• Epilepsy;
  • Mental retardation;
  • Refractory epilepsy;
  • Antiepileptic medications discontinuation


  1. Top of page
  2. Abstract

Summary: Purpose: Epilepsy is a common problem in institutionalized patients with multiple handicaps. Limited data exist on the characteristics of epilepsy in this patient population and the impact of systematic evaluation by an epilepsy service.

Methods: We evaluated 138 patients with epilepsy, institutionalized at a facility that cares for 324 patients with multiple handicaps. Evaluation included EEG, MRI, and video-EEG monitoring. The medication regimen was changed according to seizure diagnosis and the status of seizure control. Follow-up was available for ≥6 months in 110 patients, 1 year for 89, and 1.5 years for 49 patients. We analyzed the seizure and epilepsy diagnosis in this population, as well as the seizure frequency after evaluation and treatment

Results: The 76 male and 62 female patients' ages ranged from 14 to 73 years. Seventy-three patients had fewer than one seizure per month, whereas 29 patients had at least one seizure per month. Of 131 patients taking antiepileptic drugs (AEDs), 62 were receiving monotherapy, and 69 were receiving two or more AEDs. At the last follow-up, overall 55% of patients had reduced seizure frequency, including 23% who became seizure free. Two of 36 patients had spontaneous seizure recurrence after being seizure free with no AEDs for 4 months in one patient and 3 years for the other. Attempts were made to discontinue phenobarbital, primidone, and clonazepam in 21 patients. However, these were discontinued in only five patients.

Conclusions: Epilepsy is heterogeneous in institutionalized patients with multiple handicaps. It is often responsive to medical therapy. Evaluation and treatment by epilepsy specialists had an overall favorable impact on seizure control.

A significant association between epilepsy and mental retardation (MR) has been documented (Corbett, 1989), although the estimated prevalence of epilepsy in individuals with MR is dependent on the survey method (Bowley and Kerr, 2000). In community-based studies, the prevalence rate was reported as 20% (Corbett JA, 1989). In an institutionalized patient survey, the prevalence was ∼32.0% (Mariani E et al., 1993). In one community-based study, the incidence was higher in patients with cerebral palsy, affecting ∼50% of these patients and 21% of those without cerebral palsy, when followed up to age 7 years (Van den Berg and Yerushalmy, 1969). MR and/or cerebral palsy represent the second most common antecedents of epilepsy, observed in 8% of cases (Hauser et al., 1993). The epilepsy in patients with MR is considered to be refractory to antiepileptic drugs (AEDs) in most cases (Pellock and Hunt, 1996). These patients are often treated with AEDs for life and are often maintained on polypharmacy. Reluctance to discontinue medications or make changes is often due to fear of precipitating status epilepticus or unacceptable behavior (Pellock and Hunt, 1996).

Few studies have classified and characterized seizures and epilepsy in this population or evaluated simplifying or tapering of AEDs. We have used the opportunity of a seizure clinic started at an institution for individuals with multiple handicaps for a better understanding of epilepsy and its treatment in this population. This study was specifically undertaken to evaluate a large institutionalized population with MR and epilepsy, to classify seizures and epilepsy in this population, to evaluate the responsiveness of epilepsy to medical therapy, and to assess the potential for simplifying AED regimens and, in some instances, complete discontinuation of AEDs.


  1. Top of page
  2. Abstract


Starting in December 1997, the epilepsy division at Vanderbilt University Medical Center established an epilepsy clinic at the Clover Bottom Developmental Center in Tennessee, an institution for patients with MR and multiple handicaps. The institution cares for 324 residents, including 138 (42%) with proven or suspected epilepsy. We evaluated all 138 patients over a period of 2 years. As the referrals were arranged by clerical staff, the referral order was random, independent of severity of epilepsy.

MR was identified in progress notes, problem lists, and neuropsychological evaluations. MR was classified as mild, moderate, or profound. Mild MR included individuals who had IQ of 56–75 and could perform activities of daily living (ADLs). Moderate MR included individuals who had IQs of 41–55 and required some assistance with routine ADLs. Severe/profound MR was determined in individuals who required total support and had IQ of ≤40 (McDermott et al., 2005).


