Variations in pastors' perceptions of the etiology of depression by race and religious affiliation.

Depression is a major, preventable problem in the United States, yet relatively few individuals seek care in traditional mental health settings. Instead, many choose to confide in friends, family, or clergy. Thus, it is important to discover how clergy perceive the definition of and etiology of depression. The author conducted a survey with 204 Protestant pastors in California. Multinomial logistic regression revealed a statistically significant difference in how depression is perceived based on race. Caucasian American pastors more readily agreed with the statement that depression was a biological mood disorder, while African American pastors more readily agreed that depression was a moment of weakness when dealing with trials and tribulations. Also, mainline Protestants more frequently disagreed with statements about spiritual causes of depression than Pentecostals and non-denominational pastors. The findings suggest that racial and religious affiliational influences shape how pastors view, and ultimately intervene, in the area of depression.

Introduction

When given a choice to turn to mental health professionals or clergy, many individuals turn to clergy first. In the National Cormorbidity Survey, Wang, Berglund, and Kessler found that a quarter of those who ever sought treatment for mental disorders did so from a clergy member. Clergy were contacted more often than psychiatrists or general medical doctors; psychiatrists and general practitioners were each contacted about one-sixth of the time (Wang et al. 2003). In a survey conducted <1 month after September 11th, approximately 60% of all the respondents said they would likely seek help from a spiritual counselor, compared to 45% of all the respondents who would likely seek help from their physician and 40% who would seek help from a mental health care professional. According to Milstein (2003), people do not choose these patterns of help-seeking because they are unaware of mental health care resources, but because they are more familiar with clergy, clergy do not charge fees, and less stigma is involved in discussing one’s personal problems with clergy (Milstein 2003).

In addition, the clergy are often first responders to crises. In a systematic research synthesis of the psychological literature on collaboration between clergy and mental health professionals (completed on journals between 1970 and 1999), Oppenheimer et al. (2004) found that one of the major themes identified in the literature was the recognition of clergy as frontline mental health workers. Clergy handle funeral arrangements, marital conflicts, and personal crises in the lives of parishioners and community members.

Clergy intervene in families’ lives at major developmental milestones. They are involved in birth processes through Christening ceremonies and baby dedications. They are involved in marriages by providing pre-marital counseling and performing weddings. Clergy intervene when deaths occur in families; they perform funerals and provide bereavement counseling for families. Thus, clergy are very familiar with handling grief, bereavement, loss, and depression. Death, dying and loss issues differ from clinical depression, although pastors handle both issues. In his article An intensive course for clergy on death, dying, and loss, Norman Clemens stated, “… Where the response to [bereavement] crisis has been abnormal or poorly resolved, the clergyman may be the most likely of all community resources to become aware of the problem and to assist in obtaining care from the mental-health professions” (Clemens 1976). Thus, clergy may play an important role in improving the knowledge of and linkage to depression care (Kramer et al. 2007).

Although clergy have been identified as a community mental health resource, few empirical studies of clergy practices when handling depression have been conducted. This study introduces data from the author’s Clergy Depressive Counseling Survey, which is a survey of pastoral clergy who have counseled depressed individuals in the course of their ministerial careers. There are no instruments specifically addressing these issues anywhere in literature, so an original survey was created; this process is explained in detail in the methodology section.1

Clergy Views on the Etiology of Depression

Few researchers oppose the fact that clergy counsel individuals frequently. Minimal discussion exists in the literature, however, about how clergy’s views shape their decisions about mental health referral and intervention. It is logical that the counseling that clergy provide for depression will be heavily influenced by the views they have about depression. Robert Taylor et al., in their article on the role of clergy in African American churches, noted that behaviors may be defined differently by clergy. In turn, these definitions shape beliefs about the best solutions to address the behaviors (Taylor et al. 2000). Little, if any, research has been done to determine what clergy views are about handling specific issues such as depression.

A body of research exists that examines differences in views about mental health service, based on race. The literature has established that African Americans tend to terminate traditional mental health treatment earlier than Caucasians, for example. Millet et al. (1996) discusses one explanation for this behavior, proposing that members of the two groups hold different views about mental health problems and their treatment. Millet tested this hypothesis empirically and found that African American respondents in his study rated spiritual factors as more important in the etiology and treatment of the difficulties presented in vignettes than did Caucasian Americans. Millet said that African American and Caucasian Americans possess different cognitions with regard to the etiology and treatment of mental health problems. Because of this, it is expected that African Americans and Caucasian Americans would think, feel, and act differently in response to mental health problems, “either in themselves or in others”. Millet also stated that “when confronted with a mental health problem, (African Americans and Caucasians) would likely seek help at different points in its course, turn to different sets of resource people, and… expect success from different forms of assistance”. (Millet et al. 1996).

