- Split View
-
Views
-
Cite
Cite
Robert S. Wilson, Carlos F. Mendes de Leon, Julia L. Bienias, Denis A. Evans, David A. Bennett, Personality and Mortality in Old Age, The Journals of Gerontology: Series B, Volume 59, Issue 3, May 2004, Pages P110–P116, https://doi.org/10.1093/geronb/59.3.P110
- Share Icon Share
Abstract
We examined the relation of personality to mortality in 883 older Catholic clergy members (69% women). At baseline, they completed the NEO Five-Factor Inventory, which assesses the five principal dimensions of personality. They were followed for a mean of 5.1 years, during which 182 deaths occurred. Risk of death was nearly doubled in those with a high neuroticism score (90th percentile) compared with a low score (10th percentile) and was approximately halved in those with a high conscientiousness score compared with a low score. Findings for extraversion were mixed, and neither agreeableness nor openness was strongly related to mortality. The results suggest that personality is associated with mortality in old age.
THE idea that personality may somehow influence health and longevity is longstanding and is supported by the observation that many of the leading risk factors for morbidity and mortality are behavioral or lifestyle variables such as diet, exercise, or tobacco use. The association of personality with health is not securely understood, however, in part because much of the relevant research has been based on subjective indices of health, and personality is related to health complaints, making interpretation of findings difficult (Costa & McCrae, 1987; Watson & Pennebaker, 1989).
Studies of the relation of personality to mortality provide a more objective means of examining the role of personality in health. The association between personality and longevity has been examined in several studies, but results have been mixed. One reason for this inconsistency is that many past studies have focused on novel constructs (e.g., cancer-prone personality) rather than well-established personality traits. In the past decade, a five-trait model of personality has gained widespread scientific acceptance (Digman, 1990). It consists of neuroticism (also termed emotional instability, see Goldberg, 1992, and negative affectivity, see Watson & Clark, 1984), which refers to a tendency to experience psychological distress; extraversion, which denotes a tendency to be sociable, active, and experience positive emotions; openness, which indicates intellectual curiosity and preference for varied experiences; agreeableness, which indicates a tendency to be trusting, sympathetic, and cooperative; and conscientiousness, which refers to a disposition to be diligent, organized, and achievement oriented. Although the association of these traits with health has been extensively investigated (Smith & Williams, 1992), relatively few studies have examined the relation of the five-factor taxonomy to risk of death.
Of the five traits, the relation of neuroticism to mortality has been the most frequently examined. Neuroticism was associated with an increased risk of death in one study of community-dwelling adults (Almada et al., 1991), but other studies have found no evidence of this association (Costa & McCrae, 1987; Maier & Smith, 1999; Martin & Friedman, 2000) or the opposite effect (Korten et al., 1999). The apparent lack of association between neuroticism and mortality is puzzling because persons high in this trait experience more negative emotions such as depression, anxiety, and anger (De Beurs, Beekman, Deeg, Van Dyck, & Van Tilburg, 2000; Ormel & Wohlfarth, 1991; Watson & Clark, 1984) that have been associated with increased risk of mortality (Barefoot & Schroll, 1996; Kubzansky et al., 1997; Miller, Smith, Turner, Guijarro, & Hallet, 1996; Wilson, Bienias, Mendes de Leon, Evans, & Bennett, 2003).
Extraversion has not been related to mortality (Korten et al., 1999; Maier & Smith, 1999; Martin & Friedman, 2000), though there is some evidence that optimism, a component of extraversion, may be (Maruta, Colligan, Malinchoc, & Offord, 2000). Evidence regarding the other three traits is scant, with existing data suggesting that conscientiousness and possibly agreeableness (Christensen et al., 2002; Martin & Friedman, 2000), but not openness (Christensen et al., 2002; Maier & Smith, 1999), may be associated with reduced mortality.
Another factor that has complicated research in this area is socioeconomic status. Persons of lower socioeconomic status tend to have more negative emotions than persons of higher socioeconomic status (Gallo & Mathews, 2003). Low socioeconomic status is also an established risk factor for mortality (House, Kessler, & Herzog, 1990). Therefore, most studies, but not all (Costa & McCrae, 1987; Martin & Friedman, 2000), have included variables such as education or income to control for the potentially confounding effect of socioeconomic status in analyses. To the extent that these traditional indicators do not fully account for individual differences in socioeconomic status in old age (Crystal & Shea, 1990; Robert & House, 1996), however, some residual confounding may be occurring.
In this study, we examined the relation between personality and mortality in 883 older Catholic clergy members who had annual clinical evaluations for a period of up to 8 years. Because the cohort is relatively homogeneous with respect to socioeconomic status and related lifestyle factors, we reasoned that the potential for such factors to confound the association of personality with mortality might be reduced. At baseline, participants completed standard measures of neuroticism, extraversion, openness, agreeableness, and conscientiousness. We examined the relation of each trait with risk of death during the follow-up period, controlling for age, sex, education, and selected indicators of health and functional ability at baseline. Because neuroticism denotes a tendency to experience negative emotions that have been linked to mortality, we hypothesized that high levels of neuroticism would be associated with increased risk of death. We made no predictions regarding the other four traits because of the relative dearth of previous research.
Methods
Participants
Participants are from the Religious Orders Study, a longitudinal clinicopathologic investigation of aging and Alzheimer's disease in older Catholic clergy members who agreed to annual clinical evaluations and brain autopsy in the event of death. Evaluations began in January 1994 and are continuing. The study was approved by the Institutional Review Board of the Rush University Medical Center.
As of February 2003, 960 persons had completed the baseline clinical evaluation. We excluded 77 Religious Orders Study participants who met criteria for dementia at baseline (see the paragraphs that follow). Analyses are based on the remaining 883 persons. At baseline they had a mean age of 75.1 years (SD = 7.1), a mean education of 18.2 years (SD = 3.3), and a mean score of 28.4 (SD = 1.7) on the Mini-Mental State Examination (MMSE); 69.1% were women and 89.2% were White and non-Hispanic.
