The Epidemiology of the Comorbidity of Epilepsy in the General Population

Authors


Address correspondence and reprint requests to Dr. J.W. Sander at Institute of Neurology [Box 29], University College London, Queen Square, London WC1N 3BG, U.K. E-mail: lsander@ion.ucl.ac.uk

Abstract

Summary: Purpose: To describe the epidemiology of somatic and psychiatric conditions in adults with epilepsy in the community and compare it to that of people without epilepsy.

Methods: A cross-sectional population-based study extracting data from the UK General Practice Research Database for the period 1995–1998. Age- and sex-standardized prevalence rates were estimated for selected conditions and groups of conditions (categorized by ICD-9 chapters) in adults with epilepsy registered with primary care physicians. Results were compared with those in adults without epilepsy in the cohort, and prevalence ratios were calculated according to two broad age groups (16–64 and older than 64 years).

Results: Conditions common in the general population also were common in adults with epilepsy. Psychiatric disorders occurred twice as often, and the risk of somatic disorders was increased in people with epilepsy, with the exception of musculoskeletal and connective tissue disorders in older adults. The prevalence ratio of neoplasia, excluding intracranial tumors, was not increased in epilepsy. The prevalence ratio of brain tumors was particularly increased in young adults [prevalence ratio (PR), 70.7] and of meningiomas in older adults (PR, 91.9). Neurodegenerative conditions, particularly dementias and Alzheimer' disease (PR, 6.3 and 8, respectively) and Parkinson' disease (PR, 3.2), appeared more frequently in people with epilepsy. Upper gastrointestinal bleed occurred more frequently in epilepsy (PR, 4.3), as did cardio- and cerebrovascular disorders, fractures, pneumonia and chronic lung diseases, and diabetes. Eczema, osteoarthritis, and rheumatoid arthritis did not occur more frequently in epilepsy.

Conclusions: The prevalence ratio of many common psychiatric and somatic conditions is increased in adults with epilepsy who consult a primary care physician in the U.K. These findings may have implications in the diagnosis and management of epilepsy and coexisting conditions, as well as in health care provision.

Comorbidity refers to the co-occurrence of more than one condition in the same person; the term was originally used to indicate coexistent conditions in clinical trials (1). In people with epilepsy, other conditions may precede, co-occur with, or follow the diagnosis of epilepsy.

The epidemiology of comorbidity in epilepsy has not been well described. Data are limited on the cumulative incidence (2) and prevalence (3–8) of various comorbid conditions in adults with epilepsy in the community. One study compared rates of comorbid conditions in controls and adults with childhood-onset epilepsy and reported increased psychiatric, but not somatic, comorbidity in the latter (7). Case–control studies in the community have reported higher risks for various somatic disorders in epilepsy (2,9–14). Psychiatric disorders also appear to occur more frequently in people with epilepsy, although most studies have been conducted in highly selected groups of patients, mainly with temporal lobe or refractory epilepsy (15,16). Although case–control studies provide a measure of the association between epilepsy and other conditions, they do not provide comparable rates of these conditions in people with epilepsy and in the general population.

Information on comorbidity is important for planning health services (17). Disorders co-occurring with epilepsy influence the demand on health services by people with epilepsy, as we have shown recently (18). In the U.K., services for epilepsy in both primary and secondary care are often fragmented and do not adequately meet the needs of many people with epilepsy. This has been recognized, and now greater awareness exists of the need to improve health services for this group (19).

We conducted a cross-sectional population-based study to describe the prevalence and to evaluate the prevalence ratios of a number of somatic and psychiatric disorders in adults with epilepsy among a large cohort of people registered in general practices in the U.K.

METHODS

Data source

Data were extracted from the General Practice Research Database (GPRD) for the period January 1, 1995 to December 31, 1998. The GPRD is a health database established in 1987, which contains prospectively collected data from participating general practices in England and Wales. (General practices provide primary health care in the U.K. as part of the National Health System). The GPRD holds anonymous computerized patient records, including demographic information, information on significant consultations, details of all prescriptions issued, results of investigations, and hospital referrals and admissions. Participating general practices submit anonymized records at regular intervals, following agreed-on guidelines for recording clinical and prescribing data. Coding of diseases is based on Read and OXMIS systems. The completeness and accuracy of the data were previously documented (20–22).

The data for this study came from 211 general practices with a combined list size of 1.3 million patients, representing 2.6% of the population of England and Wales. The GPRD Scientific and Ethical Advisory Group approved the study.

Study population

Patients were included in the analysis if they were alive and permanently registered at the practice for the last 6 months of each analysis year from 1995 to 1998. This was to ensure better validity of data. We excluded children younger than 16 years from this study, as they tend to have different diseases from those of adults, and chronic diseases are rare in children. All Read Code and OXMIS diagnosis codes were converted to ICD-9 codes by using a table developed by the NHS Centre for Coding and Classification and the U.K. Office for National Statistics (23).

Case ascertainment was based on a clinical diagnosis of epilepsy recorded by the general practice, by using codes for the various epileptic syndromes and seizure types (ICD-9, 345). All those who received a new diagnosis of epilepsy and/or consulted for epilepsy during the study period were identified from Read and OXMIS codes and were used as the study group. The control group consisted of the remaining registered patients. The use of antiepileptic drugs (AEDs) was not taken as a surrogate for the diagnosis of epilepsy, as not all people with epilepsy take AEDs, and AEDs also can be indicated for conditions other than epilepsy.

Measurement of study variables

We identified the selected conditions in both groups; these conditions may have preceded or followed the development of epilepsy. The selection of the conditions was based on their importance in adult clinical practice, in terms of their morbidity in the general population or in people with epilepsy (Appendices 1 and 2). This selection also was informed by the analysis of the fourth national survey of morbidity in general practice (MSGP4) (18). The diagnosis of epilepsy and all other conditions was based on entries by the general practitioner, informed (where available) by information from specialists, investigations, and hospital admissions.

