Parents' Perceptions of Adversity Introduced by Upheaval and Uncertainty at the Onset of Childhood Epilepsy

Authors


  • Revision accepted August 30, 2001.

Address correspondence and reprint requests to Dr. K. J. Oostrom at Department of Child Neurology, Division of Neuropsychology (KG01.327.1), University Medical Centre, Wilhelmina Children's Hospital, PO Box 85090, 3508 AB Utrecht, The Netherlands. E-mail: K.Oostrom@wkz.azu.nl

Abstract

Summary:  Purpose: We report the parent's perceptions of and reactions to the onset of “epilepsy only” and the implications for continuity of parenting.

Methods: Content analysis was used to extract data on perceived (dis)continuity of parenting, from interviews held with parents of 69 schoolchildren in whom idiopathic or cryptogenic epilepsy (“epilepsy only”) had recently been diagnosed.

Results: Almost half of the parents (42%) perceived neither themselves nor their child as having been thrown off balance by the onset of epilepsy. Quite a few parents (33%) perceived themselves rather than their child as having been thrown off balance. More parents of children with cryptogenic than with idiopathic epilepsy perceived themselves as being off balance. However, parents' perceptions of their children's reactions to the epilepsy-related changes were not influenced by any epilepsy variable. Rather, family trouble, long-standing behavioral problems, and adolescence contributed to the child's maladaptive reaction. Three extremely off-balance parents had children with seizures of “unclassifiable” epilepsy that later were found to be nonepileptic and psychogenic.

Conclusions: The majority of the parents perceived their child as adapting well to the onset of epilepsy only. Children with seizure onset in adolescence and children with other adversities were perceived as adapting poorly to the additional adversity of epilepsy. Cryptogenic rather than idiopathic etiology leaves parents in great suspense. In behavioral studies, it is advisable to treat children with unclassifiable epilepsy as a separate group.

Children with epilepsy have a higher incidence of behavioral problems than have healthy children, but also greater than those of children with other chronic disorders. Even in schoolchildren with epilepsy of an idiopathic or cryptogenic etiology (1), who have no other chronic illness, who go to mainstream schools, live at home, and visit the neurologist, pediatrician, or family physician only on an outpatient basis, both externalizing and internalizing behavior problems are commonly reported (2–12). We refer to this subset of children with epilepsy as children with “epilepsy only”(13). Severity, cause, and course of the behavioral predicament of children with epilepsy only is as yet poorly understood. Behavioral problems are often attributed to biologic factors of the disease, including seizure type and seizure frequency (4,12). Side effects of daily antiepileptic drugs (AEDs) have been associated with negative behavioral changes (2,5). In any case, the illness variables explain only a small proportion of the variance in behavioral outcome measures, which suggests that other mechanisms underlie the behavioral predicament. It has been proposed that social factors, such as interaction between parent and child and family stress, play a role in behavioral problems of children with epilepsy (14,15), but their impact is not yet clear. Austin et al. (3) acknowledged the importance of family relationships in behavioral problems in children with benign epilepsy. Unfortunately they did not specifically report the etiology of their epilepsy sample. None of the studies in epilepsy only has simultaneously assessed the contribution of epilepsy-related and social variables.

Recently Galletti et al. (16) focused on the individual experience of having a seizure and on the feelings with respect to epilepsy that children and adolescents with epilepsy only experience, and found it was the perceived rather than the objective severity of the epilepsy that mattered. The authors also suggested that the child's perception of her or his seizures depends to a high degree on the perceptions developed by the parents. Therefore, they advise further studies to inventory the parents' experiences and emotions.

Our exploratory study sought to gain insight into parents' perceptions of their own and their children's experienced adversity and adaptation during the period of upheaval and uncertainty introduced at the onset of epilepsy. What allows some parents to maintain parenting, whereas others facing similar illness conditions are not able to do so? Parenting of children with epilepsy only has not yet been described. An approach in terms of adaptation enables one to search in several contexts for factors that influence parenting. For instance, progress at school, intellectual and social level, and family trouble, and a child's signs and symptoms of what turns out to be epilepsy can each add to or upset adaptive parenting. The child's reaction to changes in relation to the onset of epilepsy also might influence parenting. The balance between what adds to and what upsets adaptive parenting in the presence of adversity currently receives much scientific interest in that part of developmental psychology that tries to understand the mechanisms of emotional or behavioral vulnerability (17–19).