All subjects received a detailed history and physical examination by one of the epileptologists or epilepsy fellows in our group. Follow-up was for ≥6 months in 110 (80%) patients, ≥1 year in 89 (64%) patients, and ≥1.5 years in 49 (35%) patients. The evaluation included EEG, head CT, and brain MRI as necessary for diagnosis and treatment. When seizure classification was not clear, some patients were further evaluated with inpatient or outpatient video-EEG monitoring at the Vanderbilt University Medical Center Epilepsy Unit. When an inpatient evaluation was chosen, AEDs were reduced or discontinued at the onset of the study, and appropriate changes in the AED regimen were made at discharge, depending on the status of seizure control. After the evaluation was completed, all patients had their diagnosis and treatment regimen reevaluated and medication changes suggested accordingly. Seizures and epilepsy were classified according to the International League Against Epilepsy (ILAE) classifications. The final reclassification was based on the most compelling data. Ictal recordings, when available, were the main basis for reclassification. If ictal recordings were not available, then interictal epileptiform discharges determined epilepsy classification (partial vs. generalized). Interictal nonepileptiform abnormalities (focal or generalized slowing or attenuation) were usually not a basis for reclassification.

Medication changes

Each AED regimen was modified according to seizure diagnosis and the status of seizure control. For patients taking two or more AEDs, simplification of the AED regimen was considered. The first drug to be discontinued was the one with more obvious adverse effects or the one that was least suitable for the specific epilepsy or seizure classification.

Data analysis

We analyzed each patient's seizure and epilepsy diagnosis as well as the seizure outcome after evaluation and treatment. All data were entered in a computerized database, including AEDs at the start of the study, later modifications in the treatment regimen, and follow-up data including change in seizure frequency. Follow-up seizure frequency was based on seizure counts by direct-care staff. Seizure freedom was defined as no seizures for the preceding 3 months, unless the baseline seizure frequency was less than monthly, in which case, the seizure-free interval had to be at least twice the average seizure-free interval.

Fisher's exact test, χ2 test, Pearson's correlation, and the t-test were used for data analysis

IRB approval

The study was approved by Division of Mental Retardation Services (DMRS) Research Project Review Board, the scientific review committee of Tennessee Division of Mental Retardation Services.


  1. Top of page
  2. Abstract


The 62 (45%) female and 76 (55%) male patients' ages ranged from 14 to 73 years (mean, 40.5 years). The age at onset of seizures, known in 116 patients, ranged from 0 to 68 years (mean, 4.6 years; median, 1 year). MR was classified as severe/profound in 117 (84.7%) patients, moderate in 16 (11.5%) patients, and mild in five (3%) patients. One hundred two (74%) patients were totally dependent for ADLs; 30 (21%) patients were partially dependent; and six (4%) were independent. Ambulation was impaired in 123 (89%) patients (22 used a wheelchair) and normal in 15 (11%) patients. Communication skills were normal in only two patients.


The etiology of MR or epilepsy was unknown in 34 (24.6%) patients. Known risk factors for epilepsy and MR included perinatal hypoxic insult in 38 (27.5%) patients, birth trauma in 15 (11%) patients, meningitis/encephalitis in 21 (15%) patients, head injury in seven patients, hydrocephalus in four patients, Down syndrome in three patients, tuberous sclerosis in three patients, phenylketonuria in two patients, pervasive developmental disorder in two patients, Rett syndrome in two, Angelman syndrome in one, and Turner syndrome in one patient. In the unknown category, four patients had history of febrile convulsions, and one had a strong family history of MR.

Epilepsy classification

Sixty-eight (49%) patients were diagnosed with partial epilepsy, 39 (28%) patients with generalized epilepsy, seven (5%) patients with both partial and generalized seizures, and 24 (17%) patients had epilepsy that was undetermined as to whether partial or generalized.

