Research on the healing power of distant intercessory prayer: disconnect between science and faith.

Interest in non-medical treatments for illness has grown exponentially in recent years as evidenced by the expansion of health psychology and establishment of the National Center for Complementary and Alternative Medicine. One particular area of research that has generated attention and even enthusiasm from the media and Christian groups alike is distant intercessory prayer (IP) for healing. Several double blind, randomized, controlled studies have examined whether a statistically significant effect can be found when prayed for groups are compared with controls. The central premise of this article is these studies lack any theological or rational theoretical foundation and consequently produce non-interpretable findings. It is further argued that the experimental methods of science are based on important assumptions that render them ill-equipped to study divine intervention. As a result IP studies are seen as a distraction from more appropriate work that should be done in the areas of religion and health.

**********

Since the foundation of the Christian church, believers have prayed for the healing of the sick. Indeed the Bible documents many instances of divine healing (e.g., Matthew 8:1-13; 15:29-31; John 2:1-13, etc., New American Standard Version) and indicates that the prayers of the righteous may accomplish much (James 5:16). This is not unique to Christianity. Practitioners of many other major religions of the world also offer up prayer for those who are sick. Though these prayers are sometimes said in the presence of the ill person, or during enactment of religious rituals such as laying on of hands or anointing with oil, they are also offered when the sick individual is not present, i.e., they are spoken from a distance. Further, sometimes the prayers are not for a particular person as in global prayers for healing of the sick and afflicted. This presumes that at least some of the individuals being prayed for are unaware of the prayers offered on their behalf. Historically, academic study of prayer has been largely the purview of seminaries and other institutions steeped in a religious heritage. These studies were based on literary investigation of sacred texts and exegesis of important historical passages subjected to hermeneutical analysis. Recently, however, prayer has become the focus of scientific investigators at secular institutions who utilize the methods of science, namely randomized controlled clinical trials and similar experimental designs, to study questions such as whether prayer can be shown to have an impact on morbidity and mortality.

The increased scientific interest in studying prayer and health is one aspect of a contemporary trend embracing the study of health interventions and conceptualizations that are not overtly biological, chemical, or surgical (mechanical) in nature. This trend is perhaps best exemplified by the recent establishment of the National Center for Complementary and Alternative Medicine (NCCAM) within the U.S. National Institutes of Health. NCCAM reports that in 2002, 62% of Americans used some form of alternative medicine. Of the 10 most often utilized alternative medicine therapies, prayer for self (43%) and prayer for others (24.4%) were the two most commonly named therapies and being in a prayer group (9.6%) was fifth (Barnes, Powell-Griner, McFann, & Nahin, 2002). Clearly prayer is viewed by many Americans as having relevance to their own and others' health.

Health psychology. Similarly, the explosive growth over the last 15-20 years of health psychology as a professional area of specialization provides evidence that interest in behavioral and emotional variables as they influence and are influenced by health is strong. The remarkable development of health psychology has been due to many factors, too numerous to detail here, but a few examples will suffice. First, it became widely recognized that certain behaviors (e.g., smoking, overeating) are associated with negative health outcomes whereas others (e.g., regular exercise) are associated with better health. Consequently, research and intervention aimed at curbing the former and enhancing the latter has proliferated. By way of another example, advances in the field of psychoneuroimmunology demonstrated how chronic and acute emotional states influence immune system functioning rendering individuals more or less susceptible to infection and more or less likely to recover from illness (Forlenza & Baum, 2002; Ironson, Balbin, & Schneiderman, 2002; Kiecolt-Glaser, McGuire, Robles, & Glaser, 2002). Another important area receiving extensive attention is the nation's top cause of mortality, cardiovascular disease. Early research on Type A behavior, which helped establish the importance of health psychology, has given way to increased focus on cynical hostility and even depression as risk factors for coronary heart disease (Smith & Ruiz, 2002; Suchday, Tucker, & Krantz, 2002; Williams, 2001). In essence, the study of the interface between psychology and health has moved from the periphery into the mainstream in American psychology and, arguably, American medicine.

Within health psychology there has recently been a most interesting, noteworthy, and even surprising trend toward consideration of how spiritual or religious variables may influence health. The evidence for this trend is abundant. Several highly respected scientific journals recently published special issues on this topic or presented a series of articles on it (e.g., American Psychologist, Annals of Behavioral Medicine, Health Education & Behavior, Journal of Health Psychology, Psychological Inquiry, Research on Aging) and the Society of Behavioral Medicine now lists spirituality as one of the 15 topical areas for its annual meeting. The growth of spirituality as a focus of scientific investigation, like that of health psychology itself, has also likely been the result of many factors. Most notable are several important studies that demonstrated a positive relation between religiosity and health (Clark, Friedman, & Martin, 1999; Hummer, Rogers, Nam, & Ellison, 1999; McCullough, Hoyt, Larson, Koenig, & Thoresen, 2000; Powell, Shahabi, & Thoresen, 2003; Strawbridge, Cohen, Shema, & Kaplan, 1997).

