Religious Involvement and the Impact on Living Alone and Depression: An Examination of Gendered Pathways Post-Widowhood

Abstract

This study analyzed data from the nationally representative Health and Retirement Study (HRS) to examine the association between widowhood and depression. The results revealed that both men and women experienced increased depression after widowhood, with women exhibiting a better recovery pattern over time. Religiosity, especially attending religious services, was found to be a beneficial coping mechanism for both genders, although men were less religious than women. Living alone was a significant factor associated with depression in widowhood for both men and women. High religious service attendance moderated the association between living alone and depression for women, while both high and moderate religious service attendance moderated this association for men. These findings emphasize the importance of considering gender differences and the potential benefits of religious involvement in addressing depression during widowhood, highlighting the need for tailored interventions and support services for widows and widowers, particularly those living alone.

Keywords

widow/widower depression religion/religiosity social support gender

What this paper adds

• This study contributes to the existing literature on the emotional impact of widowhood by identifying gender differences in depression rates and long-term recovery patterns following the loss of a spouse.

• This paper contributes to the existing literature by identifying living alone as a significant factor associated with depression in widowhood for both men and women and further examines the impact of religiosity on these relationships.

• The findings underscore the significance of integrating religious practices into intervention and support services for widowed individuals as a means to reduce the risk of depression and effectively cope with the challenges of widowhood.

Applications of study findings

• Targeted interventions can improve emotional well-being and reduce depression risk among widowed older adults, particularly for those living alone who may be at a higher risk of depression post-widowhood.

• Healthcare professionals can use the study results to increase screening and treatment for depression among widowed older adults, particularly for men who may be less likely to seek help or engage in religious practices.

• Community-based support services, including faith-based organizations, can provide emotional and spiritual support to widowed individuals, highlighting the importance of incorporating religious and spiritual practices into the support and intervention services for this population.

Introduction

Depression is a prevalent negative health outcome associated with spousal bereavement in widowhood (Bisconti et al., 2004; Kristiansen et al., 2017, 2019; Lyu et al., 2019; Yang & Kozloski, 2012; Yang et al., 2015). Living alone after the loss of a spouse can exacerbate this vulnerability due to reduced social support and increased loneliness (Cacioppo & Hawkley, 2006; Utz et al., 2014); and research has shown that living alone plays a mediating role in the relationship between widowhood and depression in older adults (Dean et al., 1992; Guo et al., 2021). Given the growing population of older adults living alone, understanding how to promote social connectedness and emotional well-being for widowers, especially those living alone, is crucial in addressing the negative health outcomes associated with widowhood, such as depression (Hawes et al., 2022; Jeon et al., 2017; Srivastava et al., 2021).

Shielding factors that may aid in coping with loss have been identified through research (Recksiedler & Stawski, 2019; Stolzenberg & Waite, 2006), with religiosity being one such factor (Garcini et al., 2021). In the existing literature, scholars have examined the multifaceted dimensions of religiosity. These dimensions encompass various aspects of organizational and non-organizational religious involvement, including intrinsic religiosity (reflecting the acceptance of core religious doctrines), religious exclusivity (pertaining to the rigidity of faith and moral convictions), religious involvement (involving observable religious behaviors in a communal context), personal practice (focusing on individual religious activities such as private prayer), and religious salience (measuring the importance of religion relative to other life influences).

A study by Hawes et al. (2022) sheds light on the correlation between religiosity and experiences related to widowhood and depression, particularly among those adapting to a solitary lifestyle. Drawing on longitudinal data spanning from 2006 to 2016 from the Health and Retirement Study (HRS), their research focuses on individuals aged 50 and above to analyze depression trajectories in the context of widowhood. To assess religiosity, they utilized three measures: evaluating organizational religious involvement through the frequency of religious service attendance, assessing non-organizational religious engagement via private prayer frequency, and measuring intrinsic religiosity through questions capturing beliefs such as divine intervention and finding strength in religion. Their findings revealed that older adults experience a significant increase in depressive symptoms upon entering widowhood, followed by a post-widowhood decrease that doesn't fully return to pre-widowhood levels. Notably, heightened religious service attendance and intrinsic religiosity were associated with reduced depressive symptomology. Additionally, religious service attendance emerged as a moderating factor in the relationship between widowhood and depression among older adults living alone.

Building upon these insights, an examination of potential gender variations in the connections between depressive symptoms, widowhood, and solitary living is warranted. While gender has been explored in relation to some of these aspects individually, there remains a gap in our understanding when considering all these dimensions collectively. Existing research, exemplified by the work of Sasson and Umberson (2014), has emphasized gender disparities in religiosity. Women tend to exhibit higher levels of religious involvement, evident in their greater attendance at religious services and engagement in religious practices (Maselko & Kubzansky, 2006). Moreover, studies have found that women report higher levels of intrinsic religiosity, indicating a stronger personal connection to their faith and spiritual beliefs (Li et al., 2020). Recognizing these gender differences in religiosity is crucial when examining the relationship between religious practices and mental health outcomes.