Clinical Evaluation
At baseline, each person had a uniform clinical evaluation that included a medical history, a complete neurological examination, and an assessment of cognitive and motor abilities. On the basis of this evaluation, an experienced neurologist or geriatrician classified persons with respect to dementia by using the criteria of the joint working group of the National Institute of Neurological and Communicative Disorders and Stroke/Alzheimer's Disease and Related Disorders Association (McKhann et al., 1984). These require a history of cognitive decline and impairment in two or more domains of cognition. Further information on this evaluation is published elsewhere (Bennett et al., 2002; Wilson, Schneider, Bienias, Evans, & Bennett, 2003).
Assessment of Personality Traits
At baseline, we administered the NEO Five-Factor Inventory to assess personality (Costa & McCrae, 1992). Participants rated each of 60 statements on a 5-point scale (strongly disagree, disagree, neutral, agree, and strongly agree). Neuroticism (e.g., “I often feel tense and jittery”) is the disposition to experience psychological distress. Extraversion (e.g., “I like to have a lot of people around me”) refers to a tendency to be outgoing, energetic, and optimistic. Openness (e.g., “I am intrigued by the patterns I find in art and nature”) indicates an active imagination, intellectual curiosity, and independence of judgment. Agreeableness (e.g., “I try to be courteous to everyone I meet”) is the tendency to be altruistic and helpful. Conscientiousness (e.g., “I keep my belongings clean and neat”) refers to the tendency to be scrupulous, reliable, and purposeful. Item scores, which ranged from 0 to 4, were added to yield summary measures of each trait that could range from 0 to 48, with higher scores indicating a higher level of the trait. If any item score was missing, the trait score was treated as missing; this occurred in less than 1% of participants.
The NEO Five-Factor Inventory was developed as a short form of the NEO Personality Inventory (Costa & McCrae, 1992), which includes 48 items per trait rather than 12. In normative studies, correlations of traits assessed by the NEO Five-Factor scales with the same trait assessed on the NEO Personality Inventory scales ranged from a low of.75 for conscientiousness to a high of.89 for neuroticism in one cohort, and from.77 for agreeableness to.92 for neuroticism in another cohort. The internal consistency of the trait scales from the NEO Five-Factor Inventory appears acceptable, with estimates of coefficient alpha ranging from.68 for agreeableness to.86 for neuroticism. The scales also show clear evidence of convergent and discriminant validity, based on correlational evidence (Costa & McCrae, 1992). Overall, the NEO Five-Factor Inventory appears to provide psychometrically sound measures of the Big Five personality traits.
Assessment of Other Clinical Features
We gathered data on six indicators of health and function at baseline. Cognitive function was assessed with a set of 19 individual tests in an approximately 1-hr session. We formed a composite measure of global cognition by converting raw scores on each test to z scores, using the baseline mean and standard deviation, and averaging the z scores to yield the composite. On the basis, in part, of a factor analysis of the tests at baseline, we also formed summary measures of episodic memory (based on 7 tests), semantic memory (based on 4 tests), working memory (based on 4 tests), perceptual speed (based on 2 tests), and visuospatial ability (based on 2 tests) by converting raw scores to z scores and computing the average. Detailed information about the 19 individual tests and the formation of the composite measures is published elsewhere (Wilson, Barnes, et al., 2002; Wilson, Beckett, et al., 2002; Wilson, Bienias, Berry-Kravis, Evans, & Bennett, 2002; Wilson, Mendes de Leon, et al., 2002; Wilson, Schneider, et al., 2003).
We assessed lower limb function with five performance-based measures, as previously described (Wilson, Bienias, et al., 2003). Each individual measure was scaled from 0 to 5, with higher scores indicating better performance and 0 indicating an inability to perform the task. We asked participants to walk 8 ft (2.43 m) at a normal pace and measured the time and number of steps taken. Each distribution was divided into quintiles, with scores of 5 assigned to the fastest walking speeds and fewest walking steps. Participants were asked to sit up and down five times in a chair. Those who attempted but could not complete the task were given a score of 1. The time taken to successfully complete the task was divided into quartiles, and scores of 2 to 5 were assigned, with higher scores for more rapid performance. We asked persons to walk an 8-ft (2.43-m) line, heel to toe. The number of steps off the line was divided into quintiles, with scores of 5 assigned to those with the least steps off line. To assess standing balance, we asked persons to stand in each of four positions for a period up to 10 s: full tandem stand (i.e., heel to toe) with eyes open, semitandem stand (i.e., heel of one foot astride toes of other) with eyes open, and side-by-side stand (i.e., feet aligned and touching) with eyes open and with eyes closed. Total time on the four tasks was divided into quintiles, with 5 assigned to those able to maintain balance the longest. Because the resulting five measures were moderately correlated (median r =.39), we averaged them to create a global measure of lower limb function.
Seven medical conditions were present in at least 5% of persons at baseline: hypertension, diabetes, heart disease, cancer, thyroid disease, head injury, and stroke. Classification was based on medical history except for stroke, which was based on history, examination, and review of brain scan when available. We used the total number of conditions present as an index of chronic illness, as previously described (Wilson, Beckett, et al., 2003; Wilson, Bienias, et al., 2003; Wilson, Mendes de Leon, et al., 2002).
Persons were divided into those who ever versus never smoked cigarettes. Current alcohol use was expressed as the number of alcoholic drinks consumed in the past year, log transformed because of the skewed distribution. Weight divided by height squared (in kilograms divided by square meters) was used as an index of body mass.
Data Analysis
We assessed the internal consistency of each trait scale with Cronbach's coefficient alpha and the associations of the scales with each other and covariates with Pearson correlation coefficients.
We examined the relation of each trait with mortality in proportional hazards models (Cox, 1972). The core model for each trait included the trait score and terms for age, gender, and education. In a secondary analysis, we included all five traits in the same model. In variations of the core model for each separate trait, we added terms for the following: the interaction of Gender × Trait; the interaction of Age × Trait; number of medical conditions; baseline measures of global cognition and of lower limb function; history of smoking tobacco, alcoholic drinks consumed in the past year, and body mass index and a term for body mass index squared (because body mass index had a nonlinear association with mortality); and number of medical conditions, global cognition, lower limb function, tobacco use, alcohol use, and body mass index. In variations of the latter model, we substituted specific medical conditions for number of medical conditions, with a separate analysis for each of seven conditions, and we then substituted measures of specific cognitive function for the global cognitive measure, with a separate analysis for each of five cognitive measures.