Seizure diagnosis at screening

Only one seizure type was reported in 114 (82%) patients, and 24 (18%) patients had more than one seizure type (Table 1). At the first evaluation, the most common seizure diagnosis was generalized tonic–clonic. Eighty-six (62%) patients had generalized tonic–clonic seizures as a single seizure type, and 22 (16%) patients had one of two or more seizure types. One patient was thought to have behavioral attacks only.

Table 1. Seizure classification at initial visit
No. of seizure typesSeizure type(s)Number of patients
  1. Two patients had a third seizure type not listed above.

Single seizure type (114)Generalized tonic–clonic86 (75%)
Complex partial15 (13%)
Unknown12 (11%)
Tonic1 (1%)
>1 seizure type (24)Seizure type 1Seizure type 2 
Generalized tonic–clonicComplex partial12 (50%)
Simple partial1 (4%)
Tonic 3 (12%)
Absence2 (8%)
Myoclonic 3 (12%)
Unknown2 (8%)
Complex partialAtonic1 (4%)

Seizure frequency

A wide range existed for seizure frequency. Thirty-six (26%) patients had been seizure free for ≥2 years; 16 (11%) patients had fewer than one seizure per year; 57 (41%) patients had seizures less than once a month, but more than once a year; 20 (14%) patients had monthly seizures, eight patients had weekly seizures, and one had daily seizures. No significant relations was found between seizure frequency and severity of MR (mean seizure frequency tended to be less in those with severe MR) (t-test, p = 0.9498), classification of epilepsy (partial vs. generalized), or age at seizure onset. No significant correlation was noted between baseline seizure frequency and number of baseline AEDs, but a significant positive correlation was seen between baseline seizure frequency and number of AEDs at follow-up (correlation at 6 months: r= 0.46; at 1 year: r= 0.49; at 1.5 years: r= 0.42; p < 0.01 for each), possibly reflecting a combination of lesser numbers of AEDs at baseline and a greater likelihood of simplification of AED regimens in patients with a lower baseline seizure frequency. When we analyzed the correlation between change in number of AEDs at last follow-up and baseline seizure frequency, the correlation was in the correct direction (greater reduction with lower seizure frequency), but not significant (r=−0.17).

Antiepileptic drugs

At baseline, seven (5%) patients were taking no AEDs, 62 (45%) patients were taking monotherapy, 55 (40%) patients were taking two AEDs, 12 (8%) patients were taking three AEDs, and two patients were taking four AEDs. The antiepileptic medications were phenytoin (PHT) in 40 patients, carbamazepine (CBZ) in 52 patients, phenobarbital (PB) in 44 patients, valproic acid (VPA) in 43 patients, primidone (PRM) in nine patients, clonazepam (CZP) in 10 patients, gabapentin (GBP) in 11 patients, lamotrigine (LTG) in five, topiramate (TPM) in one, and tiagabine (TGB) in one patient.


Routine EEGs were performed in 10 (7%) patients; five recordings showed focal epileptiform discharges, one showed generalized epileptiform discharges, and four were normal.

MRI was performed in 16 (11.5%) patients. Nine had diffuse abnormalities, five had focal abnormalities, and two had normal MRI. No patient had a neoplasm or other lesion requiring surgical intervention. One patient with tuberous sclerosis already had an MRI study indicating two enhancing nodules in the left lateral ventricle.

EEG-CCTV evaluation was ordered in 62 (45%) patients and was performed in 52 (38%) patients. Generalized ictal or interictal epileptiform discharges were recorded in 18 patients, and focal ictal or interictal epileptiform discharges in 18 patients. Sixteen patients did not have epileptiform or ictal activity; six patients had generalized, and two patients, focal nonepileptiform abnormalities, and eight had no abnormalities. In 19 patients (36.5%), the EEG-CCTV changed or clarified the seizure and epilepsy diagnosis conclusively based on ictal or interictal epileptiform discharges (Table 2) Thirteen of these were previously given an incorrect diagnosis: three patients previously considered to have generalized seizures had evidence of partial epilepsy, eight patients considered to have partial seizures had evidence for generalized epilepsy, and two patients who were considered to have both partial and generalized seizures had evidence only for generalized epilepsy. Among six patients whose seizures were of unknown classification, four patients had evidence for partial epilepsy, and two patients had evidence for generalized epilepsy.

