Selected annotated bibliography.

MacLean CD, Susi B, Phifer N, et al. Patient preference for physician discussion and practice of spirituality. J Gen Int Med 2003; 18(1):38-43.

A substantial number of patients desire spiritual interaction with their physician, but the desire is for a limited intensity in that interaction.

The Religion and Spirituality in the Medical Encounter Study (RESPECT) was conducted at 6 academic medical centers in 3 states (NC, FL, VT) and involved 456 patients. Results may have been skewed by several factors, including: a limited geographic spread, a largely Protestant population (n = 267, 56%), and the use of hypothetical questions.

Two interesting trends were noted in this study. Firstly, the percentage interested in spiritual dialogue with their physician increased according to the severity of the illness, from 33% during a routine office visit, to 40% during general hospitalization, to 70% when dying. Secondly, patient desire for spiritual interaction decreased as the intensity of the interaction increased: although 40% wished their physician to ask about personal religious beliefs while hospitalized, 39% desired the physician to say a silent prayer for the patient, and only 29% wished the physician to pray with them.

Interestingly, during routine office visits the majority (78%) did not wish to divert time from that spent discussing medical problems to time spent discussing spiritual issues. Time constraints may have been a consideration, helping to explain the increased percentage during hospitalization. Blacks were more likely to opt for such an exchange.

Abbotts JE, Williams RGA, Sweeting HN, and West PB. Is going to church good or bad for you? Denomination, attendance and mental health of children in West Scotland. Soc Sci Med 2004; 58:645-656.

Weekly church attendance had differing effects upon Roman Catholic and Church of Scotland (CoS) children affiliated. For Catholics weekly church attendance was associated with positive mental health, raised levels of self-esteem and lowered levels of anxiety and depression. Those affiliated with the CoS, however, appeared disadvantaged by weekly attendance. Only measurements of aggression, which were lower among weekly attenders, showed a similar correlation in the two denominations.

The study included 2,586 11-year-old students who completed a questionnaire covering health, self-esteem and self-image, health-related behaviors and attitudes, family life, school, leisure activities, friends, and projections for the future. All models were controlled for sex, social class, and all possible first order interactions. In this region of the country, there is very little interaction between the two groups, socially or otherwise.

On closer examination, the decisive factor appeared to be one of conformity. Among Catholic circles, weekly attendance was normative, while among CoS circles, weekly attendance was unusual, thus marking the child as different.

Due to the specificity of the study, results cannot be generalized. However, this study does demonstrate the significant effect that religion and society mutually have upon each other: in an accepting social clime, religiosity may operate as a beneficial agent; on the other hand, in an unaccepting social clime, religious attendance may well prove detrimental to mental health.

Salsman JM, Brown TL, Brechting EH, and Carlson CR. The link between religion and spirituality and psychological adjustment: The mediating role of optimism and social support. Personality and Social Psych Bulletin 2005 Apr; 31(4):522-535.

Religiousness and spirituality, as mediated by optimism and social support, were associated with greater life satisfaction and lowered levels of psychological distress among young adults. Religiousness was assessed by intrinsic and extrinsic motivations ("living versus using one's religion"); spirituality was assessed by sense of community, prayer fulfillment, and meaning and purpose in life. Higher levels of intrinsic religiousness and higher levels of prayer were associated with higher levels of optimism, which in turn led to a greater degree of satisfaction with life. Extrinsic religiousness, on the other hand, was unrelated to either optimism or social support. Social desirability was controlled for, and psychological symptoms and symptom patterns were also assessed.

These conclusions were based on two studies conducted at the University of Kentucky. The first on 217 students taking undergraduate psychology courses (mean age 20.2 years; SD 2.8), and the second, a replication study, on 159 similar students (mean age 20.8 years, SD 3.7). Both studies yielded much the same results. Unlike Study 1, however, the effect of intrinsic religiousness upon psychological distress was not significant in Study 2.

Specific associations between religiousness and spirituality, and with optimism and social support, were beyond the scope of this study. Nonetheless, it seems unlikely that the influence of religiousness and spirituality can be reduced to the influence of optimism or social support.

Nisbet PA, Duberstein PR, Conwell Y, and Seidlitz L. The effect of participation in religious activities on suicide versus natural death in adults 50 and older. J Nervous Mental Dis 2000; 188(8):543-546.

Religious activity deters from suicide. Adjusting for sex, race, marital status, age, and frequency of social contact, the odds for having never participated in religious activities are greater among suicide victims as compared with those dying of natural deaths. This effect remains even after controlling for the frequency of social contact. In regard to their frequency of visiting or talking on the telephone with family or friends, suicides differed little from natural deaths.