Table Appendix 1..  Conditions selected for analysis, with ICD-9 codings
Mental health 
 Neuroses300
 Obsessive–compulsive disorder300.3
 Anxiety300.0
 Hysteria300.1
 Depression311
 Schizophrenia295
 Organic psychoses291–294
 Other psychoses295–298
 Alcohol dependence303
 Dementia290
Somatic
 Fractures800–829
 Diabetes mellitus250
 Neoplasia(140–208)–(191–192)
 Brain neoplasms191
 Meningiomas (malignant and benign)192.1 + 225.2
 Ischemic heart disease410–414
 Heart failure428
 Congenital cardiac abnormalities745–746
 All cerebrovascular accidents430–438
 Hemorrhagic cerebrovascular accidents431
 Occlusive cerebrovascular accidents433–434
 Transient ischemic attacks435
 Migraine346
 Parkinson's disease332
 Pneumonias480–486
 Chronic bronchitis491
 Emphysema492
 Peptic ulcers531–534
 All gastrointestinal bleeds578
 Upper gastrointestinal bleed578.0
 Lower gastrointestinal bleed578.1
 Unspecified gastrointestinal bleed578.9
 Eczema691.8
 Rheumatoid arthritis714
 Osteoarthritis715
 Cerebral degeneration(290+331)–(331.3–.4)
 Alzheimer's disease331.0
Table Appendix 2..  Coding for ICD-9 chapters
ICD-9 CHAPTERSICD-9 Coding
  1. † Excluding epilepsy (345)

I. Infections and parasitic diseases001–139
II. Neoplasms140–239
III. Endocrine, nutritional, and metabolic diseases and immunity disorders240–279
IV. Diseases of blood and blood-forming organs280–289
V. Mental disorders290–319
VI. Disease of the nervous system and sense organs†320–389 (excluding 345)
VII. Diseases of the circulatory system390–459
VIII. Diseases of the respiratory system460–519
IX. Diseases of the digestive system520–579
X. Diseases of the genitourinary system580–629
XI. Complications of pregnancy, childbirth, and the puerperium630–676
XII. Diseases of the skin and subcutaneous tissue680–709
XIII. Diseases of the musculoskeletal system and connective tissue710–739
XIV. Congenital anomalies740–759
XV. Certain conditions originating in the perinatal period760–779
XVI. Symptoms, signs, and ill-defined conditions780–799
XVII. Injury and poisoning800–999

Outcome measures and statistical analysis

The period prevalence of epilepsy for 1995 through 1998 was calculated with a 95% confidence interval (CI; estimated for a single proportion). Age- and sex-specific prevalence rates of the selected conditions and of their diagnostic groups (ICD-9 chapters) were produced for both groups in two age groups (16–64 and 64+ years) in a cross-sectional design. We calculated age-standardized prevalence ratios to allow for differences in the age distribution between the groups. Prevalence ratios were calculated per 10-year age band for the younger age group (16–24, 25–34, to 55–64 years) and per 5-year age band for the older age group (65–69, 70–74, to 80–84, and 85+ years). A pooled summary estimate for each age group was produced by the Mantel–Haenszel method (24). Prevalence ratios were calculated for all subjects (male and female patients combined) for each age group, as well as for all individuals across all ages.

RESULTS

We identified 1,041,643 people (48.9% male and 51.1% female subjects) aged 16 years and older in the study, of whom 5,834 had epilepsy. The distribution is shown in Table 1. The overall prevalence of epilepsy in the study group was 5.6 per 1,000 (95% CI, 5.5–5.7).

Table 1. Number of people identified in each subgroup
 16–64 yr>64 yr
 EpilepsyNonepilepsyEpilepsyNonepilepsy
Male2,321420,31253386,130
Female2,338410,851642118,516

Prevalence of selected conditions and diagnostic groups in epilepsy

Individual conditions

The most common psychiatric conditions in epilepsy in adults were depression (18%), neuroses (15%) especially anxiety (11%), and psychoses (9%). Neuroses were more pronounced in younger women (20%). Nonorganic psychoses (see Box) and dementia were particularly increased in the older age group (15% and 12%, respectively). Schizophrenia affected <1% of patients. Overall, 41% of people with epilepsy received a diagnosis of a psychiatric disorder during the study period.

BOX

Organic Psychoses (291–294): Alcoholic psychoses, drug psychoses, transient organic psychotic conditions, chronic organic psychotic conditions

Other Psychoses (295–298): Schizophrenic psychoses, affective psychoses, paranoid states, other nonorganic psychoses

Dementia (290): senile dementia, presenile dementia, arteriosclerotic dementia

Cerebral Degeneration (290 + 331, excluding 331.3, 331.4): Senile and presenile organic conditions (senile dementia, simple type; presenile dementia; senile dementia, depressed or paranoid type; senile dementia with acute confusional state; arteriosclerotic dementia; other; unspecified)

Other cerebral degenerations (Alzheimer's disease, Pick disease, senile degeneration of brain, Creutzfeldt–Jakob disease, progressive multifocal encephalopathy, cerebral degeneration in other diseases classified elsewhere, other cerebral degeneration, unspecified)

Excludes communicating hydrocephalus and obstructive hydrocephalus

The most common somatic conditions in adults with epilepsy were fractures (10%), particularly in women older than 64 years (17%), and asthma (9%), particularly in younger women (11%). Migraine was common in younger women (8%). In the older group, the most common diagnoses in people with epilepsy were diabetes (9%), transient ischaemic attacks (18%), ischaemic heart disease (14%), heart failure (12%), neoplasia (7%), and osteoarthritis (12%). The most common neurologic disorders in this age group were brain degenerative diseases (14%) and Parkinson's disease (PD; 4%).

Conditions grouped by ICD-9 chapter

Disorders of the respiratory system (61%) and skin (46%), injury and poisoning (45%), infections (41%), and psychiatric disorders (40%) appear more often in all adults with epilepsy. Diseases of the respiratory (61%), circulatory (59%), nervous (59%), and musculoskeletal (59%) systems are diagnosed more frequently in those older than 64 years.

Prevalence ratios of selected conditions and diagnostic groups

Individual conditions

Among psychiatric conditions in epilepsy, organic psychoses, alcohol dependence, schizophrenia, and nonorganic psychoses have a 4 to 6 times higher prevalence in people with epilepsy than in the general population without epilepsy (Tables 2 and 3). Hysteria is diagnosed 4 times more often in the younger epilepsy group than in the nonepilepsy group, with men having a higher prevalence ratio than women, although the prevalence rate is higher in women. Interestingly, with the exception of obsessive–convulsive disorder and hysteria in the older group (which have small numbers), all individual psychiatric disorders studied have a significantly higher prevalence in epilepsy in both age groups (Tables 2 and 3).