Schnittker et al. found that racial differences in etiological beliefs play a substantial part in explaining African Americans’ tendency to have more negative attitudes than Caucasian Americans toward professional mental health treatment. In their study, African Americans were more likely than Caucasian Americans to reject the idea that mental illnesses are caused by either genetics or an unhealthy family upbringing (Schnittker et al. 2000).

African American ministers may have similar views about the etiology of mental illness issues as the African American population in general. The study by Mollica et al. (1986) of the mental health counseling practices of 214 African American and Caucasian American ministers found that, compared with their Caucasian American peers, African American clergy placed greater emphasis on using religious practices (for example, church attendance) as a method for treating emotional problems (Mollica et al. 1986). Thus, the idea that African American clergy emphasize spiritual causes of mental illness more than a genetic cause has been alluded to in the literature.

There have also been a few studies that point to differences in prevalence of and views about mental illness based on religious affiliation. In a study by Meador et al. (1992) on Religious Affiliation and Major Depression, the relative risk for having major depression was three times greater for Pentecostals than other groups, when other risk factors were controlled for (Meador et al. 1992). Also, the level of conservatism of clergy was discussed by at least one study. Researchers suggested that members of the clergy with liberal theologies are more likely to make referrals to mental health agencies. In contrast, those who endorse conservative theologies are more likely to attempt to treat people with symptoms of psychiatric disorders themselves (Taylor et al. 2000).

This study examines clergy views on the definition and etiology of depression, and looks at differences based on race and religious affiliation. The hypothesis is that both clergy race and religious affiliation will each play a significant factor when determining clergy views about depression. Also, the study explores whether these differences remain when accounting for education level, maturity level, SES, or gender of the pastors.

Methods

Population and Sample Selection

The population of interest for the Clergy Depressive Counseling Survey included pastors and ministers in the State of California. “Pastors” are defined as heads of churches from 26 Protestant denominations. Ministers are licensed and/or ordained individuals licensed by an authoritative overseeing church body. Clergy is an all-inclusive term that includes both Protestant pastors and other ministers.

Inclusion/Exclusion Criteria

Pastors and ministers who had either accessible e-mail addresses or church mailing addresses were included in the study. Those excluded from the study included Christian clergy from denominations outside of the 26 listed (including Jesus Christ of Latter Day Saints, Jehovah Witness, Bahai Faith, Sith, Catholic churches), clergy whose access information could not be obtained, clergy outside of the geographic area in question, and clergy who were not English speaking.

Protestant pastors were chosen for this study, allowing an exploration of a fairly heterogeneous sample of Christian pastors from a variety of denominations. However, Catholic pastors, leaders of other Christian faiths, and leaders from other religions were not included due to sample size constraints for this study.

Sampling Methods

A proportional stratified random sampling plan was used. First, a list of the 491 cities in California was created, and the cities were placed in alphabetical order. To begin a random selection process amongst these 491 cities, a number was assigned to each city by random selection procedures. The randomizer at www.randomizer.org was used to generate 491 random sampling numbers (Urbaniak and Plous 1997). These random numbers were downloaded into Excel, and the list of alphabetical cities was assigned to each of these random numbers as they were listed. The cities were then sorted by random number assigned.

Next, a sampling of 10% of the churches from each of the cities (in random sampling order) was chosen. The 10% was taken from two different yellow page listings of churches (http://www.Superpages.com and http://www.switchboard.com), after those churches not meeting inclusion criteria were excluded. For each city, all of the churches in that city that met inclusion criteria were listed in a Word document. The total number of churches within that city was counted. Then, randomizer.org was used to obtain random numbers for 10% of the churches listed, and 10% of the churches were chosen randomly. After the 10% have been obtained for that city, the next randomized city was examined, all of the churches in that city were listed, and 10% was obtained. This continued until approximately 1,000 churches were obtained as a sampling frame.