These findings were based upon data from the 1993 National Mortality Followback Survey (National Center for Health Statistics, 1998). Final analyses used 584 suicides and 4279 natural deaths. Women comprised some 42% of the entire sample; 64% of the subjects were white and 40.7% were married at the time of death.

Exactly which elements of religious participation were responsible for reducing the risk of suicide cannot be discerned from this study as data on specific religious activity was unavailable. Although the social support and social integration afforded by religious participation has been acknowledged, attending religious services does not always convey these supports. Moreover, whites accounted for 95% of the suicides in the NMFSA data; consequently, it was impossible to determine whether the relationship of religious activity participation with suicide varied by race or ethnicity.

Eskin M. The effects of religious versus secular education on suicide ideation and suicidal attitudes in adolescents in Turkey. Soc Psychiatry Psychiatr Epidemiol 2004; 39:536-542.

Among Turkish adolescents, suicide ideation was significantly higher among the secularly educated (41.7%) than among the religiously educated (24%).

Despite the high percentage of suicide ideation, the secularly educated were less accepting of suicidal peers, regarding such individuals as mentally ill; nonetheless, they accepted the place of suicide in society. The religiously educated, on the other hand, accepted neither the concept of suicide nor its place in society, yet they were more likely to accept and support suicidal friends.

Participants were 214 (41.1% female) students enrolled in a secular high school and 206 (29.5% female) students enrolled in a high school with religious education (imam hatip lisesi) in Istanbul, Turkey. The results were attained by a three-part, anonymous questionnaire. Included were questions about family background, attitudes and opinions towards suicide, and reaction to a "fictive close friend" who revealed suicide ideation.

While a relationship between type of education and suicide ideation was found to exist, causality could not be established. Also, the type of education may not have been the determinative factor: while the authors kept age, suicide ideation and parental education as covariates, other possible confounding factors--notably socioeconomic status, rural/urban background, and family size--were not considered.

Fontana A and Rosenheck R. Trauma, change in strength of religious faith, and mental health service use among veterans treated for PTSD. J Nervous Mental Dis 2004 Sept; 92(9):579-584.

A combination of guilt and weakened religious faith is the major impetus for veterans seeking further mental health services rather than the severity of posttraumatic stress disorder (PTSD) itself. Veterans who suffered from PTSD tended to be searching for "a meaning and purpose to their traumatic experiences" which was no longer satisfied by their religious faith.

These results were based on 1,385 patients (554 outpatient, 831 inpatient) who were surveyed over 8 years. Due to incomplete surveys, however, the study sample size was reduced to 1,006.

This study makes a strong case for the importance of individual, subjective spirituality to mental health. With two caveats, the authors conclude that spirituality may help treat PTSD: firstly, as spirituality deals with existential questions, it cannot be viewed as one content domain like any other. Secondly, as therapists themselves may need to engage spiritually during psychotherapy, therapists should address such existential matters prior to counseling.

Owing to the nature of the study, there are certain unavoidable limitations. Firstly, the population was limited to veterans and results cannot be generalized to include civilian populations. Additionally, all of these veterans were receiving treatment from the VA. Secondly, the nature of the study was retrospective and is thus subject to the fallibility of perception and memory. Finally, spirituality received only a cursory assessment.

Ogata M, Ikeda M, and Kuratsune M. Mortality among Japanese Zen priests. J Epidem Comm Health 1984; 38:161-166.

Zen priests in Japan had a significantly lower standardized mortality ratio when compared with the Japanese population at large. Although not attributing such low mortality to any one issue, the authors focused upon the lifestyle to which Zen priests generally conform and identified diet as a major contributor.

This cohort study was based on 1,396 deaths among 4,352 Japanese male Zen priests from 1 January 1955 to 31 December 1978 and was followed by a questionnaire survey to investigate the current lifestyle of active priests. This questionnaire revealed that Japanese Zen priests smoke less (although 53% were regular or occasional smokers and 45% were exsmokers or had never smoked), eat less meat and fish, and live in less polluted areas than the population at large; alcohol consumption did not differ markedly.

The possibility of a selection bias was acknowledged: candidates to become Zen priests must undergo a strict training process of several months, therein imposing a high mental and physical health level upon entrants. Additionally, the priests, most of whom are college graduates, may be also more health conscious. Given the possible selection bias, mental attitude and different lifestyle, religiosity alone cannot account for the differences in mortality between Zen priests and the rest of the population in Japan.

Kaplan SD. Retrospective cohort mortality study of Roman Catholic priests. Prev Med 1988; 17:335-343.

Roman Catholic priests are at a lower risk of dying from suicide, with a standardized mortality ratio (SMR) of 13, emphysema (26), pneumonia (67), cancer of lung (59), cancer of stomach (74), and infective and parasitic diseases (29).

