Combining Formal and Informal Caregiving Roles: The Psychosocial Implications of Double- and Triple-Duty Care

Abstract

Objectives.

Women who combine formal and informal caregiving roles represent a unique, understudied population. In the literature, healthcare employees who simultaneously provide unpaid elder care at home have been referred to as double-duty caregivers. The present study broadens this perspective by examining the psychosocial implications of double-duty child care (child care only), double-duty elder care (elder care only), and triple-duty care (both child care and elder care or “sandwiched” care).

Method.

Drawing from the Work, Family, and Health Study, we focus on a large sample of women working in nursing homes in the United States ( n = 1,399). We use multiple regression analysis and analysis of covariance tests to examine a range of psychosocial implications associated with double- and triple-duty care.

Results.

Compared with nonfamily caregivers, double-duty child caregivers indicated greater family-to-work conflict and poorer partner relationship quality. Double-duty elder caregivers reported more family-to-work conflict, perceived stress, and psychological distress, whereas triple-duty caregivers indicated poorer psychosocial functioning overall.

Discussion.

Relative to their counterparts without family caregiving roles, women with combined caregiving roles reported poorer psychosocial well-being. Additional research on women with combined caregiving roles, especially triple-duty caregivers, should be a priority amidst an aging population, older workforce, and growing number of working caregivers.

Double-duty care, Healthcare employees, Psychosocial well-being, Sandwiched generation, Triple-duty care, Working caregivers.

A lthough women traditionally serve as family caregivers and predominately fill caregiving occupations in the healthcare industry ( National Alliance for Caregiving and AARP, 2009 ; Paraprofessional Healthcare Institute, 2011 ), women who combine paid, formal care and unpaid, family care roles remain an understudied population ( Ward-Griffin, 2004 ; Ward-Griffin, Brown, Vandervoort, McNair, & Dashnay, 2005 ). In the literature, paid healthcare professionals who simultaneously provide unpaid elder care at home are referred to as double-duty caregivers ( Rutman, 1996 ). Double-duty care, however, comprises a limited body of research. Conceptually, double-duty care overlooks other types of family caregiving roles. Further, double-duty care research has primarily been based on qualitative data, women working outside of the United States, and relatively small samples (e.g., Boumans & Dorant, 2014 ; Ross, Rideout, & Carson; 1994 ; Rutman, 1996 ; Scott, Hwang, & Rogers, 2006 ; Ward-Griffin, 2004 ; Ward-Griffin et al., 2005 ; Ward-Griffin, St-Amantr, & Brown, 2011 ).

Using a large sample of women working in nursing homes in the United States, the current study builds on and extends the literature by considering different combined caregiving roles among healthcare employees. That is, we broaden the definition of double-duty care to refer to healthcare employees providing a single family caregiving role at home, whether it be child or elder care, and distinguish between these roles with the terms “double-duty child care” and “double-duty elder care,” respectively. We also account for healthcare employees in the sandwiched generation (e.g., Miller, 1981 ), meaning women who provide both child and elder care at home; we call these women “triple-duty caregivers” to emphasize their simultaneous engagement in three different caregiving roles—formal care, child care, and elder care. Additionally, we examine the psychosocial implications associated with both double- and triple-duty care by applying the stress process model ( Pearlin, Mullan, Semple, & Skaff, 1990 ). Specifically, we assess subjective primary (perceived stress, psychological distress) and secondary (work and family role strains) stressors among women with combined caregiving roles.

Double- and Triple-Duty Care

Nearly 66 million adults provide some form of unpaid family care in the United States ( National Alliance for Caregiving and AARP, 2009 ), constituting the largest source of long-term care in the nation ( AARP Public Policy Institute, 2011 ). As the population rapidly ages ( Jacobsen, Kent, & Lee, Mather, 2011 ), there will be a greater need for family caregivers’ contributions. The interplay of recent societal and demographic trends, however, has complicated the provision of family care. Americans are remaining in the workforce longer and employment among older women is at a record high ( Blakely, 2011 ; Jacobsen et al., 2011 ). At home, women continue to outnumber men as family caregivers ( National Alliance for Caregiving and AARP, 2009 ). Collectively, these trends indicate that the number of women balancing work and caregiving roles is increasing ( Gordon, Pruchno, Wilson-Genderson, Murphy, & Rose, 2012 ). Caregiving is further complicated by delayed childbirth and longer life expectancy, which can leave women “sandwiched” between caring for dependent children and older adults while working (e.g., Pierret, 2006 ). Yet, one of the most complex combinations of work and caregiving roles is also one that has been neglected in the literature—women who provide formal care at work and informal care at home or double-duty care ( Ward-Griffin, 2004 ).

