Abstract
There are few population-level insights into the use of traditional healers and other forms of alternative care for the treatment of common mental disorders in sub-Saharan Africa. We examined the extent to which alternative practitioners are consulted, and predictors of traditional healer visits. A national survey was conducted with 3651 adults South African between 2002 and 2004 using the World Health Organization Composite International Diagnostic Interview (CIDI) to generate diagnoses. A minority of participants with a lifetime DSM-IV diagnosis obtained treatment from Western (29%) or alternative (20%) practitioners. Traditional healers were consulted by 9% of the respondents and 11% consulted a religious or spiritual advisor. Use of traditional healers in the full sample was predicted by older age, black race, unemployment, lower education and having an anxiety or a substance use disorder. Alternative practitioners, including traditional healers and religious advisors, appear to play an important role in the delivery of mental health care in South Africa.
INTRODUCTION
Mental disorders and their associated psychosocial disabilities are a source of considerable morbidity and impose a significant drain on national resources (Lopez et al., 2006). Results from the WHO World Mental Health Consortium demonstrate that common mental disorders are highly prevalent in both the developed and the developing world (Demyttenaere et al., 2004). The majority of the world’s 450 million people who suffer from psychiatric morbidity live in developing countries, and less than 10% have access to mental health care (WHO, 2001).
Results from the South African Stress and Health Study, the first nationally-representative study of psychiatric morbidity in South Africa, indicate that approximately 30% of adults have experienced a DSM-IV disorder in their lifetime; this includes 16% with an anxiety disorder, 10% with a mood disorder and 13% with a substance use disorder (Stein et al., in press). Most South Africans have limited access to psychiatric care. One national study reported that the overall staff/population ratio in the public sector mental health services was 19.5 per 100,000 of the population, with an inter- provincial range of 5.7–31.5 per 100,000 (Lund and Flisher, 2002). SASH found that only one-quarter of South Africans with a recent DSM-IV diagnosis received treatment in the year preceding the interview (Wang et al., 2007)
Several studies have shown that alternative practitioners may play an important role in addressing mental health care needs in South Africa by offering culturally appropriate treatment (Nattrass, 2005; Freeman et al., 1994; Mbanga et al., 2002). In many traditional African belief systems, mental health problems are perceived as due to ancestors or by bewitchment and traditional healers and religious advisors are viewed as having expertise in these areas. Furthermore, these sources of health care are often more accessible than Western forms of mental health care. It has been estimated that there are at least 200 000 healers in South Africa (or aproximately 1 per 500 South Africans) (Abdool Karim et al., 1992).
There are a range of different types of alternative practitioners in South Africa. For example, traditional healers include the herbalist and the diviner (Edwards, 1986; Freeman and Motsei, 1992). The diviners are believed to be specialists in divination within the supernatural context which gives them the ability to divine the cause of illness and misfortune. The herbalist specializes in the production of herbal medicines (Ngubane, 1977). Religious and spiritual leaders include faith healers who integrate Christian ritual and traditional practices, and belong to one of the Independent African churches (Edwards et al., 1983).
Results of small studies conducted in South Africa investigating individuals with a mental illness report that approximately one-half (41–61%) of patients have consulted a traditional healer (Freeman et al., 1994; Ensink and Robertson, 1999). Although there is a gap in the literature on predicting traditional healer use in South Africa, studies conducted in Zimbabwe and Tanzania result in conflicting findings. In Tanzania, multivariate analysis indicated that being better educated, older, widowed or separated and being of Christian faith were independently associated with consulting a traditional healer (Ngoma and Prince, 2003). In Zimbabwe, a different pattern emerged, in that patients of traditional healers were more likely to be female and also to be unemployed, with less education (Patel et al., 1997). However, these have been small studies employing a range of convenience sampling methods, leading to results that are difficult to generalise to the South African population as a whole. In turn, the extent to which alternative practioners are consulted, and predictors of traditional healer visits, remains poorly understood. We used the SASH dataset to examine the use of alternative practitioners for mental health concerns among the South African population.
METHODS
The South African Stress and Health (SASH) study (Williams et al., 2004) was undertaken as part of the World Mental Health Survey (Demyttenaere et al., 2004) to investigate the prevalence of mental disorders in the South African context. The survey was conducted between January 2002 and June 2004. The rationale and survey methods have been detailed previously (Williams et al., 2004) and are briefly summarized here. Ethical approval was provided by the University of Michigan, Harvard Medical School, and by a single project assurance of compliance from the Medical University of South Africa that was approved by the National Institute of Mental Health.
Sample Selection
The study population consisted of South Africans who resided in both households and hostels and were a minimum of 18 years old. The sample excluded those individuals living in institutions (including hospitals, prisons, mental health institutions and military bases). The sample was selected using a multi-stage area probability sample design. First, Enumerator Areas (EA; a unit of census administration) used in the 2001 national census were stratified according to province, location (rural/urban) and majority population group (African, coloured, white or Indian); 960 EAs were selected from the resulting strata, with the number of EAs selected per stratum proportional to the estimated stratum population (minimum, 1; maximum, 85). Second, within each EA a random sample of 5 households was selected and finally the third stage consisted of a random sub-selection of a single adult responded in each selected sample housing unit.
Up to three attempts were made to contact each respondent selected to participate. The overall response rate was 85% and the final sample consisted of 3651 individuals. The SASH interviewers received intensive training for one week in centralized group sessions. The face-to-face interviews lasted approximately three and a half hours, although a number of interviews required more than one visit to complete. The interviews were conducted in one of six languages: English, Afrikaans, Zulu, Xhosa, Northern Sotho, and Tswana.
Measures
Participant demographic characteristic were collected using standard questionnaire items. It is noteworthy to disclose that race was assessed using the race groups as defined by the apartheid government (Asian, black, coloured, white), as these categories are still relevant in terms of health outcomes. In this context ‘black’ refers to not being coloured, white or Indian/Asian. While the term ‘coloured’ refers to an ethnic group of people who possess some degree of sub-Saharan ancestry, but not enough to be considered Black during apartheid. Additionally, income was divided into categories of R1–5000 (US$1–$714), R5001–25 000 (US$714–$3 571)), 25 000–100 000 (US$ 571–$14 285), R 100 000+ (US$ 14 285+).