We decided to extend the work of Galletti et al. (16). Semistructured interviews were carried out to assess parents' perceptions of changes in their own parenting habits related to the onset of epilepsy and their perceptions of their child's reaction to the onset of epilepsy. In this way, we expected to gain knowledge about the adversity in parenting when facing diagnostic uncertainty and sometimes upheaval caused by signs and symptoms, hospital visits, and examinations before diagnosis. The developmental view of behavioral pathology, particularly the concept of developing proneness (20–23), guided us to trace the impact of the epilepsy as early as possible. Currently it is reasoned that behavioral pathology develops from intrinsically normal affective reactions. Recurring affects induce a proneness to react to these affects. If proneness has developed, it distorts cognition and biases action and perception. A reason for assessing perceptions of parents “here and now” rather than retrospectively was to avoid recollection bias.

The present study is, to the best of our knowledge, the first qualitative study to inventory parents' perceptions of experienced adversity in relation to the onset of epilepsy only. It relates epilepsy variables, prediagnostic child characteristics, and family trouble to continuity of parenting.

METHODS

This study was part of a multicenter neuropsychological project of the Dutch Study of Epilepsy in Childhood (DuSECh). All children participated on the basis of written consent from their parents, and, if 12 years or older, dependent on their own agreement. In children without an obvious idiopathic etiology, it was made clear to the parents and the child that neuroimaging magnetic resonance imaging (MRI) was needed to rule out a minor chance of an underlying structural cause for the child's epilepsy. Only after having communicated the normal result of the MRI and the probably favorable prognosis of the epilepsy, the doctor introduced the present study and gave the written information and forms requesting informed consent. After having received informed consent, we interviewed the parents of the children with epilepsy within 48 h after diagnosis. The Ethics and Research Committees of the 10 participating hospitals had approved the study.

Subjects

Parents of 69 children were interviewed (Table 1). Of the mothers, 27 (39%) had completed lower administrative or technical training (level I), 27 (39%) had completed higher administrative and technical training (level II), and 15 (22%) had completed university or equivalent education (level III). Twenty-four (35%) fathers had completed level I, 19 (28%) level II, and 23 (33%) level III; the educational level of three (4%) fathers remained unclear. Between January 1997 and October 1998, the 69 children were consecutively included immediately after the diagnosis of epilepsy. Inclusion criteria were two or more unprovoked seizures; an idiopathic or cryptogenic etiology according to the epidemiologic standards set forth by the International League Against Epilepsy (ILAE; 1989) (24); age between 5 and 16 years; and mainstream education. Exclusion criteria were any associated neurologic disorder (by history, physical examination, or neuroimaging) or other chronic illness (e.g., diabetes mellitus or asthma), and previous use of AEDs. Children with cryptogenic epilepsy had an epilepsy syndrome that was probably of symptomatic origin but for which no etiology has been found by history taking, neurologic examination, and neuroimaging (1,24).

Table 1.  Age and intellectual levels, grouped according to ILAE epilepsy syndromes classificationa
PatientsNo.Age in years
[mean (SD)]
IQb
[mean (SD) range]
  • a

     Epilepsy classification according the Commission on Classification and Terminology of the International League Against Epilepsy (29).

  • b

     <11 years: Colored Progressive Matrices (25); ≥11 years: Standard Progressive Matrices (26).

  • c

     Benign childhood epilepsy with centrotemporal spikes (1.1.a ILAE): generalized tonic–clonic seizures (n = 11), partial seizures (n = 4), unclassifiable seizures (n = 1).

  • d

     Childhood or juvenile absence epilepsy (2.1.e and 2.1.d ILAE): absence seizures.

  • e

     Juvenile myoclonic epilepsy (n = 3) (2.1.f ILAE) and other generalized idiopathic epilepsies (n = 1) (2.1.h ILAE): generalized tonic–clonic seizures (n = 2), unclassifiable seizures (n = 2).

  • f

     Localization-related cryptogenic epilepsy (1.3.a ILAE), with a temporal focus (temporal lobe epilepsy): partial (n = 4), generalized
    tonic–clonic (n = 3), unclassifiable seizures (n = 1).

  • g

     Localization-related cryptogenic epilepsy (1.3.a ILAE), with a frontal focus (frontal lobe epilepsy): partial (n = 3), generalized seizures (n = 7).