Distant intercessory prayer. As these many forces merged, interest in the effect of distant intercessory prayer (IP) on health has grown. News media outlets and Christian groups have enthusiastically reported the results of the relatively small number of published research articles on IP. Investigations in this area are designed to scientifically test whether praying for patients can, within the confines of a scientifically controlled study, actually be demonstrated to improve patients' health relative to a control group that is not the object of prayer. An important feature of these studies is that the patients in them do not know if they are receiving prayer (i.e., they do not know if they are in the prayer group or in the control) and, in some cases, do not even know that they are participating in a study at all. Thus, and this is quite significant, what is being tested is something specifically different from the efficacy of prayer offered in the patient's presence or even prayer offered with the certain knowledge that one is the recipient of prayer. Instead, these studies are aimed at establishing whether prayer, apart from any placebo, expectancy, or other psychological effect, can make a scientifically discernable difference in health status. In other words, is prayer, apart from any psychological mechanism, as when it is tested in double- or triple-blind studies, an efficacious treatment for illness?

In this manuscript I will first lay bare my own pre-suppositions in terms of religion and science. I will then briefly describe the existing scientific literature on IP in order to provide the reader with a background on these studies. I do not, however, intend to focus the discussion on a critical review of the experimental designs or methods of this body of literature. Instead, this article will focus on what I perceive as the incompatibility of the question (i.e., is prayer efficacious for healing?) with the method (i.e., controlled scientific study). The presentation will be structured around a series of questions with brief explication of their potential implications. In the end I hope that scientists and believers alike will be much less enthusiastic about studies of IP and more enthusiastic about investigations of important and more appropriate variables that may influence the religion/health relationship.

Personal Assumptions

Since research on IP is inherently fraught with both religious and scientific overtones, it seems appropriate in the interest of full disclosure, to briefly expose my personal positions in these areas. Essentially, I believe in both, i.e., I am an evangelical Christian and a social scientist, specifically a clinical health psychologist. Of perhaps greater importance to this discussion, I believe that the methods of science are, in nearly all cases, relevant for investigating matters pertaining to the influence of faith, beliefs, lifestyles, etc. on health. In fact, I recently published in the area of religious orientation and cardiovascular reactivity to stress (Masters, Hill, Kircher, Lensegrav-Benson, & Fallon, 2004; Masters, Lensegrav-Benson, Kircher, & Hill, 2005) and have previously published on religion and mental health (Masters & Bergin,1992; Masters, Bergin, Reynolds, & Sullivan, 1991). Thus, a priori, I believe that scientific investigation of faith based practices is nearly always appropriate, illuminating, and relevant.

Empirical Research on IP

A review of the medical and psychological literature reveals that there are presently 12 published controlled trials of IP (Aviles et al., 2001; Byrd, 1988; Cha, Wirth & Lobo, 2001; Harris et al., 1999; Collipp, 1969; Joyce & Welldon, 1965; Matthews, Marlowe, & MacNutt, 2000; Matthews, Conti & Sireci, 2001; O'Laoire, 1996; Palmer, Katerndahl, & Morgan-Kidd, 2004; Tloczynski & Fritzsch, 2002; Walker, Tonigan, Miller, Comer, & Kahlich, 1996) though other studies are likely underway. These studies investigated the effects of IP on a wide array of conditions and outcome variables including rheumatoid arthritis, leukemia in children, cardiac disease, substance abuse, kidney dialysis, successful pregnancy via in vitro fertilization-embryo transfer, and measures of mental health including self-esteem, anxiety, and depression. One study (Palmer et al., 2004) was even designed to test the effects of prayer on problem resolution among a non-clinical sample. The studies have generally followed what are typically considered best practices in research methods, i.e., procedures designed to increase internal validity. In what follows I will offer a more detailed presentation of one of the IP studies in order to familiarize readers with this literature. Specifically, the work of Harris and colleagues (1999) provides a compelling example of the most rigorous of these studies and will serve as the exemplar.

Harris and colleagues (1999) conducted a study designed as a more methodologically stringent replication of Byrd (1988), which is probably the best-known IP study to date. They utilized a "randomized, controlled, double-blind, prospective, parallel-group trial" (p. 2273) to test whether IP would reduce over-all adverse events and length of stay in a coronary care unit (CCU). At the time of admission 1013 consecutive patients were randomly assigned to either a usual care (control) group (n = 529; 52%) or a prayer group (n = 484; 48%). Twenty-three patients were subsequently eliminated from the sample because their hospital stays were less than 24 hours. Brief stays did not allow enough time for subjects to be assigned to groups and for the intercessors to pray for the patients. Consequently 466 patients remained in the prayer group and 524 in the usual care condition. It is important to note that the patients did not know to which group they were assigned; in fact, they did not even know that a study was taking place. This was also true of their physicians and other hospital staff.

