seizures had elevated scores on the Child Behavior Checklist (CBCL), suggesting a risk for behavioral problems [11]. Austin et al. found that 32% of the children with new-onset seizures had baseline CBCL scores in the clinical or at-riskrange [12].

Though most studies of behavior in children with epilepsy have been cross-sectional, several groups have assessed change over time. Mitchell et al. followed children for 18 to 30 months. They found that the severity of seizures and the prevalence of behavioral problems remained relatively stable over time [13]. We have followed 212 children with new-onset seizures over 24 months and have found a trend for behavior to improve in those children with no further seizures and to remain unchanged in those with recurrence of seizures [14]. Lendt et al., in a study of children assessed before and after seizure surgery, showed that behavior disturbance dropped from 39% of the sample at baseline to 25% at 3 months after operation [15].

2. Specific behavioral problems in children with epilepsy

Seizures in children have been associated with specific behavioral problems, with disruptive behaviors and mood disorder found most commonly. There seems to be an association between epilepsy and autistic disorder, with seizures occurring in one-third of children with autism. Tuchman and Rapin have suggested an association between seizures and regression in autistic children [16]. Psychosis in childhood epilepsy is rare, but Caplan et al. have found an increase in illogical thought processing in children with partial complex seizures [17]. Sleep disturbances appear to be common in children with epilepsy and have been associated with behavioral problems [18,19].

Of the disruptive disorders, the association between attention-deficit/hyperactivity disorder (ADHD) and epilepsy has the best support from research data. In two large population-based studies, McDermott et al. reported hyperactivity in 28% of children with epilepsy compared with 13% of children with heart disease and 5% of controls, and Carlton-Ford et al. noted impulsivity in 39% of children with a history of epilepsy versus 11% of controls [7,20]. Semrud-Clikeman and Wical, using both structured interviews and a computerized continuous performance test, found that 33% of children with partial complex seizures had ADHD as defined by DSM-III-R criteria [21]. Using either the Child or Adolescent Symptom Inventory, we noted symptoms of ADHD, combined type, in 11% and ADHD, predominantly inattentive type, in 24% of a sample of 175 children with chronic seizures [22].

Though there are anecdotal reports of oppositional behavior, aggression, and delinquency in children with epilepsy, there does not appear to be a significant in-

crease in riskof oppositional defiant disorder (ODD) or conduct disorder in children with uncomplicated epilepsy when compared with children in the general population. Caplan et al. found ODD in 3 and conduct disorder in 1 of 24 children with generalized seizures and ODD in 3 and conduct disorder in 2 of 30 children with complex partial seizures [17]. Schoenfeld et al. reported no significant increase in scores on the CBCL externalizing scale in children with complex partial seizures compared with siblings [23]. Using the Child or Adolescent Symptom Inventory, we noted symptoms of ODD in 12 and conduct disorder in 12 of 96 children with epilepsy [24].

Just as found in adults with epilepsy, there does seem to be an association between depression and epilepsy in children and adolescents. Ettinger et al. reported elevated scores on the Child Depression Inventory in 26% of a sample of children 7–18 years of age with epilepsy [25]. Dunn et al. found that 23% of a sample of adolescents with epilepsy had symptoms of depression [26]. Alwash et al., in a study of adolescents with seizures, found depression in 33% compared with 16% of controls [27]. Oguz et al. noted more symptoms of depression in adolescents 12–18 years of age with epilepsy when compared with children 9–11 years of age with seizures or healthy controls [28]. Children with complex partial seizures, in a study by Schoenfeld et al., had elevated scores on the internalizing section of the CBCL, suggesting a riskfor depression or anxiety [23].

3. Risk factors for behavioral problems associated with epilepsy

Multiple factors are associated with increased riskof behavioral problems in children with epilepsy. Additional neurological impairment, neuropsychological deficits, intractable seizures, and problems within the family have been found most consistently. Studies have shown that children with epilepsy plus additional neurological impairment have approximately twice the rate of behavioral problems as children with uncomplicated epilepsy [4]. Steffenburg et al. found that 59% of children with mental deficiency and epilepsy had psychiatric impairment [29], and Hoare noted that epilepsy plus additional disabilities had a substantial negative impact on quality of life [30]. An association between frequent poorly controlled seizures and behavioral problems has been seen repeatedly [see 23,28,31,32]. Hoare and Kerley [33], Austin et al. [32], and Mitchell et al. [34] have shown that stresses within the family are significantly associated with increased riskof behavioral problems in the child with epilepsy.

