Annales de démographie historique
Belin
I.S.B.N.270113725X
222 pages
p. 85 à 104
doi: en cours
Veille sur la revue
Veille sur l'auteur
Belin
I.S.B.N.270113725X
222 pages
p. 85 à 104
doi: en cours
Veille sur la revue
Veille sur l'auteur
Vous consultez
no 108 2004/2
Adult mortality risks and religious affiliation:
The role of social milieu in biodemographic studies
Geraldine MINEAU a Department of Oncological Sciences and Huntsman Cancer Institute,University of Utah,2000 Circle of Hope,Salt Lake City,Utah 84112. Ken SMITH b Department of Family and Consumer Studies and Huntsman Cancer Institute,University of Utah,Salt Lake City,Utah. Lee BEAN c Department of Sociology,University of Utah,Salt Lake City,Utah
L'objectif de l'article est de préciser comment l'appartenance religieuse modifie le risque de décès. La base de données de la population de l'Utah est ici utilisée afin d'étudier la mortalité générale d'un échantillon d'hommes et de femmes mariés qui ont survécu à leur quarantième anniversaire. Les individus appartiennent aux générations 1850-1919, ils sont suivis pendant toute leur vie et se situent à l'époque de la transition démographique. Les membres actifs de l'Église des Mormons présentent une mortalité inférieure à celle des membres non actifs ou des non-Mormons, différence qui se maintient si l'on tient compte des hiérarchies socio-économiques. Les écarts sont plus forts pour les âges médians et pour les générations nées le plus récemment. L'appartenance à l'Église mormone est aussi plus discriminante pour les hommes que pour les femmes. Ces observations empiriques confirment le rôle des pratiques de santé et du soutien social dans les liens entre l'engagement religieux et la plus faible mortalité.
The purpose of this study is to understand how religious affiliation affects mortality risk. The Utah Population Database is used to examine all-cause mortality for a set of married men and women who survived to age 40. Individuals in selected birth cohorts (1850-1919) are observed across their lifetime and span the period of the demographic transition. We find active members of the Church of Jesus Christ of Latter-day Saints (LDS or Mormon) have lower mortality than those who are inactive or non-LDS in all cohorts and this relationship remains after controlling for socioeconomic status. The protective influences of being an active member of the LDS Church are greatest for the middle-aged and for those born in the more recent birth cohort. Our results show that religious affiliation has stronger effects on adult mortality for men rather than women. These observations are consistent with explanations of health practices and social support factors that have been posited to understand the positive relationship between religious involvement and mortality outcome.
There is a large and growing body of literature on the relationship between religious involvement and health outcomes. For the most part, religious involvement (e.g. church attendance or church membership) is associated with better physical and mental health and longer survival (George, Ellison and Larson, 2002). A meta-analysis of data from 42 studies from diverse populations found that religious involvement was significantly associated with lower mortality and the strength of the association was similar to that found for other psycho-social factors (McCullough et al., 2000). These studies have clarified the ways by which religion benefits health and divided them into four possible mechanisms: health practices, social support, psychosocial resources and a sense of coherence or meaning. The work of Hummer and his colleagues (Hummer et al., 1999) illustrates that social ties and behavioral factors mediate the relationship between religious involvement and mortality.
Durkheim’s seminal work on suicide (Durkheim, 1951) was among the first to advance the thesis that social support and integration are important factors affecting health status. This sociological insight has been confirmed by many investigators, including Waite and Lehrer (Waite and Lehrer, 2003) who recently reviewed the health benefits of both marriage and religion. In terms of religion, they cite a comprehensive body of research that documents an association between religious involvement and improved health status, including life expectancy. They note that religious affiliation may affect health outcomes because adherents to a faith have access to a network of people who may provide social resources, behavioral norms and instrumental support. Similarly, Jarvis and Northcott (Jarvis and Northcutt, 1987) indicate that people who attend church are more likely to be married, to be involved in a network of friendships, and to participate in social activities compared to those who do not attend services regularly. These attributes all work to improve the well being of individuals who have a religious attachment in relation to those lacking such an affiliation.
Health practices show the strongest associations on health outcomes when members of religions, that have explicit prescriptions and proscriptions about health behaviors, are compared to others (George, Ellison and Larson, 2002). Studies of Seventh Day Adventists and members of the Church of Jesus Christ of Latter-day Saints (LDS or Mormons), and Orthodox Jews have reported these findings. Observations from these studies show differences for the risks of cardiovascular disease, hypertension, stroke, numerous cancer sites, general health status indicators, and overall and cause-specific mortality (Levin, 1994). For example, Enstrom (Enstrom, 1989) found active LDS men had lower risks of cancer mortality than other Mormon men and had a life expectancy at birth that was more than six years longer than other U.S. white males. He also observed that couples committed to the LDS church are more likely to live longer.