  • h

     Localization-related cryptogenic epilepsy (1.3.a ILAE), not further defined: partial (n = 9) and generalized seizures (n = 6).

BCECTSc168.7 (1.8)99 (12.5) 84–132
Absence epilepsyd138.2 (2.2)98 (13.7) 76–115
Other generalizede48.5 (3.0)100 (27.7) 69–136
TLEf89.2 (3.7)115 (18.4) 84–145
FLEg1010.0 (2.9)96 (16.9) 69–124
Cryptogenich159.3 (3.5)90 (15.8) 55–111
Unclassifiable39.7 (0.8)71 (7.4) 66–76

The median number of days between the first recognized unprovoked seizure and the interview was 18.5 days for primary or secondarily generalized seizures (range, 2–293 days); for absences, 28.0 days (4–73 days); for partial seizures, 29.0 days (2–259 days); and for unclassifiable seizures, 29.0 days (11–47 days). Seven patients had a family history of epilepsy in first-degree relatives. Three boys and three girls were not Dutch by birth. Intelligence quotients (IQs) were derived from the computerized version of the Colored Progressive Matrices (25) in children younger than 11 years and the Standard Progressive Matrices (26) in children 11 years and older (27,28).

After follow-up, it was decided that the three patients with unclassifiable epilepsy were included incorrectly. All three had nonepileptic, psychogenic seizures, as concluded after repeated EEG investigations, some with simultaneous video-registration of behavior. As their correct diagnosis became apparent after inclusion, they remained in the study according to the “intention to treat.”

Data collection

The semi-structured interview focused on the parent's perceptions:

  • 1Long-standing behavioral problems in the child: difficult behavior, perceived by the parent as still present but dating from before signs and symptoms of epilepsy were first recognized;
  • 2long-standing learning problems in the child: disappointing school results, perceived by the parents as already present before signs and symptoms of epilepsy were first recognized;
  • 3family trouble, such as marital distress, divorce, psychopathology in another member of the family;
  • 4parenting in the context of adversity caused by changes in the child that were related to the onset of epilepsy. The interviews had to yield information on whether parents managed to continue their habitual parenting in the period of uncertainty and upheaval due to the onset and diagnosis of epilepsy; and
  • 5reaction of the child to changes in relation to the onset of epilepsy. The interviews had to yield information on whether the child managed to adapt well despite exposure to epilepsy-induced uncertainty either in herself or himself, in her or his environment, or in both.

A list of questions was prepared for each of the variables. The interviews started with the question “When did you first notice that there was something the matter with your child?” Parents were given ample opportunity to talk about their experiences of seizures, medical consultations, hospital visits, and being informed about the diagnosis. The interviewer raised the preformatted questions addressing the major predetermined topics or subtopics if parents themselves did not come up with or elaborate on the issues.

Content analysis

The interviews were transcribed and processed according to Chi (30). The unit of analysis was the phrase. Every phrase containing information that pertained to one of the interview variables was tagged in accordance with the tenor of the utterance. A few examples may clarify the procedure: In response to the question, “What was it like to find out about the illness of X?” an interviewee answered, “I am so worried. He is our only son attending secondary modern school. What will happen if he has to quit school? What if he wants to marry? What girl is going to sleep next to someone who is at risk of seizures?” The italicized fragments of the phrases were tagged as both “worries about the future” and “fear of stigma.” Another interviewee said that she “thought X was dying,” which was tagged as “experiencing child's death.”

A grid was composed consisting of a row for every tag and a column for every interviewee. Three researchers (A.S., K.J.O., and A.J.S.) analyzed 10 interviews and distinguished 75 tags. Then two researchers (A.S. and K.J.O.) independently analyzed the remaining 59 interviews. Four tags were added by mutual agreement, resulting in a list of 79 tags covering all parents' utterances pertaining to the five interview variables. The scarce discrepancies (n = 7, 1%) between researchers' content evaluations were dealt with by reviewing the transcribed interview until consensus was obtained. This process resulted in a 79 × 69 grid, with entries for relevant tags.

The tags were grouped into the five interview variables (Table 2):

Table 2.  Results of content analysis, tabulated separately for children with idiopathic, cryptogenic, as well for children with unclassifiable etiology
 Idiopathic
etiologya
Cryptogenic
etiologya
Unclassifiable
epilepsya,b
Total
  • a

     Epilepsy classification according to ILAE (1989).

