• Epilepsy;
  • Perceived stigma;
  • Sub-Saharan Africa;
  • EMIC interview;
  • Goldberg Anxiety and Depression Scale


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

Purpose: Stigma is a major burden of epilepsy. In sub-Saharan Africa the few studies that addressed epilepsy stigma emphasize enacted, rather than perceived, stigma. This inattention may compromise clinical management and delay help seeking, thereby contributing to the treatment gap. We assessed perceived stigma and identified sociocultural and psychological factors explaining greater stigma among people with epilepsy (PWE) in Benin.

Methods: PWE included in this study were ascertained using a door-to-door survey in the general population in a Beninese rural area. We applied both qualitative and quantitative research methods to assess stigma and patient’s experience and beliefs. An Explanatory Model Interview Catalogue (EMIC) and verbally administered questionnaires provided data for demographic, clinical, and sociocultural features. Sociocultural features were evaluated in terms of illness-related experience and sociocultural representations of epilepsy. Depression and anxiety were also screened.

Results: Eighty PWE were included. About 68.7% reported feeling stigmatized. Multivariate regression revealed that factors independently associated with perceived stigma were experience of social isolation (p < 0.001), experience of marital problems (p < 0.01), and presence of anxiety disorder (p < 0.01).

Discussion: Perceived stigma is an important issue in epilepsy in Benin. Social factors seem to be more influential than sociocultural representation of epilepsy. Insofar as research is needed in other African countries to determine the nature and relevant features of stigma to improve treatment and control.

Epilepsy is a highly prevalent health problem in many developing countries (Preux & Druet-Cabanac, 2005; Birbeck et al., 2007), including Benin (Crepin et al., 2007). However, the proportion of people with epilepsy (PWE) who require treatment but are not receiving it can exceed 90% (Meyer & Birbeck, 2006). Uncontrolled seizures have serious consequences on morbidity and mortality rates. Therefore, PWE in sub-Saharan Africa (SSA) have high exposure to cognitive impairment, physical injuries, and disabilities (Jacoby et al., 2005). All these factors combined could exacerbate the stigma associated with epilepsy, which has long been described as a major burden to the disease.

The stigma is a sociologic concept, which includes occurred realities in all societies. Everywhere in the world, having a chronic disease is a handicap for individuals who are characterized as deviant (Goffman, 1963). What can differ from one society to another is the answer given to this deviance and the nature of stigma. Why such behavior is accepted in a society and is strongly opposed in another? Cultural beliefs and practices affect the management of epilepsy, especially in countries where epilepsy is still misunderstood (Jacoby et al., 2005; Weiss & Ramakrishna, 2006).

Because stigma contributes to the disease burden and influences illness behavior—both help-seeking and ill-advised termination of treatment (Weiss et al., 2006), it has become a major focus in public health.

Studies in the Southern hemisphere, specifically those in SSA, focus on enacted stigma, which refers to concrete experiences of discrimination against PWE solely based on identifying them with their condition (i.e., attitudes against PWE); contrary to the studies in the Northern hemisphere, which focus on perceived stigma (or felt stigma) defined by shame associated with having epilepsy and fear of anticipated discrimination (Reis & Meinardi, 2002). The two different approaches highlight the need for cross-cultural comparison to determine whether these differences result from orientation of researchers or from sociocultural differences, with implication for treatment adherence and seizure control.

This study was undertaken to assess perceived stigma in Benin inasmuch as the majority of studies emphasize enacted stigma. The second objective of this study was to identify sociocultural and psychological factors explaining greater stigma among PWE.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

Patients and procedures

A two-step survey was performed in 2005 in the rural district of Djidja Centre in Benin (area of 160 km2 with 11 villages) (Crepin et al., 2007). The first step consisted of a door-to-door survey using a validated 5-item questionnaire (95.1% of sensitivity and 65.6% of specificity) (Preux et al., 2000; Diagana et al., 2006) to screen potential PWE. This first step was performed by local general physicians trained by a neurologist in the detection of epilepsy. The second step consisted of interviewing people who had at least one positive response to the questionnaire. A diagnosis of epilepsy was verified by neurologist assessment. A total of 11,668 subjects (94% of the population) were screened for epilepsy. The neurologic evaluation confirmed 148 PWE (children, adolescents, and adults mixed), suggesting a prevalence of epilepsy of 12.7‰ [95% confidence interval (CI) 10.8–14.8‰] (Crepin et al., 2007). The definition of epilepsy was based on the occurrence of at least two seizures in more than 24 h, as recommended by the International League Against Epilepsy (ILAE) (Commission on Epidemiology and Prognosis of the ILAE, 1993).

