Advertisement

A longitudinal cohort study of childhood MMR vaccination and seizure disorder among American children

Open AccessPublished:September 24, 2020DOI:https://doi.org/10.1016/j.braindev.2020.09.001

      Abstract

      Background

      Measles (rubeola) is a highly contagious infectious disease with significant morbidity/mortality. Measles-Mumps-Rubella (MMR) is a live-attenuated vaccine used in the United States (US) to prevent measles. This retrospective longitudinal cohort study evaluated childhood MMR vaccination and the risk of a seizure episode and seizure disorder.

      Methods

      The Independent Healthcare Research Database (IHRD) composed of records prospectively generated from Florida Medicaid was analyzed using SAS to identify persons continuously enrolled from birth for 120 months. Two cohorts were examined: 23,486 persons received at least one dose of MMR vaccine between 12 and 17 months (vaccinated) and 41,725 persons not receiving a measles-containing vaccine (unvaccinated). The daily incidence rate of an initial seizure episode (ICD-9 code: 780.3x) and seizure disorder (ICD-9 code: 345.xx) following an initial seizure episode diagnoses were examined using Cox proportional hazards ratio (HR) and time-trend models post-MMR vaccination compared to unvaccinated persons and in a self-controlled case-series (SCCS).

      Results

      The daily incidence rate of an initial seizure episode diagnosed from 6 to 11 days post-MMR vaccination in comparison to 12 to17 months among unvaccinated persons was significantly increased (unadjusted HR = 5.73, p < 0.0001 and adjusted HR = 5.94, p < 0.0001) in HR models. The daily incidence rate of an eventual seizure disorder diagnosis among those diagnosed with an initial seizure episode from 6 to 11 days post-MMR vaccination was significantly increased (unadjusted HR = 17.7, p < 0.01 and adjusted HR = 17.4, p < 0.01) in comparison to the daily incidence rate of an eventual seizure disorder diagnosis among those diagnosed with an initial seizure episode from 12 to 17 months among unvaccinated persons. Time-trend analyses revealed a significantly increased rate ratio (RR) for an initial seizure episode (RR = 4.64, p < 0.0001) and seizure disorder (RR = 5.51, p < 0.0001) diagnoses. Time-trend SCCS analyses revealed a significantly increased daily incidence rate of an initial seizure episode (RR = 3.80, p < 0.0001) when comparing periods from 6 to 11 days post-MMR vaccination to 49–60 days post-MMR vaccination. The incidence rate of an eventual seizure disorder diagnosis among those with an initial seizure episode diagnosis from 6 to 11 days post-MMR vaccination compared to 49–60 days post-MMR vaccination was significantly increased (RR = 4.15, p < 0.01).

      Conclusion

      Seizure episode and seizure disorder are rare consequences of routine childhood MMR vaccination.

      Keywords

      1. Background

      As described by the United States (US) Centers for Disease Control and Prevention (CDC), measles (rubeola) is highly contagious (90% of exposed susceptible persons develop measles) rash illness [

      McLean HQ, Fiebelkorn AP, Temte JL, Wallace GS; Centers for Disease Control and Prevention. Prevention of measles, rubella, congenital rubella syndrome and mumps, 2013: summary recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep 2013;62(RR-04):1-34.

      ]. It was estimated on an annual basis in the US prior to the introduction of routine measles vaccination that there were about 500,000 reported cases of measles. Among those cases of measles, an estimated 500 persons died, 48,000 were hospitalized, and 1,000 persons developed encephalitis with permanent brain damage [
      • Bloch A.B.
      • Orenstein W.A.
      • Stetler H.C.
      • Wassilak S.G.
      • Amler R.W.
      • Bart K.J.
      • et al.
      Health impact of measles vaccination in the United States.
      ].
      Since, the late 1980s/early 1990s, the Advisory Committee on Immunization Practices (ACIP), the American Academy of Pediatrics (AAP), and American Academy of Family Practitioners (AAFP) have recommended that the first dose of measles-mumps-rubella (MMR) vaccine should be given to children aged 12 through 15 months [

      McLean HQ, Fiebelkorn AP, Temte JL, Wallace GS; Centers for Disease Control and Prevention. Prevention of measles, rubella, congenital rubella syndrome and mumps, 2013: summary recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep 2013;62(RR-04):1-34.

      ]. The MMR vaccine used in the US is the M−M−R® II vaccine (Merck & Co, Inc, Whitehouse Station, NJ) [

      MERCK & CO., INC., M-M-R-II, Measles Mumps, and Rubella Virus Vaccine Live. http://www.merck.com/product/usa/pi_circulars/m/mmr_ii/mmr_ii_pi.pdf. Revised 5/2017; Accessed on July 18, 2018.

      ].
      Comprehensive reviews of the safety of MMR vaccines were undertaken by investigators affiliated with the Cochrane Review [

      Demicheli V, Rivetti A, Debalini MG, Di Pietrantonj C. Vaccines for measles, mumps and rubella in children. Cochrane Database Syst Rev 2012:CD004407.

      ,

      Di Pietrantonj C, Rivetti A, Marchione P, Debalini MG, Demicheli V. Vaccines for measles, mumps, rubella, and varicella in children. Cochrane Database Syst Rev 2020;4:CD004407.

      ]. A meta-analysis undertaken by these researchers of published studies revealed that the onset of febrile seizures from 6 to 11 days post-MMR vaccination among children 12 to 23 months was significantly increased (relative incidence = 4.09, 95% confidence interval = 3.1 to 5.33). Despite this observation, the researchers described that the design and reporting of pre- and post-marketing safety studies of MMR vaccine was inadequate, and future studies should be undertaken to further examine the safety of MMR vaccines.
      As a result of the purported inadequacies of previous MMR vaccine safety studies, the purpose of the present retrospective longitudinal cohort study was to examine the potential relationship between childhood MMR vaccination and the risk of diagnosed seizure episodes and seizure disorders in the US during the 1990s/2000s by examining prospectively collected healthcare records.

      2. Methods

      2.1 Independent healthcare research Database (IHRD)

      The Independent Healthcare Research Database (IHRD) is composed of non-identifiable healthcare records generated from the Florida Medicaid system. The data contained within the IHRD were obtained from the Agency for Health Care Administration (AHCA) of the state of Florida and included eligibility and claim files. It is possible to link a person’s eligibility and claim records by a unique recipient identifier code. The eligibility records included detailed information for each person regarding their month and year of enrollment, gender, date of birth, and county level of residency. The claim records included detailed information for each person regarding their diagnosis status using the International Code for Disease, 9th revision (ICD-9) codes, healthcare procedure codes (medical, dental, etc.), and administered drugs using National Drug Codes (NDC). The data in the IHRD were assembled and accessed under approval by the Liberty Institutional Review Board (IRB) (Deland, FL). The SAS system for Windows, version 9.4 (Cary, NC, USA) was used to examine the IHRD.

      2.2 Study participants

      Fig. 1 presents a schematic flowchart of the IHRD data examined in the present study. A cohort of 8,440,941 persons of all ages with no changes or missing genders or dates of birth and eligible at specific times for Florida Medicaid from July 1990 through June 2009 was initially evaluated in this study. Among this cohort, a total of 1,871,728 persons were eligible for Florida Medicaid from their date of birth and among those persons a total of 193,453 persons were continuously eligible for Florida Medicaid for 120 months following birth (this is the study period examined in the current study). Finally, among the cohort of 193,453 persons continuously eligible for Florida Medicaid for 120 months following birth, a sub-cohort of 101,736 persons with ≥10 outpatient office visits during the 120 month period following birth was identified.
      Figure thumbnail gr1
      Fig. 1A schematic flowchart of the data examined in the present study.

      2.3 Vaccination status

      The vaccination status variable examined was identified from the healthcare procedure codes filed on claims for each cohort member examined. The procedure codes examined, included: measles vaccination (codes: 9945, W1941, 90705), MMR vaccine (codes: 90707, W1943, 9948), measles and rubella vaccine (code: 90708), measles, mumps, rubella, and varicella (MMRV) vaccine (code: 90710). Only persons receiving at least a single dose of MMR vaccine were included in the vaccinated cohort (codes: 90707, W1943, or 9948) and persons were considered unvaccinated, if they did not receive any measles-containing vaccine (codes: 9945, W1941, 90705, 90707, W1943, 9948, 90708, or 90710) within the first 120 months of life after birth. Among those vaccinated with MMR, the date of service for the first claim in chronological order with a procedure code specifying MMR vaccine was assumed to be the date of vaccine administration, and only persons receiving their first dose of MMR vaccine between 12 through 16 months of age were included in the present study (a total of 8,279 persons receiving their first dose of MMR vaccine outside of the age ranges examined were excluded from this study). Overall, 24,637 persons were in the MMR vaccinated cohort and 43,538 persons were in the unvaccinated cohort.
      It would be assumed that most American children born in the 1990s received their childhood MMR vaccination in accordance with the ACIP, AAP, and AAFP recommendations to routinely administered MMR vaccination between the ages of 12 through 15 months [

      McLean HQ, Fiebelkorn AP, Temte JL, Wallace GS; Centers for Disease Control and Prevention. Prevention of measles, rubella, congenital rubella syndrome and mumps, 2013: summary recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep 2013;62(RR-04):1-34.

      ]. The Medicaid system is different because there are substantial numbers of hard-to-reach under-immunized low-income infants and preschool aged children. During the 1990s, the US CDCand state health agencies mounted aggressive efforts to immunize increased numbers of children in this age group [
      • Hakim R.B.
      • Boben P.J.
      • Bonney J.B.
      Medicaid and the health of children.
      ]. As a result, the present study is relatively unique because it was possible to examine large cohorts of vaccinated/unvaccinated American children in the same medical system.