Table 2. Rates of comorbid disorders in people aged 16 to 64 years
 MenWomenOverall Rate ratio (95% CI)
Epilepsy
(n = 2,321)
Nonepilepy
(n = 420,312)
Rate ratio
(95% CI)
Epilepsy
(n = 2,338)
Nonepilepsy
(n = 410,851)
Rate ratio
(95% CI)
 No. (%)No. (%) No. (%)No. (%) 
Mental health disorders
 Neuroses281 (12.1%)22,020 (5.2%)  2.36 (2.11–2.63) 471 (20.1%)45,355 (11%)  1.87 (1.72–2.02)2.03 (1.90–2.17)
  Obsessive–compulsive disorder 5 (0.2%) 427 (0.1%)2.08 (0.86–5.02)12 (0.5%) 675 (0.2%)3.10 (1.76–5.49)2.71 (1.68–4.37)
  Anxiety219 (9.4%) 15,829 (3.8%) 2.56 (2.25–2.90)333 (14.2%)30,902 (7.5%)  1.95 (1.77–2.16)2.16 (2.0–2.34) 
  Hysteria 6 (0.3%)  207 (0.05%) 5.36 (2.38–12.06)13 (0.6%) 639 (0.2%)3.72 (2.15–6.43)4.13 (2.62–6.50)
 Depression332 (14.3%)23,661 (5.6%) 2.59 (2.35–2.86)550 (23.5%)54,850 (13.3%) 1.79 (1.67–1.93)2.04 (1.92–2.16)
  Schizophrenia19 (0.8%)1,004 (0.2%)3.45 (2.19–5.42)14 (0.6%) 562 (0.1%)4.59 (2.70–7.78)3.83 (2.72–5.40)
 Organic psychoses79 (3.4%)2,120 (0.5%)6.67 (5.35–8.31)43 (1.8%)1,233 (0.3%) 6.26 (4.63–8.46)6.48 (5.42–7.74)
 Other psychoses107 (4.6%) 4,008 (1.0%)4.89 (4.05–5.89)116 (5.0%) 6,375 (1.6%) 3.26 (2.73–3.90)3.89 (3.42–4.43)
 Alcohol dependence94 (4.0%)2,960 (0.7%)5.94 (4.86–7.25)32 (1.4%)1,173 (0.3%) 5.01 (3.53–7.10) 5.64 (4.74–6.72)
 Dementia 7 (0.3%)  57 (0.01%) 24.30 (11.13–53.08) 9 (0.4%)   72 (0.02%) 26.40 (13.19–52.84) 25.22 (15.02–42.33)
 Somatic disorders
Cerebrovascular accident (CVA)70 (3.0%)1,187 (0.3%)11.65 (9.24–14.69)72 (3.1%) 807 (0.2%) 17.75 (14.04–22.43) 14.19 (12.04–16.73)
  Hemorrhagic CVA 9 (0.4%)   69 (0.02%) 24.63 (12.31–49.26) 5 (0.2%)   40 (0.01%)24.28 (9.50–62.04) 24.43 (13.99–42.66)
  Occlusive CVA 8 (0.3%)  210 (0.05%) 7.54 (3.73–15.22)12 (0.5%)  127 (0.03%) 18.74 (10.37–33.88)11.87 (7.57–18.61)
  Transient ischemic attack (TIA)35 (1.5%) 986 (0.2%)7.09 (5.11–9.84)38 (1.6%) 867 (0.2%) 8.89 (6.47–12.22)7.95 (6.33–9.98)
 Neoplasia27 (1.2%)3,152 (0.7%)1.68 (1.16–2.44)22 (0.9%)4,788 (1.2%)0.90 (0.60–1.38)1.22 (0.9–21.61)
  Brain neoplasms15 (0.6%)   52 (0.01%) 54.69 (30.84–96.98)19 (0.8%)    38 (0.01%)  90.36 (52.19–156.46)  70.70 (47.70–104.80)
  Meningiomas0  00      1 (0.04%)  8 (0%) 23.62 (2.84-196.24) 13.57 (1.75-105.20)
 Cerebral degeneration16 (0.7%)  102 (0.02%) 31.02 (18.38–52.36)11 (0.5%)    97 (0.02%) 23.80 (12.77–44.38) 27.45 (18.40–40.96)
 Alzheimer's disease 5 (0.2%) 18 (0%)  54.59 (20.40–146.04) 3 (0.1%)    23 (0.01%)27.92 (8.38–92.94) 39.76 (18.68–84.62)
 Parkinson's disease 2 (0.1%)  230 (0.05%)1.71 (0.43–6.88) 3 (0.1%)   167 (0.04%) 3.63 (1.16–11.35) 2.51 (1.041–6.06)
 Migraine84 (3.6%)6,745 (1.6%)2.22 (1.80–2.74)182 (7.8%) 22,116 (5.4%) 1.44 (1.25–1.65)1.63 (1.45–1.83)
 Ischemic heart disease63 (2.7%)8,085 (1.9%)1.56 (1.23–1.98)33 (1.4%)3,853 (0.9%)1.76 (1.26–2.46)1.63 (1.34–1.98)
 Heart failure14 (0.6%)1,122 (0.3%)2.49 (1.48–4.20)7 (0.3% 628 (0.2%)2.31 (1.01–4.85)2.44 (1.59–3.73)
 Congenital cardiac abnormalities10 (0.4%)  155 (0.04%)11.63 (6.14–22.02) 7 (0.3%)  183 (0.04%) 6.62 (3.11–14.07) 8.89 (5.47–14.46)
 Diabetes mellitus65 (2.8%)7,654 (1.8%)1.66 (1.31–2.10)58 (2.5%)5,601 (1.4%)2.00 (1.55–2.58)1.80 (1.52–2.15)
 Pneumonias27 (1.2%)1,555 (0.4%)3.25 (2.22–4.74)36 (1.5%)1,417 (0.3%)4.71 (3.40–6.55)3.94 (3.08–5.05)
 Asthma173 (7.5%) 24,129 (5.7%) 1.26 (1.10–1.46)265 (11.3%)31,481 (7.7%) 1.45 (1.29–1.62)1.37 (1.25–1.50)
 Chronic bronchitis 5 (0.2%) 768 (0.2%)1.29 (0.54–3.10)14 (0.6%) 884 (0.2%)3.18 (1.88–5.37)2.30 (1.46–3.60)
 Emphysema 4 (0.2%) 283 (0.1%)2.82 (1.05-7.55) 3 (0.1%)  203 (0.05%)3.00 (0.96-9.39)2.92 (1.38-6.14)
 Peptic ulcers31 (1.3%)3,157 (0.8%)1.88 (1.32–2.67)25 (1.1%)  1,759 (0.4%)2.73 (1.84–4.03)2.17 (1.67–2.81)
 Gastrointestinal (GI) bleed35 (1.5%)1697 (0.4%)3.80 (2.73-5.30)27 (1.2%)1085 (0.3%)4.56 (3.12-6.66)4.09 (3.19-5.25)
  Upper GI bleed24 (1.0%)1,001 (0.2%)4.34 (2.90–6.49)22 (0.9%) 628 (0.2%)6.25 (4.10–9.54)5.06 (3.78–6.77)
  Lower GI bleed 7 (0.3%) 544 (0.1%)2.45 (1.16–5.15) 5 (0.2%) 368 (0.1%)2.57 (1.07–6.21)2.50 (1.41–4.41)
  Unspecified GI bleed 5 (0.2%) 260 (0.1%)3.66 (1.51–8.86) 3 (0.1%)  134 (0.03%) 4.34 (1.38–13.63)3.88 (1.93–7.82)
 Rheumatoid arthritis 4 (0.2%) 762 (0.2%)1.03 (0.39–2.75)11 (0.5%)1,854 (0.5%)1.16 (0.64–2.10)1.12 (0.68–1.86)
 Osteoarthritis35 (1.5%)7,569 (1.8%)0.92 (0.66–1.27)63 (2.7%)9,949 (2.4%)1.29 (1.01–1.63)1.13 (0.93–1.36)
