A bidirectional association between cognitive ability in young adulthood and epilepsy: a population-based cohort study

A bidirectional association between cognitive ability in young adulthood and epilepsy: a... Abstract Aim To investigate the bidirectional association between cognitive ability in young adulthood and epilepsy. Methods This cohort study included 1 159 076 men enrolled in the mandatory conscription board examination from the Danish Conscription Database (DCD; 658 465 men examined 1957–84), the Danish Defence Personnel Organization Database (DPOD; 216 509 men examined 1987–2005) and the Danish Conscription Registry (DCR; 284 102 men examined 2006-15). A supplementary analysis included 14 814 female volunteers. Cognitive ability was measured at conscription, and epilepsy was ascertained by physician diagnoses in the Danish National Patient Registries 1977–2016 [using International Classification of Diseases (ICD) codes: ICD-8-345; ICD-10-G40-G41]. Differences in cognitive ability in relation to epilepsy status at the time of conscription (age 19) were calculated using linear regression. The risk of epilepsy associated with cognitive ability was estimated using Cox regression models, split at age at follow-up (40 and 60 years) and adjusted for year of birth, cerebrovascular disease, traumatic brain injury and education. Results In all, 5097 (1.0%) men from the DPOD and DCR were diagnosed with epilepsy before conscription, and they had about 0.25 standard deviation (SD) lower cognitive scores than men without epilepsy. The largest difference in cognition was seen for those with the largest number of hospital contacts. A total of 22 364 (1.9%) men developed epilepsy, and cognitive ability was inversely associated with the risk of epilepsy. With the end of follow-up at age 40 years, the adjusted hazard ratio (HR)per SD increase was 0.75 (95% confidence interval = 0.73–0.77). The association attenuated with increasing age at diagnosis. The findings were replicated in female conscripts. Conclusions The cognitive impairment seen in adults with epilepsy seems to reflect combined effects of epileptic processes and lower premorbid cognitive ability. Cognitive ability, epilepsy, cohort study Key Messages Epilepsy diagnosed during childhood and adolescence was associated with about 0.25 SD lower cognitive score in young adulthood. The largest reductions in cognitive ability were seen for those with the largest number of hospital contacts. Cognitive ability in young adulthood was inversely associated with the risk of developing epilepsy later in adult life. Introduction Epilepsy, defined as recurrent unprovoked seizures, is estimated to affect more than 65 million people worldwide.1,2 The age-specific incidence tends to be high in early childhood and low between ages 20 and 50, and increases again after 50 years of age.3 Thus, the estimated yearly incidence among children and adolescents is 70 per 100 000, and in adulthood the yearly incidence is 30 per 100 000 in people aged 20 64 and 100 per 100 000 in older people.4 In most cases the aetiology of epilepsy remains unknown. The recognized risk factors include traumatic brain injuries (TBI), cerebrovascular diseases, brain infections, neurodegenerative diseases and various genetic syndromes.1 Low early-life cognitive ability might reflect brain development5 and is associated with increased risk of injuries,6 cardiovascular disease7 and dementia.8 To our knowledge, no longitudinal studies have investigated the association between early life cognitive ability and risk of incident epilepsy. Cognitive impairment is common in epilepsy, with up to 65% of all patients showing cognitive deficits. This has been observed in both child and adult onset epilepsy, and epilepsy has been suggested to accelerate the rate of cognitive decline.9 Thus, low cognitive ability in patients with epilepsy may be observed both because cognitive ability is a risk factor for developing the disorder and because of effects of the disease on cognitive function.9 In the present study, we examined whether cognitive ability in young adult men was associated with epilepsy before and after they were diagnosed with the disease. Based on the literature, we hypothesized that men diagnosed with epilepsy before age 18 would have lower cognitive ability at the conscription board examination than men without epilepsy. We also suggested that low cognitive ability in young adulthood would be associated with an increased risk of developing epilepsy later in adult life. Further, we explored whether any association between epilepsy before age 18 and cognitive ability in young adulthood varied with age of onset and number of hospital admissions. Methods Study population The study was based on conscription board data from three sources: the Danish Conscription Database (DCD),10 the Danish Defence Personnel Organization Database (DPOD) and the Danish Conscription Registry (DCR). The DCD includes conscription board examination data for 728 159 men born between 1929 and 1960 (99.8% born between 1939 and 1959) and examined from 1957 through 1984 (99.8% examined between 1958 and 1978). The DCD has information on selected variables, digitized for research purposes. The DPOD has recorded key data from conscription board examinations from 1995 through 2005 on 217 361 men considered fit for military service or limited duties. They were born between 1950 and 1988 (90.7% born between 1976 and 1986). The DCR was established in 2006 and has since then registered all data from Danish conscription board examinations. For the present study, we retrieved data for all conscription examinations between 2006 and 2015. This included 349 487 men born between 1973 and 1998 (97.1% born between 1986 and 1997). Table 1 gives an overview of the male populations covered in the three sources of conscription data. Of the 1 295 007 men, 127 680 (9.9%) had missing information on one or more key variables and were excluded from the analyses (Table 1). The DPOD only contained information on those who, based on available health information, were considered suitable for military service and had few missing data [n = 806 (0.4%)]. For the DCD and DCR, the percentage of men with missing data was higher [n = 60 547 (8.5%) and 65 327 (18.8%)]. The missing information was mainly due to lack of registry information for men who volunteered for military service before they turned 18 or who had been exempted from the conscription board examination due to medical conditions such as intellectual disability, epilepsy or type 1 diabetes. Thus, available for data analysis were 1 167 327 men. From 1962, women could volunteer for military service and from 2005 on, all women receive an invitation to conscription when they turn 18 years. However, few women attend the conscription board examination and only 14 814 women (examined from 1995 through 2015) were registered with complete data in the DPOD or DCR. The data from this relatively small and selected subsample of women were used in a supplementary analysis. The study was approved by the Danish Data Protection Agency. Table 1 Distribution of variables from conscription board examination in relation to data source Number (%) in all data sources Number (%) in Danish Conscription Database Number (%) in Danish Defence Personnel Organisation Database Number (%) in Danish Conscription Registry Number of men in data source 1 295 007 728 159 217 361 349 487 Proportion of Danish men alive at age 24a 76.0 80.4 54.6 87.2 Number with missing data 127 680 (9.9) 60 547 (8.5) 806 (0.4) 65 327 (18.8) Number died or emigrated before 1977 8251 (0.6) 8147 (1.1) 46 (0.0) 58 (0.0) Number included 1 159 076 658 465 216 509 284 102 Year of examination  1957–66 261 911 (22.6) 261 911 (39.8)  1967–76 323 513 (27.9) 323 513 (49.1)  1977–84 73 041 ( 6.3) 73 041 (11.9)  1995–96 48 205 (4.2) 48 205 (4.2)  1997–2006 193 266 (16.7) 168 304 (77.7) 24 912 (8.8)  2007–15 59 140 (22.4) 259 140 (91.2) Year of birth  1939–49 357 081 (30.8) 357 081 (54.2)  1950–59 300 600 (45.7) 300 600 (45.7)  1960–69 1225 (0.1) 784 (0.1) 441 (0.2)  1970–79 99 900 (8.6) 99 706 (46.1) 194 (0.1)  1980–89 206 679 (18.5) 116 362 (53.8) 90 317 (31.8)  1990–98 193 591 (16.7) 193 591 (68.1) Mean height (SD) 178.4 (6.9) 176.8 (6.6) 180.4 (6.6) 180.3 (6.8) Education  High 307 324 (26.5) 112 695 (17.1) 77 112 (35.7) 117 318 (41.3)  Middle 534 480 (46.1) 376 683 (57.2) 68 304 (31.6)) 897 493 (31.5)  Low 317 229 (27.4) 169 087 (25.7) 70 861 (32.7) 77 281 (27.2) Mean BPP (SD) 39.9 (11.1) 37.8 (12.0) 44.5 (8.3) 41.1 (9.6) Cerebrovascular disease before examination 285 (0.0) 3 (0.0) 99 (0.1) 183 (0.1) TBI diagnosis before examination 22 490 (1.9) 20 (0.0) 304 (0.1) 22 166 (7.8) Number (%) in all data sources Number (%) in Danish Conscription Database Number (%) in Danish Defence Personnel Organisation Database Number (%) in Danish Conscription Registry Number of men in data source 1 295 007 728 159 217 361 349 487 Proportion of Danish men alive at age 24a 76.0 80.4 54.6 87.2 Number with missing data 127 680 (9.9) 60 547 (8.5) 806 (0.4) 65 327 (18.8) Number died or emigrated before 1977 8251 (0.6) 8147 (1.1) 46 (0.0) 58 (0.0) Number included 1 159 076 658 465 216 509 284 102 Year of examination  1957–66 261 911 (22.6) 261 911 (39.8)  1967–76 323 513 (27.9) 323 513 (49.1)  1977–84 73 041 ( 6.3) 73 041 (11.9)  1995–96 48 205 (4.2) 48 205 (4.2)  1997–2006 193 266 (16.7) 168 304 (77.7) 24 912 (8.8)  2007–15 59 140 (22.4) 259 140 (91.2) Year of birth  1939–49 357 081 (30.8) 357 081 (54.2)  1950–59 300 600 (45.7) 300 600 (45.7)  1960–69 1225 (0.1) 784 (0.1) 441 (0.2)  1970–79 99 900 (8.6) 99 706 (46.1) 194 (0.1)  1980–89 206 679 (18.5) 116 362 (53.8) 90 317 (31.8)  1990–98 193 591 (16.7) 193 591 (68.1) Mean height (SD) 178.4 (6.9) 176.8 (6.6) 180.4 (6.6) 180.3 (6.8) Education  High 307 324 (26.5) 112 695 (17.1) 77 112 (35.7) 117 318 (41.3)  Middle 534 480 (46.1) 376 683 (57.2) 68 304 (31.6)) 897 493 (31.5)  Low 317 229 (27.4) 169 087 (25.7) 70 861 (32.7) 77 281 (27.2) Mean BPP (SD) 39.9 (11.1) 37.8 (12.0) 44.5 (8.3) 41.1 (9.6) Cerebrovascular disease before examination 285 (0.0) 3 (0.0) 99 (0.1) 183 (0.1) TBI diagnosis before examination 22 490 (1.9) 20 (0.0) 304 (0.1) 22 166 (7.8) a Data from statistics Denmark for the included birth cohorts alive age 24:total 1 704 127; DCD: n = 905 911 (1939-59); DPOD: n = 397 789 (1978-86); DCR: n = 400.427 (1986-97). Table 1 Distribution of variables from conscription board examination in relation to data source Number (%) in all data sources Number (%) in Danish Conscription Database Number (%) in Danish Defence Personnel Organisation Database Number (%) in Danish Conscription Registry Number of men in data source 1 295 007 728 159 217 361 349 487 Proportion of Danish men alive at age 24a 76.0 80.4 54.6 87.2 Number with missing data 127 680 (9.9) 60 547 (8.5) 806 (0.4) 65 327 (18.8) Number died or emigrated before 1977 8251 (0.6) 8147 (1.1) 46 (0.0) 58 (0.0) Number included 1 159 076 658 465 216 509 284 102 Year of examination  1957–66 261 911 (22.6) 261 911 (39.8)  1967–76 323 513 (27.9) 323 513 (49.1)  1977–84 73 041 ( 6.3) 73 041 (11.9)  1995–96 48 205 (4.2) 48 205 (4.2)  1997–2006 193 266 (16.7) 168 304 (77.7) 