serious health declines.

Adult children’s elevated responsiveness to their mothers’ caregiving preferences in the immediate period following their fathers’ death could continue as mothers age; however, this is not likely to be the case. Studies have shown that the effects of widowhood on parent–adult child relations are most pronounced in the first few years following the event, after which patterns of closeness, support, and interaction tend to return to earlier levels (Ferraro, Mutran, & Barresi, 1984; Fuller-Thomson, 2000; Guiaux, Van Tilburg, & Van Groenou, 2007; Khodyakov & Carr, 2009). This pattern mirrors the findings of research on other major life events, such as divorce and parental loss, which has shown that the negative effects of these transitions tend to become muted within 2–4 years (Amato, 2000; Li, 2005; Umberson, 2006). Thus, we anticipated that the effects of widowhood on mothers’ likelihood of receiving care for serious health events from preferred adult children would occur only in the first few years following the bereavement.

Based on these arguments, we hypothesized that mothers who had been recently widowed would be more likely to receive care from their preferred adult child caregivers when facing a major health event for which they needed assistance than would mothers who were divorced or who had become been widowed less recently.

Control Variables in the Analysis

We also included in the analysis four sociodemographic characteristics that should be considered when studying the processes that lead children to become caregivers: mother’s age, race, number of living children, and educational attainment.

Just as fathers’ death may serve as a specific priming event to increase adult children’s awareness of their mothers’ vulnerability, mothers’ age is also likely to increase such awareness and thus needs to be taken into consideration in the analysis. Further, given the focus of the present analysis on the effects of widowhood, it is important to include mothers’ age to ensure that widowhood is not simply serving as a surrogate for this factor.

Race has been found to structure patterns of intergenerational care, with support provision being shared among a larger proportion of family members in black than white families (Sarkisian and Gerstel, 2004; Suitor, Sechrist, Gilligan, & Pillemer, 2011). Thus, there may be lower expectations for any particular child to provide support in black families even if mothers have preferences regarding their care, leading mothers to be less likely to receive care from preferred offspring.

Number of living children has been found to reduce the likelihood that any particular child will provide care (Tolkacheva, van Groenou, & van Tilburg, 2010), even when mothers have specific expectations regarding which children will assume that role (Suitor, Gilligan, & Pillemer, in press). Thus, mothers with a larger number of living offspring may be less likely to receive support from preferred children than are those with smaller families.

Finally, mothers’ educational attainment has been found to play a role in the patterns of caregiving, primarily because it serves as proxy for both financial and social resources (Suitor, Sechrist, Gilligan, & Pillemer, 2011; Uhlenberg & Cheuk, 2008). Thus, we believe it is important to include in the analysis.

Method

The data used in the present analyses were collected as part of the Within-Family Differences Study (WFDS). The design of the WFDS involved selecting a sample of mothers 65–75 years of age with at least two living adult children and collecting data from mothers regarding each of their children. (For a more detailed description of the WFDS design, see Suitor & Pillemer [2006] and Suitor et al. [in press].) The first wave of interviews took place with 566 women between 2001 and 2003; the original study was expanded to include a second wave of data collection from 2008 to 2011. At T2, a total of 420 mothers were interviewed regarding 1,514 of their children. Two hundred thirty four of the mothers reported having needed assistance for a chronic condition or a serious illness or injury at some point within 2 years prior to the T2 interview; 150 of whom were widowed or divorced at T2. Because the focus of the present analysis is whether mothers received care from children whom they preferred assume this role, the 20 mothers who specified that they did not want care from their children were omitted, resulting in a final analytic sample of 130 mothers.

Procedures

Massachusetts city and town lists were used as the source of the original WFDS sample. With the assistance of the Center for Survey Research (CSR) at the University of Massachusetts, Boston, Suitor and Pillemer drew a probability sample of women aged 65–75 with two or more children from the greater Boston area. The T1 sample consisted of 566 mothers, which represented 61% of those who were eligible for participation, a rate comparable to that of similar surveys in the past decade (Dixon &Tucker, 2010; Wright & Marsden, 2010).