A modified version of the World Mental Health Composite International Diagnostic Interview (WMH-CIDI) was used to assess lifetime DSM-IV disorders and treatment (Kessler & Uston, 2004). The lifetime DSM-IV disorders included anxiety disorders (panic disorder, agoraphobia, social phobia, generalized anxiety disorder and posttraumatic stress disorder), mood disorders (major depressive disorder, dysthmia), and substance use disorders (alcohol and drug abuse and dependence), and disorders associated with impulse control intermittent explosive disorder. In addition, the lifetime experience of trauma was measured.
The mental health service utilization module of the questionnaire assessed participants treatment received in the past 12 months for problems associated with “emotions or mental health”. The list of treatment providers included a (1) psychiatrist; (2) other mental health professional (e.g., psychologist, psychotherapist, psychiatric social worker, mental health nurse, mental health counsellor; (3) a general practitioner or other medical doctor; (4) any other health professional (e.g., nurse or physician’s assistant) (5) traditional healer; (6) religious or spiritual advisor (e.g., minister, priest, or rabbi); (7) any other healer (e.g., chiropractor or spiritualist). For the purpose of this study mental health service utilization was grouped into the following categories:
Western Medicine
Psychiatrist, other mental health professionals (psychologist, psychotherapist, psychiatric social worker, a mental health nurse or mental health counsellor), a general practitioner or any other medical doctor, any other health professional, such as a nurse or physician’s assistant.
Alternative Medicine
Traditional healer, spiritual or religious advisor (e.g., minister, priest, or rabbi), or any other healer (e.g., chiropractor or spiritualist).
Data Management & Analysis
Data were analysed using Stata Version 9.0 (College Station, Texas, USA). All analyses were weighted to adjust for sample selection, non-response, and for residual discrepancies between the sample and the population on a profile of Census demographic and geographic variables. We examined the unadjusted associations between service use as the dependent variable, and lifetime DSM-IV diagnosis (any versus none) and participant demographic characteristics as independent variables. Pearson’s chi-squared test was used for the comparison of proportions and a corrected t-test was used to compare means. Statistical significance was based on 2-sided tests set at α=0.05. In addition, we developed three multiple logistic regression models of whether a traditional healer was consulted for emotional and mental health concerns on demographic variables. The first model was based on the full sample, the second model on participants with a lifetime DSM-IV diagnosis, and the third model on those with no lifetime DSM-IV diagnosis. The results of the regression models were reported as odds ratios (ORs) with 95% confidence intervals (CIs).
RESULTS
The 3651 respondents had a mean age of 37 years; 46% were male; and 76% were black, 10% coloured, 3% Indian/Asian and 10% white (Table 1). A majority of the participants lived in an urban area (62% vs. 38%). Only 40% of the respondents had more than a high school education, 51% were married and 31% were employed.
TABLE 1.
DESCRIPTION OF PARTICIPANT DEMOGRAPHIC CHARACTERISTICS, MENTAL HEALTH SEEKING BEHAVIOURS AND DSM-IV DIANOSIS IN A NATIONALLY REPRESENTATIVE SAMPLE OF SOUTH AFRICAN ADULTS (N = 3651)
| Total Sample % N= 3651 |
Individuals with any DSM-IV disorder (%) |
Individuals with no DSM-IV disorder(%) |
|
|---|---|---|---|
| Age: 18–29 | 39.1 | 36.5 | 40.3 |
| 30–39 | 22.1 | 23.3 | 21.5 |
| 40–49 | 18.1 | 20.8 | 16.8 |
| 50+ | 20.7 | 19.3 | 21.4 |
| Sex: Male | 46.3 | 47.4 | 45.8 |
| Race: Black | 76.2 | 75.2 | 76.7 |
| Coloured | 10.4 | 12.0 | 9.7 |
| White | 10.0 | 9.7 | 10.1 |
| Indian/Asian | 3.4 | 3.2 | 3.5 |
| Currently married | 50.1 | 49.7 | 50.3 |
| Location: Rural | 38.4 | 35.8 | 39.6 |
| Urban | 61.6 | 64.2 | 60.4 |
| Employed: Yes | 31.0 | 33.9 | 29.7 |
| No | 69.0 | 66.1 | 70.3 |
| Education: 0-Gr11 | 59.7 | 59.2 | 59.9 |
| High school + | 40.3 | 40.8 | 40.1 |
| Traumatic Life events (0) | 35.0 | 24.8 | 39.9 |
| 1–2 | 39.5 | 40.6 | 39.1 |
| 3–4 | 17.3 | 22.6 | 14.8 |
| 25+ | 8.1 | 12.0 | 6.3 |
| DSM-IV diagnoses | |||
| Any Disorder | 31.9 | 100.0 | - |
| Any Anxiety Disorders | 16.1 | 50.5 | - |
| Any Mood Disorder | 11.9 | 37.1 | - |
| Any Substance Disorder | 13.3 | 41.6 | - |
| Mental health seeking | |||
| No form of health care | 72.2 | 61.5 | 77.3 |
| Any form of health care* | 27.8 | 38.5 | 22.7 |
| Western Medicine |
Of the total sample, 32% had a lifetime DSM-IV disorder. Overall 28% of the total sample reported seeking treatment during the last 12 months; 39% and 23% of those with and with no lifetime DSM-IV disorder respectively reported seeking treatment during the last 12 months. The prevalence of seeking treatment was similar among those with anxiety (41%), mood (40%) and substance use disorders (41%), respectively.
The average number of visits for mental health care in the past 12 months was 2.7 for traditional healers, and 3.4 for Western medical practitioners (Table 2a). The mean duration of each visit was 43 minutes for a consultation with a traditional healer, and 32 minutes for a Western medical practitioner. The average costs in the last 12 months were R321 and R318 for consulting a traditional healer and Western health practitioner respectively.
TABLE 2.