 Fractures235 (10.1%)20,317 (4.8%) 2.01 (1.78–2.23)166 (7.1%) 11,434 (2.8%) 2.61 (2.25–3.02)2.20 (2.00–2.42)
 Eczema 3 (0.1%) 520 (0.1%)1.01 (0.33–3.14) 6 (0.3%) 992 (0.2%)0.99 (0.44–2.20)1.00 (0.52–1.94)
Table 3. Rates of comorbid disorders in people aged 64 years and older
 MenWomenOverall Rate ratio (95% CI)
Epilepsy
(n = 533)
Nonepilepsy
(n = 86,130)
Rate ratio
(95% CI)
Epilepsy
(n = 642)
Nonepilepsy
(n = 118,516)
Rate ratio
(95% CI)
 No. (%)No. (%) No. (%)No. (%) 
Mental health disorders
 Neuroses51 (9.6%) 4,104 (4.8%)2.00 (1.54–2.60) 70 (10.9%)11,473 (9.7%)  1.14 (0.92–1.43)1.36 (1.15–1.61)
  Obsessive–compulsive disorder1 (0.2%)    38 (0.04%) 4.24 (0.58–30.79)0  91 (0.1%)01.36 (0.19–9.70)
  Anxiety40 (7.5%) 3,237 (3.8%)1.99 (1.47–2.69)58 (9.0%)9,291 (7.8%) 1.17 (0.91–1.49)1.37 (1.13–1.65)
  Hysteria1 (0.2%)    29 (0.03%) 5.54 (0.76–40.49) 1 (0.2%) 102 (0.1%) 1.83 (0.26–13.12) 2.67 (0.66–10.77)
 Depression73 (13.7%)4,867 (5.7%)2.39 (1.93–2.96)108 (16.8%)13,277 (11.2%)  1.5 (1.26–1.78)1.72 (1.50–1.97)
 Schizophrenia8 (1.5%)   85 (0.1%)15.39 (7.50–31.56) 1 (0.2%) 187 (0.2%)1.00 (0.14–7.21) 5.85 (3.02–11.32)
 Organic psychoses9 (1.7%)  242 (0.3%) 5.95 (3.07–11.50)10 (1.6%) 503 (0.4%)3.46 (1.86–6.43)4.23 (2.72–6.70)
 Other psychoses66 (12.4%)2,062 (2.4%)4.94 (3.94–6.21)105 (16.4%)4,704 (4.0%) 3.75 (3.15–4.46)4.10 (3.57–4.70)
 Alcohol dependence9 (1.7%)  333 (0.4%)4.47 (2.32–8.60) 8 (1.2%) 166 (0.1%) 9.22 (4.56–18.64)6.13 (3.80–9.90)
 Dementia52 (9.8%) 1,135 (1.3%)6.92 (5.32–9.00) 87 (13.6%)2,626 (2.2%) 5.38 (4.43–6.52) 5.83 (4.99– 6.81)
Somatic disorders
 Cerebrovascular accident (CVA)129 (24.2%) 3,547 (4.1%)5.69 (4.89–6.62)134 (20.9%)4,441 (3.7%) 5.15 (4.43–5.97)5.46 (4.91–6.07)
  Hemorrhagic CVA0   80 (0.1%)0 3 (0.5%) 94 (0.1%) 5.72 (1.82–17.98)2.92 (0.93–9.13)
  Occlusive CVA20 (3.8%)   484 (0.6%) 6.62 (4.27–10.27)12 (1.9%) 424 (0.4%)5.14 (2.91–9.08)6.09 (4.30–8.62)
  Transient ischemic attack89 (16.7%)3,361 (3.9%)4.15 (3.42–5.02)125 (19.5%)4,752 (4.0%) 4.52 (3.86–5.29)4.38 (3.88–4.94)
 Neoplasia40 (7.5%) 7,078 (8.2%)0.89 (0.66–1.20)39 (6.1%)6,839 (5.8%) 1.03 (0.76–1.40)0.97 (0.78–1.20)
  Brain neoplasms2 (0.4%)    16 (0.02%)20.87 (4.83–90.12)00011.55 (2.78–48.08)
  Meningiomas1 (0.2%)    3 (0.00%) 57.88 (5.73–584.21) 1 (0.2%)  1 (0.00%)  184.67 (12.44–2,740.43) 91.88 (16.70–505.47)
 Cerebral degeneration62 (11.6%)1316 (1.5%)7.13 (5.62–9.04)102 (15.9%)2,982 (2.5%) 5.59 (4.69–6.66)6.05 (5.25–6.97)
 Alzheimer's disease8 (1.5%) 233 (0.3%)5.23 (2.6–10.54)24 (3.7%)541 (0.5%) 7.50 (5.03–11.18)6.71 (4.74–9.50)
 Parkinson's disease25 (4.7%) 1,184 (1.4%)3.28 (2.23–4.82)22 (3.4%)1,215 (1.0%)  3.2 (2.12–4.84)3.29 (2.48–4.36)
 Migraine7 (1.3%) 551 (0.6%)2.10 (1.00–4.42) 9 (1.4%)1,784 (1.5%) 0.98 (0.51–1.88)1.23 (0.76–2.00)
 Ischemic heart disease87 (16.3%)12,005 (13.9%)1.16 (0.96–1.41) 78 (12.1%)11,551 (9.7%) 1.23 (0.99–1.51)1.21 (1.05–1.40)
 Heart failure55 (10.3%)6,015 (7.0%)1.41 (1.10–1.81) 81 (12.6%)7,885 (6.7%) 1.74 (1.42–2.13)1.60 (1.37–1.88)
 Diabetes mellitus46 (8.6%) 6,893 (8.0%)1.08 (0.82–1.42) 64 (10.0%)7,179 (6.1%) 1.66 (1.31–2.10)1.37 (1.15–1.64)
 Pneumonias31 (5.8%) 2,139 (2.5%)2.25 (1.59–3.17)50 (7.8%)2,596 (2.2%) 3.19 (2.46–4.15) 2.78 (2.26–3.43)
 Asthma36 (6.8%) 5,557 (6.5%)1.05 (0.76–1.44)39 (6.1%)7,882 (6.7%) 0.94 (0.69–1.28)0.99 (0.79–1.23)
 Chronic bronchitis16 (3.0%) 1,872 (2.2%)1.36 (0.84–2.20)13 (2.0%)1,707 (1.4%) 1.43 (0.83–2.45)1.41 (0.98–2.03)
 Emphysema2 (0.4%) 651 (0.8%)0.49 (0.12–1.96) 1 (0.2%) 332 (0.3%)0.58 (0.08–4.12)0.54 (0.17–1.67)
 Peptic ulcers20 (3.8%) 1,700 (2.0%)1.89 (1.23–2.92) 8 (1.2%)1,417 (1.2%) 1.04 (0.52–2.08)1.55 (1.08–2.25)
 Gastrointestinal (GI) bleed18 (3.4%) 1,228 (1.4%)2.30 (1.46–3.64)24 (3.7%)1,402 (1.2%) 2.96 (2.0–4.39) 2.67 (1.98–3.60)
  Upper GI bleed12 (2.3%)  529 (0.6%)3.53 (2.00–6.22)12 (1.9%) 654 (0.6%)3.16 (1.80–5.55)3.36 (2.25–5.01)
  Lower GI bleed6 (1.1%) 495 (0.6%)1.92 (0.86–4.27) 5 (0.8%) 495 (0.4%)1.77 (0.73–4.25)1.88 (1.04–3.40)
  Unspecified GI bleed1 (0.2%) 301 (0.3%)0.53 (0.07–3.73) 8 (1.2%) 355 (0.3%)3.89 (1.94–7.79)2.30 (1.19–4.43)
 Rheumatoid arthritis3 (0.6%) 542 (0.6%) 0.9 (0.29–2.79) 7 (1.1%)1,552 (1.3%) 0.86 (0.41–1.80)0.85 (0.46–1.57)
 Osteoarthritis53 (9.9%)  8706 (10.1%)0.98 (0.76–1.26) 90 (14.0%)17,017 (14.4%) 0.97 (0.80–1.18) 0.96 (0.82–1.12)
 Fractures54 (10.1%)2761 (3.2%)3.10 (2.40–4.01)111 (17.3%)10,398 (8.8%) 1.89 (1.59–2.24)2.10 (1.82–2.43)

