Long-term social outcomes for children with epilepsy


Address correspondence to Carol S. Camfield, MD, FRCP(C), IWK HEALTH CENTRE, P.O. Box 9700, Halifax, Nova Scotia, Canada B3K 6R8. E-mail: Camfield@dal.ca


SummaryChildren with epilepsy often grow into adults with significant social problems including decreased employment, marriage, social relationships, and independent living arrangements. These problems are noted in population-based longitudinal and cross-sectional studies from many countries. Learning disorder and mental handicap are the most consistent predictors of poor social outcome. Epilepsy variables, even remission, appear to have little effect. The influence of epilepsy on social outcome is greater than found in other childhood chronic disease control groups. More attention and research is needed to correct these unfortunate outcomes.

At the time of diagnosis of childhood epilepsy psychosocial problems and comorbidities are common. These include: (1) mental handicap in 21–34% (Camfield & Camfield, 2007; Sillanpaa, 1992); (2) learning disorder in 51– 57% (Sillanpaa, 1990, 1992; Camfield et al., 1993; Oostrom et al., 2003); (3) mental health problems in 25–34% (Rutter et al., 1970; Oostrom et al., 2003; Austin et al., 2001), and (4) attention deficit disorder in 10% (Austin et al., 2001; Hesdorffer et al., 2004; Dunn & Kronenberger, 2005). These difficulties may reflect an underlying brain abnormality or family dysfunction, and they have important effects on later outcome children with epilepsy.

Surprisingly, the continued social problems in normally intelligent children with epilepsy are not clearly related to remission or any other biologic factor except learning disorder. The social outcomes are worse in children with epilepsy than in either community or chronic illness controls.

Published studies that address the long-term social outcome of childhood epilepsy are not easily compared because they have used varying definitions for predictor variables (such as mental handicap, neurological deficit and type of syndrome) and varying definitions of what constitutes a satisfactory long-term outcome.

Initial Suggestions of Poor Social Outcome from Population-based Studies

Two excellent British prevalence studies, the Isle of Wight study (Rutter et al., 1970) and the British Child and Adolescent Mental Health Survey (Davies et al., 2003), found that over 25% of children aged 5–14 years with epilepsy had “psychiatric difficulties” as compared with 7–9% of normal controls. The mental health survey also found that only 11% of children with diabetes, a chronic disease control, had behavioral difficulties. In addition, psychiatric difficulties were experienced by 56% of children with epilepsy plus comorbid neurological problems.

Large Population-based Long-term Follow-up Studies From Many Regions of The World

Oulu Finland (Kokkonen et al., 1997)

This population-based Finnish series began as a prevalence cohort of all noninstitutionalized children with epilepsy (n = 92) born between 1964–67. When these patients were 22 years of age, they were compared with 211 controls from same birth cohort by interview or questionnaire. Those with childhood onset epilepsy had significantly greater social immaturity and psychiatric morbidity. In addition, 20% of the epilepsy group had completed <9 years of elementary education and presumably had severe learning disorders or mental retardation (vs. 2% controls) and 27% did not graduate from high school (vs.13% controls). Although 50% of those graduating in both groups went on for further educational or vocational training, 27% with epilepsy (vs.11% controls) did not complete the professional educational course. If the vocational courses were completed, employment status and level of social status was similar in both groups. Even if they were eligible for employment, those with epilepsy were more likely on disability pension (12% vs. 2% controls) and were twice as likely to be unemployed. In addition, they were more often unmarried, living alone or with parents (69% vs. 40% controls).

Logistic regression analysis indicated that these differences were unrelated to any epilepsy factors (remission status, use of antiepileptic drugs (AEDs) social economic status, localization related epilepsy, or fine motor disturbance). Rather, the presence of cognitive or neurologic disability accounted for most of the poor social outcome.

Japan (Wakamoto et al., 2000)

This population-based cohort of 148 children developed epilepsy before age 16 years between 1961–1992 within the prefecture of Ehime. Ninety-three percent were followed up at a mean age of 26 years through medical record review and telephone interviews. They were compared to age matched controls from the same geographic area.

All of the 99 survivors with normal intelligence attended regular class. As compared to controls, an equal number graduated from high school and vocational or higher levels of education (p = 0.47). Only seven experienced psychiatric disturbances—school refusal, violent behavior, depression, or hysteria. However, among the 49 young adults with mental retardation the outcome was dismal: only 14% attended regular class; 6% entered high school; 20% were employed; and 2% married.

Employment among the normally intelligent occurred in 95%, and the marriage rate was nearly the same as controls. Again, the scenario was dismal for those with mental retardation: 6% entered high school, 20% were employed, 2% married, and a large number lived in social isolation.

