Conclusions about type 1 diabetes and hygiene hypothesis are premature

doi:10.1136/bmj.322.7299.1429

BMJ 2001;322:1429 ( 9 June )

Letters

Conclusions about type 1 diabetes and hygiene hypothesis are premature

EDITOR $---$ Johnston and Openshaw state that children are born with strong interleukin 4 based (type 2) immune responses and matureto interferon $gamma$ based (type 1) responses, and that this processis under genetic and environmental influence.¹ They go on tostate that asthma and atopy are rising in prevalence and thathaving older siblings and being exposed to infections promotesthe normal maturation of the immune system towards a type 1response.

This argument fails to take account of the evidence from type 1 diabetes, which is an interferon $gamma$ based disease. Type 1 diabetesis rising in incidence in children from Western societies andis commoner in first born children and in the children of thewell off.^2-4 Infection may have a role in the changing epidemiologyof disease, but the evidence contradicts the suggestion that thisis due to a failure of normal immune development towards a type1response.

Though consistent with current dogma, the conclusion that we should attempt to mimic the effect of childhood infection onthe immune system is premature, if not totally flawed, on twocounts. Firstly, association, not causation, has been shown,⁵and, secondly, the proposed mechanism of action is inconsistentwith availableevidence.

Wendy J A Anderson, consultant respiratory physician.
Antrim Hospital, Antrim BT41 2RL jendywane{at}hotmail.com

Lorna Watson, specialist registrar in public health medicine.
Lothian Health, Edinburgh EH8 6RE

1.	Johnston SL, Openshaw PJM. The protective effect of childhood infection. BMJ 2001; 322: 376-377[Free Full Text]. (17 February.)
2.	Patterson CC, Carson DJ, Hadden DR. Epidemiology of childhood IDDM in Northern Ireland 1989-1994: low incidence in areas with highest population density and most household crowding. Northern Ireland Diabetes Study Group. Diabetologia 1996; 39: 1063-1069[Medline].
3.	Bingley P, Douek I, Rogers C, Gale E. Influence of maternal age at delivery and birth order on risk of type1 diabetes in childhood: prospective population based family study. BMJ 2000; 321: 420-424[Abstract/Free Full Text].
4.	Green A, Gale E, Patterson C. Incidence of childhood-onset insulin dependent diabetes mellitus: the EURODIAB ACE study. Lancet 1992; 339: 905-909[CrossRef][Medline].
5.	Last JM, ed. Hill's criteria of causation. A dictionary of epidemiology. 3rd ed. Oxford: Oxford University Press, 1995.