We found no graphic or analytic evidence of nonlinearity or nonproportionality. Programming was done in SAS (SAS Institute Inc., 2000).
Results
Psychometric Information on Personality Traits
Each trait measure had an approximately symmetric distribution. Table 1 provides psychometric information on the trait measures in Religious Orders Study participants and in the normative cohort (Costa & McCrae, 1992). The mean level of each trait is roughly comparable in the two cohorts, except for a slightly lower level of neuroticism in Religious Orders Study participants. We assessed the internal consistency of each trait scale with Cronbach's coefficient alpha. As shown in Table 1, these coefficients ranged from.67 to.81 in the current study, which suggests a moderately high level of internal consistency and is consistent with estimates of internal consistency in the normative cohort.
Neuroticism was negatively related to the other four traits, with correlations ranging from −.13 to −.39, and the remaining four traits were positively intercorrelated, with correlations ranging from.04 to.29 (Table 2). Overall, the traits were not strongly correlated with age, education, or diverse measures of health and function. Women had higher levels of neuroticism, openness, agreeableness, and conscientiousness than men but did not differ from men in extraversion.
Personality and Mortality
As of February 2003, after an average of 5.1 years of observation, 182 (21%) persons had died and 701 survived. Table 3 provides crude data on these two subgroups at baseline. Those who died were older and more apt to be male than the survivors; they had a higher level of neuroticism but a lower level of the other four traits; and they had generally lower levels of health and function.
We examined the relation of each trait to mortality in separate proportional hazards analyses controlling for age, gender, and education. In these analyses (Table 4), neuroticism was positively related to risk of death, and extraversion and conscientiousness were inversely related to risk of death, with a similar trend for agreeableness and no effect for openness. Thus, a person with a high neuroticism score (score = 25, 90th percentile) was 95% more likely to die than a person with a low score (score = 11, 10th percentile); a person with a high extraversion score (score = 35, 90th percentile) was 75% less likely to die than a person with a low score (score = 20, 10th percentile); and a conscientiousness score of 40 (90th percentile) was associated with a 48% reduced risk of death compared with a score of 28 (10th percentile). Figure 1 illustrates these effects for neuroticism (upper panel) and conscientiousness (lower panel).
Because the five traits are not entirely orthogonal, we constructed a model that included all five traits, as well as age, gender, and education. In this analysis, neuroticism was related to mortality (relative risk, or RR = 1.037; 95% confidence interval or CI = 1.006, 1.068), but extraversion (RR = 0.978; 95% CI = 0.948, 1.010), openness (RR = 1.015; 95% CI = 0.981, 1.049), agreeableness (RR = 0.997; 95% CI = 0.954, 1.041), and conscientiousness (RR = 0.982; 95% CI = 0.952, 1.013) were not.
Because gender is related to both personality (Costa & McCrae, 1992) and mortality, we considered the possibility that gender might modify the association between personality and mortality risk. To test this idea, we repeated the core analysis of each trait with a term added for the interaction of gender with the trait score. Gender did not interact with any trait (all p >.30), suggesting that the association between personality and mortality does not differ in men compared with women. In a similar series of analyses, we found no evidence that age modified the relation of personality to mortality in this cohort (all p >.30).
In another series of analyses, we considered diverse indicators of health at baseline that might help account for the association of personality with mortality (Table 5). We added terms to the core model for number of chronic medical conditions at baseline (Model 1); for global measures of cognitive function and lower limb function because these functions are compromised by disease and are related to mortality (Wilson, Beckett, et al., 2003; Wilson, Schneider, Beckett, Evans, & Bennett, 2002; Model 2); for tobacco use, alcohol use, and body mass because of their association with health and possible relations to personality (Model 3); and for all of these factors simultaneously (Model 4). The associations of neuroticism and conscientiousness with mortality were not substantially changed in these analyses. The association of extraversion with mortality was not strongly affected when medical conditions (Model 1) or health-related variables (Model 3) were controlled for, but it was attenuated in models that controlled for level of cognitive and motor function (Models 2 and 4).
In a final set of models, we examined the possibility that use of global measures of chronic illness and cognitive function may have obscured a more specific confounding effect. First, we repeated Model 4 by using the presence of a specific medical condition (e.g., stroke) in place of number of conditions, with separate models for each of seven conditions. Second, we repeated Model 4, controlling for specific forms of cognition (i.e., episodic memory, semantic memory, working memory, perceptual speed, and visuospatial ability) rather than global cognition, with separate analyses for each of five cognitive measures. The results of these analyses were comparable with those of the original versions of Model 4.
Discussion
We measured the five principal dimensions of personality in a cohort of more than 800 older persons who were then followed for an average of approximately 5 years. We found that persons with a high level of neuroticism were nearly twice as likely to die as persons low in the trait. We also found that a high level of conscientiousness was associated with an approximately 50% reduction in mortality risk in comparison with a low level of conscientiousness. The results support the idea that personality is related to mortality in old age.
In persons with chronic renal conditions (Christensen et al., 2002) or cardiac conditions (Denollet, Sys, & Brutsaert, 1995), neuroticism has been associated with an increased risk of death, consistent with the results of this study. Further, neuroticism predicts the occurrence of negative emotional states that are related to mortality (Barefoot & Schroll, 1996; Kubzansky et al., 1997; Miller et al., 1996; Wilson, Bienias, et al., 2003). With few exceptions (Almada et al., 1991), however, previous studies of community-dwelling older persons have not observed an association between high neuroticism and increased mortality (Costa & McCrae, 1987; Korten et al., 1999; Maier & Smith, 1999; Martin & Friedman, 2000), as already noted. We think a combination of factors may have contributed to these mixed findings. One issue is that diverse scales have been used to assess neuroticism in prior studies, including short forms of established scales (Korten et al., 1999; Maier & Smith, 1999), a measure based exclusively on somatic symptoms (Somervell et al., 1989), and a measure reconstructed from archival data (Martin & Friedman, 2000), suggesting that inconsistent findings have been partly due to variation between studies in the definition of the trait and how adequately it was assessed. Indeed, neuroticism has multiple dimensions or facets (Costa & McRae, 1992), and it is quite possible that some facets bear a stronger relation to mortality than others. For example, in a previous study of this cohort (Wilson, Bienias, et al., 2003), measures of anxiety, suppressed anger, and depressive symptoms, suggestive of the neuroticism facets of anxiety, angry hostility, and depression from the NEO Personality Inventory (Costa & McRae, 1992), were related to mortality. In contrast, measures of other negative emotions (e.g., anger proneness) were not related.