24 912 (8.8)  2007–15 59 140 (22.4) 259 140 (91.2) Year of birth  1939–49 357 081 (30.8) 357 081 (54.2)  1950–59 300 600 (45.7) 300 600 (45.7)  1960–69 1225 (0.1) 784 (0.1) 441 (0.2)  1970–79 99 900 (8.6) 99 706 (46.1) 194 (0.1)  1980–89 206 679 (18.5) 116 362 (53.8) 90 317 (31.8)  1990–98 193 591 (16.7) 193 591 (68.1) Mean height (SD) 178.4 (6.9) 176.8 (6.6) 180.4 (6.6) 180.3 (6.8) Education  High 307 324 (26.5) 112 695 (17.1) 77 112 (35.7) 117 318 (41.3)  Middle 534 480 (46.1) 376 683 (57.2) 68 304 (31.6)) 897 493 (31.5)  Low 317 229 (27.4) 169 087 (25.7) 70 861 (32.7) 77 281 (27.2) Mean BPP (SD) 39.9 (11.1) 37.8 (12.0) 44.5 (8.3) 41.1 (9.6) Cerebrovascular disease before examination 285 (0.0) 3 (0.0) 99 (0.1) 183 (0.1) TBI diagnosis before examination 22 490 (1.9) 20 (0.0) 304 (0.1) 22 166 (7.8) Number (%) in all data sources Number (%) in Danish Conscription Database Number (%) in Danish Defence Personnel Organisation Database Number (%) in Danish Conscription Registry Number of men in data source 1 295 007 728 159 217 361 349 487 Proportion of Danish men alive at age 24a 76.0 80.4 54.6 87.2 Number with missing data 127 680 (9.9) 60 547 (8.5) 806 (0.4) 65 327 (18.8) Number died or emigrated before 1977 8251 (0.6) 8147 (1.1) 46 (0.0) 58 (0.0) Number included 1 159 076 658 465 216 509 284 102 Year of examination  1957–66 261 911 (22.6) 261 911 (39.8)  1967–76 323 513 (27.9) 323 513 (49.1)  1977–84 73 041 ( 6.3) 73 041 (11.9)  1995–96 48 205 (4.2) 48 205 (4.2)  1997–2006 193 266 (16.7) 168 304 (77.7) 24 912 (8.8)  2007–15 59 140 (22.4) 259 140 (91.2) Year of birth  1939–49 357 081 (30.8) 357 081 (54.2)  1950–59 300 600 (45.7) 300 600 (45.7)  1960–69 1225 (0.1) 784 (0.1) 441 (0.2)  1970–79 99 900 (8.6) 99 706 (46.1) 194 (0.1)  1980–89 206 679 (18.5) 116 362 (53.8) 90 317 (31.8)  1990–98 193 591 (16.7) 193 591 (68.1) Mean height (SD) 178.4 (6.9) 176.8 (6.6) 180.4 (6.6) 180.3 (6.8) Education  High 307 324 (26.5) 112 695 (17.1) 77 112 (35.7) 117 318 (41.3)  Middle 534 480 (46.1) 376 683 (57.2) 68 304 (31.6)) 897 493 (31.5)  Low 317 229 (27.4) 169 087 (25.7) 70 861 (32.7) 77 281 (27.2) Mean BPP (SD) 39.9 (11.1) 37.8 (12.0) 44.5 (8.3) 41.1 (9.6) Cerebrovascular disease before examination 285 (0.0) 3 (0.0) 99 (0.1) 183 (0.1) TBI diagnosis before examination 22 490 (1.9) 20 (0.0) 304 (0.1) 22 166 (7.8) a Data from statistics Denmark for the included birth cohorts alive age 24:total 1 704 127; DCD: n = 905 911 (1939-59); DPOD: n = 397 789 (1978-86); DCR: n = 400.427 (1986-97). Cognitive measures Cognitive ability This was assessed by the conscription board intelligence test called the Børge Prien Prøve (BPP). The BPP consists of four subtests assessing logical, verbal, numerical and spatial ability and was recorded as a total score with a range of 0–78, representing the sum of the number of correct answers to the four subtests. This test has been used without changes since 1957,11 and the total BPP score correlates well with the full-scale Wechsler Adult Intelligence Scale (r = 0.82).12 For 260 658 men in DCR we also had information on basic school grade points from the 9th class (15 years of age) recorded in the School Grade Register. This Register was established in 2002 and uses the European Credit Transfer and Accumulation System grading scale: 3,0,2,4,7,10 and 12, ranging from fail marks (-3 poor or 0 inadequate) to the best (10 very good or 12 excellent). We used the highest grade obtained in Danish and/or mathematics. Epilepsy Hospital diagnosis of epilepsy In the present study, information on any admission to a Danish hospital from 1977 until 2016 was obtained by record linkage with the Danish National Patient Registry (DNPR), and the Danish Psychiatric Central Register (PCR) using the civil registration number as key. Diagnoses for all in-hospital discharges have been classified according to 8th Revision of the International Classification of Diseases (ICD-8) for the period 1977 to 1994, and according to the 10th Revision (ICD-10) from 1995 and onwards. From 1995, the DNPR also included all outpatient and emergency contacts. Death from epilepsy without previous hospitalization was identified by linkage with the Danish Cause of Death Register (n = 86). Incident cases were defined as first hospital discharge or death with main or contributory diagnosis of epilepsy (ICD-8-345.00-345.99 and ICD-10-G40.0-G41.9). For men diagnosed with epilepsy before the conscription board examination, we also calculated the number of readmissions from disease onset until the examination. The validity of epilepsy diagnosis obtained from the DNPR has been assessed in a study of 200 patients admitted between 1977 and 2002.13 This showed that 81% of epilepsy diagnoses were confirmed by review of medical records. Epilepsy diagnosed at conscription board examination The DCR also included diagnoses registered at the conscription board based on health certificates, subjective reports or the objective health examination. These diagnoses were coded following the ICD-10. Of the young men diagnosed with epilepsy at conscription, 524 (75.6%) also had an epilepsy diagnosis in the patient registries. Supplementary Table 1a and b (available as Supplementary data at IJE online) shows the number of epilepsy cases identified in the different registers in relation to follow-up and missing conscription data. This indicated that epilepsy was most frequent among men exempted from conscription board examination. Similarly, we found that the percentage of men diagnosed with epilepsy at the conscription board registration in the DCR was higher [2.6% (n = 1704)] in men with missing information as compared with those with data from a conscription board examination [0.2% (n = 704)]. Covariables Since reports suggest that cerebrovascular disease and severe TBI4,14 are potential confounders for associations between cognitive ability and epilepsy, we identified men who had been diagnosed with these conditions before the conscription board examination, using the relevant codes for classification of cerebrovascular disease (ICD-8-430-438 and ICD-10-I60-I69, G46;G45.9) or TBI (ICD-8-851-854 and ICD-10-S06.1-S06.9). From the conscription board data we also had information on educational level, which has been shown to be associated with cognitive ability early in life15 and epilepsy.1 Educational level reported at conscription was coded into three categories comprising low (7–9th grade), medium (vocational training or 10th-11th grade) and high educational level (12th grade or more advanced). Although conscription board examinations are highly standardized, there might be some variation across the seven districts. Information on the location of the examination was available in DCD and DPOD and was included in a supplementary analysis. Statistical analyses Association between epilepsy and cognitive ability The associations between diagnosed epilepsy before conscription board examinations and BPP scores were analysed using simple and multiple linear regression. Since information on epilepsy diagnosis was available from 1977, very few men (n = 30) from the DCD had been diagnosed with epilepsy before the examination. Thus, the analyses were restricted to the men from DPOD and DCR. In adjusted models, cerebrovascular disease, TBI and year of birth were included as covariables. In the DCR, we also examined how grade points from 9th class were related to epilepsy diagnosis. Finally, we analysed the influence of age at diagnosis on cognitive ability in men with epilepsy. Association between cognitive ability and incident epilepsy The association between cognitive ability in young adulthood and subsequent incident epilepsy was analysed using Cox proportional hazard regression with age as the underlying time scale. Person-years (py) of follow-up for an epilepsy diagnosis were accumulated from age on 1 January 1977 (age 19–39 years) or on the date of the conscription board examination, and follow-up was terminated at the age of first admission for epilepsy, death, emigration or end of follow–up (1 April 2016; age 19–78 years), whichever came first. We analysed all data separately for the DCD, DPOD and DCR because the men had been followed at different ages and because the registers varied in coverage before and after 1995. BPP was analysed as a continuous variable, since the association with the risk of epilepsy seemed to be linear. The initial analysis showed that hazard ratios (HRs) varied slightly with age at follow-up, and consequently data from the DCD was split on follow-up times 40 and 60 years. Men who had died (n = 4029) or emigrated (n = 4222) before 1977 were excluded from all analysis (Table 1). In addition, 11 266 men with epilepsy before study entry were not included in the time-to-event analyses. All analyses were carried out in STATA version 12. Results The men from the oldest birth years (DCD) had lower education and BPP score than the men from the DPOD and DCR representing the youngest birth years. The men from DPOD had a higher mean and smaller SD in BPP scores because the DPOD only consists of men evaluated as being suitable for military service. The lower rows in Table 1 provide the distribution of covariables in relation to data source (DCD, DPOD or DCR). Epilepsy and cognitive ability in young adulthood A total of 2453 (1.1%) and 3046 (1.0 %) men from the DPOD and DCR, respectively, had been registered with an epilepsy diagnosis in the DNPR before the conscription board examination. In the DCR, 702 (0.3%) men had also an epilepsy diagnosis registered at the conscription board examination. Table 2 shows that in both cohorts, men with an epilepsy diagnosis before conscription had about 0.25 SD lower mean cognitive scores than those without epilepsy. The number of hospital contacts with epilepsy was associated with cognitive ability in a dose-response-like manner. The association was attenuated when adjusting for year of birth, cerebrovascular disease and TBI. In the DCR, analyses were repeated with the epilepsy recorded at the conscription board examination, and this showed the same pattern of association. Grade points from 9th class correlated well with the BPP measured 4-5 years later (r = 0.48) and 2 489 men with epilepsy diagnosed before the basic school grading had lower mean gradings [5.57 (SD 2.47)] than the 175 men diagnosed after [5.86 (SD 2.61)] or the 257 994 men with no epilepsy diagnosis before conscription [5.97 (SD 2.56)]. Age at diagnosis was not associated with BPP in men with epilepsy (data not shown). Table 2 Mean cognitive score at conscription board examination in relation to epilepsy diagnosis before or at the examination in Danish men examined between 1986 and 2015 Number BPP score mean (SD) Beta-coefficienta (95% CI) Adjustedb beta-coefficient (95%CI) Danish Defence Personel Organization Database (n = 216 509) Epilepsy diagnosed in patient registry before conscription  No 214 056 44.5 (8.5)  Yes 2453 43.3 (8.3) −1.37 (−1.70 to −1.04) −1.38 (−1.71 to −1.05) Number of hospital contacts with epilepsy  No 214 056 44.5 (8.3) 1.24 (0.90 to 1.59) 1.27 (1.05 to 1.59)  1–4 2303 43.4 (8.6) 0 0  5–22 150 41.1 (8.4) −2.13 (−3.50 to −0.75) −2.19 (−3.52 to −0.81) Danish Conscription Register (n = 284 102) Epilepsy