Unlike the other population-based cohort studies, the normally intelligent young adults generally had favorable long-term educational and employment outcomes. When comorbidity also included mental retardation, the social outcome was poor. It is likely that diverse reporting methods or differing societal norms and expectations in Japan significantly affect these social outcome results, which differ appreciably from the majority of other countries.

Turku, Finland (Sillanpaa, 1990; Jalava et al., 1997)

Sillanpaa developed a prevalence cohort of 245 children with active epilepsy from the Turku catchment in 1961–64 and compared them to 99 adult controls from a nationwide Finnish registry. After a remarkable 35-year follow-up, 11% with a learning disorder, and 33% with near normal intelligence received a disability pension (vs.1% of controls, p = 0.02). Those entirely off AEDs (n = 66) also had similar lifelong effects. Table 1 documents the catastrophic long-term psychosocial outcome of these children followed into mid-adulthood.

Table 1. Psychosocial outcome of 161 Finnish adults with childhood epilepsy compared with 99 employed individuals (Jalava et al., 1997)
 % in controls (n = 99)% with IQ >85 (n = 161)% with IQ 75–85 (n = 21)% with IQ <71 (n = 63)
<High school education2347100 100
No vocational training526695100
Unemployed 82859100
Living alone103268 97
Lower SES525795100

In 1997, 100 survivors from this cohort with “epilepsy only” were matched to 100 “employee” controls and 100 controls from a nationwide registry. Those with childhood epilepsy had reduced levels of education, employability, social economic status, and frequency of marriage (p = 0.0001). Fifteen percent were dependent on other caretakers. Overall, those with epilepsy felt that the ability to control their lives was “poor or missing” four times more frequently than “employee” controls. Not surprisingly, those on AEDs were significantly less satisfied with their present health/life than those off medication. It was concluded that many children with “epilepsy only” have persistent and significant social adjustment and competence problems in later adulthood.

Canada: Nova Scotia Childhood Epilepsy Population-Based Cohort Epilepsy (Camfield et al, 1993; Camfield & Camfield, 2007; Wirrell et al., 1997)

This incidence cohort consists of 692 children aged 1 month to 16 years with newly diagnosed epilepsy in 1977–85. The epilepsies were divided into three groups: (1) Partial and Convulsive Epilepsies (main seizure types of generalized tonic clonic, partial and secondarily generalized (n = 511, 76%); (2) Absence Epilepsies (n = 97, 14%); or (3) Secondary Generalized Epilepsies (n = 85, 12%).

For the 304 children with partial and convulsive epilepsies plus normal intelligence who were followed into school: (1) 35% repeated a grade; (2) 23% required mental health assessment; and (3) 5% were prescribed psychotropic medication. For those followed to age ≥18 years: (1) 2% had a criminal conviction; (2) 13% of females had an unplanned pregnancy; (3) 16% were socially isolated; and (4) 30% were financially dependent or unemployed. Each of these unfavorable outcomes is considerably more common than the Nova Scotian reference population. Poor social outcome was much more common in the 12% of patients with a learning disorder [multivariate analysis (p = 0.0001).]

Children with childhood absence epilepsy (CAE) were followed to age >18 years and then compared by structured interview with a chronic disease control group – children with juvenile rheumatoid arthritis (JRA). Both groups were medically similar; however, their social outcome was very disparate (Table 2). Young adults with CAE had far less education, lower working status and markedly higher rates of behavioral and psychiatric difficulty. Thus, it is clear that these social and educational problems are associated with epilepsy and/or stigma and not simply the result of a generic effect of a chronic disease.

Table 2. Social outcome of children with childhood absence compared with juvenile rheumatoid arthritis
Mean age of onset7.4 years6.7 years
Follow-up age23 years23 years
Seizures/symptoms in last year of follow up34%72%
Remain on daily medication32%38%
Behavior problems41%10%
Unplanned pregnancy34%3%
Heavy alcohol drinking39%16%
Psychiatric consultation54%31%
Unskilled labor53%16%

The social outcome was also very poor for the 46 survivors with secondary generalized epilepsy (SGE) at end of a follow-up that averaged 23 years. Mental handicap was present in 76% and 48% were unable to walk. Assistance with all activities of daily living was required by 60% and complete financial dependency on parents or the state occurred in 86%. No pregnancies occurred, although sexual behavior was problematic in 11%. Aggressive behavior was difficult in 29% and social isolation was noted by the caretakers in 21% and was not related to degree of comorbid disability. Therefore, the social outcome of children with SGE is usually disappointing; most are mentally handicapped and/or highly dependent with ongoing frequent seizures. Only 6% were seizure-free, off AEDs, living on their own and financially independent.


Children with epilepsy have very high rates of social problems in adulthood, even if they are neurologically and intellectually normal. Furthermore, those with mental retardation do less well. Remission of epilepsy does not alter this problem—even if the seizures go away there is often need for ongoing psychosocial support.

Disclosure of Conflicts of Interest

The authors have declared no conflicts of interest.