  • b

     Children with unclassifiable epilepsy were not included in statistical analyses (31).

Independent variables    
 1. Behavioral problems    
  Present713323
  Not present2620046
 2. Learning problems    
  Present64111
  Not present2729258
 3. Family trouble    
  Present913224
  Not present2420145
Dependent variables    
 4. Parenting    
  Continuous2113034
  Off balance1220335
 5. Child's reaction    
  Adapting2723252
  Off balance610117
  • 1Long-standing behavioral problems and
  • 2long-standing learning problems: cases were allocated to the categories “present” or “not present,” based on tags reflecting the presence of long-standing problems.
  • 3Family trouble was similarly categorized as “present” or “not present.”
  • 4Parents' perceptions of changes in their parenting habits, in relation to the onset of epilepsy: tags representing parenting behaviors that reflected continuity of habitual parenting (i.e., the same as before the diagnosis of epilepsy) were defined as “continuous”; tags representing disturbed parenting behaviors were defined as “discontinuous.” If the majority of entries referred to tags reflecting discontinuous parenting, the case was allocated to the “off balance” category. Conversely, if the majority of entries referred to tags reflecting continuous parenting, the case was allocated to the “continuity in parenting” category.
  • 5Parent's perceptions of the child's reaction to change in relation to the onset of epilepsy: tags representing a child's behavior reflecting adaptation to epilepsy-related changes were defined as “positive”; tags representing a child's maladaptation to the changes in relation to the epilepsy onset were defined as “negative.” If the majority of entries referred to tags reflecting maladaptive reactions, the case was allocated to the “off balance” category. Conversely, if the majority of entries referred to tags reflecting adaptive reactions, the case was allocated to the “well adapting” category.

Statistical analyses

Analyses were carried out using the S+ statistical package (S-PLUS 2000 Professional Release 3, 2000; Mathsoft, Inc., Cambridge, MA, U.S.A.), following the paths depicted in Fig. 1. Boxes represent groups of independent variables (A, demography and IQ; B, epilepsy; C, prior adversity) and dependent variables (D, continuity in parenting and child's reaction). The numbered paths indicate the order of analysis of the relations. All relations between and within boxes were systematically examined.

Figure 1.

Diagram depicting paths of statistical analysis.

Linear mixed effects analyses of variance (ANOVAs) were performed to analyze relations between categoric data (epilepsy features, dependent and independent interview variables, and the educational level of the mother or the father) and continuous data (age, IQ, days between first recognized unprovoked seizure and interview). Relations between categoric data were analyzed by using χ2 test for independence of factors. Correlations between continuous data (age, IQ, and days between first recognized unprovoked seizure and interview) were calculated by Spearman's correlation. 0.05 were deemed to be statistically significant. If a p value was between 0.05 and 0.10, a tendency to significance was presumed and will be reported. Statistically not significant relationships (p>0.10) are not reported (data can be obtained from the first author). Within the framework of this exploratory study, we did not correct for multiple comparisons.

In all tests, p values of ≤0.05 were deemed to be statistically significant. If a p value was between 0.05 and 0.10, a tendency to significance was presumed and will be reported. Statistically not significant relationships (p > 0.10) are not reported (data can be obtained from the first author). Within the framework of this exploratory study, we did not correct for multiple comparisons.

RESULTS

The analyses as presented were carried out after removal of the three children with unclassifiable epilepsy, who at follow-up appeared to have nonepileptic, psychogenic seizures, and who appeared to have a considerable impact on several relationships. We deemed it unjustifiable to ignore cases that had been included on well-considered grounds and decided to report on these three children at the end of this section.

Some of the most salient perceptions of the epilepsy and the reactions elicited by it, as disclosed by parents in the interviews, are listed (Table 3). Of 32 (48%) parents who perceived themselves as not being able to continue habitual parenting, 22 perceived their child as adapting well, and 10 perceived their child as maladapting to the changes related to the onset of epilepsy. Of 34 (52%) parents who perceived themselves as continuing habitual parenting, 28 perceived also their child as adapting well, and six perceived their child as maladapting.