diagnosed in patient registry before conscription  No 281 056 41.1 (10.2)  Yes 3 046 39.1 (9.5) −2.03 (−2.37 to −1.68) −1.95 (−2.21 to −1.60) Number of hospital contacts with epilepsy  No 281 124 41.1 (9.5) 1.44 (1.05 to 1.83) 1.42 (1.03-1.81)  1–4 2 343 39.7 (9.9) 0 0  5–46 635 37.1 (11.0) −2.58 (−3.42 to −1.74) −2.46 (−3.30 to −1.28) Epilepsy diagnosed at conscription  No 283 400 41.1 (9.5)  Yes 702 40.1 (10.2) −1.03 (−1.73 to −0.32) −1.00 (−1.69 to −0.28) Number BPP score mean (SD) Beta-coefficienta (95% CI) Adjustedb beta-coefficient (95%CI) Danish Defence Personel Organization Database (n = 216 509) Epilepsy diagnosed in patient registry before conscription  No 214 056 44.5 (8.5)  Yes 2453 43.3 (8.3) −1.37 (−1.70 to −1.04) −1.38 (−1.71 to −1.05) Number of hospital contacts with epilepsy  No 214 056 44.5 (8.3) 1.24 (0.90 to 1.59) 1.27 (1.05 to 1.59)  1–4 2303 43.4 (8.6) 0 0  5–22 150 41.1 (8.4) −2.13 (−3.50 to −0.75) −2.19 (−3.52 to −0.81) Danish Conscription Register (n = 284 102) Epilepsy diagnosed in patient registry before conscription  No 281 056 41.1 (10.2)  Yes 3 046 39.1 (9.5) −2.03 (−2.37 to −1.68) −1.95 (−2.21 to −1.60) Number of hospital contacts with epilepsy  No 281 124 41.1 (9.5) 1.44 (1.05 to 1.83) 1.42 (1.03-1.81)  1–4 2 343 39.7 (9.9) 0 0  5–46 635 37.1 (11.0) −2.58 (−3.42 to −1.74) −2.46 (−3.30 to −1.28) Epilepsy diagnosed at conscription  No 283 400 41.1 (9.5)  Yes 702 40.1 (10.2) −1.03 (−1.73 to −0.32) −1.00 (−1.69 to −0.28) a Linear regression with cognitive scores as outcome. b Adjusted for birth year, cerebrovascular disease and traumatic brain injury at conscription (DPOD also for conscription board district). Table 2 Mean cognitive score at conscription board examination in relation to epilepsy diagnosis before or at the examination in Danish men examined between 1986 and 2015 Number BPP score mean (SD) Beta-coefficienta (95% CI) Adjustedb beta-coefficient (95%CI) Danish Defence Personel Organization Database (n = 216 509) Epilepsy diagnosed in patient registry before conscription  No 214 056 44.5 (8.5)  Yes 2453 43.3 (8.3) −1.37 (−1.70 to −1.04) −1.38 (−1.71 to −1.05) Number of hospital contacts with epilepsy  No 214 056 44.5 (8.3) 1.24 (0.90 to 1.59) 1.27 (1.05 to 1.59)  1–4 2303 43.4 (8.6) 0 0  5–22 150 41.1 (8.4) −2.13 (−3.50 to −0.75) −2.19 (−3.52 to −0.81) Danish Conscription Register (n = 284 102) Epilepsy diagnosed in patient registry before conscription  No 281 056 41.1 (10.2)  Yes 3 046 39.1 (9.5) −2.03 (−2.37 to −1.68) −1.95 (−2.21 to −1.60) Number of hospital contacts with epilepsy  No 281 124 41.1 (9.5) 1.44 (1.05 to 1.83) 1.42 (1.03-1.81)  1–4 2 343 39.7 (9.9) 0 0  5–46 635 37.1 (11.0) −2.58 (−3.42 to −1.74) −2.46 (−3.30 to −1.28) Epilepsy diagnosed at conscription  No 283 400 41.1 (9.5)  Yes 702 40.1 (10.2) −1.03 (−1.73 to −0.32) −1.00 (−1.69 to −0.28) Number BPP score mean (SD) Beta-coefficienta (95% CI) Adjustedb beta-coefficient (95%CI) Danish Defence Personel Organization Database (n = 216 509) Epilepsy diagnosed in patient registry before conscription  No 214 056 44.5 (8.5)  Yes 2453 43.3 (8.3) −1.37 (−1.70 to −1.04) −1.38 (−1.71 to −1.05) Number of hospital contacts with epilepsy  No 214 056 44.5 (8.3) 1.24 (0.90 to 1.59) 1.27 (1.05 to 1.59)  1–4 2303 43.4 (8.6) 0 0  5–22 150 41.1 (8.4) −2.13 (−3.50 to −0.75) −2.19 (−3.52 to −0.81) Danish Conscription Register (n = 284 102) Epilepsy diagnosed in patient registry before conscription  No 281 056 41.1 (10.2)  Yes 3 046 39.1 (9.5) −2.03 (−2.37 to −1.68) −1.95 (−2.21 to −1.60) Number of hospital contacts with epilepsy  No 281 124 41.1 (9.5) 1.44 (1.05 to 1.83) 1.42 (1.03-1.81)  1–4 2 343 39.7 (9.9) 0 0  5–46 635 37.1 (11.0) −2.58 (−3.42 to −1.74) −2.46 (−3.30 to −1.28) Epilepsy diagnosed at conscription  No 283 400 41.1 (9.5)  Yes 702 40.1 (10.2) −1.03 (−1.73 to −0.32) −1.00 (−1.69 to −0.28) a Linear regression with cognitive scores as outcome. b Adjusted for birth year, cerebrovascular disease and traumatic brain injury at conscription (DPOD also for conscription board district). Cognitive ability in young adulthood and risk of epilepsy The 658 435, 214 056 and 281 056 men from the DCD, DPOD and DCR, respectively, without epilepsy at conscription were followed for epilepsy in the DNPR between 1977 and 2016 (medians of 45.1, 16.1 and 5.0 years in the DCD, DPOD and DCR, respectively). During that time, respectively 18 985 (2.9%), 1233 (0.6%) and 2146 (0.8%) developed epilepsy with corresponding incidence rates of 6.5 [95% confidence interval (CI) = 6.4–6.6], 3.7 (95% CI = 3.5–3.9) and 14.5 (95% CI = 13.9–15.1) per 10 000 py. The incidence increased with age at follow-up (see Supplementary Figure 1, available as Supplementary data at IJE online). Table 3 provides the crude and adjusted HRs for the association between cognitive ability in young adulthood and epilepsy in relation to age at follow-up in the three conscription cohorts. It shows that in all three cohorts combined, cognitive ability in young adulthood was inversely associated with risk of epilepsy before age 40 years (HRper SD increase in BPP 0.75, 95% CI = 0.73–0.77) after adjustment for cerebrovascular disease, TBI and educational level. In the DCD, the HRs also seemed to attenuate with increasing age at diagnosis. Further adjustment for location by district did not influence the HRs for the DCD or DPOD. The supplementary analyses for the 14 814 women indicated a similar bidirectional association between young adult cognitive ability and epilepsy as observed for men (Table 4). Table 3 The hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between cognitive ability (BPP) in young adulthood and epilepsy in men from the Danish Conscription Database (DCD), Danish Defence Personnel Organization Database (DPOD) and the Danish Conscription Register (DCR) Age at diagnosis or end of follow-up Number of outcomes HR (95%CI)a HRb (95% CI) DCD (n = 658 435 men) ≤ 40 4342 0.74 (0.71–0.75) 0.76 (0.73–0.79) 41–60 9406 0.84 (0.82–0.85) 0.86 (0.84–0.88) > 60 5236 0.95 (0.90–1.01) 0.92 (0.86–0.99) DPOD (n = 214 056 men) ≤40 1218 0.74 (0.68–0.79) 0.82 (0.74–0.89) DCR (n = 281 056 men) ≤ 40 2140 0.74 (0.72–78) 0.80 (0.75–0.84) All men ( 1 153 547) < 40 7700 0.81 (0.79–0.83) 0.75 (0.73–0.77) Age at diagnosis or end of follow-up Number of outcomes HR (95%CI)a HRb (95% CI) DCD (n = 658 435 men) ≤ 40 4342 0.74 (0.71–0.75) 0.76 (0.73–0.79) 41–60 9406 0.84 (0.82–0.85) 0.86 (0.84–0.88) > 60 5236 0.95 (0.90–1.01) 0.92 (0.86–0.99) DPOD (n = 214 056 men) ≤40 1218 0.74 (0.68–0.79) 0.82 (0.74–0.89) DCR (n = 281 056 men) ≤ 40 2140 0.74 (0.72–78) 0.80 (0.75–0.84) All men ( 1 153 547) < 40 7700 0.81 (0.79–0.83) 0.75 (0.73–0.77) a HR per SD increase in BPP. b Adjusted for cerebrovascular disease, traumatic brain injury and educational level at conscription. Table 3 The hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between cognitive ability (BPP) in young adulthood and epilepsy in men from the Danish Conscription Database (DCD), Danish Defence Personnel Organization Database (DPOD) and the Danish Conscription Register (DCR) Age at diagnosis or end of follow-up Number of outcomes HR (95%CI)a HRb (95% CI) DCD (n = 658 435 men) ≤ 40 4342 0.74 (0.71–0.75) 0.76 (0.73–0.79) 41–60 9406 0.84 (0.82–0.85) 0.86 (0.84–0.88) > 60 5236 0.95 (0.90–1.01) 0.92 (0.86–0.99) DPOD (n = 214 056 men) ≤40 1218 0.74 (0.68–0.79) 0.82 (0.74–0.89) DCR (n = 281 056 men) ≤ 40 2140 0.74 (0.72–78) 0.80 (0.75–0.84) All men ( 1 153 547) < 40 7700 0.81 (0.79–0.83) 0.75 (0.73–0.77) Age at diagnosis or end of follow-up Number of outcomes HR (95%CI)a HRb (95% CI) DCD (n = 658 435 men) ≤ 40 4342 0.74 (0.71–0.75) 0.76 (0.73–0.79) 41–60 9406 0.84 (0.82–0.85) 0.86 (0.84–0.88) > 60 5236 0.95 (0.90–1.01) 0.92 (0.86–0.99) DPOD (n = 214 056 men) ≤40 1218 0.74 (0.68–0.79) 0.82 (0.74–0.89) DCR (n = 281 056 men) ≤ 40 2140 0.74 (0.72–78) 0.80 (0.75–0.84) All men ( 1 153 547) < 40 7700 0.81 (0.79–0.83) 0.75 (0.73–0.77) a HR per SD increase in BPP. b Adjusted for cerebrovascular disease, traumatic brain injury and educational level at conscription. Table 4 The distribution of variables and the associations between cognitive ability and epilepsy before and after conscription examination in 14 814 female volunteers Distribution of variables (%) Conscription register  Danish Defence Personel Organization Data 2 427 (16.4)  Danish Conscription Register 12 387 (83.6) Birth cohort  −1979 1 054 (7.3)  1980–89 5 445 (36.8)  1990–98 8 293 (56.0) Educational level  High 8 733 (59.0)  Middle 3 930 (26.5)  Low 2 151 (14.5) Mean BPP (SD) 41.9 (8.4) Cerebrovascular disease before examination 16 (1.1) TBI before examination 969 (6.5) Epilepsy diagnosis  Before examination 161 (1.1)  After examination 145 (1.0) Epilepsy and mean cognitive ability Epilepsy diagnosis  No 42.0 (8.3)  Yes, before 39.3 (8.7) Cognitive ability and epilepsy risk (HR and 95% CI)  BPP per SD increase 0.64 (0.52–0.80)  BPPaper SD increase 0.69 (0.59–0.87) Distribution of variables (%) Conscription register  Danish Defence Personel Organization Data 2 427 (16.4)  Danish Conscription Register 12 387 (83.6) Birth cohort  −1979 1 054 (7.3)  1980–89 5 445 (36.8)  1990–98 8 293 (56.0) Educational level  High 8 733 (59.0)  Middle 3 930 (26.5)  Low 2 151 (14.5) Mean BPP (SD) 41.9 (8.4) Cerebrovascular disease before examination 16 (1.1) TBI before examination 969 (6.5) Epilepsy diagnosis  Before examination 161 (1.1)  After examination 145 (1.0) Epilepsy and mean cognitive ability Epilepsy diagnosis  No 42.0 (8.3)  Yes, before 39.3 (8.7) Cognitive ability and epilepsy risk (HR and 95% CI)  BPP per SD increase 0.64 (0.52–0.80)  BPPaper SD increase 0.69 (0.59–0.87) a Adjusted for cerebrovascular disease, traumatic brain injury (TBI) and educational level at conscription. Table 4 The distribution of variables and the associations between cognitive ability and epilepsy before and after conscription examination in 14 814 female volunteers Distribution of variables (%) Conscription register  Danish Defence Personel Organization Data 2 427 (16.4)  Danish Conscription Register 12 387 (83.6) Birth cohort  −1979 1 054 (7.3)  1980–89 5 445 (36.8)  1990–98 8 293 (56.0) Educational level  High 8 733 (59.0)  Middle 3 930 (26.5)  Low 2 151 (14.5) Mean BPP (SD) 41.9 (8.4) Cerebrovascular disease before examination 16 (1.1) TBI before examination 969 (6.5) Epilepsy diagnosis  Before examination 161 (1.1)  After examination 145 (1.0) Epilepsy and mean cognitive ability Epilepsy diagnosis  No 42.0 (8.3)  Yes, before 39.3 (8.7) Cognitive ability and epilepsy risk (HR and 95% CI)  BPP per SD increase 0.64 (0.52–0.80)  BPPaper SD increase 0.69 (0.59–0.87) Distribution of variables (%) Conscription register  Danish Defence Personel Organization Data 2 427 (16.4)  Danish Conscription Register 12 387 (83.6) Birth cohort  −1979 1 054 (7.3)  1980–89 5 445 (36.8)  1990–98 8 293 (56.0) Educational level  High 8 733 (59.0)  Middle 3 930 (26.5)  Low 2 151 (14.5) Mean BPP (SD) 41.9 (8.4) Cerebrovascular disease before examination 16 (1.1) TBI before examination 969 (6.5) Epilepsy diagnosis  Before examination 161 (1.1)  After examination 145 (1.0) Epilepsy and mean cognitive ability Epilepsy diagnosis  No 42.0 (8.3)  Yes, before 39.3 (8.7) Cognitive ability and epilepsy risk (HR and 