This article has Open Peer Review reports available.
The health impact of remarriage behavior on chronic obstructive pulmonary disease: findings from the US longitudinal survey
- Tatsuya Noda†1,
- Toshiyuki Ojima1,
- Shinya Hayasaka1,
- Akihito Hagihara†2Email author,
- Ryoichi Takayanagi3 and
- Koichi Nobutomo4
© Noda et al; licensee BioMed Central Ltd. 2009
Received: 22 May 2009
Accepted: 14 November 2009
Published: 14 November 2009
Chronic obstructive pulmonary disease (COPD) is a major disease among adults, and its deterioration was reported to be associated with psychological imbalance. Meanwhile, bereavement and divorce have proven harmful to the health status of a surviving spouse. But few studies have been conducted to evaluate the remedial effect on survivors' health outcome by remarriage after bereavement. The present study thus examined the associations between remarriage and the onset of COPD.
Our cohort was drawn from Health and Retirement Study participants in the United States, and consisted of 2676 subjects who were divorced or bereaved from 1992 to 2002. We then followed them for up to 11 years and assessed the incidence rate of COPD using a Cox proportional hazard model after adjusting for marital status, age, gender, education and the number of cigarettes smoked.
Among all subjects, 224 who remarried after bereavement or divorce tended to be younger and more male dominated. Remarriage after bereavement/divorce was associated with significantly decreased risk of COPD onset for overall subjects [hazard ratio (HR): 0.51, 95% confidence interval (95% CI): 0.28-0.94], female subjects [HR: 0.36, 95% CI: 0.13-0.98], and for those under 70 years old [HR: 0.36, 95% CI: 0.17-0.79].
This study investigates the impact of remarriage on health outcome based on a large-scale population survey and indicates that remarriage significantly correlates with reduced risk of COPD incidence, even after adjusting smoking habit.
Losing a family member is one of the most serious life transitions for many adults. Studies on the health impact of bereavement have been conducted from the nineteenth century and still remain an important issue [1–7]. The harmful effect of bereavement on a marriage partner is almost certain from previous nationwide studies [8, 9]. Although the bereavement effect is more obvious among younger than older people, more than 600,000 adults are widowed each year in the United States and the bereavement effect in later years brings about a growing impact upon aging society [10, 11]. Spousal loss in later life is irreparable and unavoidable unless one divorces or dies before a partner, so it is meaningful to seek for modifiable social factors which might soften the impact of late life spousal bereavement. In this study, we focused on remarriage after bereavement as a possible rescue factor against the bereavement effect because remarriage is a joyful event in general and remarriage is possible after the onset of bereavement.
From the viewpoint of spousal loss, divorce or separation among couples also seems to be an important event in later life. In the United States, more than 1.1 million people over 50 have been divorced or separated, and it is about 6 times greater in those who have been widowed Although the magnitude and direction of the impact by divorce varies with the individual, for those shocked by divorce, remarriage is a possible relief event. But few studies have been conducted on the relationship between remarriage after divorce and the alteration of health statistics, and it remains unclear whether remarriage would make the health of those divorced take a turn for the better [12, 13].
Several studies have examined the effect of remarriage on health status [14–17]. While two longitudinal studies suggested that remarriage might reduce mortality after bereavement, another suggested that remarriage after divorce was not associated with health status [4, 18, 19]. To our knowledge, no study has been found which focuses on the association between remarriage and incidence rates of particular diseases.
In the present study, we examined the relationship between remarriage after bereavement and the onset of disease, especially chronic obstructive pulmonary disease (COPD) which is known to have a large negative impact on health-related quality of life in the elderly [20, 21]. Although smoking habits are well known as a major risk factor of COPD, functional impairment in this disease occurs in correlation with anxiety and depression in the older adults . It remains possible that the psychological burden in pre-COPD patients aggravates their respiratory function and acts as a trigger for COPD onset.
Most previous studies on bereavement and mortality were limited because they were based on demographic data and therefore did not take into account the subjects' health status and health-related behavior. It is likely that people with unhealthy behavior are less likely to remarry than healthy people . In this study, we succeeded both in excluding those who were already COPD patients from our study population and in including subjects' health behavior (i.e. their smoking status) into our study model.