The healthcare industry is currently facing significant challenges from accelerated population aging, such as retaining an aging workforce already older than that found in other industry sectors, providing care to a rising number of older patients, and recruiting healthcare workers to counter a potentially severe nursing shortage amidst an increasing demand for services ( Harrington & Heidkamp, 2013 ; Hatcher, 2006 ; Institute of Medicine of the National Academies, 2008 ). Given that women dominate the majority of professions within the healthcare industry and typically enact family caregiving roles ( National Alliance for Caregiving and AARP, 2009 ; U.S. Bureau of Labor Statistics, 2013 ), the increasing likelihood that healthcare employees will be balancing caregiving demands at home presents an additional challenge ( Ward-Griffin et al., 2009 ). Despite this projection, empirical investigations of women simultaneously participating in both private and public domains of care were “almost non-existent” a decade ago ( Ward-Griffin et al., 2005 , p. 381) and remain limited to date ( Boumans & Dorant, 2014 ). Instead, researchers have primarily considered professional, paid caregiving and personal, unpaid caregiving to be separate domains ( Ward-Griffin, 2008 ).

We address this limitation by conducting a timely examination of the psychosocial implications associated with double- and triple-duty care. To our knowledge, only one study has examined the implications of multiple caregiving roles among healthcare employees ( Scott et al., 2006 ). That study assessed the impact of providing child, elder, and sandwiched care at home on hospital staff nurses’ fatigue and stress. Nurses with family caregiving roles reported more stress, physical fatigue, and mental fatigue relative to their nonfamily caregiving counterparts. Specifically, levels of fatigue and stress were highest among nurses engaging in sandwiched care at home. However, the Scott and coworkers (2006) study was limited by a small sample of elder and sandwiched caregivers (32 and 29, respectively). The current study is therefore the first to examine the psychosocial implications associated with different family caregiving roles among a large, unique sample of women working in nursing homes in the United States.

Conceptual Framework: The Stress Process

We adapted the stress process model ( Pearlin et al., 1990 ) to examine the perceived psychosocial implications of double- and triple-duty care. The stress process model is based on the concept of proliferation ( Pearlin & Aneshensel, 1994 ; Pearlin, Aneshensel, & LeBlanc, 1997 ). Proliferation occurs when a stressor or set of stressors experienced in the family caregiving role generate new problems in other roles or lead to an accumulation of stressors that spread to multiple life domains (e.g., work). The stress process model also distinguishes between primary and secondary stressors. Within the context of family caregiving, primary stressors are directly rooted in caregiving hardships and drive the stress process by producing other stressors. In this paper, primary stressors refer to subjective appraisals of hardships encountered in family caregiving roles. Secondary stressors are exacerbated or influenced by primary stressors; they arise as a result of caregiving demands but occur in other roles. In the current study, we pose the following questions for both sets of stressors: How do nonfamily caregiving healthcare employees differ from women with combined caregiving roles? How do women with combined caregiving roles differ from each other?

Primary Stressors

We consider two subjective indicators of primary stressors, perceived stress and psychological distress. Given that women in combined caregiving roles provide care at work and at home, they may be exposed to challenges beyond typical stressors associated with family caregiving. For instance, relative to home-based care workers, nursing home employees indicate more physical and emotional strain ( Hasson & Arnetz, 2008 ). Indeed, these nurses engage in physically demanding tasks and experience workplace stressors ranging from witnessing the decline and/or death of patients to respecting patients’ autonomy when confronted with physical resistance ( Ross et al., 1994 ; VonDras, Flittner, Malcore, & Pouliot, 2009 ). However, healthcare employees may anticipate or expect such stressors because it is likely that they self-selected into their employment role ( Pearlin & Aneshensel, 1994 ). That is, because healthcare employment is a role that women have embarked on, prepared for, or settled into, they are aware of or familiar with its stressors. Similarly, double-duty child caregivers may be balancing two expected care roles in that child care is often regarded as a more normative or planned role relative to other family care roles ( Pearlin & Aneshensel, 1994 ). Further, unlike caring for an elderly person, child care recipients become increasingly independent over time and care arrangements are often more stable ( Hessel & Keck, 2009 ). Thus, double-duty child caregivers may subjectively experience hardships more similarly to nonfamily caregivers than double-duty elder or triple-duty caregivers.

In contrast, double-duty elder and triple-duty caregivers may be balancing expected and unexpected care roles. Qualitative data from double-duty elder caregivers suggests that they are often not self-selected into their family caregiving role; rather, women describe facing strong familial pressure to provide family care because of their gender and healthcare expertise ( Ward-Griffin et al., 2005 ). Double-duty elder caregivers also mention family members’ heavy reliance on them for providing or coordinating care, being pressured to accept more care responsibilities than desired, and having high expectations of themselves to provide quality care. These findings are reminiscent of role captivity, wherein an individual unwillingly enacts or feels trapped by a family caregiving role ( Pearlin et al., 1997 ). Role captivity is associated with distress, partially because family caregivers feel that they lack control or power over aspects of their own lives ( Pearlin & Aneshensel, 1994 ). In addition, elder care has been considered more emotionally demanding as well as less predictable than child care given that elder care recipients’ needs increase over time, change suddenly, and last for an unspecified period of time ( Hessel & Keck, 2009 ). Double-duty elder and triple-duty caregivers may therefore indicate more primary stressors relative to their nonfamily caregiving counterparts and double-duty child caregivers.

