Objectives: To determine if marital status is associated with depressive symptoms in men and women, and to determine if partner satisfaction is associated with depressive symptoms.
Methods: Community-dwelling adults (n = 1751) aged 65 years and older were sampled from a representative population-based registry. Age, sex, education, and marital status were self-reported. Satisfaction with living partner was assessed with an item from the terrible-delightful scale. The Center for Epidemiologic Studies-Depression was used to assess depressive symptoms, with a score of 15 or more indicating depressive symptoms.
Results: Only 3% of the sample were dissatisfied with their living partner, and those who were dissatisfied with their living partner had higher levels of depressive symptoms. Proportions of men with depressive symptoms were: 20.6% for never married; 19.2% for separated or divorced; 17.3% for widowed; 7.3% for married (satisfied); and 38.1% for married (dissatisfied) (χ^sup 2^ = 32.98, df = 4, P < 0.001). In women, the results were different: 11.8% for never married; 23.1% for separated or divorced; 15.4% for widowed; 14.4% for married (satisfied); and 41.9% for married (dissatisfied) (χ^sup 2^ = 18.33, df = 4, P < 0.001). In logistic regression models, a significant interaction term was seen for Marital status � Sex. In stratified models, adjusting for age and education, an effect of marital status was seen in men, but not in women.
Conclusions: The effects of marriage on depressive symptoms in men and women may be different, with lower levels of depressive symptoms in married men, compared with unmarried men, but similar rates in married women, compared with unmarried women. Dissatisfaction with their partner was uncommon, but highly associated with depressive symptoms.
Can J Psychiatry. 2009;54(7):487-492.
Clinical Implications
* Married men had lower rates of depressive symptoms than unmarried men, but married women did not have lower rates of depressive symptoms than unmarried women.
* Men and women who were dissatisfied with their partner had high rates of depressive symptoms.
* Clinicians should be alert to depressive symptoms in unmarried men, and researchers should stratify analyses of depression on marital status and sex.
Limitations
* This was a cross-sectional survey, rather than prospective data, making causal inferences difficult.
* It was not possible to adjust for some known confounding variables (such as functional status) owing to small numbers in some categories.
* The measure of satisfaction with one's partner was very crude.
Key Words: sex, gender, marital status, marriage, depression, depressive symptoms
Abbreviations used in this article
CES-D Center for Epidemiologic Studies-Depression
MSHA Manitoba Study of Health and Aging
Marriage may confer health benefits for people. Previous studies have shown that married people often have less risky health behaviours1 and lower mortality rates than unmarried people.2-4 Some attribute this protective effect to a beneficial effect of marriage on mental health. Indeed, some researchers go as far as to claim that unmarried people may the of a broken heart.5 The effects of marital status on depression have been studied in clinical samples and in population-based studies. However, many previous reports of marital status and depression have often not stratified on sex, and the effects of marriage on mood may be different for men and women. This is particularly relevant in older cohorts, where marital duration, roles, and responsibilities were different than in younger cohorts. Married women in older cohorts may have had less opportunity to pursue higher education and careers because there was an expectation of maintaining a household and family. For men, being married was unlikely to have resulted in less educational or occupational opportunities. There are few Canadian studies of marital status and depression. In one recent population-based survey of depression in Canada, Patten et al6 found that depression was more prevalent in unmarried men than married men, while it was less prevalent in unmarried women than married women. However, this study was in adults of all ages, and not specific to older adults. In other countries, widowed and single people may have a higher risk of depression than married people,7 but the risk with widowhood may decrease with time.8
Satisfaction with a living partner could also be associated with depression. There are relatively few population-based studies of marital quality and depression in older adults. People in unhappy marriages or who are not satisfied with their partner may have higher levels of depressive symptoms, as has been shown in some studies of younger adults,9 and in older adults.10'11 Conversely, those with depression may be at a higher risk of poor marital relations.12 We therefore conducted a secondary analysis of an existing population-based epidemiologic study. The objectives were: to determine if marital status is associated with depressive symptoms in men and women, and if there is an interaction between marital status and sex in this effect; and to determine if satisfaction with their partner is associated with depressive symptoms in men and women.