COMPARISON OF (A) WESTERN PRACTIONERS AND TRADITIONAL HEALER CONSULATIONS FOR PARTICIPANTS WHO USED THESE SERVICES AND (B) TYPES OF ALTERNATIVE MEDICINE PRACTIONERS IN A NATIONALLY REPRESENTATIVE SAMPLE OF SOUTH AFRICAN ADULTS
| (a) | ||
|---|---|---|
| Traditional healer (n=57) | Western Medicine (n=26) | |
| Mean # of visits | 2.7 | 3.4 |
| Average duration of visits (minutes) | 42.7 | 32.2 |
| Average cost spent in the past 12 months (Rand) | 321.08 | 318.31 |
| (b) | |||
|---|---|---|---|
| Type of Practitioner | Total Sample | With DSM-IV Diagnosis |
Without DSM-IV Diagnosis |
| Traditional healer | 6.4 | 8.9 | 5.2 |
| Religious or spiritual advisor | 6.9 | 10.8 | 5.1 |
| Chiropractor or spiritualist | 2.2 | 3.2 | 1.8 |
| Traditional healer exclusively | 2.4 | 3.6 | 1.9 |
Health Seeking Behaviour in the Full Sample
Of the 28% of participants who sought treatment (regardless of whether they met lifetime DSMIV criteria for a disorder or not), 21% were treated by Western practitioners, 13% were treated by alternative practitioners, and 7% by a combination of both Western and alternative practitioners. Similar proportions were reported in participants with and without a common mental disorder (Table 1). Of the subjects who sought treatment from alternative medicine, 6% sought treatment from a traditional healer, 7% of the sample sought treatment from a spiritual or religious advisor, and 2% by another type of healer. However, while 14% of respondents sought treatment from a Western practitioner exclusively, only 2% sought treatment from a traditional healer exclusively (see Table 2b)
Being above 40 years of age and having a 12 month DSM-IV diagnosis (OR=2.3, 95% CI 1.8–2.9) was associated with seeking help from alternative practitioners (Table 3). Traditional healer use in the full sample was associated with age, black race, employment, having less than a grade education, (OR=0.5, 95% CI 0.3–0.7), experiencing a traumatic event, and having an anxiety or a substance use disorder (OR=1.8, 95% CI 1.3–2.4). Ninety seven percent of participants who consulted a traditional healer were black.
TABLE 3.
UNADJUSTED ASSOCIATIONS BETWEEN MENTAL HEALTH SEEKING BEHAVIOUR AND PARTICIPANT DEMOGRAPHIC CHARACTERISTICS OVERALL AND THE PRESENCE OR ABSENCE OF DSM-IV DEFINED MENTAL DISORDER, IN A NATIONALLY REPRESENTATIVE SAMPLE OF SOUTH AFRICAN ADULTS.
| No health Care |
Western Medicine |
OR (95% CI) | Alternative Medicine |
OR (95% CI) | Western & Alternative |
OR (95% CI) | Traditional Healer |
OR (95% CI) | |
|---|---|---|---|---|---|---|---|---|---|
| Age (mean) | 36.1 | 39.7 | 39.4 | 41.0 | 40.0 | ||||
| 18–29 | 42.2 | 29.3 | 1.0 | 30.4 | 1.0 | 24.9 | 1.0 | 28.9 | 1.0 |
| 30–39 | 21.7 | 22.6 | 1.5 (1.2–1.8) | 23.4 | 1.4 (1.0–2.0) | 23.8 | 1.8 (1.2–2.6) | 22.3 | 1.4 (0.9–2.2) |
| 40–49 | 17.1 | 21.9 | 1.8 (1.4–2.4) | 21.7 | 1.5 (1.3–2.1) | 25.3 | 2.3 (1.7–3.3) | 21.1 | 1.6 (1.0–2.6) |
| 50+ | 19.0 | 26.3 | 1.9 (1.4–2.6) | 24.5 | 1.6 (1.3–2.0) | 26.0 | 2.1 (1.4–3.0) | 27.6 | 1.9 (1.3–2.6) |
| Sex: Male | 47.8 | 44.3 | 1.0 | 41.0 | 1.0 | 44.4 | 1.0 | 45.1 | 1.0 |
| Female | 52.3 | 55.7 | 1.1 (0.9–1.4) | 59.0 | 1.3 (1.0–1.7) | 55.6 | 1.1 (0.9–1.4) | 55.0 | 1.1 (0.7–1.5) |
| Race: Black | 76.0 | 73.3 | 1.0 | 86.7 | 1.0 | 83.4 | 1.0 | 96.6 | 1.0 |
| Coloured | 11.1 | 9.2 | 0.9 (0.6–1.4) | 5.2 | 0.4 (0.3–0.6) | 5.2 | 0.4 (0.3–0.7) | 1.4 | 0.01 (0.03–0.4) |
| White | 9.8 | 12.6 | 1.4 (0.8–2.7) | 5.1 | 0.4 (0.2–1.1) | 7.1 | 0.6 (0.2–1.8) | 0.3 | 0.02 (0.003–0.1) |
| Indian/Asian | 3.1 | 4.8 | 1.7 (1.1–2.7) | 3.0 | 0.8 (0.4–1.5) | 4.3 | 1.2 (0.6–2.3) | 1.8 | 0.4 (0.2–0.9) |
| Currently married | 48.4 | 57.0 | 1.5 (1.1–1.8) | 52.6 | 1.1 (0.9–1.4) | 56.9 | 1.3 (1.0–1.8) | 55.4 | 1.3 (0.9–1.8) |
| Rural | 38.4 | 36.0 | 0.9 (0.7–1.1) | 43.3 | 1.3 (1.0–1.7) | 40.2 | 1.1 (0.8–1.5) | 48.5 | 1.6 (1.1–2.1) |
| Employed | 30.0 | 33.7 | 1.2 (0.9–1.5) | 33.3 | 1.1 (0.9–1.4) | 34.4 | 1.2 (0.9–1.6) | 34.2 | 1.2 (0.9–1.5) |
| Highs school completed | 40.7 | 41.3 | 1.1 (0.8–1.4) | 33.8 | 0.7 (0.6–0.9) | 36.2 | 0.8 (0.6–1.2) | 25.3 | 0.5 (0.3–0.7) |
| Traumatic Life Events | |||||||||
| None | 39.2 | 24.3 | 1.0 | 20.4 | 1.0 | 18.6 | 1.0 | 21.4 | 1.0 |
| 1–2 | 39.4 | 39.6 | 1.6 (1.2–2.0) | 28.0 | 1.7 (1.3–2.4) | 24.4 | 1.7 (1.1–2.6) | 36.7 | 1.6 (1.0–2.3) |
| 3–4 | 14.9 | 23.8 | 2.4 (1.9–3.0) | 27.5 | 3.2 (2.2–4.7) | 32.4 | 3.9 (2.4–6.4) | 25.4 | 2.5 (1.6–4.0) |
| 5 + | 6.5 | 12.2 | 2.7 (2.0–3.6) | 14.2 | 3.6 (2.5–5.1) | 14.7 | 3.8 (2.2–6.5) | 16.5 | 3.7 (2.6–5.3) |
| No Disorder | 72.8 | 56.3 | 1.0 | 51.4 | 1.0 | 48.2 | 1.0 | 55.3 | 1.0 |
| Any Disorder | 27.2 | 43.7 | 1.9 (1.6–2.3) | 48.6 | 2.3 (1.8–2.9) | 51.8 | 2.5 (1.7–3.5) | 44.7 | 1.8 (1.3–2.4) |
After adjusting for the effects of other variables in the model, participants older than 50 years had an increased likelihood of consulting a traditional healer compared to the 18–29 year olds (OR=1.8, 95% CI 1.3–2.7) (Table 5). Black respondents were 9.1 times more likely (OR=9.1, 95% 4.4–19.3) (data not shown) to consult traditional healers than white, coloured and Asian respondents. Having completed high school, decreased the odds of consulting a traditional healer (OR=0.7, 95% CI 0.5–1.0). Being employed (OR=1.4, 95% CI 1.0–1.8), having a substance abuse disorder (OR=1.69, 95% CI 1.1–2.5) or an anxiety disorder (OR=1.7, 95% CI 1.2–2.4) were associated with consulting a traditional healer for emotional and mental health concerns.