Brain tumors and meningiomas occur 55 and 31 times respectively more often in people with epilepsy. The prevalence ratio of brain tumors is more pronounced in the younger adult epilepsy group than in the older group. The inverse picture is observed with meningiomas, for which the prevalence ratio is higher in the older epilepsy group (Tables 2 and 3). The prevalence ratio of Alzheimer's disease (AD) or stroke is higher in people with epilepsy, considerably so in the younger adult group (Tables 2 and 3). With the exception of osteoarthritis, rheumatoid arthritis, eczema, and emphysema, all other studied conditions occur more frequently in adults with epilepsy. The prevalence of migraine is 60% higher in adults with epilepsy but only 40% higher in younger women with epilepsy, in whom prevalence of the condition is high (Tables 2 and 3).

Conditions grouped by ICD-9 chapter

In the younger adult group, congenital anomalies are encountered nearly 3 times more frequently in those with epilepsy, especially in men (Table 5). Mental health (Chapter V) and blood disorders (Chapter IV) occur twice as often in people with epilepsy and, again, the risk is higher in male patients. Men with epilepsy have a 50% increased prevalence of neoplasia, but women do not. With the exception of conditions arising in the perinatal period (Chapter XV), diseases in all other chapters occur slightly more often in people with epilepsy.

Table 5. Rates of disease groups per gender and epilepsy status in people aged 16–64 years
 MenWomenOverall Rate ratio
(95% CI)
Epilepsy
(n = 2,321)
Nonepilepsy
(n = 420,312)
Rate ratio
(95% CI)
Epilepsy
(n = 2,338)
Nonepilepsy
(n = 410,851)
Rate ratio
(95% CI)
 No. (%)No. (%) No. (%)No. (%) 
ICD-9 Chapters
I. Infections and parasitic diseases783 (33.7%) 96,840 (23.0%)1.45 (1.37–1.53)1,157 (49.5%)148,663 (36.2%)1.34 (1.29–1.40)1.39 (1.34–1.44)
II. Neoplasms128 (5.5%) 15,541 (3.7%)1.52 (1.29–1.80) 173 (7.4%)29,477 (7.2%)1.06 (0.91–1.22)1.22 (1.09–1.36)
III. Endocrine, nutritional, and metabolic diseases and immunity disorders197 (8.5%) 26,493 (6.3%)1.45 (1.27–1.65)  341 (14.6%)32,283 (7.9%)1.98 (1.80–2.18)1.75 (1.62–1.89)
IV. Diseases of blood and blood-forming organs75 (3.2%) 4,993 (1.2%)2.75 (2.19–3.44)  250 (10.7%)27,283 (6.6%)1.58 (1.41–1.78)1.78 (1.60–1.97)
V. Mental disorders840 (36.2%) 57,612 (13.7%)2.68 (2.54–2.83)1,009 (43.2%) 97,878 (23.8%)1.84 (1.76–1.93)2.15 (2.08–2.23)
VI. Disease of the nervous system and sense organs905 (39.0%)110,608 (26.3%)1.50 (1.43–1.58)1,125 (48.1%)146,322 (35.6%)1.37 (1.31–1.43)1.43 (1.38–1.47)
VII. Diseases of the circulatory system416 (17.9%) 48,281 (11.5%)1.67 (1.54–1.81)  437 (18.7%) 56,934 (13.9%)1.45 (1.34–1.57)1.55 (1.47–1.65)
VIII. Diseases of the respiratory system1255 (54.1%) 186,419 (44.4%)1.21 (1.16–1.25)1,584 (67.8%)242,827 (59.1%)1.14 (1.11–1.17)1.17 (1.14–1.20)
IX. Diseases of the digestive system748 (32.2%) 79,881 (19.0%)1.73 (1.63–1.84)  926 (39.6%)10,4843 (25.5%)1.58 (1.50–1.66)1.65(1.58–1.71)
X. Diseases of the genitourinary system306 (13.2%)40,594 (9.7%)1.39 (1.25–1.54)1,286 (55.0%)189,918 (46.2%)1.19 (1.15–1.24)1.24 (1.19–1.29)
XI. Complications of pregnancy, childbirth, and the puerperium 661 (28.3%) 94,718 (23.1%)1.19 (1.11–1.27)1.33 (1.25–1.40)
XII. Diseases of the skin and subcutaneous tissue961 (41.4%)120,743 (28.7%)1.43 (1.36–1.50)1,134 (48.5%)156,394 (38.1%)1.26 (1.12–1.32)1.34 (1.30–1.38)
XIII. Diseases of the musculoskeletal system and connective tissue985 (42.4%)157,827 (37.5%)1.16 (1.10–1.21)1,203 (51.5%)182,490 (44.4%)1.20 (1.15–1.24) 1.18 (1.15–1.22)
XIV. Congenital anomalies35 (1.5%) 1,782 (0.4%)3.48 (2.50–4.85)  34 (1.5%) 2,432 (0.6%)2.37 (1.70–3.32)2.84 (2.24–3.60)
XV. Certain conditions originating in the perinatal period  1 (0.04%)   146 (0.03%)1.22 (0.17–8.72)  10 (0.4%) 1831 (0.4%)0.89 (0.48–1.66)0.93 (0.51–1.68)
XVI. Symptoms, signs, and ill-defined conditions1744 (75.1%) 181,805 (43.3%)1.75 (1.71–1.79)1961 (83.9%)247,043 (60.1%)1.40 (1.38–1.43)1.55 (1.53–1.58)
XVII. Injury and poisoning1095 (47.2%) 132,692 (31.6%)1.47 (1.41–1.53)1022 (43.7%)110,863 (27.0%)1.62 (1.54–1.69)1.53 (1.48–1.58)

