Letters

Carriage rate of Neisseria meningitidis among university students

BMJ 2000; 321 doi: https://doi.org/10.1136/bmj.321.7257.383 (Published 05 August 2000) Cite this as: BMJ 2000;321:383

Further data are needed

  1. Anna Gilmore, senior registrar in public health medicine (anna.gilmore{at}lshtm.ac.uk),
  2. James Stuart, consultant epidemiologist
  1. London School of Hygiene and Tropical Medicine, London WC1E 7HT
  2. Communicable Disease Surveillance Centre (South West), Public Health Laboratory, Gloucester
  3. Department of Public Health Medicine and Epidemiology
  4. Department of Public Health Medicine and Epidemiology
  5. Department of Public Health Medicine and Epidemiology
  6. Meningococcal Research Group, Division of Microbiology
  7. Department of Public Health Medicine and Epidemiology
  8. Meningococcal Research Group, Division of Microbiology
  9. Department of Public Health Medicine and Epidemiology
  10. Communicable Disease Surveillance Centre Trent
  11. Meningococcal Research Group, Division of Microbiology, University of Nottingham, Queens Medical Centre, Nottingham, NG7 2UH

    EDITOR—The paper by Neal et al documenting risk factors for acquisition of Neisseria meningitidis among university students helps further our understanding of the aetiology of outbreaks of meningococcal disease at universities.1 Several outbreaks have now been linked to bars and nightclubs.2 3 By showing that social factors such as attendance at bars influence acquisition, the study helps substantiate evidence that social behaviour is important in determining outbreak occurrence.

    However, the threefold rise in carriage rates in the first four days of term is unexpected. The initial carriage rate (6.9%), as Neal et al acknowledge, is surprisingly low. In 16–24 year olds, carriage rates in both outbreak and non-outbreak situations are usually 20% or higher.4 5 The sudden increase, in a cross sectional study, to rates that would be normal for this age group suggests potential confounding. Carriage has been clearly documented to vary with age and sex, 2 4 5 yet no comparison of the age and sex of the four groups is presented, nor are these variables included in the regression analysis that examines risk factors for carriage.

    The study shows that virulent C2a strains are acquired more rapidly over the term than other strains. Rapid acquisition of disease causing strains was also suggested in the outbreaks at Cardiff and Southampton universities, where low carriage rates of serogroup C outbreak organisms were documented along with very close clustering of the cases in time.2 If a rapid increase in carriage does occur at the start of term, and particularly if virulent C2a strains are transmitted preferentially as the study suggests, one would expect the rise in carriage to be accompanied by a dramatic peak in disease incidence in the first one to two weeks of term. Yet the peak of cases usually occurs after a delay of three to five weeks.1 The alternative, albeit unlikely, explanation is that acquisition of different strains varies over time, with acquisition of C2a strains occurring later.

    It would be useful if Neal et al could present further analysis to help rule out confounding as an explanation for the study's unexpected initial findings. Although variation in sensitivity of swabbing is difficult to exclude, Neal et al have otherwise done their best to validate the data. Nevertheless, a cohort study using sensitive microbiological techniques would be required to confirm whether the rapid rise in carriage at the start of term is the result of a true rise in acquisition, whether differential acquisition of virulent and non-virulent strains occurs, or whether alternative explanations account for the findings.

    References

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

    Authors' reply

    1. Keith R Neal, senior lecturer (keith.neal{at}nott.ac.uk),
    2. Jonathan S Nguyen-Van-Tam, senior lecturer,
    3. Nicholas Jeffrey, medical student,
    4. Richard C Slack, senior lecturer,
    5. Richard J Madeley, professor,
    6. Kamel Ait-Tahar, PhD student,
    7. Katy Job, medical student,
    8. Martin C J Wale, regional epidemiologist,
    9. Dlawer A A Ala'Aldeen, reader
    1. London School of Hygiene and Tropical Medicine, London WC1E 7HT
    2. Communicable Disease Surveillance Centre (South West), Public Health Laboratory, Gloucester
    3. Department of Public Health Medicine and Epidemiology
    4. Department of Public Health Medicine and Epidemiology
    5. Department of Public Health Medicine and Epidemiology
    6. Meningococcal Research Group, Division of Microbiology
    7. Department of Public Health Medicine and Epidemiology
    8. Meningococcal Research Group, Division of Microbiology
    9. Department of Public Health Medicine and Epidemiology
    10. Communicable Disease Surveillance Centre Trent
    11. Meningococcal Research Group, Division of Microbiology, University of Nottingham, Queens Medical Centre, Nottingham, NG7 2UH

      EDITOR—We agree with Gilmore and Stuart that social factors are important in the epidemiology of meningococcal disease among university students1 and that these contribute to the high rate of disease seen in this group.2

      The threefold rise was unexpected, but, as mentioned in the paper, we have repeated this work using a cohort of 229 students who had swabs taken on their first day on arrival at university and again eight or 10 days later; this cohort showed a similar increase in carriage with overlapping confidence intervals. Given this finding, confounding is an unlikely explanation for the rise. The low initial carriage rate seen in both studies may reflect the effect of the prolonged summer holidays when many students disperse from their established social groups and go away. The increased carriage rates seen with more social mixing and residence on campus also supports the notion that the rise we described is real. Given local knowledge of how students were recruited it is unlikely that confounding could produce such a large effect.

      Sex was controlled for in the analysis of carriage risk factors (table 2), although this was not mentioned. Most of the students were aged 18 or 19 (89.2%), and 97% were aged 21 or younger. Ages were evenly distributed by day, except during the final day when the students were slightly older, but restricting the analysis to those aged 18 and 19, or under 22, shows no significant changes in carriage rates. Age was not associated with carriage, although 0/17 students aged 25 and over were negative for meningococci.

      Many of the isolates from Thursday and Friday towards the end of the first swabbing round were likely to have been acquired at university, and these are the rapidly transmitted strains. The general absence of C2a in the first week and the much higher prevalence in November, when cases of disease peak,2 suggests that this strain may spread more slowly than others. The peak of C2a strains happened at the same time as the peak in cases among university students. In addition, one of the students who became ill while part of the study had carried the pathogenic strain for at least seven weeks before becoming ill,3 which implies that longer periods of carriage before illness may contribute to the mid-term peak of disease.

      Preferential treatment of pathogenic or non-pathogenic strains could not be assessed as it is not possible to know whether a strain is pathogenic as disease is rare and there are host organism interactions involved in illness. The final point made in the letter about a cohort study has already been addressed.

      References

      1. 1.
      2. 2.
      3. 3.
      View Abstract

      Sign in

      Log in through your institution

      Subscribe