Another complication is that we assessed mortality in old age whereas some other studies have examined mortality in middle age or across the life span (Almada et al., 1991; Huppert & Whittington, 1995; Somervell et al., 1989). Although recent studies have shown that some change in neuroticism does occur in adulthood (Mroczek & Spiro, 2003; Small, Hertzog, Hultsch, & Dixon, 2003), 6-year test–retest correlations have approached unity in some studies (Costa & McCrae, 1988; Hultsch, Hertzog, Small, & Dixon, 1999), suggesting that the trait is relatively stable. Therefore, its deleterious impact on health may be due to cumulative effects that do not emerge until old age.
The bases of the association between neuroticism and risk of death are uncertain. Prior research has not suggested that neuroticism is related to specific conditions such as coronary heart disease (Shekelle, Vernon, & Ostfeld, 1991) or cancer (Schapiro et al., 2001), though it is possible that facets of neuroticism are. Neuroticism has been linked with impaired immune function (Sternberg, Chrousos, Wilder, & Gold, 1992), which might affect morbidity and mortality, especially in old age. Neuroticism has also been related to compromised regulation of the hypothalamic–pituitary–adrenal stress axis with resultant changes in brain structure and function, particularly in the limbic system, and in forms of learning and memory mediated by these systems (Frodl et al., 2002; Lupien et al., 1997; Sheline, Wang, Gado, Csemansky, & Vannier, 1996; Wilson, Evans, et al., 2003; Wilson et al., 2004). Thus, neuroticism over a lifetime may compromise multiple physiological and functional systems and thereby increase vulnerability to disease in old age.
We found that conscientiousness was associated with reduced mortality. In Terman's cohort of children with a high level of intelligence, conscientiousness in childhood and in adulthood was associated with reduced risk of death (Friedman et al., 1993; Martin & Friedman, 2000). A similar effect was observed in persons with chronic renal insufficiency (Christensen et al., 2002). In this cohort, but not in a previous study (Christensen et al., 2002), the effect of conscientiousness was attenuated after adjustment for other personality traits, suggesting that the effect may be due in part to the association of conscientiousness with other personality traits. Both conscientiousness and neuroticism have been related to adherence to medical regimens and to other health-related behaviors (Booth-Kewley & Vickers, 1994; Wiebe & Christensen, 1996). We did not find that controlling for health-related behaviors (i.e., alcohol and tobacco use) or for other indicators of health strongly affected results, but it is certainly possible that other health-related behaviors are contributing to the observed associations. Further research is needed to clarify the mechanisms underlying the association of conscientiousness with mortality.
Extraversion was associated with reduced mortality in some models. The association was no longer significant, however, when we controlled for other personality traits or for level of cognitive and motor function, suggesting, in the context of other negative studies (Maier & Smith, 1999; Martin & Friedman, 2000; Korten et al., 1999), that extraversion is not independently related to mortality.
Neither openness nor agreeableness was significantly related to mortality in any analysis, which is generally consistent with prior research (Christensen et al., 2002; Maier & Smith, 1999), though an association between an archive-based measure of agreeableness and mortality was observed in a follow-up study of Terman's cohort (Martin & Friedman, 2000).
A unique feature of this study is that the participants are older Catholic clergy members. It is possible that homogeneity within this cohort with respect to unmeasured socioeconomically related factors (e.g., access to health care, diet) may have enhanced our ability to detect an association between personality and mortality by reducing the possibly confounding influence of these factors.
The principal strength of this study is that we assessed personality in a large cohort of older persons free of dementia by using psychometrically established measures of the five principal dimensions of personality. The principal limitation of the findings is that they are based on a selected group of older persons who differ from the general population of older persons. Therefore, it will be important to replicate these findings in cohorts that are more diverse and representative of older persons in the U.S. population. Another limitation is that our use of all-cause mortality as an outcome may have obscured associations between personality and specific mortality outcomes.
In summary, the results suggest that two of the five principal dimensions of personality, neuroticism and conscientiousness, are related to mortality in old age, even after multiple indicators of physical health are controlled for. The results provide strong evidence that health in old age depends on psychological characteristics as well as physical ones. Because neuroticism and conscientiousness are traits that we all possess to greater or lesser degree, understanding the pathways linking them with morbidity and mortality in old age could have substantial public health implications.
Decision Editor: Margie E. Lachman, PhD
. | Religious Orders Study . | . | . | Normative Cohort . | . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|
Trait . | M . | (SD) . | α . | M . | (SD) . | α . | ||||
Neuroticism | 16.7 | (5.7) | .81 | 19.1 | (7.7) | .86 | ||||
Extraversion | 27.7 | (5.6) | .78 | 27.7 | (5.9) | .77 | ||||
Openness | 26.3 | (5.2) | .69 | 27.0 | (5.8) | .73 | ||||
Agreeableness | 34.0 | (3.9) | .67 | 32.8 | (5.0) | .68 | ||||
Conscientiousness | 34.0 | (5.0) | .81 | 34.6 | (5.9) | .81 |
. | Religious Orders Study . | . | . | Normative Cohort . | . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|
Trait . | M . | (SD) . | α . | M . | (SD) . | α . | ||||
Neuroticism | 16.7 | (5.7) | .81 | 19.1 | (7.7) | .86 | ||||
Extraversion | 27.7 | (5.6) | .78 | 27.7 | (5.9) | .77 | ||||
Openness | 26.3 | (5.2) | .69 | 27.0 | (5.8) | .73 | ||||
Agreeableness | 34.0 | (3.9) | .67 | 32.8 | (5.0) | .68 | ||||
Conscientiousness | 34.0 | (5.0) | .81 | 34.6 | (5.9) | .81 |
Note: The alpha denotes the coefficient alpha, a measure of internal consistency.
