This essay provides a review of forty years of follow-up data from the Charleston Heart Study (CHS) in their article Divorce and Death: Forty Years of the Charleston Heart Study. This includes a longitudinal CHS, which included data collected from more than 1300 adults from 1960 to 2000, Sbarra and Nietert explored the relationship between social connectedness and health using the CHS data and attempted to provide insight into the long term health consequences of becoming separated or divorced (2009). Being separated or divorced during the follow-up window appeared to be one of the strongest predictors of early mortality. In addition to the review of the actual study, a critique is included that provides a critical analysis of the quality of the researchers’ study and article as published in the Psychological Science journal. The critique addresses such items as ethics, usefulness, sample size and diversification as well as a plethora of other interesting items useful to provide collegiate feedback of the work by Sbarra and Nietert.
Early Mortality: Review of the Charleston Heart Study
The researchers posed that recent research in social epidemiology has spurred advances into the association between interpersonal relationships and health (Sbarra & Nietert, 2009) but there were limits to the overall findings. It is suggested that relationship can play a vital role in an individual’s lifespan development. Sbarra and Nietert share this sentiment. More specifically, that social integration (i.e. a close relationship in which one feels close to others and that the relationship is reciprocally dependable) is positively correlated with mortality; however, a life deficient in social integration may be a strong predictor of early mortality (Sbarra & Nietert, 2009). The researchers carefully reviewed the data from the CHS to help explain the mechanism linking social connectedness and health.
The authors were interested in providing new insights into the long-term health consequences of divorce or separation. Many previous studies on the link between divorce and health have failed to present marital status as a dynamic variable. However, the researchers took into account both the length of time the participants were divorced and eventual remarriage rates which appears to have significantly improved their ability to make the link.
In order to present marital status as dynamically as possible, the researchers classified participants in the study in three ways. First, marital status (married, separated-divorced, widowed, or never married) at the first assessment was examined as a predictor of long-term mortality (Sbarra & Nietert, 2009). Second, the researchers calculated the hazard ratio (HR) of early death for adults who were separated or divorced at each assessment relative to all other participants who contributed at least one marital-status entry (Sbarra & Nietert, 2009). Finally, to determine if the effect of remaining separated or divorced differed from the effect of having ever experienced a separation or divorce, the researchers reclassified the sample to calculate the HR of adults who simply experienced a marital separation or divorce at some point during the CHS follow-up period relative to all other participants (Sbarra & Nietert, 2009).
The researchers used data collected in the Charleston Heart Study to address the possible link between marital status and health. The CHS was a community-based cohort study designed to assess the normal course of health and aging for adults over age 35 residing in Charleston County, South Carolina and while the study began in 1960, marital-status data were collected during several follow-up phases: 1962–1964, 1974–1975, 1984–1985, 1987–1989, and 1990–1991 (Sbarra & Nietert, 2009). Mortality data were updated throughout the study, with the final revision spanning the entire 41-year period.
The CHS began with an initial sample of 2181 adults, 1195 women and 986 men, 61% were of Caucasian decent, while the remainder of the sample was of African American decent (Sbarra & Nietert, 2009). At the onset of the study, the average age of the participants were 48 years 9 months. The fact that the average age of the participants was not exceptionally young, combined with the extended length of the study, provided an opportunity for the researchers to capture data to the end of life for the majority of the participants (Sbarra & Nietert, 2009).
In the CHS, information on marital status was collected by self-report, and participants were classified as married, widowed, separated, divorced, or never married at each assessment. In the researchers’ analysis, the separated and divorced categories were merged to create a single category defined by the experience of marital separation (Sbarra & Nietert, 2009). At the 1962–1964 assessment, the first-time marital status data was captured, 81.2% of the sample was married, 10% was widowed, 5.5% was separated or divorced, and 3.3% had never been married (Sbarra & Nietert, 2009).
In-person medical interviews at the beginning of the CHS provided a large amount of medical data, such as blood pressure, smoking status, and body mass index, that helped the researchers predict mortality. To predict mortality over the 41-year study period, the researchers used Cox proportional hazards models, a regression approach commonly referred to as survival analysis. The researchers were able to create predictors for each variable to isolate the effect of the variable of interest – time spent divorced or separated on age of death.
Of the 1,376 adults in the restricted sample assessed at baseline, 74% had died by 2000, with the remaining 26% excluded from the research (Sbarra & Nietert, 2009). The researchers found that separated and divorced participants had a 55% greater likelihood of death in the follow-up period than participants from the other three categories of relationship status (Sbarra & Nietert, 2009).
The analysis of the CHS data by the researchers suggests that longevity of life is positively correlated to healthy and cherished relationships and that living a large portion of life as a separated or divorced adult may add considerable risk for all-cause of mortality (as cited in Sbarra & Nietert). It may be further suggested that the process of a marital break may not present the highest mortality risk; rather it is more likely the time spent without a significant other that is the most crucial factor. The results do not lend to reasons of cause but suggest results of significant correlation. This agrees with Sbarra and Nietert, who also suggest that additional research is necessary to fully understand the mechanisms behind the elevated risk of early mortality associated with time spent as a separated or divorced individual (2009).
The Charleston Heart Study, on which the article is based, appeared to be conducted in an ethical and scientific manner. In addition, the researchers appeared to be interested in conducting a study that included a diverse sample. After the initial 1960 phase of the study, 102 Black men of high socioeconomic status were added to increase the diversity component, which could be considered “forward-thinking” for 1962-era America. The fact that this was a longitudinal study (i.e. conducted over a span of 41 years) proved to be beneficial. Since the original study was conducted with cardiovascular health in mind, the fact that the authors had the insight of gathering other useful data for later use, is truly impressive.
Still, there are some shortcomings with the data, as Sbarra and Nietert admit (2009). The incident of divorce was less customary at the onset of the CHS, though it would become significantly more common as the study progressed. Consequently, this presents one of the limits of a cohort study (i.e. how the customs of a single generation can affect the results of a study unlike another generation). The number of participants that remained separated or divorced throughout the follow-up period was relatively small, limiting the ability of the sample to be generalized across the entire population. It may have also been more efficacious to delineate between those who were separated and those who were divorced in the original study. Separating this group into two variables may have expanded the results and provided more insight.
Despite these limits, the work of the authors has provided contemporary researchers with a good start in understanding the link between marriage status and longevity of life. A more modern approach to this type of study, with more nuanced relationship status categories and more frequent follow-up occasions, might benefit the subject matter. A new study, with a more contemporary cohort sample and the opportunity to collect data specifically useful to psychological research, may proffer more insight for the avid researcher. Additional follow up studies may also include individuals from select cultures or ethnicities as well as individuals from a variety of socioeconomic statuses.
Sbarra, D. A. & Nietert, P. J. (2009). Divorce and death: Forty years of the Charleston heart study. Psychological Science, 20(1), pp. 107-113.