      2.4 Outcomes

      The outcome variable examined in this study was identified from the ICD-9 codes filed on claims for each cohort member examined. All persons diagnosed with an initial seizure episode (code: 780.3x) were examined. No information was available regarding the methods utilized to confirm seizure episode diagnoses. Persons were considered to have a seizure episode if they had any seizure episode diagnoses and persons were not considered to have seizures if they did not have any seizure diagnoses within the first 120 months of life after birth. Among those with a seizure episode diagnosis, the date of service for the first claim in chronological order with a seizure episode diagnosis was assumed to be the date of onset. An initial seizure episode was considered to be an acute event with unknown long-term sequelae.
      In order to evaluate the potential for an initial seizure episode to manifest as a chronic, long-term condition with sequelae, persons diagnosed with an initial seizure episode followed by a seizure disorder diagnosis (code: 345.xx) were examined. No information was available regarding the methods utilized to confirm seizure disorder diagnoses. Persons were considered to have a seizure disorder, if they had any seizure disorder diagnoses, and persons were not considered to have a seizure disorder, if they did not have any seizure disorder diagnoses within the first 120 months of life after birth. Among those with a seizure disorder diagnosis, the date of service for the first claim in chronological order with an initial seizure episode diagnosis was assumed to be the initial date for the manifestations of symptoms that could ultimately be associated with the seizure disorder diagnosis. Only persons diagnosed with a seizure disorder after an initial seizure episode diagnosis were considered in this study.
      A total of 779 persons were removed from the study because they were diagnosed with a seizure disorder before their seizure episode diagnosis. In the unvaccinated cohort, a total of 1,331 persons diagnosed with a seizure episode before 12 months were removed. Thus, the overall size of the unvaccinated cohort was reduced to 41,725 persons. In the MMR vaccinated cohort, a total of 1,151 persons were diagnosed with a seizure episode or a seizure disorder before receipt of their first dose of MMR vaccination administered between 12 months through 16 months of age. Thus, the overall size of the vaccinated cohort was reduced to 23,486 persons.

      2.5 Statistical analyses

      In all statistical analyses, the statistical package in SAS or StatsDirect, version 3.3.3 (Merseyside, UK) was utilized, and a two-sided p-value < 0.05 was considered statistically significant. The null hypothesis was that MMR vaccination would have no impact on the incidence rate of diagnosed seizure episodes or seizure disorders.
      In order to evaluate the relationship between MMR vaccination and a seizure episode diagnosis, person-days of follow-up were calculated in the vaccinated and unvaccinated cohorts examined in this study. In the unvaccinated cohort, person-days of follow-up began at 12 months of age and were continued through 16 months of age (a maximum of 152 days) or until the date of first seizure episode diagnosis. In the vaccinated cohort, in keeping with previous studies showing the clustering of the onset of a seizure episode from 6 to 11 days post-MMR vaccination, person-days of follow-up began on day 7 post-MMR vaccination and continued to day 11 post-MMR vaccination (a maximum of 4 days) or until the date of a first seizure episode diagnosis. If a seizure episode diagnosis was made from the day of MMR vaccination through the 6th day post-MMR vaccination, these persons were determined to have contributed no person-days of follow-up. A similar analysis was undertaken for seizure disorder diagnoses, except for those with a seizure episode diagnosis and a subsequent seizure disorder diagnosis, the date of onset of seizure disorder was assumed to be the first date of a seizure episode diagnosis.
      A regression analysis of a diagnosed seizure episode or seizure disorder based on the Cox proportional hazards model was used to evaluate over time in days the potential relationship between MMR vaccination and the outcomes examined in this study. Ties in the failure times were handled using the exact method. All models were constructed without adjustment for covariates (Model I) and with adjustment for the covariates of gender (categorical variable), date of birth (continuous variable), and county of residence (continuous variable) (Model II).
      Person-time modeling was used to evaluate the rate over time of a diagnosed seizure episode or seizure disorder among those in the MMR vaccinated cohort as compared to the unvaccinated cohort. The same methods were employed for hazard models. Namely, in the unvaccinated cohort, person-days of follow-up began at 12 months of age and were continued through 16 months of age (a maximum of 152 days) or until the date of a first seizure episode diagnosis. Overall, there were a total of 6,311,571 person-days in the unvaccinated cohort. In the vaccinated cohort, person-days of follow-up began on day 7 post-MMR vaccination and continued to day 11 post-MMR vaccination (a maximum of 4 days) or until the date of a first seizure episode diagnosis. If a seizure episode diagnosis was made from the day of MMR vaccination through the 6th day post-MMR vaccination, these persons were determined to have contributed no person-days of follow-up. Overall, there were a total of 93,849 person-days in the MMR vaccinated cohort.
      Finally, person-time modeling was utilized to evaluate within the MMR vaccinated cohort a self-controlled case-series for the rate over time of a diagnosed seizure episode or seizure disorder with an initial onset from 6 to 11 days-post MMR vaccination in comparison to onset from 49 to 60 days post-MMR vaccination (a maximum of 10 days). When examining the 49 to 60 days post-MMR vaccination, if a seizure episode diagnosis was made from the day of MMR vaccine through the 49th day post-MMR vaccination, these persons were determined to have contributed no person-days of follow-up. Overall, there were a total of 233,581 person-days in the MMR vaccinated cohort when examining a seizure episode or seizure disorder with onset from 49 to 60 days post-MMR vaccination.

      3. Results

      Table 1 displays the demographic characteristics of the population of persons examined. Overall, there were a total of 23,486 persons in the MMR vaccinated cohort contributing a total of 93,849 person-days and 41,725 persons in the unvaccinated cohort contributing a total of 6,311,571 person-days. The gender distribution was similar in the MMR vaccinated cohort (male/female ratio: 1.12) and unvaccinated cohort (male/female ratio = 1.14). The mean age of administration of the first dose of MMR was 14.26 months (range: 12 through 16 months) in the MMR vaccinated cohort. The overall mean of examination was 14.49 months (range: 12 through 16 months) in the unvaccinated cohort. The overall mean date of birth in the MMR vaccinated cohort (1996) was significantly (p < 0.0001) different than the unvaccinated cohort (1994).
      Table 1Demographic characteristics of the persons examined in this study
      All persons examined in this study were enrolled from their date of birth for 120 consecutive months. All persons had non-changing dates of birth and gender status. All persons had ≥ 10 outpatient office visits.
      Parameter ExaminedMMR Vaccinated Cohort
      Persons received their first dose of MMR vaccine from 12 months through 16 months of age.


      (n = 23,486)
      UnvaccinatedCohort
      Persons received no doses of any measles-containing vaccine.


      (n = 41,725)
      Gender (%)
      Male12,402 (52.81%)22,194 (53.19%)
      Female11,084 (47.19%)19,531(46.81%)
      Date of Birth
      mean ± std (range)1996 ± 2.5(1990–1999)1994 ± 2.5(1990–1999)
      Age of Examination (months)
      mean (range)14.49(12–17)
      Age of Vaccination (months)
      mean ± std (range)14.26 ± 1.5(12–17)
      Number Diagnosed with a Seizure Episode

      (ICD-9 Code: 7803x)
      29 (0.12%)
      Persons were examined for onset of a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination.
      420 (1.01%)
      Persons were examined for a seizure episode diagnoses with an initial onset from 12 months through 16 months of age.
      Number Diagnosed with a Seizure Disorder

      (ICD-9 Code: 345xx)
      10 (0.04%)
      Persons were examined for a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination and subsequent diagnosis of a seizure disorder within the first 10 years of life.
      122 (0.29%)
      Persons were examined for a seizure episode diagnoses with an initial onset from 12 months through 16 months of age and subsequent diagnosis of seizure disorder within the first 10 years of life.
      Number of Person-Days93,849
      Person-days were from day 6 to 11 post their first dose of MMR. If a person was diagnosed with a seizure episode, person-days were counted from day 6 until day of diagnosis.
      6,311,571
      Person-days were from 12 months through 16 months of age. If a person was diagnosed with a seizure episode, person-days were counted from 12 months of age until day of diagnosis.
      ICD-9 = International Code of Disease, 9th revision, MMR = measles-mumps-rubella, std = standard deviation.
      1 All persons examined in this study were enrolled from their date of birth for 120 consecutive months. All persons had non-changing dates of birth and gender status. All persons had ≥ 10 outpatient office visits.
      2 Persons received their first dose of MMR vaccine from 12 months through 16 months of age.
      3 Persons received no doses of any measles-containing vaccine.
      4 Persons were examined for onset of a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination.
      5 Persons were examined for a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination and subsequent diagnosis of a seizure disorder within the first 10 years of life.
      6 Persons were examined for a seizure episode diagnoses with an initial onset from 12 months through 16 months of age.
      7 Persons were examined for a seizure episode diagnoses with an initial onset from 12 months through 16 months of age and subsequent diagnosis of seizure disorder within the first 10 years of life.
      8 Person-days were from day 6 to 11 post their first dose of MMR. If a person was diagnosed with a seizure episode, person-days were counted from day 6 until day of diagnosis.
      9 Person-days were from 12 months through 16 months of age. If a person was diagnosed with a seizure episode, person-days were counted from 12 months of age until day of diagnosis.
      Table 2 shows the demographic characteristics of the 449 persons diagnosed with a seizure episode examined. Overall, more males than females (male/female ratio = 1.30) were diagnosed with a seizure episode, but the increased male/female ratio was much less pronounced in the MMR vaccinated cohort (male/female ratio = 1.07) than the unvaccinated cohort (male/female ratio = 1.32). The mean date of birth and age at diagnosis among persons diagnosed with a seizure episode was similar in both cohorts examined. Among the types of seizure episode diagnoses examined, Table 2 reveals that most initial seizure episode diagnoses were made in unspecified diagnostic categories (780.3 or 780.39). However, it was observed that the percentage of febrile seizure episode diagnoses were significantly more common among those children diagnosed with a seizure episode in the MMR vaccinated cohort as compared to those children diagnosed with a seizure episode in the unvaccinated cohort (risk ratio = 2.52, 95% confidence interval = 1.28 to 4.52, p < 0.05).
      Table 2Demographic summary of the persons diagnosed with a seizure episode examined in this study
      All persons examined in this study were enrolled from their date of birth for 120 consecutive months. All persons had non-changing dates of birth and gender status. All persons had ≥ 10 outpatient office visits.
      Parameter ExaminedAll Cohorts CombinedMMR Vaccinated Cohort
      Persons received their first dose of MMR vaccine from 12 through 16 months of age and were diagnosed with a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination.
      Unvaccinated Cohort
      Persons received no doses of measles-containing vaccine and a seizure episode diagnosis with an initial onset from 12 months through 16 months of age.
      Gender (%)
      Male254 (56.57%)15 (51.72%)239 (56.90%)
      Female195 (43.43%)14 (48.28%)181 (43.10%)
      Date of Birth
      mean ± std (range)1994 ± 2.5