Methods
We conducted a secondary analysis of the MSHA, an epidemiologic study conducted in Manitoba in conjunction with the Canadian Study of Health and Aging.13 During 1991-1992, 1751 adults aged 65 years and older were interviewed in their homes by trained interviewers. The original sampling frame, which was comprehensive and representative, was from a list provided by Manitoba Health. Older groups were oversampled. People residing in institutions (nursing homes, hospitals, and prisons) were not included in these analyses. Initially, 2890 people were selected. Among these, 443 refused to participate, 480 were not eligible (had died, had entered a nursing home, or were too ill), 162 could not be located, and 54 did not complete the screening questionnaire. This left a sample of 1 75 1 participants. The MSHA received ethics approval from the Research Ethics Committee of the University of Manitoba, and adhered to the Helsinki Declaration. Informed consent was obtained from the participants or from the appropriate proxy.
Measurements
Age, sex, education, and marital status were all self-reported. For the purpose of these analyses, marital status was categorized as never married, currently married, widowed, separated, or divorced. An item regarding satisfaction with current living partner was taken from the terrible-delightful scale.14'15 This is a scale that measures satisfaction with various life circumstances. The participant was asked to rate their satisfaction with their current living partner on a 7-point Likert-type scale from terrible to delightful. Responses of mixed, dissatisfying, very dissatisfying, and terrible were categorized as dissatisfied, while those who rated their satisfaction with their current partner as being delightful, very satisfying, or satisfying were considered to be satisfied with their partner.
The CES-D scale16 was used to measure depressive symptoms. This is a reliable, valid measure of depressive symptoms in older adults. However, it is not a diagnostic tool for major depression. It consists of 20 items, each scored from 0 to 3, giving a possible score from 0 to 60. Participants with a score of 16 or greater were considered to have depressive symptoms. In sensitivity analyses, we considered the entire score as a continuous variable.
Analysis
Student t tests (assuming unequal variance) and analysis of variance were used to compare continuous variables. Chi-square tests were used for categorical variables. The outcome variable was the dichotomized CES-D score. Stratified analysis and logistic regression models were used to determine the effects of marital status and sex on depressive symptoms. Analyses stratified on sex were conducted. As well, a logistic regression model with the main effects terms marital status, sex, and the interaction term of Marital status ? Sex was constructed. As this interaction term was highly significant, all analyses were stratified on sex, and separate logistic regression models were constructed for men and women. Regression diagnostics were performed. AU analyses were performed using SPSS software, version 11.5 (SPSS Ine, Chicago, IL).
Results
Characteristics of the sample are shown in Table 1. Women were more likely than men to be widowed, while men were more likely to be married. Women who had never married were more educated and older than those who were married. Conversely, men who had never married were younger and less educated than married men. Most people were satisfied with their living partner: 95% of married or common-law men, and 89.2% of married or common-law women reported being satisfied with their partner.
Overall, married people who were dissatisfied with their partner were most likely to have depressive symptoms (Figure 1 ). People who were divorced or separated were also more likely to experience depressive symptoms. This was true both in men and in women. However, there was a difference in depressive symptoms in married and never-married men and women. Married men were less likely to experience depressive symptoms, compared with their never married counterparts. In women, marital status was not associated with depressive symptoms.
Regression models with interaction terms showed a large and statistically significant interaction between sex and marital status. Therefore, separate models for men and for women were constructed (Table 2). In men, the effect of marital status on depressive symptoms persisted, even after accounting for age (Model 1), and age and education (Model 2); married men were less likely to have depressive symptoms than their never-married counterparts. In women, being dissatisfied with their living partner was strongly associated with depressive symptoms after accounting for age (Model 1) and age and education (Model T). As living arrangements (living alone, compared with not living alone) was collinear both with marital status and with sex, we did not include living arrangements in the logistic regression models. Sensitivity analyses considering the CES-D as a continuous, rather than dichotomous, score gave similar results.