95% CI)  BPP per SD increase 0.64 (0.52–0.80)  BPPaper SD increase 0.69 (0.59–0.87) a Adjusted for cerebrovascular disease, traumatic brain injury (TBI) and educational level at conscription. Discussion In this nation-wide cohort of 1 153 076 Danish male conscripts we found, corroborating our previous hypothesis, that epilepsy diagnosed during childhood and adolescence was associated with about 0.25 SD lower cognitive ability in young adulthood. The largest differences were seen for men with the largest number of hospital contacts. The associations persisted after adjustment for year of birth, cerebrovascular disease and TBI. The study also showed that cognitive ability in young adulthood was inversely associated with the risk of developing epilepsy when cerebrovascular disease, TBI and educational level were adjusted for. This was also in accordance with our hypothesis, and the findings also were replicated in a smaller cohort of female conscripts volunteering for military service. The proportion of men who developed epilepsy before age 40 years (0.7%) corresponded to the estimates for men in Sweden8 and Europe4 of 0.6%. In the present study, the proportion of epilepsy was higher in the DCR than in the DPOD and DCD. This might reflect that the DPOD represented men suitable for military employment, whereas the follow-up of most of the men in the DCD was from after the age of 18, and thus follow-up during young adulthood was based on in-hospital contacts only. However, these differences did not seem to influence the associations between variables estimated within the three cohorts, as they were rather similar. A few smaller follow-up studies of children with epilepsy (n < 400) have suggested that higher seizure frequency is associated with cognitive impairments or poor educational achievements.16–18 These studies did not compare the cognitive outcomes with children who grew up without epilepsy. Unfortunately, we did not have repeated measures of cognitive ability in our cohort and do not know whether the lower cognitive ability had been present before disease onset or had in fact decreased due to the effect of epileptic seizures or treatment. However, the supplementary analysis with school grades obtained 4-5 years before conscription indicated that those with onset after the 9th class examination had higher grades than those diagnosed before. On the other hand, it is also possible that this relatively small number of events emerging during adolescence represent milder cases. Thus, our analysis of the relation between cognitive ability and epilepsy indicated that lower cognitive ability might be present before epilepsy, as a potential marker of brain dysfunction acquired early in life. The influence of cognitive ability in young adulthood on later epilepsy could be explained by confounding from genetic predisposition, familial factors and poor health during early childhood or influences mediated through cardiovascular diseases, depression and injuries.1,6,7,19 In our study, educational level and TBI were not important confounders. Furthermore, information on cognitive ability was not available for those exempted from conscription due to severe somatic, neurodegenerative and psychiatric disorders. This suggests that the reported associations are not due to confounding from severe mental disorders. The strength of the present analysis is that it was based on a large sample and that it covers nearly 60 birth years (1939–99). In addition, cognitive ability was assessed in young adulthood when it can be expected to be relatively stable. Outcome information was retrieved from national registers and we were able to include all hospitalizations for epilepsy from 1977 with sensitivity measures above 80%. However, for men in the DCD born between 1939 and 1959, this information was first available at age 19–39 years. Unfortunately, most of the diagnoses were based on non-specific ICD codes and it was not possible to look specifically at epilepsy with focal or generalized seizures. Further, our study did not include those exempted from conscription due to epilepsy, and our analysis of non-participants showed that subjects with missing data had higher risk of epilepsy than those available for the present study. This could suggest that our estimates of the relation between epilepsy and cognitive ability may not apply to these potentially more severe cases. In the DCD, study participants were followed for a median of 45 years. During this time, ICD coding has changed from ICD-8 to ICD-10, and the coverage of the hospital registers has increased due to the inclusion of outpatient and emergency contacts from 1995. However, we did not find any major difference in the risk estimates for the association between cognitive ability and epilepsy across several birth years. Our outcome measure may not include the presumed milder cases of epilepsy diagnosed and treated only in primary care in Denmark. We did consider using information on purchase of anti-epileptic medication (AED) from the Danish Prescription register. However, it is not recommended to identify patients with epilepsy through filled prescriptions for AED20 as it would lead to misclassification, since approximately 50% of all anti-epileptic drugs are prescribed for other purposes than epilepsy.21 Data from the DCR indicated that around 25% of young men with epilepsy would not be registered in the DNPR. Our main results were based on a cohort of men, but a supplementary analysis on a selected subsample of female conscripts indicated that the findings could be generalizable to women. In conclusion, hospital contacts for epilepsy were associated with decreased cognitive ability in young adulthood in a dose-response-dependent manner. Low cognitive ability in young adulthood was also associated with higher risk of developing epilepsy later in life. Thus, the cognitive impairment seen in adult patients with epilepsy may reflect combined effects of epileptic processes and lower premorbid cognitive ability, which might be a marker of brain dysfunction inherited or acquired early in life. Funding The work was supported by the Danish Medical Research Council (grant number 09-063599 and 09-069151) and the Velux Foundation (grant number 31205), the Jascha Foundation (grant 6200) and Doctor Sofus Carl Emil Friis and Olga Doris Friis grant. Acknowledgement We thank Esther Mortensen, MA in English, for linguistic revision. Conflict of interest: None declared. References 1 Abramovici S , Bagic A. Epidemiology of epilepsy . Handb Clin Neurol 2016 ; 138 : 159 – 71 . Google Scholar Crossref Search ADS PubMed 2 Singh A , Trevick S. The epidemiology of global epilepsy. Neurol Clin 2016 ; 34 : 837 – 847 . Google Scholar Crossref Search ADS PubMed 3 Christensen J , Vestergaard M , Pedersen MG , Pedersen CB , Olsen J , Sidenius P. Incidence and prevalence of epilepsy in Denmark . Epilepsy Res 2007 ; 76 : 60 – 65 . Google Scholar Crossref Search ADS PubMed 4 Forsgren L , Beghi E , Oun A , Silanpaa M. The epidemiology of epilepsy in Europe - a systematic review . Eur J Neurol 2005 ; 12 : 245 – 53 . Google Scholar Crossref Search ADS PubMed 5 Richards M , Deary IJ. A life course approach to cognitive reserve: a model for cognitive aging and development? Ann Neurol 2005 ; 58 : 617 – 22 . Google Scholar Crossref Search ADS PubMed 6 Osler M , Andersen AM , Laursen B , Lawlor DA. Cognitive function in childhood and early adulthood and injuries later in life: the Metropolit 1953 male birth cohort . Int J Epidemiol 2007 ; 36 : 212 – 19 . Google Scholar Crossref Search ADS PubMed 7 Dobson KG , Chow CH , Morrison KM , Van Lieshout RJ. Associations between childhood cognition and cardiovascular events in adulthood: a systematic review and meta-analysis . Can J Cardiol 2017 ; 33 : 232 – 42 . Google Scholar Crossref Search ADS PubMed 8 Nyberg J , Aberg MA , Schioler L et al. Cardiovascular and cognitive fitness at age 18 and risk of early-onset dementia . Brain 2014 ; 137 : 1514 – 23 . Google Scholar Crossref Search ADS PubMed 9 Breuer LE , Boon P , Bergmans JW et al. Cognitive deterioration in adult epilepsy: Does accelerated cognitive ageing exist? Neurosci Biobehav Rev 2016 ; 64 : 1 – 11 . Google Scholar Crossref Search ADS PubMed 10 Christensen GT , Molbo D , Angquist LH et al. Cohort Profile: The Danish Conscription Database (DCD): A cohort of 728 160 men born from 1939 through 1959 . Int J Epidemiol 2015 ; 44 : 432 – 40 . Google Scholar Crossref Search ADS PubMed 11 Teasdale TW. The Danish draft board's intelligence test, Borge Priens Prove: psychometric properties and research applications through 50 years . Scand J Psychol 2009 ; 50 : 633 – 38 . Google Scholar Crossref Search ADS PubMed 12 Mortensen EL , Reinisch JM , Teasdale TW. Intelligence as measured by the WAIS and a military draft board group test . Scand J Psychol 1989 ; 30 : 315 – 18 . Google Scholar Crossref Search ADS 13 Christensen J , Vestergaard M , Olsen J , Sidenius P. Validation of epilepsy diagnoses in the Danish National Hospital Register . Epilepsy Res 2007 ; 75 : 162 – 70 . Google Scholar Crossref Search ADS PubMed 14 St Germaine-Smith C , Liu M , Quan H , Wiebe S , Jette N. Development of an epilepsy-specific risk adjustment comorbidity index . Epilepsia 2011 ; 52 : 2161 – 67 . Google Scholar Crossref Search ADS PubMed 15 Deary IJ , Batty GD. Cognitive epidemiology . J Epidemiol Community Health 2007 ; 61 : 378 – 84 . Google Scholar Crossref Search ADS PubMed 16 Geerlings RP , Aldenkamp AP , Gottmer-Welschen LM et al. Developing from child to adult: Risk factors for poor psychosocial outcome in adolescents and young adults with epilepsy . Epilepsy Behav 2015 ; 51 : 182 – 90 . Google Scholar Crossref Search ADS PubMed 17 Camfield PR , Camfield CS. What happens to children with epilepsy when they become adults? Some facts and opinions . Pediatr Neurol 2014 ; 51 : 17 – 23 . Google Scholar Crossref Search ADS PubMed 18 Bourgeois BF , Prensky AL , Palkes HS , Talent BK , Busch SG. Intelligence in epilepsy: a prospective study in children . Ann Neurol 1983 ; 14 : 438 – 44 . Google Scholar Crossref Search ADS PubMed 19 Kjeldsen MJ , Kyvik KO , Christensen K , Friis ML. Genetic and environmental factors in epilepsy: A population-based study of 11,900 Danish twin pairs . Epilepsy Res 2001 ; 44 : 167 – 78 . Google Scholar Crossref Search ADS PubMed 20 Kee VR , Gilchrist B , Granner MA , Sarrazin NR , Carnahan RM. A systematic review of validated methods for identifying seizures, convulsions, or epilepsy using administrative and claims data . Pharmacoepidemiol Drug Saf 2012 ; 21(Suppl 1) : 183 – 93 . Google Scholar Crossref Search ADS PubMed 21 Baftiu A , Johannessen Landmark C , Rusten IR , Feet SA , Johannessen SI , Larsson PG. Changes in utilisation of antiepileptic drugs in epilepsy and non-epilepsy disorders - a pharmacoepidemiological study and clinical implications . Eur J Clin Pharmacol 2016 ; 72 : 1245 – 54 . Google Scholar Crossref Search ADS PubMed © The Author(s) 2018; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Epidemiology Oxford University Press

A bidirectional association between cognitive ability in young adulthood and epilepsy: a population-based cohort study

Loading next page...