In addition, mounting evidence suggests that men and women have different reactions to widowhood, and the depressive influence of living alone may also be experienced differently (Sasson & Umberson, 2014). However, studies investigating gender disparities in depression following widowhood have yielded inconsistent results. Some studies have reported a more pronounced negative impact on men (Luhmann et al., 2012; Stroebe et al., 2001; van Grootheest et al., 1999), others have found that women are more susceptible to depression after the loss of a spouse (Chou & Chi, 2000). While research has yielded inconsistent results in examining gender differences in depression following widowhood, no known studies have explored potential differences in the relationship between religiosity and depression by gender—in particular for those living alone. This gap in the literature underscores the need for more comprehensive research that utilizes longitudinal, nationally representative data to explore trends in depression among widowed individuals and to compare the role of religiosity for men and women in the years following spousal loss. Such research may have important implications for developing targeted interventions aimed at improving mental health outcomes for widowed individuals living alone.

Based on existing literature that highlights the protective association between religious practice and depression symptoms (Recksiedler & Stawski, 2019; Stolzenberg & Waite, 2006), as well as the higher vulnerability to depression among widowed individuals living alone (Garcini et al., 2021; Hawes et al., 2022), the purpose of this study is to understand (1) gender specific trends of depressive symptoms and (2) the gender disparity in the effect of living alone and religiosity on depressive symptoms during the transition to widowhood and post-widowhood. By using longitudinal data from the Health and Retirement Study (HRS), we can document long-term trajectories of common symptoms of depression post-widowhood while additionally testing the hypothesis of gender variation.

Methods

Data Source

This study employs a longitudinal design using data from seven waves spanning from 2006 to 2018, which are publicly available through the Health and Retirement Study (HRS). This panel survey is nationally representative and includes community-dwelling adults aged 50 years and older. Since its inception in 1992, the HRS has been conducted biennially and comprises a sample of around 20,000 participants, offering extensive information on various demographic, health, economic, and social factors. The Leave-Behind Questionnaire (LBQ) section of the HRS, initially administered in 2006, was utilized. The LBQ encompasses comprehensive inquiries on psychosocial characteristics and enabled us to assess religiosity more comprehensively by offering information on both organizational religious participation and non-organizational engagement (e.g., private prayer outside of religious services) as well as intrinsic religiosity.

Study Sample

The sample for our study consisted of HRS participants aged 50 and older who were not widowed at baseline (2006) and became widowed between 2006 and 2018. Participants included in the sample must have completed both the HRS Core Survey and LBQ during the study period for a final sample of 1840 (1356 females and 484 males). Because the HRS includes the spouses/partners of participants in the survey even if they are under the age of 50, 43 participants in our sample who were widowed between 2006 and 2018 were aged 41 to 49 at baseline rather than 50 and older.

Measurement

The HRS assesses depression using the 8-item version of the Center for Epidemiological Studies-Depression scale (CES-D 8; Radloff, 1977). To measure depression, this study created two variables: Depression, a binary variable that indicates the presence of depression when three or more depressive symptoms are reported on the CES-D 8 (depressed = 1; non-depressed = 0); and Depressive Symptoms, a tally of the number of depressive symptoms reported on the CES-D scale (ranging from 0 to 8).

The independent variable in this study is marital status, which was assessed by including the HRS question that asks respondents about their current relationship status. This question was coded as a dichotomous variable (widowed = 1; non-widowed = 0). A dichotomous variable for whether an individual re-married/re-partnered was also created for post-widowhood analysis. To measure the duration of widowhood, we calculated the time elapsed from the initial report of widowhood to the study's endpoint in 2018.

Based on prior literature, we assessed religiosity as a moderator variable using three primary measures. Religion variables in the HRS were part of the LBQ which was collected every four years rather than the usual two-year intervals of the core survey. A sensitivity analysis revealed that the religion measures did not change substantially over time prior to widowhood. Thus, for those who did not have religion data for the wave in which widowhood was reported, we utilized the data from the closest wave year prior to widowhood. For post-widowhood analysis, we used the latest reported religion measures to examine the change from the point of widowhood to post-widowhood.

The initial measure examined organizational religious involvement via frequency of attendance at religious services. From this HRS question, we derived a set of dichotomous variables categorized as follows: high (attendance once a week or more), moderate (attendance two to three times monthly), and low/none (attendance one or more times a year or not at all). To measure change in the religious service attendance post-widowhood, a categorical variable was created (0 = no/low attendance, 1 = moderate attendance, and 2 = high attendance) and an additional variable measuring the difference in this categorical variable from time of widowhood to the latest data point was examined. The second variable, Private Prayer Frequency, refers to non-organizational religious involvement and was measured based on the frequency of private prayer outside of religious services (a scale of 0–8 where 8 was the highest frequency of private prayer); the difference in this measure from the point of widowhood to the latest data point was used in post-widowhood analysis. The final measure evaluated intrinsic religiosity and was determined by the average of four HRS questions: “I believe God watches over me,” “The events in my life unfold according to a divine/greater plan,” “I carry religious beliefs into all dealings in life,” and “I find strength in religion.” Respondents answered on a scale ranging from 1 (strongly disagree) to 6 (strongly agree), and the average score across the four items was calculated with higher scores indicating greater intrinsic religiosity; the difference in this measure from the point of widowhood to the latest data point was used in post-widowhood analysis.

Additionally, respondents were asked about their religious affiliation, which included Protestant, Catholic, Jewish, or none/other religion.