Next we performed a survey in this area from February to May 2006, on the basis of these 148 PWE. Only adults aged 18 years or older were included (48 children and adolescents were excluded). All types of seizure were accepted, but subjects were not included if they had major cognitive impairments or mental deficiency that would impair their judgment or have a negative impact on their quality of life beyond the effects caused by epilepsy. Eighty PWE were included.

The Ministry of Health of Benin approved this work. The written consent form was provided and was read and discussed orally in the patient’s preferred language (Fon, the principal local language in Benin, or French). Signed consent forms could include a formal signature or a thumbprint based on responders’ preference. The principal investigator accompanied by a local interpreter questioned those who agreed to participate.


Demographic and clinical data were collected. Seizure-related variables included age at onset, seizure frequency, and history of seizure-related injuries in the previous year.

Perceived stigma

Perceived stigma was assessed using two validated instruments. The first was an Explanatory Model Interview Catalogue (EMIC), based on the framework of cultural epidemiology (Weiss et al., 1992). Initially developed in India, EMIC interviews were subsequently adapted for use in North America, Western Europe, Africa, and East Asia (Weiss, 1997). This instrument is relevant for cross-cultural studies because it is not a fixed scale but a catalog of explanatory model interviews based on an adaptable operational framework. It addresses illness experience (patterns of distress), perceived causes, general illness beliefs, preferences for help seeking, and lastly, the stigma caused by the illness. The EMIC was adapted to the context of Benin and allowed us to study sociocultural features of epilepsy including stigma.

To assess stigma from the semistructured EMIC interview, locally coherent indicators of stigma were studied. PWE were asked concerning several aspects of stigma, and each response was coded yielding a 4-point scale from 0–3 (higher values indicating a more definite indication of stigma). The internal consistency of all items was assessed to validate the index of stigma with the Cronbach alpha statistic, which guided the evaluation and the decision to retain or drop an item. This strategy was based on an approach previously developed for analysis of indicators of stigma (Vlassof et al., 2000; Weiss et al., 2001; Raguram et al., 2004).

The second instrument used for assessing perceived stigma was the Jacoby’s stigma scale (1994), a three-item scale. This scale has been used typically in Western countries to measure perceived stigma in epilepsy. Few studies were conducted in SSA with this scale, but Birbeck and colleagues have shown its relevance to African culture (Birbeck et al., 2007). To measure stigma from the Jacoby’s stigma scale, subjects were asked to select a response from a 4-point hierarchy ranging from 1–4 (Abetz et al., 2000). The overall score was the sum of each response and ranged from 3–12. A higher score was indicative of a high level of perceived stigma. This second measure allowed us both to compare the results using both instruments and to do cross-cultural comparisons.

Sociocultural features of epilepsy illness

Sociocultural features of epilepsy were evaluated in terms of illness-related experience and meaning of epilepsy, as formulated in the EMIC interview (Weiss et al., 1992). Illness-related experience was specified by patterns of distress (PD) and refers to the full range of problems associated with a patient’s condition, including clinical symptoms and social and economic disruptions. The meaning of epilepsy was specified by perceived causes (PC) and refers to patient’s ideas about the causes and the sources of their clinical problem.

Psychological factors: Depression and anxiety

Depression and anxiety were screened with the Goldberg Anxiety and Depression Scale (GADS), an 18-item scale (nine related to anxiety and nine related to depression). Each item required a dichotomic response (yes/no) coded to 1 or 0. The total score on each subscale was calculated, and a threshold score of 5 determined the probability of having anxiety disorder and a threshold score of 2 determined the probability of having a major depressive disorder (Goldberg et al., 1988). The GADS has a cross-cultural validity and seems to be a good instrument for epidemiologic studies in French-speaking populations, with a specificity of 91% and a sensitivity of 86% (Huber et al., 1999). This instrument has already been used in former studies in Benin and a neighboring country, Togo (Nubukpo et al., 2004a).

