*3.5.1. Seasonal changes in the incidence of type 1 diabetes mellitus*

The seasonality of onset or diagnosis of type 1 diabetes has been extensively studied and the re‐ sults, so far, are conflicting (Moltchanova, Schreier et al. 2009). However, an increment of type 1 diabetes incidence during the winter has been reported by manifold studies (for details: Padaiga, Tuomilehto et al. 1999) from different countries, e.g. Australia (Elliott, Lucas et al. 2010), the Unit‐ ed States (Gorham, Barrett-Connor et al. 2009), Chile (Durruty, Ruiz et al. 1979; Santos, Carrasco et al. 2001), Sweden (Samuelsson, Carstensen et al. 2007; Ostman, Lonnberg et al. 2008), Norway (Joner and Sovik 1981), Greece (Kalliora, Vazeou et al. 2011), and the Czech Republic (Cinek, Sumnik et al. 2003). Recently, Jarosz-Chobot et al. (2011) reported that a significant increase in type 1 diabetes incidence among children over 4 years of age was observed in the autumn–winter season (p = 0.137 for the age group 0–4 years and p < 0.001 for the age groups 5-9 and 10-14 years). These findings were confirmed by other studies in Poland (Pilecki, Robak-Kontna et al. 2003; Zubkiewicz-Kucharska and Noczynska 2010). Other, partially incomparable, studies revealed no seasonal pattern in the onset or diagnosis of type 1 diabetes mellitus (Levy-Marchal, Papoz et al. 1990; Muntoni and Songini 1992; Ye, Chen et al. 1998) or reported seasonal changes only for subgroups (Michalkova, Cernay et al. 1995; Douglas, McSporran et al. 1999; Padaiga, Tuomileh‐ to et al. 1999). Moltchanova et al. (2009) analyzed data from 105 centers in 53 countries: however, only 42 centers exhibited significant seasonality (p < 0.05) in the incidence of type 1 diabetes when the data were pooled for age and sex (Moltchanova, Schreier et al. 2009). Centers further away from the equator were on average more likely to exhibit seasonality (p < 0.001). Although the ma‐ jority of the published data suggests seasonal-dependent changes in the incidence of type 1 dia‐ betes mellitus, further research is needed to complete the picture. Especially populations living below the 30th parallel north should be studied, the populations themselves should be investi‐ gated more deeply, and the sample sizes should be increased to gain adequate power to detect seasonal changes in low-incidence populations.

betes type 1 had most births during these months (1972; Neu, Kehrer et al. 2000). A Ukrainian group found that type 1 diabetes was some 30% more common among persons born in April than among persons born in December (Vaiserman, Carstensen et al. 2007). McKinney et al. analyzed data from 19 European countries, but found no uniform seasonal pattern of birth in childhood diabetes patients across European populations, either overall or according to sex and age (McKinney 2001). Small Turkish studies did not reveal any significant differences of the season of birth in type 1 diabetic vs. metabolically healthy children (Evliyaoglu, Ocal et al. 2002; Karaguzel, Ozer et al. 2007). The controversial results might be explained by the composition of most study samples: Laron et al. found a pattern in the seasonality of month of birth only in ethnically homo‐ genous populations (such as Ashkenazi Jews, Israeli Arabs, individuals in Sardinia and Canter‐ bury, New Zealand, and Afro-Americans), but not in heterogeneous populations (such as in Sydney, Pittsburgh and Denver; Laron, Lewy et al. 2005). Thereby, it becomes likely that ethni‐ cally heterogeneous populations comprising a mixture of patients with various genetic back‐ grounds and environmental exposures mask the different seasonality pattern of month of birth that many children with diabetes present when compared to the general population (Laron,

The Epidemiology of Type 1 Diabetes Mellitus

http://dx.doi.org/10.5772/52893

11

Authors describing a relationship between season of birth and susceptibility for type 1 dia‐ betes have attributed this to intrauterine infections, dietary intake of certain nutrients and possible toxic food components, short duration of breastfeeding, early exposure to cows' milk proteins, and vitamin D deficiency (Vaiserman, Carstensen et al. 2007). Since most of these factors vary with season, one would expect a difference in the seasonal birth pattern between the general population and those children who develop diabetes. A possible link between environmental factors and type 1 diabetes mellitus manifestation was provided by Badenhoop et al. They found HLA susceptibility genes to be in different proportions of pa‐ tients either born in different seasons of the year or having manifested their disease in differ‐

