**3. Insulin and lipoprotein metabolism**

Insulin plays a central role in the regulation of lipid metabolism (Vergès, 2001). The main sites of action of insulin on lipoprotein metabolism are shown in Figure 2.

In adipose tissue, insulin inhibits the hormone-sensitive lipase. Thus, insulin has an antilipolytic action, promoting storage of triglycerides in the adipocytes and reducing release of free fatty acids from adipose tissue in the circulation.

Insulin inhibits VLDL production from the liver. In normal subjects, it has been shown that insulin induces a 67% decrease of VLDL-triglyceride production and a 52% decrease of VLDL-apoB production (Lewis et al., 1993; Malmström et al., 1998). Insulin reduces VLDL production by diminishing circulating free fatty acids (due to its antilipolytic effect), which

Lipid Disorders in Type 1 Diabetes 49

Triglyceride-rich lipoproteins (chylomicrons, VLDLs) are increased leading to hypertriglyceridemia. This is mainly due to decreased lipoprotein lipase activity (Vergès, 2001; Dullaart, 1995). Diabetic ketoacidosis is a situation of severe insulin deficiency with reduced lipoprotein lipase activity as a consequence, because insulin usually stimulates its activity. Decreased lipoprotein lipase activity leads to profound reduction of triglyceriderich lipoprotein catabolism (Taskinen, 1987). In this condition of severe insulin deficiency, reduced catabolism of triglyceride-rich lipoproteins is, by far, the main factor involved in hypertriglyceridemia. This hypertriglyceridemia resolves rapidly after well titrated insulin

LDL-cholesterol is decreased during diabetic ketoacidosis (Weidman et al., 1982). This fall in plasma LDL-cholesterol level is the direct consequence of the reduction of triglyceride-rich

 In diabetic ketoacidosis, HDL-cholesterol level is significantly decreased (Weidman et al., 1982). This is a consequence of hypertrigliceridemia observed in this condition. Indeed, the augmented level of plasma triglyceride-rich lipoproteins drives, through CETP, the transfer of triglycerides from triglyceride-rich lipoproteins to HDLs leading to the formation of triglyceride-rich HDL particles. HDLs enriched in triglycerides become very good substrate for hepatic lipase, leading to increase their catabolism and, thus, to decrease plasma HDLcholesterol level. This low HDL-cholesterol condition resolves rapidly after well titrated

Patients with treated type 1 diabetes may show quantitative lipid disorders. In a prospective study performed in 895 young subjects with type 1 diabetes, 20.1% had plasma triglycerides above 1.7 mmol/l, 9.6% had LDL-cholesterol above 3.4 mmol/l and 25.9% had non-HDL cholesterol above 3.4 mmol/l (Marcovecchio et al., 2009). It has been shown that abnormal lipid levels, in type 1 diabetes, predict worse cardiovascular outcomes (Soedamah-Muthu et al., 2004). HbA1c has been shown to be independently correlated with LDL-cholesterol, non-HDL cholesterol and triglyceride levels, indicating that these disorders were mostly observed in patients with poor glycemic control (Marcovecchio et al., 2009). In a British follow-up study of 229 children with type 1 diabetes, LDL cholesterol and non-HDL cholesterol values increased with duration of diabetes (Edge et al., 2008). In that study, total cholesterol, triglycerides and non-HDL cholesterol were positively correlated with HbA1c and around 10% of the patients had lipid values outside recommendations (Edge et al., 2008). In a large study performed in 29 979 patients with type 1 diabetes, multivariate analyses showed a significant positive association between HbA1c and total cholesterol (p<0.0001), LDL cholesterol (p<0.0001) and a significant negative association between HbA1c and HDL cholesterol (p<0.0001) (Schwab et al., 2009). In the Diabetes Control and Complications Trial (DCCT), HbA1c correlated positively with total cholesterol, LDLcholesterol and triglycerides at baseline (The DCCT Research Group, 1992). Data from the Coronary Artery Calcification in type 1 diabetes (CACTI) study, which examined 652 patients with type 1 diabetes, have shown, in patients not using hypolipidemic agents, that a higher HbA1c was associated with significantly higher levels of total cholesterol, triglycerides, LDL cholesterol and non-HDL cholesterol (Maahs et al., 2010). In that study, 1% change in HbA1c was associated with an increase of 0.101 mmol/l (4 mg/dl) for total cholesterol, of 0.052 mmol/l (4.5 mg/dl) for triglycerides, of 0.103 mmol/l (4 mg/dl) for LDL cholesterol and of 0.129 mmol/l (5 mg/dl) for non-HDL cholesterol (Maahs et al.,

lipoprotein catabolism, due to decreased lipoprotein lipase activity (see above).

therapy (Weidman et al., 1982).

insulin therapy (Weidman et al., 1982).