The examination of religion and mortality risks presents interesting issues because of endogamy and the fact that members of certain religions may have arisen from isolated populations. This means that, when examining the effects of religious affiliation on health, it is possible that any association will be a function of some shared genetic risk (e.g., Ashkenazi Jews and breast cancer) as well as the environmental influences affecting health due to the behavioral practices embraced by adherents to a specific faith. It is therefore important to examine the stability of a religion-health association over time since such as assessment may reveal how socio-environmental factors directly alter mortality risks through time or through their interaction with underlying genetic susceptibilities.
The purpose of this article is to study how religious affiliation affects adult mortality risk. Our analyses focus on all-cause mortality for adults who marry and survive to age 40 and will examine the differences between active and inactive members of the LDS Church as well as those who are non-LDS. We use an historical and longitudinal data set that allows us to follow prospectively individuals born during a 70-year period (1850-1919) from age 40 to their death or to the present. The vast majority of studies investigating the association between religion and mortality have focused on contemporary U.S. populations using year-specific data that generate synthetic cohort estimates of differential mortality. A strength of the analysis presented here is our use of extinct (or nearly so) sex-specific cohorts. These cohorts span three key historical periods of time encompassing the demographic transition, thereby allowing us to compare the strength of association between religion and mortality in the distant past compared to a more contemporary period.
Study population
The LDS Church was established in 1830 in the state of New York. In the years immediately following, its members migrated and created settlements in Ohio, Missouri, Illinois and Iowa. The Mormons entered Salt Lake Valley in 1847; this was the beginning of an organized migration between 1846 and 1870 when over 60,000 pioneers and adherents of the LDS Church migrated from eastern and midwestern United States, as well as from Western Europe, into the U.S. intermountain west (Wahlquist, 1974). By the 1880 census there were 143,963 residents enumerated in Utah Territory and the 1900 census enumerated 276,749 residents. The state of Utah was created in 1896. Over time the LDS proportion of the population has varied from a peak in the initial settlement period with about 97 percent in 1870 to lower levels of around 56 percent in 1890 (Allen, 1989). About seventy percent of Utah’s current 2.3 million residents are members of the LDS Church.
We hypothesize that the health and longevity of active LDS members in relation to others individuals is partly due to their adherence to what is known as the “Word of Wisdom”. The Word of Wisdom contains language prohibiting the use of tobacco, alcohol or other stimulants (Alexander, 1996, 1981). The doctrine is attributed to a divine revelation to the founder of the LDS Church—Joseph Smith—in 1833; however, “(c)urrently available evidence indicates that adherence to the Word of Wisdom in the nineteenth century was sporadic” (Alexander, 1996). Emphasis on strict adherence to the Word of Wisdom appears to date from the early 20th century. Church authorities agreed not to “fellowship” (convert to membership in the church) “anyone who operated or frequented saloons.” (Alexander, 1996). Strict adherence may have remained somewhat problematic. In 1921 the Church authorities decreed that strict adherence was required to gain admission to an LDS Temple (Alexander, 1996).
Merrill’s (Merrill, 2004) recent research highlights the benefits of following these behavioral factors. He concludes that a portion of the difference in higher life expectancy among LDS and non-LDS is related lower use of tobacco and that “(t)obacco-related deaths had a larger impact on the difference in life expectancy when conditioned on older ages because of the latency period often involved with tobacco-related diseases”(p. 78). However, this factor did not explain all of the difference in life expectancy between LDS and non-LDS members, clearly indicating that other factors associated with the LDS faith contributes to their greater life expectancy. A second study that focused on suicide among young adult males in Utah compared different levels of affiliation with the LDS Church along with U.S. males in general. Active LDS men consistently had lower suicide rates for ages 20 through 34 while less active and nonmembers had fluctuating and higher rates than all U.S. males (Hilton, Fellingham and Lyon, 2001). Enstrom (Enstrom, 1989) showed that mortality risks are significantly lower among active male members of the LDS Church who hold major religious positions for a broad range of serious disease including cancer and cardiovascular disease in relation to U.S. men in general. Collectively, these findings are consistent with the hypothesis that health practices, social integration, and social support associated with religious involvement reduce the risk of a broad spectrum of adverse health outcomes.
The depth of commitment to religion is crucial in addressing behavioral versus familial effects. For example, siblings born into a faith are likely to share some genetic risk of disease/mortality. Suppose one of the siblings later rejects the faith but the other does not. Also imagine that the sibling who is less religious fares poorly in relation to sibling who remains faithful to the religion. This would suggest that behavioral aspects associated with religious adherence are important even in the presence of a shared genetic susceptibility. For these reasons, this analysis considers level of religious commitment and its effects on mortality risk.