Priests also appear to be moderately protected against prostatic cancer by virtue of their general sexual inactivity (SMR 81); when men ordained after the age of 30--many of whom were widowers and thus more likely to have engaged in frequent sexual activity--are removed from the sample, however, the result attains significance (SMR 71). No discussion was made of the possible effects of masturbation. This population's high proclivity for obesity may have been somewhat accountable for cases of prostate cancer. If celibacy is indeed protective for cancer of the prostate, the effect is considerably less than the effect of celibacy in nuns for cancer of the cervix.

Although the study involved 10,026 eligible priests in 33 United States dioceses, these were concentrated in the northeast and north-central states and in California. Additionally, three large dioceses contributed to over 20% of the cohort, possibly distorting the results.

The mortality experience of Roman Catholic priests resembles that of a population with a high alcohol and caloric intake but a relatively low exposure to tobacco. This study did not investigate the effect of religion or spirituality.

Michalek AM, Mettlin C, and Priore RL. Prostate cancer mortality among Catholic priests. J Surg Oncol 1981; 17:129-133.

Among Roman Catholic priests, the mortality ratios for all causes of death and for cancer were about 15% less and 30% less, respectively, than in the general population. Cancer mortality was significantly lower for cancers of the lung, colo-rectum, stomach, and prostate, but higher for malignant melanoma and "unspecified respiratory organs." Death rates from influenza and pneumonia were also much lower in priests. On the other hand, priests were at a higher risk of death from heart disease, diseases of the arteries, arteriosclerosis, cerebrovascular disease, cirrhosis, and peptic ulcer.

Causes of mortality were analyzed among 6,226 Roman Catholic clergymen in eight dioceses in New York State from 1965 to 1977. A total of 1,060 clerics were found to have died during this period, 1,006 of whom the cause of death was determined.

This study is limited in geography and did not investigate the effects of religion or spirituality but rather those resulting from a different lifestyle. Additionally, the state's two major religious orders, the Jesuits and the Franciscans, were included in the study, although their lifestyle may differ significantly from diocesan priests, especially as they are prone to be sent abroad.

Kendler KS, Gardner CO, and Prescott CA. Religion, psychopathology, and substance use and abuse: A multimeasure, genetic-epidemiologic study. Am J Psychiatry 1997; 154(3): 322-329.

In twins, religiosity was not strongly associated with psychiatric symptoms and disorders, although it did appear to lessen substance dependance, reduce the depressogenic effects of stress, and promote a healthy lifestyle.

Three dimensions were used to measure religiosity: personal devotion, personal conservatism and institutional conservatism in white, female same-sex twins (n = 1,902 pairs) both monozygous and dizygous.

While the preponderance of evidence indicates that religiosity lowers levels of depression, in this study, personal devotion alone attained levels of significance. All three dimensions of religiosity were, however, significantly and inversely related to current alcohol and tobacco use, although the association was slightly weaker with personal conservatism. Levels of correlation among the three measures far exceeded chance expectations, indicating the strong familial-environmental influence. No strong association was found between genetics and religion.

Although it is impossible to say conclusively, analyses suggest that at least some of the relationship between religiosity, depressive symptoms and drug use and abuse is causal. The study is further limited as assessment was only made at a single point in time (by three successive interviews) and failed to include parental measures of religiosity. Additionally, only female twins were included for study despite evidence that there is a gender difference in the relationship between psychopathology and religiosity.

Koenig HG, George LK, Meador KG, et al. Religious practices and alcoholism in a southern adult population. Hosp Comm Psych 1994; 45(3):225-231.

Religion appears to lessen alcohol disorders. Three relationships were uncovered:

1) Rates of alcohol abuse were significantly lower in both those who frequently attended church services and among those who frequently engaged in private religious activity (e.g., prayer or Bible study). Conversely, alcoholism was more than twice as common among frequent users of religious media (watching or listening to religious television or radio).

2) Rates of alcoholism were lower among reported born-again Christians.

3) Lifetime, rates of alcohol abuse were highest among Pentecostal denominations (17.4%) and lowest among mainline (8.6%) and conservative (8.8%) Protestants.

These observations were based on 2,969 persons who participated in the 1983-1984 National Institute of Mental Health Epidemiologic Catchment Area survey at its Piedmont, North Carolina location. Data was collected in two waves, spaced out by six months. Respondents between the ages of 60 and 97 constituted the largest group.

Causal relationships could not be determined due to the cross-sectional nature of the data. Church attendance, therefore, could not be definitely said to lessen alcohol abuse. Nor was it possible to determine whether those utilizing religious media are more isolated and therefore more susceptible to alcoholism, or whether they are largely alcoholics looking for solace. Lastly, Pentecostal churches may attract persons with alcohol problems into their congregations by providing better support, hope and alternative coping behavior rather than actually encouraging alcohol disorders.