. | Religious Orders Study . | . | . | Normative Cohort . | . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|
Trait . | M . | (SD) . | α . | M . | (SD) . | α . | ||||
Neuroticism | 16.7 | (5.7) | .81 | 19.1 | (7.7) | .86 | ||||
Extraversion | 27.7 | (5.6) | .78 | 27.7 | (5.9) | .77 | ||||
Openness | 26.3 | (5.2) | .69 | 27.0 | (5.8) | .73 | ||||
Agreeableness | 34.0 | (3.9) | .67 | 32.8 | (5.0) | .68 | ||||
Conscientiousness | 34.0 | (5.0) | .81 | 34.6 | (5.9) | .81 |
. | Religious Orders Study . | . | . | Normative Cohort . | . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|
Trait . | M . | (SD) . | α . | M . | (SD) . | α . | ||||
Neuroticism | 16.7 | (5.7) | .81 | 19.1 | (7.7) | .86 | ||||
Extraversion | 27.7 | (5.6) | .78 | 27.7 | (5.9) | .77 | ||||
Openness | 26.3 | (5.2) | .69 | 27.0 | (5.8) | .73 | ||||
Agreeableness | 34.0 | (3.9) | .67 | 32.8 | (5.0) | .68 | ||||
Conscientiousness | 34.0 | (5.0) | .81 | 34.6 | (5.9) | .81 |
Note: The alpha denotes the coefficient alpha, a measure of internal consistency.
Trait . | Extrav. . | Open. . | Agree. . | Consc. . | Age . | Educ. . | Cog. F. . | Lower Limb F. . | Chron. Illness . | Alc. Use . | Ever Smoked . | Body Mass . |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Neuroticism | −.39 | −.13 | −.34 | −.30 | .04 | −.15 | −.19 | −.12 | .06 | −.15 | −.01 | −.00 |
Extraversion | .28 | .26 | .27 | −.13 | .19 | .12 | .18 | −.03 | .08 | −.03 | .04 | |
Openness | .16 | .04 | −.17 | .32 | .29 | .15 | .05 | .12 | .04 | .03 | ||
Agreeableness | .29 | −.06 | .13 | .17 | .06 | −.05 | .05 | −.02 | .05 | |||
Conscientiousness | −.07 | .04 | .08 | .06 | −.07 | −.02 | −.09 | −.10 |
Trait . | Extrav. . | Open. . | Agree. . | Consc. . | Age . | Educ. . | Cog. F. . | Lower Limb F. . | Chron. Illness . | Alc. Use . | Ever Smoked . | Body Mass . |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Neuroticism | −.39 | −.13 | −.34 | −.30 | .04 | −.15 | −.19 | −.12 | .06 | −.15 | −.01 | −.00 |
Extraversion | .28 | .26 | .27 | −.13 | .19 | .12 | .18 | −.03 | .08 | −.03 | .04 | |
Openness | .16 | .04 | −.17 | .32 | .29 | .15 | .05 | .12 | .04 | .03 | ||
Agreeableness | .29 | −.06 | .13 | .17 | .06 | −.05 | .05 | −.02 | .05 | |||
Conscientiousness | −.07 | .04 | .08 | .06 | −.07 | −.02 | −.09 | −.10 |
Note: For correlations with an absolute value of.07 or more, p < 0.05. Extrav. = extraversion; Open. = openness; Agree. = agreeableness; Consc. = conscientiousness; Educ. = education; Cog. F. = cognitive function; Lower Limb F. = lower limb function; Chron. Illness = chronic illness; Alc. Use = alcohol use.
Trait . | Extrav. . | Open. . | Agree. . | Consc. . | Age . | Educ. . | Cog. F. . | Lower Limb F. . | Chron. Illness . | Alc. Use . | Ever Smoked . | Body Mass . |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Neuroticism | −.39 | −.13 | −.34 | −.30 | .04 | −.15 | −.19 | −.12 | .06 | −.15 | −.01 | −.00 |
Extraversion | .28 | .26 | .27 | −.13 | .19 | .12 | .18 | −.03 | .08 | −.03 | .04 | |
Openness | .16 | .04 | −.17 | .32 | .29 | .15 | .05 | .12 | .04 | .03 | ||
Agreeableness | .29 | −.06 | .13 | .17 | .06 | −.05 | .05 | −.02 | .05 | |||
Conscientiousness | −.07 | .04 | .08 | .06 | −.07 | −.02 | −.09 | −.10 |
Trait . | Extrav. . | Open. . | Agree. . | Consc. . | Age . | Educ. . | Cog. F. . | Lower Limb F. . | Chron. Illness . | Alc. Use . | Ever Smoked . | Body Mass . |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Neuroticism | −.39 | −.13 | −.34 | −.30 | .04 | −.15 | −.19 | −.12 | .06 | −.15 | −.01 | −.00 |
Extraversion | .28 | .26 | .27 | −.13 | .19 | .12 | .18 | −.03 | .08 | −.03 | .04 | |
Openness | .16 | .04 | −.17 | .32 | .29 | .15 | .05 | .12 | .04 | .03 | ||
Agreeableness | .29 | −.06 | .13 | .17 | .06 | −.05 | .05 | −.02 | .05 | |||
Conscientiousness | −.07 | .04 | .08 | .06 | −.07 | −.02 | −.09 | −.10 |
Note: For correlations with an absolute value of.07 or more, p < 0.05. Extrav. = extraversion; Open. = openness; Agree. = agreeableness; Consc. = conscientiousness; Educ. = education; Cog. F. = cognitive function; Lower Limb F. = lower limb function; Chron. Illness = chronic illness; Alc. Use = alcohol use.