Much of our previous mortality and longevity research has emphasized the need to include information on religious affiliation as a way to reduce the variability in social support and behavioral factors. In a study familial excess longevity (Kerber et al., 2001), affiliation with the LDS Church was found to be associated with better survival past the age of 65 and used as a covariate to reduce possible confounding in a model of inheritance. Kerber found about 10 to 20 percent of the variability in excess longevity could be heritable (Kerber et al., 2001). Similarly in analyzing the relationship between reproductive behavior and adult longevity from an evolutionary biology approach, we (Smith, Mineau and Bean, 2002) controlled for the commitment to the LDS Church and found that bearing and rearing children does affect the mortality risks of post-reproductive mothers and fathers.
Utah Population Database
This mortality study draws upon the Utah Population Database (UPDB) for information on a range of demographic, family, and religious characteristics and their influences on the risk of mortality across time periods. The UPDB contains over seven million records. It includes the genealogies of the founders of Utah and their Utah descendants. These records were computerized in the mid to late 1970s (Skolnick et al., 1979; Skolnick et al., 1981) and have been linked to other data sets, including birth and death certificates, cancer records, driver license records, census records and records from the Social Security Death Index. The UPDB is the only such database in the United States and one of few such resources in the world.
These genealogical records originated as “Family Group Sheets” filled out by members of LDS Church. Records were selected from the Family History Library of the LDS Church in 1975-76 and again in 1978-79 and computerized (Bean, Mineau and Anderton, 1990). The criterion for selection was that one or more family members was born or died on the Mormon Pioneer Trail or in Utah. The purpose was to represent migrants to Utah and their Utah descendants. The genealogy records for early migrants and their families represent birth cohorts that date back to about 1760. These 170,000 family group sheets (containing about 1.6 million individuals) have been linked across generations and, in some instances, the records encompass as many as seven generations. While these records may be similar to those available on the web with FamilySearch or other publicly available genealogical databases, they are not the same database and have been maintained as a resource only for biomedical and health research.
The UPDB is a dynamic database and receives annual updates for Utah births, deaths, cancer records and driver license records. Several projects have computerized and linked older sets of vital records and these data are now part of the UPDB. The database now includes approximately 1.8 million Utah birth certificates from 1915-1921 and 1947-2002 and 666,878 death certificates from 1904-2002. Much of the value of this resource depends on the ability to match records on individuals from two or more data sets, known as record linking, and to create longitudinally linked data that are able to capture many events associated with an individual. These record linking activities provide quality control for demographic information and eliminate duplicate records in the genealogical data.
The Utah Resource for Genetic and Epidemiologic Research (RGE) at the University of Utah administers access to these data through a review process of the project proposal. The protection of privacy and confidentiality of individuals represented in these records has been negotiated with agreements between RGE and the data contributors. All research projects require approval from an Institutional Review Board and RGE Review Committee (Wylie and Mineau, 2003).
Previous Research Using UPDB
The representativeness of the genealogy file has been demonstrated in a variety of demographic studies on infant mortality (Lynch, Mineau and Anderton, 1985; Bean, Mineau and Anderton, 1990) and maternal mortality (Bardet et al., 1981) that have compared Utah rates and patterns to other populations. Other studies have analyzed fertility (Bean, Mineau and Anderton, 1990), birth spacing (Anderton and Bean, 1985), and widowhood (Mineau, 1988; Mineau, Smith and Bean, 2002). This population is biologically representative of a broad spectrum of the white U.S. population and is genetically similar to other Northern European-derived populations. The population has a low inbreeding rate that is very similar to that of the U.S. white population due to a large founding population and high rates of immigration from a diverse group of outside populations (Jorde, 1989, 2001).
Description of data and methods
Death Information
Death dates and places are available from the genealogy whether the death occurred in Utah or in other states/countries and are nearly complete through the 1960s. Additional follow-up information comes from the Utah death certificates that are complete through 2002 or the Social Security Death Index (SSDI) that are available through fall 2003. The death certificates provide information on deaths within Utah and the SSDI provides death dates and places that occur outside of Utah. Information on individuals who are still living was determined by a driver license or identification card that is issued through the Driver License Division.
Religion Affiliation
The UPDB contains individual religion information relating to the LDS Church, specifically dates of baptism and endowment. These two formal religious rites define affiliation with the LDS Church. Baptism for children typically occurs at about age eight and therefore does not imply a high degree of personal choice. Individuals who convert or select membership in the LDS Church may be baptized at older ages. Regardless of the age at which baptism occurs, it is an essential prerequisite for the second, more formal religious rite, endowment. Endowments take place in an LDS temple or during the xixth century in a Sealing House. Endowments occur for young men (and increasingly young women) when they are called to serve an LDS proselytizing mission. Endowments may also occur at the time of a temple marriage or years after a secular marriage as members secure access to a temple. Regardless of the time or reason, an endowment implies a strong commitment to the LDS Church because each person must be certified by a religious authority. Their worthiness is evaluated during a personal interview to determine church attendance, payment of tithing, maintenance of the Word of Wisdom, and professed belief in the theology of the LDS Church.