      (1990–1999)
      1995 ± 2.7

      (1991–1999)
      1994 ± 2.5

      (1990–1999)
      Age at Diagnosis in Years
      mean ± std (range)1.21 ± 0.12

      (1.00–1.43)
      1.16 ± 0.11

      (1.05–1.43)
      1.21 ± 0.12

      (1.00–1.41)
      Year of Diagnosis
      mean ± std (range)1995 ± 2.53

      (1991–2001)
      1997 ± 2.68

      (1992–2000)
      1995 ± 2.50

      (1991–2001)
      Seizure Episode Diagnostic Codes (ICD-9)
      Seizure (780.3)359 (79.96%)19 (65.52%)340 (80.95%)
      Febrile Seizure (780.31)54 (12.03%)8 (27.59%)46 (10.95%)
      Other Seizure (780.39)36 (8.02%)2 (6.90%)34 (8.10%)
      ICD-9 = International Code of Disease, 9th revision, MMR = measles-mumps-rubella, std = standard deviation.
      1 All persons examined in this study were enrolled from their date of birth for 120 consecutive months. All persons had non-changing dates of birth and gender status. All persons had ≥ 10 outpatient office visits.
      2 Persons received their first dose of MMR vaccine from 12 through 16 months of age and were diagnosed with a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination.
      3 Persons received no doses of measles-containing vaccine and a seizure episode diagnosis with an initial onset from 12 months through 16 months of age.
      Table 3 examines the demographic characteristics of the 132 persons diagnosed with a seizure disorder. A non-significant increase in the mean age at diagnosis of seizure disorder was observed in the unvaccinated cohort (2.63 years-old) as compared to the MMR vaccinated cohort (1.96 years-old) and a non-significant increase in the mean onset of a seizure disorder diagnosis after a seizure episode diagnosis in the unvaccinated cohort (521 days) as compared to the MMR vaccinated cohort (274 days) was also observed. The percentage of febrile seizure episode diagnoses were significantly more common among those children diagnosed with a seizure disorder in the MMR vaccinated cohort as compared to those children diagnosed with a seizure disorder in the unvaccinated cohort (risk ratio = 4.44, 95% confidence interval = 1.62 to 10.2, p < 0.05).
      Table 3Demographic summary of the persons diagnosed with a seizure disorder examined in this study
      All persons examined in this study were enrolled from their date of birth for 120 consecutive months. All persons had non-changing dates of birth and gender status. All persons had ≥ 10 outpatient office visits.
      Parameter ExaminedAll Cohorts CombinedMMR Vaccinated Cohort
      Persons received their first dose of MMR vaccine from 12 through 16 months of age. Persons with a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination and subsequent diagnosis of seizure disorder within the first 10 years of life.
      Unvaccinated Cohort
      Persons received no doses of measles-containing vaccine. Persons with a seizure episode diagnosis with an initial onset from 12 months through 16 months of age and subsequent diagnosis of seizure disorder within the first 10 years of life.
      Gender (%)
      Male75 (56.82%)5 (50.00%)70 (57.38%)
      Female57 (43.18%)5 (50.00%)52 (42.62%)
      Date of Birth
      mean ± std (range)1994 ± 2.7

      (1990–1999)
      1996 ± 2.8

      (1991–1996)
      1994 ± 2.6

      (1990–1999)
      Age at Diagnosis in Years
      mean ± std (range)2.58 ± 1.96

      (1.02–9.94)
      1.96 ± 1.67

      (1.08–6.60)
      2.63 ± 1.98

      (1.02–9.94)
      Year of Diagnosis
      mean ± std (range)1996 ± 2.65

      (1992–2001)
      1997 ± 2.74

      (1993–2000)
      1996 ± 2.63

      (1992–2001)
      Onset in Days After Seizure Diagnosis
      mean ± std (range)502 ± 713

      (0–3,229)
      274 ± 582

      (0–1,889)
      521 ± 721

      (0–3,229)
      Seizure Disorder Diagnostic Codes (ICD-9)
      Generalized nonconvulsive epilepsy (345.0x)11 (8.33%)1 (10.00%)10 (8.20%)
      Generalized convulsive epilepsy (345.1x)26 (19.70%)1 (10.00%)25 (20.49%)
      Grand mal status (345.3x)22 (16.67%)3 (30.00%)19 (15.57%)
      Localization-related (focal) (partial) epilepsy and epileptic syndromes with complex partial seizures (345.4x)15 (11.36%)0 (0%)15 (12.29%)
      Localization-related (focal) (partial) epilepsy and epileptic syndromes with simple partial seizures (345.5x)6 (4.55%)1 (10.00%)5 (4.10%)
      Infantile spams (345.6x)1 (0.76%)0 (0%)1 (0.82%)
      Epilepsia partialis continua (345.7x)2 (1.51%)0 (0%)2 (1.64%)
      Other forms of epilepsy and recurrent seizures (345.8x)1 (0.76%)0 (0%)1 (0.82%)
      Epilepsy, unspecified (345.9x)48 (36.36%)4 (40.00%)44 (36.07%)
      Seizure Episode Diagnostic Codes (ICD-9)
      Seizure (780.3)99 (75.00%)5 (50.00%)94 (77.05%)
      Febrile Seizure (780.31)15 (11.36%)4 (40.00%)11 (9.02%)
      Other Seizure (780.39)18 (13.64%)1 (10.00%)17 (13.93%)
      ICD-9 = International Code of Disease, 9th revision, MMR = measles-mumps-rubella, std = standard deviation.
      1 All persons examined in this study were enrolled from their date of birth for 120 consecutive months. All persons had non-changing dates of birth and gender status. All persons had ≥ 10 outpatient office visits.
      2 Persons received their first dose of MMR vaccine from 12 through 16 months of age. Persons with a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination and subsequent diagnosis of seizure disorder within the first 10 years of life.
      3 Persons received no doses of measles-containing vaccine. Persons with a seizure episode diagnosis with an initial onset from 12 months through 16 months of age and subsequent diagnosis of seizure disorder within the first 10 years of life.
      Table 4 reveals the Cox proportional hazards model results examining the impact of childhood MMR vaccination on the incidence rate of a diagnosed seizure episode or seizure disorder as compared to unvaccinated persons. The risk of being diagnosed with an initial seizure episode from 6 to 11 days post-MMR vaccination in comparison to not being vaccinated with MMR at a similar age was associated with a significantly elevated hazard ratio = 5.73 in the unadjusted model and hazard ratio = 5.94 in the adjusted model. Similarly, the risk of eventually being diagnosed with a seizure disorder after an initial seizure episode diagnosed from 6 to 11 days post-MMR vaccination was associated with significantly increased (hazard ratio = 17.7 in the unadjusted model and hazard ratio = 17.4 in the adjusted model). None of the covariates examined were significantly related to the outcomes of seizure episode or seizure disorder.
      Table 4Cox proportional hazards model results examining the relationship between MMR vaccination and the outcomes examined with an initial onset from 6 to 11 days post-MMR vaccination.
      ModelOutcomeVariableHazard Ratio

      (95% CI)
      p-valueχ
      Persons received no doses of measles-containing vaccine.
      I
      Seizure EpisodeVaccinated
      Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      (Initial Diagnosis from 6 to 11 Days Post-Vaccination)

      vs

      Unvaccinated
      Persons received no doses of measles-containing vaccine.


      (Initial Diagnosis from 12 through 16 Months of Age)
      5.73

      (2.71 to 12.1)
      < 0.000120.9
      Seizure DisorderVaccinated
      Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      (Seizure Disorder Diagnosed following an initial Seizure Episode Diagnosed from 6 to 11 Days Post-Vaccination)

      vs

      Unvaccinated
      Persons received no doses of measles-containing vaccine.


      (Seizure Disorder Diagnosed following an initial Seizure Episode Diagnosed from 12 through 16 Months of Age)
      17.7

      (2.27 to 138)
      0.00617.53
      II
      Seizure EpisodeVaccinated
      Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      (Initial Diagnosis from 6 to 11 Days Post-Vaccination)

      vs

      Unvaccinated
      Persons received no doses of measles-containing vaccine.


      (Initial Diagnosis from 12 through 16 Months of Age)
      5.94

      (2.81 to 12.58)
      < 0.000121.7
      Gender (Female vs Male)0.87

      (0.72 to 1.054)
      0.161.98
      County of Residence1.00

      (0.99 to 1.005)
      0.880.02
      Date of Birth1.00

      (1.00 to 1.00)
      0.181.84
      Seizure DisorderVaccinated
      Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      (Seizure Disorder Diagnosed following an initial Seizure Episode Diagnosed from 6 to 11 Days Post-Vaccination)

      vs

      Unvaccinated
      Persons received no doses of measles-containing vaccine.