The Health and Retirement Study (HRS) is a nationally representative longitudinal study sponsored by the National Institute on Aging . The HRS sample is selected using a multi-stage area probability sample design. The HRS consisted of two cohorts. The original HRS cohort began in 1992 noninstitutionalized adults in the 48 states of the United States who were born during 1931-1941, with oversampling of blacks and Hispanics and Florida residents. The other cohort is Asset and Health Dynamics Among the Oldest Old cohort started in 1993 with adults born in 1923 or earlier. If a subject married, a new spouse was included in the cohort. In both cohorts, a survey has been conducted every 2 or 3 years (usually biennially). As a consequence, HRS included 30888 participants from 1992 to 2004. All respondents provided oral informed consent, and internal ethics review board approval was obtained. More detailed information regarding the HRS design can be available in the special feature paper . After applying for registration as a data user and obtaining permission from the Health and Retirement Study group, we conducted analyses with the use of publicly available data files.
From this HRS cohort, we selected 4100 subjects who experienced divorce or bereavement and then excluded 302 subjects who dropped out of the HRS shortly after they divorced or were bereaved, 522 subjects who did not answer all the questions required and 600 subjects who had already developed COPD at study entry. Thus, we eventually identified 2676 subjects for the analysis.
Outcome and exposures
While subjects who would have never remarried until censoring or the end of follow-up were included in the not-remarried group, subjects who remarried in the follow-up period entered the remarried group. Subjects were censored when they died or were lost to follow-up (in both groups), or when they were widowed by or divorced from their partner (in the remarried group). As each survey of the HRS was performed every two or three years, the midpoint between the two consecutive surveys was set as the time of event occurrence.
Our primary endpoint was the incidence rate of COPD. Marital status after divorce/bereavement, age at study entry, gender, race/ethnicity (white/Caucasian, or other), education (none/basic/secondary, or high school and higher), and the number of cigarettes smoked at study entry were included in the study model as potential confounding variables. Age at study entry was treated as a continuous variable, and the number of cigarettes was classified into 3 categories (non-smoker, 1-20 cigarettes per day and over 20 cigarettes per day), while the remaining variables were modeled as binary covariates.
The HRS questionnaire asked respondents whether a doctor had ever told them that they had chronic lung disease such as chronic bronchitis or emphysema. Although precise definitions of COPD vary, COPD most commonly refers to patients with chronic bronchitis and emphysema  We thus determined the onset of COPD based on the questionnaire noted above. The HRS questionnaire, in addition, equated a common-law partnership with official marriage.
Significance tests were performed for age at entry and all other variables using Student's t test and chi-square tests, respectively. The Cox proportional hazard model was used to calculate the hazard ratios of COPD onset. To test for parallelism, the proportional assumption was graphically verified by plotting log [-log (survivor function)] against time in groups identified by each covariate. Statistical analyses were performed using SAS version 9.1.3.
Baseline characteristics of subjects
(n = 224)
(n = 2452)
Age at study entry
mean age (± SD)
61.8 (± 9.7)
69.9 (± 11.6)
high school or higher
Number of cigarettes smoked at study entry
> 20 cigarettes/day
COPD onset during follow-up
Relative morbidity hazard estimates; overall, stratified by sex and the age group
By age group
(n = 2676)
(n = 871)
(n = 1806)
< 70 yo
(n = 1323)
≥ 70 yo
(n = 1354)
Marital status after divorce/bereavement
Age at study entry
high school or higher
Number of cigarettes at study entry
In stratified analysis, female subjects and under 70-year-old subjects manifested a significant difference in the incidence of COPD between those who remarried and those who remained single (hazard ratios: 0.36 and 0.36, respectively). In these subject groups, white/Caucasian race, lower education level and smoking more cigarettes had a significant increase of hazard ratios. Smoking cigarettes also showed a soaring incidence rate in subjects 70 or older.