Individuals who have records containing either baptism or endowment dates (but not posthumously) are treated as affiliated with the LDS Church. Records in the genealogy that do not contain these dates are classified as having less or no affiliation with the LDS Church. Religious commitment is based on a classification scheme developed by Mineau (Mineau, 1980) and Bean (Bean, Mineau and Anderton, 1990) using the timing of endowment. Individuals who have records containing endowment dates before age 40 are treated as religiously committed to the LDS church; all others are less committed or non-LDS members. Three categories are used: 1) Active LDS are persons who were baptized and endowed before the age of 40 (Females: N=85,117 or 53.2% of all females; Males: N=84,267 or 48.5% of all males). 2) Inactive LDS are persons who were baptized before the age of 40 but not endowed by age 40. They may have been endowed later or never endowed (Females: N=53,689 or 33.5% of all females; Males: N=58,869 or 33.9% of all males). 3) Non-LDS are persons with no information about baptism or endowment before age 40. These individuals have little or no affiliation with the LDS Church; however, they may be active members of other religious groups (Females: N=21,249 or 13.3% of all females; Males: N=30,463 or 17.6% of all males).
We define converts to the LDS Church as individuals who are endowed after age 40 and are a subset of individuals who were initially classified as Inactive LDS or non-LDS. For females, 30.6% (22,949 of 74,938) have an endowment date at 40 years of age or older while for men, this figure is 21.2% (18,960 of 89,332).
Sample Selection
A sample was selected for individuals who have complete data on all key variables necessary for the analysis. There are several eligibility criteria for inclusion into the final analysis sample. An individual’s record must contain complete birth and death dates. In instances where an individual is known to be alive, it is also required that a date last known alive is available. Given the importance of categorizing individuals by their religious affiliation, it is necessary to have information on baptism and endowment within the LDS Church. Endowment often occurs at the time of marriage, particularly for women. Consequently, we restricted the sample to include ever-married individuals only. This restriction reduces the problem of categorizing an individual as inactive LDS (i.e., lacking an endowment date) because they were unmarried. As described above, persons who lack both baptism and endowment dates are classified as non-LDS. Persons who are inactive LDS are those who baptized but not endowed prior to age 40. Converts to the LDS faith were those with endowment dates after age 40.
Our focus is on adult mortality after age 40 so all individuals in the analysis are required to have survived to this age. This survival threshold was selected in large measure because it represents an age by which the majority of individuals have married. For those that marry, we are in a position to observe whether the individual has the requisite endowment dates needed to determine their religiosity.
The subjects’ birth years are also restricted to the years 1850 to 1919. This restriction was imposed because it represents, at the lower end, an historical period following the establishment of the LDS Church in 1830 and initial Utah settlement in 1847. Individuals born in 1850 would be age 40 in 1890 and would have had access to religious facilities throughout the region where ceremonies could take place. We do not consider births past 1919 because we wanted to observe mortality rates for as many years past age 40 as possible. Accordingly, someone born in 1919 would reach age 40 in 1959 giving us approximately 40 years of adult life-span after age 40 in which to observe mortality rates.
Analysis
Our sample, shown in Table 1, comprises 333,654 individuals, with female and male sub-samples of 160,055 and 173,599, respectively. There are more males in this sample than females for at least two reasons. The first is related to differential mortality after marriage and before age 40. This results in the exclusion of female deaths due to maternal mortality; for example, more women than men were excluded in the 1850-79 cohort because they died before age 40. The second reason is related to our criteria that we know either a death date or a verification of “living” status to be included in the sample. Women are excluded from the sample at a slightly higher rate than men, because we may lack information on some women who remarry and have a different last name on their death certificate, driver license, or other records. This is particularly apparent in the 1909-1919 cohort where the disproportion of men to women is the greatest.