TABLE 5.
RESULTS OF MULTIVARIATE LOGISTIC REGRESSION ANALYSIS OF FACTORS ASSOCIATED WITH CONSULTATION WITH TRADITIONAL HEALER, OVERALL AND BY THE PRESENCE OF DSM-IV-DEFINED MENTAL DISORDERS, IN A NATIONALLY REPRESENTATIVE SAMPLE OF SOUTH AFRICAN ADULTS
| Model 1 Total sample |
Model 2 DSM-IV Diagnosis |
Model 3 No DSM-IV Diagnosis |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Variables | OR | p | CI | OR | p | CI | OR | p | CI |
| Age 30–39 | 1.29 | 0.29 | 0.78–2.11 | 1.12 | 0.76 | 0.53–2.37 | 1.41 | 0.19 | 0.86–2.40 |
| Age 40–49 | 1.46 | 0.09 | 0.95–2.26 | 1.82 | *0.03 | 1.05–3.16 | 1.17 | 0.64 | 0.58–2.27 |
| Age 50 | 1.83 | 0.00 | 1.25–2.69 | 1.46 | 0.19 | 0.83–2.57 | 2.16 | *0.00 | 1.35–3.38 |
| Sex | 1.12 | 0.49 | 0.80–1.58 | 1.15 | 0.60 | 0.67–2.00 | 1.11 | 0.64 | 0.71–1.75 |
| Race | 0.11 | *0.00 | 0.05–0.22 | 0.07 | *0.00 | 0.02–0.26 | 0.13 | *0.00 | 0.06–0.30 |
| Religion (< once a month) | 0.95 | 0.85 | 0.58–1.56 | 1.12 | 0.75 | 0.56–2.23 | 0.72 | 0.31 | |
| Religion (> once a month | 0.83 | 0.25 | 0.60–1.14 | 0.73 | 0.29 | 0.41–1.31 | 0.86 | 0.60 | |
| Location | 1.17 | 0.33 | 0.85–1.64 | 1.31 | 0.27 | 0.81–2.45 | 1.01 | 0.93 | 0.66 |
| Employment | 1.36 | *0.03 | 1.02–1.80 | 1.52 | 0.08 | 0.94–2.46 | 1.25 | 0.19 | 0.92 |
| Education | 0.66 | *0.04 | 0.45–0.98 | 0.57 | 0.09 | 0.31–1.08 | 0.79 | 0.44 | 0.41 |
| DSM-IV: Any Anxiety | 1.71 | *0.00 | 1.19–2.44 | ||||||
| DSM-IV: Any Mood | 1.02 | 0.92 | 0.67–1.56 | 0.77 | 0.27 | 0.47–1.24 | |||
| DSM-IV: Any Substance | 1.69 | *0.01 | 1.13–2.53 | 1.17 | 0.59 | 0.66–2.06 | |||
statistically significant difference at p < 0.05
Health Seeking Behaviour Among Participants with DSM-IV Diagnosis
Similar findings were apparent in those with a lifetime DSM-IV diagnosis. Overall, 62% reported no treatment in the past 12 months (see table 1) and while 18% of respondents sought treatment from a Western practitioner exclusively, only 4% sought treatment from a traditional healer exclusively (Table 4). Patients with a DSM-IV disorder who consulted a traditional healer for their emotional and mental health concerns were more likely to be located in an urban setting (OR=2.0, 95% CI 1.3–3.2), be between 40–49 years old and had experienced more than three traumatic events in their lifetime. Patients with more than a grade 12 education (OR=0.42, 95% CI 0.22–0.80) and coloured, whites and Indian/Asians (OR = 0.06, 95% CI 0.01–0.4) were less likely to call upon the services of a traditional healer, while 97% of participants who consulted a traditional healer were black.
TABLE 4.
UNADJUSTED ASSOCIATIONS BETWEEN MENTAL HEALTH SEEKING BEHAVIOUR AND PARTICIPANT DEMOGRAPHIC CHARACTERISTICS IN THE PRESENCE OR ABSENCE OF A DSM-IV DEFINED MENTAL DISORDER, IN A NATIONALLY REPRESENTATIVE SAMPLE OF SOUTH AFRICAN ADULTS.