It has been proposed that much of the current variation in the incidence of type 1 diabetes is due in part to differing distributions of ethnicity throughout the world. Many large studies of type 1 diabetes have provided evidence that the ethnic background is one of the most im‐ portant risk factors for type 1 diabetes (Vehik and Dabelea 2010). It can be assumed that there is a genetically founded – and thereby ethnically associated – varying susceptibility for type 1 diabetes. The onset of the disease is then triggered by ubiquitous environmental fac‐ tors (Knip, Veijola et al. 2005; Knip and Simell 2012). In general, susceptibility to type 1 dia‐ betes is attributable to genes that link disease progression to distinct steps in immune

One half of the genetic susceptibility for type 1 diabetes is explained by the HLA (human leukocyte antigen) genes (Knip and Simell 2012). It becomes conclusive that the main re‐ search focus is on ethnic variances in HLA-haplotypes and its association with type 1 diabe‐ tes (Lipton, Drum et al. 2011; Noble, Johnson et al. 2011). Based on the presence of two highrisk HLA-DQA1/B1 haplotypes, an investigation in the United States revealed that

ent historical periods over time (Badenhoop, Kahles et al. 2009).

activation, expansion, and regulation (Nepom and Buckner 2012).

Lewy et al. 2005).

**3.6. Ethnic differences**

According to the published literature, the seasonal changes in the incidence of type 1 diabe‐ tes are likely to be caused by changes of the (auto-)immune activity. The first point is that a reduced ultraviolet radiation exposure during the winter months may lead to reduced vita‐ min D levels. Thereby, the inhibitory effect of vitamin D on Th1-lymphocytes decreases. The second point is the stimulation of the immune system especially by viral infections during the winter months. The result of both could be a higher (auto-)immune activity that causes ß-cell destruction.

#### *3.5.2. Effects of the season of birth on the incidence of type 1 diabetes mellitus*

Possible influences of the season of birth are discussed for many autoimmune diseases, e.g. mul‐ tiple sclerosis, Hashimoto thyreoditis, or Grave's disease (Krassas, Tziomalos et al. 2007). Spring births were associated with increased likelihood of type 1 diabetes, but possibly not in all United States regions (Kahn, Morgan et al. 2009). An Egyptian group reported that 48.3% of diabetic chil‐ dren were delivered during summer months (Ismail, Kasem et al. 2008). A German investigation showed children and adolescents with diabetes being significantly less often born during the months April-June and July-September (Neu, Kehrer et al. 2000). This seasonality pattern was different from those registered in Israel, Sardinia and Slovenia, in which the population with dia‐ betes type 1 had most births during these months (1972; Neu, Kehrer et al. 2000). A Ukrainian group found that type 1 diabetes was some 30% more common among persons born in April than among persons born in December (Vaiserman, Carstensen et al. 2007). McKinney et al. analyzed data from 19 European countries, but found no uniform seasonal pattern of birth in childhood diabetes patients across European populations, either overall or according to sex and age (McKinney 2001). Small Turkish studies did not reveal any significant differences of the season of birth in type 1 diabetic vs. metabolically healthy children (Evliyaoglu, Ocal et al. 2002; Karaguzel, Ozer et al. 2007). The controversial results might be explained by the composition of most study samples: Laron et al. found a pattern in the seasonality of month of birth only in ethnically homo‐ genous populations (such as Ashkenazi Jews, Israeli Arabs, individuals in Sardinia and Canter‐ bury, New Zealand, and Afro-Americans), but not in heterogeneous populations (such as in Sydney, Pittsburgh and Denver; Laron, Lewy et al. 2005). Thereby, it becomes likely that ethni‐ cally heterogeneous populations comprising a mixture of patients with various genetic back‐ grounds and environmental exposures mask the different seasonality pattern of month of birth that many children with diabetes present when compared to the general population (Laron, Lewy et al. 2005).

Authors describing a relationship between season of birth and susceptibility for type 1 dia‐ betes have attributed this to intrauterine infections, dietary intake of certain nutrients and possible toxic food components, short duration of breastfeeding, early exposure to cows' milk proteins, and vitamin D deficiency (Vaiserman, Carstensen et al. 2007). Since most of these factors vary with season, one would expect a difference in the seasonal birth pattern between the general population and those children who develop diabetes. A possible link between environmental factors and type 1 diabetes mellitus manifestation was provided by Badenhoop et al. They found HLA susceptibility genes to be in different proportions of pa‐ tients either born in different seasons of the year or having manifested their disease in differ‐ ent historical periods over time (Badenhoop, Kahles et al. 2009).