Tab. 1
Mortality and Religious Information for Individuals who Survived to Age 40 Specified for Males and Females and by Birth Cohort
All of the survival models are estimated with Cox Proportional Hazard Models based on this sample. Other models address the question about how religiosity affects the likelihood of reaching an advanced age or not. For these models, the outcome variable is a dichotomous variable that equals zero if you died before age 85 and equals one if you lived to age 85. Approximately 80% of the sample (all of whom survived to age 40) died before age 85. Accordingly, persons dying prior to age 85 are used as a standard comparison group for this set of analyses. We therefore code persons as zero for deaths occurring before age 85 and create two additional dichotomous variables for longevity, whether you live to age 90, and whether you lived to age 95. For these latter two variables (“living to age 90 versus dying before 85” and “living to age 95 versus dying before 85”), the sample size declines because of the intentional exclusion of individuals who died between the two threshold ages used to construct the variables. The sample sizes for analyses using these variables vary from one another by definition. The sample sizes for the outcomes for living to age 85, 90, and 95 are 199,619, 176,364, and 162,188, respectively.
Statistical Methods
Survival models for mortality past age 40 were estimated using Cox proportional hazards (PH) regression models where the outcome is the hazard rate for all-cause mortality. Several PH models are estimated to address different aspects of the association between religiosity and mortality. The first set of models assesses how religiosity affects the mortality hazard rate imposing the standard proportionality assumption that states that the effects of religiosity on mortality do not change with age. A second class of models is estimated that relaxes this assumption to allow for the effects of religiosity to change with age (non-proportionality models). Finally, we consider models that provide for the effects of converting to the LDS Church after age 40 by allowing for a time-varying covariate that accounts for the ages after 40 where an individual was and was not a member of the LDS Church.
The PH models focus on hazard rates but they do not show directly how religiosity affects survival to an advanced age. A set of “exceptional survival” (ES) models provides estimates for such effects. We rely on logistic regressions to examine this discrete version of survival. The dependent variables here are binary that take on a value of one if they exceed a threshold age (live to at least 85, 90, or 95) and a value of zero if they did not live to age 85. Only persons born prior to 1900 are included in these analyses.
Results
The risks of mortality are significantly lower for active LDS members than for either in-active or non-LDS individuals (Table 2). The protective influences of being Mormon are found to be greater for men than for women. Among males, inactive and non-LDS members have nearly identical risks of mortality in relation to active church members. For females, the benefits of being active LDS are weaker (compared to males) but the largest difference is with non-LDS women.
Tab. 2
Effects of Religiosity and Fertility, by Gender, on Mortality Risks Based on Cox Proportional Hazards Models
We have investigated whether the effects of religiosity may be confounded with fertility given that active Mormons have larger families (Mineau et al., 1984; Bean, Mineau and Anderton, 1990) and that increasing parity could affect mortality risks of parents (Smith Mineau, and Bean, 2002). We define parity as the number of children ever born to a couple. To address this possibility, we re-estimated our PH models using a sub-sample comprising persons who had complete parity data. Using this sub-sample, we estimated two equations (two for each gender): one with our set of religiosity codes and birth year as covariates and a second that added parity as a covariate. As shown in Table 2, the statistical adjustment for fertility had no appreciable influence on the estimated impact of religiosity on mortality risk.
The LDS Church has grown in myriad ways since its inception in 1830. These changes have generated conditions that could alter the manner in which being Mormon affects health outcomes (e.g., proscription of alcohol and tobacco, social safety net offered through church services). In Table 3, we find that the protective effects of being an active LDS Church member have grown with time. This trend holds for men and women alike. The weaker effects of religiosity in the earliest cohort may reflect the ubiquitous risks faced by everyone, risks that were more difficult to minimize through the actions of religious involvement (e.g., food shortages, infectious disease outbreaks) in relation to the risks more prevalent today (e.g., smoking).
Tab. 3
Effects of Religiosity, by Gender and Birth Cohort, on Mortality Risks Based on Cox Proportional Hazards Models. Panel A is for the Full Sample
It is possible that members of the LDS Church may have lower mortality risks because they tend to have higher social standing and better access to resources that also have salutary effects in relation to non-LDS individuals. We therefore examined how religious affiliation affects the mortality hazard rate after adjusting for the effects of socioeconomic status (SES) as measured by occupation obtained from Utah death certificates. Occupations were coded to the 1980 U.S. Census categories and assigned a socioeconomic status score based on Nam and Powers (Nam and Powers 1983). Higher scores represent occupations with greater socioeconomic status ranging from 1 to 100. For men, SES is measured based on their own occupation while for women it is based on their husband’s occupation. Table 3 shows that for this subset of records, active members of the LDS Church are still found to have lower mortality rates than inactive members and non-LDS before and after controlling for the influence of SES (Panels B and C in Table 3). The cohort of 1910-19 is excluded from these SES analyses because of a significant proportion still living.
Adult mortality risks change dramatically with age and so it is useful to determine how religiosity affects these changing risks as individuals age. Table 4 shows that the greatest beneficial effects of being an active Mormon occur during the middle years (40-49, 50-59) and steadily decline with advancing age. For example, among inactive LDS men the risk of dying between age 40 and 49 is 1.63 times greater than for active LDS men and among non-LDS men it is 1.70 times greater. For older ages (70+), the benefits of active involvement in the LDS Church are about half what they were for those under age 60 indicating that religion has attenuated effects on forestalling mortality as chronic, degenerative disease become more prevalent.