The present study revealed that remarriage after bereavement or separation from one's spouse correlated with a reduced incidence rate of COPD. Our study model contained the number of cigarettes smoked for evaluating the impact of the remarriage effect in a more direct manner. And we included not only bereaved people but also divorced/separated people. Divorce/separation is difficult to confirm from death records or other national health statistics, and hence it has seldom been taken into account in previous studies. In addition, the marriage status variable was based upon de facto marital status, not legal status. Since marital status sometimes changes without changing legal status, our study model would be reflected in the actual realities of subject's marital situations.
In this study, white/Caucasian race, lower education level and smoking more cigarettes also significantly correlated with higher incidence of COPD. Previous studies revealed that smoking was a dominant risk factor for COPD and white/Caucasian race was prone to this disease because of a genetic susceptibility . In addition, lower education level was also related to the high prevalence of COPD . These results of earlier studies were consistent with the results of this study.
While the prevalence of COPD in this study was 14.6% (600/4100) just before the study entry, the estimated prevalence of COPD based on well-designed epidemiologic studies was in the range of 4% to 10% among people aged over 15 years . In national statistics, the estimated prevalence of self-reported COPD was about 10% among people over 65 years . Considering the majority of HRS participants are older adults, it would be safe to say the prevalence in this study was close to a reasonable range and indicated the external validity of this study to a certain extent. In addition, it could not be denied that misclassification might occur because the prevalence estimates of COPD based on self-reports. However, epidemiological surveillance of COPD was often conducted not only by physiologic pulmonary function tests (e.g. spirometry) but also by patient reports based on the diagnosis by attending physician, we, therefore, used self-reports in this study .
In regard to the association between marital status after bereavement and the incidence of diseases (e.g. COPD), two hypotheses are widely accepted in the bereavement studies: selection mechanisms hypothesis and social causation mechanisms hypothesis . The selection mechanisms hypothesis proposes that the relatively good health status of married persons is the result of the selection of "healthy" persons into and "unhealthy" persons out of the married state, thus the relative amount of unhealthy person increases in the unmarried group. Applying this hypothesis to our study, bereaved or widowed people who were heavy smoker or have poor respiratory health tended not to remarry and, at the same time, are prone to COPD. On the other hand, the social causation mechanisms hypothesis advocates that marriage itself has a health promoting or a health protective effect while being unmarried would have adverse health effect. In this hypothesis, the effect of marital status on health is assumed to be intermediated by psychosocial factors (e.g. grief), the change of socio-economic status (e.g. income, burden of housework) and the alteration of health behaviors (e.g. smoking cessation). According to this hypothesis, remarried people after bereavement or widowhood tended not to develop COPD due to relief of stress, smoking cessation, and so on.
While two hypotheses are not mutually exclusive, it seems reasonable to suppose that social causation mechanisms hypothesis is more credible in this study. Because there was no significant difference of smoking rates between the unmarried and remarried group and the number of cigarette was included as a covariate in the model, we, therefore, considered that the influence of smoking which was a dominant confounding factor on the selection mechanisms hypothesis was not so great in this study. Although it is difficult to provide concrete descriptions concerning social causation mechanisms hypothesis on COPD, we assumed that changing health behavior after remarriage is the most probable cause. Remarriage after bereavement or widowhood may bring smoking rates lower because smokers who remarry after bereavement tend to quit smoking with consideration for or request from their new partner. This leads to tobacco cessation in later life which may slow the progression of airway damage and hamper a 'final push' to the onset of COPD. Whereas the development of COPD is the result of life-long progression of pulmonary damage, it was previously reported that stress itself was a risk factor of worsened pulmonary obstruction [22, 31]. It is, therefore, appropriate to adopt the social causation mechanisms as an explanation for the results in this study.
This study has some limitations. The HRS has been conducted every 2 or 3 years and, therefore, we could not identify the subjects who developed COPD within 2 or 3 years since they divorced or were bereaved. Second, the majority of subjects in this study were retirement age and so the remarriage effect in middle age and youth remains unclear. Third, some confounding factors remain unadjusted. In this study, we adjusted for race/ethnicity and education background as confounding factors, but other potential confounding factors such as income, bronchial hyper-responsiveness and so on.