#### **3.6. Ethnic differences**

*3.5.1. Seasonal changes in the incidence of type 1 diabetes mellitus*

seasonal changes in low-incidence populations.

ß-cell destruction.

10 Type 1 Diabetes

The seasonality of onset or diagnosis of type 1 diabetes has been extensively studied and the re‐ sults, so far, are conflicting (Moltchanova, Schreier et al. 2009). However, an increment of type 1 diabetes incidence during the winter has been reported by manifold studies (for details: Padaiga, Tuomilehto et al. 1999) from different countries, e.g. Australia (Elliott, Lucas et al. 2010), the Unit‐ ed States (Gorham, Barrett-Connor et al. 2009), Chile (Durruty, Ruiz et al. 1979; Santos, Carrasco et al. 2001), Sweden (Samuelsson, Carstensen et al. 2007; Ostman, Lonnberg et al. 2008), Norway (Joner and Sovik 1981), Greece (Kalliora, Vazeou et al. 2011), and the Czech Republic (Cinek, Sumnik et al. 2003). Recently, Jarosz-Chobot et al. (2011) reported that a significant increase in type 1 diabetes incidence among children over 4 years of age was observed in the autumn–winter season (p = 0.137 for the age group 0–4 years and p < 0.001 for the age groups 5-9 and 10-14 years). These findings were confirmed by other studies in Poland (Pilecki, Robak-Kontna et al. 2003; Zubkiewicz-Kucharska and Noczynska 2010). Other, partially incomparable, studies revealed no seasonal pattern in the onset or diagnosis of type 1 diabetes mellitus (Levy-Marchal, Papoz et al. 1990; Muntoni and Songini 1992; Ye, Chen et al. 1998) or reported seasonal changes only for subgroups (Michalkova, Cernay et al. 1995; Douglas, McSporran et al. 1999; Padaiga, Tuomileh‐ to et al. 1999). Moltchanova et al. (2009) analyzed data from 105 centers in 53 countries: however, only 42 centers exhibited significant seasonality (p < 0.05) in the incidence of type 1 diabetes when the data were pooled for age and sex (Moltchanova, Schreier et al. 2009). Centers further away from the equator were on average more likely to exhibit seasonality (p < 0.001). Although the ma‐ jority of the published data suggests seasonal-dependent changes in the incidence of type 1 dia‐ betes mellitus, further research is needed to complete the picture. Especially populations living below the 30th parallel north should be studied, the populations themselves should be investi‐ gated more deeply, and the sample sizes should be increased to gain adequate power to detect

According to the published literature, the seasonal changes in the incidence of type 1 diabe‐ tes are likely to be caused by changes of the (auto-)immune activity. The first point is that a reduced ultraviolet radiation exposure during the winter months may lead to reduced vita‐ min D levels. Thereby, the inhibitory effect of vitamin D on Th1-lymphocytes decreases. The second point is the stimulation of the immune system especially by viral infections during the winter months. The result of both could be a higher (auto-)immune activity that causes

Possible influences of the season of birth are discussed for many autoimmune diseases, e.g. mul‐ tiple sclerosis, Hashimoto thyreoditis, or Grave's disease (Krassas, Tziomalos et al. 2007). Spring births were associated with increased likelihood of type 1 diabetes, but possibly not in all United States regions (Kahn, Morgan et al. 2009). An Egyptian group reported that 48.3% of diabetic chil‐ dren were delivered during summer months (Ismail, Kasem et al. 2008). A German investigation showed children and adolescents with diabetes being significantly less often born during the months April-June and July-September (Neu, Kehrer et al. 2000). This seasonality pattern was different from those registered in Israel, Sardinia and Slovenia, in which the population with dia‐

*3.5.2. Effects of the season of birth on the incidence of type 1 diabetes mellitus*

It has been proposed that much of the current variation in the incidence of type 1 diabetes is due in part to differing distributions of ethnicity throughout the world. Many large studies of type 1 diabetes have provided evidence that the ethnic background is one of the most im‐ portant risk factors for type 1 diabetes (Vehik and Dabelea 2010). It can be assumed that there is a genetically founded – and thereby ethnically associated – varying susceptibility for type 1 diabetes. The onset of the disease is then triggered by ubiquitous environmental fac‐ tors (Knip, Veijola et al. 2005; Knip and Simell 2012). In general, susceptibility to type 1 dia‐ betes is attributable to genes that link disease progression to distinct steps in immune activation, expansion, and regulation (Nepom and Buckner 2012).