Tab. 4
Age-Specific Effects of Religiosity and Later-Life Conversion, by Gender, on Mortality Risks Based on Cox Proportional Hazards Models
Table 4 also shows that persons joining the LDS Church later in life have dramatic reductions in mortality in relation to active LDS members. This may reflect a greater adherence to the faith that translates into stronger health benefits. It is also possible that “late converts” are not so much converts but rather life-long members of the church who happened to marry late in life, possibly demonstrating a resolve and depth of commitment to their faith that translated into lower mortality risks.
We elaborate on the findings further in Table 5 by stratifying the sample by birth cohort. These results are not surprising if we extend and combine the findings obtained from Table 3 and 4. The protective influences of being an active LDS member are greatest for the middle-aged and for those born in the most recent birth cohort. Indeed, among men, the mortality risks of inactive and non-LDS members are twice that of active members for the middle-aged in the last cohort. We also see an increasing benefit of joining the LDS church over time.
Tab. 5
Age-Specific Effects of Religiosity and Later-Life Conversion, by Gender and Birth Cohort, on Mortality Risks Based on Cox Proportional Hazards Models
The exceptional survival models support the PH models in nearly every respect, as shown in Table 6. Active LDS Church members have a greater likelihood of reaching advanced ages than others and this benefit is greater for men than women. While these models assess these influences for only two birth cohorts, we observe again that the benefits of being an active Mormon are greater among more recent cohorts.
Tab. 6
Effects of Religiosity, by Gender and Birth Cohort, on Exceptional Survival (Living to ages 85, 90, 95) Based on Logistic Regressions.
Discussion/Conclusion
This article is part of a series of studies using a unique database that provide the opportunity to explore major biodemographic as well as social issues related to mortality and longevity. In order to clarify the social factors that may obscure genetic and biological factors determining longevity, these analyses are specifically designed to ascertain the degree to which a critical variable—religious affiliation—influences mortality risk in the Utah population.
There are several factors that can be used to explain our observation that active LDS have lower mortality than inactive or non-LDS. Both the health practices that were officially mandated by the LDS Church and the social support mechanism associated with church affiliation are more directly interpretable given our data sources and historical context. Because we studied only individuals who married, the systems of social support would be related more specifically to religious involvement that might be observed in other studies.
Some contemporary studies suggest that the protective effect of religious affiliation and mortality is stronger for women than men (Strawbridge and Cohen, 2000). Our results show the opposite, that religious affiliation has stronger effects on adult mortality for men than women and that this is the case across all three cohorts. We considered the possibility that fertility and its differential adverse effect on women (Smith, Mineau and Bean, 2002) might explain this finding, but we have rejected this mechanism as a likely explanatory factor.
Another explanation would be that active LDS have higher socioeconomic status and thus have access to better social conditions than others. However, our results indicate that the positive effect of being active LDS and particularly a greater protective effect for men is not the result of the man’s occupation.
A plausible alternative explanation relates to the health practices among active LDS members and others and how these differences vary by gender. During the earliest birth cohorts, both LDS and non-LDS men were more alike than in relation to men today. This observation is based on the fact the behaviors described in the Word of Wisdom were not institutionalized and thus were not adhered to as they are today. Accordingly, behaviors affecting health (e.g., tobacco and alcohol consumption) may have been more commonly practiced in LDS and non-LDS men alike in these early years. With the passage of time, the differences in the rates of these important risk behaviors would have diverged resulting in rising relative health benefits among active LDS men. Women, both Mormons and non-Mormons, were likely to have been more similar from the beginning because their rates of alcohol and tobacco consumption were low. Despite their low prevalence, active LDS women would have engaged in them even less as the Word of Wisdom went from being a principle to a requirement (Alexander, 1981) starting around 1920. This would account for the increasing (albeit weaker) relative health benefits observed among active LDS women. It is also possible that the potential positive effects of being active LDS are greater for men than women because of the greater social standing men have within the LDS Church in relation to women.
Smoking may partially explain the difference between active Mormons and others, especially for men, but even after statistical controls have been made for this confounder, active LDS members still enjoy lower rates of mortality (Merrill, 2004). A likely explanation for this protective effect, therefore, is the greater access to social support and networks available to LDS members than others. This explanation is consistent with other large-scale prospective studies of social support and mortality. The Alameda County study was among the first of such studies that demonstrated that those with few social ties (including the lack of church membership) had two to three times the risk of mortality than persons with many social contacts (Berkman and Syme, 1979). Our risk estimates are somewhat smaller than the Alameda study partly because our sample was restricted to ever-married persons, a data restriction that results in some minimum level of social integration for everyone in the sample. The mortality-social connection has since been demonstrated in a broad range of studies that concur with the general findings reported here. There is explicable variation in the effect sizes given varying measures of social support/integration, length of follow-up, study design, and sample eligibility criteria (House, Landis and Umberson, 1988) (Berkman and Glass, 2000).