Statistical analysis

Data were processed using STATA Version 10 (Stata Corp., College Station, TX, U.S.A.) and SPSS (Statistical Package for the Social Sciences), version 14.0 (SPSS Inc., Chicago, IL, U.S.A.). Results for quantitative variables are expressed as mean ± SD (standard deviation) and results for qualitative variables as percentages. The index of stigma measured by the EMIC interview was the principal outcome. We used the Mann-Whitney test to compare the means and the Spearman correlation test to analyze the correlation between the two scores of stigma and the correlation between the index of stigma and the score of depression and anxiety.

We performed multivariate analyses using linear regressions for individual PD, PC, and psychological, adjusted for confounding with a backward stepwise procedure. All variables with a p ≤ 0.20 in the crude bivariate analysis were retained for the multivariate analysis. At each step, nonsignificant explanatory variables were removed and only variables with p ≤ 0.05 were retained in the model. In the end, all variables retained in each model were included for the final grouped PD, PC, and psychological multivariate logistic regression model.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

Sample characteristics

The study sample of 80 PWE consisted of 40 men and 40 women. The mean age was 35.2 ± 14.6 years (range 18–76 years) and the mean age at onset was 20.8 ± 15.9 years. Most of PWE lived with their families, but 35.0% had never been married (Table 1), with a proportion of men never married greater than women (55% vs. 15%; p < 0.001).

Table 1.   Demographic and clinical characteristics of people with epilepsy in Benin (n = 80)
CharacteristicsNumber of responders (%)
Demographic Marital status
 Never married28 (35.0%)
 Married/cohabit34 (42.5%)
 Separated/divorced13 (16.2%)
 Widowed5 (6.3%)
 Christian42 (52.5%)
 Voodoo29 (36.2%)
 Muslim3 (3.8%)
 Other6 (7.5%)
 Illiteracy65 (81.2%)
 Number of years of scholarship
 None65 (81.2%)
 < 5 years8 (10.0%)
 ≥ 5 years7 (8.8%)
Employment status
 No job14 (17.5%)
 Farmer54 (67.5%)
 Trade/business5 (6.2%)
 Student1 (1.2%)
 Other6 (7.5%)
Economic status
 Regular income25 (31.2%)
 No regular income40 (50.0%)
 No income15 (18.8%)
Seizure type
Generalized seizures
 Tonic–clonic seizures39 (48.8%)
 Absence seizures10 (12.5%)
Partial seizures
 Simple partial seizures9 (11.2%)
 Partial seizures evolving secondarily in21 (26.2%)
  generalized seizures
Undetermined epileptic seizures1 (1.2%)
Seizure frequency
 No seizure in the last year8 (10.0%)
 1 seizure per month17 (21.3%)
 2–4 seizures per month30 (37.5%)
 More than 5 seizures per month25 (31.2%)
Experiencing an aura43 (53.8%)
Physical stigmata of epilepsy present11 (13.7%)

Generalized seizures totalled 61.3%, partial seizures 37.4%, and 1 seizure (1.2%) was of undetermined seizure type. Of the 37.4% of patients with partial seizures, 26.2% had partial seizures that evolved secondary in generalized tonic–clonic seizure. Therefore, 75.0% of patients had experiences of tonic–clonic seizures (Table 1). A considerable proportion of respondents (45.0%) had first- or second-degree relatives with epilepsy. About 11.8% denied that they had epilepsy and referred instead to another disorder, such as hypertension for example or malaria (malaria was the fourth most-mentioned disease). Of those who have accepted the diagnosis, 73.7% said that at the beginning of their disease they had not identified their problem as epilepsy but as “a simple disease,”“a disease linked to the cool air,” or “God snake’s disease.” A total of 97.5% described their disease as a serious problem, and the majority had in the past consulted a traditional healer to solve their problem. The prescribed treatments were principally based on herbal tea and specific diet (forbidden food).

Almost all of respondents were not treated by a doctor and did not take antiepileptic drugs (AEDs). Therefore, all participants were subsequently treated by medical staff with AEDs: 75.0% were treated using phenobarbitone, 11.3% carbamazepine, and 13.8% diazepam.