One half of the genetic susceptibility for type 1 diabetes is explained by the HLA (human leukocyte antigen) genes (Knip and Simell 2012). It becomes conclusive that the main re‐ search focus is on ethnic variances in HLA-haplotypes and its association with type 1 diabe‐ tes (Lipton, Drum et al. 2011; Noble, Johnson et al. 2011). Based on the presence of two highrisk HLA-DQA1/B1 haplotypes, an investigation in the United States revealed that Caucasians are at the highest and Latinos are at the second-highest risk for developing type 1 diabetes compared to all other ethnic groups (Lipton, Drum et al. 2011). However, there is accumulating evidence that the proportion of subjects with newly diagnosed type 1 diabetes and high-risk HLA genotypes has decreased over the last decades, whereas the proportion of those with low-risk or even protective HLA genotypes has increased (Hermann, Knip et al. 2003; Gillespie, Bain et al. 2004).

tional Diabetes Federation 2011), there 116,100 cases of type 1 diabetes in the Europe, 64,900 in the Middle East and North Africa region, 36,100 in the Africa, 94,700 in the North America and Carib‐ bean, 36,100 in the South and Central America, 111,500 in the South-East Asia and 30,700 in the Western Pacific region. In accordance with incidence rates differing regionally within countries and also among different countries, the prevalence of type 1 diabetes mellitus varies in a broad

The Epidemiology of Type 1 Diabetes Mellitus

http://dx.doi.org/10.5772/52893

13

**Country Sampling Period Prevalence** Finland 2000–2005 427.5 Sweden 2001–2005 270.5 Norway 1999–2003 182.4 United Kingdom 1989–2003 158.3 Canada 1990–1999 146.7 Denmark 1996–2005 141.2 Australia 1999–2008 137.8 United States 2002–2003 135.6 Germany 1989–2003 126.7 Netherlands 1996–1999 124.8 Czech Republic 1989–2003 117.5 New Zealand 1999–2000 115.9 Belgium 1989–2003 107.7 Ireland 1997 107.3 Austria 1989–2003 97.6 Portugal 1994–1998 95.5 Luxembourg 1989–2003 94.9 Slovak Republic 1989–2003 94.2 Iceland 1994–1998 91.1 Poland 1989–2003 85.7 France 1998–2004 84.5 Greece 1995–1999 80.2 Hungary 1989–2003 76.5 Spain 1989–2003 74.6 Switzerland 1991–1999 61.1 Italy 1990–1999 59.9 Turkey 1992–1996 19.8 Japan 1998–2001 15.7 Mexico 1990–1993 8.1 Korea 1990–1991 6.7

**Table 3.** The prevalence of type 1 diabetes in children younger than 15 years in different OECD countries. Data are based on estimations of the International Diabetes Federation (2009) and related to 100,000 children (0 to 14 years of

age) of each country.

range. The prevalence of type 1 diabetes in different countries is summarized in Table 3.

The second half of the genetic susceptibility for type 1 diabetes is caused by more than 50 non-HLA genetic polymorphisms (Knip and Simell 2012). Nowadays, there are more than 60 gene loci contributing to the susceptibility of developing type 1 diabetes (Morahan 2012), but this overwhelming list of type 1 diabetes risk genes exerts little influence on the clinical management of children that are at high risk. Conclusively, it is necessary to place the genet‐ ics of type 1 diabetes in a more amenable clinical context (Morahan 2012).

Despite the fact that there is consensus about the different genetic type 1 diabetes suscepti‐ bility among different ethnic groups, these differences cannot explain the complete variance of type 1 diabetes incidence and prevalence. Furthermore, the annual increment of type 1 diabetes incidence cannot be explained by changing genetic susceptibility. Together with the fact that many individuals are genetically highly susceptible for type 1 diabetes, it becomes conclusive that environmental factors play a crucial role in the onset of the disease and its epidemiology (Knip and Simell 2012).