An advantage of our analysis is that religiosity is measured earlier in life where initial health status would have been excellent for most persons. We have also restricted our sample to those who marry and survive to age 40, thereby enriching the sample with healthy individuals. These features of our analysis greatly minimize the possibility that health status affected religious participation (e.g., the healthy are better able to participate in church functions; or the ill are more religiously active as they confront the prospects of their own mortality). And as indicated by George (George et al., 2002), the mediating effect of factors such as health practices appear somewhat smaller in longitudinal studies compared to cross-sectional studies; however, estimates based on longitudinal studies “are likely to be more accurate because at least part of the effects of social selection have been excluded (p. 193)”.
Our results based on the UPDB and active members of the LDS Church suggest that exceptional life expectancies can be achieved at a population level under very positive health conditions accompanied with high levels of social integration. The exceptional longevity of active LDS members has been noted by others in the discussion of the upper limits of human longevity. Mormons are likely to be closer to some theoretical maximum than the larger U.S. population and life expectancy among active LDS members may well represent a target for longevity that might be achievable in other populations. From a public health prospective, this study adds to the body of literature that seeks to understand the mechanisms that promote better health outcomes as well as those associated with maximizing longevity (Manton, Stallard, and Tolley, 1991).
ADH_id270113725X_pu2004-02s_sa06_art06_img007.jpg
Fig. 1
Survival Curves for Males who Survived to Age 40 by Birth Cohort
Fig. 2
Survival Curves for Females who Survived to Age 40 by Birth Cohort
ACKNOWLEDGEMENT
We wish to thank the Pedigree and Population Resource funded by the Huntsman Cancer Foundation, University of Utah for providing the data and valuable computing support. The work was supported by NIH grant AG13748 (Kinship and Socio-demographic Determinants of Mortality) and NIH grant AG022095 (The Utah Study of Fertility, Longevity and Aging). In addition, we wish to acknowledge the contributions of Alison Fraser and Diana Lane Reed in enabling access to data and support of the project.
BIBLIOGRAPHIE
·
Alexander, Thomas G. (1981), “The Word of Wisdom: From Principle to Requirement”, Dialogue, 14 (3), 78-88.
·
Alexander, Thomas G. (1996), Mormonism in Transition: A History of the Latter-day Saints 1890-1930, Urbana and Chicago, University of Illinois Press.
·
Allen, Richard D. (1989), “Religion in Twentieth-Century Utah”, Utah's History, edited by R. D. Poll, T. G. Alexander, E. E. Campbell and D. E. Miller, Logan, Utah, Utah State University Press.
·
Anderton, D. L., Bean L. L. (1985), “Birth Spacing and Fertility Limitation; a Behavioral Analysis of a Nineteenth Century Frontier Population”, Demography, 22, 169-183.
·
Bardet, J.-P., Lynch, K. A., Mineau, G. P., Hainsworth, M. and M. Skolnick (1981), « La mortalité maternelle autrefois : une étude comparée », Annales de Démographie Historique, 31-49.
·
Bean, L. L., Mineau, G. P. and D. L. Anderton (1990), Fertility Change on the American Frontier, Adaptation and Innovation, Berkeley, University of California Press.
·
Berkman, L. F., Glass, T. (2000), Social Integration, Social Networks, Social Support and Health, edited by L. Berkman and I. Kawachi, Social Epidemiology, N. Y., Oxford.
·
Berkman, L. F., Syme, S. L. (1979), “Social Networks, Host Resistance, and Mortality: a Nine-Year Follow-up Study of Alameda County Residents”, American Journal of Epidemiology, 109 (2), 186-204.
·
Durkheim, E. (1951), Suicide, Glencoe, Ill, The Free Press.
·
Enstrom, J. E. (1989), “Health Practices and Cancer Mortality among Active California Mormons”, Journal of the National Cancer Institute, 81, 1807-14.
·
George, L. K., Ellison, C. G. and D. B. Larson (2002), “Explaining the Relationships between Religious Involvement and Health”, Psychological Inquiry, 13 (3), 190-200.
·
Hilton, S. C., Fellingham, G. W. and J. Lyon (2001), “Suicide Rates and Religious Commitment in Young Adult Males in Utah”, American Journal of Epidemiology, 155 (5), 413-419.