Group assignment was made on a random basis, daily, by the secretary in the chaplain's office using the computer record of new admissions to the coronary care unit. For patients assigned to the prayer group, the secretary contacted a prayer team leader and informed him/her of the patient's first name only. No other demographic or medical information was provided to intercessors. The prayer leader subsequently called the other 4 persons on the intercessor team. All total there were 15 intercessor teams of 5 members each. The intercessors could be associated with any denomination but had to agree with the following statement: "I believe in God. I believe that He is personal and is concerned with individual lives. I further believe that He is responsive to prayers for healing made on behalf of the sick" (p. 2274). The intercessors were primarily women (87%) and all reported at least weekly church attendance.

As the name implies, the usual care patients received their normal medical care. They also received any other usual care that would be provided, notably, if patients in this group happened to request prayer from the chaplain during their stay it was provided for them. The authors indicate that typically about 5% of CCU patients request prayer (though the number for this particular sample was not provided) and, remember, the chaplain was also blinded to experimental group membership of the patients. The members of the prayer group received daily prayer over a period of 28 days from the 5 intercessors assigned to their case. Specifically, the intercessors were instructed to pray for "a speedy recovery with no complications" (p. 2274).

Several measures of outcome were included. The most important was an overall, global, measure of the course of patient hospital stays. This measure used a weighted (for severity) total of negative events that occurred during the stay so that a lower score indicated a stay with fewer comorbid conditions or complications. The authors also reported an unweighted count of these same conditions or complications. The outcome measure used in the Byrd (1988) study was also employed as was a determination of length of stay. Finally, the individual components of the global measure were tested for between group differences using a statistical probability correction to adjust the alpha level for multiple comparisons (something Byrd did not do).

The results from this study are typical of the results from other studies that found a beneficial effect for IP (not all did) in that the benefits were not consistent across measures. For example, a statistically significant difference favoring the prayer group was found on the global measure of outcome but none of the 34 individual components of this measure, including death, showed this same difference. Further, the study specifically failed to replicate the Byrd study as no differences were found on the measure used by Byrd. It is important to also note that, despite the directive to pray for a speedy recovery, there were no differences in length of stay between the groups. Nevertheless, the study did find an 11% reduction in scores on the weighted global measure of outcome in the prayer group when compared to the control and a 10% reduction on the unweighted global measure.

Questions Regarding IP Research

What follows are a series of questions regarding IP research. Again, the focus of this discussion is on the philosophical and theological implications of the research and will only touch on methodological issues as they pertain to these larger questions.

1. What do significant results mean?

The ultimate purpose of scientific experimentation is to provide the empirical basis for the development and refinement of explanatory theories and concepts that better elucidate the functioning of the natural world and therefore allow scientists to better understand and predict natural events. The end of any particular experiment is only the beginning of the scientific process. Statistically significant results must be properly integrated into theories that provide the context for the "real world" significance, not statistical significance, of the findings. As anyone who has read doctoral dissertations can attest, it is precisely at this point that many authors cannot find the forest for the trees, i.e., the careful attention to methodological rigor is not matched with thoughtful integration of data into theory.

What is the theoretical basis of IP studies that provides the context for integration of empirical results? If there is no established theory that is applicable, what novel theories guide the research? I believe that, in both cases, there simply are no explicit theories. This has practical and methodological consequences, for example, there are no explanations offered in IP studies for the choices of outcome measures, types of intercessors, wording of intercessory prayers, choices of patient conditions studied, etc. In short, these are studies that lack any theoretical guidance so when it is determined that, for example, coronary patients in the prayed for group improved on the global measure when compared to the controls, there is nothing more to say that is coherent with any philosophically or theologically accepted position. Confusion over the nature of the independent variable is evident in the discussion section of the Harris et al. study. They state, "... we have not proven that God answers prayer or that God even exists. It was intercessory prayer, not the existence of God, that was tested here" (p. 2277). Yet, their intercessors had to believe in a personal God that is responsive to prayers for healing of the sick. If the study had nothing to do with testing God and was only about "intercessory prayer" then why were these the criteria for choice of intercessors (see below)? In other words, what is the precise nature of the independent variable? If the independent variable is not clearly articulated then it will be very difficult to properly explain the findings. The preponderance of the IP literature has decidedly Christian underpinnings but no concurrent Christian theory in which to integrate the results. (There is a separate body of literature concerning distant healing, not IP, that steers clear of Christian associations).