In the older group, the prevalence ratio of a mental disorder in people with epilepsy is twice that of the background nonepilepsy population (Table 6). Similarly, the prevalence ratio of infections, blood and genitourinary disorders, and injuries and poisoning is ∼50% higher. The prevalence ratio of cardiovascular, nervous, and digestive disorders, although common in older adults with epilepsy, is only 20 to 40% higher. Musculoskeletal conditions do not occur more often in people with epilepsy in this age group (Table 6).

Table 6. Rates of disease groups per gender and epilepsy status in people aged 64 years and older
 MenWomenOverall rate ratio (95% CI)
Epilepsy
(n = 533)
Nonepilepsy
(n = 86,130)
Rate ratio
(95% CI)
Epilepsy
(n = 642)
Nonepilepsy
(n = 118,516)
Rate ratio
(95% CI)
 No. (%)No. (%) No. (%)No. (%) 
ICD-9 chapters
I. Infections and parasitic diseases192 (36.0%)20,940 (24.3%)1.47 (1.31–1.65)279 (43.5%)34,049 (28.7%)1.50 (1.38–1.64)1.48 (1.38–1.59)
II. Neoplasms 65 (12.2%) 9,595 (11.1%)1.08 (0.86–1.35) 78 (12.1%)10,772 (9.1%) 1.33 (1.08–1.64)1.21 (1.04–1.41)
III. Endocrine, nutritional and metabolic diseases and immunity disorders113 (21.2%)16,513 (19.2%)1.11 (0.94–1.31)193 (30.1)  21,579 (18.2%)1.67 (1.48–1.88) 1.41 (1.28–1.55)
IV. Diseases of blood and blood-forming organs 54 (10.1%)4,935 (5.7%)1.71 (1.33–2.21) 82 (12.8%)9,623 (8.1%)1.51 (1.23–1.84)1.57 (1.34–1.84)
V. Mental disorders218 (40.9%)12,974 (15.1%)2.69 (2.42–2.98)297 (46.3%)27,338 (23.1%)1.96 (1.81–2.14)2.19 (2.05–2.33)
VI. Disease of the nervous system and sense organs324 (60.8%)40,164 (46.6%)1.29 (1.21–1.38)373 (58.1%)58,219 (49.1%)1.17 (1.10–1.25)1.22 (1.17–1.28)
VII. Diseases of the circulatory system335 (62.9%)39,743 (46.1%)1.35 (1.26–1.44)393 (61.2%)54,863 (46.3%) .31 (1.23–1.39)1.33 (1.27–1.39)
VIII. Diseases of the respiratory system335 (62.9%)44,285 (51.4%)1.22 (1.14–1.30)391 (60.9%)63,663 (53.7%)1.13 (1.07–1.21)1.17 (1.12–1.22)
IX. Diseases of the digestive system285 (53.5%)33,204 (38.6%)1.37 (1.27–1.49)346 (53.9%)46,993 (39.7%)1.34 (1.25–1.44)1.36 (1.29–1.43)
X. Diseases of the genitourinary system203 (38.1)  21,275 (24.7%)1.52 (1.37–1.70)305 (47.5%)36,745 (31.0%)1.52 (1.40–1.65)1.51 (1.42–1.62)
XI. Complications of pregnancy, childbirth, and the puerperium138 (21.5%)19,382 (16.4%)1.33 (1.15–1.55) 1.23 (1.1–1.37)
XII. Diseases of the skin and subcutaneous tissue264 (49.5%) 33249 (38.6%)1.28 (1.17–1.39)341 (53.1%)50,071 (42.2%)1.25 (1.16–1.34)1.26 (1.19–1.33)
XIII. Diseases of the musculoskeletal system and connective tissue302 (56.7%)45,985 (53.4%)1.06 (0.98–1.14)390 (60.7%)73,074 (61.7%)0.99 (0.93–1.05)1.01 (0.96–1.06)
XIV. Congenital anomalies 2 (0.4%)  378 (0.4%)0.85 (0.21–3.41) 8 (1.2%) 564 (0.5%)2.64 (1.32–5.28)1.85 (0.99–3.44)
XV. Certain conditions originating in the perinatal period000 1 (0.2%)  80 (0.1%) 2.27 (0.32–16.22) 1.51 (0.21–10.82)
XVI. Symptoms, signs, and ill-defined conditions481 (90.2%)58,142 (67.5%)1.33 (1.29–1.37)580 (90.3%)86,137 (72.7%)1.24 (1.20–1.27)1.27 (1.25–1.30)
XVII. Injury and poisoning216 (40.5%)21,105 (24.5%)1.64 (1.48–1.82)319 (49.7%)38,921 (32.8%)1.49 (1.37–1.61)1.53 (1.44–1.63)

DISCUSSION

Principal findings

In this large population-based study, we describe the frequency of specific disorders, as well as groups of disorders, in a primary care cohort of adults with and without epilepsy, stratified by age and sex. As expected, conditions common in the general population also were common in adults with epilepsy. However, the prevalence ratio of most of the studied disorders was increased in adults with epilepsy; psychiatric disorders occurred twice as often. The prevalence ratio of somatic disorders across categories was increased in people with epilepsy, with the exception of musculoskeletal and connective tissue disorders in older adults. Congenital anomalies were increased in younger adults with epilepsy, but perinatally acquired conditions were not (probably related to poor survival in this group).