The present study revealed that remarriage after bereavement or separation from a spouse was associated with a significantly reduced incidence rate of COPD. Remarriage might have a positive effect on the health status concerning COPD, it is, however, uncertain how remarriage affects the changes in the health status in individual diseases. To accommodate an aging society, further studies will be needed to figure out the potential relationship between remarriage behavior and the alteration of the health status.
We are grateful to Dr. Hiroo Miyazaki, Hamamatsu University School of Medicine for clinical advice. Moreover, we greatly appreciate the efforts of the Health and Retirement Study staff at the Institute for Social Research, University of Michigan.
- Stroebe MS: Handbook of bereavement research: consequences, coping, and care. 2001, Washington, DC: American Psychological Association, 1View ArticleGoogle Scholar
- Farr W: Vital statistics: a memorial volume of selections from the reports and writings of William Farr. 1975, Metuchen, N.J.: Scarecrow PressGoogle Scholar
- Clayton PJ: The effect of living alone on bereavement symptoms. Am J Psychiatry. 1975, 132 (2): 133-137.View ArticlePubMedGoogle Scholar
- Mineau GP, Smith KR, Bean LL: Historical trends of survival among widows and widowers. Soc Sci Med. 2002, 54 (2): 245-254. 10.1016/S0277-9536(01)00024-7.View ArticlePubMedGoogle Scholar
- Pitt B: Loss in late life. BMJ. 1998, 316 (7142): 1452-1454.View ArticlePubMedPubMed CentralGoogle Scholar
- Kaprio J, Koskenvuo M, Rita H: Mortality after bereavement: a prospective study of 95,647 widowed persons. Am J Public Health. 1987, 77 (3): 283-287. 10.2105/AJPH.77.3.283.View ArticlePubMedPubMed CentralGoogle Scholar
- Smith KR, Zick CD: Risk of mortality following widowhood: age and sex differences by mode of death. Soc Biol. 1996, 43 (1-2): 59-71.PubMedGoogle Scholar
- Schaefer C, Quesenberry CP, Wi S: Mortality following conjugal bereavement and the effects of a shared environment. Am J Epidemiol. 1995, 141 (12): 1142-1152.PubMedGoogle Scholar
- Turvey CL, Carney C, Arndt S, Wallace RB, Herzog R: Conjugal loss and syndromal depression in a sample of elders aged 70 years or older. Am J Psychiatry. 1999, 156 (10): 1596-1601.View ArticlePubMedGoogle Scholar
- Martikainen P, Valkonen T: Mortality after the death of a spouse: rates and causes of death in a large Finnish cohort. Am J Public Health. 1996, 86 (8): 1087-1093. 10.2105/AJPH.86.8_Pt_1.1087.View ArticlePubMedPubMed CentralGoogle Scholar
- Detailed tables - number, timing and duration of marriages and divorces: 2004. [http://www.census.gov/population/www/socdemo/marr-div/2004detailed_tables.html]
- Helsing KJ, Szklo M, Comstock GW: Factors associated with mortality after widowhood. Am J Public Health. 1981, 71 (8): 802-809. 10.2105/AJPH.71.8.802.View ArticlePubMedPubMed CentralGoogle Scholar
- Parkes CM, Prigerson HG: Bereavement: studies of grief in adult life. 2009, Hove, East Sussex, UK; New York: Routledge, 4Google Scholar
- Gentry M, Shulman AD: Remarriage as a coping response for widowhood. Psychol Aging. 1988, 3 (2): 191-196. 10.1037/0882-7922.214.171.124.View ArticlePubMedGoogle Scholar
- Hurd Clarke L: Remarriage in later life: older women's negotiation of power, resources and domestic labor. J Women Aging. 2005, 17 (4): 21-41. 10.1300/J074v17n04_03.View ArticlePubMedGoogle Scholar
