Attendance at Religious Services, Interleukin-6, and Other Biological
Parameters of Immune Function in Older Adults
International Journal of Psychiatry in Medicine 1997; 27 27:233-250
Harold G. Koenig M.D., M.H.Sc.
Associate Professor of Psychiatry
Director, Center for the Study of Religion/Spirituality and Health
DukeUniversityMedicalCenter
Durham, North Carolina
Harvey Jay Cohen, M.D.
Professor of Medicine
Director, Center for the Study of Aging and Human Development
DukeUniversityMedicalCenter
Linda K. George, Ph.D.
Professor of Sociology
Associate Director, Center for the Study of Aging and Human Development
DukeUniversityMedicalCenter
Judith C. Hays, R.N., Ph.D.
Assistant Research Professor of Geriatric Psychiatry
Senior Fellow, Center for the Study of Aging and Human Development
DukeUniversityMedicalCenter
David B. Larson, M.D., M.S.P.H.
Adjunct Professor of Psychiatry
DukeUniversityMedicalCenter
Dan G. Blazer, M.D., Ph.D.
JB Gibbons Professor of Psychiatry
Dean of Medical Education
DukeUniversityMedicalCenter
Duke University School of Medicine
The research upon which this publication is based was performed pursuant to Contract number N01-AG-1-2102 with the National Institute on Aging, in support of the Established Populations for Epidemiologic Studies of the Elderly (Duke). Funding also in part provided by the John Templeton Foundation, Radnor, Penn, Monarch Pharmaceuticals, a division of King Pharmaceuticals, Bristol, TN (Dr. Larson), and a NIMH Clinical Mental Health Academic Award MH01138 (Dr. Koenig). Correspondence to Dr. Koenig, Box 3400, DukeMedicalCenter, Durham, NC27710.
Abstract
Objective: First, to examine and explain the relationship between religious service attendance and plasma Interleukin-6 (IL-6) levels, and second, to examine the relationship between religious attendance and other immune-system regulators and inflammatory substances. Methods: During the third in-person interview (1992) of the Establishment of Populations for Epidemiologic Studies of the Elderly (EPESE) project, Duke site, 1718 subjects age 65 or over had blood drawn for analysis of immune regulators and inflammatory factors, including IL-6 measurements. IL-6 was examined both as a continuous variable and at a cutoff of 5 pg/ml. Information on attendance at religious services was available from the 1992 interview and two prior interviews (1986 and 1989). Results: Religious attendance was inversely related to high IL-6 levels (> 5 pg/ml), but not to IL-6 measured as a continuous variable. Bivariate analyses revealed that high religious attendance in 1989 predicted a lower proportion of subjects with high IL-6 in 1992 (beta -.10, p=.01). High religious attendance in 1992 also predicted a lower proportion of subjects with high IL-6 levels in 1992 (beta -.14, p=.0005). When age, sex, race, education, chronic illnesses, and physical functioning were controlled, 1989 religious attendance weakened as a predictor of high IL-6 (beta -.07, p=.10), but 1992 religious attendance retained its effect (beta=-.10, p=.02). When religious attenders were compared to non-attenders, they were only about one-half as likely to have IL-6 levels greater than 5 ng/ml (OR 0.58, 95% CI 0.40-0.84, p<.005). Religious attendance was also related to lower levels of the immune-inflammatory markers alpha-2 globulin, fibrin d-dimers, polymorphonuclear leukocytes, and lymphocytes. While controlling for covariates weakened most of these relationships, adjusting analyses for depression and negative life events had little effect. Conclusions: There is a weak relationship between religious attendance and high IL-6 levels that could not be explained by other covariates, depression, or negative life events. This finding provides some support for the hypothesis that older adults who frequently attend religious services have healthier immune systems, although mechanism of effect remains unknown.
Religious beliefs and behaviors, particularly church attendance, have been associated with better physical health in the elderly.1-3 These studies indicate that frequent church attenders have lower blood pressure4-6, lower rates of stroke7, and survive longer8-10 than infrequent attenders. There have been few attempts, however, to understand why church-goers appear healthier than non-attenders. Some investigators have attributed this association to the fact that only the healthy are able to get to church; e.g., physical disability prevents attendance.11 What is not clear is whether it is the physical disability that prevents church attendance, or the frequent church attendance that prevents physical disability and health problems. Indeed, Idler and Kasl have reported that attendance at religious services is a predictor of better physical functioning 8 to 12 years later.12,13
If frequent church attendance somehow enhances physical health, then it is imperative to understand the physiological and biological mechanisms responsible for such effects. Psychosocial factors affecting physiology have been hypothesized. For example, investigators have found a strong relationship between church attendance and social support.14 Higher social support, in turn, has been associated with fewer emotional problems and improved coping.15 Greater religious involvement has also been directly linked with better adjustment to stress16 and lower rates of depression17-21 (independent of social support). High psychological stress and depressive disorder are associated with the release of hormones such as cortisol and other biological substances that impair immune function and increase susceptibility to disease.22-24 By helping to improve stress control, greater religious involvement may help keep down the production of biological substances that impair the body's capacity to fend off disease.