| No health Care |
Western Medicine |
OR (95% CI) | Alternative Medicine |
OR (95% CI) | Western & Alternative |
OR (95% CI) | Traditional Healer |
OR (95% CI) | |
|---|---|---|---|---|---|---|---|---|---|
| DSM-IV DIAGNOSIS | |||||||||
| Any anxiety disorder | 59.1 | 30.2 | 1.1 (0.8–1.5) | 19.9 | 1.3 (1.0–1.8) | 11.3 | 1.3 (0.9–2.0) | 8.9 | 1.4 (1.0–2.1) |
| Any mood disorder | 59.9 | 30.1 | 1.1 (0.8–1.5) | 21.3 | 1.1 (0.9–1.5) | 12.1 | 1.1 (0.8–1.6) | 8.0 | 0.8 (0.5–1.3) |
| Any substance disorder | 59.0 | 32.3 | 1.3 (1.0–1.8) | 21.7 | 1.2 (0.9–1.6) | 13.8 | 1.5 (1.1–2.1) | 9.7 | 1.2 (0.7–1.9) |
| Age 18–29 | 43.0 | 23.2 | 1.0 | 27.8 | 1.0 | 23.2 | 1.0 | 27.8 | 1.0 |
| 30–39 | 22.1 | 26.3 | 2.15 (1.47–3.14) | 23.4 | 1.40 (0.86–2.28) | 25.0 | 1.79 (0.93–3.45) | 21.0 | 1.20 (0.65–2.22) |
| 40–49 | 18.9 | 25.1 | 2.37 (1.55–3.62) | 27.0 | 1.95 (1.22–3.10) | 30.2 | 2.54 (1.38–4.66) | 29.3 | 1.97 (1.18–3.28) |
| 50+ | 16.1 | 26.6 | 2.74 (1.73–4.35) | 21.8 | 1.62 (1.10–2.38) | 21.6 | 1.87 (1.07–3.29) | 21.9 | 1.54 (0.94–2.55) |
| Sex: Female | 51.0 | 55.8 | 1.20 (0.88–1.63) | 55.2 | 1.14 (0.82–1.59) | 56.4 | 1.19 (0.86–1.64) | 53.0 | 1.02 (0.60–1.72) |
| Race: Black | 76.1 | 69.7 | 1.0 | 83.6 | 1.0 | 79.8 | 1.0 | 97.4 | 1.0 |
| Coloured | 12.0 | 12.6 | 1.19 (0.76–1.87) | 6.8 | 0.45 (0.32–0.63) | 6.6 | 0.49 (0.27–0.88) | 1.7 | 0.10 (0.02–0.60) |
| White | 9.0 | 13.1 | 1.75 (1.20–2.54) | 7.4 | 0.63 (0.31–1.29) | 10.4 | 1.01 (0.40–2.58) | 0.6 | 0.44 (0.01–0.30) |
| Indian/Asian | 2.8 | 4.7 | 2.06 (0.93–2.55) | 2.2 | 0.56 (0.18–1.61) | 3.2 | 0.94 (0.27–3.30) | 0.3 | 0.06 (0.01–0.38) |
| Currently married | 47.0 | 56.8 | 1.50 (1.12–2.00) | 50.6 | 1.05 (0.78–1.42) | 55.0 | 1.27 (0.90–1.79) | 51.3 | 1.07 (0.72–1.61) |
| Location: Rural | 34.6 | 36.9 | 1.1 (0.8–1.4) | 42.1 | 1.4 (0.9–2.1) | 41.6 | 1.3 (0.8–2.2) | 49.0 | 2.0 (1.3–3.2) |
| Employed | 32.2 | 35.1 | 1.1 (0.9–1.5) | 36.1 | 1.1 (0.8–1.5) | 33.8 | 1.0 (0.6–1.6) | 27.6 | 1.2 (0.7–1.9) |
| Graduated High School | 40.9 | 43.1 | 1.14 (0.84–1.56) | 34.8 | 0.73 (0.54–0.98) | 37.7 | 0.97 (0.53–1.41) | 23.7 | 0.42 (0.22–0.80) |
| Traumatic Life Events - None | 29.5 | 17.0 | 1.0 | 14.7 | 1.0 | 12.7 | 1.0 | 14.9 | 1.0 |
| 1–2 | 40.4 | 41.0 | 1.7 (1.1–2.5) | 39.0 | 1.8 (1.2–2.5) | 37.0 | 1.9 (1.1–3.1) | 36.3 | 1.5 (0.9–2.5) |
| 3–4 | 19.3 | 28.3 | 2.3 (1.6–3.4) | 31.3 | 2.8 91.6–5.0) | 36.5 | 3.6 (1.9–7.0) | 29.0 | 2.3 (1.2–4.2) |
| 5+ | 10.8 | 13.7 | 2.0 (1.2–3.2) | 15.0 | 2.5 (1.4–4.5) | 13.8 | 2.4 (1.2–5.0) | 19.8 | 3.0 (1.5–6.1) |
| NO DSM-IV DIAGNOSIS | |||||||||
| Age 18–29 | 41.9 | 34.0 | 1.0 | 32.8 | 1.0 | 26.7 | 1.0 | 29.8 | 1.0 |
| 30–39 | 21.6 | 19.7 | 1.1 (0.8–1.5) | 23.5 | 1.4 (1.0–2.0) | 22.5 | 1.6 (1.0–2.6) | 23.4 | 1.5 (0.9–2.5) |
| 40–49 | 16.5 | 19.4 | 1.5 (1.0–2.2) | 16.7 | 1.2 (0.8–1.9) | 20.1 | 1.9 (1.1–3.2) | 14.6 | 1.2 (0.6–2.4) |
| 50+ | 20.0 | 26.9 | 1.6 (1.1–2.4) | 27.0 | 1.6 (1.3–2.1) | 30.7 | 2.3 (1.4–3.6) | 32.2 | 2.1 (1.4–3.2) |
| Sex: Female | 52.8 | 55.7 | 1.1 (0.8–1.4) | 62.6 | 1.5 (1.0–2.1) | 54.7 | 1.0 (0.7–1.6) | 56.6 | 1.1 (0.7–1.7) |
| Race: Black | 76.0 | 76.1 | 1.0 | 89.7 | 1.0 | 87.2 | 1.0 | 95.8 | 1.0 |
| Coloured | 10.7 | 6.6 | 0.6 (0.3–1.3) | 3.6 | 0.3 (0.1–0.7) | 3.7 | 0.3 (0.1–0.9) | 1.2 | 0.1 (0.02–0.4) |
| White | 10.1 | 12.3 | 1.3 (0.6–3.0) | 2.9 | 0.2 (0.07–0.8) | 3.6 | 0.3 (0.1–1.2) | 0 | - |
| Indian/Asian | 3.2 | 5.0 | 1.6 (0.9–2.9) | 3.8 | 0.9 (0.5–1.8) | 5.5 | 1.4 (0.8–2.5) | 3.0 | 0.7 (0.3–1.6) |
| Currently married | 48.9 | 57.1 | 1.4 (1.0–1.9) | 54.4 | 1.2 (0.9–1.6) | 59.0 | 1.4 (0.9–2.3) | 58.8 | 1.4 (0.9–2.3) |
| Location: Rural | 39.8 | 35.3 | 0.8 (0.6–1.1) | 44.4 | 1.2 (0.9–1.7) | 38.6 | 1.0 (0.6–1.5) | 46.6 | 1.4 (0.9–2.0) |
| Employed | 29.2 | 32.6 | 1.2 (0.9–1.5) | 30.6 | 1.0 (0.8–1.4) | 35.1 | 1.3 (0.9–1.8) | 31.5 | 1.1 (0.8–1.5) |
| Graduated High School | 40.6 | 39.9 | 1.0 (0.7–1.4) | 32.8 | 0.7 (0.5–1.0) | 34.5 | 0.8 (0.5–1.3) | 26.6 | 0.5 (0.3–0.9) |
| Traumatic Life Events: None | 42.8 | 30.0 | 1.0 | 25.8 | 1.0 | 24.8 | 1.0 | 26.6 | 1.0 |
| 1–2 | 39.0 | 38.6 | 1.4 (1.0–1.8) | 37.0 | 1.5 (1.0–2.3) | 31.6 | 1.3 (0.7–2.4) | 37.1 | 1.4 (0.8–2.5) |
| 3–4 | 13.3 | 20.3 | 2.1 (1.5–2.8) | 23.8 | 2.8 (1.6–4.6) | 28.0 | 3.2 (1.7–6.1) | 22.5 | 2.4 (1.2–4.6) |
| 5+ | 4.9 | 11.1 | 2.9 (1.9–4.4) | 13.4 | 3.9 (2.2–7.0) | 15.6 | 4.4 (2.2–8.9) | 13.9 | 3.6 (1.7–7.7) |
Being between 40–49 years old was independently associated with an increased likelihood of consulted a traditional healer compared to the 18–29 year olds (OR=1.8 95% CI 1.1–3.2) (Table 5). Blacks were more likely to consult a traditional healer for emotional and mental health concerns (OR=0.06, 95% CI 0.02–0.3), than whites, Indians and Asians. Although not significant, a higher level of education (grade 12 +) was associated with a 43% decrease in consulting a traditional healer (OR=0.6, 95% CI 0.3–1.1) compared to less educated respondents, and being employed increased the likelihood of consulting a traditional healer for emotional and mental health concerns (OR=1.5, CI 0.9–2.5). Gender, marital status and religion were not associated with traditional healer consultation.
Health Seeking Behaviour Among Participants without a DSM-IV Diagnosis
Similar findings were apparent in those without a lifetime DSM-IV diagnosis. However, of the subjects who sought treatment from alternative medicine, 5% sought treatment from a traditional healer, 5% of the sample sought treatment from a spiritual or religious advisor, and 2% by another type of healer (e.g. chiropractor or spiritualist) (see Table 3).