      (Seizure Disorder Diagnosed following an initial Seizure Episode Diagnosed from 12 through 16 Months of Age)
      17.4

      (2.23to 136)
      0.00647.42
      Gender (Female vs Male)0.86

      (0.61 to 1.22)
      0.400.71
      County of Residence0.998

      (0.99 to 1.007)
      0.600.27
      Date of Birth1.00

      (1.00 to 1.00)
      0.800.06
      Italicized results are statistically significant. Model I = unadjusted, Model II = adjusted for gender, county of residence, and date of birth.
      CI = confidence interval
      1 Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      2 Persons received no doses of measles-containing vaccine.
      Fig. 2 is a Cox proportional hazards survival plot evaluating the incidence rate of an initial seizure episode diagnoses over the period persons were followed in the MMR vaccinated cohort and the unvaccinated cohort. The plot reveals that during the 4 day follow-up period post-MMR vaccination (from 6 to 11 days post-MMR vaccination) there were a greater number of initial seizure episode diagnoses than in the unvaccinated cohort. The overall incidence of an initial seizure episode diagnosis during the 152 follow-up days (between 12 months through 16 months of age) in the unvaccinated cohort remained relatively constant. Fig. 3 is a Cox proportional hazards survival plot evaluating the incidence rate of seizure disorder diagnoses following an initial seizure episode diagnosis over the period persons were followed in the MMR vaccinated cohort and the unvaccinated cohort. For those diagnosed with seizure disorder after an initial seizure episode diagnosis during the first 10 years of life, the age of initial seizure disorder onset was considered to be the age of initial seizure episode diagnosis. The plot shows during the 4 day follow-up period post-MMR vaccination (from 6 to 11 days post-MMR for initial seizure diagnosis) there were a greater number of initial onsets of seizure disorder diagnoses than in the unvaccinated cohort. The overall incidence rate of initial seizure disorder diagnoses was lower in the first 25 days of follow-up in the unvaccinated cohort than in the subsequent period between 25 days to 152 days. During the latter period, the incidence rate of initial seizure disorder diagnoses remained relatively constant. As a result, the hazard ratios observed in Table 4 for seizure disorder diagnoses in the MMR vaccinated cohort as compared to the unvaccinated cohort, were significantly increased.
      Figure thumbnail gr2
      Fig. 2A summary of an initial seizure episode diagnosis by days of follow-up in the MMR vaccinated cohort1 in comparison to the unvaccinated cohort2. 1 Persons were examined for a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination. 2 Persons were examined for a seizure episode diagnosis with an initial onset from 12 months through 16 months of age.
      Figure thumbnail gr3
      Fig. 3A summary of seizure disorder diagnoses by days of follow-up in the MMR vaccinated cohort1 in comparison to the unvaccinated cohort2. 1 Persons were examined for a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination and subsequently diagnosed with a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be age of initial seizure episode diagnosis). 2 Persons were examined for a seizure episode diagnosis with an initial onset from 12 months through 16 months of age and subsequently diagnosed with a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be age of initial seizure episode diagnosis).
      Table 5 presents person-time models examining the frequency of the outcomes evaluated in the MMR vaccinated as compared to the unvaccinated cohort. The incidence rate for an initial seizure episode diagnosis (rate ratio = 4.64, p < 0.0001) was significantly higher from 6 to 11 days post-MMR vaccination in comparison to the incidence rate for an initial seizure episode diagnosis made between the ages of 12 through 16 months in the unvaccinated cohort. It was also observed that the incidence rate for eventual seizure disorder diagnoses among those with an initial seizure episode diagnosis made in the 6 to 11 days post-MMR vaccination in comparison to the rate of eventual seizure disorder diagnoses among those with an initial seizure episode diagnosis made between the ages of 12 through 16 months (rate ratio = 5.51, p < 0.0001) were significantly higher from 6 to 11 days post-MMR vaccination in comparison to the unvaccinated cohort.
      Table 5Person-time model results examining the frequency of the outcomes examined in the MMR vaccinated cohort in comparison to the unvaccinated cohort.
      Parameter ExaminedMMR Vaccinated Cohort
      Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.


      (initial onset from 6 to 11 days post MMR vaccination)
      Unvaccinated Cohort
      Persons received no doses of measles-containing vaccine.


      (initial onset from 12 through 16 months of age)
      Seizure Episode
      Number Diagnosed with the Outcome29
      Persons were examined for a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination.
      420
      Persons were examined for a seizure episode diagnosis with an initial onset from 12 months through 16 months of age.
      Total Number of Person-Days93,8496,311,571
      Rate (per 100,000 person-days)30.96.654
      Rate Ratio (95% CI)4.64 (3.07 to 6.77)
      Rate Difference (per 100,000 person-days)

      (95% CI)
      24.2 (18.9 to 29.6)
      p-value< 0.0001
      Seizure Disorder
      Number Diagnosed with the Outcome10
      Persons were examined for a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination and a subsequent diagnosis of a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be age of an initial seizure episode diagnosis).
      122
      Persons were examined for a seizure episode with an initial onset from 12 months through 16 months of age and a subsequent diagnosed of a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be the age of an initial seizure episode diagnosis).
      Total Number of Person-Days93,8496,311,571
      Rate (per 100,000 person-days)10.661.933
      Rate Ratio (95% CI)5.51 (2.58 to 10.5)
      Rate Difference (per 100,000 person-days)

      (95% CI)
      8.7 (5.8 to 11.6)
      p-value< 0.0001
      Italicized results are statistically significant. CI = confidence interval, MMR = measles-mumps-rubella,
      1 Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      2 Persons received no doses of measles-containing vaccine.
      3 Persons were examined for a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination.
      4 Persons were examined for a seizure episode diagnosis with an initial onset from 12 months through 16 months of age.
      5 Persons were examined for a seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination and a subsequent diagnosis of a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be age of an initial seizure episode diagnosis).
      6 Persons were examined for a seizure episode with an initial onset from 12 months through 16 months of age and a subsequent diagnosed of a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be the age of an initial seizure episode diagnosis).
      Table 6 reveals person-time models examining the frequency of the outcomes examined in a self-controlled case-series in the MMR vaccinated cohort. The incidence rate of initial seizure episode diagnoses were significantly higher from the 6 to 11 days post-MMR vaccination period as compared to 49 to 60 days post-MMR vaccination period (rate ratio = 3.80, p < 0.0001). Similarly, the incidence rate of eventual seizure disorder diagnoses among those diagnosed with an initial seizure episode from 6 to 11 days post-MMR vaccination in comparison to 49 to 60 days post-MMR vaccination (rate ratio = 4.15, p < 0.01).
      Table 6Person-time model results examining the frequency of the outcomes in a self-controlled case-series in the MMR vaccinated cohort.
      Parameter ExaminedMMR Vaccinated Cohort

      (with initial onset from 6 to 11 Days post-MMR Vaccination)
      Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      MMR Vaccinated Cohort

      (with initial onset from 49 to 60 Days post-MMR Vaccination)
      Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      Seizure Episode
      Number Diagnosed with the Outcome2919
      Total Number of Person-Days93,849233,581
      Rate (per 100,000 person-days)30.98.134
      Rate Ratio (95% CI)3.80 (2.06 to 7.17)
      Rate Difference (per 100,000 person-days)

      (95% CI)
      22.8 (13.6 to 31.9)
      p-value< 0.0001
      Seizure Disorder
      Persons were examined for subsequent diagnosis of a seizure disorder within the first 10 years of life (age of seizure disorder onset was considered to be the age of an initial seizure episode diagnosis).
      Number Diagnosed with the Outcome106
      Total Number of Person-Days93,849233,581
      Rate (per 100,000 person-days)10.662.569
      Rate Ratio (95% CI)4.15 (1.37 to 13.9)
      Rate Difference (per 100,000 person-days)

      (95% CI)
      8.1 (2.8 to 13.4)
      p-value< 0.01
      Italicized results are statistically significant. CI = confidence interval, MMR = measles-mumps-rubella.
      1 Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      2 Persons were examined for subsequent diagnosis of a seizure disorder within the first 10 years of life (age of seizure disorder onset was considered to be the age of an initial seizure episode diagnosis).

      4. Discussion

      The results of this retrospective longitudinal cohort study of prospectively collected healthcare data provide important and compelling new epidemiological quantitative data regarding the association between childhood MMR vaccine routinely administered to American children and the risk of children being diagnosed with a seizure episode and seizure disorder. This study utilized hazard ratio and person-time models with adjustment for covariates to examine vaccinated as compared to unvaccinated cohorts and a self-controlled case-series. This study also utilized long-term follow-up of the cohorts examined to determine the eventual diagnostic status of the persons examined.
      As described previously, comprehensive reviews of the safety of MMR vaccines were undertaken by investigators affiliated with the Cochrane Review [

      Demicheli V, Rivetti A, Debalini MG, Di Pietrantonj C. Vaccines for measles, mumps and rubella in children. Cochrane Database Syst Rev 2012:CD004407.

      ,

      Di Pietrantonj C, Rivetti A, Marchione P, Debalini MG, Demicheli V. Vaccines for measles, mumps, rubella, and varicella in children. Cochrane Database Syst Rev 2020;4:CD004407.

