Time Use and Well-being in Older Widows: Adaptation and Resilience

Abstract

Research on widowhood mostly emphasizes negative outcomes, but some studies suggest widows are resilient. This study evaluated the in-depth daily experiences of older widows compared to married women. Participants included 75 widows and 125 married women from the National Study of Daily Experiences II who completed daily diaries over eight days. For long-term widows, there were no differences compared to married women in daily and general well-being, and a few differences in time use. Results add to the literature by using daily diary data to illustrate the general resilience of older women many years after the death of a spouse.

Widowhood is common among older women; 52% of women 75–84 years of age and 76.2% over age 85 are widows (Federal Interagency Forum on Aging-Related Statistics, 2008). Additionally, over 50% of women over age 70 have ever been widowed (Krieder, 2005). There is a common gender-related stereotype of elderly widows as being frail, vulnerable, and requiring extensive assistance from others (Utz, Reidy, Carr, Nesse, & Wortman, 2004). This “frail-old-woman” stereotype may actually lead others to consider older widows as more dependent and vulnerable than they really are (Utz et al., 2004).

Widowhood does provide a number of challenges, both emotionally and in terms of role changes. The Dual Process Model suggests bereavement requires both loss- and restoration-oriented coping for successful adjustment (Stroebe & Schut, 1999; Utz, 2006). The emotional changes women experience during bereavement require loss-oriented coping, while the inevitable changes in daily life necessitate restoration-oriented coping.

Most research on widows focuses on declines in well-being (loss-oriented coping) after the death of a spouse. These studies associate widowhood with lower morale and social engagement (Bennett, 2005), increased anxiety or anger if the spouse was in physical pain prior to death (Carr, 2003), increased depressive symptoms (Fry, 2001; Lee, deMaris, Bavin, & Sullivan, 2001), and clinically significant symptoms a year later for an estimated 15% of recently bereaved widows (Bonanno & Kaltman, 2001). Contrary to this emphasis on problems during widowhood, Bonanno and colleagues’ evaluation of widowhood trajectories demonstrates that the highest percentage (45.9%) of widows in their study exhibited resilient characteristics with low depressive symptoms during the 18 months after the death of their spouse (Bonanno et al., 2002; Bonanno, 2009). Some widows may actually show improvements in well-being after the death of a spouse, which may be associated with declines in strains such as caregiving (Bonanno et al., 2002; Schulz et al., 2003). Based on this research, it is not well understood to what extent psychological resilience might characterize women beyond the first couple years of widowhood. Additionally, widows who were, on average, 12.6 years post loss reported that although they thought about their spouse once every week or so, they only “sometimes” or “rarely” were upset by these thoughts (Carnelley, Wortman, Bolger, & Burke, 2006). Qualitative research comparing the short and long-term effects of widowhood has begun to indicate that the initial short-term negative effects may be accompanied by longer-term positive effects, or resiliency (Feldman, Byles, & Beaumont, 2000). Recent widows (less than one year) are more likely to have worse social functioning and depressed mood compared to women who have been widowed for more than a year (Wilcox et al., 2003). Preliminary research evaluating loss-oriented coping suggests widowhood may be more benign over the long term than is typically found in studies assessing distress very recently after the death of a spouse, but research is limited beyond the first few years post loss.

Research on ways that older women who are widows develop new activities and roles after loss (restoration-oriented coping) is less developed (Carr & Utz, 2001; Utz, 2006; Utz, Reidy et al., 2004). However, psychological interventions for depression and grief often include an emphasis on altering daily activities (Qualls & Knight, 2008; Worden, 2008) and some researchers suggest that the daily life changes may also be an important area for research in widowhood (Feldman et al., 2000). Past studies suggest that widows may decrease time doing chores after their spouse dies (Utz et al., 2004). Shortly after a spouse’s death, a bereaved partner may spend more time receiving support and less time giving support to their children (Ha, Carr, Utz, & Nesse, 2006). Widows may require more assistance in daily activities, whereas married women may have more resources available to provide support to their children. Utz and colleagues found evidence suggesting that widows and non-widows had similar levels of “formal social participation,” but that widows spent more time in “informal social participation” than did married women (Utz, Carr, Nesse & Wortman, 2002). The authors included volunteering, organizational meetings, and religious services together as “formal social participation” and getting together or talking on the phone with friends/relatives as “informal social participation” (Utz et al., 2002). Informal social participation could also encompass other activities, such as exchanging social support or spending time with children. Other individuals in addition to children, such as friends, neighbors, or religious groups, may also provide assistance to widows, and research examining these relationships is limited. Past research provides an idea of changes in widows’ time use during adjustment to widowhood using retrospective measures that consider a range of activities describing formal and informal participation (Utz et al., 2002). Although research has begun to understand some of the general changes in time use (e.g., Utz et al., 2004; Utz et al., 2002; Ha, 2001; Ha et al., 2006), further research evaluating the specific day-to-day activities of non-recent widows may provide a better understanding of the long-term well-being of widows.