For the follow-up study, the survey team attempted to contact each mother who participated in the original study. At T2, interviews were conducted with 420 of the mothers who had participated at T1. Of the 146 mothers who participated at only T1, 78 had died between waves, 19 were too ill to be interviewed, 33 refused, and 16 could not be reached. Thus, the 420 represent 86% of mothers who were living at T2. Comparisons between the mothers alive at T2 who did and not participate revealed that they differed on only education and subjective health; those who participated at both T1 and T2 were better educated and in better health.

Analytic Sample Characteristics

Table 1 presents the demographic characteristics of the analytic sample of 130 mothers. Although the mean number of living children in this subsample is higher than would be found in a nationally representative sample of women in this age group, this feature is due primarily to the criterion that all participants must have at least two living adult children. The mean number of children of women in the subsample is similar to that found in national samples, such as the National Survey of Families and Households (Sweet & Bumpass, 1996), when compared specifically to mothers in the same age group who have two or more children.

Table 1.

Demographic Information on Mothers (n = 130)

Age in years (; SD)	78.2 (3.2)
Race (in %)
White	61.5
Not white	38.5
Marital status (in %)
Divorced/Separated	23.1
Widowed 4+ years	60.0
Widowed within 4 years	16.9
Education (in %)
Less than high school	26.2
High school graduate	47.7
Some college	12.3
College graduate	13.9
Number of children (; SD)	3.8 (1.6)

Measures

Dependent variable: congruence between mothers’ caregiver preferences and outcomes.

Creating a measure of congruence between mothers’ preferences and outcomes regarding caregivers required comparing which children they preferred at T1 as their future caregivers with which children, if any, actually provided them care when they were seriously ill or injured or needed assistance with Activities of Daily Living (ADL) in the 2-year period prior to the T2 interview.

We began by creating measures of mothers’ caregiving preferences at T1 and actual provision of assistance at T2. To assess mothers’ future caregiver preferences at T1, each mother was asked to select which child she would prefer help from if she (the mother) became ill or disabled. Mothers were asked about their care preferences from adult children only and not about their preferences for care from other formal or informal sources. In the current analytic sample, 88% of the mothers named a particular child at T1 from whom they would prefer help if they became ill or disabled. Based on the mothers’ responses, each child in these families was coded as having been chosen or not having been chosen as their mother’s preferred future caregiver (1 = preferred; 0 = not preferred). In the case of the 12% of mothers who reported that they would be comfortable with any of their children as caregivers and could not choose any child over the other(s), each child was coded as preferred (1). At T2, each mother was asked whether any of their children had provided help for their recent health problem(s), and if so, which child(ren) had provided help; each child was then classified as having or not having provided assistance for those problems.

The congruence measure was created by comparing the mothers’ stated caregiver preferences at T1 to which children actually provided care for a health problem within the 2 years prior to T2. First, we developed a four-category variable measuring congruence between mothers’ caregiver preferences and actual caregiving provided by children for the recent health event. The categories were as follows: (a) received no care from children (n = 29; 22.3%); (b) received care exclusively from children preferred at T1 (n = 30; 23.1); (c) received care from preferred children in combination with not preferred children (n = 48; 36.9%); and (d) received care exclusively from children not preferred at T1 (n = 23; 17.7%). Because previous research has shown that mothers’ psychological well-being is affected by receiving care exclusively from not preferred children but is not affected by receiving care from a combination of preferred and not preferred children (Suitor et al., 2013), we combined “care from preferred children only” and “care from a combination of preferred and not preferred children.”

We use all three categories of preference/outcomes congruence for the descriptive analysis. Ideally, we would have preferred to conduct the multivariate analysis using the three categories separately. However, the relatively small number of cases in which mothers received care only from not preferred children resulted in too few cases in several cells to use multinomial logit modeling. Thus, we combined the categories “no care from children” and “care from only not preferred children” in a single “not preferred caregiver” category and conducted the analysis using binary logit modeling. Although this approach loses some information and therefore is less than ideal, we can nevertheless address our central question of whether mothers’ recent widowhood affected whether they received care from preferred children.