      ]. These researchers, after applying various inclusion/exclusion criteria regarding study quality and design, included in their review a total of three epidemiological studies examining the relationship between childhood MMR vaccination and the risk of seizures [
      • Vestergaard M.
      • Hviid A.
      • Madsen K.M.
      • Wohlfahrt J.
      • Thorsen P.
      • Schendel D.
      • et al.
      MMR vaccination and febrile seizures: evaluation of susceptible subgroups and long-term prognosis.
      ,
      • Ward K.N.
      • Bryant N.J.
      • Andrews N.J.
      • Bowley J.S.
      • Ohrling A.
      • Verity C.M.
      • et al.
      Risk of serious neurologic disease after immunization of young children in Britain and Ireland.
      ,
      • Miller E.
      • Andrews N.
      • Stowe J.
      • Grant A.
      • Waight P.
      • Taylor B.
      Risk of convulsion and aseptic meningitis following measles-mumps-rubella vaccination in the United Kingdom.
      ].
      Vestergaard et al. (2004) undertook a population-based cohort study of 537,171 children born in Denmark between 1991 through 1998 and who were followed-up through 1999 [
      • Vestergaard M.
      • Hviid A.
      • Madsen K.M.
      • Wohlfahrt J.
      • Thorsen P.
      • Schendel D.
      • et al.
      MMR vaccination and febrile seizures: evaluation of susceptible subgroups and long-term prognosis.
      ]. Overall, 82% of the children examined received MMR vaccination. Within the first 2 weeks after MMR vaccination, the rate ratio for febrile seizures was significantly increased by 2.75-fold as compared to unvaccinated children. After the first 2 weeks post-MMR vaccination, the rate of febrile seizures was similar in MMR vaccinated and unvaccinated children. On a long-term basis (up to 105 months of follow-up), the investigators compared the rate of recurrent febrile seizures and seizure disorder among those that experienced febrile seizures within 2 weeks of childhood MMR vaccination to those with febrile seizures among unvaccinated children. They observed that the rate ratio for recurrent febrile seizures was significantly increased in the childhood MMR vaccination cohort by 1.19-fold in comparison to the unvaccinated children, but no difference was observed in the rate of seizure disorder.
      In considering the results of the present study with those observed by Vestergaard et al. (2004) there are several similarities and differences. Among the differences between the two studies are that the present study examined a more precise onset window for initial seizure episode diagnoses post-MMR vaccination than the Vestergaard et al (2004) study (from 6 to 11 days vs 0 to 14 days). Namely, the period examined in the current study is more biologically plausible for when measles virus can negatively impact vaccine recipients (i.e., < 7 days is too short a period for measles virus incubation), and, hence, Vestergaard et al. (2004) by including a longer onset window may have weakened the association between MMR vaccination and seizure episode diagnoses. This appears to be the case because the rate ratio for seizure episode diagnoses post-MMR vaccination in the current study was from about 4 to 6-fold significantly increased, whereas in the Vestergaard et al. (2004) study the rate ratio for seizure episode diagnoses post-MMR vaccination was 2.75-fold significantly increased. Another difference between the two studies is the method utilized to examine long-term seizure disorder status. The current study examined the rate of eventual seizure disorder diagnoses among those with an initial seizure episode diagnosis in the MMR vaccinated group (with an initial seizure episode diagnosis onset from 6 to 11 days post-MMR vaccination) in comparison to either unvaccinated persons (with an initial seizure episode diagnosis onset from 12 through 16 months of age) or in the self-controlled case-series analysis (with an initial seizure episode diagnosis onset from 49 to 60 days post-MMR vaccination). Among those with an eventual seizure disorder diagnosis, it was assumed in the current study that the initial onset of seizure disorder occurred with the onset of the initial seizure episode diagnosis. Therefore, the rates were calculated based upon the specific temporal windows examined. By contrast, Vestergaard et al. (2004), evaluated the long-term rate of seizure disorder diagnoses among those with febrile seizures in the vaccinated and unvaccinated cohorts. The current study methodology revealed that the rate ratio of seizure disorder diagnoses were from about 4- to 18-fold significantly increased with onset of initial seizure episode diagnoses from 6 to 11 days post-MMR vaccination. The Vestergaard et al. (2004) study revealed no difference in the rate ratio for seizure disorder following a febrile seizure diagnosis occurring post-MMR vaccine as compared to those diagnosed with a febrile seizure in the unvaccinated cohort. This observation is consistent with the observation made in the current study that about 1/3rd of persons diagnosed with an initial seizure episode went onto be diagnosed with seizure disorder regardless of the original etiology (i.e., an initial seizure episode diagnosed post-MMR vaccination has a similar long-term risk for a seizure disorder prognosis as a seizure episode diagnosed in the absence of MMR vaccination), but, unfortunately, such an examination does not test the hypothesis of whether MMR vaccine administration is associated with seizure disorder.
      Ward et al. (2007) conducted a prospective survey of children using a self-controlled case-series method to evaluate the relationship between childhood MMR vaccination and the risk of neurological symptoms manifesting as children admitted to the hospital with fever and seizures lasting >30 min [
      • Ward K.N.
      • Bryant N.J.
      • Andrews N.J.
      • Bowley J.S.
      • Ohrling A.
      • Verity C.M.
      • et al.
      Risk of serious neurologic disease after immunization of young children in Britain and Ireland.
      ]. They described the risk period from 6 to 11 days post-MMR vaccination, and diagnoses made outside of this period were considered to be the background incidence rate. These investigators observed that the risk of the neurological symptoms examined was significantly increased by 5.68-fold from 6 to 11 days post-childhood MMR vaccination as compared to diagnoses made outside this period. A total of 6 persons with the aforementioned symptoms were identified from 6 to 11 days post-MMR vaccination, and the investigators concluded based upon examination of the symptoms reported in these persons that many met the criteria for complex febrile seizures.
      The results observed by Ward et al (2007) are consistent with those observed in the current study, but are significantly limited. The results reported by Ward et al. (2007) are limited because a very narrow definition developed by the investigators themselves to describe neurological symptoms was utilized. It is not easy to extrapolate the neurological symptoms described by the investigators to ICD-9 codes employed in routine medical care. Further, by examining only persons presenting with such neurological symptoms to specific hospital centers, and as described by the investigators themselves, it seems likely that the number of cases collected was an underestimate of the true number. As a result, it is not possible to utilize the data collected by Ward et al (2007) to estimate the true population rate of such occurrences. Finally, Ward et al. (2007) did not undertake long-term follow-up of the children examined. The investigators reported on the symptoms experienced by the children while they were in the hospital, and whether they were recovered from the symptoms at the time of discharge, but no long-term investigations were made of their eventual diagnostic status. Also, given that only 6 persons were examined, the statistical power of the study was relatively low. Based upon the information presented by the investigators, they described that out of the 6 children examined, 1 was described as not recovered at the time of discharge from the hospital (16.7%).
      Miller et al. (2007) evaluated children between 12 and 23 months of age with a hospital discharge diagnosis of febrile seizure [
      • Miller E.
      • Andrews N.
      • Stowe J.
      • Grant A.
      • Waight P.
      • Taylor B.
      Risk of convulsion and aseptic meningitis following measles-mumps-rubella vaccination in the United Kingdom.
      ]. They were admitted to the hospital between January 1, 1998 and June 30, 2002 and were linked with their MMR immunization records. A self-controlled case-series methodology was utilized to determine the frequency of febrile seizure diagnoses in the at risk period from 6 to 11 days post-MMR vaccination in comparison to other periods before or after MMR vaccination. These investigators observed a 4.09-fold significantly increased risk of diagnosed febrile seizure from 6 to 11 days post-MMR vaccination as compared to other periods. The investigators estimated that 1 in 1,150 doses resulted in febrile seizures attributably related to MMR vaccination from 6 to 11 days post-MMR vaccination.
      The results observed by Miller et al. (2007) are consistent with those observed in the current study. Similarly, in both studies, during period from 6 to 11 days post-MMR the risk of a seizure episode diagnosis was significantly increased by about 4-fold. In addition, the estimated attributable rate of a diagnosed seizure episode from 6 to 11 days post-MMR vaccination in the current study at 1 in 1,098 MMR vaccine doses is comparable to that observed by Miller et al. (2007) at 1 in 1,150 MMR vaccine doses. The current study is differentiated from the Miller et al. (2007) by the fact that American children were examined in the current study and British children in the Miller et al. (2007) study. In addition, the Miller et al. (2007) study, in contrast to the current study, did not examine the long-term potential sequalae of the seizure episode diagnoses among the children examined. Finally, the current study utilized ICD-9 codes to examine the outcomes of children evaluated, whereas Miller et al. (2007) utilized ICD-10 hospital discharge diagnosis codes.
      In addition to the aforementioned studies examined by the Cochrane Review, there are several other previous studies examining the relationship between childhood MMR vaccination and the risk of seizures. One such study was a cohort study of 137,457 children immunized with MMR vaccine and the risks of febrile and nonfebrile seizures based upon review of medical records from four large health maintenance organizations [
      • Barlow W.E.
      • Davis R.L.
      • Glasser J.W.
      • Rhodes P.H.
      • Thompson R.S.
      • Mullooly J.P.
      • et al.
      The risk of seizures after receipt of whole-cell pertussis or measles, mumps, and rubella vaccine.
      ]. The researchers observed in the 8 to 14 day post-MMR vaccination period that the risk for diagnosed febrile seizure was significantly increased 2.83-fold (13 children diagnosed with febrile seizures were identified), but the risk for diagnosed nonfebrile seizures was not increased significantly (1 child diagnosed with a nonfebrile seizure was identified). The investigators reported that the attributable risk for diagnosed febrile seizures in the 8 to 14 day period post-MMR vaccination was between 25.0 and 34.2 per 100,000 children. In a subsequent analysis, of automated records (the investigators could not distinguish between diagnosed febrile seizures or nonfebrile seizures based upon ICD-9 coding), they observed a 2.68-fold significant increased risk for diagnosed seizures in the 8 to 14 day post-MMR vaccination period. They also reported that upon follow-up children diagnosed with a febrile seizure after MMR vaccination were no more likely to have a subsequent seizure diagnosis than children diagnosed with a febrile seizure in the absence of vaccination. As for a long-term diagnosis of seizure disorder, the investigators reported none of the children diagnosed with a febrile seizure after MMR vaccination were diagnosed with seizure disorder, whereas 5% of those children diagnosed with a febrile seizure in the absence of vaccination were diagnosed with seizure disorder.
      There are several important points to consider when evaluating the results observed by Barlow et al. (2001) relative to the present study. First, in both studies, when examining ICD-9 seizure episode diagnoses following childhood MMR vaccination, the risk was similarly elevated. Second, unlike the present study, the only comparison group utilized by Barlow et al. (2001) was an unvaccinated cohort (there was no self-controlled case-series analysis). Third, the size of the sample of children diagnosed with a seizure episode (febrile seizure + nonfebrile seizure = 14 children in the 8 to 14 days post-MMR vaccination period) was significantly smaller than the current study (29 children in the 6 to 11 days post-MMR vaccination period). As a result, the rate of a diagnosed seizure episode in the Barlow et al. (2001) study was 1 in 9,818 children. This was significantly lower than the rate observed in the present study at 1 in 810 children. The most likely cause, as described by Barlow et al. (2001), was that there limitations in collection methods of diagnoses among patients at the various health maintenance organization records (i.e., lack of outpatient data), and as such, the identification of all cases of seizure episodes at the health maintenance organizations examined was not complete. Further, the under ascertainment of outcomes in the Barlow et al. (2001) study significantly limited their ability to evaluate the potential relationship between childhood MMR vaccination and long-term seizure disorder diagnoses. Namely, using the rate of subsequent seizure disorder diagnoses among children diagnosed with a seizure episode from the current study (at about 1 in 3 children), this would mean that only about 5 children would be expected to be diagnosed with seizure disorder from the 14 children diagnosed with a seizure in the 8 to 14 days post-MMR vaccination period. Barlow et al. (2001) reported among 41 children examined with a febrile seizure soon after any of the vaccines in their study (this includes 22 after receipt of the MMR vaccine only) none were given a seizure disorder diagnosis. Barlow et al. (2001) also reported among 521 children with a febrile seizure in the absence of vaccination that only 5% of those children during a two year follow-up period were diagnosed with seizure disorder. As a result, the Barlow et al. (2001) study was underpowered to be able to detect an association between childhood MMR vaccination and the long-term risk of a child being diagnosed with a seizure disorder. Interestingly, they did report that children who had a febrile seizure after MMR vaccination were equally as likely to have a subsequent seizure as unvaccinated children. This observation by Barlow et al. (2001) is consistent with the results observed in the current study finding that the rate of seizure disorder diagnoses was similar in those diagnosed with a seizure episode attributably-associated with MMR vaccination or not.
      The Institute of Medicine of the US National Academy of Sciences following an extensive review of the literature determined that the evidence convincingly supports a causal relationship between MMR vaccination and febrile seizures and the mechanistic evidence supporting such an association in the 7 to 14 days post-MMR vaccination period as intermediate [

      US Institute of Medicine. Adverse Effects of Vaccines: Evidence and Causality. Washington, DC: National Academy Press, 2012, pgs 1-866.