Daily diary methods evaluate individuals every day for a period of days, allowing researchers to look at well-being through what is described as the “microscope” of social science as an alternative to traditional retrospective measures that may risk recall bias (Almeida, 2005; Bolger, Davis, & Rafaeli, 2003). The current study hopes to better understand the specific daily activities in widowhood by comparing daily time use (Aim 1) and both daily and global well-being (Aim 2) in widowed and married women. In our comparison of widows and married womens’ daily time use, we hypothesized that widows would spend less time doing chores, more time with children, and more time receiving support than married women. Although research is lacking in the comparison of daily well-being for widows and married women, Bonanno’s (2009) research led us to hypothesize that this sample of non-recent widowed women would exhibit resilience with similar levels of daily and global well-being compared to married women.

Method

Participants

Data were from the second wave (2004–2006) of the National Study of Daily Experiences (NSDE II), a satellite study of the second wave of the Midlife in the United States Survey (MIDUS II; for descriptions of the NSDE and MIDUS project, see Almeida, Wethington, & Kessler, 2002: Mroczek & Kolarz, 1998). NSDE II participants were randomly selected from the larger MIDUS II study, and the response rate for the NSDE II sample was 86%. Respondents in this study included widows and married women, aged 60 and older, from the initial sample of 2,022 adults who had participated in NSDE II.

Demographic variables included in the analyses were age (measured continuously), education (percent with high school degree), number of children (measured continuously), race (White or non-White), and employment status (working or not working). Seventy-five widows and a demographically-similar comparison group of 125 married women were included in the analyses. Time since widowhood was evaluated among the widows, measured continuously as the number of years since the spouse’s death.

Sample Characteristics

The study evaluated demographic information for the two groups, including age, education, number of children, race, time since widowhood (for widows), and employment status (Table 1). Widows were slightly older than married women, and the other demographic variables were not significantly different between the two groups. Most of the married and widowed women were White and not working. Widows were not recently bereaved.

Table 1.

Demographic Information of Widowed (n = 75) and Married Women (n = 125)

	Widows	Married Women	t or χ2
	M (SD) or %	M (SD) or %
Age	72.37 (6.67)	70.03 (5.97)	−2.57*
Percent with high school degree	82.6	91.2	3.23
Number of children	3.48 (1.90)	3.39 (1.80)	−.33
Percent with no children	1.3	1.6	0.02
Race (percent White)	87.8	95.2	−1.72
Years widowed	11.12 (8.26)	n/a	n/a
Percent working	21.3	16.0	0.98
Full-time (of percent working)	31.8	6.1	2.33

Note.

^*p < .05.

Respondents participated in eight consecutive days of daily telephone interviews where they reported on their daily time use and daily well-being. In addition to these daily interviews, respondents also completed self-report questionnaires as part of the larger MIDUS II study. Demographic information and global well-being were assessed as part of MIDUS II. Global well-being was measured by asking respondents, “In general, would you say your mental/emotional health is excellent, very good, good, fair or poor?” Global well-being was measured with a 5-point Likert-type scale that was reverse-coded, ranging from 1 (poor) to 5 (excellent; Idler & Kasl, 1991).

Daily Measures

Daily time use

Each day, respondents were asked how much time they spent with children, doing housework, doing work or schoolwork, relaxing or doing leisure activities, watching television, volunteering, giving or receiving unpaid assistance, giving or receiving emotional support, providing assistance to someone with a disability, and sleeping. To gain a better understanding of the nature and meaning of the exchange of support in widows and married women, we also report the relationship of the person who is either providing support/assistance or receiving support/assistance with the older women in our sample. Daily time use was measured for eight consecutive days, and aggregate measures of each type of time use were used in the current analyses (McDonald & Almeida, 2004).

Daily measures of well-being

Daily positive and negative affect were measured using an adapted inventory of emotions from the Non-Specific Psychological Distress Scale developed from other validated scales for use in MIDUS I (Mroczek & Kolarz, 1998). In NSDE II, the psychological distress scale was supplemented with the Positive and Negative Affect Schedule (PANAS; Watson, Clark & Tellegen, 1988) and calibrated for daily affect. Daily negative affect items included how often during the past day they experienced fourteen different negative emotions (e.g. worthless, hopeless, and restless or fidgety), and daily positive affect items included how often during the past day they experienced thirteen different positive emotions (e.g., in good spirits, cheerful, and calm and peaceful). Respondents rated their responses on a 5-point scale ranging from 0 (none of the time) to 4 (all of the time), and greater scores reflected higher levels of positive and negative affect. Scores across the positive and negative affect items were summed and averaged across the interview week to produce two scores for analysis. Cronbach’s alpha was .91 for negative affect and .96 for positive affect.