Independent variable: mothers’ marital status at T2.

To create the measure of mothers’ marital status at T2, we combined information provided by the mothers at T1 and T2 regarding their marital status at the time of the interviews. Because the changes in interpersonal relationships that often follow widowhood typically occur within the first 4 years after the death of the spouse (Ferraro et al., 1984; Guiaux et al., 2007), we used this as the cutoff point for “recent widowhood.” As shown in Table 1, 60% (n = 78) had been widowed more than 4 years prior to T2, 16.9% (n = 22) had become widowed within 4 years prior to the T2 interview, and 23.1% (n = 30) were divorced prior to T1 and did not marry between waves. Those who had been widowed within 4 years prior to T2 had been widowed for approximately 2 years ( = 2.1; SD = 1.1); those who had been widowed longer had lost their spouse from 5 to 49 years prior to T2 ( = 19; SD = 10.7). A set of dummy variables was created to reflect these three marital status categories; widowed 4 or more years at T2 was used as the reference category.

Control variables.

Race was measured by asking the mothers to select from a card listing several races and ethnicities (e.g., white, black, or African American, Hispanic or Latina, Native American, Asian). They were instructed that they could choose more than one race or ethnicity. The analytic sample for this article included 80 mothers who identified themselves as white, 47 who identified as black, 1 as Hispanic, and 2 as Native American. Based on the literature on later life families (Suitor et al., 2011), which has shown greater filial responsibility in black, Asian, and Hispanic than white families, we coded race as white = 1 and not white = 0.

Age at T2 was the age mothers provided at T1 plus 7 (the number of years between interviews). We measured family size by asking mothers for the names of each of their children at T1 and asking them to confirm this at T2. Mothers’ educational attainment was assessed by asking the highest level of education completed.

Entry into the analytic sample could occur in two ways: (a) If the mother had a chronic condition for which she needed assistance with ADLS or (b) if she had a serious acute illness or injury that led to the need for care. To take this factor into consideration we created a measure of source of care need. Each of the mothers was asked an extensive set of questions about the existence of chronic conditions that limited her activities, her need for assistance with activities of ADLs, and whether she had experienced any serious illnesses or injuries specifically for which she needed assistance within the previous 5 years. Mothers who reported serious illnesses or injuries in the preceding 5 years were asked in which year they had experienced the health event, which allowed us to determine whether the event had occurred within 2 years of the T2 interview. Approximately 35% (n = 46) of the mothers reported that they had difficulty performing one or more ADLs; 37.7% (n = 49) reported that they had experienced a serious illness or injury for which they needed assistance within the previous 2 years; and 26.9% (n = 35) reported that they had experienced both a serious illness or injury within 2 years and had difficulty with ADLs. Because our intent was to differentiate between those with serious acute care needs and those with chronic care needs, we combined “serious illness/injury only” with “both serious illness/injury and ADL limitations” to create a dichotomous measure (1 = experienced a chronic illness or injury, with or without ADL needs; 0 = ADL needs only).

Analytic plan.

We conducted the multivariate analyses using binary logistic regression. There were no missing cases on any of the variables. The analyses were conducted using SPSS 21.

Results

Descriptive Analysis

Table 2 presents the distribution of care from adult children by mothers’ marital status. Consistent with our hypothesis, recently widowed mothers were substantially more likely to receive assistance from the children they identified as their preferred caregivers than were mothers in other marital status categories. In fact, they were almost twice as likely to receive care from preferred children than were divorced mothers (p < .05), and about 50% more likely than mothers who had been widowed more than 4 years (p < .05). The analysis revealed relatively similar patterns for mothers who were divorced and those who had been widowed more than 4 years.

Table 2.

Children Favored at T1 Versus T2 in Percentages (n = 130)

Care provided by adult children	Divorced T1 and T2	Widowed more than 4 years	Widowed 4 or fewer years
No care from children	30.0 (9)	24.4 (19)	4.5 (1)
Care from not preferred only	23.3 (7)	17.9 (14)	9.1 (2)
Care from preferred children only or in combination	46.7 (14)	57.7 (45)	86.4* (19)

Note. *Differences between column 3 and both columns 1 and 2 are statistically significant at p < .05.