Strengths and weaknesses of the study

This study uses information from the GPRD, a large and well-validated general practice–derived database that has been used for epidemiologic research for >10 years (25). The population of participating practices was large: 1.3 million in 1998 (>1 million adults), and the age/sex distribution of this population was broadly similar to that of England and Wales in the same year. All practices passed quality tests on the data they supplied to the database. Selection criteria for the study population resulted in underrepresentation of more mobile (and possibly healthier) population groups and exclusion of people who died. Because prevalence is the result of both the condition's incidence and its mortality rate, factors influencing survival may lead to spurious associations between disorders. However, we consider it unlikely that the studied conditions affected survival differently in adults with and without epilepsy. An underrepresentation of healthy individuals in the study cohort also may result from the absence of young people without illness who are not registered with a general practitioner (GP). This would lead to overestimation of disease rates in the general population and, subsequently, to underestimation of PRs in epilepsy. Ninety-eight percent of the U.K. population, however, is registered with a GP (26), and the degree of any such underestimation of ratios would be small.

The calculated prevalence ratios are weighted averages of the age band–specific PR estimates. The stratum-specific PRs appear to be consistent across the 10-year age bands (although small numbers prohibit statistical testing). We cannot assess consistency of estimated ratios across different practices (to look for any regional or urban–rural variations), as the information on practice characteristics is not released for confidentiality purposes. However, the population is representative of the population of England and Wales.

The main limitation of this study is that, because of resource constraints, we did not validate the diagnosis, for example, by examining whether patients had their diagnosis confirmed by a specialist. Misdiagnosis is an important problem in the management of epilepsy, with as many as 20% of cases inaccurately diagnosed (27,28). Another limitation is that the study includes only people coming into contact with health services and would not include cases that had not come to primary care or specialist attention. People with prevalent epilepsy, the elderly, and people whose seizures are well controlled may not consult a doctor (29). This can introduce an admission-rate bias, with more severe cases with greater morbidity being included in the study. People with epilepsy are known to make higher use of health services than do people without epilepsy (18); however, we do not know what proportion of the health services requirement is due to the severity of epilepsy, its treatment, aetiology, or other factors.

Comparison with other studies

The period prevalence of epilepsy between 1995 and 1998 in our study was 5.6 per 1,000 (95% CI, 5.5–5.7/1,000). This is lower than the 1998 age-standardized prevalence reported previously by our group (7.4/1,000) by using the same GPRD practices (30). This discrepancy arose because of differences in case selection. In the previous study, the prevalence figure included people who had a diagnosis of epilepsy recorded several years before the study period (30). In the current analysis, we had no prior information available. Hence only diagnoses recorded during the study period would have been included, and the method used resulted in a lower prevalence of epilepsy.

Our study is the first of large scale in an unselected population. Its findings are consistent with those of previous community-based and unselected population case–control studies that suggest a higher risk of vascular disorders (strokes, myocardial infarction, peripheral vascular disease, hypercholesterolemia, left ventricular hypertrophy) (13), migraine (2), hypertension (11,14), dementia (12), brain tumors (31), fractures (9), and depression (31) in people with epilepsy. The size of the risk cannot be compared, however, as some of these studies were conducted in incident cases of epilepsy (11,12,14,31) and/or studied the prevalence of conditions before the development of epilepsy (11,12,31) or assessed the incidence of co-occurring conditions (2,9).

The higher risk of somatic disorders in our study contrasts with the results from a population-based cohort study, which reported increased psychiatric but not somatic comorbidity compared with that in controls (7). This study, however, was conducted in adults with childhood-onset epilepsy and a mean age of 35.6 years at the end of follow-up, who have different baseline characteristics from our younger adult epilepsy group. In addition, the sample size of only 220 people with epilepsy may not have been large enough to detect any real differences. A population-based prevalence study in 713 adults with epilepsy in Sweden reported a 5.9% prevalence of psychiatric disorders and a 50% prevalence of somatic disorders and disabilities (3). This compares with a 40% period prevalence of psychiatric disorders in our study and a much higher prevalence of somatic disorders.

Previous work by our group showed that a higher proportion of people with epilepsy consulted for neoplasms, haematologic and mental health disorders, diabetes, ischemic heart disease, heart failure, dementia, stroke, and gastrointestinal (GI) bleeding than did people without epilepsy in the community (18). Although these estimates were not age standardized, and rate ratios compared proportions rather than prevalence rates, results were very similar to those of this study, adding further support to our findings. It is noteworthy that in both studies, the prevalence of GI bleeding was significantly increased [risk ratio (RR), 4.93 in the MSGP4 study and PR 3.37 in this study]. In the GPRD, the PR is higher in the younger adult group and is attributed mainly to upper GI bleeding. A possible explanation is increased use of aspirin, nonsteroidal antiinflammatory drugs, or excess alcohol consumption by people with epilepsy, factors that contribute significantly to upper GI bleed (32). Aspirin is an important part of treatment in conditions such as ischemic heart disease (IHD), occlusive cerebrovascular accident (CVA), and transient ischemic attack (TIA), which occurred more frequently in people with epilepsy in this study (Tables 2–4). Alcohol dependence also is more common in this group (Tables 2–4). AEDs are not known to increase the risk of GI bleeding.