Survey Instrument

To develop a survey instrument sensitive to the religious “cultural” language used by pastors and church members, the researcher first engaged in qualitative research. About 2-h unstructured interviews with church leaders were conducted to find out about views of depression from a church leader’s perspective. The raw data that was analyzed included the verbatim transcripts of recorded, unedited interviews, and the Atlas.ti software program was used to assist the data analysis. Audio-taped interviews were transcribed, and units of text were assigned to coding categories which were conceptually related to the issues surrounding depression and mental health treatment. The categories that emerged were then reviewed by three pastors who had not participated in the initial interviews. These pastors affirmed that the categories which emerged resonated with issues they found salient about depression and mental health treatment. This process aided in strengthening the authenticity of the themes of focus.

Based on those interviews, a preliminary close-ended pilot questionnaire was created and tested by e-mail via a sampling of pastors and leaders all across the US. About 35 responses were obtained, which aided in fine tuning of the instrument. After changes were noted to the questionnaire, it was then administered in person (in a hardcopy fashion) to ten ministers at a local church in Los Angeles, and feedback was solicited regarding the structure and wording of the survey. Following this, the survey was pilot tested again by e-mail with 45 English-speaking pastors of various ethnicities nationwide. None of the pastors which engaged in the pilot testing or interviews were part of the actual survey.

In the actual study, pastors responded to a survey instrument with 45 items. The pastors were asked various demographic questions; further detail is given in a table describing descriptive statistics located elsewhere in this paper. When asked about what depression meant to them and their views on the causes of depression, pastors were presented with questions that they answered via a 5-point Likert scale with the following choices; almost always true, usually true, occasionally true, usually not true, and almost never true. The UCLA Office for the Protection of Research Subjects reviewed and approved all IRB requirements for this study, including pilot testing. Data collection began July 2006 and concluded in August 2007.

Data Collection Procedure

The study involved both e-mail and mail out attempts at recruitment; a mixed mode survey strategy was utilized, where e-mail was used as a first choice strategy and mailed letters were sent when e-mail was not available (Schaefer and Dillman 1998).

E-Mail Surveys

Pastors and church leaders were chosen to participate in the study by random sample. Assignment to an e-mailed or mail out survey was not random, however. Instead, it was based on availability of e-mail addresses, given that some pastors had e-mail contact information and others did not.

E-mail addresses were obtained via a meticulous procedure. First, a sampling of 10% of the churches from each of the cities (in random sampling order) was chosen, as stated previously, using two different yellow page listings of churches (http://www.Superpages.com and http://www.switchboard.com. Besides physical address information, both of these websites also include information about church website locations for those churches that had their websites listed. After obtaining a random sample, those churches with websites listed were separated from those who had no website. The website for each church was explored, and an e-mail address for the pastor in question was obtained from their church website.

Where e-mail addresses were obtained, an e-mail letter of invitation was sent. In the e-mail, pastors were invited to participate and also told that their participation was voluntary. If they chose to participate, they were able to click a link on the e-mail which led them to the informed consent form and survey, located at http://surveymonkey.com. If they clicked the link on the e-mail choosing not to participate, their e-mail was automatically removed from the research invitation list. Non-responders received a follow-up e-mail in 2 weeks. After two non-responses, their e-mails were also removed from the invitation list.

Mailed Out Surveys

For pastors who received mailed out surveys, a written letter was sent addressed to the subject with the following information inside; an introductory letter of invitation to participate in the study, a written informed consent form, a copy of the survey, and a stamped self-addressed stamped envelope (addressed with the investigator’s address and information only).

Responses Obtained

A total of 212 responses were obtained from pastors who took the e-mail or mailed survey. Table 1 shows that 1,126 pastors were initially sampled from 61 cities, and out of those, 89 pastors were never contacted. There were 54 pastors who declined participation. In addition, there were 771 non-responses of pastors who did not respond to a first or second mailing/e-mailing. Thus, 1037 pastors were sampled, reflecting an overall response rate of 20%. There was a low response rate from mailed pastors of 14%, and a higher response from e-mailed pastors of 34%. Although fewer pastors were surveyed by e-mail due to difficulty in locating valid e-mail addresses, the response rate of pastors who were e-mailed was over twice that of those who were mailed. The average response rate for e-mail surveys with a single contact is 28.5 and 41% for two contacts (Schaefer and Dillman 1998). Mailed surveys with one follow-up yield an average response rate of 30–35% (Kaplowitz et al. 2004). One reason for this survey’s low response rate is likely due to high turnover rates of pastors over congregations; new ministers or pastors over churches are not reflected in Yellow Page listings, so the mailed survey may have been addressed to a pastor no longer over the congregation. Additional discussion about response rate will occur in the limitations section.

