Volume 11 - Março de 2006
Editor: Giovanni Torello
Setembro de 1997 - Vol.2 - Nş 9
Psychiatric Morbidity in Elderly Epileptics
Osvaldo P. Almeida, MD, PhD
Vinte e nove pacientes idosos atendidos em um ambulatório especializado em epilepsia (Reino Unido) foram avaliados com o objetivo de investigar a presença de sintomas psiquiátricos de acordo com uma entrevista semi-estruturada para a avaliação do estado mental de idosos (GMSS).
Sintomas psiquiátricos estavam presentes em 89,7% dos pacientes. Sintomas depressivos (55,2%), ansiedade (20,7%), disfunção somática (20,7%), ansiedade situacional (13,8%), dificuldade de pensamento (34,5%), lentificação (17,2%), desconforto social (27,6%), irritabilidade (20,7%) e sintomas psicóticos (17,2%) foram todos freqüentemente observados entre os idosos avaliados. Esses sintomas foram considerados graves o suficiente para preencherem os critérios diagnósticos da CID-10 em 75,6% dos casos.
A avaliação cognitiva geral destes pacientes através do NART, MMSE, e CAMCOG mostraram que o desempenho deles estava dentro dos limites da normalidade.
O diagnóstico de epilepasia focal foi feito em 69% dos casos. Convulsões parciais complexas estavam presentes em 69% dos pacientes. A epilepsia foi considerada secundária a doença cerebral em 41,4% dos entrevistados. Não houve associação significativa entre a presença de sintomas psiquiátricos, tipos de convulsão, gravidade da epilepsia e lesões cerebrais.
Until recently, epilepsy with onset in old age was considered uncommon. Epidemiological surveys published during the past 20 years have proved this assumption wrong. There is a marked increase in the incidence of epilepsy in the elderly. Hauser and Kurland (22) reported that the annual incidence of the disorder rose from 8:100,000 in those aged 40-59 to 75.7:100,000 in subjects aged 60 or older. Similar figures were found in Denmark (27), although an even higher rate of 127:100,000 were later described in Southwest France (26). A recent UK survey of a primary care computerised database established that the overall incidence of epilepsy in subjects over the age of 60 was 117:100,000 (40), 24% to 35.5% of the new cases requiring treatment for epilepsy were older than 60 (36,40).
This rise in the incidence of epilepsy in the elderly associated with the relentless ageing of the world?s population led to an obvious growth in the prevalence of epilepsy in old age. This was well documented in Rochester-USA, where the prevalence of epilepsy increased from 2.9:1,000 in 1940 to 9.4:1,000 in 1980 in those over the age of 55 (23). In the UK, the prevalence of epilepsy in subject's aged 60 or over was recently reported as 11.8:1,000 (40).
Epilepsy with onset late in life is partial in 50% to 80% of cases (23, 27, 39). Acute confusional states (11, 17) and post-ictal paresis (18) may be the most distinct manifestation of epileptic activity in some patients. Other features such as cognitive impairment, psychological and behavioural disturbances are probably common in old age, but their prevalence in elderly epileptics has not yet been established.
In fact, subjects with focal brain lesions and partial seizures seem particularly vulnerable to the development of psychiatric symptoms (8, 41, 42). These risk factors are more frequent in late onset cases, particularly when seizures start after the 6th decade of life (23, 27, 39). It is, therefore, surprising that no studies have so far looked at the prevalence of psychiatric symptoms in a population of elderly epileptics.
We performed this study in order to address this issue.
Subjects and Methods
Patients were recruited from the Epilepsy Outpatient Clinic at King?s College Hospital in South London. The clinic is a specialised referral service for the Camberwell Health Authority, which includes the London Borough of East Lambeth and South Southwark. All patients aged 55 or older attending the clinic from 1 September 1991 to 1 August 1992 were approached. Thirty-two potential subjects satisfied Gunn and Fenton?s (20) operational criteria for the definition of epilepsy: at least 3 seizures in any two-year period. Three patients were later excluded; one for not agreeing to participate and two for not having had seizures for a period longer than 5 years.
The classification of seizures and epileptic syndromes followed the guidelines proposed by the Commission on Classification and Terminology of the International League Against Epilepsy (2, 3, 4). If patients presented more than one type of seizure, that occurring more frequently was used for classification purposes. Epilepsy severity was ranked as mild (less than 3 seizures in the year prior to assessment), moderate (between 4 and 12 seizures in the year prior to assessment), or severe (13 or more seizures in the year prior to assessment).