Not only is there no existing theory but it seems quite unlikely that a theologically cogent theory could be developed that would provide sufficient context for these studies and their results. For example, I invite the reader to develop a theory to explain why God would respond more favorably to the prayers made on behalf of a group of people who were chosen at random to be in the prayed for group? This of course means that God did not respond as favorably (i.e., at the p < .05 level, more below) to those randomly placed in the control group. What kind of God would this be? It is also quite likely that some in the control group were prayed for by friends, family, etc though there is no way of knowing this for sure or to how many this may apply. Another potential question arises, why would the prayers of distant intercessors who only know the patients' first names and nothing else potentially carry influence with God relative to the prayers of those who intimately know and care for these patients? What current or even potential theological, philosophical, psychological, etc. theory can explain, or even address, these questions raised by significant IP findings?

2. What does it mean if results are not statistically significant?

If significant findings cannot be explained, perhaps non-significant ones can. If so, perhaps this can provide a rationale for continuing IP studies. But what do non-significant findings mean? Should readers conclude that prayer is useless, God does not hear, God does not exist, the wrong intercessors were chosen, the wrong words were used in the prayer, the prayers should have been offered more times daily, etc.? None of these conclusions seems warranted. Again, since there is no explicit theory to be tested there is no importance that can be assigned to non-significant findings.

3. What does it mean if some measures show significant results and others do not?

The findings of IP studies show either no significant results or some significant results. What they do not show are consistent significant results across measures within the same study. In Harris et al. (1999), for example, significant findings were obtained for the global measures but not for the specific measures, including mortality, and not for the seemingly important length of stay variable. Similarly, Byrd (1988) obtained several times as many non-significant results as significant ones. So again what does this mean? Given that there was no theory to guide the choice of measures, readers will have difficulty interpreting the findings. In the Harris study, however, a particular outcome variable was specified as the target of the intervention. The intercessors were specifically assigned to pray for "a speedy recovery" (p. 2274). It is striking, therefore, that the variable most directly targeted by this intervention, length of hospital stay, showed no significant findings. This seems to indicate that there was no effect for IP at precisely the point where an effect was specifically sought. In other words, at the point where a specific outcome measure was linked with a particular intervention, the findings were not significant. One other study deserves note. Cha et al. (2001) had intercessors specifically pray for increased rate of pregnancy in patients undergoing in vitro fertilization. They did find increased pregnancy rates in the prayer group, but only for women over the age of 30. At the risk of being redundant, what theological principles can account for this finding? Does God generally favor women over 30 when responding to prayers for fertility and therefore we have learned something predictable and characteristic about God, or is this just a fluke finding? Is there any way to develop a theological or scientific principle from this finding?

It seems difficult indeed to explain why IP would be efficacious for some measures and some samples but not others. In this light, the failure to find significant differences for mortality is noteworthy. Although mortality is not the only variable of interest in any health outcome study, it is surely a very important one. Presumably family and friends deem it important to actually have their loved one alive. Yet the failure to find differences in mortality is not discussed in the research. From a Christian perspective it is difficult to conceive why God would care or be more capable of intervening regarding some physiological and health variables than others.

Taking the question of outcome measures a step further leads to consideration of the populations and samples utilized in these studies. Clearly choice of outcome measures, if not directed by theory, is at least to some extent dictated by the conditions under study. Thus, if patients have heart disease it makes sense to have measures of cardiac function, or, if patients are attempting to stop drinking, measures of alcohol intake are sensible. But what theory guides the choice of patient populations in the first place? Why, for example, have there been no IP studies of paralyzed individuals or those suffering blindness? Choices of outcome measures in these cases seem straightforward and there can be no argument that many patients afflicted with these conditions desire a cure. In fact, one could argue that in the absence of effective medical treatments, it is precisely these conditions where IP may be shown to have its greatest effects. Yet, the studies have not been done. Why?

4. What if some people in the prayer group show improvement and others do not?

In studies that compare the mean of one group with the mean of another there is always some spread of scores around those means. It is also true for most, if not all, studies that some individual members of the control group fare better in terms of magnitude/severity of symptom reduction or final level of disease/dysfunction than do some members of the treatment group. So what is to be made of the fact that some patients in the prayer groups do not get better, some get worse, and some do not do as well as some members of the control group? Again, what coherent theoretical or theological propositions relating to the prayer intervention can be invoked to account for these differences?