Respondents were asked their gender (male/female). We coded the answers to this question as a dichotomous variable. Living alone was measured as a dichotomous variable with a value of 1 for participants who reported living alone and 0 for those who did not.

The control variables in this study consist of demographic covariates such as age, education (in years), race/ethnicity (Hispanic, non-Hispanic White, non-Hispanic Black, non-Hispanic, and other), sex (male/female), retired (yes = 1 and no = 0), and net total wealth (the natural log was used because the variable was not normally distributed). We assessed health covariates using several indicators. These included a self-reported health rating scale, with values ranging from 1 (excellent) to 5 (poor). Additionally, we used a dichotomous variable to distinguish between individuals who reported poor or fair self-rated health (yes = 1) and those who reported excellent, very good, or good health (no = 0). Furthermore, we considered a count of eight self-reported chronic conditions, encompassing heart disease, lung disease, cancer, stroke, high blood pressure, diabetes, arthritis, and psychological problems, with scores ranging from 0 to 8. Activities of daily living limitations (ADLs) were measured by a count of five self-reported limitations, including bathing, dressing, eating, walking across the room, and getting in or out of bed (0–5). Instrumental activities of daily living limitations (IADLs) were also accounted for, including using a telephone, taking medication, handling money, shopping, and preparing meals (0–5). Additionally, we used separate dichotomous variables to track the use of anti-depressants (yes = 1 and no = 0), smoking status (smoker = 1 and non-smoker = 0), alcohol abuse (defined as consuming three or more drinks per day, yes = 1 and no = 0), and regular moderate to vigorous exercise (yes = 1 and no = 0).

Analytic Plan

We examined longitudinal data from the HRS using 2006 as the baseline to investigate the relationship between widowhood, depressive symptomology, religiosity, and gender. The study tracked participants for up to 10 years beyond their transition into widowhood. Descriptive statistics, including percentages and averages, were reported for two gender-stratified groups. We used chi-square tests and t tests to examine significant within- and between-group gender differences in characteristics prior to and at the time of widowhood.

We used two models for our analysis. Model 1 used ordinary least squares (OLS) regression to analyze the relationship between widowhood and depression in the transition to widowhood. Model 2 used OLS to analyze that same relationship in the post-widowhood period. Specifically, Model 2 examined the difference in all covariates from widowhood to the most recent 2018 data point with the exception of race/ethnicity variables that do not change over time and were measured at baseline. We stratified the models by gender and tested them with and without interaction terms for widowhood and religiosity. OLS regression model assumptions were maintained in the analysis such that there was a random sample of observations and coefficients, and error terms were linear with no multi-collinearity among independent variables used in the models. Further, variance of errors remained consistent for all observations with a conditional mean of zero. Multi-collinearity was assessed by examining variance inflation factors (VIFs) in the regression models and using a conservative VIF cutoff of less than 5.0 for all variables included in the models. The independent variable for cognitive impairment had a VIF of 9.32 and was highly correlated with age. A sensitivity analysis showed that the cognitive impairment measure was insignificant across all of the regression models and that removing it made no observable impact on the other coefficients in the models. Thus, to avoid any potential multi-collinearity issues, the cognitive impairment measure was removed from the regression analyses. The LBQ cross-wave sampling weights provided by the HRS were applied for all analyses to account for any selection bias. IBM SPSS version 27 was used for all statistical analyses.

Results

Descriptive Statistics and Bivariate Analysis

Table 1 displays the descriptive statistics for the sample for widows and widowers prior to widowhood (baseline 2006) in comparison to the time of reported widowhood. Results show that women are more likely to be depressed compared to men prior to widowhood (25.3% for women and 17.7% for men). At the point of becoming widowed, depression significantly increased for both men and women, with men displaying slightly higher depression rates at the time of widowhood (34.8% for men compared to 34% for women). The depression spike after widowhood from 25.3% to 34% for women is significant (about a 40% relative increase); but for men, the increase from 17.7% to 34.8% is a relative increase of almost 200%, indicating that men are much more likely to be depressed post-widowhood compared to women. Furthermore, results show that men are significantly less religious than women both before and after widowhood (for all categories of religiosity: intrinsic, attendance, and private prayer). Men are less likely to have high religious service attendance and are more likely to not go to attend at all. However, men post-widowhood showed a slightly higher likelihood of high religious service attendance, whereas for women, high religious service attendance decreased at the time of widowhood.

Table 1.

Gender Differences in Descriptive and Bivariate Statistics at Baseline and Widowhood.

