This cross sectional study examined the prevalence and risk factors of anxiety and depression among children and adolescents with epilepsy aged of 7 to 19 years attending in neurology department of Yalgado Ouedraogo University Teaching Hospital.
Anxiety and related risk factors
The prevalence of anxiety in children with epilepsy (42%) was similar to those observed by Falcone (40%)  but higher than that observed in litterature witch is ranging from 17% in the study of Russ (17%)  to 30 - 35% in the study Reilly . The high prevalence of anxiety in our context could be explained by the fear of failing to have a seizure in public. Forty two percent of the children had mild to severe anxiety, as seen in the study of Baki et al in Algeria , but higher than in the study of Williams et al in USA (23%) . , supporting by sociocultural factors. The prevalence of anxiety among adolescents (10-19 years) and young children was 52.4% and 33.3% respectively. Accoding to our study, there was no relation between the age of child and the presence of anxiety, as reported in other studies[6.17]. Only Caplan et al in USA  found a significant link between age of child and presence of anxiety. The prevalence of anxiety among boys and girls was quite similar (57.8% and 52%). Indeed, our study did'nt found a link between gender and the presence of anxiety. In contrast, BİLGİÇ in Turkey had reported that boys had significantly higher scores of trait anxiety in comparison to girls.The prevalence of anxiety was respectively 80% in children with secondary level, 45.4% in children with primary level and 0% in children who were dropped out of school. This finding could be explained by the fact that school-aged children are able to understand the disease. In oue study, educational status was associated to anxiety showing that children with secondary level were more anxious than children with primary level and those who were dropped out of school. This type of result was found by Yang in China. Our study had found that prevalence of anxiety was similar in children with generalized (46.7%) and focal epilepsy (40%). However, our study did not found association between type of epilepsy and anxiety, as seen in the study of Baki et al in Algeria. . The prevalence of anxiety was similar in children with long duration of epilepsy (42.9%) and those with short duration of epilepsy (40%). However, our study had found association between epilepsy duration and anxiety. Children with short duration of their epilepsy have more anxiety than those with long duration of their epilepsy, in contrast with the study of Oguz et al in Turkey. . This result could be explained by the fact that children were not experencied with the seizures. The prevalence of anxiety in children with free-seizure during the last 3 months was 51.6%, higher than for children with uncontrolled seizures. Children are afraid that the seizure will wake up at any time. Our study did not found association between anxiety and the control of seizures, in contrast with the study of Oguz et al. [ who reported significant link between anxiety and non-seizure control. Children treated by dual AED had more anxiety than those who were treated by AED monotherapy (34.8% vs. 18.2%) without difference type of treatment and anxiety, as seen in the study of Etting et al (USA)  . However, polytherapy (being on more than one AED) has been associated with increased symptoms of anxiety [6.18]. Some studies have been reports suggesting the role of AEDs in causing or sometimes exaggerating anxiety[23.24]. Besides, withdrawal from long-term dependence on AEDs could also result in the development of anxiety as well as increase seizures in patients
Prevalence of depression and associated risk factors
Twenty six percent of children and adolescents with epilepsy had depression symptoms according to Child depression Inventory. This frequency is consistent with several clinical studies on depression in children with epilepsy, ranging from 12.7 to 36.5%[16.26]. For those using Child depression Inventory, the prevalence of depression was ranging from 12 to 32%[27.28]. However, our results were higher than those observed by Yang in China (16.9%) and Russ in USA (17%) . The differences between studies could be explained by the characteristics of the population studies (size, inclusion criterias, depression scales and sociocultural characteristics). Twenty six percent of depressed children had mild depression (100%), higher than in the study of Adewuya et al in Nigeria (28.43%). Sociocultural factors could explained theses differences. The prevalence of depression in adolescents (11-19 years) was higher than those of youg children (7-10 years): 42.8% vs 12.5%). This result was similar to those of Oguz et al. in Turkey who reported more depression in the 12- to 18-year-old patients compared to children aged of 9-10 years by using Children Depression Inventory scale. According to several authors, depression score was more frequent in old children (≥11 years)[6.19.30]. While our study did not find a link between depression and children's age, several authors reported that depression was higher in adolescents and older groups[6.19.30]. According to gender, the prevalence of depression in girls was higher than to boys (27.8% vs 25.9%), in line with the study of Caplan in USA. This situation could be explained by psychosocial and behavioral problems are more likely in epileptic girls than boys, especially in adolescents. In our context, adolescent girls were exposed to excision, painful menstruations and forced marriages. There was no association between gender and depression, in line with several pediatric studies. In a study, Bilgiç et al in Turkey had found more depression score in boys than in boys. Although depression was most frequent in children residing in towns than in rural areas (30.3% vs 16.7%), without differences between the two regions. This result suggest that children residing in towns may have additional anxiety coming from parent's stress. The prevalence of depression was significantly similar in children with short epilepsy duration (<5 years) than those with long duration of epilepsy (≥5 years): 28.6% vs 20%. Only two studies had reported that long duration of epilepsy is likely correlated factor for developing depression in children[6.33]. This result could be explain by the fact of an inaugiral post diagnosis depression (hope of healing). Regarding the seizure types, children with focal and generalized epilepsy had equal prevalence of depression level (26.7%). As some studies, we did not found link between depression and seizure types[6.16.17]. In contrast, several authors reported that focal epilepsy is more significantly associated with depression than generalized epilepsy[184.108.40.206]. Children with controlled seizures during the last 3 months had more depression than those who had uncontrolled seizures (35.5% versus 7.1%), in line with the study of Oguz et al. .However, our study did not found significant relationship between seizures control and depression level. The prevalence of depression in children treated by polytherapy was elevated than those of chiltren treated by monotherapy (34.8% vs 18.2%) without difference between the two groups. Several authors had reported that depression was significantly higher in patients receiving more than one AED, compared to monotherpay[6.29.33]. The possible justification might be due to serious adverse effect, cost of drug and drug interaction. Indeed, Plioplys stated that the depressogenic effect of AEDs cannot explain development of depression in children perse; suggesting this effect to be related to the side effects of individual drugs. In addition, Brent et al demonstrated a much higher prevalence of major depression in epileptic children treated with phenobarbital comparatively to carbamazepine. In our study, this medication was used alone or in combination in a significant proportion of children with epilepsy (22.2%).
Strenght and limitations of the study
This cross-sectional study is the kind one of studies on anxiety and depression in children with epilepsy in Burkina Faso. Howerevr, he had several limitations, according to the small size of the population study, the nature of the study (cross-sectional) and the nature of the questionnaire who is hetero-administered. Our study did not determine parental factors associated with children depression (familly history of mental diseases, familly history of epilepsy, number of children in the familly, socioeconomic characteristics of the familly, ect… ..).