Afin d’évaluer la prévalence de la dépression et de l’anxiété parmi les épileptiques et de les comparer à une population témoin, une enquête cas-témoin transversale utilisant l’échelle d’anxiété et de dépression de Goldberg a été conduite au Bénin (Afrique de l’Ouest) auprès de 98 épileptiques âgés de 18 ans et plus, ainsi que 98 témoins appariés selon le sexe, l’âge ( 10 ans) et le milieu de vie. L’âge moyen est de 32,6 ans 11,5 et le sex-ratio 1,28. 93 % des personnes interrogées vivent en famille, sont mariées ou vivent maritalement (témoins : 98,2 % ; cas : 87,9 %) ; 57,4 % sont mariés (témoins : 70 % ; cas : 44 %) ; 97 % des épileptiques affirment avoir eu une crise dans les deux années précédant l’enquête ; 48 % en ont eu 2 à 5 et 41,5 %, plus de 10. Il n’y a pas de corrélation significative entre la fréquence des crises et le niveau d’anxiété et de dépression, en revanche la présence d’un traitement antiépileptique (77,5 %) diminue l’anxiété et la dépression. Les épileptiques présentent plus souvent une anxiété sévère (79,8 %) ou une dépression sévère (89,6 %) que les témoins (12,3 % et 46,9 %) (p ≪ 0,0001). La comparaison des scores moyens pour l’anxiété et la dépression obtenus à l’échelle de Goldberg confirme la différence (p ≪ 0,0001) entre cas (5,8 2,0 et 2,3 1,9) et témoins (4,7 2,4 et 2,0 2,1). Ni le sexe, ni l’âge, ni le milieu de vie (urbain/rural), n’ont d’influence significative sur l’anxiété et la dépression. La forte proportion de troubles anxieux et dépressifs chez les épileptiques souligne l’importance de la dimension psychologique de la maladie et la nécessité d’une prise en charge spécifique.

In order to assess prevalence of depression and anxiety among epileptic patients and to compare it to a control population, a matched case-control survey was performed in 196 persons above 18 year old (98 epileptics and 98 controls matched according to sex, age 10 and social environment) in Republic of Benin (West Africa), using Goldberg’s Depression and Anxiety scale. Two main investigators helped by 5 sociology students were trained on a questionnaire by a psychiatrist skilled with public health matters. People taking part in the survey are epileptic patients who already used health services. Inclusions took place within 17 communes of four departments (Mono, Zou, Ouémé, Atlantique) located in Southern part of Benin. The questionnaire used an Identity sheet and the Goldberg Depression Scale. Results are shown as mean standard deviations, for quantitative values, and percents for qualitative ones. Comparisons of proportions in qualitative variables are carried out using c² test or Fisher’s exact test. Comparisons of means rates between subject’s groups are carried out with a Student t test or variance analysis. The correlations between two quantitative variables were assessed by linear correlation coefficient. Significance threshold chosen for the whole set of statistics analysis is 0.05. The majority of interviewed epileptic patients is young (average 32.6 11.5 years old). A male predominance exists (sex ratio 1.28). 93 % of interviewed persons live within their family, are married or cohabit (controls : 98.2 % ; cases : 87.9 %) ; 57.4 % are married (controls : 70 % ; cases : 44 %). The most represented professional categories are craftsmen and shopkeepers (29.2 %) as well as farmers (19.5 %). Most of recruited patients live in an urban setting (55.4 %) and 63.6 % of interviewed persons had been living in the area of survey for over 10 years. The most represented religion within the sample is Christian religion (67.7 %), Animists (23.3 %) and Muslims (5.8 %). 97 % of epileptic patients reported they had one fit during the two years before the survey ; roughly one half (48 %) had 2 to 5 fits and 41.5 % of them had more than 10 fits ; only 14 % say they have had an EEG. Presence of an Anti Epileptic Drug (77.5 %) reduces anxiety and depression. Considering a severity threshold of 5 for anxiety and 2 for depression (8), proportions of epileptic patients displaying a severe anxiety (79.8 %) or a severe depression (89.6 %) are significantly higher (p ≪ 0.0001) than in control subjects (12.3 % and 46.9 %). Comparison of average scores confirms the difference (p ≪ 0.0001) between cases (5.8 2.0 and 2.3 1.9) and controls (4.7 2.4 and 2.0 2.1) regarding anxiety and depression. Neither the sex, nor age, nor life environment (urban/rural), nor frequency of fits hold significant influence over anxiety and depression. However, results in this survey include higher average rates of anxiety and depression for women (6.3 1.8) than men (5.5 1.8), though such statement is only nearly significant (p = 0.06). Results of the survey confirm the other works on this topic about characteristics of depression for epileptic patients, though results here are higher than usual. Anxiety and depression are common troubles found in epileptic patient, both often occurring at the same time. Two distinct theories about this fact are opposed, first explain the connection of anxiety and depression with epilepsy because of the social and cultural burden upon an epileptic patient in those countries, the second theory is about depression and epilepsy sharing some neuroaminergical dysfunctions ; these facts were not considered in this survey. When taking into account the thresholds of seriousness on Goldberg’s scale <sup>[8GOLDBERG D, BRIDGES K, DUNCAN-JONES P et al. Detecting anxiety and depression in general medical settings. Br Med J 1988 ; 297 : 897-9.

Cliquez ici pour aller à la section Références]</sup>, proportion of controls displaying a probably deeper depression is high (46.9 %) compared to usual hospital prevalence rates (4 to 25 %) found in Western Africa where survey in general population are scarce <sup>[2

Cliquez ici pour aller à la section Références, 5

Cliquez ici pour aller à la section Références et

Cliquez ici pour aller à la section Références]</sup>. It could either be a bias in the selection or the confirmation that family and relatives of an epileptic patient are enduring a great psychic pain too ; finally, this high rate of severe depression within controls may suggest a need to adapt thresholds of Goldberg’s scale to match African populations. This hypothesis is interesting regarding the results of some authors who validated this scale in population of aged people ; they highlighted the fewer precision of two items (lost of interest, focus difficulties) and also a slight discriminatory weight for other few items <sup>[12HUBER P, MULIGAN R, MACKINNON A et al. Detecting anxiety and depression in hospitalised elderly patients using a brief inventory. Eur Psychiatry 1999 ; 14 :11-6.

Cliquez ici pour aller à la section Références]</sup>. However the psychometric characteristics of Goldberg’s scale are accurate (for depression subscale, a sensitivity of 88 % and a specificity of 68 % ; for the whole scale, a sensitivity of 84 % and a specificity of 68 %) <sup>[16MACKINNON A, CHRISTENSEN H, JORM AF et al. A latent trait analysis of an inventory designed to detect symptoms of anxiety and depression using an elderly community sample. Psychol Med 1994 ; 24 : 977-86.

Cliquez ici pour aller à la section Références]</sup>. It would be pertinent to carry out a study to validate Goldberg’s scale within some African populations. This study also highlights the importance of psychological factors in epilepsy and suggests a specific global care of the disease.