Other factors are noted less consistently, which may be partially due to sample limitations. Many of the studies of behavioral problems in children with epilepsy

have a limited range of age and often exclude children with a mental handicap. Gender has had a very inconsistent association, with Stores finding boys had more trouble [35], Hoare and Kerley noting no difference [33], and Austin et al. reporting more trouble in girls [32]. Age at onset of seizures has been associated with cognitive problems but not with behavioral trouble. Seizure type or syndrome was not related to behavior problems in the studies of Hermann et al. [31], Austin et al. [32], Schoenfeld et al. [23], or Lendt et al. [15]. Behavioral problems may be one of the complications of infantile spasms and Lennox–Gastaut syndrome but would not be seen in studies in which mental handicap is an exclusionary criterion. There are anecdotal reports of behavioral disturbances due to antiepileptic drugs, but most large series have found minimal if any effects [36–38].

4. Assessment

Since behavioral disturbance occurs in both new-onset seizures and epilepsy, monitoring for emotional difficulties should begin as soon as the diagnosis is made and should be a part of ongoing care for the child with epilepsy. Several instruments are available for assessment. The Child Behavior Checklist (CBCL) and the Child and Adolescent Symptom Inventories are general behavioral screening questionnaires that have been used to assess children with chronic health problems [39,40]. Qualityof- life scales specific for children with epilepsy cover physical, psychological, social, and academic functioning. Examples include the Quality of Life in Epilepsy for Adolescents (QOLIE-AD-48), the Quality of Life in Childhood Epilepsy (QOLCE), the Impact of Pediatric Epilepsy Scale (IPES), and the Impact of Childhood Illness Scale [41–44]. These scales vary in length from 11 to 73 items and, with the exception of the QOLIE-AD-48, have been used for both children and adolescents.

5. Treatment of behavioral problems in children with epilepsy

Treatment of behavioral problems should begin with education. Lewis et al. have described a group approach for educating children and parents about seizures and their management. After four sessions the children who participated in the sessions were more knowledgeable, felt more competent and better behaved, and needed fewer restrictions than children in the control group. Parents showed increased understanding of epilepsy and less anxiety [45]. Austin et al. devised the Parent Report of Psychosocial Care Scale and the Child Report of Psychosocial Care to assess concerns and fears and also questioned parents on their knowledge of epilepsy [46].

Austin et al. then provided an intervention designed to address these individual worries, concerns, and information needs. Families were contacted by telephone on two occasions and were given the opportunity to participate in a group telephone conference with families having similar problems. At the time of follow-up, both parents and children were more knowledgeable and had fewer information and support needs. Children had fewer concerns and were happier with family relationships [47].

Once behavioral problems become established, individual, group, or family therapies may be needed. There are limited data in this area since few intervention trials have been conducted. Nevertheless, guidelines are available. Ziegler et al. suggest ongoing psychosocial assessment and treatment coordinated with neurological care. Specific issues that might be addressed are the initial reactions of grief, anxiety, or anger related to the diagnosis of epilepsy, decreased competence and autonomy in the preschool child, attentional and learning problems in the school-age child, and self-esteem and social issues in the adolescent [48]. We found that adolescents with epilepsy and symptoms of depression had negative attitudes toward their illness, poor communication within the family, and an external or unknown locus of control, all issues that could be a focus in individual psychotherapy [26]. In a review of the current literature on epilepsy and the family, Ellis et al. identified negative attitudes, overprotectiveness, and hostility as factors in the familial response to epilepsy that were associated with behavioral problems in children and could be amenable to family therapy [49]. Individual and group therapies have been used to assess for emotionally based seizure precipitants and then to teach self-relaxation techniques or self-control [50,51]. The outcomes in these studies have included both improved self-esteem and reduced seizure frequency.

Psychopharmacology may be another option for therapy of behavioral problems in children with seizures. The choice of psychotropic agent will depend on likely effectiveness in treatment of the specific behavioral problem, potential interactions between antiepileptic drugs (AEDs) and the psychotropic agent, and the presumed effect of the psychotropic drug on the seizure threshold. Psychopharmacological management should also include an assessment of potential behavioral side effects of the antiepileptic drugs employed in treating the child.