Table 3.  Most salient parental perceptions of their own and their child's reaction to the onset of epilepsy, data on long-standing behavioral and learning problems and on family trouble (disregarding interviews held with parents of children with unclassifiable etiology)
Categoriesn (%)Tags:n
Parents thrown off balance32 (48%)Emotionally overwhelmed by the signs, symptoms, and diagnosis27
  Excessive fear of recurrence of seizures20
  Undifferentiated fear for epilepsy14
  Experiencing child's death when witnessing a seizure10
  Clearly distorted expectations regarding the child's future (including marriage, driving license)8
  Excessive fear of adverse AED effects8
  Excessive fear of stigmatization7
  Steering clear of any information concerning epilepsy5
  Extreme distrust with regard to neurologist's expertise3
  Overshoot in downplaying signs and symptoms4
Continuity of parenting34 (52%)Comforting/confident parenting after the onset of epilepsy17
  Diagnosis, epilepsy-related information, parenting, and environment in equilibrium27
Children thrown off balance16 (24%)Excessive feelings of shame and withdrawal, interfering with going to school or other social activities12
  Excessive fear of recurrence of seizures8
  Excessive fear of adverse effects and refusal of AED2
  Death wish2
Children positively adapted50 (76%)Benefit from being ill3
Long-standing behavioral problems23 (35%)Outsider at school14
  Extremely pushy7
  Noticeably sad6
  Noticeably hyperactive and dominant6
  History of mental health services6
Long-standing learning problems10 (15%)Disappointing school results with no clear cause10
  Due to lack of concentration, alertness/speed8
  Due to school absenteeism attributed to epilepsy3
Family trouble22 (33%)One-parent family12
  Substantial difficulties threatening happy family circumstances10
  New partner moved in7
  • 1Parent's perception of the child's reaction to changes in relation to the onset of epilepsy was statistically significantly related to age [F(1, 67) 16.69, p ≤ 0.01]. On average, children who were perceived as having been thrown off balance by the onset of epilepsy (11.5 ± 2.79 years) were significantly older than children who were perceived to adapt well to the onset of epilepsy (8.5 ± 2.27 years).
  • 2Self-perceived parenting was statistically significantly related to etiology (χ2, 3.88; df, 4; p ≤ 0.05). Among the parents of children with cryptogenic epilepsy, 20 (61%) of 33 were off-balance, as compared with 12 (36%) of 33 parents of children with idiopathic epilepsy. It should be noted that the seven children with a first-degree relative with epilepsy were perceived as adapting well to the onset of epilepsy, whereas not all parents with a family history of epilepsy perceived themselves as adapting well.
  • 3Parents' perceptions of the child's reaction to the onset of epilepsy tended to a statistically significant relationship (χ2, 3.58; df, 4; p ≤ 0.10) with long-standing behavioral problems of the child. Among the 20 children with long-standing behavioral problems, parents of seven (35%) perceived their child to be thrown off balance by the onset of epilepsy. Yet of 46 children who were not perceived as having long-standing behavioral problems, nine (20%) were perceived as thrown off balance. The child's reaction to the onset of epilepsy was statistically significantly related to family trouble (χ2, 3.88; df, 4; p ≤ 0.05). Of the 22 children belonging to families in trouble, parents perceived eight (36%) children as adapting poorly to the onset of epilepsy, compared with only eight (18%) of 44 children who belonged to happy families.
  • 4Age was statistically significantly related to seizure type [F(3, 65) 3.40, p ≤ 0.05]. Children with tonic–clonic seizures were statistically significantly older than both children with absence seizures (p ≤ 0.05) and children with partial seizures (p ≤ 0.00). Only the child's IQ was statistically significantly related to ILAE epilepsy syndrome [F(7, 59) 2.60, p ≤ 0.05]. On average, children with temporal lobe epilepsy (TLE) had a statistically significantly higher IQ than children with benign childhood epilepsy with centrotemporal spikes (BCECTS), absence epilepsy, and frontal lobe epilepsy (FLE) (p values ≤0.05), and than children with otherwise cryptogenic epilepsy (p ≤ 0.001).
  • 5Long-standing learning problems tended to be statistically significantly related to ILAE epilepsy syndrome (χ2, 9.83; df, 10; p ≤ 0.10). Five of the 13 children with absence epilepsy, as compared with none of four with other generalized idiopathic epilepsies, one of 16 with BCECTS, four of 33 with cryptogenic epilepsy (one of 10 FLE and none of eight TLE) had long-standing learning problems. The relation between long-standing learning problems and main seizure type tended also to statistical significance (χ2, 7.14; df, 8; p ≤ 0.10): five of the 13 children with absence seizures, as compared with three of 29 children with tonic–clonic seizures and two of 20 children with partial seizures had long-standing learning problems.
  • 6Long-standing learning problems were statistically significantly related to IQ [F(1, 65) 5.27, p ≤ 0.05]. The mean IQ of children with long-standing learning problems (87.6 ± 15.49) was statistically significantly lower than that of children without learning problems (100.5 ± 16.54). Family trouble was statistically significantly related to the educational level of the mother (χ2, 7.23; df, 5; p ≤ 0.05). Of the 15 mothers who had completed university or equivalent education, only one (7%) had family trouble, statistically significantly less than in the other groups; family trouble was present in eight (31%) of 26 mothers with lower and in 11 (44%) of 25 mothers with higher administrative or technical schooling.