This raises the related issue of the use of inferential statistics in IP studies. Statements of probability based on inferential statistics are likewise based on knowledge of the theoretical sampling distribution of the statistic, the probability distribution of the variable in question within the particular population, etc. These considerations are themselves based on the assumptions of a mechanistic universe that operates in ways that yield predictable probabilities and replications thereof. Based on knowledge of these probabilities, it is agreed that when statistical values differ by a certain amount it is relatively unlikely that this result happened by chance alone. Thus significance levels, themselves arbitrarily and conventionally defined, establish the cutoff point for determining that the difference is large enough to render it highly unlikely that it is simply due to chance. But is this model applicable when the question has to do with the intervention of God via prayer? In fact, is not God's very nature such that universal probabilities and estimates of chance do not apply? God is not constrained by the "rules" of things created: After all, God is the Creator. Neither is there any articulated reason to suppose that God would operate according to conventional significance levels. So upon what basis would one infer that God would operate by these rules in the context of IP studies and therefore theistic prayer intervention effects could be revealed on the basis of the probabilistic outcomes of controlled scientific investigation? A more thorough and excellent discussion of this point is offered by Chibnall, Jeral, and Cerullo (2001).

5. Upon what basis are intercessors chosen?

In traditional intervention studies the characteristics of those providing the intervention are important. By way of analogy, the history of psychotherapy outcome research is replete with examples of controversy over the adequacy of the therapists. Many early therapy outcome studies relied upon graduate students as therapists. If the treatment under consideration failed to show an effect, one explanation that was often offered was that the therapists were not competent or experienced enough to provide a valid test of the particular therapeutic intervention. This controversy could be resolved, however, by careful deliberations leading to a priori delineations as to who would be qualified to offer the intervention and then conducting an experiment with those qualified persons as the therapists.

Can this model be applied to IP studies? Is it possible to determine the necessary qualifications for conducting IP? Moving further with this same analogy, some psychotherapy studies compared results from experienced therapists with those having less experience in order to test the impact of experience as a therapist variable. Other studies compared adherents to one school of therapy with adherents of another (e.g., psychoanalysis vs. cognitive-behavior therapy). Thus characteristics of the therapists themselves could be studied based on theoretical propositions depicting what constitutes an effective therapist.

Similarly, it seems reasonable to suggest that some intercessors are more qualified than others (remember the prayer of the righteous cited above) and that, therefore, this is a potentially important variable, certainly one worth studying. But is anyone ready to conduct a study where one group of intercessors is Jewish, another is Christian, and another is Muslim? For that matter, would anyone really want to compare intercessors who were Presbyterian with those who were Methodist and with those who were Baptist? It seems unlikely that these studies will appear anytime soon, yet there are essential theological principles that differentiate these groups that may be important. Further, as mentioned above, there have been minimal criteria established regarding the qualifications of the intercessors in at least some studies (e.g., must believe in God, power of prayer to heal, etc.; see Walker et al., 1996 for another example). So why not study these qualifications themselves? One could argue, as was suspected in the therapy outcome studies, that the failure to find strong effects for IP could be due to the inadequacy of the intercessors. But to make this case would require a clear delineation of the conditions believed to be necessary for IP to be effective. This would also imply that these conditions or characteristics could be reliably and validly measured, at least to the level of determining their presence or absence. To this point in the literature only the most general of criteria have been applied to the question of what qualifies one to be an adequate intercessor: This leaves a seemingly important factor in the research unaddressed. My personal belief is that it will not be possible to articulate and measure the necessary and sufficient characteristics of intercessors that are predictive of their effectiveness because, in the end, it is God's purposes that determine outcomes. Nevertheless, in principle researchers could specify, based on religious or other doctrine, what they believe to be the essential characteristics of intercessors and then measure these in relation to their prediction of outcome. I challenge anyone intent on conducting IP studies to take this aspect of the research seriously.

Wrong Method to Address the Question

Christians, and other people of faith, have since the beginning of time offered up prayers for the sick with the fervent belief that, at least on some occasions, these prayers are answered with healing. In the scientific tradition, as far back as 1883 Sir Francis Galton (as cited in Palmer et al., 2004) suggested that whether sick persons who are prayed for recovered more rapidly than those not prayed for was a proper topic for empirical study. Since the scientific method has proven its worth in terms of collecting data about the workings of the world, why not use it to test whether there is evidence to support these long held beliefs?

It is my contention, however, that a major source of confusion in IP studies is the result of applying the wrong method to the question of the efficacy of IP. The scientific method is not appropriate or equipped to resolve questions that concern the intervention of deity which is, I believe, the implicit, ill-defined, and sometimes denied "theory" behind IP research. The basic premise of science is the functioning of a mechanistic and predictable world but the basic premise of the Biblical deity is that God acts according to God's own purposes and is not constrained by physical limits. God is metaphysical, science is physical. Natural processes are the proper domain of science but supernatural processes are the domain of theology. Further, God indicates that God's ways are not known to humans nor should they be questioned or tested (Romans 8:26-9:33; 11:33-36). There is no theological principle to suggest that God's ability to heal can ever be tested by controlled, scientific methods. In fact, quite the opposite seems to be the case. Scriptural passages warn to not tempt, test, or question God (Deuteronomy 6:16; Matthew 4:1-11; Luke 4:1-13).