The stimulants methylphenidate and dextroamphetamine are first-line agents for therapy of ADHD and may have some utility in treatment of conduct disorders. In children with complex partial seizures and ADHD, Semrud-Clikeman and Wical have shown improvement on continuous performance tasks after methylphenidate [21]. Neither Feldman et al. [52] nor Gross-Tsur et al. [53] found deterioration in seizure control in children

given methylphenidate. Stimulants have caused an elevation in blood levels of phenobarbital and phenytoin.

The serotonin reuptake inhibitors seem to be first-line agents for the treatment of depression and anxiety in children with epilepsy, though no controlled trials are available in this population. Kanner et al. used sertraline in adults with seizures and found reduction in seizure control in 6 of 100 patients, 5 of whom improved after adjustment of antiepileptic drug doses [54]. Blumer recommends tricyclic antidepressants for adults with depression and epilepsy, though there may be a marginal lowering of seizure threshold [55]. SSRIs may cause inhibition of the cytochrome P450 enzyme system, resulting in an increase in phenytoin, carbamazepine, and valproate levels. A lowering of the seizure threshold has been seen with higher-dose bupropion, clomipramine, and maprotiline [56].

The antipsychotic agents have been used for children with psychosis, pervasive developmental disorders, and tics. Chlorpromazine and clozapine are best avoided because of their tendency to lower seizure threshold. Haloperidol has little effect on seizure threshold. Other than clozapine, the newer atypical antipsychotics have not been reported to significantly affect seizure threshold though very little information is yet available. The enzyme- inducing antiepileptic agents carbamazepine, phenobarbital, and phenytoin may cause a decrease in serum concentration of most of the antipsychotic medications.

6. Case discussion

There are multiple factors in the present case of Peter¹ that indicate a riskfor behavioral problems. Peter has intractable seizures occurring three or four times per day. Though he has no obvious additional neurological deficit, his psychoeducational and neuropsychological evaluations have shown learning disability in math and written expression, language impairment, delay in fine motor skills, and problems with attention. His mother is described as overprotective and his father hostile. Peter has a negative attitude toward his illness and poor selfesteem. In addition, there is a history of depression in Peter
s mother, placing Peter at riskfor both mood disorder and externalizing symptoms [57,58]. Other factors that may be increasing his riskof behavioral problems are early age of onset and sleep disruption from recurrent seizures.

Peter has been given three DSM-IV diagnoses [59]. He has six of nine symptoms of a major depressive episode. His problems with fatigue, impaired concentration, and trouble sleeping could be due to seizures or AEDs. However, his irritability, withdrawal, and thoughts of

¹ See the article by Shafer and Dean [61] in this issue.

death are more consistent with a mood disorder. The prominence of his seizure disorder suggests using a diagnosis of mood disorder due to epilepsy instead of major depressive disorder. Peter also has eight of nine symptoms of inattention and three of nine symptoms of hyperactivity–impulsivity. There are sufficient criteria and evidence of significant impairment to allow a diagnosis of ADHD, predominantly inattentive type. He displays five or six of the eight symptoms of oppositional defiant disorder. If these symptoms occur only when Peter has evidence of a mood disorder, a diagnosis of oppositional defiant disorder would not be made.

Psychiatric interventions to be tried with Peter should include psychotherapy and medication, preferably given in an integrated fashion. Both Peter and his family should receive continuing education on epilepsy and its treatment. Peter should benefit from individual psychotherapy that helps enhance his sense of control over his environment and especially his epilepsy. A cognitive–behavioral approach could reduce negative thinking and cognitive distortions with resultant lessening of depression. The presence of right hemisphere weakness on neuropsychological testing is often associated with difficulty understanding social cues, suggesting Peter may benefit from a social skills group of similar-age children. Both Peter and his family could participate in an assessment to define stresses that may be increasing seizure frequency and in counseling to learn relaxation or self-control techniques to manage stress. The parents should be encouraged to take a more neutral response to seizures and to develop more adaptive responses to Peter, including lessening the current overprotectiveness and hostility.

Pharmacological intervention should begin with a review of AEDs. Neither the felbamate nor valproate that Peter currently receives causes significant depression or disruptive behavior and thus can be continued. Paroxetine has been effective in treating depression in adolescents [60] and does not significantly lower the seizure threshold. It does inhibit cytochrome P450 enzyme systems and thus valproate blood levels should be monitored. Paroxetine has not been helpful in improving attention and the addition of a stimulant may improve academic functioning.

Because of the intractability of Peter's seizures, he will need to receive comprehensive care that attends to seizure control, academic progress, behavioral problems, and social concerns. The team treating Peter will need to remain realistic without being unreasonably pessimistic.

References