·
House, J.-S., Landis, K. R. and D. Umberson (1988), “Social Relationships and Health”, Science, 241, 540-545.
·
Hummer, R. A., Rogers, G. R., Nam, C. B. and C.G. Ellison (1999), “Religious Involvement and U.S. Adult Mortality”, Demography, 36 (2), 273-285.
·
Jarvis, G. K., Northcutt, H. C. (1987), “Religion and Differences in Morbidity and Mortality”, Social Science and Medicine, 25 (7), 813-24.
·
Jorde, L. B. (1989), “Inbreeding in the Utah Mormons: An Evaluation of Estimates Based on Pedigrees, Isonymy, and Migration Matrices”, Annals of Human Genetics, 52, 339-355.
·
Jorde, L. B. (2001), “Consanguinity and Prereproductive Mortality in the Utah Mormon Population”, Human Heredity, 52, 61-5.
·
Kerber, R. A., O'Brien, E., Smith, K. R. and R.M. Cawthon (2001), “Familial Excess Longevity in Utah Genealogies”, Journal of Gerontology: Biological Sciences.
·
Levin, J. (1994), “Religion and Health: Is there an association, is it valid, and is it causal”, Social Science and Medicin, 38 (11), 1475-1482.
·
Lynch, K. A., Mineau, G. P. and D. L. Anderton (1985), “Estimates of Infant Mortality on the Western Frontier: the Use of Genealogical Data”, Historical Methods, 18, 155-164.
·
Manton, K. G., Stallard, E. and H. D. Tolley (1991), “Limits to Human Life Expectancy: Evidence, Prospects, and Implications”, Population and Development Review, 17, 603-637.
·
McCullough, M. E., Hoyt, W. T., Larson, D. B., Koenig, H. G. and C. Thoresen (2000), “Religious Involvement and Mortality: a Meta-Analytic Review”, Health Psychology, 29 (3), 211-222.
·
Merrill, R. (2004), “Life Expectancy among LDS and non-LDS in Utah”, Demographic Research, 10, 61-82.
·
Mineau, G. P. (1980), “Fertility on the Frontier: an Analysis of the Nineteenth- Century Utah Population”, Sociology Dissertation, University of Utah.
·
Mineau, G. P. (1988), “Utah Widowhood: a Demographic Profile”, On Their Own: Widows and Widowhood in the American Southwest 1848-1939, edited by University of Illinois. Press. Chicago.
·
Mineau, G. P., Anderton, D. L., Bean, L. L. and J.-D. Willigan (1984), « Évolution différentielle de la fécondité et groupes sociaux religieux: l'exemple de l'Utah au xixe siècle », Annales de Démographie Historique, 219-236.
·
Mineau, G. P., Smith, K. R. and L. L. Bean (2002), “Historical Trends of Survival among Widows and Widowers”, Social Science and Medicine, 54, 245-254.
·
Nam, C.B., Powers, M. G. (1983), The Socioeconomic Approach to Status Measurement, Houston, Cap and Gown Press.
·
Skolnick, M., Bean, L. L., Dintelman, S. and G. P. Mineau (1979), “A Computerized Family History Database System”, Social Science Research, 63, 506-523.
·
Skolnick, M., Bishop, D., Carmelli, D. et al. (1981), “A Population-Based Assessment of Familial Cancer Risk in Utah Mormon Genealogies”, Gènes, Chromosomes, and Neoplasia, New York, N. Y. Raven Press.
·
Smith, K.R., Mineau, G. P. and L. L. Bean (2002), “Fertility and Post-Reproductive Longevity”, Social Biology, 49 (1), 55-75.
·
Strawbridge, W. J., Cohen, R. D. (2000), “Comparative Strength of Association between Religious Attendance and Survival”, Int'l J. Psychiatry in Medicine, 30 (4), 299-308.
·
Wahlquist, W. L. (1974), Settlement Progress in the Mormon Core Area: 1847-1890, Dissertation, University of Nebraska.
·
Waite, L. J., Lehrer, E. L. (2003), “The Benefits from Marriage and Religion in the United States: a Comparative Analysis”, Population and Development Review, 29 (2), 255-275.
·
Wylie, J. E., Mineau, G. P. (2003), “Biomedical Databases: Protecting Privacy and Promoting Research”, Trends in Biotechnology, 21, 113-116.
© Cairn 2007
Vie privée |
Conditions d’utilisation |
Conditions générales de vente
À propos | Éditeurs | Bibliothèques | Aide à la navigation | Plan du site | Raccourcis
À propos | Éditeurs | Bibliothèques | Aide à la navigation | Plan du site | Raccourcis
Retour en haut de page
Ici s'achève le contenu de la page, la suite ne concerne que les attributs d'affichage