 
/lp/ou_press/a-bidirectional-association-between-cognitive-ability-in-young-7RipBsVYgp
Publisher
Oxford University Press
Copyright
© The Author(s) 2018; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association
ISSN
0300-5771
eISSN
1464-3685
D.O.I.
10.1093/ije/dyy018
Publisher site
See Article on Publisher Site

Abstract

Abstract Aim To investigate the bidirectional association between cognitive ability in young adulthood and epilepsy. Methods This cohort study included 1 159 076 men enrolled in the mandatory conscription board examination from the Danish Conscription Database (DCD; 658 465 men examined 1957–84), the Danish Defence Personnel Organization Database (DPOD; 216 509 men examined 1987–2005) and the Danish Conscription Registry (DCR; 284 102 men examined 2006-15). A supplementary analysis included 14 814 female volunteers. Cognitive ability was measured at conscription, and epilepsy was ascertained by physician diagnoses in the Danish National Patient Registries 1977–2016 [using International Classification of Diseases (ICD) codes: ICD-8-345; ICD-10-G40-G41]. Differences in cognitive ability in relation to epilepsy status at the time of conscription (age 19) were calculated using linear regression. The risk of epilepsy associated with cognitive ability was estimated using Cox regression models, split at age at follow-up (40 and 60 years) and adjusted for year of birth, cerebrovascular disease, traumatic brain injury and education. Results In all, 5097 (1.0%) men from the DPOD and DCR were diagnosed with epilepsy before conscription, and they had about 0.25 standard deviation (SD) lower cognitive scores than men without epilepsy. The largest difference in cognition was seen for those with the largest number of hospital contacts. A total of 22 364 (1.9%) men developed epilepsy, and cognitive ability was inversely associated with the risk of epilepsy. With the end of follow-up at age 40 years, the adjusted hazard ratio (HR)per SD increase was 0.75 (95% confidence interval = 0.73–0.77). The association attenuated with increasing age at diagnosis. The findings were replicated in female conscripts. Conclusions The cognitive impairment seen in adults with epilepsy seems to reflect combined effects of epileptic processes and lower premorbid cognitive ability. Cognitive ability, epilepsy, cohort study Key Messages Epilepsy diagnosed during childhood and adolescence was associated with about 0.25 SD lower cognitive score in young adulthood. The largest reductions in cognitive ability were seen for those with the largest number of hospital contacts. Cognitive ability in young adulthood was inversely associated with the risk of developing epilepsy later in adult life. Introduction Epilepsy, defined as recurrent unprovoked seizures, is estimated to affect more than 65 million people worldwide.1,2 The age-specific incidence tends to be high in early childhood and low between ages 20 and 50, and increases again after 50 years of age.3 Thus, the estimated yearly incidence among children and adolescents is 70 per 100 000, and in adulthood the yearly incidence is 30 per 100 000 in people aged 20 64 and 100 per 100 000 in older people.4 In most cases the aetiology of epilepsy remains unknown. The recognized risk factors include traumatic brain injuries (TBI), cerebrovascular diseases, brain infections, neurodegenerative diseases and various genetic syndromes.1 Low early-life cognitive ability might reflect brain development5 and is associated with increased risk of injuries,6 cardiovascular disease7 and dementia.8 To our knowledge, no longitudinal studies have investigated the association between early life cognitive ability and risk of incident epilepsy. Cognitive impairment is common in epilepsy, with up to 65% of all patients showing cognitive deficits. This has been observed in both child and adult onset epilepsy, and epilepsy has been suggested to accelerate the rate of cognitive decline.9 Thus, low cognitive ability in patients with epilepsy may be observed both because cognitive ability is a risk factor for developing the disorder and because of effects of the disease on cognitive function.9 In the present study, we examined whether cognitive ability in young adult men was associated with epilepsy before and after they were diagnosed with the disease. Based on the literature, we hypothesized that men diagnosed with epilepsy before age 18 would have lower cognitive ability at the conscription board examination than men without epilepsy. We also suggested that low cognitive ability in young adulthood would be associated with an increased risk of developing epilepsy later in adult life. Further, we explored whether any association between epilepsy before age 18 and cognitive ability in young adulthood varied with age of onset and number of hospital admissions. Methods Study population The study was based on conscription board data from three sources: the Danish Conscription Database (DCD),10 the Danish Defence Personnel Organization Database (DPOD) and the Danish Conscription Registry (DCR). The DCD includes conscription board examination data for 728 159 men born between 1929 and 1960 (99.8% born between 1939 and 1959) and examined from 1957 through 1984 (99.8% examined between 1958 and 1978). The DCD has information on selected variables, digitized for research purposes. The DPOD has recorded key data from conscription board examinations from 1995 through 2005 on 217 361 men considered fit for military service or limited duties. They were born between 1950 and 1988 (90.7% born between 1976 and 1986). The DCR was established in 2006 and has since then registered all data from Danish conscription board examinations. For the present study, we retrieved data for all conscription examinations between 2006 and 2015. This included 349 487 men born between 1973 and 1998 (97.1% born between 1986 and 1997). Table 1 gives an overview of the male populations covered in the three sources of conscription data. Of the 1 295 007 men, 127 680 (9.9%) had missing information on one or more key variables and were excluded from the analyses (Table 1). The DPOD only contained information on those who, based on available health information, were considered suitable for military service and had few missing data [n = 806 (0.4%)]. For the DCD and DCR, the percentage of men with missing data was higher [n = 60 547 (8.5%) and 65 327 (18.8%)]. The missing information was mainly due to lack of registry information for men who volunteered for military service before they turned 18 or who had been exempted from the conscription board examination due to medical conditions such as intellectual disability, epilepsy or type 1 diabetes. Thus, available for data analysis were 1 167 327 men. From 1962, women could volunteer for military service and from 2005 on, all women receive an invitation to conscription when they turn 18 years. However, few women attend the conscription board examination and only 14 814 women (examined from 1995 through 2015) were registered with complete data in the DPOD or DCR. The data from this relatively small and selected subsample of women were used in a supplementary analysis. The study was approved by the Danish Data Protection Agency. Table 1 Distribution of variables from conscription board examination in relation to data source Number (%) in all data sources Number (%) in Danish Conscription Database Number (%) in Danish Defence Personnel Organisation Database Number (%) in Danish Conscription Registry Number of men in data source 1 295 007 728 159 217 361 349 487 Proportion of Danish men alive at age 24a 76.0 80.4 54.6 87.2 Number with missing data 127 680 (9.9) 60 547 (8.5) 806 (0.4) 65 327 (18.8) Number died or emigrated before 1977 8251 (0.6) 8147 (1.1) 46 (0.0) 58 (0.0) Number included 1 159 076 658 465 216 509 284 102 Year of examination  1957–66 261 911 (22.6) 261 911 (39.8)  1967–76 323 513 (27.9) 323 513 (49.1)  1977–84 73 041 ( 6.3) 73 041 (11.9)  1995–96 48 205 (4.2) 48 205 (4.2)  1997–2006 193 266 (16.7) 168 304 (77.7) 24 912 (8.8)  2007–15 59 140 (22.4) 259 140 (91.2) Year of birth  1939–49 357 081 (30.8) 357 081 (54.2)  1950–59 300 600 (45.7) 300 600 (45.7)  1960–69 1225 (0.1) 784 (0.1) 441 (0.2)  1970–79 99 900 (8.6) 99 706 (46.1) 194 (0.1)  1980–89 206 679 (18.5) 116 362 (53.8) 90 317 (31.8)  1990–98 193 591 (16.7) 193 591 (68.1) Mean height (SD) 178.4 (6.9) 176.8 (6.6) 180.4 (6.6) 180.3 (6.8) Education  High 307 324 (26.5) 112 695 (17.1) 77 112 (35.7) 117 318 (41.3)  Middle 534 480 (46.1) 376 683 (57.2) 68 304 (31.6)) 897 493 (31.5)  Low 317 229 (27.4) 169 087 (25.7) 70 861 (32.7) 77 281 (27.2) Mean BPP (SD) 39.9 (11.1) 37.8 (12.0) 44.5 (8.3) 41.1 (9.6) Cerebrovascular disease before examination 285 (0.0) 3 (0.0) 99 (0.1) 183 (0.1) TBI diagnosis before examination 22 490 (1.9) 20 (0.0) 304 (0.1) 22 166 (7.8) Number (%) in all data sources Number (%) in Danish Conscription Database Number (%) in Danish Defence Personnel Organisation Database Number (%) in Danish Conscription Registry Number of men in data source 1 295 007 728 159 217 361 349 487 Proportion of Danish men alive at age 24a 76.0 80.4 54.6 87.2 Number with missing data 127 680 (9.9) 60 547 (8.5) 806 (0.4) 65 327 (18.8) Number died or emigrated before 1977 8251 (0.6) 8147 (1.1) 46 (0.0) 58 (0.0) Number included 1 159 076 658 465 216 509 284 102 Year of examination  1957–66 261 911 (22.6) 261 911 (39.8)  1967–76 323 513 (27.9) 323 513 (49.1)  1977–84 73 041 ( 6.3) 73 041 (11.9)  1995–96 48 205 (4.2) 48 205 (4.2)  1997–2006 193 266 (16.7) 168 304 (77.7) 24 912 (8.8)  2007–15 59 140 (22.4) 259 140 (91.2) Year of birth  1939–49 357 081 (30.8) 357 081 (54.2)  1950–59 300 600 (45.7) 300 600 (45.7)  1960–69 1225 (0.1) 784 (0.1) 441 (0.2)  1970–79 99 900 (8.6) 99 706 (46.1) 194 (0.1)  1980–89 206 679 (18.5) 116 362 (53.8) 90 317 (31.8)  1990–98 193 591 (16.7) 193 591 (68.1) Mean height (SD) 178.4 (6.9) 176.8 (6.6) 180.4 (6.6) 180.3 (6.8) Education  High 307 324 (26.5) 112 695 (17.1) 77 112 (35.7) 117 318 (41.3)  Middle 534 480 (46.1) 376 683 (57.2) 68 304 (31.6)) 897 493 (31.5)  Low 317 229 (27.4) 169 087 (25.7) 70 861 (32.7) 77 281 (27.2) Mean BPP (SD) 39.9 (11.1) 37.8 (12.0) 44.5 (8.3) 41.1 (9.6) Cerebrovascular disease before examination 285 (0.0) 3 (0.0) 99 (0.1) 183 (0.1) TBI diagnosis before examination 22 490 (1.9) 20 (0.0) 304 (0.1) 22 166 (7.8) a Data from statistics Denmark for the included birth cohorts alive age 24:total 1 704 127; DCD: n = 905 911 (1939-59); DPOD: n = 397 789 (1978-86); DCR: n = 400.427 (1986-97). Table 1 Distribution of variables from conscription board examination in relation to data source Number (%) in all data sources Number (%) in Danish Conscription Database Number (%) in Danish Defence Personnel Organisation Database Number (%) in Danish Conscription Registry Number of men in data source 1 295 007 728 159 217 361 349 487 Proportion of Danish men alive at age 24a 76.0 80.4 54.6 87.2 Number with missing data 127 680 (9.9) 60 547 (8.5) 806 (0.4) 65 327 (18.8) Number died or emigrated before 1977 8251 (0.6) 8147 (1.1) 46 (0.0) 58 (0.0) Number included 1 159 076 658 465 216 509 284 102 Year of examination  1957–66 261 911 (22.6) 261 911 (39.8)  1967–76 323 513 (27.9) 323 513 (49.1)  1977–84 73 