Discussion
In this sample, marital status is associated with depressive symptoms in men but not in women. Specifically, married men were less likely to have depressive symptoms than unmarried men. As well, being dissatisfied with one's living partner is associated with depressive symptoms in women.
Marriage has been reported to confer health advantages to people. Others have reported that married people have less risky health behaviours,1 and better survival than unmarried people.2"4 There may also be differences in depression, as has been demonstrated in younger Canadians.6 However, it is possible that the effect of marriage is different in men and women. This may be for several reasons; there may be marital selection. Unmarried women in this cohort may have pursued careers and (or) education which may have positively influenced their health throughout life. Indeed, the unmarried women in this cohort had higher educational levels than married women. Conversely, men who had never married may have had characteristics which impeded marriage. A second possible explanation is that marriage itself has effects on mood, which are protective for depression in men but not in women.
In the multivariable models, widowers had a similar risk for depressive symptoms, compared with other groups. In a pan-European study, it was found that unmarried people had higher rates of depression than other groups,7 but the effect was fairly small. In addition, the effect of widowhood may lessen over time.8 We do not have longitudinal data, and so are unable to follow the temporal course of depressive symptoms with widowhood or separation. Others have reported that widowhood is associated with depressive symptoms in women more than men, and that this effect is due to the financial strain associated with losing one's spouse (and their income).17 We were unable to examine the effect of financial strain on depressive symptoms in this study.
Strengths and Limitations
There are numerous limitations to this analysis. First, this was a cross-sectional analysis. While the MSHA is a cohort study, the number of people surviving to the follow-up assessment did not allow for a prospective analysis (as many cells did not contain 5 or more people). Thus the temporal effect of widowhood and separation could not be studied. However, population-based longitudinal studies over the entire course of adult life are rare. Second, the numbers in many of the categories were small. Notably, there were very few separated and divorced people in the sample, limiting the inferences that can be drawn in these groups. The small sample size also did not allow us to control for some known potential confounding factors: in particular, functional impairment and cognitive impairment are associated with depressive symptoms. While activities of daily living and cognitive status were measured in the MSHA, there were very few people with disabilities in some of the marital status categories. We therefore could not adjust for these factors. The third limitation is the measurement of marital status and satisfaction. We crudely grouped people into marital status categories, likely oversimplifying the complex effects of marriage. Further, the measure of partner satisfaction was even more limited, consisting only of a single item drawn from a larger scale. Fourth, we considered only current marital status. It is possible that some participants had been married on several occasions, with different effects in different marriages throughout their life. Another final limitation is our measure of depressive symptoms. While the CES-D is a valid, reliable measure of depressive symptoms, it is not a measure of major depression. As such, it is difficult to draw inferences about the prevalence of major depression in the various marital status groups.
There are also strengths to our study. First, it is a large representative sample consisting of the entire province of Manitoba, except for extremely remote areas. It contains a large urban and a large rural sample. Second, the sampling frame was extensive and representative. Third, the measure of depression was a reliable, valid measure of depressive symptoms that has been widely used in other studies.
Conclusions
Being married appears to be associated with a lower risk of depressive symptoms in men, but not in women. It is not clear if this effect is limited to this cohort of older adults, where marital roles and responsibilities are different than in younger cohorts. As these younger cohorts age, it will be relevant to study the effects of marital status on depression over their life course. As well, further study is needed to verify these findings in other settings. It is possible that the effects we observe for depressive symptoms are limited to older Canadian adults, and may not be true in other settings.
Our findings have implications for clinicians and health policy-makers. Older men who are not married should be considered carefully for evidence of depression. This group has a particularly high rate of depressive symptoms. Socially isolated men are also at an increased risk of suicide, and it is possible that attention to depressive symptoms in this group may lessen this risk.18,19 Health policy-makers should consider unmarried men in planning and delivering health promotion and prevention activities for depression.
