**2. Obesity as a accelerate factor to type 1 diabetes mellitus development**

Studies with streptozotocin-induced diabetic baboons showed that to have an abnormal glucose tolerance it is necessary an isolated huge loss of beta-cell mass or a moderate loss of these cells associated to an IR (McCulloch et al., 1991), that could be in humans the physiologically IR of adolescence (Acerini et al., 2000) or gestation (Buschard et al., 1987), periods with higher incidence of T1D, or pathological situations like infection (usually one of the triggering factors of T1D) or weight gain.

Others studies suggest that the increase in the body mass index (BMI) and the consequent IR may accelerate the β cell destruction process in individuals predisposed to T1D, due to the release of obesity-related cytokines that show inflammatory and/or immunomodulatory properties (Aldhahi & Hamdy, 2003), triggering diabetes. This hypothesis may be reinforced by one study that correlated high anti-GAD levels with high BMI (Rolandsson et al., 1999). Two interesting data from studies with non-obese diabetic (NOD) mice are that hyperinsulinemia, an IR marker, precede clinical T1D (Armani et al., 1998) and that T1D incidence falls after treatment with rosiglitazone, an insulin sensitizer drug (Beales & Pozzili, 2002).

The IR, autoimmunity and apoptosis of the β cells constitutes the three factors of the called "accelerator hypothesis", proposed by Wilkin (Wilkin, 2001), that contemplate the factors presented in both more common types of diabetes, that is, T2D and T1D. There is a constitucional (intrinsic) high speed of apoptosis of β cells that is necessary to the development of diabetes, but rarely enough. The other two factors, extrinsic, that can speed the apoptosis of beta-cells are IR (result of weight gain and/or physical inactivity) and autoimmunity against beta-cells.

It is known that obese individuals have elevated serum levels of leptin, a cytokine secreted by adipocytes in proportion to adipose tissue mass and that is responsible, among other functions, for regulating food intake and thus BMI. Moreover, leptin controls the cellular immune response and is involved in the pathogenesis of autoimmune diseases (Lord, 2002). Studies have shown that administration of leptin in NOD mice promoted an early inflammatory infiltrate in the pancreatic islets, increased production of interferon gamma (IFN-gamma) by T lymphocytes, which accelerated the establishment of a T1D (Matarese, 2002 e 2005).

On the other hand, adiponectin, another important cytokine produced by adipose tissue, inversely proportional to its fat mass, can decrease the systemic and pancreatic islets inflammatory process, acting as a protective factor in the development of T1D, in addition to reducing IR (Kadowaki et al., 2006; Wellen & Hotamisligil, 2005).

However, development report (OECD, 2009) from 16 countries does not show any obvious relationship between national estimates of childhood obesity prevalence and incidence rates of T1D (Table 2). Therefore, obesity does not account for the wide between-country differences in T1D incidence, which range from 0.57 per 100 000 person-years in China to more than 48 per 100 000 person-years in Sardinia and Finland in the 0- to 14-year age group (Daneman, 2006).

On the other hand, in a meta-analysis of nine studies (eight case–control studies and one cohort study) comprising a total of 2658 cases (Verbeeten et al., 2011), seven reported a significant association between childhood obesity, BMI or %weight-for-height and increased risk for T1D. Four of these studies reported childhood obesity as a categorical exposure and

Obesity in the Natural History of Type 1 Diabetes Mellitus: Causes and Consequences 185

Fig. 3. Meta-analysis (random-effects inverse variance model) of studies of childhood BMI as