	Baseline 2006		At Time of Widowhood
Mean/Percentage	Women (N = 1356)	Men (N = 484)	Women (N = 1356)	Men (N = 484)
Depressive symptoms (mean)	1.8^a,b	1.3^a,b	2.6^b	2.7^b
Depression	25.3%^a,b	17.7%^a,b	34.0%^b	34.8%^b
Frequency of religious service attendance
High religious service attendance	50.2%^a,b	38.9%^a	46.9%^a,b	39.4%^a
Moderate religious service attendance	12.3%	11.1%	13.2%	10.0%
Low/no religious service attendance	37.6%^a	50%^a	39.8%^a	50.5%^a
Religiosity index (mean) (1–6; 6 = highest)	5.3^a,b	4.7^a	5.5^a,b	4.8^a
Private prayer frequency (mean) (1–6; 6 = highest)	5.3^a,b	4.3^a	4.7^a,b	4.1^a
Religious denomination
Protestant	66.6%	62.8%
Catholic	26.3%	29.5%
Jewish	1.9%	2.2%
Other denomination	.4%	.6%
No denomination	4.8%	4.9%
Age (mean years)	69.9^b	70.8^b	75.0^a,b	77.8^a,b
Race/ethnicity
Non-Hispanic White	73.4%	75.9%
Non-Hispanic Black	13.3%	14.6%
Non-Hispanic other	3.0%	1.5%
Hispanic	10.3%	8.0%
Education (mean years)	12.1	11.9
Household income (mean)	US$51,414^a,b	US$54,963^a	US$43,882^a,b	US$51,358^a
Net wealth (mean)	US$525,202^a,b	US$538,477^a,b	US$408,449^a,b	US$480,279^a,b
Retired	48.4%^a,b	61.6%^a,b	70.5%^a,b	77.3%^a,b
Lives alone	6.6%^a,b	3.8%^a,b	68.8%^a,b	72.3%^a,b
Self-reported fair/poor health	29.7%^b	28.9%^b	34.5%^a,b	37%^b
Chronic conditions (mean)	2.1^a,b	2.7^a	2.6^b	2.8
Anti-depressant medication	23.8%^a,b	11.0%^a,b	31.0%^a,b	20.5%^a,b
Cognitive impairment	1.5%^b	1.9%^b	2.8%^a,b	4.8%^a,b
ADLs (mean)	0.3^b	0.3^b	0.6^b	0.7^b
IADLs (mean)	0.6^b	0.5^b	1.2^b	1.4^b
Smoker	12.5%	13.6%	10.4%	12.0%
Moderate/vigorous exercise regularly	68.7%^a,b	77.9%^a,b	57.9%^b	60.4%^b
Alcohol use 3+ drinks/day	3.4%	4.5%	2.0%	3.7%

^aSignificant t test difference at p < .05 between women and men.

^bSignificant t test difference at p < .05 between baseline and within 2 years of widowhood.

Comparable results were observed for intrinsic religiosity, with higher religiosity reported for both men (4.8 compared to 4.7) and women (5.5 compared to 5.3) at the time of widowhood. In contrast, the frequency of private prayer decreased among widows and widowers during the transition into widowhood compared to baseline (2006).

At the time of widowhood, men reported slightly older ages compared to their women counterparts (77.8 compared to 75 for women). Both men and women experienced significant income and net wealth deficits, with women experiencing income and net worth loss at a higher rate than men. Beyond financial deficiencies, a general decline in health measures were reported, along with higher reports of chronic conditions. Chronic conditions outlined included ADLs, IADLs, fair or poor health, and cognitive impairment. Men more often presented with poorer health outcomes than women post-widowhood. Additionally, decreased levels of moderate-to-vigorous exercise, a greater likelihood of retirement, use of SSRIs like anti-depressants, and living alone were also observed as contributing factors to depression. However, living alone was not always due to widowhood. Of the sample, 5.8% disclosed they were already living alone due to their spouse living in a nursing home or other care facility.

Figure 1 displays the average scores of depressive symptoms from 2006 to 2018. The findings reveal a substantial increase in depressive symptoms during the transition to widowhood. However, women exhibit a more favorable pattern of recovery over time after widowhood. Their depression gradually declines and ten years post-widowhood, their scores are lower than the baseline. Conversely, men also demonstrate a gradual reduction in depressive symptoms as the years go by after widowhood, but their scores remain higher than the baseline even ten years later.

Figure 1.

Depressive symptoms mean comparison for men and women during and after transition into widowhood.

Multivariate Analyses

Table 2 presents the results of the ordinary least squares (OLS) analysis examining the relationship between changes in depressive symptomology and religiosity for both genders during the transition to widowhood. Model 1 shows that high religious service attendance during widowhood was associated with lower depressive symptoms for women (β = −.58, p < .05). For men, both high (β = −.33, p < .01) and moderate (β = −.46, p < .05) religious service attendance were related to decreased depressive symptoms. Among widows, higher religiosity ratings (β = −.10, p < .05) and private prayer (β = −.08, p < .05) at widowhood were associated with lower depressive symptoms from 2006 to widowhood. There were no significant findings for private prayer frequency or intrinsic religiosity for widowers.

Table 2.

Longitudinal OLS Regression Analysis of the Relationship Between Changes in Depressive Symptoms and Religiosity During the Transition Into Widowhood for Men and Women.