Secondary Stressors

The designated secondary stressors for our investigation focus on role strains within the major institutions of work and family. With women increasingly balancing employment and family caregiving, interest in work–family issues and the implications of multiple roles has grown ( Gordon et al., 2012 ). Yet, research to date has largely not considered the work–family interface for women with combined caregiving roles ( Ward-Griffin, 2004 ). Acknowledging that secondary stressors typically arise when there are conflicts between social roles ( Knussen, Tolson, Swan, Stott, & Brogan, 2005 ), the current study examines healthcare employees’ appraisals of work–family conflict (bidirectional) and work-to-family (unidirectional) positive spillover. Although positive spillover may seem discrepant in comparison to our other psychosocial indicators, we consider this a work–family strain or subjective secondary stressor to the degree in which work-to-family positive spillover is diminished or absent ( Pearlin et al., 1990 ).

While we are unaware of any studies that have assessed work–family stressors among triple-duty caregivers, prior evidence suggests that double-duty elder caregivers indicate both positive and negative work–family outcomes, whereas double-duty child caregivers report negative effects. For instance, a previous study of hospital-based nurses found that double-duty child care was correlated with work–family conflict, but double-duty elder care was not ( Gottlieb, Kelloway, & Martin-Matthews, 1996 ). In contrast, Boumans and Dorant (2014) found that more hours of family care provision was associated with greater work–family conflict as well as family-to-work positive spillover among double-duty elder caregivers. These work–family conflict findings are consistent with qualitative data from double-duty elder caregivers in which women express feelings of prolonged caregiving, the perception of being on-call all day with limited or no time off, and a continuous negotiation and subsequent renegotiation of professional and personal role boundaries ( Ward-Griffin et al., 2011 ). Boumans and Dorant (2014) attributed their positive spillover findings to the role expansion perspective, which posits that individuals benefit from balancing multiple roles by accumulating or enhancing resources such as mastery, social support, and personal accomplishment. Indeed, double-duty elder caregivers identify recognition, self-esteem, and opportunities for personal and family growth as rewarding aspects of family caregiving in qualitative interviews ( Ross et al., 1994 ). Boumans and Dorant (2014) also noted that role expansion may be especially applicable to double-duty elder caregivers because their care roles are essentially an extension of each other, with similar knowledge or skills used in both roles to manage comparable care demands.

The current study also utilizes a subsample of healthcare employees in a wider family network through their roles as a spouse or partner ( n = 893). The presence of a spouse or partner does not mean an individual automatically benefits from partner support; rather, relationship quality is more critical for support ( Pearlin, Lieberman, Menaghan, & Mullan, 1981 ). Thus, women who positively appraise partner relationship quality may be able to capitalize on partner support to avoid, eliminate, or reduce stressors when confronted with caregiving hardships. For instance, support encompassing both personal and professional resources has been negatively correlated with the degree of blurring between professional and personal care roles ( Ward-Griffin et al., 2011 ). We therefore assess healthcare employees’ perceptions of partner relationship quality (partner support and strain) to understand whether women with combined caregiving roles view partners as resources or stressors. Similar to work-to-family positive spillover, we considered partner support to be a stressor to the degree in which it is diminished or absent. We also decided not to examine partner support as a moderating resource to explain stressor variation; rather, we opted for a comprehensive assessment of proliferated stressors among women with combined caregiving roles. That is, we sought to assess differences in exposure to proliferated stressors and examine whether healthcare employees with varying family caregiving roles are different in their exposure to stressors rather than assume similar exposure across roles ( Pearlin et al., 1997 ).

Method

Sample and Procedures

Data derive from the Work, Family, and Health Study (WFHS), which examined work and family life outcomes among healthcare employees ( Bray et al., 2013 ). Employees were recruited from 30 nursing homes spanning six states, all owned by the same long-term health and specialized care company. Eligible employees worked a minimum of 22.5hr per week in direct patient care, were not exclusively night shift workers, and received study information through recruiting materials and informed consent documents. Of 1,783 eligible employees, 1,524 (85%) enrolled in the WFHS. Participants completed 1-hr long computer-assisted personal interviews regarding work experiences, personal well-being, and family relationships at a private location in the workplace.

The current study focuses on the 1,399 women who agreed to be interviewed. Women had a mean age of 38.7 ( SD = 12.7, range = 18–72), 72% were non-Hispanic White, and 64% were married or in a cohabiting relationship. Most (67%) held positions as certified nursing assistants (CNA) and registered nurses or licensed practical nurses (28%). Sixty-one percent of women had some college education or were college graduates and the modal household income was $60,000 or more. Just over half (52%) worked a regular daytime shift and the average company tenure was 6.33 years ( SD = 6.61, range = 1–42).