Interleukin-6 (IL-6) is an inflammatory cytokine and possible immune system regulator. IL-6's main role appears to be the induction of the acute inflammatory response and the induction of B-lymphocyte proliferation and differentiation.25,26 Changes in IL-6 regulation may constitute one of the fundamental aging processes and contribute to a broad spectrum of age-related diseases27,28 including B cell lymphomas,29,30 multiple-myeloma,31 autoimmune diseases,32 malignancies,32 viral diseases,32 abnormal amyloid deposition as seen in Alzheimer's dementia,33,34 and post-menopausal osteoporosis.35,36 We have previously reported an association between high IL-6 levels and increasing age, impaired physical functioning, cancer, heart disease, and high blood pressure in older adults.37 Other studies indicate high levels of IL-6 in patients with AIDS, particularly those who develop Kaposi's sarcoma.38,39
Besides IL-6, other substances in blood that are linked to inflammatory and immune responses are fibrin d-dimers (markers of coagulation and fibrinolytic activity), alpha-1 globulin (containing the acute inflammatory marker alpha-1 antitrypsin), alpha-2 globulin (containing the acute inflammatory marker haptoglobin and the chronic marker alpha-2 macroglobulin), beta globulin (containing transferrin which is reduced during inflammation), and gamma globulin (containing immunoglobulins which tend to increase during inflammation).
Psychological stress is known to be a strong inducer of IL-6 and other inflammatory cytokines.40 Increased IL-6 and related cytokines have been found in depressed persons (in vivo) and in culture supernatant assays of mitogen-stimulated peripheral leukocytes from such individuals (in vitro).41-43 IL-6 has also been shown to stimulate production of corticotropin releasing hormone.44 This effect is thought to occur through activation of the hypothalamic-pituitary-adrenal (HPA) axis, and IL-6 has been shown to stimulate each of these three organs independently and in series.44-46
One mechanism by which frequent religious attendance may convey better physical health is by lowering psychological stress, thereby reducing the production of IL-6 and the release of cortisol and other substances that adversely affect the immune system. The Established Populations for Epidemiologic Studies in the Elderly (EPESE), Duke site, is a multi-wave longitudinal study begun in 1986 of a cohort of over 4,000 adults age 65 or over living in central North Carolina. We have previously shown an inverse relationship between religious attendance and depression in this sample.21 During the third in-person interview in 1992, blood was drawn to examine the association between biological makers, aging, and physical functioning.37 Frequency of attendance at religious services was available for all three in-person interviews. This provided a unique opportunity to examine the association between religious attendance, IL-6, and other biological markers of immune function and regulation.
We hypothesize that (1) attendance at religious services in 1986 and 1989 will predict lower levels of serum IL-6 in 1992, independent of age, sex, race, education, chronic illness, and physical functioning; (2) the inverse association between religious attendance and IL-6 level will be greatest when both are measured concurrently (1992 interview); (3) these associations will weaken when depression and negative life events (stress) are covariates; and (4) religious attendance will also be related to lower levels of other biological markers of inflammation and immune system dysregulation, following the same pattern of relationships as seen with IL-6 (including weakening when depression and negative life events are controlled).
Methods and Procedures
The procedures involved in recruitment of the inception cohort and acquisition of biological markers in Wave III of the study have been described elsewhere.37,47,48 We will briefly summarize these procedures here. In 1986, a sample of 4,162 persons age 65 or older were selected based on a random household sampling in a five-county area, including and adjacent to Durham, North Carolina. The sample was selected specifically for comparison of racial groups. Participants were interviewed every 12 months, either by telephone or in-person. For the remainder of this paper we will refer to the 1986 in-person interview as Wave I, 1989 in-person interview as Wave II, and 1992 in-person interview as Wave III.