	Women (N = 1356)				Men (N = 484)
	Model 1		Model 2		Model 1		Model 2
At Time of Reported Widowhood	β	SE	β	SE	β	SE	β	SE
Age	.02	.01	.02	.01	−.03*	.01	−.03*	.01
Non-Hispanic Black	.18	.37	.18	.39	−.48**	.17	−.48**	.47
Non-Hispanic other	−.56	.74	−.55	.74	−.90	1.25	−.91	1.23
Hispanic	1.18**	.47	1.18**	.44	−1.16	1.00	−1.16	.98
Education	.07	.04	.07	.04	.01	.05	.02	.05
Net wealth (logged)	−.05*	.02	−.06*	.02	−.10	.09	−.10	.09
Retired	.04	.25	.04	.25	−.04	.37	−.04	.37
Lives alone	.97***	.26	.97***	.24	1.47***	.20	1.47***	.20
Self-reported health	.26*	.04	.26*	.03	.21	.17	.21	.17
Chronic conditions	.09*	.02	.09*	.03	.10	.14	.10	.13
ADLs	.12	.19	.12	.19	.05	.21	.05	.22
IADLs	.01	.20	.01	.20	.25	.32	.25	.30
Anti-depressant medication	.05*	.01	.05*	.01	.13*	.04	.13*	.04
Smoker	.15	.38	.16	.39	.48	.49	.49	.49
Moderate/vigorous exercise regularly	−.34	.23	−.34	.24	−.37	.33	−.37	.33
Alcohol use 3+ drinks/day	−.99	.83	−.99	.83	−.31	.61	−.31	.61
High religious service attendance	−.58*	.26	−.58*	.26	−.33**	.11	−.33**	.11
Moderate religious service attendance	−.69	.36	−.69	.37	−.46*	.19	−.46*	.19
Religiosity index	−.10*	.02	−.10*	.02	−.11	.12	−.11	.12
Private prayer frequency	−.08*	.02	−.08*	.02	−.06	.08	−.06	.08
High religious service attendance/live alone interaction			−.21*	.07			−.20*	.06
Moderate attendance/live alone interaction			−.07	.08			−.11*	.03
Religiosity/live alone interaction			−.09	.11			−.15	.12
Private prayer/live alone interaction			−.13	.23			−.12	.16
Adjusted R²	.31		.31		.28		.28

*p < .05; **p < .01; ***p < .001.

Covariates measured at reported widowhood.

Dependent variable is change in number of depressive symptoms from baseline to the HRS wave year immediately following reported widowhood.

Reference groups: Non-Hispanic White, not retired, lives with others, no anti-depressant medication, non-smoker, no/light regular exercise, none or less than 3 alcoholic drinks/day, and no/low religious service attendance.

The study's control variables were analyzed, and the results indicated that living alone and taking anti-depressants were associated with increased depressive symptoms for both men and women. The increases in depressive symptoms associated with living alone were greater for men than for women but were significant for both genders. Among women, having chronic conditions, poorer health, and identifying as Hispanic were each linked to an increase in depressive symptomology. On the other hand, net wealth was associated with a decrease in depressive symptoms for women. For men, identifying as non-Hispanic Black and higher age were related to decreased depressive symptoms.

Model 2 (presented in Table 2) included interaction terms between living alone and various dimensions of religiosity, such as high and moderate religious service attendance, private prayer, and intrinsic religiosity. The aim was to investigate whether religiosity acted as a moderator in the relationship between living alone and depression for both men and women. The results indicated that only high religious service attendance was a significant moderator for women, weakening the association between living alone and depression (β = −.21, p < .05). For men, both high (β = −.20, p < .05) and moderate (β = −.11, p < .05) religious service attendance were significant moderators, also weakening the association between living alone and depression.

Table 3 presents an ordinary least squares (OLS) analysis examining the association between changes in depressive symptomology and religiosity stratified by gender in post-widowhood (from widowhood to 2018). This regression analysis takes a similar approach to that shown in Table 2, but with the change in depressive symptoms and change in covariates measured in the period post-widowhood. Model 1 (Table 3) shows that increases in religious service attendance post-widowhood were associated with a lowering of depressive symptoms (and vice versa) over the same period for both women (β = −.61, p < .01) and men (β = −.46, p < .01). Much like during the transition into widowhood, for widows, increases in intrinsic religiosity ratings (β = −.14, p < .05) and private prayer (β = −.010, p < .05) post-widowhood were associated with a lowering of depressive symptomology from widowhood to 2018. Once again, there were no significant associations found for intrinsic religiosity ratings or private prayer frequency among widowers.

Table 3.

Longitudinal OLS Regression Analysis of the Relationship Between Changes in Depressive Symptoms and Religiosity Post-widowhood for Men and Women.