Relations within A (IQ, age, gender and educational level of the mother or father) or within B (epilepsy variables) did not yield unexpected information. In C, long-standing behavioral problems were statistically significantly related to family trouble (χ2, 8.82; df, 4; p ≤ 0.00). Ten (45%) of 22 children with family trouble had long-standing behavioral problems, compared with seven (16%) of 44 children who belonged to a happy family.

Children with unclassifiable epilepsy

The three children who were initially diagnosed as having unclassifiable epilepsy whose diagnoses were revised into psychogenic nonepileptic seizures functioned at the lower end of the IQ distribution and were perceived by their parents as having severe long-standing behavioral problems without long-standing learning problems. None of the parents of the three children was able to continue habitual parenting.

DISCUSSION

Parenting and changes therein are sources of behavioral maladaptation in children (32). In particular, stress in the parent(s) has been shown to be potently related to behavioral problems in children with chronic diseases including epilepsy (3,14). We hypothesized that the onset of epilepsy might be an uncertainty-provoking factor that might influence parenting and studied parents' perceptions of the medical and other adversities experienced in the period of onset and diagnosis of epilepsy, a period characterized for many families by uncertainty and often upheaval caused by the signs and symptoms of the epilepsy, the necessary hospital visits and examinations, and ultimately the diagnosis. This subject has not yet been investigated for childhood idiopathic or cryptogenic epilepsy (epilepsy only). The present study is meant to be a first step in unraveling parents' feelings concerning their own as well as their children's adaptation to the changed situation. Relations with illness variables are addressed.

An important finding is that 52% of the parents perceived themselves and 76% perceived their child as adapting well to the changes in relation to the onset of epilepsy. Thirty-three percent of the parents perceived only themselves—and not their child—as being thrown off balance since the onset of epilepsy. Another 15% of the parents perceived both themselves and their child as not adapting adequately to the situation. A minority of parents (9%) perceived themselves as to continuing habitual parenting but their child as being thrown off balance. Forty-three percent of the parents perceived themselves and their child as adapting well to the onset of epilepsy only.

With respect to parent's perception of their child's adaptation, two findings are of interest. First, a context of long-standing adversity significantly influenced the parents' perception of their children's adaptations to the epilepsy-induced adversity. Divorce and other threats to family happiness, but also long-standing difficult behavior of the children are important in the parents' perceptions of their children's adaptation to the onset of epilepsy. Studies in other chronic illnesses are in line with this finding. Correlations between high levels of family stress and maladaptive behavior have been reported for children with insulin-dependent diabetes and for those with brain tumors (33,34). In children with closed-head injury, pretraumatic behavior has been acknowledged to be of even greater importance than the severity of the head injury itself (35). Austin et al. (36) used a questionnaire to rate behavioral problems that existed before the first recognized seizure. They found 26% of children with newly diagnosed benign epilepsy to be rated by their parents as having had clinically relevant behavioral problems. The authors suggested that having parents rate their children's behavior problems before the first recognized seizure will substantially reduce the possibility that side effects of medication or a negative family response to a first seizure are held to account for the higher rates of behavior problems. We now add that quite a proportion of parents are thrown off balance by the signs and symptoms of the onset of epilepsy, to a degree that they discontinue their habitual parenting. Prior experience with epilepsy does not seem to safeguard against this, as even those with a family history of epilepsy are thrown off balance.

