Twin  and family studies indicate that genetic factors alone cannot explain the etiology of T1DM. Seasonality, increasing incidence and epidemics of T1DM as well as numerous ecological, cross-sectional and retrospective studies suggest a critical role of environmental factors, such as infections with certain viruses (especially enteric infections in early life) and effects of early childhood diet. Natural history studies that follow children at increased risk of T1DM provide the best opportunity to study environmental triggers.
Viruses: Herpes viruses,mumps, rubella, and retroviruses  have been implicated. Viral infections appear to initiate autoimmunity and perhaps also precipitate diabetes in subjects with autoimmunity. ICA or IAA have been detected after mumps, rubella, measles, chickenpox, Coxsackie, ECHO-4, and rotavirus  infections.
While several viruses have been linked to T1DM, studies have provided the strongest overall evidence for enteroviruses, although results have been somewhat conflicting . In one longitudinal birth cohort study (DAISY), progression from islet autoimmunity to T1DM seemed to increase after an enterovirus infection . In the Finnish type 1 Diabetes Prediction and Prevention (DIPP) study, enteroviral infections are seen more frequently in prediabetic children and prior to the onset of islet autoimmunity, implying a temporal relationship between enterovirus infections and the induction of β-cell autoimmunity . Although several studies report that enteroviruses may play a role in the pathogenesis of T1DM, the evidence that enterovirus infections are associated with initiation or progression of islet autoimmunity is still inconsistent .
Effect of childhood infections and daycare exposure: Early infectious exposure may play a role in the development of immunoregulatory mechanisms that protect against diabetes. Social mixing through attendance at daycare in early infancy appears to confer protection against the development of childhood diabetes . Although several other well-designed case-control studies show a statistically significant protective effect of day care exposure on T1DM, meta-analysis reveals too much heterogeneity to accept overall synthesis .
Improvement in hygiene: Genetic models are unable to explain the apparent temporal changes in the incidence of T1DM . Alternative explanations look at environmental factors and some invoke the congenital rubella model. Briefly, increased hygiene in the Western world has led to a decline in immunity to common infections among women in child-bearing age. These women are more likely to develop viremia during pregnancy resulting in congenital persistent infection of beta cells and early onset T1DM in the offspring. This model could explain both the increasing incidence of diabetes and the decreasing age of disease onset.
Routine childhood immunization: Neither type nor quantity or timing of vaccinations, including BCG vaccine, HIB vaccine, diphtheria, tetanus and pertussis vaccine, measles, mumps and rubella vaccine, hepatitis B vaccine, varicella vaccine, or tick-born encephalitis vaccine, have been associated with the development of islet antibodies and diabetes [86,89,90]. At least two studies even showed a possible protective effect of the measles-mumps-rubella vaccine .
Perinatal factors: Environmental risk factors may play a role early in life, possibly in utero. Several studies have investigated perinatal determinants for developing T1DM. Offsprings of a T1DM parent have an increased risk of developing diabetes, the risk being higher if the father has diabetes. Although the relation of maternal age and birth order to risk of T1DM is complex, several studies found that maternal age over 35–40 years  and/or increasing birth order  were associated with an increase in T1DM. There also seems to be a relatively weak but significant association between increasing birth weight and increasing risk of T1DM [94,95], although several other case-control studies have not found any association .
Interestingly, several studies suggest that early weight gain and/or rapid linear growth are risk factors for development of not only type 2 but also T1DM in children .
Dietary factors: Cow’s milk or wheat introduced at weaning trigger insulitis and diabetes in animal models perhaps through a molecular mimicry . Breast-feeding may be viewed as a surrogate for the delay in the introduction of diabetogenic substances present in formula or early childhood diet. Several human studies suggested an association between short duration of breastfeeding and increase in T1DM [99,100], while cohort studies failed to find an association between cow’s milk and beta-cell autoimmunity [90,101] (Figure 2.4). Interestingly, a study from Finland suggested that current cow’s milk consumption was more closely linked to prediabetic autoimmunity and diabetes than infant exposure . To resolve this controversy, a dietary intervention trial to prevent T1DM by a short-term elimination of cow’s milk from infant diet (TRIGR) is underway .
Concerning the introduction of cereals, the DAISY  and the German BabyDiab  studies found that early introduction of gluten before the age of 3 months increases the risk of development of β-cell autoimmunity (Figure 2.4). However, delaying gluten exposure until the age of 12 months in the BabyDiet cohort did not substantially reduce the risk for islet autoimmunity in genetically at-risk children .
Vitamins and dietary supplements: Studies in vitro have shown that vitamin D3 is immunosuppressive or immunomodulating and studies in experimental models of autoimmunity have shown vitamin D to be protective . Results from clinical studies have been somewhat conflicting. While a European study reported a protective effect of vitamin D supplementation in infancy against T1DM[108,109], the DAISY cohort study did not show any association between vitamin D intake or 25(OH)D levels throughout childhood with the risk of islet autoimmunity or progression to T1DM  (Figure 2.4). On the other hand, maternal intake of vitamin D through food during pregnancy was associated with a protective effect on the appearance of islet autoimmunity in DAISY offspring ; however, these findings were not confirmed in a more recent Finnish study (DIPP) . A study in Norway  found that cod liver oil taken during pregnancy was associated with reduced risk of T1DM, suggesting that vitamin D and/or the omega-3 fatty acids in the cod liver oil have a protective effect. In the prospective DAISY study, dietary intake of omega-3 fatty acids and the omega-3 fatty acid content of erythrocyte membranes were associated with reduced risk of islet autoimmunity  (Figure 2.4). However, neither intake nor membrane levels of omega-3 or omega-6 fatty acids were associated with risk of developing T1DM in those children with islet autoimmunity . A Nutritional Intervention to Prevent T1D (NIP) is currently underway to examine whether nutritional supplements with docosahexaenoic acid (DHA), given during the last trimester of pregnancy and the first few years of life, can prevent development of islet autoantibodies .
Gene–environment interaction in clinical type 1 diabetes
T1DM is likely caused by an interactive effect of genetic and environmental factors within a limited age-window. While both the susceptibility genes and the candidate environmental exposures appear to be quite common, the likelihood of these causal components meeting within the susceptibility age-window is low. To investigate the environmental causes of T1DM, the study subjects have to be screened for known susceptibility gene markers so that gene–environment interactions can be accounted for.
Interaction between HLA Class II alleles and viral infection:
Susceptibility to diabetogenic enteroviruses in humans appears to be genetically restricted by HLA-DR and DQ alleles . However, the allelic specificity is controversial and may depend on the viral type and epidemicity. In general, the HLA-DR3 allele, present in most patients with T1DM, is associated with viral persistence.
Interaction between HLA Class II alleles and infant diet:
Few studies to date have examined a possibility of an interaction between the HLA genes and dietary exposures . The epidemiologic data are limited, but suggest that an early exposure to cow’s milk in relatives with HLA-DR3/4,DQB1*0302, DR3/3 or DRx/4,DQB1*0302 is not associated with development of beta-cell autoantibodies .