Change from Widowhood to 2018	Women (N = 1356)				Men (N = 484)
	Model 1		Model 2		Model 1		Model 2
	β	SE	β	SE	β	SE	β	SE
Years widowed	−.48**	.13	−.48**	.13	−.62***	.17	−.62***	.17
Age	−.03	.02	−.03	.03	−.05*	.01	−.05*	.02
Non-Hispanic Black	.21	.24	.20	.24	−.39*	.12	−.39*	.12
Non-Hispanic other	−.41	.51	−.41	.51	−.82	.99	−.81	.99
Hispanic	1.01**	.38	1.02**	.39	−1.07	.74	−1.07	.74
Education	.04	.03	.04	.03	.02	.08	.02	.08
Net wealth (logged)	−.04*	.02	−.04*	.02	−.06*	.02	−.06*	.02
Retired	.03	.27	.03	.26	−.03	.19	−.05	.17
Re-married/partnered	−.05	.04	−.06	.04	−.09	.06	−.09	.06
Lives alone	1.01***	.18	1.01***	.18	1.62***	.21	1.63***	.22
Self-reported health	.41*	.09	.40*	.09	.31	.19	.31	.19
Chronic conditions	.08*	.03	.08*	.03	.15	.14	.15	.15
ADLs	.09	.15	.09	.16	.07	.19	.07	.19
IADLs	.02	.10	.02	.10	.12	.20	.12	.21
Anti-depressant medication	.06*	.02	.06*	.01	.10*	.03	.10*	.04
Smoker	.20	.26	.22	.25	.34	.31	.34	.31
Moderate/vigorous exercise regularly	−.25	.19	−.25	.19	−.33	.26	−.33	.26
Alcohol use 3+ drinks/day	−.71	.48	−.71	.48	−.28	.55	−.28	.55
Religious service attendance	−.61**	.17	−.60**	.16	−.46**	.15	−.46**	.14
Religiosity index	−.14*	.05	−.14*	.05	−.10	.09	−.10	.09
Private prayer frequency	−.10*	.03	−.10*	.03	−.05	.05	−.05	.05
Religious service attendance/live alone interaction			−.18*	.05			−.25*	.10
Religiosity/live alone interaction			−.04	.09			−.07	.08
Private prayer/live alone interaction			−.15	.19			−.17	.24
Adjusted R²	.32		.32		.29		.30

*p < .05; **p < .01; ***p < .001.

Covariates measured as the difference from time of widowhood to latest data point in 2018 (for those with no 2018 religion data, a 2016 data endpoint is used for religion variables), with the exception of race/ethnicity which is measured at baseline.

Dependent variable is change in number of depressive symptoms from time of widowhood to 2018.

Reference groups: Non-Hispanic White, not retired, not remarried/partnered, lives with others, no anti-depressant medication, non-smoker, no/light regular exercise, and none or less than 3 alcoholic drinks/day.

Associations between changes in depressive symptoms and the control variables during post-widowhood (Table 3, Model 1) were substantially the same as those in the regression analysis examining the transition into widowhood discussed above (from Table 2). For the post-widowhood regression analysis, we added control variables for the length of time participants were widowed as well as whether participants re-married/re-partnered. A higher number of years widowed was significantly associated with lower depressive symptoms for both women (β = −.48, p < .01) and men (β = −.62, p < .001). There were no significant findings for being re-married/re-partnered in post-widowhood.

In Model 2 (Table 3), we further investigated whether religion acted as a moderator in the relationship between living alone and depressive symptoms post-widowhood for both men and women. Results showed similar patterns as in the transition into widowhood (Table 2, Model (2)) such that religious service attendance was a significant moderator for both widows (β = −.18, p < .05) and widowers (β = −.25, p < .05), weakening the association between living alone and depression. No moderation effects were found between intrinsic religiosity or private prayer and living alone.

Discussion

Our study, in line with previous research (Luhmann et al., 2012; Stroebe et al., 2001), reveals gender differences in depression during widowhood. Our findings highlight that men experienced a significant increase in depressive symptoms during the transition to widowhood and continued to have higher depression levels compared to women even 10 years after the loss.

Our study further adds to the existing body of literature by highlighting the effectiveness of religious practices, particularly attending religious services, in coping with widowhood and reducing depression for both genders. The results indicate that high religious service attendance during widowhood was associated with a decrease in depressive symptoms for women; while for men, both high and moderate religious service attendance were related to a reduction in depressive symptomology. Our study also revealed a reduction in religious service attendance following widowhood. It's plausible that after losing a spouse, religious attendance decreased because attending services had been a shared activity. The deceased spouse may have been the primary motivator, and without them, motivation waned. Additionally, attending services may have become emotionally challenging due to condolences and reminders of the past. This loss could also prompt the surviving partner to question their faith across various religious measures. In addition, the observed differential effect between widowed women and men may be attributed to the social support and community engagement provided by high religious service attendance. It is possible that widowed women require stronger social bonds and higher levels of social engagement, which may not be adequately fulfilled by moderate religious service attendance. Interestingly, our findings suggest that men are less inclined to engage in religious practices across various dimensions of religiosity compared to women, providing support for prior research (Li et al., 2020). Specifically, men are less likely to attend religious services frequently and more likely to refrain from attending them altogether.

The differential impact of intrinsic religiosity and private prayer on depression between women and men may be attributed to various factors. Women often exhibit higher levels of involvement in religious activities (Miller & Hoffmann, 1995) and a greater sense of internalization of religious beliefs (Wink & Dillon, 2003), which may provide them with a stronger protective effect against depressive symptoms. Intrinsic religiosity, with its personal and internalized aspects of faith, may hold greater significance for women by offering meaning, support, and coping mechanisms. Women have also been shown to engage more in private prayer and experience a deeper emotional connection to their spirituality (Maselko & Kubzansky, 2006). This emotional engagement may provide women with comfort, solace, and support, leading to a reduction in depressive symptoms. Additionally, societal and cultural factors may contribute to the differential impact, as women may face unique stressors or experiences that are alleviated by the psychological benefits of private prayer.

Our regression analysis revealed that living alone was significantly associated with higher depression levels for both men and women, which is consistent with previous research (Lee & DeMaris, 2007). Healthcare providers should consider the living arrangement of older adults when assessing their risk for depression and provide targeted interventions for those who live alone. Living alone had a greater impact on men than on women.