Perceived stigma

Index of stigma according to the EMIC interview

Sixteen items were retained to assess locally indicators of stigma. The index of stigma was computed for each individual by adding the responses to each question. The mean score was 21.9 ± 11.8, with values for the sample ranging from 0–48. The lower quartile, median, and upper quartile values were, respectively, 9.7, 24.0, and 30.0. The internal consistency of this 16-item index of stigma was good, with a Cronbach alpha coefficient at 0.84 (Table 2).

Table 2.   Indicators of stigma assessed in EMIC interview among people with epilepsy in Benin (n = 80)
ItemsNumber of “yes” responses (%)
  1. aReverse code.

Keep others from knowing if possible59 (73.7%)
Disclosed to confidenta34 (42.5%)
Think less of yourself because of this problem36 (45.0%)
Feelings of shame or embarrassment because of your epilepsy64 (80.0%)
Others would think less of you because of your problem40 (50.0%)
Your disease would have bad effects on others11 (13.7%)
Others have avoided you because of your problem43 (53.7%)
Others might refuse to visit your home36 (45.0%)
Others would think less of your family because of your problem20 (25.0%)
Problem for your children in the community21 (26.2%)
Difficult for you to marry32 (40.0%)
Expectation of support of spousea23 (28.7%)
Problem would cause trouble in sexual life24 (30.0%)
Problem would cause trouble in marital life23 (28.7%)
Problem would make it difficult for relative to marry22 (27.5%)
Presumed health problems34 (42.5%)

Narrative accounts of various aspects of stigma, given by this EMIC interview, elaborate the nature of stigma and its impact. Patients emphasized concerns about disclosing the illness, difficulties in getting married, or problems in an ongoing marriage. Men seem more affected by their chances to get married than women. A young 30-year-old man explained:

“Because of this illness, I am not fulfilled. I have no home. I live with my uncle and I cannot woo a woman. Everybody make fun of me.”

Narratives also indicate that it was difficult for patients to clearly distinguish enacted and perceived stigma. When they were asked for perceived stigma during the EMIC interview they also reported concrete experiences (enacted stigma), as examples.

The Jacoby’s stigma scale

The mean score of perceived stigma with the Jacoby’s stigma scale was 6.9 ± 3.6, with values ranging from 3–12. The lower quartile, median, and upper quartile values were, respectively, 3.0, 6.0, and 11.7 (Table 3). The proportion of PWE who felt stigmatized by their condition was 68.7%, and 38.8% of them reported high feelings of stigma.

Table 3.   Indicators of stigma assessed by the Jacoby scale among people with epilepsy in Benin (n = 80)
Because of my epilepsyNumber of “yes” responses (%)
I feel some people are uncomfortable with me27 (33.7%)
I feel some people treat me like an inferior person27 (33.7%)
I feel some people would prefer to avoid me36 (45.0%)

The two scales, EMIC and Jacoby’s scale, were positively and strongly correlated (r = 0.74; p < 0.001), but strangely, the proportion of yes respondents were sometimes different when PWE were asked about the same domain. For example, with the Jacoby’s stigma scale 33.7% felt that some people have less regard for them because of their epilepsy (item 2) versus 50.0% with the EMIC interview (item 5).

Sociocultural features

Illness experience: Patient distress (PD)

A total of 63.7% of respondents complained of financial problems and 83.7% reported emotional distress because of their epilepsy (Table 4). The most troubling features of illness-related experience reported by PWE were worried about the course of their disease, constitutional weakness, and reduced social status.

Table 4.   Relationship between the mean score of stigma given by the EMIC and categories of distress reported by people with epilepsy in Benin (n = 80)
Pattern of distressNumber of “yes” responses (%)Mean score of stigmap-value
  1. aPatient distress (PD) variables retained for the multivariable analysis.

