Papers

MMR vaccine and Crohn's disease: ecological study of hospital admissions in England, 1991 to 2002

BMJ 2005; 330 doi: http://dx.doi.org/10.1136/bmj.38449.476759.AE (Published 12 May 2005) Cite this as: BMJ 2005;330:1120
  1. Valerie Seagroatt, statistician (valerie.seagroatt{at}dphpc.ox.ac.uk)1
  1. 1 Unit of Health-Care Epidemiology, Department of Public Health, University of Oxford, Oxford OX3 7LF
  • Accepted 29 March 2005

Introduction

It has been hypothesised that the measles, mumps, and rubella vaccine (MMR vaccine) increases the risk of autism and Crohn's disease. Although a possible link with autism has been extensively studied and refuted,1 a link with Crohn's disease has not. I tested this hypothesis by analysing trends in age specific admission rates for Crohn's disease in children and adolescents to determine if the introduction of MMR vaccine in 1988 increased rates in those populations that were offered the vaccine as infants.

Methods and results

Counts of admissions, taken as the first consultant episode in a hospital stay, in patients aged &880418; years with a main diagnosis of Crohn's disease in England (population 50 million) were available for the 12 years from April 1991 to March 2003.w1I restricted the analysis to emergency admissions as these were probably less susceptible to changes in thresholds for admission and clinical practice than elective admissions. In the first two years of the MMR vaccination programme, the percentages of children completing a primary course of MMR vaccine in their second year of life were 7% and 68%; thereafter the percentage was at least 84%.w2

Initially, temporal trends in age specific rates were plotted, differentiating between the rates for those born before and after the introduction of the vaccine. (Rates for patients aged 16 to 18 years—too old to have been offered the vaccine as infants—provided information on underlying trends unaffected by MMR vaccination in infancy.) Data for those born in 1987-8, of whom only 68% were vaccinated as infants, were excluded from the analysis. The MMR vaccination programme was then modelled as a variable with two levels (vaccination rate of=≥ 84% and of ≤ 7%) using Poisson regression, with adjustment for year of admission and age in single years (as categorical variables).

There were 4463 admissions for Crohn's disease, 923 of which occurred in populations with a vaccination rate of ≥84% (those born in 1988-9 or later). Although the age specific rates increased over the study period, no obvious changes occurred that coincided with the introduction of MMR vaccine (figure). The estimated rate ratio for the MMR vaccination programme (rates in populations with a vaccination rate of ≥84% compared with those with a rate of ≤ 7%) was 0.95 (95% confidence interval 0.84 to 1.08).

Figure1

Age specific rates per 10 000 population per year for emergency hospital admissions for Crohn's disease in England, 1991 to 2002. Rates in children aged <4 years were relatively low and so were excluded from the figure. Three-year groups, rather than the more conventional five-year groups, were used in order to discriminate between rates in children born before and after the introduction of MMR

Comment

The introduction of MMR vaccine, replacing the single measles vaccine, was not associated with an increase in Crohn's disease. Given the precision of the rate ratio, all but a small risk would have been detected. This was an ecological study, and findings from such studies generally need to be treated cautiously because of potential for confounding. Could the negative finding from this analysis be due to confounding? If so, some factor(s) would have to be negatively associated with Crohn's disease, be introduced over the same three year period, and be targeted at the same population of infants as MMR vaccine to mask a true association. This seems highly unlikely.

What is already known on this topic

It has been hypothesised that the measles, mumps, and rubella vaccine (MMR vaccine) increases the risk of Crohn's disease, though the evidence base for this hypothesis is sparse

What this study adds

An ecological analysis of national data on hospital admissions found no increase in Crohn's disease associated with the introduction of the MMR vaccination programme, providing strong evidence against the hypothesis that MMR vaccine increases the risk of Crohn's disease

A smaller ecological study2 and two case-control studies3 w3also found no increased risk of Crohn's disease associated with MMR vaccine.

Initially, measles vaccine was reported to be associated with higher rates of Crohn's disease,4 but this was not confirmed by subsequent studies.3 5 Natural infections with measles and mumps within one year were also associated with an increased risk of Crohn's disease.w4Although that finding has yet to be confirmed, it raised the possibility that infections with multiple viruses in MMR vaccine might increase the risk of Crohn's disease. My study provides strong evidence against that hypothesis and adds to the evidence that MMR vaccine is no less safe in this respect than the single measles vaccine.

See also p 1119, and Clinical review p 1132

Embedded ImageAdditional references (w1-w4) are on bmj.com

This article was posted on bmj.com on 18 April 2005:http://bmj.com/cgi/doi/10.1136/bmj.38449.476759.AE

The Unit of Health-Care Epidemiology is funded by the National Co-ordinating Centre for Research Capacity Development at the Department of Health. I thank Steven Morgan at the South East Public Health Observatory for providing recent data from hospital episode statistics, and Paddy Farrington and Irene Stratton for comments on earlier versions of this manuscript.

Footnotes

  • Contributors VS is the sole contributor.

  • Funding No special funding.

  • Competing interests None declared.

  • Ethical approval Not required.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.