Measures

Combined caregiving roles.

We categorized employees into four groups based on existing child and elder care measures from the WFHS, which are comparable to those found in other studies (e.g., Dautzenberg et al., 2000 ; Neal & Hammer, 2007 ; Tement & Korunka, 2013 ). Dependent child care was defined as having children 18 years of age or younger living with the employee for at least 4 days per week. Elder care was defined as providing care (e.g., help with shopping, medical care, or assistance in financial/budget planning) for at least 3hr per week in the past 6 months to an adult relative, regardless of their living situation. Triple-duty caregivers satisfied criteria for both the child and elder care measures. Two women were excluded from family care classifications for indicating they had children but not providing their ages. Overall, there were 498 (36%) nonfamily caregivers, 475 (34%) double-duty child caregivers, 228 (16%) double-duty elder caregivers, and 196 (14%) triple-duty caregivers. On average, both double-duty child ( SD = 0.95) and triple-duty ( SD = 1.02) caregivers reported having 1.93 children living at home. The average age of dependent children was 6.81 ( SD = 5.18) and 8.29 ( SD = 5.31) in double-duty child and triple-duty caregiving households, respectively.

Primary stressors.

We used a global measure of perceived stress ( Cohen, Kamarck, & Mermelstein, 1983 ) and the K6 measure of psychological distress ( Kessler et al., 2003 ). Perceived stress comprised four items (e.g., “How often have you felt that things were going your way?”) pertaining to the last 30 days. Responses ranged from 1 ( very often ) to 5 ( never ). We reverse-coded two items and summed responses to compute perceived stress scores ranging from 4 to 20, with higher scores reflecting more stress (α = .76). Six items examined psychological distress (e.g., “How much of the time did you feel hopeless?”) during the past 30 days. Responses ranged from 1 ( none of the time ) to 5 ( all of the time ). We summed responses to compute a composite distress score ranging from 6 to 30, with higher scores denoting greater distress (α = .84).

Secondary stressors.

We used the work–family conflict scale from Netermeyer, Boles, and McMurrian (1996) . Five items pertained to work-to-family conflict (e.g., “The demands of your work interfere with your family or personal time”) and five items assessed family-to-work conflict (e.g., “Family-related strain interferes with your ability to perform job-related duties”) in the past 6 months. Responses ranged from 1 ( strongly disagree) to 5 ( strongly agree). Higher scores reflect more work-to-family (α = .88) and family-to-work (α = .79) conflict. The WFHS also included the affective spillover subscale from Hanson, Hammer, and Colton (2004) to examine work-to-family positive spillover. The scale comprised four items (e.g., “Being happy at work helps you to be happy at home”), with responses ranging from 1 ( strongly disagree) to 5 ( strongly agree). Higher scores indicate greater work-to-family positive spillover (α = .85).

We adopted a measure of partner relationship quality from Schuster, Kessler, and Aseltine (1990) . Five items evaluated support (e.g., “Does your partner appreciate you?”) and five items examined strain (e.g., “Does your partner criticize you?”) within the past month. Responses ranged from 1 ( not at all ) to 4 ( a lot ). We summed all items for each subscale, which resulted in a total score ranging from 5 to 20. Higher scores reflect greater support (α = .92) and strain (α = .83).

Covariates.

We selected a number of demographic, employment, and health covariates motivated by the stress process model and past research. The model considers caregivers’ social and economic background characteristics, specific features of the family caregiving context, and the caregivers’ health status to be key characteristics influencing stressors ( Pearlin et al., 1990 ). Caregivers’ ascribed statuses, such as age (in years) and race (White, other), along with educational attainment (postsecondary education or not) and income (measured in $5,000 increments up to $60,000 or more), are embedded in the stress process. Further, adults’ perceived stress, subjective well-being, and caregiving strain have differed by demographic characteristics in prior research ( Jivraj, Nazroo, Vanhoutte, & Chandola, 2014 ; Perkins et al., 2013 ; Vasunilashorn, Lynch, Glei, Weinstein, & Goldman, 2014 ); we therefore examine these background characteristics as potential statistical controls. We also consider marital status (married/cohabiting versus single) because partners are part of employees’ networks and may provide support; indeed, the continuously married are considered advantaged in psychological well-being ( Sasson & Umberson, 2014 ). Additionally, we select features of the employment role controlled for in previous double-duty care and healthcare employee studies, including hours worked per week, company tenure, and occupational status (CNA vs. other) ( Boumans & Dorant, 2014 ; Schrijnemaekers et al., 2003 ). We also include the psychological job demands scale from Karasek and coworkers (1998) because work demands are linked to greater work–family conflict ( Gordon et al., 2012 ), and elder care is associated with work-related strain ( Trukeschitz, Schneider, Mühlmann, & Ponocny, 2013 ); higher scores indicate greater job demands (α = .59).