Characteristic . | Survived (n = 701) . | Died (n = 182) . | p Value . |
---|---|---|---|
Age, years | 73.7 (6.5) | 80.5 (6.4) | <.001 |
Education, years | 18.1 (3.3) | 18.4 (3.5) | .337 |
Women, % | 72.8 | 55.0 | <.001 |
Neuroticism | 16.4 (5.5) | 17.7 (6.4) | .020 |
Extraversion | 28.1 (5.7) | 26.3 (5.2) | <.001 |
Openness | 26.6 (5.3) | 25.5 (4.8) | .014 |
Agreeableness | 34.2 (3.9) | 33.4 (3.7) | .011 |
Conscientiousness | 34.3 (5.0) | 32.7 (4.5) | <.001 |
Cognitive function | 0.17 (0.49) | −0.15 (0.50) | <.001 |
Lower limb function | 3.33 (1.09) | 2.57 (1.24) | <.001 |
Chronic illnessa | 1.04 (0.99) | 1.42 (1.05) | <.001 |
Current alcohol useb | 0.63 (1.86) | 0.58 (1.75) | .784 |
Ever smoked, % | 17.7 | 25.8 | .014 |
Body mass index, kg/m2 | 27.6 (5.4) | 26.6 (5.4) | .031 |
Characteristic . | Survived (n = 701) . | Died (n = 182) . | p Value . |
---|---|---|---|
Age, years | 73.7 (6.5) | 80.5 (6.4) | <.001 |
Education, years | 18.1 (3.3) | 18.4 (3.5) | .337 |
Women, % | 72.8 | 55.0 | <.001 |
Neuroticism | 16.4 (5.5) | 17.7 (6.4) | .020 |
Extraversion | 28.1 (5.7) | 26.3 (5.2) | <.001 |
Openness | 26.6 (5.3) | 25.5 (4.8) | .014 |
Agreeableness | 34.2 (3.9) | 33.4 (3.7) | .011 |
Conscientiousness | 34.3 (5.0) | 32.7 (4.5) | <.001 |
Cognitive function | 0.17 (0.49) | −0.15 (0.50) | <.001 |
Lower limb function | 3.33 (1.09) | 2.57 (1.24) | <.001 |
Chronic illnessa | 1.04 (0.99) | 1.42 (1.05) | <.001 |
Current alcohol useb | 0.63 (1.86) | 0.58 (1.75) | .784 |
Ever smoked, % | 17.7 | 25.8 | .014 |
Body mass index, kg/m2 | 27.6 (5.4) | 26.6 (5.4) | .031 |
Note: All data are presented as mean (SD) unless otherwise indicated.
aNumber of medical conditions present at baseline.
bNumber of alcoholic drinks consumed in the past year, log transformed because of the skewed distribution.
Characteristic . | Survived (n = 701) . | Died (n = 182) . | p Value . |
---|---|---|---|
Age, years | 73.7 (6.5) | 80.5 (6.4) | <.001 |
Education, years | 18.1 (3.3) | 18.4 (3.5) | .337 |
Women, % | 72.8 | 55.0 | <.001 |
Neuroticism | 16.4 (5.5) | 17.7 (6.4) | .020 |
Extraversion | 28.1 (5.7) | 26.3 (5.2) | <.001 |
Openness | 26.6 (5.3) | 25.5 (4.8) | .014 |
Agreeableness | 34.2 (3.9) | 33.4 (3.7) | .011 |
Conscientiousness | 34.3 (5.0) | 32.7 (4.5) | <.001 |
Cognitive function | 0.17 (0.49) | −0.15 (0.50) | <.001 |
Lower limb function | 3.33 (1.09) | 2.57 (1.24) | <.001 |
Chronic illnessa | 1.04 (0.99) | 1.42 (1.05) | <.001 |
Current alcohol useb | 0.63 (1.86) | 0.58 (1.75) | .784 |
Ever smoked, % | 17.7 | 25.8 | .014 |
Body mass index, kg/m2 | 27.6 (5.4) | 26.6 (5.4) | .031 |
Characteristic . | Survived (n = 701) . | Died (n = 182) . | p Value . |
---|---|---|---|
Age, years | 73.7 (6.5) | 80.5 (6.4) | <.001 |
Education, years | 18.1 (3.3) | 18.4 (3.5) | .337 |
Women, % | 72.8 | 55.0 | <.001 |
Neuroticism | 16.4 (5.5) | 17.7 (6.4) | .020 |
Extraversion | 28.1 (5.7) | 26.3 (5.2) | <.001 |
Openness | 26.6 (5.3) | 25.5 (4.8) | .014 |
Agreeableness | 34.2 (3.9) | 33.4 (3.7) | .011 |
Conscientiousness | 34.3 (5.0) | 32.7 (4.5) | <.001 |
Cognitive function | 0.17 (0.49) | −0.15 (0.50) | <.001 |
Lower limb function | 3.33 (1.09) | 2.57 (1.24) | <.001 |
Chronic illnessa | 1.04 (0.99) | 1.42 (1.05) | <.001 |
Current alcohol useb | 0.63 (1.86) | 0.58 (1.75) | .784 |
Ever smoked, % | 17.7 | 25.8 | .014 |
Body mass index, kg/m2 | 27.6 (5.4) | 26.6 (5.4) | .031 |
Note: All data are presented as mean (SD) unless otherwise indicated.
aNumber of medical conditions present at baseline.
bNumber of alcoholic drinks consumed in the past year, log transformed because of the skewed distribution.
Personality Trait . | Relative Risk . | 95% Confidence Interval . |
---|---|---|
Neuroticism | 1.049 | 1.022, 1.077 |
Extraversion | 0.964 | 0.937, 0.991 |
Openness | 1.005 | 0.973, 1.038 |
Agreeableness | 0.964 | 0.927, 1.002 |
Conscientiousness | 0.968 | 0.940, 0.996 |
Personality Trait . | Relative Risk . | 95% Confidence Interval . |
---|---|---|
Neuroticism | 1.049 | 1.022, 1.077 |
Extraversion | 0.964 | 0.937, 0.991 |
Openness | 1.005 | 0.973, 1.038 |
Agreeableness | 0.964 | 0.927, 1.002 |
Conscientiousness | 0.968 | 0.940, 0.996 |
Note: Relative risk is for a 1-point increase in each personality trait measure.