Table 1

Response patterns of pastor survey respondents (n = 204)

	Mailed (%)	E-mailed (%)		Totals (%)
Initial sample	782	344		1,126
No contact	53	36		89
Actual sample (minus no contact)	729	308		1,037
Declined participation	16 (2%)	38 (12%)		54 (5%)
Non-response	612 (84%)	159 (52%)		771 (74%)
Responses with incomplete data	2	6		8
Full responses obtained	99 (14%)	105 (34%)		204 (20%)

Percentages of the sample in each category is given in parenthesis

There was a difference between those who actively declined participation in the study versus those who chose not to respond to the invitation. Of those who actively declined participation, 80% ran churches in primarily Caucasian areas.2 On the other hand, of those who chose not to respond to invitations received, a majority (56%) ran churches in primarily African American areas. Those pastors serving African American areas may be less comfortable with saying “no” to participation in a survey than those pastors who serve Caucasian areas. When looking briefly at socioeconomic status (based on zip code), those in higher socioeconomic areas were more likely to actively decline participation or to participate. Those in lower socioeconomic areas were more likely to refrain from responding to invitations. Also, pastors declined participation more readily by e-mail than by mail. A small number of pastors stated why they declined, citing “I’m too busy” as the primary reason for declining participation.

Data Analysis Procedure

Operationalization of Variables

Table 2 discusses the operationalization of religious affiliation for this study. Religious affiliation was defined based upon the definition of religious affiliation in the study by Meador et al. (1992) on Religious Affiliation and Major Depression. Race was defined by pastor’s self-report based on a list of close ended options and one open ended “other—please explain” option.

Table 2

Religious affiliation

Mainline protestants(Presbyterian, Lutheran, Congregational, Reformed, United Church of Christ, Episcopal, Methodist, African Methodist Episcopal, Christian Methodist Episcopal, Disciples of Christ, Christian Church, Salvation Army, Quaker, Community Church)

Conservative protestants(Baptist, United Missionary, Nazarene, Church of Christ, Primitive Baptist, Freewill Baptist, Seventh Day Adventist, Southern Baptist, Christian and Missionary Alliance, and other Fundamentalists)

Pentecostals(Church of God, Assemblies of God, Church of God in Christ, Holiness, Apostolic, Charismatic, Foursquare, Evangelical, Vineyard, Full Gospel)

Non-denominational(Self declared as Non-Denominational)

Other(Latter-day Saints, Unitarian, Others not noted)

Adapted based on Meador et al. study, 1992

Analysis Method

A total of eight respondents were not used in the analysis due to missing data, resulting in 204 analyzed results. Descriptive statistics, multinomial logistic regression and Wald/LR analyses were generated with Stata Software (Version 10). Table 2 provides the descriptive statistics that were used to profile the characteristics of the pastors in the study. Multinomial logistic regression was used to examine the effects of race and religious affiliation on views about the definition and causes of depression (P < .05 was considered statistically significant). Additional logistic regressions were run on other variables, including gender, age, secular education, theological education, number of years in the ministry, and other variables to determine if they were significantly associated with views about the definition and causes of depression.

Results

Church and Clergy Characteristics

Table 3 shows the descriptive information for the pastors in the study. Of the 204 responding pastors, 29 were women and 175 were men, ranging in age from 20 to over 65 years of age. Women pastors were either Caucasian or African American; there were no Asian or Hispanic women pastors. In all other demographics, women were comparable to men (in congregation size, age, etc.).

Table 3

Descriptive statistics of pastor survey respondents and their churches (n = 204)