To account for family caregiving context, we control for developmental disabilities, physical health problems, or long-term, serious mental health problems among dependent children. With regard to health status, we assess employees’ diagnoses of chronic health conditions (diabetes, cancer, and high blood pressure) as such conditions have been differentially associated with caregiving strain ( Perkins et al., 2013 ). Similar to prior research, history of chronic health conditions was obtained by dichotomous indicators (e.g., “Has a doctor ever told you that you have cancer?”) ( Perkins et al., 2013 ). Given its relationship to emotional well-being over the life course ( Windsor, Burns, & Byles, 2013 ), we also examine employees’ functional disabilities with the physical functioning subscale from the SF-36, which assesses the ability to perform various physical activities (e.g., climbing stairs); higher scores indicate better physical functioning (α = .87; Ware & Sherbourne, 1992 ).

Analytic Strategy

We first examine employees’ demographic, employment, and health characteristics for potential inclusion as covariates in multivariate analysis by conducting analysis of variance tests with Tukey post hoc comparisons to identify mean differences among nonfamily, double-duty child, double-duty elder, and triple-duty caregivers. We include any variables on which the groups differ significantly, in addition to child disability, as covariates. Next, we perform multiple linear regression analysis to predict the psychosocial implications (primary stressors: perceived stress, psychological distress; secondary stressors: work–family conflict, partner relationship quality) of double- and triple-duty care with nonfamily caregivers as the reference group. We also analyze adjusted mean differences with analysis of covariance (ANCOVA) tests using Tukey post hoc comparisons to assess how women with combined caregiving roles differ from each other. For relationship quality analyses, we restrict the sample to married and cohabiting employees ( n = 893).

Results

Background and Context of the Stress Process

Table 1 provides information about demographic, work, and health characteristics by combined caregiving roles. Employees differed by age, marital status, race, company tenure, psychological job demands, functional disability, and high blood pressure. We therefore used these variables, as well as child disability, as covariates. On average, double-duty child and triple-duty caregivers were younger than nonfamily and double-duty elder caregivers, married or cohabiting to a greater extent than double-duty elder caregivers, and were more racially diverse and had shorter company tenure than nonfamily caregivers. Additionally, double-duty child caregivers were married or cohabiting more and had better physical functioning scores relative to nonfamily caregivers as well as had shorter company tenure and reported high blood pressure to a lesser extent than double-duty elder caregivers. Both double-duty elder and triple-duty caregivers indicated greater psychological job demands than nonfamily caregivers.

Table 1.

Healthcare Employees’ Background and Caregiving Context Characteristics

	Healthcare employees
Characteristics, n (%)	No family care n = 498 (36%)	Child care(double-duty) n = 475 (34%)	Elder care(double-duty) n = 228 (16%)	Sandwiched care(triple-duty) n = 196 (14%)
Age (in years)	42 (14.94) c,t	34 (8.36) n,e	43 (14.21) c,t	36 (8.97) n,e
Married or cohabiting	0.61 c	0.70 n,e	0.54 c,t	0.67 e
White	0.77 c,t	0.69 n	0.77 t	0.60 n,e
Postsecondary education	0.64	0.57	0.62	0.60
Household income
$34,999 or less	0.31	0.37	0.30	0.39
$35,000–59,999	0.30	0.31	0.37	0.32
$60,000 or more	0.39	0.32	0.33	0.29
Care for disabled child	—	0.17	—	0.26
Certified nursing assistant	0.64	0.70	0.66	0.72
Hours worked per week	37 (7.27)	36 (7.62)	36 (7.72)	37 (7.33)
Years worked for company	7.25 (7.80) c,t	5.29 (4.89) n,e	6.97 (7.37) c	5.73 (5.62) n
Psychological job demands	3.74 (.78) e,t	3.82 (.71)	3.94 (.72) n	3.96 (.76) n
Physical functioning ability	0.91 c	0.94 n	0.90	0.91
Diabetes	0.07	0.08	0.12	0.07
Cancer	0.04	.04	0.07	0.03
High blood pressure	0.27	0.20 e	0.30 c	0.21

Note. Means (and SD s) or proportions are shown. Two women were excluded from care classifications for reporting they had children but not providing information about their ages. Analysis of variance tests with Tukey post hoc comparisons were conducted to identify mean differences across groups. Subscript letters represent family caregiving roles and denote significant differences among groups: n = no family care, c = double-duty child care, e = double-duty elder care, t = triple-duty care.

Psychosocial Implications

Multiple regression and ANCOVA results are presented in Tables 2 and 3 , respectively. Company tenure and high blood pressure were nonsignificant predictors across models and subsequently removed; this did not affect model fit or alter results. For each psychosocial indicator we present regression results first, in which no family care is the reference group, and then discuss adjusted mean differences in outcome scores across all four groups of employees.

Table 2.

