I have had an interest in the relationship between diabetes
and infant feeding for some time.

First, on the topic of whether there is an association
between lack of breastfeeding and incidence of diabetes.
Unfortunately, as is often the case, many of the studies
so far have had methodological flaws.  The evidence was
reviewed by Tamir M Ellis and Mark Atkinson recently in
the Lancet, (Ellis TM, Atkinson MA. Early infant diets
and insulin dependent diabetes Lancet 1996;347:1464-5).

"A meta-analysis of most of these studies identified a
weak but significant association (odds ratio 1.5) between
infant nutrition and the risk of IDD. (Gerstein H.
Cow's milk exposure and type I diabetes mellitus.
Diabetes Care 1994;17:13-19.)  That study showed that IDD
patients were more likely to have earlier exposure to cow's
milk and were likely to have been breastfed for a shorter
time than healthy (non-diabetic) individuals."

The rest of the Ellis, Atkinson article focussed on the
possible role of cow's milk formula in triggering an
auto-immune process that leads to diabetes, and concluded
that there is insufficient evidence to warrant changes in
dietary advice.

I believe that juvenile onset (insulin dependent) diabetes
is likely to have several causes, one of which is infection
in utero or first year of life with Coxsackie B virus.
There are several studies that have found an association
between infection with Coxsackie B and onset of diabetes
(Clements GB, Galbraith DN, Taylor KW  Coxsackie B virus
infection and onset of childhood diabetes
Lancet 1995;346:221-23).  Coxsackie B virus is an enterovirus
(or gut virus) that is transmitted by the faecal-oral route.
These studies were reinforced by a recent study by Rothwell
et al that children with diabetes were more likely to be born
in the spring or early summer. (Rothwell PM, Staines A,
Smail P, Wadsworth E, Mckinney P.  Seasonality of birth of
patients with childhood diabetes in Britain.
BMJ 1996;312:1456-7) since enterovirus infection shows a
marked seasonal pattern, with infection more likely in the
summer months in temperate climates (Jenista JA, Powell K,
Menegus MA.   Epidemiology of neonatal enterovirus infection.
J Pediatr 1984;104:685-6909).  Breastfeeding has been found
to protect infants from infection with enteroviruses,
(Jenista JA, Powell K, Menegus MA.   Epidemiology of neonatal
enterovirus infection.  J Pediatr 1984;104:685-6909)
probably as a result of direct neutralisation of virus in the
gut by antibodies in the milk and the increased production of
IgA by breastfed babies.  However I have not seen any
suggestion in the literature that this is the most plausible
explanation for the association between infant diet and
diabetes mellitus in childhood.  I have sent this as a letter
to JAMA.

Hope this helps
Wendy Holmes, Melbourne, Australia