If on one hand intensive insulin prevents microvascular and macrovascular complications associated with poor glycemic control, the other brings an increased risk of severe hypoglycemia and weight gain, traditionally viewed as a normalization of weight, i.e. the correction of glycosuria, diuresis, and wasting with the initiation of insulin therapy. Insulin stimulates lipogenesis, inhibits protein catabolism, and slows basal metabolism. Other important aspect is the abnormal physiological route of insulin via its peripheral administration in those with T1D, which is also associated with reduced energy metabolism (Charlton & Nair, 1998). Classically normal or underweight, the phenotype of the T1D individuals is thus changing. A follow-up of 18 years of 589 individuals from the Pittsburgh Epidemiology of Diabetes Complications Study (EDC), a cohort of childhood-onset T1D, showed an increase in the prevalence of overweight by 47% (from 28.6% at baseline to 42%) and of obesity by sevenfold (from 3.4% at baseline to 22.7%), concomitantly with the highest prevalence of intensive insulin therapy - 7% and 82% were on intensive insulin therapy (≥ 3 insulin injections per day or on insulin pump) at baseline and 18 years after, respectively (Conway et al., 2010). Although injection frequency increased, total daily insulin dose decreased from 0.76 to 0.62 Ukgday. Figure 4 shows the temporal patterns in the prevalence of being overweight and obese and the use of intensive insulin treatment, and these data was not influenced by the aging of the cohort and survivorship, as can be seen on Table 3. (age-group-specific prevalence for the 40–49-year-old age group by time period): overweight or obesity were present in 25% of the T1D individuals in 1986–1988 and in 68.2% in 2004–

a risk factor for subsequent T1D. Odds ratios correspond to a 1-unit increase in BMI

standard deviation score (SDS)(Verbeeten et al., 2011).

2007 (Conway et al., 2010).

**3. Obesity after clinical Type 1 diabetes diagnostic** 

produced a pooled odds ratio of 2.03 (95% CI 1.46–2.80) for subsequent T1D, but with age at obesity assessment varying from age 1 to 12 years (Figure 2). A dose–response relationship was supported by a continuous association between childhood BMI and subsequent T1D in a meta-analysis of five studies (pooled odds ratio 1.25 (95%CI 1.04–1.51) per 1 SD higher BMI) (Figure 3).


Table 2. Relationship between Type 1 diabetes incidence and prevalence of childhood overweight or obesity in 16 Organization for Economic Co-Operation and Development (OECD) countries, from Health at a Glance 2009: OECD Indicators (OECD, 2009).

Fig. 2. Meta-analysis (fixed-effects inverse variance model) of studies of childhood obesity as a risk factor for subsequent T1D (Verbeeten et al., 2011).

produced a pooled odds ratio of 2.03 (95% CI 1.46–2.80) for subsequent T1D, but with age at obesity assessment varying from age 1 to 12 years (Figure 2). A dose–response relationship was supported by a continuous association between childhood BMI and subsequent T1D in a meta-analysis of five studies (pooled odds ratio 1.25 (95%CI 1.04–1.51) per 1 SD higher

Table 2. Relationship between Type 1 diabetes incidence and prevalence of childhood overweight or obesity in 16 Organization for Economic Co-Operation and Development (OECD) countries, from Health at a Glance 2009: OECD Indicators (OECD, 2009).

Fig. 2. Meta-analysis (fixed-effects inverse variance model) of studies of childhood obesity as

a risk factor for subsequent T1D (Verbeeten et al., 2011).

% of children aged 11-15 years overweight or obese

Country T1D incidence rate in children aged 0-14 years (per 100.000 person-years)

Finland 57,4 15,8 Sweden 41 10,5 Norway 27,9 10 UK 24,5 12 Denmark 22,2 9,7 Canada 21,7 21,3 USA 20,8 29,8 Netherlands 18,8 8 Germany 18 12 Ireland 16,3 14,2 Iceland 14,7 14,5 Spain 13 16,7 Poland 12,9 11,2 France 12,2 10,5 Greece 9,9 18,8 Italy 8,4 18,3

BMI) (Figure 3).

Fig. 3. Meta-analysis (random-effects inverse variance model) of studies of childhood BMI as a risk factor for subsequent T1D. Odds ratios correspond to a 1-unit increase in BMI standard deviation score (SDS)(Verbeeten et al., 2011).