Pastor’s gender		Regions where churches are located
Male	86% (175)	Central Ca	19% (39)
Female	14% (29)	Northern Ca	7% (14)
		Southern Ca	74% (151)
Pastor’s age
20–34 years	7% (15)	SES (based on zip code of church location)
35–49 years	28% (57)	Under 35,000 a year	25% (51)
50–64 years	50% (102)	31–45,000 a year	30% (61)
65 years and up	15% (30)	46–65,000 a year	20% (40)
		Over 65,000 a year	25% (52)
Pastor’s race
Black	25% (51)	Religious affiliation
White	65% (133)	Mainline protestants	29% (60)
Other	10% (20)	Conservative protestants	25% (51)
		Pentecostals	27% (56)
Pastor’s years of ministry		Non-denominational	15% (31)
1–10	11% (23)	Other	3% (6)
11–20	33% (67)
21–30	32% (65)	Congregation size (members)
31–40	17% (34)	<50	14% (29)
Over 40	7% (15)	51–150	39% (79)
		151–250	16% (33)
Pastor’s secular degrees		251–350	5% (11)
None	18% (37)	351–450	5% (11)
Some college, no degree	19% (38)	Over 450	20% (40)
Associate degree	7% (14)
Bachelors degree	31% (64)	Areas of pastor’s secular education
Masters degree	14% (29)	General education	22% (45)
Doctorate degree	9% (19)	Physical or natural sciences	2% (5)
Other (J.D., M.D., LVN)	1% (3)	Social or applied sciences	37% (75)
		Business	14% (28)
Formal pastoral counseling training		Humanities	20% (40)
Had training	25% (50)	Medical/health care	1% (3)
No training	75% (154)	Trade school	1% (3)
Senior or head pastor of the church		Pastor’s theological degrees
Senior pastor	87% (177)	No theological training	4% (9)
Associate minister	13% (27)	Some bible college, no degree	26% (53)
		Theological bachelors degree	14% (29)
		Theological masters degree	36% (74)
		Theological doctoral degree	19% (38)

Sample sizes (N’s) given in parenthesis

Caucasian pastors responded most frequently to the survey (65%), followed by African American clergy (25%). There was a limited response from Asian and Hispanic clergy due to the sampling decision to include English-speaking clergy only. When pastors were asked about their congregational makeup, 41% stated that their churches served a primarily Caucasian congregation; 22% stated that they served a primarily African American congregation; 4% stated that they served primarily Asian, Hispanic, or Native American congregations; and 33% of the pastors stated that their congregations were multi-ethnic.

The churches of these pastors were mainly located in Southern California; 74% of the respondents had churches in this location, despite outreach to other areas of California. The sample represents churches from 80 different cities in California. The cities were randomly chosen: based on the randomization, 21 Northern California Cities, 10 Central California cities, and 30 Southern California cities were chosen. When examining the cities chosen, it happens that the cities in Northern and Central California that were chosen had less numbers of churches in them than the cities that were chosen from Southern California. For example, the city with the largest number of churches chosen from Central California was San Jose, with 374 churches. The city with the largest number of churches chosen from Southern California was Los Angeles, with 1,321 usable listings. Thus, after taking a sample of 10% of churches from each of the selected cities, a greater amount of Southern California churches was randomly selected into this sample.

The majority of pastors (50%) were between 50 and 64 years of age, followed by 28% of pastors who were between 35 and 49 years of age. Most pastors in the sample had quite a bit of experience in the ministry; 33% of the sample had been in the ministry between 11 and 20 years, followed by 32% of the sample which had between 21 and 30 years of ministry experience. In terms of congregational size, most respondents had between 50 and 150 members in their congregation (39%), followed by pastors with large congregations of over 450 members (20%).

A large number of the pastors in the sample (82%) had some level of secular college education. Fifty-six percent of the pastors had at least a Bachelors Degree from a secular university. Nine percent of the pastors had secular Ph.D’s, followed by 14% which had a masters degree from a secular college or university. Their areas of education were diverse; there were at least 35 different types of areas of study that the pastors had received degrees in, including history, anthropology, health care, biology, business, social work, communications, computer science, law, astronomy, economics, education, engineering, liberal arts, psychology, sociology, urban planning, and a host of other disciplines.

Ninety-six percent of the pastors had some type of theological training. Most pastors had more than one type of theological training, which included training given to obtain a minister’s or ordination license, training in seminary, training in bible schools, chaplaincy training, priesthood training, and other types of theological training. Many pastors (69%) held at least a BA in theology, Christian education, or pastoral care from a theological training institution. Nineteen percent of the pastors had theological doctoral degrees, followed by 36% which had a master’s degree in divinity or another theological discipline. Surprisingly, only one-fourth of the pastors surveyed had pastoral counseling training.

Pastors were asked “What does the word depression mean to you?” and “What is the cause of depression?” They were then asked to respond “almost always true”, “usually true”, “occasionally true”, “usually not true”, or “almost never true” to a series of statements associated with these two questions. For the purposes of this analysis, “almost always true” and “usually true” were collapsed into a category called “agree”, and “almost never true” and “usually not true” were collapsed into a category called “disagree”. A chi-square test determined that the relation between race and religious affiliation was statistically significant (Pearson χ2[3] = 11.76, P = .008). However, there were no statistically significant interactions in any of the models tested, so interaction terms were eliminated from the models.