Black respondents were 7.4 times more likely (OR 7.4, 95% CI 3.4–16.0) to seek help from a traditional healer as compared to the coloured, white or Asian /Indian participants (Table 4). The proportion of patients without a DSM-IV disorder who consulted a traditional healer for their emotional and mental health concerns are more likely to be above 50 years of age (OR= 2.1, 95% CI 1.4–3.2) and had experienced more than three traumatic events in their lifetime. Participants 9 were less likely to see traditional healers if they had more than a grade 12 education (OR=0.5, 95% CI 0.3–0.9), or were coloured or white when compared to the black participants (OR=0.1, 95% CI 0.02–0.4).
The independent effects of participant demographic characteristics on mental health seeking behaviour are presented in table 5. Further investigations of the variables shown in Table 4 were examined using a number of multivariate logistic models, to control for demographics and socioeconomic variables (including gender, age, religion race, location, employment and education (See table 4). Being above 50 years old increased the likelihood of consulting a traditional healer (OR=2.2, 95% CI 1.4–3. 5), compared to younger respondents (18–29 years old). Being black was associated with an 87% increase of consulting a traditional healer for emotional and mental health concerns (OR=0.1, 95% CI 0.1–0.3) compared to the coloured, white, or Asian/Indian participants. Gender, location, religion, employment and education were not associated with traditional healer consultation.
DISCUSSION
The main results of this study are that 1) A minority of participants with a lifetime DSM-IV disorder obtained treatment in the past 12-months from Western (29%) or alternative (20%) practitioners; 2) Traditional healers were consulted by 9% and religious or spiritual advisor by 11% of the respondents; and 3) Alternative practitioner and traditional healer use in the full sample were predicted by older age, black race, unemployment, lower education and having an anxiety or a substance use disorder.
The initial finding that South Africans consult Western health practioners more frequently than alternative practitioners, including traditional healers, for their mental health care needs is consistent with a number of previous studies assessing the pathways to care in sub-Saharan Africa. For example, a study on pathways to psychiatric care in Zimbabwe indicated that individuals with an acute illness would be more inclined to first consult a biomedical health professional, and then only seek traditional care if this treatment failed (Patel et al., 1997). However, the frequency with which individuals consult a traditional healer prior to presenting to mental health services has been found to vary significantly. A study in Nigeria reported that 26% of patients suffering from a mental illness visited a traditional healer prior to presenting themselves to a mental health services (Abiodun, 1995). Another study conducted in Ghana reported that only 6% of patients consulted a traditional healer prior to presenting themselves to mental health care services (Appiah- Poku et al., 2004)
Particularly striking is the comparatively low proportion that used traditional healers versus the higher proportion that used religious and spiritual advisors in the black population. Looking at the entire sample (irrespective of DSM-IV diagnosis), 6.4% reported consulting with a traditional healer exclusively (8.1% of Blacks) and 6.8% see a religious or spiritual advisor exclusively (7.2% of Blacks). In the total sample 11% of the respondents consulted with a religious or spiritual advisor and 9% consulted with a traditional healer. Thus religious and spiritual advisors may be consulted more frequently than traditional healers for mental health care concerns. These results are similar to those found in a study conducted in Ghana (Appiah-Poku, 2004) that reported 14% of patients had consulted with a pastor, compared to 6% consulting traditional healers before presentation to mental health services.
Little research has been conducted on attitudes of church pastors in urban or rural Africa, although the growth of their involvement has been noted (Pfeiffer, 2006). Although anecdotal, it is believed that the advent of the faith healer can be seen as an outgrowth of the influence of urbanization, acculturation, Christianity, and the African independent church movement. It has also been argued that many of the traditional roles of the isangoma have been assumed by the faith or spiritual healer (Edwards et al., 2003) Evidence for this change, is provided by a study conducted in Zimbabwe in 2002–2003 as part of a larger study of African Independent Church (AIC) and Pentecostal expansion. Members and recent converts of ten churches were interviewed in addition to in depth interviews with 80 pastors, prophet healers, traditional healers and other community leaders. In these interviews the common theme of payment and how it influenced the authenticity of the traditional healer’s practices emerged frequently (Pfeiffer, 2006). Therefore, spiritual advisors place themselves at the cusp between what they refer to as the backward and outdated traditional healers and the modern, scientifically based Western medicine (Freeman and Motsei, 1992).