Multivariate Linear Regression Analysis: Psychosocial Implications Associated With Double- and Triple-Duty Care

	Subjective primary stressors		Secondary stressors
	Perceived stress( n = 1391)	Psychological distress ( n = 1389)	Work-to-family conflict( n = 1388)	Family-to-work conflict( n = 1390)	Work-to-family positive spillover( n = 1391)	Partner support( n = 882)	Partner strain( n = 883)
	B ( SE )	B ( SE )	B ( SE )	B ( SE )	B ( SE )	B ( SE )	B ( SE )
Age	−.04 (0.01)***	−.05 (0.01)***	−.01 (0.00)***	−.01 (0.00)***	.001 (0.00)	−.01 (0.01) †	.0004 (0.01)
Caucasian	−.41 (0.18)*	−.99 (0.25)***	.05 (0.05)	−.08 (0.03)*	−.01 (0.04)	.36 (0.21) †	−.51 (0.27) †
Married	−.53 (0.17)***	−.86 (0.23)***	.02 (0.05)	−.08 (0.03)**	.11 (0.04)**	—	—
Child disability	1.03 (0.30)***	1.35 (0.40)***	.24 (0.08)**	.06 (0.06)	−.06 (0.07)	−.83 (0.32)**	.19 (0.41)
Psychological job demands	.45 (0.11)***	.71 (0.15)***	.36 (0.03)***	.03 (0.02)	.13 (0.03)***	−.03 (0.12)	.27 (0.16) †
Physical functioning ability	−.03 (0.01)***	−.05 (0.01)***	−.01 (0.00)***	−.003 (0.00)**	−.0002 (0.00)	.01 (0.01) †	−.02 (0.01)**
Double-duty child care	.39 (0.21) †	.07 (0.28)	.01 (0.06)	.12 (0.04)**	−.04 (0.05)	−.55 (0.22)*	1.08 (0.29)***
Double-duty elder care	.53 (0.24)*	1.25 (0.33)***	.10 (0.07)	.16 (0.05)***	.01 (0.06)	−.22 (0.28)	.50 (0.37)
Triple-duty care	.67 (0.27)*	1.25 (0.37)***	.18 (0.08)*	.15 (0.05)**	-.04 (0.06)	-.80 (0.29)**	1.36 (0.38)***
R2	.10	.12	.15	.06	.03	.04	.04

Note . Unstandardized regression coefficients are shown. If one item was missing from an outcome measure, then the remaining items were averaged; employees with additional missing information were excluded from models. Single employees were excluded from couple relationship quality analyses.

^†p < .10, * p < .05, ** p < .01, *** p < .001.

Table 3.

Adjusted Mean Differences in Psychosocial Scores by Combined Caregiving Roles

	Healthcare employees
	No family care n = 498 (36%)	Child care(double-duty) n = 475 (34%)	Elder care(double-duty) n = 228 (16%)	Sandwiched care(triple-duty) n = 196 (14%)	F
Subjective primary stressors	Mean	Mean	Mean	Mean
Perceived stress	9.81 e,t	10.21	10.34 n	10.49 n	2.98*
Psychological distress	12.56 e,t	12.52 e,t	13.82 n,c	13.74 n,c	8.53***
Secondary stressors
Work-to-family conflict	2.84 t	2.85 t	2.94	3.02 n,c	2.62*
Family-to-work conflict	2.04 c,e,t	2.16 n	2.19 n	2.18 n	5.98***
Work-to-family positive spillover	3.94	3.90	3.95	3.90	0.40
Partner support	17.94 c,t	17.39 n	17.72	17.14 n	3.19*
Partner strain	9.27 c,t	10.35 n	9.77	10.63 n	5.91**

Note . Two women were excluded from care classifications for reporting they had children but not providing information about their ages. Covariates include age, race, marital status, child disability, psychological job demands, and physical functioning ability. If one item was missing from an outcome measure, then the remaining items were averaged; employees with additional missing information were excluded from models. Single employees were excluded from partner relationship quality analyses. Subscript letters represent family care roles and denote significant differences among groups: n = no family care, c = double-duty child care, e = double-duty elder care, t = triple-duty care.

* p < .05, * p < .01, *** p < .001.

Nonfamily Caregivers versus Double-Duty Child, Double-Duty Elder, and Triple-Duty Caregivers

Primary stressors.

Double-duty elder and triple-duty care were associated with greater perceived stress and psychological distress in regression analysis. A subsequent ANCOVA complemented these findings; double-duty elder and triple-duty caregivers had higher perceived stress ( p < .05 and p < .05, respectively) and psychological distress ( p < .001 and p <.001, respectively) scores. Additionally, employees who were older, White, married or cohabiting, and had better physical functioning indicated fewer subjective primary stressors, whereas child disability and more psychological job demands were linked to more primary stressors.

Secondary stressors.