Table 4. Prevalence ratios of comorbid disorders in all individuals, irrespective of age or gender

All individuals
Rate ratio
(95% CI)
Mental health disorders
 Neuroses1.90 (1.79–2.02)
 Obsessive–compulsive disorder2.57 (1.61–4.10)
 Anxiety1.99 (1.85–2.14)
 Hysteria3.92 (2.55–6.04)
 Depression1.98 (1.87–2.09)
 Schizophrenia4.13 (3.05–5.61)
 Organic psychoses6.05 (5.13–7.14)
 Other psychoses3.98 (3.62–4.38)
 Alcohol dependence5.70 (4.84–6.71)
 Dementia6.34 (5.47–7.35)
Somatic disorders
 Cerebrovascular accident (CVA)6.96 (6.38–7.60)
 Hemorrhagic CVA10.62 (6.52–17.32)
 Occlusive CVA7.49 (5.69–9.86)
 Transient ischemic attack4.94 (4.44–5.50)
 Neoplasia1.05 (0.89–1.25)
 Brain neoplasms 55.05 (38.00–79.75)
 Meningiomas 31.44 (9.16–107.91)
 Cerebral degeneration6.80 (5.96–7.76)
 Alzheimer's disease 8.05 (5.89–11.00)
 Parkinson's disease3.19 (2.44–4.18)
 Migraine1.60 (1.43–1.80)
 Ischemic heart disease1.34 (1.19–1.50)
 Heart failure1.68 (1.45–1.95)
 Congenital cardiac abnormalities 7.34 (4.58–11.75)
 Diabetes mellitus1.57 (1.39–1.78)
 Pneumonias3.19 (2.72–3.74)
 Asthma1.30 (1.19–1.41)
 Chronic bronchitis1.67 (1.26–2.21)
 Emphysema1.25 (0.67–2.34)
 Peptic ulcers1.92 (1.55–2.37)
 Gastrointestinal (GI) bleed3.37 (2.78–4.08)
 Upper GI bleed4.31 (3.41–5.46)
 Lower GI bleed2.16 (1.43–3.25)
 Unspecified GI bleed2.85 (1.77–4.59)
 Rheumatoid arthritis0.99 (0.67–1.47)
 Osteoarthritis1.02 (0.91–1.15)
 Fractures2.17 (2.00–2.35)
 Eczema0.90 (0.47–1.74)

Implications for practice

Our results show higher comorbidity in adults with epilepsy than in those without. A comorbid condition can be the cause of epilepsy (e.g., strokes, brain tumors) or can be associated with epilepsy via common etiologic or shared risk factors (either genetic or environmental), as well as treatment (33). An example of a shared environmental factor is head injury, a risk factor for both epilepsy and migraine that may account for part of the relation between the disorders (33). Depression may share a common pathogenic mechanism with epilepsy that facilitates the occurrence of one in the presence of the other (34). Epilepsy and psychosis may arise out of cerebral dysfunction common to both, or psychosis may be a consequence of seizure activity (due to the effects of seizures on amygdala, hippocampal and septal areas) or AEDs (16,35). In addition, people with epilepsy may develop depression as a reaction to stresses in life imposed by the condition or may have an injury after a seizure.

The association of chronic bronchitis and emphysema with epilepsy can perhaps be explained through smoking. The commonest cause for these conditions is cigarette smoking (36), which also is a known risk factor for cerebrovascular disease and cancer (37), which can themselves cause epilepsy. A similar mechanism also may explain the association between epilepsy and asthma, as smoking is a risk factor for persistence of asthma into adulthood (38). Epilepsy is known to be more common in people from lower socioeconomic groups (39), and this also may contribute to the higher prevalence of comorbidity in people with epilepsy.

We found a considerably higher PR for AD in the younger adult epilepsy group than the older group (PR 40 vs. 7, respectively), with similar findings for dementia (PR ∼25 vs. ∼6) and cerebral degeneration (PR 27 vs. 6). Both a diagnosis of AD and a diagnosis of nonorganic dementia have been associated with a sixfold increased risk of unprovoked seizures, in the absence of other prior neurologic insult (12). The discrepancies between the age groups in our study may be related to the severity of the underlying condition associated with epilepsy and, therefore, reflect more aggressive early-onset disease. One study in patients with uncomplicated, definite AD on autopsy found that patients with new-onset, unprovoked seizures had a younger age at dementia onset than did the AD patients without seizures and that, at seizure onset, they had advanced dementia (40).

PD, a neurodegenerative condition, occurs about 3 times more commonly in people with epilepsy. The association between the two disorders is intriguing because they involve different parts of the CNS. The diagnosis of PD is made clinically; however, other disorders with prominent symptoms and signs of parkinsonism, such as drug-induced and arteriosclerotic parkinsonism, may be confused with PD until the diagnosis is confirmed at autopsy (41). This may particularly be the case in a primary care setting, such as the one used for this study. Cerebrovascular disease and dementia are risk factors for parkinsonism, and this may explain the apparent association between epilepsy and PD.

Neoplasms, in particular brain tumors, cerebrovascular and ischemic heart disease, and pneumonias are associated with increased mortality in people with epilepsy (42). The prevalence ratio of these conditions was found to be higher in our epilepsy cohort, which might explain the higher mortality observed, particularly in those with remote symptomatic epilepsy (42). Co-occurrence of conditions in a person also can complicate diagnosis, as symptom overlap might occur. For example, patients with epilepsy often have medically undiagnosed and untreated migraine (43), depression (15), or anxiety (16).

The generalization of our findings is potentially limited because of the case selection that possibly favored more severe cases and cases of symptomatic epilepsy being included in the cohort; people with epilepsy in remission and no comorbidity may not consult a doctor (29,44,45). Our results are, however, valid for adults with epilepsy who come to medical attention. In this setting, increased awareness is needed for the diagnosis and treatment of diseases co-occurring with epilepsy, as this may have profound consequences for the quality of life and mortality of patients, as well as for the burden on health services. Information on the comorbidity of epilepsy can be used in the design and organisation of health services for patients with the condition.

We were unable to look at the temporal association between epilepsy and comorbid disorders. For example, where epilepsy is associated with a brain tumor, we cannot be certain whether the brain tumor was the cause of epilepsy or whether epilepsy was present before the development of the tumor.

Conclusions

Our findings suggest that the risk of many common psychiatric and somatic disorders is increased in adults with epilepsy who consult a primary care physician. This implies that the presence of epilepsy should increase, not reduce, the suspicion that other disorders may be present. Conditions commonly seen in primary care also are common in people with epilepsy; however, physicians should also have a high degree of suspicion for other disorders. Our findings also emphasise the importance of taking a holistic view of the health of people with epilepsy. This requires treating not only the epilepsy but also any other present conditions, as well as ensuring that patients are given adequate advice on general aspects of their health.

Acknowledgments

Acknowledgment:  We thank the Morbidity and Health Care Team at the UK Office for National Statistics for their help. We also thank Dr. Gail Bell for her careful review of the manuscript and useful suggestions. A.M. is the holder of a National Primary Care Scientist Award, funded by the Department of Health. The UK National Society for Epilepsy funded A.G.

Ancillary