During Wave III, 2569 participants were visited in their homes and informed consent was requested to draw a blood sample for analysis. Of these, 269 were either too cognitively impaired or otherwise sick to give informed consent and were excluded. Of the remaining 2300 participants, 573 did not have blood drawn because of refusal, medical contraindications to blood drawing, or unusable blood samples. This left 1727 participants from whom blood was drawn and available for analysis; information on religious attendance was obtained at one or more of the three in-person interviews in 1718 of these subjects.
A comparison of baseline characteristics (1986) of participants on whom blood was drawn with those on whom it was not (n=2,281), revealed differences in age (26.9% >75 years vs. 45.5%, X2 p=.001), education (52.9% > grade school vs. 43.4%, p=.001), chronic medical illness (82.6% present vs. 85.7%, p=.007), and religious attendance (60.2% vs. 47.8%, p=.001), but no differences on race (53.4% Black vs. 55.3%) or sex (65.0% female vs 65.1%). Measures
Demographics. Demographic variables controlled for in the current analyses included age (65-74 years vs. 75 years or over), sex, race (White vs. Black), and education (1 to 8 years vs. 9 years or more).
Physical Functioning. Measures of physical functioning included the Katz49, Rosow-Breslau50, and Fillenbaum51 activities of daily living (ADL) scales. The number of impaired ADLs assessed by these three scales was summed, and scores were dichotomized into no ADL impairment vs. 1 or more ADL impairments.
Chronic illnesses. Chronic medical conditions recorded at Wave I included five physical health conditions: heart problems, hypertension, diabetes, stroke, and cancer. For analysis purposes, subjects with one or more chronic illnesses were compared to those with none of these conditions.
Depressive Symptoms. Depressive symptoms were assessed using the 20-item Center for Epidemiologic Studies - Depression (CES-D) scale.52 The items on the CES-D were presented to respondents in a yes-no format; positive responses were summed to create a scale ranging from 0 to 20.48
Negative Life Events. Negative life events experienced during the past year were recorded at Wave I. For analysis purposes, subjects with at least one negative life event were compared to those experiencing no negative life events.
Religious Attendance. Religious attendance was assessed with the following question: "How often do you attend religious services or other religious meetings?" The six response options were: (1) "never/almost never," (2) "once or twice a year," (3) "every few months," (4) "once or twice a month," (5) "once a week," and (6) "more than once a week."
Laboratory Methods for IL-6 Measurement. Blood was collected in EDTA-containing vacutainer tubes, placed on ice and taken to the laboratory where it was centrifuged and the plasma was frozen at -70 degrees in 0.5 ml aliquots.37 Plasma IL-6 was measured by ELISA (Quantikine, R&D Systems, Minneapolis, MN).53 This method has previously been shown to have a high degree of reliability and reproducibility in measuring plasma IL-6 over time in elderly subjects. The intra-class correlation coefficient for one measurement of IL-6 in 8 blood samples from a subject over a period of 36 days was 0.87, suggesting stability of IL-6 level over time.54 The distribution of IL-6 serum values was skewed towards the lower end of the scale (median 1.7, range 0-201 pg/ml) (Figure 1). For this reason, IL-6 level was examined both as a continuous variable and as a dichotomous variable (0-5 pg/ml vs. 5 pg/ml). Prior work has demonstrated the importance of examining IL-6 levels in the higher range (e.g., greater than 5 pg/ml) as a predictor of health outcomes in elderly persons.37
Other Biological Variables. Other biological variables indicative of immune function or inflammation measured at Wave III by serum protein electrophoresis were: (1) alpha-1 globulin (mostly alpha-1 antitrypsin in gm/dl), alpha-2 globulin (mostly alpha-2 macroglobulin and haptoglobin in gm/dl), beta globulin (predominantly transferrin in gm/dl), and gamma globulin (predominantly immunoglobulins in gm/dl). Also measured at Wave III were fibrin d-dimers (ng/ml) by ELISA (DimertestTM Stripwell EIA Kit, American Diagnostica, Greenwich, CT) and absolute neutrophil and lymphocyte counts (per cubic mm) by a coulter counter at Nichols Institute, San Juan Capistrano, CA. As for IL-6, d-dimer was examined as both a continuous and categorical variable.
Missing Values at Wave I. For covariates measured at Wave I (education, chronic illness, activities of daily living, negative life events, depressive symptoms), missing data was handled in the following manner. For measures with fewer than 2 percent missing values, the mean value was imputed. For measures with between 2 and 5 percent missing data, imputed versions were created using more complicated stochastic regression techniques; the only exception to this was activities of daily living for which mean values were imputed for the 5% with missing data. For Waves II and III, missing values on any covariate resulted in exclusion of the case from the analysis.