Pastor’s Race and Views About Depression

Table 4 presents the relative risk ratios of pastors’ agreement with the definition and the cause of depression by race. There was a statistically significant difference, based on race, when pastors responded to the statement “Depression is hopelessness that happens when one does not trust God”. Caucasian pastors were significantly less likely to agree with that statement than African American pastors (RR = .31, P = .024), and less likely to say that the statement is occasionally true (RR = .36, P = .012).

Table 4

Relative risk ratios of pastors’ agreement with definition and causes of depression by race (based on Caucasian or African American pastoral race, n = 184)

Questions asked	Agree (vs. disagree)			Occasionally (vs. disagree)
	RR	95% CI	P value	RR	95% CI	P value
Spiritual definition: “Depression is hopelessness that happens when one does not trust God”
African American (base group)	1.00	–	–	1.00	–	–
Caucasian	.31	(.11, .857)	0.024*	.36	(.16, .802)	0.012*
Biological definition: “Depression is a biological mood disorder”
African American (base group)	1.00	–	–	1.00	–	–
Caucasian	6.51	(2.1, 19.95)	0.001*	3.62	(1.3, 9.79)	0.011*
Situational definition: “Depression is a moment of weakness when dealing with trials and tribulations”
African American (base group)	1.00	–	–	1.00	–	–
Caucasian	.86	(.37, 2.01)	.727	1.04	(.48, 2.27)	.922
Intrinsic definition: “Depression is due to a person feeling worthless or having low self-esteem”
African American (base group)	1.00	–	–	1.00	–	–
Caucasian	.30	(.06, 1.48)	.139	.45	(.09, 2.20)	.321
Moral cause: “Depression is due to a moral problem in one’s life”
African American (base group)	1.00	–	–	1.00	–	–
Caucasian	.88	(.25, 3.06)	.837	.75	(.25, 2.27)	.613
Medical cause: “Depression is due to a medical or biological condition”
African American (base group)	1.00	–	–	1.00	–	–
Caucasian	1.80	(.54, 5.98)	.336	1.35	(.47, 3.88)	.572
Spiritual cause: “Depression is due to a lack of faith in God”
African American (base group)	1.00	–	–	1.00	–	–
Caucasian	.49	(.18, 1.30)	.152	.77	(.35, 1.65)	.496

* P < .05

On the other hand, when asked to respond to the statement “Depression is a biological mood disorder”, Caucasian pastors were significantly more likely to both occasionally agree (RR = 3.6, P = .011) and to wholeheartedly agree (RR = 6.5, P = .001) with the statement than were African American pastors.

Religious Affiliation and Views About Depression

Table 5 presents the relative risk ratios of pastors’ agreement with the definition and the cause of depression by religious affiliation. Consistently, mainline protestant pastors and Pentecostal pastors had differing views about depression. Mainline protestant pastors disagreed with the statement “Depression is a moment of weakness when dealing with trials and tribulations” more often than Pentecostals, who were more likely to agree with the statement (RR = 3.6, P = .034). When asked if depression is “due to a spiritual or moral problem in one’s life”, mainline protestants disagreed with this statement more often than Pentecostals, who both wholeheartedly agreed (RR = 13.5, P = .004) and occasionally agreed (RR = 6.03, P = .028). When asked if depression is “due to a lack of faith in God”, Pentecostals wholeheartedly agreed with this statement (RR = 11.7, P = .004), and non-denominational pastors also wholeheartedly agreed (RR = 5.84, P = .045), in comparison to mainline protestant pastors, who disagreed more often.