      ]. In addition, a prospective, placebo-controlled clinical trial in twins observed that within the first 21 days following MMR vaccination most adverse events commenced 5 to 7 days following immunization and peaked on day 10 [
      • Virtanen M.
      • Peltola H.
      • Paunio M.
      • Heinonen O.P.
      Day-to-day reactogenicity and the healthy vaccinee effect of measles-mumps-rubella vaccination.
      ]. As described in both of the aforementioned studies, the incubation period for natural measles virus (10 to 12 days) is significantly shorter than for mumps (16 to 18 days) and rubella (12 to 23 days). As a result, by examining seizures diagnosed from 6 to 11 days post-MMR vaccination, the present study most likely examined the contribution of the measles component of the MMR vaccine. It is possible that by examining later risk periods following MMR vaccine that contributions of the rubella and/or mumps components of the MMR vaccine to the risk of seizure diagnoses may be further elucidated.
      In further considering the importance of the 6 to 11 days post-MMR vaccination period, additional analyses were undertaken using the same person-time modeling employed in our previous analyses to evaluate the onset of a seizure episode or a seizure disorder regardless of the onset period post-MMR vaccination as compared to a similar potential onset period in the unvaccinated cohort. As revealed in Table 7, there was a small, but significant increased risk for seizure episode (risk ratio = 1.11), but not for seizure disorder (risk ratio = 0.98) in the MMR vaccinated cohort as compared to the unvaccinated cohort. These observations in the IHRD are consistent with those previously discussed by Vestergaard et al. (2004) [
      • Vestergaard M.
      • Hviid A.
      • Madsen K.M.
      • Wohlfahrt J.
      • Thorsen P.
      • Schendel D.
      • et al.
      MMR vaccination and febrile seizures: evaluation of susceptible subgroups and long-term prognosis.
      ]. Namely, they observed that the rate ratio for recurrent febrile seizures was significantly increased in the childhood MMR vaccination cohort by 1.19-fold in comparison to the unvaccinated children, but no difference was observed in the rate of seizure disorder. By contrast, when isolating the onset period from 6 to 11 days post MMR vaccination in comparison to unvaccinated persons using the same analysis, the increased risk for an initial seizure episode diagnosis (risk ratio = 4.64) and an initial seizure episode diagnosis followed by a subsequent seizure disorder diagnosis (risk ratio = 5.51) were significantly higher than observed in the aforementioned analyses. These results further affirm the significant importance in temporality (6 to 11 days post-MMR vaccination) in mediating the apparent acute measles viral etiology of an initial seizure episode and the potential for a long-term seizure disorder post-MMR vaccination.
      Table 7Person-time model results examining the frequency of the outcomes with examined in the MMR vaccinated cohort in comparison to the unvaccinated cohort.
      Parameter ExaminedMMR Vaccinated Cohort
      Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.


      (any time post-MMR Vaccination)
      Unvaccinated Cohort
      Persons received no doses of measles-containing vaccine.


      (any time from 12 months of age)
      Seizure Episode
      Number Diagnosed with the Outcome1,897
      Persons were examined for a seizure episode diagnosis with an initial onset anytime following their first dose of MMR vaccination.
      3,125
      Persons were examined for a seizure episode diagnosis with an initial onset anytime from 12 months of age.
      Total Number of Person-Years195,773357,154
      Rate (per 1,000 person-years)9.698.75
      Rate Ratio (95% CI)1.11 (1.05 to 1.17)
      Rate Difference (per 1,000 person-years)

      (95% CI)
      0.94 (0.42 to 1.47)
      p-value< 0.001
      Seizure Disorder
      Number Diagnosed with the Outcome519
      Persons were examined for an seizure episode diagnosis with an initial onset any time after their first dose of MMR vaccination and were subsequently diagnosed with a seizure disorder within the first 10 years of life (age of seizure disorder onset was considered to be age of an initial seizure episode diagnosis).
      968
      Persons were examined for a seizure episode diagnosis with an initial onset any time after 12 months of age and subsequently diagnosed with a seizure disorder within the first 10 years of life (age of seizure disorder onset was considered to be the age of an initial seizure episode diagnosis).
      Total Number of Person-Years195,773357,154
      Rate (per 1,000 person-years)2.652.71
      Rate Ratio (95% CI)0.98 (0.88 to 1.09)
      Rate Difference (per 1,000 person-years)

      (95% CI)
      −0.06 (-0.35 to 0.23)
      p-value0.70
      Italicized results are statistically significant. CI = confidence interval, MMR = measles-mumps-rubella.
      1 Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      2 Persons received no doses of measles-containing vaccine.
      3 Persons were examined for a seizure episode diagnosis with an initial onset anytime following their first dose of MMR vaccination.
      4 Persons were examined for a seizure episode diagnosis with an initial onset anytime from 12 months of age.
      5 Persons were examined for an seizure episode diagnosis with an initial onset any time after their first dose of MMR vaccination and were subsequently diagnosed with a seizure disorder within the first 10 years of life (age of seizure disorder onset was considered to be age of an initial seizure episode diagnosis).
      6 Persons were examined for a seizure episode diagnosis with an initial onset any time after 12 months of age and subsequently diagnosed with a seizure disorder within the first 10 years of life (age of seizure disorder onset was considered to be the age of an initial seizure episode diagnosis).
      The results of the present study also revealed a significant time-lag between an initial seizure episode diagnosis and a subsequent seizure disorder diagnosis. As shown in Table 3, the average time-lag between initial seizure episode diagnosis and a subsequent seizure disorder diagnosis is 502 days (or 1.4 years), with some being diagnosed > 8 years later. There is the potential for a variable and sometimes lengthy latent period between injury and subsequent neurological and neuropsychiatric dysfunction [
      • DeKosky S.T.
      • Blennow K.
      • Ikonomovic M.D.
      • Gandy S.
      Acute and chronic traumatic encephalopathies: pathogenesis and biomarkers.
      ]. As a result, it can be difficult to identify a cause and effect relationship for delayed sequelae from a brain insult. Fortunately, the present study was able to identify an initial date of a seizure episode diagnosis, and then, follow-up these children to determine their long-term seizure disorder diagnosis status.
      There are several factors involved in delayed brain pathologies that depend upon on the initial source of brain injury. For example, studies show that cases of progressive brain pathogenesis resulting in delayed sequelae from a brain insult can include (but are not limited to): neuroinflammation, oxidative stress, exitotoxicity, altered cell function and signaling, demyelination, immune activation/dysfunction, and/or neuronal and glial damage [
      • Thom S.R.
      • Bhopale V.M.
      • Fisher D.
      • Zhang J.
      • Gimotty P.
      Delayed neuropathology after carbon monoxide poisoning is immune-mediated.
      ,
      • Hendriksen E.
      • van Bergeijk D.
      • Oosting R.S.
      • Redegeld F.A.
      Mast cells in neuroinflammation and brain disorders.NeurosciBiobehav.
      ,
      • Giovannoni G.
      • Cutter G.
      • Pia-Sormani M.
      • Belachew S.
      • Hyde R.
      • Koendgen H.
      • et al.
      Is multiple sclerosis a length-dependent central axonopathy? The case for therapeutic lag and the asynchronous progressive MS hypotheses.
      ,
      • Neubauer R.A.
      • Neubauer V.
      • Nu A.K.C.
      • Maxfield W.S.
      Treatment of late neurologic sequelae of carbon monoxide poisoning with hyperbaric oxygenation: a case series.
      ,
      • Fike J.R.
      Physiopathology of radiation-induced neurotoxicity.
      ]. In cases of delayed sequalae, research revealed that even when an initial single-incident can be identified, and may even be considered mild, later neurological dysfunction is often reported and can be substantial and life disruptive [
      • DeKosky S.T.
      • Blennow K.
      • Ikonomovic M.D.
      • Gandy S.
      Acute and chronic traumatic encephalopathies: pathogenesis and biomarkers.
      ]. There is limited research on the relationship between an initial seizure episode diagnosis and a later seizure disorder diagnosis [
      • Shinnar S.
      • Berg A.T.
      • Moshe S.L.
      • O’Dell C.
      • Alemany M.
      • Newstein D.
      • et al.
      The risk of seizure recurrence after a first unprovoked afebrile seizure in childhood: an extended follow-up.
      ], but the result of the current study suggest, for a minority of children diagnosed with an initial seizure episode between 12 through 16 months of age, an ongoing progressive encephalopathy and/or neurodegeneration process ultimately results in a seizure disorder diagnosis.
      Another important aspect of the present study is to consider the findings with broader public health policy regarding the safety and effectiveness of MMR vaccination in the US. Namely, the current study suggests that about 1 in 1,100 doses of childhood MMR administered from 12 through 16 months of age is attributably associated with an initial seizure episode diagnosis from the self-controlled case-series analysis. The current study also suggests that about 1 in 3,100 doses of childhood MMR administered from 12 through 16 months of age is attributably associated with a seizure disorder diagnosis from the self-control case-series analysis. As a result, it can be estimated assuming about 4 million children per birth cohort year in the US, and an average childhood MMR vaccine uptake of about 80%, that about 1,000 children per year will be diagnosed with a seizure disorder attributably associated with MMR vaccination. By way of comparison, a previous study evaluated 5,940 cases of measles over a 25-year period for neurological manifestations [
      • Boenheim C.
      Uber nervose komplikationen bei spezifisch infections krankheiten kindlichen.
      ]. A total of 11 children were diagnosed with seizure disorder (1 in 540 children). As a rough estimate, this means that the childhood MMR vaccination as compared to natural measles infection is associated with about a 6-fold reduction (>80% reduction) in the rate of seizure disorder diagnoses (MMR vaccine = 1 in 3,100 children vs natural measles infection = 1 in 540 children). It should be pointed out that the aforementioned direct comparisons are somewhat clouded by the fact that both MMR vaccination and natural measles infection are associated with other neurological symptoms (i.e., encephalitis, meningitis, etc.) that may overlap or even have concurrent seizures [
      • Tyler H.R.
      Neurological complications of rubeola (measles).
      ], so these may influence the exact numbers.
      A recently published 10-year longitudinal cohort study of prospectively collected healthcare records examined the effectiveness of routine childhood MMR vaccination on American children born in the 1990s [
      • Geier D.A.
      • Kern J.K.
      • Geier M.R.
      Childhood MMR vaccination and the incidence rate of measles infection: a ten year longitudinal cohort study of American children born in the 1990s.
      ]. When comparing a cohort of children receiving a single dose of MMR vaccine to a cohort of children never receiving any measles-containing vaccine, the incidence rate of being diagnosed with measles was significantly reduced by 80% to 90%. These investigators concluded that their study reveals that routine childhood MMR vaccination is an important public health tool to control cases of measles. This recent finding on American children is consistent with a previous meta-analysis conducted by investigators from the Cochrane Review [