A 16 channel electroencephalographic recording (EEG) and a computerised tomography (CT) using the GE 9800 whole body CT scanner had been obtained for all participants at the Maudsley Hospital within a 3 year period from examination. The recording and images were qualitatively analysed and reported by consultant specialists.
Cognitive and Psychiatric Assessment
Cognitive assessment included the National Adult Reading Test (NART) (28) which provides an estimate of IQ as measured by the Wechsler Adult Intelligence Scale-Revised (WAIS-R) (43), and de Cambridge Cognitive Assessment Schedule (CAMCOG). The CAMCOG is the cognitive battery from the CAMDEX (34) and consists of 72 questions assessing various aspects of intellectual abilities: orientation, language, memory, attention, praxis, perception, calculation and abstract thinking. Scores may range from 0 to 107. Additional questions have been included in the battery to allow for the computation of the Mini Mental State Examination Score (MMSE) (15).
Psychopathology was assessed using the Geriatric Mental State Schedule (GMSS) (6, 21). The GMSS is a semi-structured psychiatric interview for the elderly and its development was based on the Present State Examination (PSE-9) (44). It is divided in 25 different sections assessing symptoms of worry, anxiety, depression, physical health, tension, somatic dysfunction, situation anxiety, autonomic dysfunction, thinking difficulties, thought disorder, slowness, elation, social discomfort, delusions, self-depreciation, irritability, obsessions, interest, concentration, depersonalisation, perceptual distortion, hallucinations, drug and alcohol abuse (21). A positive response to any items of the above sections was regarded as indicative of the presence of symptoms associated with these factors.
Psychiatric diagnoses followed the Diagnostic Criteria for Research of Mental and Behavioural Disorders of the ICD-10 (45).
The data was analysed with SPSS/PC+ 4.0 (29). Two main methods of statistical analysis were employed. Nonparametric tests (Mann-Whitney U test) were used in the analysis of discrete numerical data (age, age of onset, length of illness and the scores of cognitive tests). Likelihood ratio analysis of contingency tables with continuity correction was used in the investigation of categorical data (the statistical result is distributed as chi-squared). Fisher?s Exact Test (FET, two-tail) was used for 2x2 tables with expected frequency lower than 5 in 2 or more cells. Cohen?s kappa was calculated to measure agreement between CT and EEG reports. Ninety-five percent confidence intervals (CI) were calculated for the estimated means.
Patients? basic demographic features are summarised in table 1. The patients? mean age was 66.0 (CI=63.0 to 70.4). They were more frequently women (65.5%), unmarried (58.6%), British (86,2%), unemployed/retired (75.9%), and right handed (93,1%).
1: DEMOGRAPHIC AND COGNITIVE FEATURES
Epilepsy: diagnosis, classification, aetiology and treatment
The clinical features associated with epilepsy are reported in table 2. The mean age at onset of seizures was 45.0 (CI=37.4 to 52.5), with the duration of the disorder ranging broadly from 1 to 64 years (mean=20.9, CI=14.2 to 27.5). Six subjects (20.6%) exhibited more than one type of seizure, which was partial complex in 69% of all cases. The aetiology of seizures was not identified (i.e., idiopathic seizures) in 58.6% of patients. Carbamazepine was the chosen treatment for 62.1% of subjects, other drugs being used much less frequently. The clinical picture was considered mild, moderate and severe in 38.0%, 31.0% of cases respectively.
TABLE 2: EPILEPTIC PATIENTS - CLINICAL FEATURES
Epilepsy: CT and EEG findings
The majority (58.6%) of brain scans were considered within the normal limits for age. Generalised brain pathology, mainly cortical atrophy, was reported in 20.7% of patients. Focal brain abnormalities affecting temporal and frontal lobes were described in 20.7% of cases. The agreement between CT and EEG reports in terms of localisation of lesions was poor (kappa=0.21). These results are described in table 2.
The NART mean score was 102.2 (CI=96.8 to 107.0), whereas the MMSE and the CAMCOG mean scores were 26.9 (CI=25.6 to 28.3) and 87.7 (CI=82.7 to 92.6) respectively, all well within the normal limits for this age group. These results are summarised in table 1.
Figure 1 shows the frequency distribution of patients? psychiatric symptoms. Many patients scored positively for more than one GMS-factor (mean number of positive scores for patients=4.11, CI=3.0 to 5.2). In fact, only 10.3% of patients showed no psychiatric symptoms. Factors associated with depression and anxiety were particularly common among patients. Depressive symptoms were present in 55.2% of patients. Other frequent factors included anxiety (20.7, ?somatic dysfunction? (20.7%), ?situation anxiety? (13.8%), ?thinking difficulties? (37.9%), ?slowness? (17.2%), ?social discomfort? (27.6%) and ?irritability? (20.7%). Psychotic symptoms (including perceptual distortion, auditory hallucinations or delusions) were present in 17.2% of patients.