Table 5

Relative risk ratios of pastors’ agreement with definition and causes of depression by religious affiliation (n = 204)

Questions asked	Agree (vs. disagree)			Occasionally (vs. disagree)
	RR	95% CI	P value	RR	95% CI	P value
Spiritual definition: “Depression is hopelessness that happens when one does not trust God”
Mainline protestants (base group)	1.00	–	–	1.00	–	–
Conservative protestants	1.44	(.36, 5.77)	.611	1.004	(.42, 2.42)	.993
Pentecostals	2.33	(.58, 9.39)	.234	1.36	(.54, 3.43)	.513
Non-denominational	3.01	(.74, 12.26)	.124	1.83	(.71, 4.70)	.208
Biological definition: “Depression is a biological mood disorder”
Mainline protestants (base group)	1.00	–	–	1.00	–	–
Conservative protestants	.31	(.07, 1.39)	.126	.79	(.17, 3.53)	.753
Pentecostals	.37	(.07, 2.01)	.252	1.67	(.33, 8.50)	.535
Non-denominational	.47	(.09, 2.43)	.364	1.34	(.26, 6.83)	.725
Situational definition: “Depression is a moment of weakness when dealing with trials and tribulations”
Mainline protestants (base group)	1.00	–	–	1.00	–	–
Conservative protestants	2.85	(.90, 9.05)	.075	1.75	(.70, 4.38)	.233
Pentecostals	3.60	(1.1, 11.74)	.034*	1.75	(.71, 4.94)	.207
Non-denominational	2.20	(.67, 7.23)	.196	1.75	(.45, 3.15)	.716
Intrinsic definition: “Depression is due to a person feeling worthless or having low self-esteem”
Mainline protestants (base group)	1.00	–	–	1.00	–	–
Conservative protestants	.76	(.17, 3.36)	.720	.66	(.16, 2.83)	.581
Pentecostals	3.76	(.38, 37.36)	.259	2.88	(.29, 28.22)	.363
Non-denominational	.51	(.11, 2.34)	.390	.61	(.14, 2.60)	.500
Moral cause: “Depression is due to a moral problem in one’s life”
Mainline protestants (base group)	1.00	–	–	1.00	–	–
Conservative protestants	2.58	(.66, 10.05)	.173	2.19	(.74, 6.42)	.155
Pentecostals	13.48	(2.3, 78.73)	.004*	6.03	(1.21, 30.0)	.028*
Non-denominational	3.61	(.85, 15.22)	.081	2.43	(.74, 7.97)	.143
Medical cause: “Depression is due to a medical or biological condition”
Mainline protestants (base group)	1.00	–	–	1.00	–	–
Conservative protestants	.28	(.05, 1.67)	.161	.67	(.12, 3.91)	.659
Pentecostals	.13	(.02, .77)	.025*	.55	(.10, 3.08)	.493
Non-denominational	.17	(.03, 1.04)	.055	.61	(.10, 3.53)	.578
Spiritual cause: “Depression is due to a lack of faith in God”
Mainline protestants (base group)	1.00	–	–	1.00	–	–
Conservative protestants	3.43	(.63, 18.69)	.155	1.29	(.55, 3.03)	.564
Pentecostals	11.70	(2.2, 62.67)	.004*	2.45	(.94, 6.38)	.066
Non-denominational	5.84	(1.03, 32.9)	.045*	2.18	(.87, 5.48)	.097

* P < .05

In comparison, when asked if depression is “due to a medical or biological condition”, mainline protestant pastors wholeheartedly agreed with the statement more often than Pentecostals, who disagreed (RR = .13, P = .025). Thus, mainline protestant pastors more likely disagreed with a spiritual or moral cause for depression, while Pentecostal pastors were more likely to disagree with a medical cause.

Discussion

The results of this study are important, because they show that mental health practitioners and researchers must be aware that attributions about depression in first responders such as pastors are affected by a number of factors, including race and denomination. The study results were in line with past research stating that African Americans think differently about mental health issues such as depression. For instance, Cooper and her colleagues noted that African American patients used spirituality to help cope with their depression more often than Caucasian American patients, and African American patients cited spirituality as a coping mechanism more frequently than Caucasian American patients. African American patients also discussed utilizing church members for support more frequently than Caucasian American patients (Cooper-Patrick et al. 1997).

It should be noted that Caucasian and African American pastors were congruent in answering many questions in this study. The area of divergence, however, is that African American pastors in the study were more open to the idea that depression can be defined on a spiritual basis—that is, it is hopelessness resulting from a lack of trust in God. African American pastors were much less likely to agree with the idea of depression being defined as a “biological mood disorder” than Caucasian pastors. It is very interesting that, for this study, race influenced how pastors defined depression.

Yet, when examining beliefs about the etiology of depression, answers were influenced more by religious affiliation factors. Mainline Protestants in the study were very committed to their view that depression is caused by medical or biological conditions rather than spiritual causes. This is a significantly different belief than that of Pentecostals, who were more likely to believe that depression was caused by spiritual problems or moral problems rather than biological reasons.