041 ( 6.3) 73 041 (11.9)  1995–96 48 205 (4.2) 48 205 (4.2)  1997–2006 193 266 (16.7) 168 304 (77.7) 24 912 (8.8)  2007–15 59 140 (22.4) 259 140 (91.2) Year of birth  1939–49 357 081 (30.8) 357 081 (54.2)  1950–59 300 600 (45.7) 300 600 (45.7)  1960–69 1225 (0.1) 784 (0.1) 441 (0.2)  1970–79 99 900 (8.6) 99 706 (46.1) 194 (0.1)  1980–89 206 679 (18.5) 116 362 (53.8) 90 317 (31.8)  1990–98 193 591 (16.7) 193 591 (68.1) Mean height (SD) 178.4 (6.9) 176.8 (6.6) 180.4 (6.6) 180.3 (6.8) Education  High 307 324 (26.5) 112 695 (17.1) 77 112 (35.7) 117 318 (41.3)  Middle 534 480 (46.1) 376 683 (57.2) 68 304 (31.6)) 897 493 (31.5)  Low 317 229 (27.4) 169 087 (25.7) 70 861 (32.7) 77 281 (27.2) Mean BPP (SD) 39.9 (11.1) 37.8 (12.0) 44.5 (8.3) 41.1 (9.6) Cerebrovascular disease before examination 285 (0.0) 3 (0.0) 99 (0.1) 183 (0.1) TBI diagnosis before examination 22 490 (1.9) 20 (0.0) 304 (0.1) 22 166 (7.8) Number (%) in all data sources Number (%) in Danish Conscription Database Number (%) in Danish Defence Personnel Organisation Database Number (%) in Danish Conscription Registry Number of men in data source 1 295 007 728 159 217 361 349 487 Proportion of Danish men alive at age 24a 76.0 80.4 54.6 87.2 Number with missing data 127 680 (9.9) 60 547 (8.5) 806 (0.4) 65 327 (18.8) Number died or emigrated before 1977 8251 (0.6) 8147 (1.1) 46 (0.0) 58 (0.0) Number included 1 159 076 658 465 216 509 284 102 Year of examination  1957–66 261 911 (22.6) 261 911 (39.8)  1967–76 323 513 (27.9) 323 513 (49.1)  1977–84 73 041 ( 6.3) 73 041 (11.9)  1995–96 48 205 (4.2) 48 205 (4.2)  1997–2006 193 266 (16.7) 168 304 (77.7) 24 912 (8.8)  2007–15 59 140 (22.4) 259 140 (91.2) Year of birth  1939–49 357 081 (30.8) 357 081 (54.2)  1950–59 300 600 (45.7) 300 600 (45.7)  1960–69 1225 (0.1) 784 (0.1) 441 (0.2)  1970–79 99 900 (8.6) 99 706 (46.1) 194 (0.1)  1980–89 206 679 (18.5) 116 362 (53.8) 90 317 (31.8)  1990–98 193 591 (16.7) 193 591 (68.1) Mean height (SD) 178.4 (6.9) 176.8 (6.6) 180.4 (6.6) 180.3 (6.8) Education  High 307 324 (26.5) 112 695 (17.1) 77 112 (35.7) 117 318 (41.3)  Middle 534 480 (46.1) 376 683 (57.2) 68 304 (31.6)) 897 493 (31.5)  Low 317 229 (27.4) 169 087 (25.7) 70 861 (32.7) 77 281 (27.2) Mean BPP (SD) 39.9 (11.1) 37.8 (12.0) 44.5 (8.3) 41.1 (9.6) Cerebrovascular disease before examination 285 (0.0) 3 (0.0) 99 (0.1) 183 (0.1) TBI diagnosis before examination 22 490 (1.9) 20 (0.0) 304 (0.1) 22 166 (7.8) a Data from statistics Denmark for the included birth cohorts alive age 24:total 1 704 127; DCD: n = 905 911 (1939-59); DPOD: n = 397 789 (1978-86); DCR: n = 400.427 (1986-97). Cognitive measures Cognitive ability This was assessed by the conscription board intelligence test called the Børge Prien Prøve (BPP). The BPP consists of four subtests assessing logical, verbal, numerical and spatial ability and was recorded as a total score with a range of 0–78, representing the sum of the number of correct answers to the four subtests. This test has been used without changes since 1957,11 and the total BPP score correlates well with the full-scale Wechsler Adult Intelligence Scale (r = 0.82).12 For 260 658 men in DCR we also had information on basic school grade points from the 9th class (15 years of age) recorded in the School Grade Register. This Register was established in 2002 and uses the European Credit Transfer and Accumulation System grading scale: 3,0,2,4,7,10 and 12, ranging from fail marks (-3 poor or 0 inadequate) to the best (10 very good or 12 excellent). We used the highest grade obtained in Danish and/or mathematics. Epilepsy Hospital diagnosis of epilepsy In the present study, information on any admission to a Danish hospital from 1977 until 2016 was obtained by record linkage with the Danish National Patient Registry (DNPR), and the Danish Psychiatric Central Register (PCR) using the civil registration number as key. Diagnoses for all in-hospital discharges have been classified according to 8th Revision of the International Classification of Diseases (ICD-8) for the period 1977 to 1994, and according to the 10th Revision (ICD-10) from 1995 and onwards. From 1995, the DNPR also included all outpatient and emergency contacts. Death from epilepsy without previous hospitalization was identified by linkage with the Danish Cause of Death Register (n = 86). Incident cases were defined as first hospital discharge or death with main or contributory diagnosis of epilepsy (ICD-8-345.00-345.99 and ICD-10-G40.0-G41.9). For men diagnosed with epilepsy before the conscription board examination, we also calculated the number of readmissions from disease onset until the examination. The validity of epilepsy diagnosis obtained from the DNPR has been assessed in a study of 200 patients admitted between 1977 and 2002.13 This showed that 81% of epilepsy diagnoses were confirmed by review of medical records. Epilepsy diagnosed at conscription board examination The DCR also included diagnoses registered at the conscription board based on health certificates, subjective reports or the objective health examination. These diagnoses were coded following the ICD-10. Of the young men diagnosed with epilepsy at conscription, 524 (75.6%) also had an epilepsy diagnosis in the patient registries. Supplementary Table 1a and b (available as Supplementary data at IJE online) shows the number of epilepsy cases identified in the different registers in relation to follow-up and missing conscription data. This indicated that epilepsy was most frequent among men exempted from conscription board examination. Similarly, we found that the percentage of men diagnosed with epilepsy at the conscription board registration in the DCR was higher [2.6% (n = 1704)] in men with missing information as compared with those with data from a conscription board examination [0.2% (n = 704)]. Covariables Since reports suggest that cerebrovascular disease and severe TBI4,14 are potential confounders for associations between cognitive ability and epilepsy, we identified men who had been diagnosed with these conditions before the conscription board examination, using the relevant codes for classification of cerebrovascular disease (ICD-8-430-438 and ICD-10-I60-I69, G46;G45.9) or TBI (ICD-8-851-854 and ICD-10-S06.1-S06.9). From the conscription board data we also had information on educational level, which has been shown to be associated with cognitive ability early in life15 and epilepsy.1 Educational level reported at conscription was coded into three categories comprising low (7–9th grade), medium (vocational training or 10th-11th grade) and high educational level (12th grade or more advanced). Although conscription board examinations are highly standardized, there might be some variation across the seven districts. Information on the location of the examination was available in DCD and DPOD and was included in a supplementary analysis. Statistical analyses Association between epilepsy and cognitive ability The associations between diagnosed epilepsy before conscription board examinations and BPP scores were analysed using simple and multiple linear regression. Since information on epilepsy diagnosis was available from 1977, very few men (n = 30) from the DCD had been diagnosed with epilepsy before the examination. Thus, the analyses were restricted to the men from DPOD and DCR. In adjusted models, cerebrovascular disease, TBI and year of birth were included as covariables. In the DCR, we also examined how grade points from 9th class were related to epilepsy diagnosis. Finally, we analysed the influence of age at diagnosis on cognitive ability in men with epilepsy. Association between cognitive ability and incident epilepsy The association between cognitive ability in young adulthood and subsequent incident epilepsy was analysed using Cox proportional hazard regression with age as the underlying time scale. Person-years (py) of follow-up for an epilepsy diagnosis were accumulated from age on 1 January 1977 (age 19–39 years) or on the date of the conscription board examination, and follow-up was terminated at the age of first admission for epilepsy, death, emigration or end of follow–up (1 April 2016; age 19–78 years), whichever came first. We analysed all data separately for the DCD, DPOD and DCR because the men had been followed at different ages and because the registers varied in coverage before and after 1995. BPP was analysed as a continuous variable, since the association with the risk of epilepsy seemed to be linear. The initial analysis showed that hazard ratios (HRs) varied slightly with age at follow-up, and consequently data from the DCD was split on follow-up times 40 and 60 years. Men who had died (n = 4029) or emigrated (n = 4222) before 1977 were excluded from all analysis (Table 1). In addition, 11 266 men with epilepsy before study entry were not included in the time-to-event analyses. All analyses were carried out in STATA version 12. Results The men from the oldest birth years (DCD) had lower education and BPP score than the men from the DPOD and DCR representing the youngest birth years. The men from DPOD had a higher mean and smaller SD in BPP scores because the DPOD only consists of men evaluated as being suitable for military service. The lower rows in Table 1 provide the distribution of covariables in relation to data source (DCD, DPOD or DCR). Epilepsy and cognitive ability in young adulthood A total of 2453 (1.1%) and 3046 (1.0 %) men from the DPOD and DCR, respectively, had been registered with an epilepsy diagnosis in the DNPR before the conscription board examination. In the DCR, 702 (0.3%) men had also an epilepsy diagnosis registered at the conscription board examination. Table 2 shows that in both cohorts, men with an epilepsy diagnosis before conscription had about 0.25 SD lower mean cognitive scores than those without epilepsy. The number of hospital contacts with epilepsy was associated with cognitive ability in a dose-response-like manner. The association was attenuated when adjusting for year of birth, cerebrovascular disease and TBI. In the DCR, analyses were repeated with the epilepsy recorded at the conscription board examination, and this showed the same pattern of association. Grade points from 9th class correlated well with the BPP measured 4-5 years later (r = 0.48) and 2 489 men with epilepsy diagnosed before the basic school grading had lower mean gradings [5.57 (SD 2.47)] than the 175 men diagnosed after [5.86 (SD 2.61)] or the 257 994 men with no epilepsy diagnosis before conscription [5.97 (SD 2.56)]. Age at diagnosis was not associated with BPP in men with epilepsy (data not shown). Table 2 Mean cognitive score at conscription board examination in relation to epilepsy diagnosis before or at the examination in Danish men examined between 1986 and 2015 Number BPP score mean (SD) Beta-coefficienta (95% CI) Adjustedb beta-coefficient (95%CI) Danish Defence Personel Organization Database (n = 216 509) Epilepsy diagnosed in patient registry before conscription  No 214 056 44.5 (8.5)  Yes 2453 43.3 (8.3) −1.37 (−1.70 to −1.04) −1.38 (−1.71 to −1.05) Number of hospital contacts with epilepsy  No 214 056 44.5 (8.3) 1.24 (0.90 to 1.59) 1.27 (1.05 to 1.59)  1–4 2303 43.4 (8.6) 0 0  5–22 150 41.1 (8.4) −2.13 (−3.50 to −0.75) −2.19 (−3.52 to −0.81) Danish Conscription Register (n = 284 102) Epilepsy diagnosed in patient registry before conscription  No 281 056 41.1 (10.2)  Yes 3 046 39.1 (9.5) −2.03 (−2.37 to −1.68) −1.95 (−2.21 to −1.60) Number of hospital contacts with epilepsy  No 281 124 41.1 (9.5) 1.44 (1.05 to 1.83) 1.42 (1.03-1.81)  1–4 2 343 39.7 (9.9) 0 0  5–46 635 37.1 (11.0) −2.58 (−3.42 to −1.74) −2.46 (−3.30 to −1.28) Epilepsy diagnosed at conscription  No 283 400 41.1 (9.5)  Yes 702 40.1 (10.2) −1.03 (−1.73 to −0.32) −1.00 (−1.69 to −0.28) Number BPP score mean (SD) Beta-coefficienta (95% CI) Adjustedb beta-coefficient (95%CI) Danish Defence Personel Organization Database (n = 216 509) Epilepsy diagnosed in patient registry before conscription  No 214 056 44.5 (8.5)  Yes 2453 43.3 (8.3) −1.37 (−1.70 to −1.04) −1.38 (−1.71 to −1.05) Number of hospital contacts with epilepsy  No 214 056 44.5 (8.3) 1.24 (0.90 to 1.59) 1.27 (1.05 to 1.59)  1–4 2303 43.4 (8.6) 0 0  5–22 150 41.1 (8.4) −2.13 (−3.50 to −0.75) −2.19 (−3.52 to −0.81) Danish Conscription Register (n = 284 102) Epilepsy diagnosed in patient registry before conscription  No 281 056 41.1 (10.2)  Yes 3 046 39.1 (9.5) −2.03 (−2.37 to −1.68) −1.95 (−2.21 to −1.60) Number of hospital contacts with epilepsy  No 281 124 41.1 (9.5) 1.44 (1.05 to 1.83) 1.42 (1.03-1.81)  1–4 2 343 39.7 (9.9) 0 0  5–46 635 37.1 (11.0) −2.58 (−3.42 to −1.74) −2.46 (−3.30 to −1.28) Epilepsy diagnosed at conscription  No 283 400 41.1 (9.5)  Yes 702 40.1 (10.2) −1.03 (−1.73 to −0.32) −1.00 (−1.69 to −0.28) a Linear regression with cognitive scores as outcome. b Adjusted for birth year, cerebrovascular disease and traumatic brain injury at conscription (DPOD also for conscription board district). Table 2 Mean cognitive score at conscription board examination in relation to epilepsy diagnosis before or at the examination in Danish men examined between 1986 and 2015 Number BPP score mean (SD) Beta-coefficienta (95% CI) Adjustedb beta-coefficient (95%CI) Danish Defence Personel Organization Database (n = 216 509) Epilepsy diagnosed in patient registry before conscription  No 214 056 44.5 (8.5)  Yes 2453 43.3 (8.3) −1.37 (−1.70 to −1.04) −1.38 (−1.71 to −1.05) Number of hospital contacts with epilepsy  No 214 056 44.5 (8.3) 1.24 (0.90 to 1.59) 1.27 (1.05 to 1.59)  1–4 2303 43.4 (8.6) 0 0  5–22 150 41.1 (8.4) −2.13 (−3.50 to −0.75) −2.19 (−3.52 to −0.81) Danish Conscription Register (n = 284 102) Epilepsy diagnosed in patient registry before conscription  No 281 056 41.1 (10.2)  Yes 3 046 39.1 (9.5) −2.03 (−2.37 to −1.68) −1.95 (−2.21 to −1.60) Number of hospital contacts with epilepsy  No 281 124 41.1 (9.5) 1.44 (1.05 to 1.83) 1.42 (1.03-1.81)  1–4 2 343 39.7 (9.9) 0 0  5–46 635 37.1 (11.0) −2.58 (−3.42 to −1.74) −2.46 (−3.30 to −1.28) Epilepsy diagnosed at conscription  No 283 400 41.1 (9.5)  Yes 702 40.1 (10.2) −1.03 (−1.73 to −0.32) −1.00 (−1.69 to −0.28) Number BPP score mean (SD) Beta-coefficienta (95% CI) Adjustedb beta-coefficient (95%CI) Danish Defence Personel Organization Database (n = 216 509) Epilepsy diagnosed in patient registry before conscription  No 214 056 44.5 (8.5)  Yes 2453 43.3 (8.3) −1.37 (−1.70 to −1.04) −1.38 (−1.71 to −1.05) Number of hospital contacts with epilepsy  No 214 056 44.5 (8.3) 1.24 (0.90 to 1.59) 1.27 (1.05 to 1.59)  1–4 2303 43.4 (8.6) 0 0  5–22 150 41.1 (8.4) −2.13 (−3.50 to −0.75) −2.19 (−3.52 to −0.81) Danish Conscription Register (n = 284 102) Epilepsy diagnosed in patient registry before conscription  No 281 056 41.1 (10.2)  Yes 3 046 39.1 (9.5) −2.03 (−2.37 to −1.68) −1.95 (−2.21 to −1.60) Number of hospital contacts with epilepsy  No 281 124 41.1 (9.5) 1.44 (1.05 to 1.83) 1.42 (1.03-1.81)  1–4 2 343 39.7 (9.9) 0 0  5–46 635 37.1 (11.0) −2.58 (−3.42 to −1.74) −2.46 (−3.30 to −1.28) Epilepsy diagnosed at conscription  No 283 400 41.1 (9.5)  Yes 702 40.1 (10.2) −1.03 (−1.73 to −0.32) −1.00 (−1.69 to −0.28) a Linear regression with cognitive scores as outcome. b Adjusted for birth year, cerebrovascular disease and traumatic brain injury at conscription (DPOD also for conscription board district). Cognitive ability in young adulthood and risk of epilepsy The 658 435, 214 056 and 281 056 men from the DCD, DPOD and DCR, respectively, without epilepsy at conscription were followed for epilepsy in the DNPR between 1977 and 2016 (medians of 45.1, 16.1 and 5.0 years in the DCD, DPOD and DCR, respectively). During that time, respectively 18 985 (2.9%), 1233 (0.6%) and 2146 (0.8%) developed epilepsy with corresponding incidence rates of 6.5 [95% confidence interval (CI) = 6.4–6.6], 3.7 (95% CI = 3.5–3.9) and 14.5 (95% CI = 13.9–15.1) per 10 000 py. The incidence increased with age at follow-up (see Supplementary Figure 1, available as Supplementary data at IJE online). Table 3 provides the crude and adjusted HRs for the association between cognitive ability in young adulthood and epilepsy in relation to age at follow-up in the three conscription cohorts. It shows that in all three cohorts combined, cognitive ability in young adulthood was inversely associated with risk of epilepsy before age 40 years (HRper SD increase in BPP 0.75, 95% CI = 0.73–0.77) after adjustment for cerebrovascular disease, TBI and educational level. In the DCD, the HRs also seemed to attenuate with increasing age at diagnosis. Further adjustment for location by district did not influence the HRs for the DCD or DPOD. The supplementary analyses for the 14 814 women indicated a similar bidirectional association between young adult cognitive ability and epilepsy as observed for men (Table 4). Table 3 The hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between cognitive ability (BPP) in young adulthood and epilepsy in men from the Danish Conscription Database (DCD), Danish Defence Personnel Organization Database (DPOD) and the Danish Conscription Register (DCR) Age at diagnosis or end of follow-up Number of outcomes HR (95%CI)a HRb (95% CI) DCD (n = 658 435 men) ≤ 40 4342 0.74 (0.71–0.75) 0.76 (0.73–0.79) 41–60 9406 0.84 (0.82–0.85) 0.86 (0.84–0.88) > 60 5236 0.95 (0.90–1.01) 0.92 (0.86–0.99) DPOD (n = 214 056 men) ≤40 1218 0.74 (0.68–0.79) 0.82 (0.74–0.89) DCR (n = 281 056 men) ≤ 40 2140 0.74 (0.72–78) 0.80 (0.75–0.84) All men ( 1 153 547) < 40 7700 0.81 (0.79–0.83) 0.75 (0.73–0.77) Age at diagnosis or end of follow-up Number of outcomes HR (95%CI)a HRb (95% CI) DCD (n = 658 435 men) ≤ 40 4342 0.74 (0.71–0.75) 0.76 (0.73–0.79) 41–60 9406 0.84 (0.82–0.85) 0.86 (0.84–0.88) > 60 5236 0.95 (0.90–1.01) 0.92 (0.86–0.99) DPOD (n = 214 056 men) ≤40 1218 0.74 (0.68–0.79) 0.82 (0.74–0.89) DCR (n = 281 056 men) ≤ 40 2140 0.74 (0.72–78) 0.80 (0.75–0.84) All men ( 1 153 547) < 40 7700 0.81 (0.79–0.83) 0.75 (0.73–0.77) a HR per SD increase in BPP. b Adjusted for cerebrovascular disease, traumatic brain injury and educational level at conscription. Table 3 The hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between cognitive ability (BPP) in young adulthood and epilepsy in men from the Danish Conscription Database (DCD), Danish Defence Personnel Organization Database (DPOD) and the Danish Conscription Register (DCR) Age at diagnosis or end of follow-up Number of outcomes HR (95%CI)a HRb (95% CI) DCD (n = 658 435 men) ≤ 40 4342 0.74 (0.71–0.75) 0.76 (0.73–0.79) 41–60 9406 0.84 (0.82–0.85) 0.86 (0.84–0.88) > 60 5236 0.95 (0.90–1.01) 0.92 (0.86–0.99) DPOD (n = 214 056 men) ≤40 1218 0.74 (0.68–0.79) 0.82 (0.74–0.89) DCR (n = 281 056 men) ≤ 40 2140 0.74 (0.72–78) 0.80 (0.75–0.84) All men ( 1 153 547) < 40 7700 0.81 (0.79–0.83) 0.75 (0.73–0.77) Age at diagnosis or end of follow-up Number of outcomes HR (95%CI)a HRb (95% CI) DCD (n = 658 435 men) ≤ 40 4342 0.74 (0.71–0.75) 0.76 (0.73–0.79) 41–60 9406 0.84 (0.82–0.85) 0.86 (0.84–0.88) > 60 5236 0.95 (0.90–1.01) 0.92 (0.86–0.99) DPOD (n = 214 056 men) ≤40 1218 0.74 (0.68–0.79) 0.82 (0.74–0.89) DCR (n = 281 056 men) ≤ 40 2140 0.74 (0.72–78) 0.80 (0.75–0.84) All men ( 1 153 547) < 40 7700 0.81 (0.79–0.83) 0.75 (0.73–0.77) a HR per SD increase in BPP. b Adjusted for cerebrovascular disease, traumatic brain injury and educational level at conscription. Table 4 The distribution of variables and the associations between cognitive ability and epilepsy before and after conscription examination in 14 814 female volunteers Distribution of variables (%) Conscription register  Danish Defence Personel Organization Data 2 427 (16.4)  Danish Conscription Register 12 387 (83.6) Birth cohort  −1979 1 054 (7.3)  1980–89 5 445 (36.8)  1990–98 8 293 (56.0) Educational level  High 8 733 (59.0)  Middle 3 930 (26.5)  Low 2 151 (14.5) Mean BPP (SD) 41.9 (8.4) Cerebrovascular disease before examination 16 (1.1) TBI before examination 969 (6.5) Epilepsy diagnosis  Before examination 161 (1.1)  After examination 145 (1.0) Epilepsy and mean cognitive ability Epilepsy diagnosis  No 42.0 (8.3)  Yes, before 39.3 (8.7) Cognitive ability and epilepsy risk (HR and 95% CI)  BPP per SD increase 0.64 (0.52–0.80)  BPPaper SD increase 0.69 (0.59–0.87) Distribution of variables (%) Conscription register  Danish Defence Personel Organization Data 2 427 (16.4)  Danish Conscription Register 12 387 (83.6) Birth cohort  −1979 1 054 (7.3)  1980–89 5 445 (36.8)  1990–98 8 293 (56.0) Educational level  High 8 733 (59.0)  Middle 3 930 (26.5)  Low 2 151 (14.5) Mean BPP (SD) 41.9 (8.4) Cerebrovascular disease before examination 16 (1.1) TBI before examination 969 (6.5) Epilepsy diagnosis  Before examination 161 (1.1)  After examination 145 (1.0) Epilepsy and mean cognitive ability Epilepsy diagnosis  No 42.0 (8.3)  Yes, before 39.3 (8.7) Cognitive ability and epilepsy risk (HR and 95% CI)  BPP per SD increase 0.64 (0.52–0.80)  BPPaper SD increase 0.69 (0.59–0.87) a Adjusted for cerebrovascular disease, traumatic brain injury (TBI) and educational level at conscription. Table 4 The distribution of variables and the associations between cognitive ability and epilepsy before and after conscription examination in 14 814 female volunteers Distribution of variables (%) Conscription register  Danish Defence Personel Organization Data 2 427 (16.4)  Danish Conscription Register 12 387 (83.6) Birth cohort  −1979 1 054 (7.3)  1980–89 5 445 (36.8)  1990–98 8 293 (56.0) Educational level  High 8 733 (59.0)  Middle 3 930 (26.5)  Low 2 151 (14.5) Mean BPP (SD) 41.9 (8.4) Cerebrovascular disease before examination 16 (1.1) TBI before examination 969 (6.5) Epilepsy diagnosis  Before examination 161 (1.1)  After examination 145 (1.0) Epilepsy and mean cognitive ability Epilepsy diagnosis  No 42.0 (8.3)  Yes, before 39.3 (8.7) Cognitive ability and epilepsy risk (HR and 95% CI)  BPP per SD increase 0.64 (0.52–0.80)  BPPaper SD increase 0.69 (0.59–0.87) Distribution of variables (%) Conscription register  Danish Defence Personel Organization Data 2 427 (16.4)  Danish Conscription Register 12 387 (83.6) Birth cohort  −1979 1 054 (7.3)  1980–89 5 445 (36.8)  1990–98 8 293 (56.0) Educational level  High 8 733 (59.0)  Middle 3 930 (26.5)  Low 2 151 (14.5) Mean BPP (SD) 41.9 (8.4) Cerebrovascular disease before examination 16 (1.1) TBI before examination 969 (6.5) Epilepsy diagnosis  Before examination 161 (1.1)  After examination 145 (1.0) Epilepsy and mean cognitive ability Epilepsy diagnosis  No 42.0 (8.3)  Yes, before 