Table 2. Epilepsy classification before and after evaluation
Epilepsy classificationNumber at screeningNumber after evaluation (including EEG-CCTV in 52 patients)
  1. EEG-CCTV changed the epilepsy diagnosis in 19 (36.5%) patients.

Both partial and generalized 7 5

Outcome at follow-up

Outcome by follow-up interval

Six months: Although follow up was for ≥6 months in 110 (80%) patients, only 80 patients actually returned for a reevaluation visit at 6 months. AEDs at screening and at last follow-up for the 110 patients with ≥6 months of follow-up were analyzed. A trend was seen toward using newer AEDs (Table 3). Four patients had worse outcomes at 6-month follow-up. At 6 months, AEDs had been changed (AED added and/or AED removed) in 20 patients (25%). Only AED dose(s) was changed in 38 patients. Nineteen patients (23%) had been switched to monotherapy, but no patients had been taken off AEDs. Because PB, PRM, and CZP can have adverse cognitive effects, a taper was attempted as appropriate (Table 3). Fifty percent were improved in seizure frequency, including 12 who became seizure free (Table 4). Among the fifty patients who returned for a 1.5-year follow-up visit, 56% had fewer seizures or were seizure free

Table 3. AEDs at screening and at last follow-up for the 110 patients with ≥6 months of follow-up
AEDsNo. of patients at screeningNo. of patients at follow-up
Phenytoin33 (30%)40 (36%)
Carbamazepine42 (38%)36 (33%)
Phenobarbital37 (34%)33a (30%) 
Valproic acid32 (29%)31 (28%)
Primidone8 (7%)6a (5%) 
Clonazepam9 (8%)7a (6%) 
Gabapentin10 (9%) 14 (13%)
Lamotrigine5 (4%)9 (8%)
Felbamate0  1 (0.9%)
Topiramate02 (2%)
  1. aClonazepam was started de novo in two; phenobarbital, in one; and primidone in two.

AED changed36 (33%)
 Decreased no. of AEDs10 (9%) 
 Increased no. of AEDs10 (9%) 
 AED(s) switched, no. unchanged16 (14%)
Only AED doses changed45 (41%)
Taken off AEDs2 (2%)
No AEDs at screening; added AED3 (3%)
Table 4. Seizure outcome at follow-up (number of patients and percentages between parentheses)
Follow-up visit (number of patients)Improved: seizure freeSame: still seizure freeImproved: still having seizuresSame: still having seizuresWorse
6 mo (110)12 (15%) 14 (17.5%)27 (34%)23 (29%)4 (5%) 
12 mo (89)14 (19%)20 (27%) 20 (27%)15 (20%) 5 (6.8%)
18 mo (49)11 (22%)8 (16%)17 (34%)11 (22%)3 (6%) 
Last follow-up (110)25 (23%)19 (17%) 35 (32%)19 (17%)11 (10%) 
Seizure frequency and AED status at the last follow-up

We investigated outcome at the last follow-up for all 110 patients who returned for follow-up (Table 4). Overall, 55% of patients were improved, including 23% who became seizure free. Thirty-six patients had their AEDs changed, and 45 patients had only their AED doses changed (Table 3). Two patients had spontaneous seizure recurrence after being seizure free with no AEDs for 4 months for one patient and 3 years for the other, and were given PHT. Five patients were converted to monotherapy. One patient was taken off AEDs altogether. AEDs were changed in 36 patients. Five patients previously maintained on monotherapy had worsening of their seizures and were given dual therapy.

Attempts were made to discontinue PB, PRM, and CZP in 21 patients. Five (24%) patients were successfully taken off these medications. In the remaining patients, these medications were continued or restarted.