This confusion over the natural vs. supernatural is inadvertently evident in the limitations section of the Palmer et al. (2004) study. They indicated that their study was limited by the fact that the participants were largely well educated and white, and therefore generalizability was limited. Clearly this would be the case were they studying natural mechanisms that may be influenced by one's level of education or ethnicity status. But how would prayer fit into this scheme? What mechanism associated with prayer to God suggests that differences in ethnicity or education are moderating variables of prayer's effectiveness? Is God likely to be influenced in some manner more or less by individuals who vary in their level of education or are of a different ethnic identification? God does not respect one person more than another (Acts 10:34).

Some may argue, as do Harris and colleagues (1999) and more recently Palmer et al. (2004), that imputation of interpretations for this research that incorporate God are not necessary and go beyond what the studies assess. These authors claim that only the natural properties of prayer as it relates to healing are under consideration and therefore the studies are perfectly congruent with scientific principles and assumptions. They further point out that researchers need to be open to testing heretofore undiscovered or unexplained processes. If only natural processes were under study, I completely agree. In fact, reference to God is not necessary and is not made in studies of distant healing, energy fields or energy medicine, etc. But such is clearly not the case in the extant IP literature. The evidence points to an agenda that invokes God into the process.

As evidence I offer the following: First, the term prayer itself is strongly associated with theistic religious traditions. Second, the IP studies are full of both explicit and implicit references to the role of God as the "mechanism" of healing. For example, in the best known IP investigation, Byrd (1988) states that the study was designed to answer the question: "Does intercessory prayer to the Judeo-Christian God have any effect on the patient's medical condition ...?" (p. 826) and those who did the praying in this study were all born-again Christians. Harris et al. (1999), as noted above, required that the intercessors believe in a personal and healing God. Cha et al. (2001) used members of "Christian denominations" (p. 783) as their intercessors. Matthews et al. (2001) utilized an on-going Christian prayer group. Walker et al. (1996) incorporated Protestant, Catholic, or Jewish volunteers who reported more than 5 years of regular intercessory prayer experience. Matthews et al. (2000) drew their intercessors from lay, volunteer prayer ministers from Christian Healing Ministries. Aviles and colleagues (2001) utilized only self-professed Christians as intercessors. Palmer et al. (2004) drew from a local church's prayer chain and from retired women living in a Christian retirement home. Tlocynski and Fritsch (2002) specifically included reference to God in their instructions for the prayers. If God is not assumed to be an integral part of the process through which prayer operates, then why are the intercessors believers, and in some cases, the prayer interventions designed specifically to appeal to God? If the natural power of IP is all that is under study with no necessity of God, then the IPs of atheists would be expected to be as effective as the IPs of the devout. Yet, atheists are not included as intercessors. Is it possible that all these Christians and some Jews really prayed without invoking God or expecting that their prayers were being heard by God? This really seems nonsensical. I believe it is extraordinarily safe to say that existing IP studies are based on the notion of God.

Concluding Thoughts on Prayer and the Role of Science

So what then to make of IP for healing? Should the weak scientific evidence for the potency of IP and the substantial questions regarding the appropriateness of scientific testing of IP weaken enthusiasm for IP among believers? Of course not. In fact, understanding that IP is not amenable to scientific testing should have no effect whatsoever on the particular beliefs or practices of religious persons. There are many very important issues that have dramatic consequences for human existence that are not amenable to scientific study: Ethics, art, and morality are three areas that quickly come to mind. In fact, the results of scientific experimentation are never normative of behavior, they are only statements of observations given certain conditions. Christians should continue, as they always have, to offer up prayers on the basis of their belief and understanding that a sovereign God hears and answers according to God's will. The practice of prayer by believers is not based on controlled scientific evidence. It is instead the response of thankful hearts to a loving God. Believers offer IP because they receive a sense of self-being from God and, in return, have a keen desire to offer themselves and their concerns. Thus for many IP involves an act of self-offering that demands no scientific proof; in fact, the search for such proof may be spiritually inauthentic.