Physical distress during the seizure
 Convulsion69 (86.2%)21.9 ± 11.419.3 ± 14.80.6
 Absence11 (13.7%)33.5 ± 7.319.8 ± 11.20.0005a
 Loss of consciousness during the seizure68 (85.0%)21.9 ± 11.416.0 ± 13.40.2a
 Issuance of urine32 (40.0%)24.1 ± 10.720.0 ± 12.20.1
 Issuance of dribble47 (58.7%)22.4 ± 10.721.2 ± 13.30.6
 Tongue bite20 (25.0%)27.5 ± 8.819.7 ± 12.00.005a
 Abnormal movement3 (3.7%)32.0 ± 6.121.3 ± 11.7-
Physical distress caused by seizure
 Seizure-related injury51 (63.7%)22.5 ± 11.420.2 ± 12.40.3
 Concentration/memory or learning trouble45 (56.2%)22.9 ± 10.619.2 ± 13.20.2a
 Other physical problems11 (13.7%)
Social problems
 Social isolation29 (36.2%)28.0 ± 9.218.3 ± 11.80.0006a
 Stigma—reduced social status27 (33.7%)28.1 ± 11.418.3 ± 10.40.0002a
 Marital problems13 (16.2%)29.1 ± 8.420.2 ± 11.70.007a
Financial problems
 Ostracized from work41 (51.2%)25.0 ± 11.618.0 ± 11.20.009a
 Loss of job and wages27 (33.7%)26.2 ± 9.419.3 ± 12.10.02a
 Reduced personal or family income30 (37.5%)21.3 ± 12.821.8 ± 11.00.7
 Sadness, anxiety, or worry59 (73.7%)23.1 ± 11.918.7 ± 13.10.17a
 Concern about course of illness50 (62.5%)21.9 ± 11.022.2 ± 12.90.92
 Other social/emotional problems6 (7.5%)

The categories of distress (PD) best explaining self-perceived stigma after adjustment in a stepwise regression analysis included social isolation and marital problems (Table 6). The statistical model accounted for 23.4% of the variance of stigma.

Table 6.   Multivariate analysis of stigma among people with epilepsy in Benin (n = 80)
VariablesORp-value95% CI
  1. aCategories specifying Patient Distress (PD).

  2. CI, confidence interval; OR, odds ratio.

Marital problems (PD)a8.30.0082.3–14.4
Social isolation (PD)a9.30.0004.5–14.1
Perceived causes (PC)

Patients reported several possible causes of their epilepsy but the majority of them said that their problem had a supernatural cause (Table 5). The PC variables associated with stigma before adjustment in a stepwise regression were heredity, witchcraft, smoking, drug abuse, or alcohol. Two PC variables remained in the final model, including witchcraft and health problems (which include blood problems, prior illness, and continuous fatigue).

Table 5.   Relationship between the mean score of stigma given by the EMIC and perceived causes reported by people with epilepsy in Benin (n = 80)
Perceived causesNumber of “yes” responses (%)Mean score of stigmap-value
Reported causeNot reported cause
  1. aPerceived causes (PC) variables retained for the multivariable analysis.

 Food/water9 (11.2%)25.5 ± 12.321.4 ± 11.70.3
 Alcohol/smoking/drug abuse9 (11.2%)29.3 ± 10.220.9 ± 11.70.04a
Health problem
 Blood problem/prior illness15 (18.7%)31.5 ± 7.620.3 ± 11.60.006a
 Heredity16 (20.0%)28.0 ± 10.620.3 ± 11.60.01a
Traditional-cultural causes
 Evil eye/sorcery45 (56.25%)24.9 ± 11.218.0 ± 11.50.007a
 Fate21 (26.25%)21.7 ± 12.521.9 ± 11.60.9
 Sexual contact4 (5.0%)
 Mental–emotional stress4 (5.0%)29.7 ± 8.221.5 ± 11.8
Other perceived cause11 (13.7%)

Psychological morbidity

The GADS showed that one-fourth of PWE presented anxiety and 66.2% presented depressive disorder. The mean score was 2.9 ± 2.0 for the Goldberg anxiety subscale and 2.5 ± 2.0 for the Goldberg depression subscale. Higher scores of the EMIC’s index of stigma were significantly associated with the presence of anxiety (28.5 ± 9.1 vs. 19.7 ± 11.8; p ≤ 0.001) and depression (23.9 ± 11.3 vs. 17.9 ± 11.9; p ≤ 0.05). Anxiety was more associated with stigma than depression.