In regression analysis, triple-duty care was associated with greater work-to-family conflict, with the same findings emerging in a follow-up ANCOVA ( p < .05). Child disability and more psychological job demands were also related to greater work-to-family conflict, whereas better physical functioning and older age were linked to less work-to-family conflict. All three combined caregiving roles predicted more family-to-work conflict, but none of the combined caregiving roles were associated with work-to-family positive spillover; ANCOVA results complemented these findings in that double-duty child ( p < .01), double-duty elder ( p < .001), and triple-duty ( p < .01) caregivers had higher family-to-work conflict scores, and there were no group differences for positive spillover. Employees who were older, White, married, and had better physical functioning also reported less family-to-work conflict. As for partner relationship quality, both double-duty child and triple-duty care were associated with less support and greater strain. Similarly, double-duty child and triple-duty caregivers had lower support ( p < .05 and p < .01, respectively) and higher strain ( p < .001 and p < .001, respectively) scores. Child disability was also associated with less partner support while better physical functioning predicted less partner strain.

Double-Duty Child Care versus Double-Duty Elder Care versus Triple-Duty Care

ANCOVA results indicated that triple-duty caregivers had higher psychological distress (primary stressor, p < .001) and work-to-family conflict (secondary stressor, p < .05) scores than double-duty child caregivers. Additionally, double-duty elder caregivers had higher psychological distress scores than double-duty child caregivers ( p < .001). Double-duty elder and triple-duty caregivers did not significantly differ on any psychosocial indicators.

Discussion

Women with combined caregiving roles represent a unique group of working caregivers. Given their invaluable societal contributions and an increasing need for their services, it is essential to acquire additional information about this understudied population. The present study builds on existing literature by broadening the perspective of double-duty care, focusing on women working in nursing homes in the United States, and examining the psychosocial implications of double- and triple-duty care.

Double-Duty Caregivers

Our findings suggest that double-duty child caregivers experience fewer stressors, particularly subjective primary stressors, in comparison to employees balancing elder and sandwiched care roles. This group, however, indicated greater secondary stressors in the form of family-related strains relative to nonfamily caregivers. Our results are consistent with a study on hospital-based nurses that found positive correlations between child care and strain- and time-based family-to-work conflict ( Gottlieb et al., 1996 ). Further, parenthood has been linked to lower marital satisfaction, a finding attributed to role conflicts and restriction of parents’ freedom, which reflects double-duty child caregivers’ poorer relationship quality ( Twenge, Campbell, & Foster, 2003 ). Our findings also align with prior research suggesting that child care is a more normative role and that working child caregivers are least likely to experience self-loss or depression ( Hessel & Keck, 2009 ; Pearlin & Aneshensel, 1994 ). Moreover, these results are consistent with our expectation that double-duty child caregivers’ subjectively experience hardships more similarly to nonfamily caregivers than double-duty elder and triple-duty caregivers.

Our results also suggest that double-duty elder caregivers experience more negative subjective primary stressors and family-related strains conflicting with work than their nonfamily caregiving counterparts. These findings complement a prior study by Boumans and Dorant (2014) , which found that more family care hours were associated with poorer mental health, more emotional exhaustion, and increased home-to-work conflict among double-duty elder caregivers. Further, our results align with qualitative evidence from double-duty elder caregivers in which women speak about heightened stress from comprehension of care recipients’ medical conditions; guilt, powerlessness, and helplessness if they lack needed clinical expertise or make poor medical decisions; and blurred boundaries between professional and familial caregiving roles ( Ward-Griffin, 2004 ), all of which may psychologically transfer between work and home domains and manifest through stress and distress.

Triple-Duty Caregivers

Overall, triple-duty care was associated with poorer psychosocial well-being relative to no family care; triple-duty caregivers also had greater work-to-family conflict and psychological distress scores than double-duty child caregivers. These results are consistent with previous work by Scott and coworkers (2006) , who found that hospital staff nurses with sandwiched care obligations at home indicated greater mental fatigue and increased stress compared with their nonfamily caregiving counterparts. Interestingly, psychosocial implications did not differ between double-duty elder and triple-duty caregivers; rather, these caregivers both reported more family-related strain and negative primary stressors compared with nonfamily caregivers. Unlike double-duty elder caregivers, however, triple-duty caregivers reported greater work-related strain and poorer relationship quality than nonfamily caregivers, which raises questions about the availability and utilization of professional and personal support among this group.

Limitations

Our study is the first to use a large sample of women working in nursing homes in the United States to examine the psychosocial implications of double- and triple-duty care. Nevertheless, we utilize self-report data and a cross-sectional, correlational design, which constrains the ability to identify causal relationships or detect changes in stressors. Considering our focus on nursing home workers, our findings may also not be generalizable to other healthcare employees (e.g., physicians). Although we used a diverse sample, everyone worked for the same corporation, further limiting generalizability. We were also unable to exhaustively examine covariates with the potential to influence stressors, such as employees’ perceptions of the division of household labor or specific features of caring for an older adult. Given our use of multiple regressions, our analysis does not account for the bias of unobserved variables. Further, although we assessed negative and positive work–family outcomes, the WFHS did not include a family-to-work positive spillover measure.