Funding and Support
The MSHA was funded primarily by Manitoba Health, with additional funding provided through the Canadian Study of Health and Aging and by the Seniors Independence Research Program of the National Health Research and Development Program of Health Canada. The second wave of the MSHA was funded primarily by Manitoba Health's Healthy Communities Development Fund, with additional funding provided through the Canadian Study of Health and Aging and by the Seniors Independence Research Program of the National Health Research and Development Program of Health Canada (Project 6606-3954-MC[S]). The results and conclusions are those of the authors and no official endorsement by Manitoba Health or other funding agencies is intended or should be inferred.
[Sidebar]
R�sum� : L'�tat matrimonial, Ia satisfaction � l'�gard du partenaire, et les sympt�mes d�pressifs chez les femmes et les hommes �g�s
Objectifs : D�terminer si l'�tat matrimonial est associ� aux sympt�mes d�pressifs chez les femmes et les hommes, et d�terminer si la satisfaction � l'�gard du partenaire est associ�e aux sympt�mes d�pressifs.
M�thodes : Des adultes (n = 1751) r�sidant dans la communaut� �g�s de 65 ans et plus ont �t� �chantillonn�s � partir d'un registre repr�sentatif de la population. L'�ge, le sexe, le niveau d'instruction, et l'�tat matrimonial �taient auto-d�clar�s. La satisfaction � l'�gard du partenaire de vie a �t� �valu�e � l'aide d'un item de l'�chelle terrible-enchanteur. L'�chelle du Centre d'�tudes �pid�miologiques-D�pression (CES-D) a �t� utilis�e pour �valuer les sympt�mes d�pressifs, et � un score de 15 ou plus, les sympt�mes d�pressifs �taient consid�r�s.
R�sultats : Seulement 3 % de l'�chantillon �taient insatisfaits � l'�gard de leur partenaire de vie, et ceux qui �taient insatisfaits � l'�gard de leur partenaire de vie avaient des niveaux plus �lev�s de sympt�mes d�pressifs. Les proportions d'hommes pr�sentant des sympt�mes d�pressifs �taient : 20,6 % chez les jamais mari�s; 19,2 % chez les s�par�s ou divorc�s; 17,3 % chez les veufs; 7,3 % chez les mari�s (satisfaits); et 38,1 % chez les mari�s (insatisfaits) (χ^sup 2^ = 32,98, dl = 4, P< 0,001). Chez les femmes, les r�sultats �taient diff�rents : 1 1,8 % chez les jamais mari�es; 23,1 % chez les s�par�es ou divorc�es; 15,4 % chez les veuves; 14,4 % chez les mari�es (satisfaites); et 41,9 % chez les mari�es (insatisfaites) (χ^sup 2^ = 18,33, dl = 4, P < 0,001). Dans les mod�les de r�gression logistique, un terme d'interaction significatif pour �tat matrimonial x sexe a �t� observ�. Dans les mod�les stratifi�s, apr�s ajustement de l'�ge et du niveau d'instruction, un effet de l'�tat matrimonial se voyait chez les hommes, mais pas chez les femmes.
Conclusions : Les effets du mariage sur les sympt�mes d�pressifs des hommes et des femmes peuvent �tre diff�rents; les hommes mari�s ont des taux plus faibles de sympt�mes d�pressifs, compar� aux hommes non mari�s, mais les taux sont semblables chez les femmes mari�es et les femmes non mari�es. L'insatisfaction � l'�gard du partenaire de vie n'�tait pas fr�quente, mais hautement associ�e aux sympt�mes d�pressifs.
[Reference]
References
1. Joung IM, Stronks K, van de MH, et al. Health behaviours explain part of the differences in self reported health associated with partner/marital status in the Netherlands. J Epidemiol Community Health. 1995;49(5):482-488.