Relationship with clinical variables

Scores of stigma were not significantly correlated with the duration of epilepsy (r = 0.02; p = 0.90) and the age at onset (r = −0.15; p = 0.24). No relationship was found between the seizure type (p = 0.19) and the frequency of seizures (p = 0.97).

Potential factors influencing stigma

The multivariate regression analysis associating all categories of variables revealed that factors independently associated with feelings of stigma were experience of social isolation, experience of marital problems, and presence of anxiety disorder (Table 6). The statistical model accounted for 45.3% of the variance for stigma. No PC variables were retained in this model.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

This study measured perceived stigma and evaluated sociocultural and psychological factors contributing to higher levels of stigma in PWE in Benin. We have estimated the perceived stigma using two measures. The first was given by an instrument specially adapted to the studied country (EMIC interview); the second measure was given by a three-item scale specific to epilepsy-related stigma (the Jacoby’s stigma scale). The strategy we developed with the EMIC interview has been validated previously and the relevance of the second has already been shown in SSA (Weiss et al., 2001; Birbeck et al., 2007). Our results showed that 68.7% of PWE in this region felt stigmatized by their condition. This high prevalence of perceived stigma was close to findings of previous studies led in Western countries and using the Jacoby’s stigma scale. More particularly with a large European study, which suggested that more than half of PWE felt stigmatized by their epilepsy (Buck et al., 1999), with a higher percentage for people living in France (66%). Doughty and colleagues reported significant cross-cultural differences between percentages of European PWE who felt stigmatized, with values ranging from 27–62% (Doughty et al., 2003). In less-developed countries few studies were made using the Jacoby’s stigma scale, but results were in line of a lower prevalence of perceived stigma. Therefore, in Iran, results suggested a lower proportion of perceived stigma, that 25% of PWE reported feeling stigmatized by their condition (Baker et al., 2005). A Korean study has shown that 31% of individuals felt stigmatized by their epilepsy and only 9% of them regarded such stigma as serious (Lee et al., 2005). More recently, a cross-sectional Zambian study focused on epilepsy stigma versus other chronic diseases, revealed that PWE had higher mean perceived stigma scores than that of a comparison group (Birbeck et al., 2007). All these substantial differences in reported amounts of perceived stigma highlight at least the existence of high prevalence of perceived stigma in both Northern and in Southern hemisphere. That suggests, despite differences in clinical advances, that stigma remains everywhere an important contributor to illness burden. These results also show the importance of providing a cultural perspective in stigma research.

In our sample, 75.0% of PWE experienced tonic–clonic seizures and almost 70.0% had high frequency of seizures (at least two seizures per month). In addition, almost all respondents were not treated and did not take any AEDs at the time of the questionnaires’ administration. All these factors could have influenced the stigma and could in part explain the high prevalence observed in this population. Previous studies suggested that PWE were more likely to feel stigmatized if they had a combination of seizure types and frequent seizures (Buck et al., 1999; Doughty et al., 2003; Lee et al., 2005). The homogeneity of clinical symptomatology in our sample, however, limited our power to assess this factor. However, tonic–clonic generalized seizures are the most diagnosed seizure types in the Benin care system and the presence of high seizure frequency can in part be explained by the huge treatment gap described for a long time in SSA. Therefore, it seemed more appropriate to measure perceived stigma before initiating treatment, as it corresponds to the realities of the country.

Our results were in line with a high prevalence of perceived stigma in this population, but these data are directly linked to the Beninese context. It is true that in our study there was no control group and that we cannot compare the perceived stigma of PWE with another healthy person or a person with another disease. Our aim was not to do a comparative study but to measure stigma in PWE and to point out which factors might be important for epilepsy-related stigma.

The use of an EMIC interview acquires complementary qualitative and quantitative data concerning: stigma, illness-related experience, and the patient’s meanings. Results showed that the most frequently reported concern about stigma was strong feelings of shame (80.0%); that had not been asked by the Jacoby’s stigma scale. The proportion of men never married was higher than women, and patient’s narrative also suggested that epilepsy affects more men in terms of their chances to get married. However, no significant difference was found between gender and the perception of stigma. We found this point intriguing, and future studies should be performed in depth.