      Demicheli V, Rivetti A, Debalini MG, Di Pietrantonj C. Vaccines for measles, mumps and rubella in children. Cochrane Database Syst Rev 2012:CD004407.

      ].

      5. Strengths/Limitations

      An important strength of this study is that the data examined were generated independently of the current study design. The retrospective observations made in the IHRD were derived from eligibility and claims records prospectively generated as part of the routine healthcare provided for persons in the Florida Medicaid system. The healthcare providers submitting claims for MMR vaccine administration, seizure episode, or seizure disorder diagnoses were most likely not thinking about the possible relationship between MMR vaccination and seizure episode or seizure disorder diagnoses. As result, there should be negligible impacts from potential recall biases between exposures and outcomes.
      Another important strength of this study is the comprehensive follow-up of the cohorts examined. All persons examined within the IHRD were eligible for Florida Medicaid from birth for 120 months (no gaps in eligibility were allowed). In addition, in order to ensure that the cohort of persons examined was actively utilizing healthcare services from the Florida Medicaid system, all persons examined in this study had to have ≥ 10 outpatient office visit claims submitted (that averages to at least one outpatient office visit per person per year). These requirements helped to significantly reduce possible enrollment factors or differences in healthcare-seeking behaviors among the persons examined in this study.
      A further strength of the current study is that exposure and outcomes were ascertained with precision for each person examined. The vaccination status was determined by examining procedure codes and dates of service for claims submitted on behalf of each person. In order for a person to become a member of the vaccinated cohort, the claims records for a person revealed that at least one dose of MMR vaccine was administered within the period from 12 through 16 months of age. Those persons with procedure codes specifying other measles-containing vaccine(s) were excluded from the present analyses. Members of the unvaccinated cohort were confirmed to have no claims submitted on their behalf specifying receipt of any type of measles-containing vaccine from birth through 10 years of age. The outcome status of each person was determined using ICD-9 diagnosis codes and dates of service for claims submitted on behalf of each person. In order for a person to be recognized as having a seizure episode diagnosis, the initial date of service specifying a seizure episode diagnosis (780.3x) was identified. Similarly, in order for a person to be recognized as having a seizure disorder diagnosis, the initial date of service specifying a seizure disorder diagnosis (345.xx) was identified.
      A final strength of the current study is that Cox proportional hazard survival plot and person-time statistical modeling were both used to evaluate the potential relationship between exposure to childhood MMR vaccination and seizure episode and seizure disorder diagnoses in a defined temporal period post-MMR vaccination. These models were of particular importance because they allowed for isolation of a potential temporal specific risk interval (i.e., from 6 to 11 days post-MMR vaccination) as compared to control intervals (i.e., the complete absence of measles-containing vaccine exposure or long enough after MMR vaccination that the vaccine should not have impacted the rate of the diagnosed outcomes examined). The present study also employed a self-controlled case-series data analysis method, which was specifically developed to handle potential confounding in vaccine safety studies [
      • Farrington C.P.
      Relative incidence estimation from case series for vaccine safety evaluation.
      ]. The self-controlled case-series design adjusts for the effects of time-invariant confounders by allowing each person to act as their own control.
      Furthermore, the present study was not only able to identify the onset of an initial seizure episode diagnosis, but persons in the cohorts examined were followed until they were 10 years of age. As a result, it was possible to determine their eventual residual seizure disorder status by examining them for a potential long-term seizure disorder diagnosis. This is important because as revealed in the current study a diagnosis of seizure disorder may occur many months or even years following an initial seizure episode diagnosis. The current study observed among children diagnosed with an initial seizure episode that about 29%–33% were subsequently diagnosed with seizure disorder (regardless, if their initial seizure episode diagnosis was associated with MMR vaccination or related to other etiologies). This result is consistent with previous estimates indicating that of children diagnosed with an initial febrile seizure between 25% and 32% went on to be diagnosed with a seizure disorder [
      • Nelson K.B.
      • Ellenberg J.H.
      Prognosis in children with febrile seizures.
      ,
      • Graves R.C.
      • Oehler K.
      • Tingle L.E.
      Febrile seizures: risks, evaluation, and prognosis.
      ].
      A potential limitation of the current study is the lack of diagnostic precision of the initial seizure episode diagnoses. Unfortunately, the initial seizure episode ICD-9 diagnosis codes utilized by healthcare providers did not always identify whether a seizure episode was associated with fever or not (febrile versus afebrile seizure). Table 2 and Table 3 show that most initial seizure episode diagnoses were made in unspecified diagnostic categories (ICD-9 codes: 780.3 or 780.39). This is not unexpected, since a healthcare provider at initial presentation of a child with a seizure does not necessarily have all the clinical information to make complete judgement as to the pathology of a child’s seizure episode. Despite this limitation, it was observed in both Table 2 and Table 3 that of all initial seizure episode diagnoses, the percentage of children diagnosed with a febrile seizure episode was significantly higher among those in the MMR vaccinated cohort as compared to the unvaccinated cohort. Person-time models were employed to evaluate on those children diagnosed with an initial febrile seizure diagnosis (ICD-9 code: 780.31). Table 8 shows children with an initial febrile seizure diagnosis were at increased risk for an initial seizure episode with an onset from 6 to 11 post-MMR vaccination as compared to onset from 12 through 16 months of age in the unvaccinated cohort (rate ratio = 8.58). Similarly, children were at an increased risk of a seizure disorder diagnosis following an initial seizure episode with an onset from 6 to 11 days post-MMR vaccination as compared to onset from 12 through 16 months of age in the unvaccinated cohort (rate ratio = 15.4). These risks were significantly higher than those children diagnosed with one of several different types of seizure episode diagnoses (ICD-9 codes: 780.3, 780.31, or 780.39), which revealed a rate ratio = 4.64 for an initial seizure episode and a rate ratio = 5.51 for a seizure disorder diagnosis. These observations are compatible with well-established measles virus fever-associated seizure etiology, and suggest that most initial seizure episodes examined in the MMR vaccinated cohort involved a febrile component. It is recommended that future studies further elucidate the potential etiology for a febrile/afebrile seizure episode post-MMR vaccination.
      Table 8Person-time modeling for the frequency of the outcomes examined in the MMR vaccinated cohort in comparison to the unvaccinated cohort.
      Parameter ExaminedMMR Vaccinated Cohort
      Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.


      (initial onset from 6 to 11 days post MMR vaccination)
      Unvaccinated Cohort
      Persons received no doses of measles-containing vaccine.