Mainline Protestants were more likely to view depression in line with mental health professionals; they were more likely to see depression as having a biological component, and more likely to see it as being separate from a religious issue. Pentecostals in particular were more likely to view depression as an issue that depends on the situation and felt depression was strongly influenced by spiritual causes.

Thus, despite major media coverage framing depression as a pharmacological issue, some religious and cultural groups retain alternative explanations of depression. Because clergy are often the first contact many individuals with depression have, it is vital to understand that these differences in views about depression exist.

Race and SES are often correlated, so SES (based on the zip code where the pastor’s church is located) was controlled for. Yet, SES did not explain the variance for the views of depression as a biological mood disorder or as an issue of trusting God; the models were still statistically significant. In addition, controlling for gender, age, and pastoral counseling training had no significant affect on race and religious affiliation differences when defining depression.

When looking at the breakdown of race by denomination, the majority of African Americans in the sample are from conservative churches (35%), followed by Pentecostal churches (33%) and non-denominational churches (22%). Very few African Americans (10%) were from mainline protestant churches. On the other hand, the majority of Caucasian pastors ran mainline churches (34%), followed by conservative churches (27%), Pentecostal churches (22%), and lastly non-denominational (16%). As stated previously, there was a statistically significant relation between race and religious affiliation in the sample, yet no statistically significant interactions between race and religious affiliation were present for the research questions explored. Still, culture is a complex issue that cannot easily be defined by individual variables such as race or religious affiliation. For example, there is much history that has motivated African Americans and Caucasians to gravitate toward some denominations and not others. The reasons why and how the divergence is specifically manifested is unknown. However, it is important for researchers and mental health practitioners to be aware that such a divergence exists, and that attribution about depression causes and treatment may be different as a result.

A pastor’s beliefs about the spiritual definition and etiology of depression can both facilitate and hinder treatment for the community members they serve. Pastors who have balanced beliefs in both the biological and spiritual aspects of depression can serve as strong advocates for their counselees. For example, Biebel and Koenig name four basic types of depression—situational depression, developmental depression, spiritual depression, and biological depression; they note that these can overlap (Biebel and Koenig 2004). Pastors who are able to utilize their spiritual expertise, and also refer out when needed, can prove to be extremely effective service providers. On the other hand, pastors who are limited in their views can potentially hinder growth in those they serve. Those who are not open to spiritual views can alienate those who come to them, and those who are not open to biological views can hinder those they serve from a referral they might need. In addition, the messages that proceed from some pulpits may inadvertently help to delay treatment seeking for those congregants who are suffering from clinical depression (Payne 2008).

Limitations and Implications

There are some limitations with this study. First, the study is based on self-report data from the pastor’s perspective that may be affected by recall bias, self-selection for the study, and pastors’ possible desire to please the experimenter, which can all affect accurate reporting. Secondly, this is a cross-sectional close-ended survey. Thus, there is an inability to examine the actual temporal relationships between pastoral interventions and congregational receipt of care. Another limitation is that denominational variation in the study was much greater than the four categories utilized; pastors reported being a part of over 25 different denominations. Also, even within denominations there are differences in emphasis on doctrine, practices, and routines. Lastly, the low response rate was a limitation of this study: based on the response rate, it is unknown if the pastors who self-selected to the study adequately represent pastors in California. Due to language capability limitations, the study was not able to tap into the rich diversity in California, particularly in regards to its Hispanic and Asian American populations.

The strengths of this study include the fact that it is the first of its kind to look in detail at the attributions of clergy on depressive issues in this format. The Clergy Depressive Counseling Survey has an additional component; there are a number of questions pastors were asked surrounding how they actually counsel depression, including how often they counsel depression and suicide, what are their referral practices, and what would they do in situations where they encounter depression or suicidal ideation. The counseling practices of these pastors will be discussed in a later article.

Clergy are often the first responders to mental health crises such as depression and suicide. For example, in the African American community, “only 4.3% of those needing professional help enter directly into the professional system without informal consultation” (Neighbors and Jackson 1984, p. 633). Thus, it is extremely important that researchers continue to engage in empirical studies on the important topic of how clergy define and treat depression. Future directions for research than can enhance empirically based knowledge of depression etiology may include exploring the psychometric properties of questionnaires such as the one used in this study, moving toward clinical measures or scales tailored to first-responders in the community, and translational research; translating scientific discoveries about depression into practical applications to be used by the community at large.