39.3 (8.7) Cognitive ability and epilepsy risk (HR and 95% CI)  BPP per SD increase 0.64 (0.52–0.80)  BPPaper SD increase 0.69 (0.59–0.87) a Adjusted for cerebrovascular disease, traumatic brain injury (TBI) and educational level at conscription. Discussion In this nation-wide cohort of 1 153 076 Danish male conscripts we found, corroborating our previous hypothesis, that epilepsy diagnosed during childhood and adolescence was associated with about 0.25 SD lower cognitive ability in young adulthood. The largest differences were seen for men with the largest number of hospital contacts. The associations persisted after adjustment for year of birth, cerebrovascular disease and TBI. The study also showed that cognitive ability in young adulthood was inversely associated with the risk of developing epilepsy when cerebrovascular disease, TBI and educational level were adjusted for. This was also in accordance with our hypothesis, and the findings also were replicated in a smaller cohort of female conscripts volunteering for military service. The proportion of men who developed epilepsy before age 40 years (0.7%) corresponded to the estimates for men in Sweden8 and Europe4 of 0.6%. In the present study, the proportion of epilepsy was higher in the DCR than in the DPOD and DCD. This might reflect that the DPOD represented men suitable for military employment, whereas the follow-up of most of the men in the DCD was from after the age of 18, and thus follow-up during young adulthood was based on in-hospital contacts only. However, these differences did not seem to influence the associations between variables estimated within the three cohorts, as they were rather similar. A few smaller follow-up studies of children with epilepsy (n < 400) have suggested that higher seizure frequency is associated with cognitive impairments or poor educational achievements.16–18 These studies did not compare the cognitive outcomes with children who grew up without epilepsy. Unfortunately, we did not have repeated measures of cognitive ability in our cohort and do not know whether the lower cognitive ability had been present before disease onset or had in fact decreased due to the effect of epileptic seizures or treatment. However, the supplementary analysis with school grades obtained 4-5 years before conscription indicated that those with onset after the 9th class examination had higher grades than those diagnosed before. On the other hand, it is also possible that this relatively small number of events emerging during adolescence represent milder cases. Thus, our analysis of the relation between cognitive ability and epilepsy indicated that lower cognitive ability might be present before epilepsy, as a potential marker of brain dysfunction acquired early in life. The influence of cognitive ability in young adulthood on later epilepsy could be explained by confounding from genetic predisposition, familial factors and poor health during early childhood or influences mediated through cardiovascular diseases, depression and injuries.1,6,7,19 In our study, educational level and TBI were not important confounders. Furthermore, information on cognitive ability was not available for those exempted from conscription due to severe somatic, neurodegenerative and psychiatric disorders. This suggests that the reported associations are not due to confounding from severe mental disorders. The strength of the present analysis is that it was based on a large sample and that it covers nearly 60 birth years (1939–99). In addition, cognitive ability was assessed in young adulthood when it can be expected to be relatively stable. Outcome information was retrieved from national registers and we were able to include all hospitalizations for epilepsy from 1977 with sensitivity measures above 80%. However, for men in the DCD born between 1939 and 1959, this information was first available at age 19–39 years. Unfortunately, most of the diagnoses were based on non-specific ICD codes and it was not possible to look specifically at epilepsy with focal or generalized seizures. Further, our study did not include those exempted from conscription due to epilepsy, and our analysis of non-participants showed that subjects with missing data had higher risk of epilepsy than those available for the present study. This could suggest that our estimates of the relation between epilepsy and cognitive ability may not apply to these potentially more severe cases. In the DCD, study participants were followed for a median of 45 years. During this time, ICD coding has changed from ICD-8 to ICD-10, and the coverage of the hospital registers has increased due to the inclusion of outpatient and emergency contacts from 1995. However, we did not find any major difference in the risk estimates for the association between cognitive ability and epilepsy across several birth years. Our outcome measure may not include the presumed milder cases of epilepsy diagnosed and treated only in primary care in Denmark. We did consider using information on purchase of anti-epileptic medication (AED) from the Danish Prescription register. However, it is not recommended to identify patients with epilepsy through filled prescriptions for AED20 as it would lead to misclassification, since approximately 50% of all anti-epileptic drugs are prescribed for other purposes than epilepsy.21 Data from the DCR indicated that around 25% of young men with epilepsy would not be registered in the DNPR. Our main results were based on a cohort of men, but a supplementary analysis on a selected subsample of female conscripts indicated that the findings could be generalizable to women. In conclusion, hospital contacts for epilepsy were associated with decreased cognitive ability in young adulthood in a dose-response-dependent manner. Low cognitive ability in young adulthood was also associated with higher risk of developing epilepsy later in life. Thus, the cognitive impairment seen in adult patients with epilepsy may reflect combined effects of epileptic processes and lower premorbid cognitive ability, which might be a marker of brain dysfunction inherited or acquired early in life. Funding The work was supported by the Danish Medical Research Council (grant number 09-063599 and 09-069151) and the Velux Foundation (grant number 31205), the Jascha Foundation (grant 6200) and Doctor Sofus Carl Emil Friis and Olga Doris Friis grant. Acknowledgement We thank Esther Mortensen, MA in English, for linguistic revision. Conflict of interest: None declared. References 1 Abramovici S , Bagic A. Epidemiology of epilepsy . Handb Clin Neurol 2016 ; 138 : 159 – 71 . Google Scholar Crossref Search ADS PubMed 2 Singh A , Trevick S. The epidemiology of global epilepsy. Neurol Clin 2016 ; 34 : 837 – 847 . Google Scholar Crossref Search ADS PubMed 3 Christensen J , Vestergaard M , Pedersen MG , Pedersen CB , Olsen J , Sidenius P. Incidence and prevalence of epilepsy in Denmark . Epilepsy Res 2007 ; 76 : 60 – 65 . Google Scholar Crossref Search ADS PubMed 4 Forsgren L , Beghi E , Oun A , Silanpaa M. The epidemiology of epilepsy in Europe - a systematic review . Eur J Neurol 2005 ; 12 : 245 – 53 . Google Scholar Crossref Search ADS PubMed 5 Richards M , Deary IJ. A life course approach to cognitive reserve: a model for cognitive aging and development? Ann Neurol 2005 ; 58 : 617 – 22 . Google Scholar Crossref Search ADS PubMed 6 Osler M , Andersen AM , Laursen B , Lawlor DA. Cognitive function in childhood and early adulthood and injuries later in life: the Metropolit 1953 male birth cohort . Int J Epidemiol 2007 ; 36 : 212 – 19 . Google Scholar Crossref Search ADS PubMed 7 Dobson KG , Chow CH , Morrison KM , Van Lieshout RJ. Associations between childhood cognition and cardiovascular events in adulthood: a systematic review and meta-analysis . Can J Cardiol 2017 ; 33 : 232 – 42 . Google Scholar Crossref Search ADS PubMed 8 Nyberg J , Aberg MA , Schioler L et al. Cardiovascular and cognitive fitness at age 18 and risk of early-onset dementia . Brain 2014 ; 137 : 1514 – 23 . Google Scholar Crossref Search ADS PubMed 9 Breuer LE , Boon P , Bergmans JW et al. Cognitive deterioration in adult epilepsy: Does accelerated cognitive ageing exist? Neurosci Biobehav Rev 2016 ; 64 : 1 – 11 . Google Scholar Crossref Search ADS PubMed 10 Christensen GT , Molbo D , Angquist LH et al. Cohort Profile: The Danish Conscription Database (DCD): A cohort of 728 160 men born from 1939 through 1959 . Int J Epidemiol 2015 ; 44 : 432 – 40 . Google Scholar Crossref Search ADS PubMed 11 Teasdale TW. The Danish draft board's intelligence test, Borge Priens Prove: psychometric properties and research applications through 50 years . Scand J Psychol 2009 ; 50 : 633 – 38 . Google Scholar Crossref Search ADS PubMed 12 Mortensen EL , Reinisch JM , Teasdale TW. Intelligence as measured by the WAIS and a military draft board group test . Scand J Psychol 1989 ; 30 : 315 – 18 . Google Scholar Crossref Search ADS 13 Christensen J , Vestergaard M , Olsen J , Sidenius P. Validation of epilepsy diagnoses in the Danish National Hospital Register . Epilepsy Res 2007 ; 75 : 162 – 70 . Google Scholar Crossref Search ADS PubMed 14 St Germaine-Smith C , Liu M , Quan H , Wiebe S , Jette N. Development of an epilepsy-specific risk adjustment comorbidity index . Epilepsia 2011 ; 52 : 2161 – 67 . Google Scholar Crossref Search ADS PubMed 15 Deary IJ , Batty GD. Cognitive epidemiology . J Epidemiol Community Health 2007 ; 61 : 378 – 84 . Google Scholar Crossref Search ADS PubMed 16 Geerlings RP , Aldenkamp AP , Gottmer-Welschen LM et al. Developing from child to adult: Risk factors for poor psychosocial outcome in adolescents and young adults with epilepsy . Epilepsy Behav 2015 ; 51 : 182 – 90 . Google Scholar Crossref Search ADS PubMed 17 Camfield PR , Camfield CS. What happens to children with epilepsy when they become adults? Some facts and opinions . Pediatr Neurol 2014 ; 51 : 17 – 23 . Google Scholar Crossref Search ADS PubMed 18 Bourgeois BF , Prensky AL , Palkes HS , Talent BK , Busch SG. Intelligence in epilepsy: a prospective study in children . Ann Neurol 1983 ; 14 : 438 – 44 . Google Scholar Crossref Search ADS PubMed 19 Kjeldsen MJ , Kyvik KO , Christensen K , Friis ML. Genetic and environmental factors in epilepsy: A population-based study of 11,900 Danish twin pairs . Epilepsy Res 2001 ; 44 : 167 – 78 . Google Scholar Crossref Search ADS PubMed 20 Kee VR , Gilchrist B , Granner MA , Sarrazin NR , Carnahan RM. A systematic review of validated methods for identifying seizures, convulsions, or epilepsy using administrative and claims data . Pharmacoepidemiol Drug Saf 2012 ; 21(Suppl 1) : 183 – 93 . Google Scholar Crossref Search ADS PubMed 21 Baftiu A , Johannessen Landmark C , Rusten IR , Feet SA , Johannessen SI , Larsson PG. Changes in utilisation of antiepileptic drugs in epilepsy and non-epilepsy disorders - a pharmacoepidemiological study and clinical implications . Eur J Clin Pharmacol 2016 ; 72 : 1245 – 54 . Google Scholar Crossref Search ADS PubMed © The Author(s) 2018; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Journal

International Journal of EpidemiologyOxford University Press

Published: Aug 1, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off