Regarding science, my hope is that scientists who are Christian, as well as those of other faith traditions and those with no religious connection at all, will stop expending valuable resources on scientifically controlled and blinded studies of IP and will instead focus on important spiritual and religious variables that are the proper subject of scientific inquiry. Much of this work is underway but more needs to be done. In the area of prayer, for example, analysis of psychophysiological variables during and after prayer could prove enlightening. Studies on the effects of individuals' own personal prayers for their own healing are completely appropriate as are studies of group or individual prayers for individuals who are aware that they are the recipients of prayer. A particularly noteworthy population to study are the intercessors themselves, i.e., what effect does prayer for others have on the health and well-being of the pray-ers? Measurement of personality and other patient variables that may interact with prayer to influence functioning would add specificity to this literature. In fact, prayer and other religious phenomena, when contextualized within the natural order, make most interesting and befitting subjects of scientific inquiry. It is ironic that to the extent that IP studies successfully incorporate blinding of patients and intercessors, and thus move away from naturalistic explanations for their effects, they remove themselves from the arena of appropriate scientific studies. Continued investigation of religious coping, study of the unique role of social support provided in a Christian context, investigation of how religious beliefs interact with cognitive patterns to influence functioning, or further analysis of how religious orientation influences psychophysiological indicators of stress response are all examples of topics worthy of programmatic in depth examination by serious scholars using the methods of science. (See Koenig, McCullough, & Larson, 2001 and Spilka, Hood, Hunsberger, & Gorsuch, 2003 for excellent overviews of much of the research on religion and psychological and physical health). Indeed there are innumerable questions regarding faith and mental and physical for which science is not only appropriate but is singularly qualified to address. Let us wisely choose among these for continued study.

REFERENCES

Aviles, J. M., Whelan, S. E., Hernke, D. A., Williams, B. A., Kenny, K. E., O'Fallon, W. M., et al. (2001). Intercessory prayer and cardiovascular disease progression in a coronary care unit population: A randomized controlled trial. Mayo Clinical Proceedings, 76, 1192-1198.

Barncs, P., Powell-Griner, E., McFann, K., & Nahin, R. (2002). CDC advance data report #343: Complementary and alternative medicine use among adults: United States, 2002. U.S. Government: Washington, DC.

Byrd, R. C. (1988). Positive therapeutic effects of intercessory prayer in a coronary care unit population. Southern Medical Journal, 81, 826-829.

Cha, K. Y., Wirth, D. P., & Lobo, R. A. (2001). Does prayer influence the success of in vitro fertilization-embryo transfer? The Journal of Reproductive Medicine, 46, 781-787.

Chibnall, J. T., Jeral, J. M., & Cerullo, M. A. (2001). Experiments on distant intercessory prayer. Archives of Internal Medicine, 161, 2529-2536.

Clark, K. M., Friedman, H. S., & Martin, L. R. (1999). A longitudinal study of religiosity and mortality risk. Journal of Health Psychology, 4, 381-392.

Collipp, P. J. (1969). The efficacy of prayer: A triple-blind study. Medical Times, 97, 201-204.

Forlenza, M. J., & Baum, A. (2002). Psychoneuroimmunology. In T. J. Boll (Series Ed.) & R. G. Frank, A. Baum, & J. L. Wallander (Vol. Eds.), Handbook of clinical health psychology: Vol. 3. Models and perspectives in health psychology (pp. 81-114)). Washington, DC: American Psychological Association.

Harris, W. S., Gowda, M., Kolb, J. W., Strychacz, C. P., Vacek, J. L., Jones, P. G., et al. (1999). A randomized, controlled trial of the effects of remote, intercessory prayer on outcomes in patients admitted to the coronary care unit. Archives of Internal Medicine, 159, 2273-2278.

Hummer, R. A., Rogers, R. G., Nam, C. B., & Ellison, C. G. (1999). Religious involvement and U.S. adult mortality. Demography, 36, 273-285.

Ironson, G., Balbin, E., & Schneiderman, N. (2002). Health psychology and infectious diseases. In T. J. Boll (Series Ed.) & S. B. Johnson, N. W. Perry, Jr., & R. H. Rozensky (Vol. Eds.), Handbook of clinical health psychology: Vol. 1. Medical disorders and behavioral applications (pp. 5-36). Washington, DC: American Psychological Association.

Joyce, C. R., & Welldon, R. M. (1965). The objective efficacy of prayer: A double-blind clinical trial. Journal of Chronic Diseases, 18, 367-377.

Kiecolt-Glaser, J. K., McGuire, L., Robles, T. F., & Glaser, R. (2002). Psychoneuroimmunology: Psychological influences on immune function and health. Journal of Consulting and Clinical Psychology, 70, 537-547.

Koenig, H. G., McCullough, M. E., & Larson, D. B. (2001). Handbook of religion and health. New York: Oxford.

Masters, K. S., & Bergin, A. E. (1992). Religious orientation and mental health. In J. F. Schumaker (Ed.), Religion and mental health (pp. 221-232). London: Oxford.