Personality Trait . | Relative Risk . | 95% Confidence Interval . |
---|---|---|
Neuroticism | 1.049 | 1.022, 1.077 |
Extraversion | 0.964 | 0.937, 0.991 |
Openness | 1.005 | 0.973, 1.038 |
Agreeableness | 0.964 | 0.927, 1.002 |
Conscientiousness | 0.968 | 0.940, 0.996 |
Personality Trait . | Relative Risk . | 95% Confidence Interval . |
---|---|---|
Neuroticism | 1.049 | 1.022, 1.077 |
Extraversion | 0.964 | 0.937, 0.991 |
Openness | 1.005 | 0.973, 1.038 |
Agreeableness | 0.964 | 0.927, 1.002 |
Conscientiousness | 0.968 | 0.940, 0.996 |
Note: Relative risk is for a 1-point increase in each personality trait measure.
. | Model 1 . | . | Model 2 . | . | Model 3 . | . | Model 4 . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Personality Trait . | RR . | 95% CI . | RR . | 95% CI . | RR . | 95% CI . | RR . | 95% CI . | ||||
Neuroticism | 1.044 | 1.017, 1.071 | 1.042 | 1.014, 1.070 | 1.046 | 1.018, 1.074 | 1.035 | 1.007, 1.064 | ||||
Extraversion | 0.968 | 0.941, 0.995 | 0.970 | 0.943, 0.997 | 0.964 | 0.936, 0.992 | 0.973 | 0.946, 1.002 | ||||
Openness | 1.003 | 0.971, 1.036 | 1.016 | 0.984, 1.050 | 1.004 | 0.971, 1.038 | 1.011 | 0.977, 1.046 | ||||
Agreeableness | 0.964 | 0.927, 1.003 | 0.964 | 0.928, 1.002 | 0.966 | 0.930, 1.004 | 0.968 | 0.931, 1.007 | ||||
Conscientiousness | 0.967 | 0.938, 0.996 | 0.966 | 0.938, 0.994 | 0.970 | 0.943, 0.998 | 0.967 | 0.939, 0.997 |
. | Model 1 . | . | Model 2 . | . | Model 3 . | . | Model 4 . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Personality Trait . | RR . | 95% CI . | RR . | 95% CI . | RR . | 95% CI . | RR . | 95% CI . | ||||
Neuroticism | 1.044 | 1.017, 1.071 | 1.042 | 1.014, 1.070 | 1.046 | 1.018, 1.074 | 1.035 | 1.007, 1.064 | ||||
Extraversion | 0.968 | 0.941, 0.995 | 0.970 | 0.943, 0.997 | 0.964 | 0.936, 0.992 | 0.973 | 0.946, 1.002 | ||||
Openness | 1.003 | 0.971, 1.036 | 1.016 | 0.984, 1.050 | 1.004 | 0.971, 1.038 | 1.011 | 0.977, 1.046 | ||||
Agreeableness | 0.964 | 0.927, 1.003 | 0.964 | 0.928, 1.002 | 0.966 | 0.930, 1.004 | 0.968 | 0.931, 1.007 | ||||
Conscientiousness | 0.967 | 0.938, 0.996 | 0.966 | 0.938, 0.994 | 0.970 | 0.943, 0.998 | 0.967 | 0.939, 0.997 |
Notes: Relative risk is for a 1-point increase in each personality trait measure. RR = relative risk; CI = confidence interval. Model 1 controlled for number of medical conditions at baseline; Model 2 controlled for baseline levels of global cognition and lower limb function; Model 3 controlled for past use of alcohol and tobacco and for baseline body mass; and Model 4 controlled for all six indicators of health and function.
. | Model 1 . | . | Model 2 . | . | Model 3 . | . | Model 4 . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Personality Trait . | RR . | 95% CI . | RR . | 95% CI . | RR . | 95% CI . | RR . | 95% CI . | ||||
Neuroticism | 1.044 | 1.017, 1.071 | 1.042 | 1.014, 1.070 | 1.046 | 1.018, 1.074 | 1.035 | 1.007, 1.064 | ||||
Extraversion | 0.968 | 0.941, 0.995 | 0.970 | 0.943, 0.997 | 0.964 | 0.936, 0.992 | 0.973 | 0.946, 1.002 | ||||
Openness | 1.003 | 0.971, 1.036 | 1.016 | 0.984, 1.050 | 1.004 | 0.971, 1.038 | 1.011 | 0.977, 1.046 | ||||
Agreeableness | 0.964 | 0.927, 1.003 | 0.964 | 0.928, 1.002 | 0.966 | 0.930, 1.004 | 0.968 | 0.931, 1.007 | ||||
Conscientiousness | 0.967 | 0.938, 0.996 | 0.966 | 0.938, 0.994 | 0.970 | 0.943, 0.998 | 0.967 | 0.939, 0.997 |
. | Model 1 . | . | Model 2 . | . | Model 3 . | . | Model 4 . | . | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Personality Trait . | RR . | 95% CI . | RR . | 95% CI . | RR . | 95% CI . | RR . | 95% CI . | ||||
Neuroticism | 1.044 | 1.017, 1.071 | 1.042 | 1.014, 1.070 | 1.046 | 1.018, 1.074 | 1.035 | 1.007, 1.064 | ||||
Extraversion | 0.968 | 0.941, 0.995 | 0.970 | 0.943, 0.997 | 0.964 | 0.936, 0.992 | 0.973 | 0.946, 1.002 | ||||
Openness | 1.003 | 0.971, 1.036 | 1.016 | 0.984, 1.050 | 1.004 | 0.971, 1.038 | 1.011 | 0.977, 1.046 | ||||
Agreeableness | 0.964 | 0.927, 1.003 | 0.964 | 0.928, 1.002 | 0.966 | 0.930, 1.004 | 0.968 | 0.931, 1.007 | ||||
Conscientiousness | 0.967 | 0.938, 0.996 | 0.966 | 0.938, 0.994 | 0.970 | 0.943, 0.998 | 0.967 | 0.939, 0.997 |
Notes: Relative risk is for a 1-point increase in each personality trait measure. RR = relative risk; CI = confidence interval. Model 1 controlled for number of medical conditions at baseline; Model 2 controlled for baseline levels of global cognition and lower limb function; Model 3 controlled for past use of alcohol and tobacco and for baseline body mass; and Model 4 controlled for all six indicators of health and function.
This research was supported by the National Institute on Aging under Grants R01 AG15819 and P30 AG10161.