Finally, our results showed that high religious service attendance was a significant moderator for women, weakening the association between living alone and depression. Similarly, for men, both high and moderate religious service attendance were significant moderators, also weakening the association between living alone and depression. These findings suggest that the level of religious service attendance may play a role in mitigating the impact of living alone on depressive symptoms for both men and women. The findings did not reveal significant moderating effects of private prayer and intrinsic religiosity on the relationship between living alone and depression for both men and women. This could be due to various factors. Private prayer, being a more individualistic religious practice, might not provide the same level of social support and community engagement as religious service attendance. Additionally, the influence of intrinsic religiosity, which relates to personal beliefs and values, may be more internally focused and less directly connected to the social support network that could impact the association between living alone and depression. Further research is necessary to explore these dimensions of religiosity and their potential moderating effects in the context of living alone and depressive symptoms.

Implications

The findings suggest that there may be a need for tailored interventions for men and women who are living alone and experiencing depression after the loss of a spouse. Healthcare professionals could consider promoting religious involvement, particularly religious service attendance, as a coping strategy for both genders, as it appears to be effective in reducing depression levels. However, given that men are less likely to attend religious services than women, interventions aimed at increasing men's engagement in religious practices could be especially beneficial. Additionally, the results highlight the importance of social support for individuals who are living alone after the death of a spouse, as living alone was significantly associated with higher depression levels for both genders. Healthcare professionals could consider providing support and resources for widowers and widows living alone, such as support groups or community programs, to help mitigate the negative effects of living alone on mental health.

Limitations

Our study has several limitations. First, since we conducted a secondary analysis, we were unable to examine changes in depressive symptoms at the exact time of spousal loss, but rather within a 2-year period due to the limited data collection schedule of the HRS. Similarly, the 4-year data collection interval of the religion measures compared to the 2-year data collection interval of the core HRS survey meant that for some participants we had to use religion data from the wave year prior to widowhood creating a 2-year gap. Although sensitivity analysis revealed that religion measures did not significantly change in the time prior to widowhood, it would be ideal to have a more consistent measure of religion at widowhood for all participants. Moreover, our measures of social support were limited to living alone and did not include other potentially important aspects of social support such as isolation. Although the HRS does collect additional measures, we were not able to include them due to the high amount of missing data on these items. Lastly, our findings may have limited generalizability to older adults in institutional settings, as the HRS is representative of community-dwelling older adults even though it does include those who have transitioned into institutional living during the survey.

Future Research

Future studies designed in a way that captures widowhood at the time it occurs may be important for determining the exact impact of widowhood on mental health as well as the way in which religiosity affects this relationship in real time. Future research should also replicate our study to further examine depressive symptomology among widows over time by analyzing data from other large-scale studies (e.g., National Health and Nutrition Examination Survey and Midlife in The United States). Relatedly, although we were able to examine three dimensions of religiosity using the HRS, there are other datasets that may capture more extensive measures of religiosity (e.g., National Survey of American Life). Future research could also include older adults in a variety of settings (e.g., assisted living and senior housing) to better understand whether similar patterns can be observed in older adults residing in congregative living where they are living/regularly interacting with others.

Conclusion

The present findings add to the existing evidence that widowhood elevates the likelihood of depression in older adults, especially in men. Living alone has a more substantial detrimental effect on men than women following spousal loss. Our results demonstrate that religiosity, especially religious service attendance, is a useful coping strategy for both men and women to manage widowhood-related depression. Nonetheless, men are significantly less religious than women across all religiosity dimensions evaluated. These results could assist healthcare professionals in providing more specific assistance to widowers, who are at a greater risk of depression post-widowhood.

Footnotes

Author Contributions

All authors contributed to the study conception and design. Data analyses were performed by Dr. Jane Tavares. The draft of the manuscript was written by Dr. Frances Hawes and all authors commented on versions of the manuscript. All authors read and approved the final manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the College Business Summer Research Funding University of Wisconsin Eau Claire.

Ethical Statement

ORCID iD

Frances M. Hawes

References

Bisconti

T. L.

Bergeman

C. S.

Boker

S. M.

(2004). Emotional well-being in recently bereaved widows: A dynamical systems approach. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 59(4), 158–167. https://doi.org/10.1093/geronb/59.4.P158

Cacioppo

J. T.

Hughes

M. E.

Waite

L. J.

Hawkley

L. C.

Thisted

R. A.

(2006). Loneliness as a specific risk factor for depressive symptoms: Cross-sectional and longitudinal analyses. Psychology and Aging, 21(1), 140.