2. Ben-Shlomo Y, Smith GD, Shipley M, et al. Magnitude and causes of mortality differences between married and unmarried men. J Epidemiol Community Health. 1993;47(3):200-205.
3. Manzoli L, Villari P, Pirone M, et al. Marital status and mortality in the elderly: a systematic review and meta-analysis. Soc Sci Med. 2007;64(l):77-94. Epub 2006 Sept 29.
4. Ebrahim S, Wannamethee G, McCallum A, et al. Marital status, change in marital status, and mortality in middle-aged British men. Am J Epidemiol. 1995;142(8):834-842.
5. Kaplan RM, Kronick RG. Marital status and longevity in the United States population. J Epidemiol Community Health. 2006;60(9):760-765.
6. Patten SB, Wang JL, Williams JV, et al. Descriptive epidemiology of major depression in Canada. Can J Psychiatry. 2006;51(2):84-90.
7. Prince MJ, Beekman AT, Deeg DJ, et al. Depression symptoms in late life assessed using the EURO-D scale. Effect of age, gender and marital status in 14 European centres. Br J Psychiatry. 1999;174:339-345.
8. Harlow SD, Goldberg EL, Comstock GW. A longitudinal study of the prevalence of depressive symptomatology in elderly widowed and married women. Arch Gen Psychiatry. 1991;48(12):1065-1068.
9. Balog P, Janszky I, Leineweber C, et al. Depressive symptoms in relation to marital and work stress in women with and without coronary heart disease. The Stockholm Female Coronary Risk Study. J Psychosom Res. 2003;54(2): 113-119.
10. Bookwala J, Jacobs J. Age, marital processes, and depressed affect. Gerontologist. 2004;44(3):328-338.
11. Bookwala J. The role of marital quality in physical health during the mature years. J Aging Health. 2005;17(1):85-104.
12. Kivela SL, Luukinen H, Sulkava R, et al. Marital and family relations and depression in married elderly Finns. J Affect Disord. 1999;54(1- 2): 177- 182.
13. Canadian Study of Health and Aging Working Group. Canadian study of health and aging: study methods and prevalence of dementia. CMAJ. 1994;150(6):899-913.
14. Andrews FM, Withey FR. Social Indicators of well-being. New York (NY): Plenum Press; 1976.
15. Michalos AC. Satisfaction and happiness. Soc Indie Res. 1980;8:385-422.
16. Radloff L. The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement. 1977;1:385-401.
17. Umberson D, Wormian CB, Kessler RC. Widowhood and depression: explaining long-term gender differences in vulnerability. J Health Soc Behav. 1992;33(l):10-24.
18. Bruce ML, Ten Have TR, Reynolds CF III, et al. Reducing suicidal ideation and depressive symptoms in depressed older primary care patients: a randomized controlled trial. JAMA. 2004;291(9):1081-1091.
19. Szanto K, Kalmar S, Hendin H, et al. A suicide prevention program in a region with a very high suicide rate. Arch Gen Psychiatry. 2007;64(8):914-920.
[Author Affiliation]
Philip Donald St John, MD, MPH, FRCPC1; Patrick Roy Montgomery, MD, FRCPC2
[Author Affiliation]
Manuscript received March 2008, revised, and accepted July 2008.
This research was presented at the Gerontological Society of America meeting in November 2007 in San Francisco, California.
1 Assistant Professor and Acting Head, Section of Geriatric Medicine, Department of Medicine, Centre on Aging, University of Manitoba, Winnipeg, Manitoba.
2 Associate Professor, Section of Geriatric Medicine, Department of Medicine, Centre on Aging, University of Manitoba, St Boniface General Hospital, Winnipeg, Manitoba.
Address for correspondence: Dr PD St John, Section of Geriatrics, Department of Medicine, and the Centre on Aging, University of Manitoba, GE 545 Health Sciences Centre, 820 Sherbrook Street, Winnipeg, MB R3A 1R9; pstjohn@hsc.mb.ca