Service use by participants in the total sample was predicted by older age, black race, lower education, employment, and having an anxiety or a substance use disorder. The findings that older participants were more likely to seek care from traditional healers, is consistent with previous research suggesting that younger participants in Ethiopia were more likely to seek health care from modern health services (Berhane et al., 2001). The finding that blacks are more likely to consult with traditional healers than whites, Asians or Indians is in keeping with the beliefs that many black South Africans may hold about traditional healers (Edwards et al., 1983). For example, in a retrospective study at a hospital in Durban, the first hundred files of psychiatric patients treated during 1980, who explained the etiology of their problems in traditional terms were analyzed. The sample constituted 29% of the total Black psychiatric population treated at the hospital over the time period. Of these patients, 81% had their traditional beliefs confirmed by a traditional healer (Edwards et al., 1983). Additionally, individuals with little or no formal education were more likely to consult traditional healers than those South Africans who were more educated. This may be because respondents with more education are educated in the Western biomedical model and are therefore more likely to pursue Western biomedical help.
We found the cost of treatment was relatively expensive and was similar between both Western (R318) and traditional healing systems (R321). Similarly, a study conducted in KwaZulu-Natal examined patterns of health seeking behaviour prior to the death of 1282 individuals. Information on the health care choices and expenditures of these individuals, who died between January 2003 and July 2004, were provided by the primary care giver. Fees for consulting a traditional healer ranged from 0- R4000, with an average of R433 (Case et al., 2005). Given the expense, it is not surprising that being employed was associated with traditional healer use. Furthermore, traditional healer service use was predicted by the presence of a mental illness, with those suffering from an anxiety or substance abuse disorder being more likely to seek traditional healer services. These findings are consistent with previous data from clinical samples indicating that the majority of patients and their families consulted with a wide variety of health professionals in an attempt to obtain effective treatment, irrespective of cost (Appiah-Poku et al., 2004; Ensink and Robertson, 1999).
These data suggest that living in a rural area was not a predictor of traditional healer use. While it has been argued that the belief in witchcraft is more common in the rural areas, traditional healing practices continue in urban settings. In many settings the notion of witchcraft may be fuelled by income inequality, competition and insecurity (Nattrass, 2005); if this is the case one would expect to find a similar portion of both rural and urban South Africans who hold beliefs in witchcraft (Ashforth, 2005) and consequently consult with traditional healers.
Furthermore, 5% of the respondents indicated that they received treatment from a traditional healer despite not having a lifetime DSM-IV diagnosis. This is consistent with a study conducted in the United States as part of the World Mental Health Survey. Results indicated that a majority (73.5%) of the respondents without a DSM-IV diagnosis sought treatment from the human services sector (30.7%) or the complementary and alternative medicine (CAM) sector (42.8%). The human service sector included religious or spiritual advisors, social workers and counsellors not specialized in mental health, while the CAM sector included any other type of healer (e.g. Chiropractor), internet support groups and self help groups. Only 8% of respondents with a low-need for treatment received treatment from the mental health or general medical sectors (Druss et al., 2007). Additionally, a study conducted in Nigeria as part of the World Mental Health Survey found low rates of use of complimentary medicine in those with a DSM-IV diagnosis (4%), but much higher rates were found in those seeking treatment without a DSM-IV diagnosis (57%) (Gureje and Lasebikan, 2006).
The results of the present study have implications for addressing the issue of limited psychiatric care in South Africa. Given the relatively low number of Western psychiatric practitioners, there may be value in working with traditional healers and spiritual advisors in this regard. Acknowledging the possible role of churches in providing care together with educating and working with church pastors may be an important way forward in improving mental health care if their practices have been shown to effective and safe (Appiah-Poku et al., 2004). Traditional healers and religious and spiritual advisors are widely dispersed throughout South Africa, are knowledgeable of the culture norms, and their advice is sought, believed and acted upon by community members. Organized and well trained traditional healers and religious and spiritual advisors have the potential to be play a significant role in mental health treatment and as a referral resource in the South African context (Peltzer et al., 2006).
Several limitations of this study must be considered when interpreting these findings. Firstly, the study relied on respondents’ recall of treatment of the past 12 months. Although the evidence is anecdotal, stigma about consulting traditional healers may have led to an underestimation of traditional healer use. Secondly, the study did not include South Africans who live in institutional settings including mental hospitals and therefore these findings may not hold for severely mentally ill patients. Thirdly, we did not have data on which service provider the subject saw first. More information on this would be helpful in understanding pathways to treatment. Fourthly, respondents may be reluctant to admit to consulting a traditional healer to a research assistant who is perceived as representing Western medicine. Finally, the categories of treatment providers created for the purpose of this study resulted in a perhaps over simplified approach that does not necessarily conform to the realities of these complex relationships.
Despite these limitations, this data is the first to describe the use of alternative medicine for mental health care in a nationally-representative population sample in South Africa. The study revealed that alternative medicine is widely used amongst South Africans, and therefore there is potential to educate both traditional healers and religious and spiritual advisors on common mental disorders. Furthermore, it is important to educate the South African population about the positive and negative consequences of alternative practices to ensure they are making fully informed decisions when choosing the appropriate health practitioners. Poor knowledge (low mental health literacy), negative attitudes, and fear of stigma are likely to influence the help-seeking behaviour of South Africans, and education has the potential to have a positive impact on the mental health situation in South Africa (Peltzer et al., 2006). Future research should focus more exclusively on the roles that religious and spiritual advisors and traditional healers play in the South African context, in addition to an assessment of the effectiveness of their mental health practices.