Chou

K. L.

Chi

(2000). Stressful events and depressive symptoms among old women and men: A longitudinal study. The International Journal of Aging and Human Development, 51(4), 275–293. https://doi.org/10.2190/VNJC-TQ4W-6T3R-6V9K

Dean

Kolody

Wood

Matt

G. E.

(1992). The influence of living alone on depression in elderly persons. Journal of Aging and Health, 4(1), 3–18. https://doi.org/10.1177/089826439200400101

Garcini

L. M.

Brown

R. L.

Chen

M. A.

Saucedo

Fite

A. M.

Fagundes

C. P.

(2021). Bereavement among widowed latinos in the United States: A systematic review of methodology and findings. Death Studies, 45(5), 342–353. https://doi.org/10.1080/07481187.2019

Guo

Mei

Wang

(2021). Widowhood and health status among Chinese older adults: The mediation effects of different types of support. Frontiers in Public Health, 9, Article 745073. https://doi.org/10.3389/fpubh.2021.745073

Hawes

F. M.

Tavares

J. L.

Ronneberg

C. R.

Miller

E. A.

(2022). The effects of religiosity on depression trajectories after widowhood. OMEGA-Journal of Death and Dying, Online ahead of print. https://doi.org/10.1177/00302228211051509

Jeon

G. S.

Choi

Cho

S. I.

(2017). Impact of living alone on depressive symptoms in older Korean widows. International Journal of Environmental Research and Public Health, 14(10), 1191. https://doi.org/10.3390/ijerph14101191

Kristiansen

C. B.

Bojesen

A. B.

Munk-Jørgensen

Andersen

(2017). The widowhood effect–mortality and adverse health effects when losing a spouse in old age. European Psychiatry, 41(S1), S656–S656. https://doi.org/10.1016/j.eurpsy.2017.01.1101

10.

Kristiansen

C. B.

Kjær

J. N.

Hjorth

Andersen

Prina

A. M.

(2019). The association of time since spousal loss and depression in widowhood: A systematic review and meta-analysis. Social Psychiatry and Psychiatric Epidemiology, 54(7), 781–792. https://doi.org/10.1007/s00127-019-01680-3

11.

Lee

G. R.

DeMaris

(2007). Widowhood, gender, and depression: A longitudinal analysis. Research on Aging, 29(1), 56–72. https://doi.org/10.1177/0164027506294098

12.

Y. I.

Woodberry

Liu

Guo

(2020). Why are women more religious than men? Do risk preferences and genetic risk predispositions explain the gender gap? Journal for the Scientific Study of Religion, 59(2), 289–310. https://doi.org/10.1111/jssr.12657

13.

Luhmann

Hofmann

Eid

Lucas

R. E.

(2012). Subjective well-being and adaptation to life events: A meta-analysis. Journal of Personality and Social Psychology, 102(3), 592. https://doi.org/10.1037/a0025948

14.

Lyu

Min

Kim

(2019). Trajectories of cognitive decline by widowhood status among Korean older adults. International Journal of Geriatric Psychiatry, 34(11), 1582–1589. https://doi.org/10.1002/gps.5168

15.

Maselko

Kubzansky

L. D.

(2006). Gender differences in religious practices, spiritual experiences and health: Results from the US General Social Survey. Social Science and Medicine, 62(11), 2848–2860. https://doi.org/10.1016/j.socscimed.2005.11.008

16.

Miller

A. S.

Hoffmann

J. P.

(1995). Risk and religion: An explanation of gender differences in religiosity. Journal for the Scientific Study of Religion, 34(1), 63–75. https://www.jstor.org/stable/1386523

17.

Radloff

L. S.

(1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401.

18.

Recksiedler

Stawski

R. S.

(2019). Marital transitions and depressive symptoms among older adults: Examining educational differences. Gerontology, 65(4), 407–418. https://doi.org/10.1159/000493681

19.

Sasson

Umberson

D. J.

(2014). Widowhood and depression: New light on gender differences, selection, and psychological adjustment. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 69(1), 135–145. https://doi.org/10.1093/geronb/gbt058

20.

Srivastava

Debnath

Shri

Muhammad

(2021). The association of widowhood and living alone with depression among older adults in India. Scientific Reports, 11(1), 1–13. https://doi.org/10.1038/s41598-021-01238-x

21.

Stolzenberg

R. M.

Waite

L. J.

(2006) Effects of marriage, divorce, and widowhood on health. In Work, family, health, and well-being (pp. 355–371), Routledge. https://doi.org/10.4324/9781410613523-35

22.

Stroebe

Schut

(2001). Gender differences in adjustment to bereavement: An empirical and theoretical review. Review of General Psychology, 5(1), 62–83. https://doi.org/10.1037/1089-2680.5.1.62

23.

Utz

R. L.

Swenson

K. L.

Caserta

Lund

DeVries

(2014). Feeling lonely versus being alone: Loneliness and social support among recently bereaved persons. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 69(1), 85–94.

24.

van Grootheest

D. S.

Beekman

A. T.

Van Groenou

M. B.

Deeg

D. J.

(1999). Sex differences in depression after widowhood. Do men suffer more? Social Psychiatry and Psychiatric Epidemiology, 34(7), 391–398. https://doi.org/10.1007/s001270050160

25.

Wink

Dillon

(2003). Religiousness, spirituality, and psychosocial functioning in late adulthood: Findings from a longitudinal study. Psychology and Aging, 18(4), 916. https://doi.org/10.1037/0882-7974.18.4.916

26.

Yang

Kozloski

(2012). Change of sex gaps in total and cause-specific mortality over the life span in the United States. Annals of Epidemiology, 22(2), 94–103. https://doi.org/10.1016/j.annepidem.2011.06.006

27.

Yang

Y. C.

Boen

Mullan Harris

(2015). Social relationships and hypertension in late life: Evidence from a nationally representative longitudinal study of older adults. Journal of Aging and Health, 27(3), 403–431. https://doi.org/10.1177/0898264314551172