Statistical Analysis

Widows who remarried were not included in the sample, and 125 married women were selected using a dataset with demographic characteristics (age, race, highest level of education achieved, number of children, employment status) to create a similar comparison group. For the two research aims, analysis of covariance (ANCOVA), controlling for age, was run by marital status on all daily time use variables in one model, and on daily and global well-being measures in a second model. Independent samples t-tests were run comparing the relationship of the person involved in the exchange or either emotional support or instrumental assistance between the two groups, married and widowed women.

Results

Before conducting analyses comparing widows and married women, initial tests were run to evaluate any differences between recent and non-recent widows within the widows sample. The small group of widows in the sample with less than two years since their spouse’s death (n = 18) were generally similar to the non-recent widows (n = 57). However, recent widows were slightly younger (M = 69.39, SD = 5.95 years vs. M = 73.30, SD = 6.70 years), t(72) = 2.21, p < .05. Controlling for the age difference between non-recent and recent widows, there were no differences in daily time use, daily affect, or global well-being, so results were pooled for all widows in subsequent analyses.

Daily Time Use for Married and Widowed Women

ANCOVA analyses revealed similarities in most of the daily time use measures (Table 2). Both widows and married women spent two hours doing chores and less than an hour doing work or school activities. On average, widows and married women spent over three hours participating in leisure activities and approximately 30 minutes engaging in vigorous activity. Widows and married women spent approximately one hour volunteering, almost two hours giving unpaid assistance, and 15–20 minutes receiving assistance. Widows and married women spent almost an hour giving emotional support, 30–40 minutes receiving emotional support, and from 45 minutes to an hour assisting someone with a disability. Consistent with our hypothesis, widows spent significantly more time with children than married women. Widows spent significantly less time sleeping and more time watching TV than married women.

Table 2.

Analysis of Covariance for Time Use Variables in Widows (n = 75) and Married Women (n = 125)^a

	Widows	Married Women	F
	M (SD) or %	M (SD) or %
Time Use Variable
With Children	0.93 (1.67)	0.55 (1.19)	3.97*
Doing Chores	1.68 (1.06)	1.99 (1.17)	2.71
Doing Work/School Activities	0.77 (1.57)	0.65 (1.43)	1.10
Leisure	3.21 (1.80)	3.08 (1.70)	0.54
Watching TV	2.51 (1.41)	2.05 (1.28)	6.20*
Engaging in Vigorous Activity	0.52 (.61)	0.60 (.67)	0.47
Volunteering	1.04 (1.65)	0.83 (1.35)	1.55
Giving Unpaid Assistance	1.66 (3.18)	2.03 (3.24)	0.24
Receiving Unpaid Assistance	0.37 (1.32)	0.17 (.85)	0.75
Giving Support	0.96 (1.42)	0.88 (1.07)	0.30
Receiving Support	0.56 (.79)	0.69 (2.22)	0.21
Assisting Disabled	1.09 (2.89)	0.80 (1.54)	1.35
Sleeping	6.91 (1.07)	7.22 (.94)	4.71*

Note.

^aAnalysis of covariance controlled for age. Daily time use measured in hours.

^*p < .05.

Daily Exchanges of Support for Married and Widowed Women

Exchanges of emotional support

Although total time spent giving and total time spent receiving support was similar for married women and widows, results of independent samples t-tests revealed some differences in the person involved in the exchange of emotional support between the two groups (see Table 3). Widows both gave and received significantly more daily emotional support from friends/neighbors than did married women. Widows also received significantly more daily emotional support from religious groups than did married women in our sample. While married women in our sample received the highest amount of daily emotional support from spouses, widows, in contrast, received equally high amounts of daily emotional support from both children and friends/neighbors. In sum, while married women receive most of their emotional support from their spouses, widows may have supplemented their need for support from other sources (children, friends, neighbors, and religious groups).

Table 3.