The ICD-10 psychiatric diagnoses associated with epilepsy are listed in table 3. Mood disorder was diagnosed in 20.7% of patients and included 13.8% of cases with mild depressive episode, 3.4% with severe depressive episode and 3.4% with an organic manic disorder. The 17.2% neurotics were diagnosed as agoraphobia without panic disorder (3.4%), generalised anxiety disorder (10.4%) and mixed anxiety and depressive disorder (3.4%). Two patients (6.7%) received the diagnosis of persistent delusional disorder. Alzheimer?s disease with late onset (6.7%), mild cognitive disorder (6.7%), organic hallucinosis (10.3%) and organic personality disorder (6.7%) were the remaining psychiatric diagnoses.
TABLE 3: PSYCHIATRIC DIAGNOSES AND AETIOLOGY
Patients with onset of epilepsy in old age (55 or over) were less likely to exhibit psychotic features than their younger counterparts (FET, p<0.05), although partial seizures (FET, p<0.05) and organic brain disease (FET, p<0.05) were more frequent in later onset cases.
Psychiatric symptoms have been frequently described in association with epilepsy (24). Pond (32) suggested that no less than 15-20% of epileptics exhibit minor psychiatric symptoms, whereas psychosis was reported in 1.4% to 8% (1, 19, 37) of patients attending to a neurological outpatient?s clinic. A more recent general practice?s survey (10) found that 48% of epileptic patients exhibit psychiatric symptoms. Minor psychiatric disorders, mostly anxiety and depression, were reported in 37.5% of patients; psychoses were noticed in 4.5% (9).
Ninety percent of our patients scored positively for at least one psychiatric symptom. These were severe enough to fulfil ICD-10 diagnostic criteria for mental disorder in 75.6% of cases. There are many possible explanations for this high prevalence of psychiatric symptoms in our sample:
Epilepsy in old age is clearly associated with a high vulnerability to psychiatric morbidity. Symptoms associated with anxiety and depression were the most prevalent in our study. Slowness and thinking difficulties have also been related to epilepsy (12). Irritability was another symptom frequently found among patients, although we could not find any correlation between the presence of the symptom, the type of epilepsy or the presence of brain lesion. In fact, two comprehensive reviews of the literature on irritability/violent behaviour and epilepsy concluded that this association is probably due to non-specific factors (25, 38). Psychotic symptoms were found in 17.2% of patients and were more likely to be present in those with early onset of seizures. This finding is in line with Trimble?s (42) suggestion that psychosis is more likely to be present in patients with onset of seizures during adolescence.
Our failure to demonstrate an association between psychiatric symptoms and structural abnormalities may imply that functional brain problems are more important than structural lesions. Flor-Henry (14), for example, argued that seizure variables were more influential in predicting the presence of psychiatric symptoms than the underlying organic damage. This was later confirmed by Fallhofer et al. (16) PET study of epileptic patients with psychosis, which showed that changes of brain function may be more obvious than changes of structure (5, 30).
The brain dysfunction associated with epilepsy of old age is apparent from the higher prevalence of psychiatric symptoms in this population. Our failure to demonstrate cognitive decline among our patients (excluding cases of dementia) might have resulted from the low sensitivity of the instruments used, as well as from the absence of a comparable control population. A thorough neuropsychological evaluation of this population may reveal specific cognitive impairments in association with psychiatric symptoms.
It is not yet clear whether elderly epileptics with onset in later life are psychiatrically different from those with onset in adolescence and childhood. The relative small numbers of our sample did not allow a direct comparison. Issues related to management and outcome of these patients also await further investigation.
In summary, we found a high prevalence of psychiatric morbidity among elderly epileptics. Symptoms associated with depression, anxiety and organic disorders were the most frequent. Complex partial seizures and organic brain damage were associated with psychiatric symptoms, although this association did not seem to be specific in psychopathological terms.
I am grateful to Dr. E.H. Reynolds for allowing me assess his patients and for his kind support and encouragement during the data collection period. I am also grateful to Professor R. Levy and Professor H. Förstl for helpful comments on an early draft of this paper. Dr. Osvaldo P.Almeida is partially supported by CNPq (Brazil).