We are indebted to the hundreds of nuns, priests, and brothers from the following groups who participated in the Religious Orders Study: Archdiocesan priests of Chicago, Dubuque, and Milwaukee; Benedictine Monks, Lisle, IL and Collegeville, MN; Benedictine Sisters of Erie, Erie, PA; Benedictine Sisters of the Sacred Heart, Lisle, IL; Capuchins, Appleton, WI; Christian Brothers, Chicago, IL and Memphis, TN; Diocesan priests of Gary, IN; Dominicans of River Forest, IL; Felician Sisters, Chicago; IL; Franciscan Handmaids of Mary, New York, NY; Franciscans, Chicago, IL; Holy Spirit Missionary Sisters, Techny, IL; Maryknolls, Los Altos, CA, and Maryknoll, NY; Norbertines, DePere, WI; Oblate Sisters of Providence, Baltimore, MD; Passionists, Chicago, IL; Presentation Sisters, BVM, Dubuque, IA; Servites, Chicago, IL; Sinsinawa Dominican Sisters, Chicago, IL, and Sinsinawa, WI; Sisters of Charity, BVM, Chicago, IL, and Dubuque, IA; Sisters of the Holy Family, New Orleans, LA; Sisters of the Holy Family of Nazareth, Des Plaines, IL; Sisters of Mercy of the Americas, Chicago, IL, Aurora, IL, and Erie, PA; Sisters of St. Benedict, St. Cloud and St. Joseph, MN; Sisters of St. Casimir, Chicago, IL; Sisters of St. Francis of Mary Immaculate, Joliet, IL; Sisters of St. Joseph of LaGrange, LaGrange Park, IL; Society of Divine Word, Techny, IL; Trappists, Gethsemane, KY, and Peosta, IA; and Wheaton Franciscan Sisters, Wheaton, IL. We also thank Julie Bach, MSW, coordinator of the Religious Orders Study; Liping Gu, MS, for programming; George Dombrowski, MS, and Greg Klein for data management; and Valerie J. Young for preparing the manuscript.
References
Almada, S. J., Zonderman, A. B., Shekelle, R. B., Dyer, A. R., Daviglus, M. L., Costa, P. T., et al.
Barefoot, J. C., Schroll, M.
Bennett, D. A., Wilson, R. S., Schneider, J. A., Evans, D. A., Beckett, L. A., Aggarwal, N. T., et al.
Booth-Kewley, S., Vickers, R. R.
Christensen, A. J., Ehlers, S. L., Wiebe, J. S., Moran, P. J., Raichle, K., Ferneyhough, K., et al.
Costa, P. T., McCrae, R. R.
Costa, P. T., McCrae, R. R.
Costa, P. T., McCrae, R. R.
Cox, D. R.
Crystal, S., Shea, D.
De Beurs, E., Beekman, A. T. F., Deeg, D. J. H., Van Dyck, R., Van Tilburg, W.
Denollet, J., Sys, S. U., Brutsaert, D. L.
Digman, J. M.
Friedman, H. S., Tucker, J. S., Tomlinson-Keasey, C., Schwartz, J. E., Wingard, D. L., Criqui, M. H.
Frodl, T., Meisenzahl, E. M., Zetche, T., Born, C., Groll, C., Jager, M., et al.
Gallo, L. C., Mathews, K. A.
Goldberg, L. R.
House, J. S., Kessler, R. C., Herzog, A. R.
Hultsch, D., Hertzog, C., Small, B., Dixon, R.
Huppert, F. A., Whittington, J. E.
Korten, A. E., Jorm, A. F., Jiao, Z., Letenneur, L., Jacomb, A., Henderson A. S., et al.
Kubzansky, L. D., Kawachi, I., Spiro, A., Weiss, S. T., Vokonas, P. S., Sparrow, D.
Lupien, S. J., Gaudreau, S., Tchiteya, B. M., Maheu, F., Sharma, S., Nair, N. P. V., et al.
Maier, H., Smith, J.
Martin, L. R., Friedman, H. S.
Maruta, T., Colligan, R. C., Malinchoc, M., Offord, K. P.
McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., Stadlan, E.
Miller, T. Q., Smith, T. W., Turner, C. W., Guijarro, M. L., Hallet, A. J.
Mroczek, D. K., Spiro, A.
Ormel, J., Wohlfarth, T.
Robert, S., House, J. S.
Schapiro, I. R., Ross-Petersen, L., Saelan, H., Garde, K., Olsen, J. H., Johansen, C.
Shekelle, R. B., Vernon, S. W., Ostfeld, A. M.
Sheline, Y. I., Wang, P. W., Gado, M. H., Csemansky, J. G., Vannier, M. W.
Small, B. J., Hertzog, C., Hultsch, D. F., Dixon, R. A.
Smith, T. W., Williams, P. G.
Somervell, P. D., Kaplan, B. H., Heiss, G., Tyroler, H. A., Kleinbaum, D. G., Obrist, P. A.
Sternberg, E. M., Chrousos, G. P., Wilder, R. L., Gold, P. W.
Watson, D., Clark, L. A.
Watson, D., Pennebaker, J. W.
Wiebe, J. S., Christensen, A. J.
Wilson, R. S., Barnes, L. L., Mendes de Leon, C. F., Aggarwal, N. T., Schneider, J. A., Bach, J., et al.
Wilson, R. S., Beckett, L. A., Barnes, L. L., Schneider, J. A., Bach, J., Evans, D. A., et al.
Wilson, R. S., Beckett, L. A., Bienias, J. L., Evans, D. A., Bennett, D. A.
Wilson, R. S., Bienias, J. L., Berry-Kravis, E., Evans, D. A., Bennett, D. A.
Wilson, R. S., Bienias, J. L., Mendes de Leon, C. F., Evans, D. A., Bennett, D. A.
Wilson, R. S., Evans, D. A., Bienias, J. L., Mendes de Leon, C. F., Schneider, J. A., Bennett, D. A.
Wilson, R. S., Fleischman, D. A., Myers, R. A., Bennett, D. A., Bienias, J. L., Gilley, D. W., et al.
Wilson, R. S., Mendes de Leon, C. F., Barnes, L. L., Schneider, J. A., Bienias, J. L., Evans, D. A., et al.
Wilson, R. S., Schneider, J. A., Beckett, L. A., Evans, D. A., Bennett, D. A.