      (initial onset from 12 through 16 months of age)
      Febrile Seizure Episode Only (ICD-9 code: 780.31)
      Number Diagnosed with the Outcome8
      Persons were examined for a febrile seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination.
      61
      Persons were examined for a febrile seizure episode diagnosis with an initial onset from 12 months through 16 months of age.
      Total Number of Person-Days96,9906,342,542
      Rate (per 1,000,000 person-days)82.59.62
      Rate Ratio (95% CI)8.58 (3.54 to 17.9)
      Rate Difference (per 1,000,000 person-days) (95% CI)73 (52 to 94)
      p-value< 0.0001
      Seizure Disorder
      Number Diagnosed with the Outcome4
      Persons were examined for a febrile seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination and a subsequent diagnosis of a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be age of an initial seizure episode diagnosis).
      17
      Persons were examined for a febrile seizure episode with an initial onset from 12 months through 16 months of age and a subsequent diagnosed of a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be the age of an initial seizure episode diagnosis).
      Total Number of Person-Days96,9906,342,542
      Rate (per 1,000,000 person-days)41.22.68
      Rate Ratio (95% CI)15.4 (3.77 to 47.2)
      Rate Difference (per 1,000,000 person-days) (95% CI)39 (27 to 50)
      p-value< 0.001
      Italicized results are statistically significant. CI = confidence interval, ICD-9 = International Code of Disease, 9th revision, MMR = measles-mumps-rubella,
      1 Persons received at least 1 dose of MMR vaccine between 12 through 16 months of age.
      2 Persons received no doses of measles-containing vaccine.
      3 Persons were examined for a febrile seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination.
      4 Persons were examined for a febrile seizure episode diagnosis with an initial onset from 12 months through 16 months of age.
      5 Persons were examined for a febrile seizure episode diagnosis with an initial onset from day 6 to 11 post their first dose of MMR vaccination and a subsequent diagnosis of a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be age of an initial seizure episode diagnosis).
      6 Persons were examined for a febrile seizure episode with an initial onset from 12 months through 16 months of age and a subsequent diagnosed of a seizure disorder within the first 10 years of life (age of initial seizure disorder onset was considered to be the age of an initial seizure episode diagnosis).
      It is possiblethat the observed associations were the result of statistical chance or cofounders/unknown biases in the data. Statistical chance seems unlikely given that a limited number of statistical tests were performed, and most results were highly statistically significant. In addition, the significant effects observed in unadjusted models remained significant even when adjusting for potential covariates such as gender, date of birth, and county of residence. The results also remained significant in the self-controlled case-series models where time-invariant cofounders were minimized because each person served as their own control. Finally, the results observed this study were consistent with previous epidemiological observations on different populations and were biologically plausible.
      It is also possible that some of the persons examined in the IHRD may have had symptoms of seizures that were so slight that they were not noted by their healthcare providers, or healthcare providers may have misdiagnosed or misclassified vaccination status for some persons. However, these potential limitations, while possible, should not have affected the results appreciably because it is uncertain how differential application would have occurred in the vaccinated and unvaccinated cohorts examined. Importantly, any misclassification with respect to diagnostic or vaccination status, would in all likelihood bias the findings towards the null hypothesis because persons examined would have been put into the wrong vaccination and/or diagnostic category and result in diminished statistical power to establish the accurate relationship between vaccination and outcomes.
      A further potential limitation of the present study was that only receipt of the first dose of MMR vaccination was examined. It is possible that other vaccines administered concurrently or before/after the first dose of MMR vaccine examined in this study may have impacted the children examined. This would seem unlikely given that only outcomes diagnosed in a priori identified, biologically plausible time period for measles vaccine-associated adverse effects from 6 to 11 days post-MMR vaccination were examined, but future studies should evaluate the potential impact of other vaccines on the phenomena observed in the current study. In addition, future studies should examine whether concurrent infections/environmental exposures [
      • Miller E.
      • Andrews N.
      • Stowe J.
      • Grant A.
      • Waight P.
      • Taylor B.
      Risk of convulsion and aseptic meningitis following measles-mumps-rubella vaccination in the United Kingdom.
      ] or genetic susceptibility factors [
      • Feenstra B.
      • Pasternak B.
      • Geller F.
      • Carstensen L.
      • Wag T.
      • Huang F.
      • et al.
      Common variants associated with general and MMR vaccine-related febrile seizures.
      ] mediate the relationship between MMR vaccine and seizure diagnoses.

      6. Conclusion

      The present study is the first retrospective longitudinal cohort study evaluating the relationship childhood MMR vaccination and the long-term risk of seizure disorder in American children utilizing Cox proportional hazards ratio and time-trend models post-MMR vaccination compared to unvaccinated persons and in a self-controlled case-series. The incidence rate of an initial seizure episode with onset of symptoms from 6 to 11 days post-MMR vaccination and the incidence rate of an eventual seizure disorder among those with an initial seizure episode with onset of symptoms from 6 to 11 days post-MMR vaccination were significantly increased in all analyses undertaken. The current study revealed that about 1 in 3,100 doses of MMR vaccine administered to children from 12 through 16 months of age are attributably associated with a seizure disorder diagnosis following an initial seizure episode with an onset of symptoms from 6 to 11 days post-MMR vaccination. The observed rate of seizure disorder diagnosis post-MMR vaccination is estimated to be a > 80% reduction as compared to natural measles infection. Further, a recent long-term longitudinal cohort study of American children revealed that childhood MMR vaccination reduced the rate of diagnosed measles by 80% to 90%. On balance, despite the adverse effects observed in the current study, the routine childhood MMR vaccination program in the US is an important and effective means to control natural measles infection, but on rare occasions is associated with an increased risk of seizure disorder.

      Declaration of Competing Interest

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      References

      1. McLean HQ, Fiebelkorn AP, Temte JL, Wallace GS; Centers for Disease Control and Prevention. Prevention of measles, rubella, congenital rubella syndrome and mumps, 2013: summary recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep 2013;62(RR-04):1-34.

        • Bloch A.B.
        • Orenstein W.A.
        • Stetler H.C.
        • Wassilak S.G.
        • Amler R.W.
        • Bart K.J.
        • et al.
        Health impact of measles vaccination in the United States.
        Pediatrics. 1985; 76: 524-532
      2. MERCK & CO., INC., M-M-R-II, Measles Mumps, and Rubella Virus Vaccine Live. http://www.merck.com/product/usa/pi_circulars/m/mmr_ii/mmr_ii_pi.pdf. Revised 5/2017; Accessed on July 18, 2018.

      3. Demicheli V, Rivetti A, Debalini MG, Di Pietrantonj C. Vaccines for measles, mumps and rubella in children. Cochrane Database Syst Rev 2012:CD004407.

      4. Di Pietrantonj C, Rivetti A, Marchione P, Debalini MG, Demicheli V. Vaccines for measles, mumps, rubella, and varicella in children. Cochrane Database Syst Rev 2020;4:CD004407.

        • Hakim R.B.
        • Boben P.J.
        • Bonney J.B.
        Medicaid and the health of children.
        Health Care Financ Rev. 2000; 22: 133-140
        • Vestergaard M.
        • Hviid A.
        • Madsen K.M.
        • Wohlfahrt J.
        • Thorsen P.
        • Schendel D.
        • et al.
        MMR vaccination and febrile seizures: evaluation of susceptible subgroups and long-term prognosis.
        JAMA. 2004; 292: 351-357
        • Ward K.N.
        • Bryant N.J.
        • Andrews N.J.
        • Bowley J.S.
        • Ohrling A.
        • Verity C.M.
        • et al.
        Risk of serious neurologic disease after immunization of young children in Britain and Ireland.
        Pediatrics. 2007; 120: 314-321
        • Miller E.
        • Andrews N.
        • Stowe J.
        • Grant A.
        • Waight P.
        • Taylor B.
        Risk of convulsion and aseptic meningitis following measles-mumps-rubella vaccination in the United Kingdom.
        Am J Epidemiol. 2007; 165: 704-709
        • Barlow W.E.
        • Davis R.L.
        • Glasser J.W.
        • Rhodes P.H.
        • Thompson R.S.
        • Mullooly J.P.
        • et al.
        The risk of seizures after receipt of whole-cell pertussis or measles, mumps, and rubella vaccine.
        N Engl J Med. 2001; 345: 656-661
      5. US Institute of Medicine. Adverse Effects of Vaccines: Evidence and Causality. Washington, DC: National Academy Press, 2012, pgs 1-866.

        • Virtanen M.
        • Peltola H.
        • Paunio M.
        • Heinonen O.P.
        Day-to-day reactogenicity and the healthy vaccinee effect of measles-mumps-rubella vaccination.
        Pediatrics. 2000; 106: E62
        • DeKosky S.T.
        • Blennow K.
        • Ikonomovic M.D.
        • Gandy S.
        Acute and chronic traumatic encephalopathies: pathogenesis and biomarkers.
        Nat Rev Neurol. 2013; 9: 192-200
        • Thom S.R.
        • Bhopale V.M.
        • Fisher D.
        • Zhang J.
        • Gimotty P.
        Delayed neuropathology after carbon monoxide poisoning is immune-mediated.
        Proc Natl Acad Sci USA. 2004; 101: 13660-13665
        • Hendriksen E.
        • van Bergeijk D.
        • Oosting R.S.
        • Redegeld F.A.
        Mast cells in neuroinflammation and brain disorders.NeurosciBiobehav.
        Rev. 2017; 79: 119-133
        • Giovannoni G.
        • Cutter G.
        • Pia-Sormani M.
        • Belachew S.
        • Hyde R.
        • Koendgen H.
        • et al.
        Is multiple sclerosis a length-dependent central axonopathy? The case for therapeutic lag and the asynchronous progressive MS hypotheses.
        Mult Scler Relat Disord. 2017; 12: 70-78
        • Neubauer R.A.
        • Neubauer V.
        • Nu A.K.C.
        • Maxfield W.S.
        Treatment of late neurologic sequelae of carbon monoxide poisoning with hyperbaric oxygenation: a case series.
        J Am Phys Surg. 2006; 11: 56-59
        • Fike J.R.
        Physiopathology of radiation-induced neurotoxicity.
        Rev Neurol (Paris). 2011; 167: 746-750
        • Shinnar S.
        • Berg A.T.
        • Moshe S.L.
        • O’Dell C.
        • Alemany M.
        • Newstein D.
        • et al.
        The risk of seizure recurrence after a first unprovoked afebrile seizure in childhood: an extended follow-up.
        Pediatrics. 1996; 98: 216-225
        • Boenheim C.
        Uber nervose komplikationen bei spezifisch infections krankheiten kindlichen.
        Ergeb Inn Med Kinderheilkd. 1925; 28: 598-637
        • Tyler H.R.
        Neurological complications of rubeola (measles).
        Medicine (Baltimore). 1957; 36: 147-167
        • Geier D.A.
        • Kern J.K.
        • Geier M.R.
        Childhood MMR vaccination and the incidence rate of measles infection: a ten year longitudinal cohort study of American children born in the 1990s.
        BMC Pediatr. 2019; 19: 325
        • Farrington C.P.
        Relative incidence estimation from case series for vaccine safety evaluation.
        Biometrics. 1995; 51: 228-235
        • Nelson K.B.
        • Ellenberg J.H.
        Prognosis in children with febrile seizures.
        Pediatrics. 1978; 61: 720-727
        • Graves R.C.
        • Oehler K.
        • Tingle L.E.
        Febrile seizures: risks, evaluation, and prognosis.
        Am Fam Physician. 2012; 85: 149-153
        • Feenstra B.
        • Pasternak B.
        • Geller F.
        • Carstensen L.
        • Wag T.
        • Huang F.
        • et al.
        Common variants associated with general and MMR vaccine-related febrile seizures.
        Nat Genet. 2014; 46: 1274-1282