Predictors of favorable change at follow-up

We divided the patient group with follow-up into those that improved (including seizure freedom) and those that did not improve or worsened. Comparison of baseline seizure frequency in the two groups showed no significant difference, although a trend was noted for a lower seizure frequency in the unchanged/worse group. This may reflect that patients with low seizure frequency may have been more likely to be kept on the same schedule or to have medications discontinued (if seizure free). No difference between the groups was found with respect to number of AEDs (p = 0.74), age at seizure onset (p = 0.29), or age at first evaluation (p = 0.87).


  1. Top of page
  2. Abstract

This study was conducted to evaluate the impact of an epileptology clinic at an institution that cares for patients with MR and epilepsy. We included all patients with a diagnosis of epilepsy, including those not taking AEDs (all were referred to the epileptology clinic). Seizure control was improved in 55% of patients. The key specific changes that occurred after implementation of this clinic were improved classification of seizures and epilepsies and subsequent medication adjustments, including initiation of new drugs, reduction of polypharmacy, and discontinuation of sedating AEDs.

The seizure and epilepsy classifications were corrected in 36.5% of our patients, with the help of video-EEG monitoring in many patients. Classification is important in seizure management. Many patients with MR have abnormal behavior episodes that may resemble epileptic seizures but that are not epileptic in origin (Crews et al., 1994). Pseudoseizures, movement disorders, breathholding spells, gastroesophageal reflux, sudden aggression, and self abuse are some of examples of such behavior, which can mimic epilepsy in these patients. Video-EEG monitoring enables any behavior to be analyzed in relation to the EEG changes and has proven to be an excellent tool to distinguish between epileptic seizures and nonepileptic spells and to classify seizures (Porter et al., 1976), Sutula et al., 1981). We performed EEG-CCTV on 52 (38%) patients, and of those, 36 (69%) had either ictal or interictal abnormalities. Video-EEG monitoring is more often productive in this population, and short-term monitoring may be sufficient to classify frequent events (Thirumalai et al., 2001). In the institutionalized patients with MR and epilepsy, partial seizures with or without secondary generalization are reported to be more prevalent than other seizure types (Mariani et al., 1993). Other seizure types in this population are absence, generalized tonic, clonic, tonic–clonic, and myoclonic seizures. In our study, the vast majority of patients had a single seizure type, and partial seizures were more frequent than generalized seizures. Seizure type was based on the EEG, MRI, and video-EEG monitoring. Because we were unable to videotape seizure occurrences in the residential areas, it is possible that the four patients categorized as having generalized tonic seizures was an underestimate. Without recorded events, the seizure diagnosis in these patients can be difficult. The clinician has to rely on sparse history and on the account of the caretakers and nursing staff who have limited knowledge of seizures and epilepsy.

After evaluation and definitive seizure diagnosis, medication changes were made in 82 (74%) of 110 patients who returned for follow-up. One important medication change is medication discontinuation. Medication discontinuation was attempted predominantly to remove sedating drugs and alleviate polypharmacy, although it was occasionally considered for patients seizure free for many years if it was suspected that therapy was no longer needed. In our study we were able to discontinue PB, PRM, or CZP in five of 21 (24%) patients without compromising seizure control. The benefits of withdrawal of sedating AEDs are improved alertness, improved mobility, and better cognitive functioning. Conversely, discontinuation of AEDs has associated risks, including relapse or exacerbation of seizures with associated risk of injuries, status epilepticus and sudden unexpected death. Other studies have reported greater success and suggested that the treatment of many patients with epilepsy can often be best accomplished with the use of fewer drugs or monotherapy (Callaghan et al., 1984; Ferngren et al., 1991; Poindexter et al., 1993). In one study, 48% of seizure-free patients with MR and epilepsy remained seizure free after AED discontinuation, even after a follow-up period of 8 years (Alvarez, 1989). In another study, 12% of patients had an AED discontinued as part of an attempt to simplify polypharmacy, but 45% of patients who had an AED discontinued required restarting that AED over a 10-year follow-up period (Pellock and Hunt, 1996). Nevertheless, this suggests that precautions should be taken with tapering these medications in this patient group.