Moreover, we conducted a secondary analysis of data from the WFHS, which was not specifically designed to study caregiving. Similar to Tement and Korunka (2013) and Scott and coworkers (2006) , we applied proxy measures of family caregiving roles to determine role occupancy and did not possess information regarding objective primary stressors or caregiving demands (e.g., elder care recipients’ health status). However, subjective reactions to caregiving have played a larger role in prior research on stress proliferation among caregivers than objective physical demands ( Knussen et al., 2005 ). Additionally, a subsample of employees in the current study participated in qualitative interviews in which they were asked about elder care obligations. Employees mentioned caring for an aging relative who is chronically ill, in declining health, or has multiple health problems (end-stage renal disease, dementia, diabetes), monitoring parents’ health status, assisting with recovery from adverse health events (post-stroke, hip replacement surgery, paralysis), helping with daily activities (bathing, dressing, transportation), managing parents’ estates, and overseeing professional or hospice care. Employees also mentioned being emotionally preoccupied with care recipients’ health when at work. Thus, even though we do not assess objective primary stressors, qualitative evidence from the WFHS suggests that employees cared for older adults with a variety of health conditions and engaged in a range of caregiving tasks. One advantage of the current approach is that, given the diversity of care situations reported, our sample may be more representative of working caregivers than a sample selected for a certain threshold of care or diagnosis (e.g., dementia). Indeed, our heterogeneous sample may actually underestimate the psychosocial implications of double- and triple-duty care.

Future Directions

We have several suggestions for future double- and triple-duty care research, which should be a priority amidst an aging population, older workforce, and growing number of working caregivers. First, researchers should assess whether these caregivers find existing workplace programs, practices, and policies relevant or useful for their unique work–family demands. This is particularly important for single healthcare employees with family care obligations. Nearly half of double-duty elder caregivers in the WFHS sample were single, automatically leaving them without access to the potential resource of partner support. To compensate for a lack of resources at home, single double- and triple-duty caregivers may rely on workplace supports more heavily. Alternatively, supportive policies may be in place but family caregivers may fear workplace penalties for utilization or overutilization of available formal support. Prospective qualitative research could explore these issues further by asking double- and triple-duty caregivers to identity perceived barriers to workplace supports and also discuss key resources at home for facilitating work-life balance in the absence of partner support.

Further, healthcare employers can administer workforce and workplace assessments to examine as well as monitor the changing demographics, skills, and work-life demands of their current workforce ( Harrington & Heidkamp, 2013 ; Neal & Hammer, 2007 ; Ward-Griffin et al., 2009 ). The subjective stressors associated with double- and triple-duty care will likely become a greater concern for the healthcare industry as it seeks to retain or hire employees with an increased likelihood of family care obligations at home. Workplace assessments may provide important insight regarding effective strategies for attracting and retaining employees as well as identifying alternate pathways into retirement ( Sweet, Pitt-Catsouphes, Besen, Hovhannisyan, & Pasha, 2010 ). Feedback from these assessments could also be used to develop appropriate and targeted work-life strategies, flexible or customized scheduling practices, family-friendly policies, health-promoting practices, and supervisor behavior training programs with an emphasis on work–family support ( Kossek, Hammer, Kelly, & Moen, 2014 ). In turn, such changes could reduce turnover, increase job satisfaction, enhance health, and facilitate work-life balance ( Harrington & Heidkamp, 2013 ), all of which may benefit care recipients by decreasing caregivers’ stressors and increasing resources. Moreover, engaging employees in decision-making processes regarding workplace practices and policies may provide family caregivers with the opportunity to inform employers about aspects of their job that create work–family conflict and also propose ways to ameliorate such situations ( Sweet et al., 2010 ). Additional research on the prevalence, types, and implications of double- and triple-duty care among both women and men employed in different occupations within the healthcare industry is also warranted.

Funding

This research was conducted as part of the Work, Family and Health Network ( http://www.WorkFamilyHealthNetwork.org ), which is funded by a cooperative agreement through the National Institutes of Health and the Centers for Disease Control and Prevention: Eunice Kennedy Shriver National Institute of Child Health and Human Development (U01HD051217, U01HD051218, U01HD051256, U01HD051276), National Institute on Aging (U01AG027669), Office of Behavioral and Social Sciences Research, and National Institute for Occupational Safety and Health (U01OH008788, U01HD059773). Grants from the William T. Grant Foundation, Alfred P. Sloan Foundation, and the Administration for Children and Families provided additional funding. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of these institutes and offices.

Acknowledgments

Special acknowledgement goes to Extramural Staff Science Collaborator, Rosalind Berkowitz King, PhD and Lynne Casper, PhD for design of the original Workplace, Family, Health and Well-Being Network Initiative. We also wish to express our gratitude to the worksites, employers, and employees who participated in this research.

Author notes

Decision Editor: Bob G. Knight, PhD