- Vinick BH: Remarriage in old age. J Geriatr Psychiatry. 1978, 11 (1): 75-77.PubMedGoogle Scholar
- Davidson K: Late life widowhood, selfishness and new partnership choices: a gendered perspective. Ageing and Society. 2001, 21 (3): 297-317. 10.1017/S0144686X01008169.View ArticleGoogle Scholar
- Burks V, Lund D, Gregg C, Bluhm H: Bereavement and remarriage for older adults. Death studies. 1988, 12: 51-60. 10.1080/07481188808252219.View ArticleGoogle Scholar
- Joung IM, Mheen van de HD, Stronks K, van Poppel FW, Mackenbach JP: A longitudinal study of health selection in marital transitions. Soc Sci Med. 1998, 46 (3): 425-435. 10.1016/S0277-9536(97)00186-X.View ArticlePubMedGoogle Scholar
- Yohannes AM, Roomi J, Waters K, Connolly MJ: Quality of life in elderly patients with COPD: measurement and predictive factors. Respir Med. 1998, 92 (10): 1231-1236. 10.1016/S0954-6111(98)90426-7.View ArticlePubMedGoogle Scholar
- Peruzza S, Sergi G, Vianello A, Pisent C, Tiozzo F, Manzan A, Coin A, Inelmen EM, Enzi G: Chronic obstructive pulmonary disease (COPD) in elderly subjects: impact on functional status and quality of life. Respir Med. 2003, 97 (6): 612-617. 10.1053/rmed.2003.1488.View ArticlePubMedGoogle Scholar
- Kim HF, Kunik ME, Molinari VA, Hillman SL, Lalani S, Orengo CA, Petersen NJ, Nahas Z, Goodnight-White S: Functional impairment in COPD patients: the impact of anxiety and depression. Psychosomatics. 2000, 41 (6): 465-471. 10.1176/appi.psy.41.6.465.View ArticlePubMedGoogle Scholar
- Goldman N: Marriage selection and mortality patterns: inferences and fallacies. Demography. 1993, 30 (2): 189-208. 10.2307/2061837.View ArticlePubMedGoogle Scholar
- Juster F, Richard S: An overview of the health and retirement study. Human Resources. 1995, 30 (Suppl): 7-56. 10.2307/146277.View ArticleGoogle Scholar
- Mannino DM: COPD: epidemiology, prevalence, morbidity and mortality, and disease heterogeneity. Chest. 2002, 121 (5 Suppl): 121S-126S. 10.1378/chest.121.5_suppl.121S.View ArticlePubMedGoogle Scholar
- Seltzer JA: Cohabitation in the United States and Britain: Demography, kinship, and the future. Journal of Marriage and Family. 2004, 66 (4): 921-928. 10.1111/j.0022-2445.2004.00062.x.View ArticleGoogle Scholar
- Chatila WM, Wynkoop WA, Vance G, Criner GJ: Smoking patterns in African Americans and whites with advanced COPD. Chest. 2004, 125 (1): 15-21. 10.1378/chest.125.1.15.View ArticlePubMedGoogle Scholar
- Bakke PS, Hanoa R, Gulsvik A: Educational level and obstructive lung disease given smoking habits and occupational airborne exposure: a Norwegian community study. Am J Epidemiol. 1995, 141 (11): 1080-1088.PubMedGoogle Scholar
- Halbert RJ, Isonaka S, George D, Iqbal A: Interpreting COPD prevalence estimates: what is the true burden of disease?. Chest. 2003, 123 (5): 1684-1692. 10.1378/chest.123.5.1684.View ArticlePubMedGoogle Scholar
- Mannino DM, Homa DM, Akinbami LJ, Ford ES, Redd SC: Chronic obstructive pulmonary disease surveillance--United States, 1971-2000. MMWR Surveill Summ. 2002, 51 (6): 1-16.Google Scholar
- Goreczny AJ, Brantley PJ, Buss RR, Waters WF: Daily stress and anxiety and their relation to daily fluctuations of symptoms in asthma and chronic obstructive pulmonary disease (COPD) patients. Journal of Psychopathology and Behavioral Assessment. 1988, 10 (3): 259-267. 10.1007/BF00962549.View ArticleGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2458/9/412/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.