REFERENCES
- Abdool K, Ziqubu-Page T. Bridging the Gap: Potential for a health care partnership between African traditional healers and biomedical personnel in South Africa: South African Medical Association. 2004 www.samedical.org. [Google Scholar]
- Abiodun O. Pathways to mental health care in Nigeria. Psychiatric Services. 1995;46:823–826. doi: 10.1176/ps.46.8.823. [DOI] [PubMed] [Google Scholar]
- Appiah-Poka J, Laugharne R, Mensah E, Osei Y, Burns T. Previous help sought by patients presenting to mental health services in Kumasi, Ghana. Social Psychiatry and Psychiatric Epidemiology. 2004;39:208–211. doi: 10.1007/s00127-004-0725-9. [DOI] [PubMed] [Google Scholar]
- Ashforth A. Witchcraft, Violence and Democracy in South Africa. Chicago: Chicago University Press; 2005. [Google Scholar]
- Berhane Y, Gossaye Y, Emmelin M, Hogberg U. Women's health in rural setting in society transition in Ethiopia. Social Science and Medicine. 2001;53:1525–1539. doi: 10.1016/s0277-9536(00)00441-x. [DOI] [PubMed] [Google Scholar]
- Case A, Menendez A, Ardington C. Health Seeking Behaviour in Northern KwaZulu Natal. University of Cape Town: Centre for Social Science Research. 2005 [Google Scholar]
- Demyttenaere K, Bruffaerts R, Posada-Villa J, Gasquet I, Kovess V, Lepine JP, et al. Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. The Journal of the American Medical Association. 2004;291:2581–2590. doi: 10.1001/jama.291.21.2581. [DOI] [PubMed] [Google Scholar]
- Druss B, Wang P, Sampson N, Olfson M, Pincus H, Wells K, Kessler R. Understanding Mental Health Treatment in Persons without Mental Diagnoses: Results from the National Comorbidity Survey Replication. Archives of General Psychiatry. doi: 10.1001/archpsyc.64.10.1196. (in press) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Edwards SD. Traditional and Modern Medicine in South Africa: A Research Study. Social Science and Medicine. 1986;22(11):1273–>1276. doi: 10.1016/0277-9536(86)90194-2. [DOI] [PubMed] [Google Scholar]
- Edwards SP, Grobbelaar PW, Sibaya PT, Nene LM, Kunene ST, Magwaza AS. Traditional Zulu Theories of Illness in Psychiatric Patients. The Journal of Social Psychology. 1983;121:213–221. doi: 10.1080/00224545.1983.9924491. [DOI] [PubMed] [Google Scholar]
- Ensink K, Robertson B. Patient and Family Experiences of Psychiatric Services and Indigenous Healers. Transcultural Psychiatry. 1999;36(1):23–43. [Google Scholar]
- Freeman M, Motsei M. Planning health care in South Africa--is there a role for traditional healers? Social Science & Medicine. 1992;35(11):1183–1190. doi: 10.1016/0277-9536(92)90311-d. [DOI] [PubMed] [Google Scholar]
- Freeman M, Lee T, Vivian W. Parktown, South Africa: Department of Community Health, University of the Witwatersrand Medical School; 1994. Evaluation of mental health services in the Orange Free State. [Google Scholar]
- Gureje O, Lasebikan V. Use of mental health services in a developing country. Social Psychiatry and Psychiatric Epidemiology. 2006;41:44–49. doi: 10.1007/s00127-005-0001-7. [DOI] [PubMed] [Google Scholar]
- Lopez A, Mathers C, Ezzati M, Jamison D, Murray C. Washington: Oxford University Press and the World Bank; 2006. Global burden of disease and risk factors. [PubMed] [Google Scholar]
- Lund C, Flisher AJ. Staff/population ratios in South African public sector mental health services. South African Medical Journal. 2002;92(2):161–164. [PubMed] [Google Scholar]
- Mbanga I, Niehaus DJH, Mzamo NC. Attitudes towards and beliefs about schizophrenia in Xhosa families with affected pro-bands. Curationis. 2002;25:69–74. doi: 10.4102/curationis.v25i1.718. [DOI] [PubMed] [Google Scholar]
- Nattrass N. Who consults Sangomas in Khayelitsha? An Exploratory Quantitative Analysis. Social Dynamics. 2005;31(2):161–182. [Google Scholar]
- Ngoma M, Prince M. Common mental disorders among those attending primary health clinics and traditional healers in urban Tanzania. British Journal of Psychiatry. 2003;183:349–355. doi: 10.1192/bjp.183.4.349. [DOI] [PubMed] [Google Scholar]
- Ngubane H. Body and Mind in Zulu medicine. London: Academic Press; 1977. [Google Scholar]
- Patel V, Simunya E, Gwanzura F. The pathways to primary mental health care in high density suburbs in Harare, Zimbabwe. Social Psychiatry and Psychiatric Epidemiology. 1997;32:97–103. doi: 10.1007/BF00788927. [DOI] [PubMed] [Google Scholar]
- Peltzer K, Mngqundaniso N, Petros G. A Controlled Study of an HIV/AIDS/STI/TB Intervention with Traditional Healers in KwaZulu-Natal, South Africa. Aids Care. 2006;18(6):608–613. doi: 10.1080/09540120500294206. [DOI] [PubMed] [Google Scholar]
- Pfeiffer J. Money, Modernity, and Morality. In: Luedke T, West H, editors. Borders and Healers: Brokering Therapeutic Resources in Southeast Africa. Bloomington: Indiana University Press; 2006. pp. 81–100. [Google Scholar]
- Stein DJ, Seedat S, Herman A, Moomal H, Heeringa SG, Kessler RC, Williams DR. Lifetime prevalence of psychiatric disorders in South Africa. British Journal of Psychiatry. doi: 10.1192/bjp.bp.106.029280. (in press) [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang P, Aguilar-Gaxiola S, Alonso J, Angermeyer M, Borges G, Bromet E, Bruffaerts R, de Girolamo G, de Graaf R, Gureje O, Haro J, Karam E, Kessler R, Kovess V, Lane M, Lee S, Levinson D, Ono Y, Petukhova M, Posada-Villa J, Seedat S, Wells E. Use of mental health services for anxiety, mood, and substance disorders in 17 countries in the WHO world mental health surveys. Lancet. 2007;370:841–850. doi: 10.1016/S0140-6736(07)61414-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williams DR, Sonnega J, Seedat S, Stein DJ, Moomal H, Wilson C. The South African Stress and Health Study: Rationale and design. Metabolic Brain Disease. 2004;19(1–2):135–147. doi: 10.1023/b:mebr.0000027424.86587.74. [DOI] [PubMed] [Google Scholar]
- WHO. Geneva: World Health Organization; 2001. World health report. [Google Scholar]
- WHO. Geneva: World Health Organization; 2005. Mental health atlas. [Google Scholar]