Masters, K. S., Bergin, A. E., Reynolds, E. M., & Sullivan, C. E. (1991). Religious life-styles and mental health: A follow-up study. Counseling and Values, 35, 211-224.

Masters, K. S., Hill, R. D., Kircher, J. C., Lensegrav-Benson, T. L., & Fallon, J. A. (2004). Religious orientation, aging, and blood pressure reactivity to interpersonal and cognitive stressors. Annals of Behavioral Medicine, 28, 171-178.

Masters, K. S., Lensegrav-Benson, T. L., Kircher, J. C., & Hill, R. D. (2005). Effects of religious orientation and gender on cardiovascular reactivity among older adults. Research on Aging, 27, 221-240.

Matthews, D. A., Marlowe, S. M., & MacNutt, F. S. (2000). Effects of intercessory prayer on patients with rheumatoid arthritis. Southern Medical Journal, 93, 1177-1186.

Matthews, W. J., Conti, J. M., & Sireci, S. G. (2001). The effects of intercessory prayer, positive visualization, and expectancy on the well-being of kidney dialysis patients. Alternative Therapies in Health and Medicine, 7,42-52.

McCullough, M. E., Hoyt, W. T., Larson, D., Koenig, H. G., Thoresen, C. (2000). Religious involvement and mortality: A meta-analytic review. Health Psychology, 19, 211-222.

O'Laoire, S. (1996). An experimental study of the effects of distant, intercessory prayer on self-esteem, anxiety, and depression. Alternative Therapies in Health & Medicine, 3, 38-53.

Palmer, R. F., Katerndahl, D., & Morgan-Kidd, J. (2004). A randomized trial of the effects of remote intercessory prayer: Interactions with personal beliefs on problem-specific outcomes and functional status. The Journal of Alternative and Complementary Medicine, 10, 438-448.

Powell, L. H., Shahabi, L., & Thoresen, C. E. (2003). Religion and spirituality: Linkages to physical health. American Psychologist, 58, 36-52.

Smith, T. W., & Ruiz, J. M. (2002). Psychosocial influences on the development and course of coronary heart disease: Current status and implications for research and practice. Journal of Consulting and Clinical Psychology, 70, 548-568.

Spilka, B., Hood, R. W., Jr., Hunsberger, B., & Gorsuch, R. (2003). The psychology of religion: An empirical approach (3rd ed.). New York: Guilford.

Strawbridge, W. J., Cohen, R. D., Shema, S. J., & Kaplan, G. A. (1997). Frequent attendance at religious services and mortality over 28 years. American Journal of Public Health, 87, 957-961.

Suchday, S., Tucker, D. L., & Krantz, D. (2002). Diseases of the circulatory system. In T.J. Boll (Series Ed.) & S. B. Johnson, N. W. Perry, Jr., & R. H. Rozensky (Vol. Eds.), Handbook of clinical health psychology: Vol. 1. Medical disorders and behavioral applications (pp. 203-238). Washington, DC: American Psychological Association.

Tloczynski, J., & Fritzsch, S. (2002). Intercessory prayer in psychological well-being: Using a multiple-baseline, across-subjects design. Psychological Reports, 91, 731-741.

Walker, S. R., Tonigan, J. S., Miller, W. R., Comer, S., & Kalich, L. (1996). Intercessory prayer in the treatment of alcohol abuse and dependence: A pilot investigation. Alternative Therapies in Health & Medicine, 3,79-87.

Williams, R. B. (2001). Hostility (and other psychosocial risk factors): Effects on health and the potential for successful behavioral approaches to prevention. In A. Baum, T. A. Revenson, & J. E. Singer (Eds.). Handbook of health psychology (pp. 661-668). Mahwah, NJ: Erlbaum.

AUTHORS

MASTERS, KEVIN S.: Address: Department of Psychology, 430 Huntington Hall, Syracuse University, Syracuse, NY 13244-2340. Title: Associate Professor and Director of Clinical Training. Degrees: BA, Cedarville College; MA, University of Dayton; PhD, Brigham Young University. Specializations: Health psychology specifically religious/spiritual variables and cardiovascular reactivity to stress; psychological variables and exercise; psychological predictors of lumbar surgery outcome.

KEVIN S. MASTERS

Syracuse University

I wish to thank Jason T. Goodson and Glen I. Spielmans for their help with this project. Additionally, I want to thank Craig K. Ewart for his insightful and provocative thoughts in response to an earlier version of the manuscript. Correspondence concerning this article may be sent to Kevin S. Masters, Ph.D., Department of Psychology, 430 Huntington Hall, Syracuse University, Syracuse, NY 13244-2340. Email: [e-mail]