We found no positive correlation between seizure frequency and severity of MR. The patients with severe MR tended to have lesser seizure frequency. This is different from the current literature and may be attributable to a selection bias related to the criteria for institutionalization at the study facility. It is possible that patients with both severe MR and severe epilepsy were more likely to be in a different type of institution. We did not find any significant correlation with seizure frequency and classification of epilepsy (partial vs. generalized) or age at seizure onset.

In conclusion, epilepsy in institutionalized patients with MR can frequently be well controlled with accurate seizure diagnosis. Simplification of the medication regimen is possible, but caution and careful weighing of risks and benefits are needed during discontinuation of sedating AEDs.


  1. Top of page
  2. Abstract
  • Alvarez N. (1989) Discontinuance of antiepileptic medications in patients with developmental disability and diagnosis of epilepsy. American Journal of Mental Retardation 93:593599.
  • Bowley C, Kerr M. (2000) Epilepsy and intellectual disability. Journal of Intellectual Disabilility Research 44:529543.
  • Callaghan N, O'Dwyer R, Keating J. (1984) Unnecessary polypharmacy in patients with frequent seizures. Acta Neurologica Scandinavica 69:1519.
  • Corbett JA. (1990) Epilepsy and mental retardation. In DamM, GramL (Eds.) Comprehensive epileptology. Raven Press, New York , pp. 271280.
  • Hauser WA, Annegers J, Kurland L. (1993) Incidence of epilepsy and unprovoked seizures in Rochester, Minnesota: 1935–1984. Epilepsia 34:453468.
  • Crews WD Jr, Bonaventura S, Rowe F. (1994) Dual diagnosis: prevalence of psychiatric disorders in a large state residential facility for individuals with mental retardation. American Journal of Mental Retardation 98:724731.
  • Ferngren H, Akerstrom I, Rane A. (1991) Mono- or polypharmacotherapy in institutionalized epileptic children with severe mental retardation? A team approach for optimizing antiepileptic therapy. Acta Paediatricia Scandinavica 80:458465.
  • Mariani E, Ferini-Strambi L, Sala M, Erminio C, Smirne S. (1993) Epilepsy in institutionalized patients with encephalopathy: clinical aspects and nosological considerations. American Journal of Mental Retardation 98(suppl):2733.
  • McDermott S, Moran R, Platt T, Wood H, Isaac T, Dasari S. (2005) Prevalence of epilepsy in adults with mental retardation and related disabilities in primary care. American Journal of Mental Retardation 110:4856.
  • Pellock JM, Hunt PA. (1996) A decade of modern epilepsy therapy in institutionalized mentally retarded patients. Epilepsy Research 25:263268.
  • Poindexter AR, Berglund JA, Kolstoe PD. (1993) Changes in antiepileptic drug prescribing patterns in large institutions: preliminary results of a five-year experience. American Journal of Mental Retardation 98(suppl):3440.
  • Porter RJ, Penry JD, Wolf AA. (1976) Simultaneous documentation of clinical and electroencephalographic manifestations of epileptic seizures. In KellawayP, PetersonI (Eds.) Quantitative analytical studies in epilepsy. Raven Press, New York pp. 253268.
  • Sutula TP, Sackellares JC, Miller JQ, Dreifuss FE. (1981) Intensive monitoring in refractory epilepsy. Neurology 31:243247.
  • Thirumalai S, Abou-Khalil B, Fakhoury T, Suresh G. (2001) Video-EEG in the diagnosis of paroxysmal events in children with mental retardation and in children with normal intelligence. Developmental Medicine and Child Neurology 43:731734.
  • Van den Berg BJ, Yerushalmy J. (1969) Studies on convulsive disorders in young children, I: incidence of febrile and nonfebrile convulsions by age and other factors. Pediatric Research 3:298304.