Statistical Analysis
For categorical outcomes (% IL-6 levels > 5 pg/ml and % d-dimer levels > 500 ng/ml), three-step logistic regression was used to examine the association with religious attendance. First, religious attendance was examined alone in the model. Second, the covariates age, sex, race, education, chronic illness, and physical functioning were added to the model to control for their effects. Third, depressive symptoms and negative life events were added to the model with the other covariates to determine if the relationships observed could be explained by depression or negative life stress as mediating variables. Standardized betas, p-values, odds ratios and 95% confidence intervals were obtained from the logistic regression output. For continuous outcome variables (all other biological variables), 3-step linear regression was used in the same manner as the logistic regression analyses above. Statistical significance was defined as p .05. Significance level was not corrected for multiple comparisons due to the exploratory nature of these analyses.
Results
Characteristics of the sample are described in Table 1. In 1986, only 26.9% of the sample was age 75 years or older; by 1992, 70.8% of the sample was over age 75. In 1992, two thirds of the sample were female (65.0%) and one-half of the sample (53.4%) was Black; a little over half of the sample (53.0%) had at least some high school education or beyond. The proportion of subjects with impairment of physical functioning (dependency on 1 or more ADLs) increased from 42.1% in 1986 to 57.6% in 1992. Despite this, approximately 60% of respondents attended church weekly or more often at all three waves. The proportion of subjects with depression increased slightly from 6.1% in 1986 to 8.8% in 1992. A little over one in 10 subjects (11.1%) had IL-6 levels greater than 5 pg/ml.
Hypothesis #1. Attendance at religious services in 1986 and 1989 will predict lower levels of serum IL-6 in 1992. While not true for IL-6 measured as a continuous variable, bivariate analyses did support the hypothesis for high IL-6 levels (>5 pg/ml) (Table 2, Figure 2). Religious attendance in 1986 weakly predicted a lower proportion of subjects with high IL-6 levels in 1992 (std beta=-.06, 1 df, p=.11). Religious attendance in 1989 was a stronger predictor of high IL-6 levels in 1992 (beta=-.10, 1 df, p=.01). When covariates were added, however, the effects of 1986 and 1989 religious attendance both weakened (beta=
-.04, 7 df, p=.41, and beta=-.07, 7 df, p=.10). Thus, there was only weak evidence that religious service attendance could predict future IL-6 levels.
Hypothesis #2. The inverse association between religious attendance and IL-6 level will be greatest when both are measured concurrently (1992 interview). If a true relationship existed between religious attendance and IL-6 levels, then the likelihood of observing such an effect would be greatest when religious attendance and IL-6 levels are measured at the same time (unless there were a delayed effect of religious attendance on IL-6 levels -- e.g., hypothesis #1 were true, for which we found only minimal support). Hypothesis #2, while again not supported for IL-6 as a continuous variable, was supported for high IL-6 levels (> 5 pg/ml) (Table 2). Bivariate analysis revealed that high 1992 religious attendance was associated with a lower proportion of subjects with high IL-6 levels in 1992 (beta=-.14, 1 df, p=.0005). Among subjects who never or almost never attended religious services, 15.7% had high IL-6 levels; among those who attended services between 1-2 times per year and 1-2 times/month, 11.7% had high IL-6 levels; and of those who attended services once per week or more, only 8.8% had IL-6 levels greater than 5 pg/ml (Figure 2). Controlling for covariates using logistic regression weakened but did not eliminate the association (beta -.10, 7 df, p=.02). Thus, the strongest relationship between low religious attendance and high IL-6 level was observed for the cross-sectional analysis in 1992. When attenders were compared to non-attenders or rare attenders, attenders were only about one-half as likely to have high IL-6 levels (odds ratio 0.58, 95% confidence interval 0.40-0.84, p<.005), after controlling for covariates.
Hypothesis #3. The associations between religious attendance and IL-6 will weaken when depression and negative life events are included as covariates. This hypothesis was not supported. Controlling for depression and negative life events in the above analyses had no effect on the relationships between religious attendance in 1986, 1989, and 1992, and high IL-6 level in 1992. The depression variable was then dichotomized to high (>8) and low (0-8) scores, and the analyses were re-run -- operating under the theory that controlling for subjects with more severe depressive symptoms might have the hypothesized effect. The results remained unchanged.