For children with new-onset epilepsy, even if the medical prognosis is favorable, the doctor should be aware of the risk of discontinuation of parenting, which results in a decrease of parenting scope and in fear-related overprotectiveness at the time of the medical trajectory that leads to the diagnosis.

Second, age at onset is another major determinant of the parents' perceptions of their children's adaptation to adversity induced by the onset of epilepsy. Children growing into adolescence rather than younger ones are perceived as having been thrown off balance. Presumably, the parents feel that being different is a serious adversity in adolescence, a period when fitting in with peers is of paramount importance for the developing child. The behavioral vulnerability of adolescents with chronic illnesses, including epilepsy, has been extensively reported (37,38).

For both parents and doctors, a practical conclusion may be that adolescents should be credited with possibilities of judging for themselves. One manner of respecting autonomy is for the doctor to communicate with the child with epilepsy herself or himself, rather than about the child with her or his parents.

Finally, the parents perceive their own adaptation to the epilepsy-induced adversity as not appreciably influenced by long-standing adversities. Rather, etiology of the epilepsy influences their perceptions. Cryptogenic epilepsy appears to be felt as a major adversity, leaving the parents in great suspense. To our surprise, seizure type is not significantly related to perceived adaptation. Further study is needed to explain the divergent effects of etiology and seizure type. We surmise that contextual differences between idiopathic and cryptogenic etiologies are relevant. In the case of cryptogenic epilepsy, the doctor's prediagnostic message may convey confusion; the epilepsy may be described as in all likelihood benign, but at the same time, the parents are told that brain imaging is necessary to rule out an underlying disorder of the child's epilepsy. The prediagnostic trajectory in the case of idiopathic epilepsy may be more straightforward; on the basis of the signs and symptoms, the prediction is that the epilepsy is benign and either the remedy is an AED, or a remedy may not even be considered necessary.

Although the explication remains only a surmise, doctors may take advantage of the finding by paying even more attention to preventing unnecessary worries and by conveying the message concerning neuroimaging with due care.

The exceptional effect of three children, diagnosed as having unclassifiable epilepsy, merits discussion. Both neurologic examinations and brain imaging had excluded the diagnosis of a symptomatic etiology. The parents of the three children were seriously upset. Furthermore, the parents perceived the three children as having had considerable behavioral difficulties long before the onset of epilepsy. The intelligence levels of the three children were among the poorest found in our group, but there were no complaints of long-standing learning problems. The children appeared later to have been included incorrectly. They had nonepileptic, psychogenic seizures. This experience taught us that unclassifiable epilepsy with a nonsymptomatic etiology is an insidious diagnosis that should be suspected. We are not aware of other studies addressing behavior or cognition that separate unclassifiable “epilepsy only” or of studies that verify later the initial diagnosis of unclassifiable “epilepsy only.” In these children, extensive multidisciplinary diagnostics are warranted to trace underlying nonepileptic factors and to prevent therapeutic blind alleys.

A final comment addresses the conspicuously wide range of intelligence levels (IQs) found. Mental retardation was not the basis of the poor IQs. The children attended mainstream education. The IQs of healthy controls, also attending mainstream education, had a similar wide range (39). Outlying IQs have been reported previously in groups of children with idiopathic epilepsy who were, according to the authors, not mentally retarded and who attended mainstream education (40).

Notwithstanding the good medical prognosis of “epilepsy only,” the proportion of parents with maladaptive reactions to the onset and diagnosis of epilepsy only is far from negligible. Future research of behavioral pathology in children with epilepsy only should take this into account.

Acknowledgment: We are most grateful to the parents and children who participated in our study. The study was subsidized by the Dutch Epilepsy Foundation (NEF), JANIVO Foundation, and Peugeot Holland NV. We thank Dr. T. J. Olthof for critical comments on an earlier draft of the article and Mrs. B. Vollers-King for language editing. Participants of the Dutch Study of Epilepsy in Childhood (DuSECh): W.F.M. Arts, J. H. Begeer, O. F. Brouwer, C. A. van Donselaar, A. T. Geerts, E. A. J. Peeters, H. Stroink, G. Hageman, R. ten Houten, J. F. de Rijk-van Andel, L. M. E. Smit, and M. J. Wennekens.

Ancillary