Multivariate Analysis

Next, we examined the association between mothers’ marital status and caregiving patterns while controlling on mothers’ demographic characteristics, family size, and source of care needs. In this analysis, the outcome variable is a binary measure of whether the mothers (a) were provided care by preferred children (alone or in combination with other children) or (b) received no care from children or received care only from children mothers did not prefer as their caregivers.

As shown in Table 3, the overall pattern of findings is similar to the descriptive analysis. For mothers who had been widowed recently, the odds of receiving assistance from their preferred caregivers were more than four and a half times greater than the odds for mothers who had been widowed longer (odds ratio = 4.62). In contrast, there was little difference in the odds of receiving care from preferred children for mothers who were divorced compared with those widowed 4 or more years.

Table 3.

Binomial Logistic Regression Predicting Whether Mothers’ Caregiving Preferences Were Met (n = 130)

	Odds ratio	95% confidence interval
Age at T2	1.01	0.90–1.14
Education	0.84	0.65–1.09
Race (white)	1.97	0.83–4.69
Mother divorced prior to T1	0.80	0.33–1.98
Mother widowed within 4 years prior to T2	4.62*	1.19–17.88
Source of care need (acute)	1.08	0.47–2.50
Number of living children at T2	0.92	0.72–1.17
−2 Log likelihood	161.11
χ2	13.87*

Note. *p < .05.

None of the other factors that we anticipated might shape whether mothers received care from preferred offspring were significant predictors, further highlighting the salient role that recent widowhood plays in congruence between mothers’ preferences and the care that she receives when in need of assistance.

Discussion

Recent research has demonstrated the importance of mothers’ expectations regarding care when facing serious health events. When mothers’ preferences for which children will care for them are violated, mothers are at an increased risk for lower psychological well-being (Suitor et al., 2013). It is, therefore, important to identify the conditions under which older mothers are the most likely to receive care from children they would prefer assume this role. Our central aim in this article was to address this issue by investigating whether recent widowhood played a role in these processes. Specifically, we asked whether mothers were more likely to receive care from their preferred adult children when they had recently experienced the loss of their spouse.

Drawing from theoretical and empirical literature on both widowhood and socioemotional selectivity, we argued that recently widowed mothers would be more likely to receive support from preferred caregivers than would mothers who were divorced or had been widowed for a longer period. Our basis for this position was twofold. First, we suggested that because adult children have more contact with their mothers and with one another following the loss of the father (Fuller-Thomson, 2000; Khodyakov & Carr, 2009), they would have more accurate knowledge of their mothers’ preferences. Second, socioemotional selectivity theory (Carstensen et al., 2003; Fung & Carstensen, 2006) suggests that when adult children lose one of their parents, they are more likely to acknowledge that their time with their surviving parent is finite, leading them to be highly responsive to that parent’s needs and desires, particularly when the parent is facing serious health declines. Thus, adult children would be more likely to know their mothers’ preferences regarding care and attempt to adhere to her wishes.

The findings provided strong evidence that recent widowhood shapes patterns of caregiving in ways consistent with our hypotheses. Both bivariate and multivariate analyses revealed that mothers who had been widowed within 4 years prior to the interview were substantially more likely than other groups to receive care from the adult children whom they had identified as preferred caregivers several years earlier. As shown by the multivariate analysis, the odds of recent widows receiving care from preferred children were approximately four and a half times greater than the odds of their counterparts who had been widowed 4 or more years earlier. In contrast, the findings revealed little difference in the odds of divorced and longer widowed mothers receiving care from preferred children, suggesting that the experiences of these two groups during caregiving may be more similar than might have been expected.