Independent Samples T-test of Support and Assistance Exchanged in Widows (n = 75) and Married Women (n = 125)^a

	Widows (%)	Married Women (%)
Giving emotional support
Spouse	0.00***	14.93
Child	19.45	24.70
Sibling/Other Relative	19.98	21.95
Friend/neighbor	54.68***	34.15
Religious Group	3.11	2.41
Other	12.46	10.32
Receiving emotional support
Spouse	1.04	38.76
Child	30.01	19.68
Sibling	18.39	12.11
Friend/neighbor	40.93*	25.20
Religious Group	12.75*	3.24
Other	6.55	12.95
Giving unpaid assistance
Spouse	0.00	0.00
Child	44.35	47.68
Sibling/Other Relative	25.70	21.69
Friend/neighbor	39.49*	22.16
Religious Group	5.69	2.36
Other	4.22*	11.79
Receiving unpaid assistance
Spouse	0.00	6.23
Child	24.99*	3.13
Sibling/Other Relative	4.16	18.75
Friend/neighbor	9.38	12.47
Religious Group	0.00	6.25
Other	36.45*	6.25

Note.

^*p < .05,

^***p < .001.

^aBecause women in this sample could have reported giving or receiving either support or assistance to multiple people within one day, each of the four categories of support/assistance exchanges do not total 100%.

Exchanges of instrumental support

With regard to giving and receiving daily instrumental assistance, similar patterns emerged in independent samples t-test results between married women and widows. Widows gave significantly more assistance to their friends/neighbors than did married women, and widows received more daily assistance from children and others (strangers, non-profit groups, etc.) than did married women. Overall, married women gave the most instrumental assistance to their children. Widows, however, both gave and received comparable amounts of instrumental assistance with friends, neighbors, siblings, and other relatives in addition to their children.

Daily and Global Well-Being for Married and Widowed Women

As anticipated, there were no statistically significant differences in either daily or global well-being between the two groups. Widows and married women, on average, reported experiencing daily negative affect “none” or “a little of the time” (M = 0.15, SD = 0.16 vs. M = 0.14, SD = 0.15, respectively) and they reported experiencing daily positive emotions “most of the time” (M = 2.93, SD = 0.60 vs. M = 2.89, SD = 0.57, respectively). Global measures of well-being suggest that widows and married women self-report their mental/emotional health as either “average” or “good” (M = 3.77, SD = 0.91 vs. M = 3.78, SD = 0.85).

Discussion

This study sought to evaluate potential differences and similarities in daily time use and both daily and global well-being between widows and married women. Overall, we found little evidence to support the stereotype of older women who are long-term widows as frail and dependent, and considerable evidence that these widows were psychologically healthy and functioning similarly to women of similar age who were still married. Widows and married women showed no differences in daily or general well being, and similarities in most aspects of time use. Differences that were found were small in magnitude, but of interest in understanding how widowhood affects daily life. Widows spent more time watching TV than married females, which may be reflective of increased solitary activities as a result of living alone. The difference in time spent sleeping (widows sleeping less, on average) might be related to increased TV time. Controlling for age, widows spent significantly more time with children, suggesting that widows continue to appreciate companionship from children over time. We predicted that widows would spend more time doing chores and receive more support than married women; however, these hypotheses were not supported. Because these widows have had time to cope with both the loss-and restoration-oriented aspects of widowhood, similarities in time use and well-being may be indicative of resilience over many years of adjustment. Consistent with Bonnano et al.’s research, most women may return to normal functioning after the death of a spouse (Bonnano et al., 2002; Bonanno, 2009), and widows who live on average 11 years after bereavement may be generally healthy and return to pre-bereavement daily activities.

The differences in the exchange of daily emotional support and unpaid daily assistance may reflect adaptation in both loss- and restoration-oriented coping. Widows exchange more daily emotional support with their friends, neighbors, siblings, other relatives, and children; this exchange in multiple relationships may help women cope with the loss of the close relationship they may have had with their spouse (loss-oriented coping). Conversely, the increased unpaid daily assistance that widows receive from non-spousal relationships may reflect the widows’ need for increased support to cope with changes in daily life post loss (restoration-oriented coping).

As predicted, neither daily nor global well-being were significantly different between the two groups. The current study adds information regarding daily well-being to the current field consisting primarily of global well-being measures and retrospective studies. This research supports findings that non-recent widows as a whole may be doing well, and adds to this research with similar results regarding daily assessments of time use and well-being. Despite the large number of studies identifying potential declines related to well-being (Bennett, 2005; Fry, 2001; Lee et al., 2001; Carr, 2003), daily aspects of widowhood may be generally less negative than previously thought.

Overall, results refute stereotypes of widows as frail, dependent, and distressed, and support recent research showing that most widows do not have clinically significant distress with the passage of sufficient time. Results indicate that many widows are resilient and find ways to accommodate to the loss of a spouse through day-to-day activities. In working with older women, it is important to acknowledge both the vulnerabilities that widowhood can provide, but also to recognize the strengths that many older women have in facing this very common late life transition. Specifically, assistance from friends or neighbors (e.g., yard work, cooking) may help widows adjust to the changes in daily life that may become stressful over time.