In this study, social factors (social isolation and marital problems) and anxiety appear to mainly influence stigma rather than clinical features or sociocultural representations of epilepsy. The social stress theory emphasizes the fact that the anticipation of negative treatment and the accompanying chronic stress results in a permanent state of vigilance (Goldstein et al., 2005). In this scenario, perceived stigma tends to intensify stressful circumstances and compromise a patient’s ability to cope with those circumstances (Goldstein et al., 2005; Meyer et al., 2008). Therefore, PWE approach interactions in society with anxiety. This argument can also be supported by the fact that 74.7% of patient themselves declared feeling sad, anxious, or worried about their situation, and 63.3% had concern about the course of their epilepsy (for 28.8% of them it was the most troubling concerns about their disease). Similar results had already been shown (Buck et al., 1999; Jacoby et al., 2005). The fact that anxiety appears closely linked to stigma can explain why the majority of studies have shown high anxiety scores among PWE compared with controls in various countries (Nubukpo et al., 2004b; Adewuya & Ola, 2005).

Another aspect of this study was that PWE did not regard stigma as extending to their family. Further study of this question of transferred stigma, however, is needed. Ideas about the importance of the meaning, rather than experience, of epilepsy as major determinants of stigma were called into question by the finding that no PC variables remained after controlling for confounding in the multivariate analysis.

This study reveals a high prevalence of stigma among PWE in Benin and also underscores the importance of considering personal attributes and patient’s illness experience. Our experience in the field also showed us that stigma can affect help-seeking behavior. Indeed, during the field work many PWE refused to come to the health care center. When we went to their home, the accounts given by these patients actually revealed the fear that their disease is put into the open. Others denied clearly that they have epilepsy. As previously suggested by Weiss and Ramakrishna, stigma tends to delay appropriate help-seeking or terminate treatment, contributing to the deficit of treatment (Weiss & Ramakrishna, 2006). And this occurs especially in countries where the health systems are weaker and where PWE are not adequately followed up.

Stigma has become an increasingly important topic for research in public health (Weiss, 2008). Despite the numerous studies concerning epilepsy-associated stigma, results are heterogeneous and factors that contribute to stigma are not yet well established. Some authors even suggested that perceived stigma is less prevalent than enacted stigma in SSA (Reis & Meinardi, 2002; Jacoby et al., 2005). Narratives indicate that it is difficult for patients to clearly distinguish enacted stigma and perceived stigma. When they were asked for perceived stigma during the EMIC interview they also reported concrete experiences (enacted stigma), as examples. Even if it seems to be difficult to present comparisons of enacted and perceived stigma in this study, our findings emphasize the relevance of previously neglected perceived stigma in an African context. If enacted stigma is considered as the greater concern, perceived stigma will never head to the forefront, making it difficult to control seizures in these areas.

The stigma evaluation, factors associated with stigma as well as its consequences, will allow programs to be established to improve patients’ quality of life. Our research fit perfectly with the strategy of the Global Campaign Against Epilepsy, improving the knowledge of epilepsy and reducing the stigma toward PWE. This is the first study to our knowledge combining both qualitative and quantitative approaches for the analysis of stigma associated with epilepsy. This approach could be extended to other countries, for cross-cultural comparisons and, more particularly, in less-developed countries, where very few studies have been done and where epilepsy remains a neglected disease.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

We thank the PWE for their cooperation and the local team for their help, most particularly Gislhaine Ghilto and Fabien Adjovi for their leading role in this project. We are grateful to our colleagues for their support in writing this article, particularly, Bertrand Courtioux, Marie-Claude Lise, and Zhou Zhou. The study was financially supported by the CORUS Benin project no. 02 341 138 (French Foreign Ministry) and by the Limoges Doctoral School “Sciences Technologie et Santé.” PN, DH, and PMP participated in the conceptualization and design of the research. FR, PN, PO, and MGW participated in instrument development. FR, PN, DH, and PMP directed enrollment of study participants. FR, PN, DST, CMD, PO, JPC, MGW, and PMP provided input into the analytical plan, and reviewed and edited the manuscript. All authors read and approved the final version.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References

We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. We certify that none of the authors had any conflict of interest to disclose.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Disclosure
  8. References