These analyses provide further evidence that, consistent with the literature (Fuller-Thomson, 2000; Khodyakov & Carr, 2009), the effects of widowhood on parent–adult child relations are most pronounced in the first few years following the event, after which patterns of closeness, support, and interaction tend to return to earlier levels. This finding suggests that mothers who experience a major health event when they are divorced or have been widowed for several years face greater obstacles to receiving care from preferred adult children than do recently widowed mothers. Given that the “window” of recent widowhood is quite narrow, most women may be beyond this window when they experience major health events, placing them at greater risk for poor caregiving outcomes.

Although the present analysis has shed light on the ways in which recent widowhood shapes whether mothers receive care from children whom they had identified as their preferred future caregivers, several closely related questions regarding the intersection of widowhood and caregiving should be pursued. First, an intriguing question that we could not address was the extent to which differences in caregiving patterns by marital status could be attributed to changes in the accuracy of adult children’s perceptions of their mothers’ preferences, to children’s greater responsiveness to these preferences, or to both. Mothers’ preferences are highly stable across time (Suitor et al., in press) and are not affected by becoming widowed; thus, they can be measured several years prior to the transition. However, to explore how the transition affects children’s perceptions and responsiveness requires collecting data shortly before and after the mothers become widowed. We recommend that future studies use designs that will provide this opportunity.

A second question is whether the patterns we uncovered might differ by parents’ gender. The literature has demonstrated that fathers are as likely as mothers to have preferences regarding care from particular children (Suitor & Pillemer, 2013); however, their potential sources of support following the loss of the spouse are more limited than are mothers’ sources (Lee, DeMaris, Bavin, & Sullivan, 2001). Thus, receiving care from preferred children may be even more important for fathers than mothers.

A third direction we suggest for future research is whether similar patterns would be found following recent divorce. Given that the average age of the present sample is nearly 80, it is not surprising that none of the mothers became divorced during the course of the study. Thus, we could not address this question. However, it is possible that the patterns we have found might be similar for midlife recently divorced mothers compared with those who had been divorced for longer periods. Studying whether recent divorce shapes patterns of care from adult children may shed important light on support processes when women face both serious illness and marital loss in midlife.

An additional question deserving further study is whether widowhood has a priming effect for adult children in cases in which the deceased spouse was a stepparent who did not raise the children. Based on socioemotional selectivity theory, it may well be that mothers’ widowhood, regardless of whether the deceased spouse was the children’s father, could serve as a priming event for offspring. In this study, nearly 90% of the deceased husbands were the fathers of all of the children in the family; thus, in all but a few cases, the adult children experienced the loss of their fathers when their mothers became widowed. Further, we focused on the loss of fathers who were married to the children’s mothers at the time of their death, yet the death of a father who was divorced from the mother could also be a priming event for adult children. We cannot address these questions, but we believe that they are of both theoretical and practical importance to pursue in future research.

In sum, the findings demonstrated clear differences in patterns of caregiving by marital status, suggesting varying implications for women in these subgroups. When recent widows face a second negative life event (serious illness or disability) shortly after the loss of their husbands, they are likely to be cared for by the adult children whom they preferred for this role. In contrast, mothers who are divorced or have been widowed for a longer period are much less likely to have their caregiving preferences met. Thus, greater attention should be directed toward soliciting information regarding caregiving preferences from these women because they may face health events at points when their children are less likely to be aware of their needs and thus less responsive to them.

Finally, one of the most important implications of the findings is that widowhood, like most status transitions, is dynamic rather than static in its consequences. It is, therefore, necessary to take into consideration not only whether individuals are widowed but also where they stand in the widowhood “career.” This point has been well recognized in research on marriage, retirement, and parenthood (Kim & Moen, 2002) but is often neglected when studying widowhood. Our finding that the length of time from widowhood affects the fulfillment of caregiving preferences points to temporal differences in the process that are worthy of future exploration.

Funding

This project was supported by a grant from the National Institute on Aging (2RO1 AG18869) to J. J. Suitor and K. Pillemer (Co-Principal investigators). K. Pillemer also acknowledges support from an Edward R. Roybal Center grant from the National Institute on Aging (1 P50 AG11711-01).