Hays JC, Landerman LR, Blazer DG, et al. Aging, health and the 'electronic church.' J Aging Health 1998; 10(4):458-482.

Elderly people with deteriorating health do not increase their participation in religious media (radio and television) as their disability forces them to reduce their participation in organizational religious activities. The use of religious media is more sensitive to cultural heritage (race, socioeconomic status and rural upbringing) than to either declining health or service attendance.

These observations are based on surveys of respondents from North Carolina's north-central Piedmont area (composed of four counties, one urban and four rural). A preliminary wave of interviews was used as a baseline (n = 4,162); followup interviews were conducted three years later for those respondents who could be located (n = 2,971). The effective sample was largely female, with an average age in the early seventies.

All analyses were performed separately by gender. Five categories of covariates were evaluated as possible confounders: 1) two other dimensions of religiousness (devotional and experiential); 2) demographic variables; 3) social environment; 4) measures of health (in addition to functional ability); and 5) negative life events.

Generalization is limited by certain regional factors, notably the population's relatively high degree of religiosity.

Strawbridge WJ et al. (1997). Frequent attendance at religious services and mortality over 28 years. Amer J Pub Health, 87(6):957-961.

Frequent attendees at religious services showed significantly lower mortality rates over nearly 3 decades when compared with infrequent attendees; this result remained significant even when accounting for mental and physical health during followup. Adjustment for social connections (marital status, close social contact and group memberships) made little difference. The relationship between frequent attendance and mortality was stronger for females than for males.

These results were based upon data collected from Alameda County, California of people aged 16 to 94 at baseline since 1965 (n = 6,928) with followups in 1974 (n = 4,864), 1983 (n = 1,798) and 1994 (n = 2,729). Response rates for the four surveys were 86%, 85%, 87% and 93%, respectively.

Changes is health, health practices and social connections were all assessed in an attempt to discover whether these factors--problematic in earlier studies--might explain any observed association. Although the covariates weakened the relationship between frequent attendance and mortality, the results still remained statistically significant.

Females, Blacks and, interestingly, the mobility impaired were all more likely to be frequent attendees. While marriage was not associated with frequent attendance, both social contacts and group memberships were; also, frequent attendees were more likely to stay married to the same person. No association was found with exercise, while smokers and heavy drinkers were much less likely to attend frequently, and the overweight more likely to attend frequently.

Oman D, Kurata JH, Strawbridge WJ, and Cohen RD. Religious attendance and cause of death over 31 years. Int'l J Psychiatry Med 2002; 32(1) 69-89.

Religious involvement promotes health through multiple causal pathways, acting in much the same sense as higher socioeconomic status. Specifically, associations between religion and cancer were weak, while those with circulatory diseases, especially cerebrovascular, were "modest yet robust."

The authors examined data from Alameda County, California, collected in 1965 (n = 6,928) with followups in 1974 (n = 4,864), 1983 (n = 1,798, a 50% subsample), and 1994 (n = 2,729). Analyses were based upon subjects aged 21 or older in 1965 who had no missing variables for at least one of the four examinations.

Religious attendance affected young men the most, presumably by acting as a buffer against work-related stressors, to which this group is the most highly exposed. Older men, by contrast, were to be afforded the least protection from their religious involvement, perhaps due to a cessation of work-related stressors.

Religious involvement promotes health through at least three means: 1) members tend to adopt and maintain healthy lifestyles (a particularly significant attenuating factor), 2) social support, and 3) psychological buffers, such as greater faith, stronger sense of coherence, and means of coping.

Krucoff MW, Crater SW, Gallup D, et al. Music, imagery, touch, and prayer as adjuncts to interventional cardiac care: The monitoring and adcualisation of noetic trainings (MANTRA) II randomized study. The Lancet 2005 July 16; 366:211-217.

Neither prayer nor noetic intervention (music, imagery and touch [MIT] therapy) had any significant effect upon the prognosis of heart patients as determined by in-hospital major adverse cardiovascular events or death or any readmission to hospital within the following 6 months.

Between May 1999 and December 2002, 748 patients undergoing percutaneous coronary intervention or elective catheterization in 9 centres across the USA were randomly assigned to one of the following groups: standard care (n = 192), prayer only (n = 182), bedside MIT therapy (n = 185), or prayer and MIT therapy (n = 189). Off-site prayer groups (double blinded) included established Christian, Muslim, Jewish and Buddhist congregations.

Baseline measures of mood, anxiety, spirituality and quality of life were well balanced across the four treatment groups. However, 613 of 688 (89%) patients were aware of prayer on their behalf beyond that of the protocol (e.g., the prayers of family and friends). As the authors admit, the presence of placebo effect in 67% and off-protocol prayer in almost 90% of the study cohort may have impaired the study's ability to detect incremental changes.

Conrad C. Daily, MTh