**4.2 Treated type 1 diabetes** 

are substrates for VLDL, but also by a direct inhibitory effect in the hepatocyte (Malmström et al., 1998). Insulin is a potent activator of lipoprotein lipase (LPL), promoting the catabolism of triglyceride-rich lipoproteins and reducing, as a consequence, plasma triglyceride level. Insulin not only enhances LPL activity (Brunzell et al., 1998), but has also a direct positive effect on LPL gene, promoting LPL synthesis (Fried etal., 1993). Insulin promotes the clearance of LDL, by increasing LDL B/E receptor expression and activity (Chait et al., 1979, Mazzone et al., 1984).

Insulin acts also on HDL metabolism by activating LCAT and hepatic lipase activities (Ruotolo et al., 1994).

**Figure 2** *VLDL*: Very Low Density Lipoprotein; *IDL*: Intermediate Density Lipoprotein, *LDL*: Low Density Lipoprotein; *HDL*: High Density Lipoprotein; *LPL*: LipoProtein Lipase; *HL*: Hepatic Lipase; *CETP*: Cholesteryl Ester Transfer Protein; *LCAT*: Lecithin-Cholesterol Acyl Transferase; *FFA*: Free Fatty Acids ; *B/E rec*.: receptor B/E (LDL receptor); *TG*: Triglycerides; *CE*: Cholesterol Esters. **1**: insulin inhibits hormone-sensitive lipase. **2** : insulin activates LipoProtein Lipase (LPL). **3**: insulin inhibits hepatic VLDL production. **4**: insulin increases LDL B/E receptor expression. **5**: insulin activates LCAT. **6**: insulin activates Hepatic Lipase (HL).

Fig. 2. Main effects of insulin on lipoprotein metabolism.

#### **4. Quantitative lipid abnormalities in type 1 diabetes**

#### **4.1 Untreated (diabetic ketoacidosis) type 1 diabetes**

In type 1 diabetic patients with diabetic ketoacidosis, quantitative lipid abnormalities are observed, due to insulin deficiency.

are substrates for VLDL, but also by a direct inhibitory effect in the hepatocyte (Malmström et al., 1998). Insulin is a potent activator of lipoprotein lipase (LPL), promoting the catabolism of triglyceride-rich lipoproteins and reducing, as a consequence, plasma triglyceride level. Insulin not only enhances LPL activity (Brunzell et al., 1998), but has also a direct positive effect on LPL gene, promoting LPL synthesis (Fried etal., 1993). Insulin promotes the clearance of LDL, by increasing LDL B/E receptor expression and activity

Insulin acts also on HDL metabolism by activating LCAT and hepatic lipase activities

**+**

**HDL2**

**INSULIN**

**Figure 2**

*VLDL*: Very Low Density Lipoprotein; *IDL*: Intermediate Density Lipoprotein, *LDL*: Low Density Lipoprotein; *HDL*: High Density Lipoprotein; *LPL*: LipoProtein Lipase; *HL*: Hepatic Lipase; *CETP*: Cholesteryl Ester Transfer Protein; *LCAT*: Lecithin-Cholesterol Acyl Transferase; *FFA*: Free Fatty Acids ; *B/E rec*.: receptor B/E (LDL receptor); *TG*: Triglycerides; *CE*: Cholesterol Esters. **1**: insulin inhibits hormone-sensitive lipase. **2** : insulin activates LipoProtein Lipase (LPL). **3**: insulin inhibits hepatic VLDL production. **4**: insulin increases LDL B/E receptor expression. **5**: insulin activates LCAT. **6**: insulin

In type 1 diabetic patients with diabetic ketoacidosis, quantitative lipid abnormalities are

**HDLn**

**LIVER**

*LRP*

**Chylomicrons**

**LPL**


**HL**

**LCAT FFA**

**VLDL LDL**

**INSULIN**

**+**

**IDL**

**CE TG**

**+**

**LPL**

**CETP**

**HDL3**

**CHU**

**Dijon**

**+**

**liver**

*B/E rec.*

**Cell Periph.**

*B/E rec.*

(Chait et al., 1979, Mazzone et al., 1984).

**Remnants Chylomicron-**

(Ruotolo et al., 1994).

**adipose tissue**

**Lipase**

activates Hepatic Lipase (HL).

observed, due to insulin deficiency.


**INSULIN**

Fig. 2. Main effects of insulin on lipoprotein metabolism.

**4.1 Untreated (diabetic ketoacidosis) type 1 diabetes** 

**4. Quantitative lipid abnormalities in type 1 diabetes** 

Triglyceride-rich lipoproteins (chylomicrons, VLDLs) are increased leading to hypertriglyceridemia. This is mainly due to decreased lipoprotein lipase activity (Vergès, 2001; Dullaart, 1995). Diabetic ketoacidosis is a situation of severe insulin deficiency with reduced lipoprotein lipase activity as a consequence, because insulin usually stimulates its activity. Decreased lipoprotein lipase activity leads to profound reduction of triglyceriderich lipoprotein catabolism (Taskinen, 1987). In this condition of severe insulin deficiency, reduced catabolism of triglyceride-rich lipoproteins is, by far, the main factor involved in hypertriglyceridemia. This hypertriglyceridemia resolves rapidly after well titrated insulin therapy (Weidman et al., 1982).

LDL-cholesterol is decreased during diabetic ketoacidosis (Weidman et al., 1982). This fall in plasma LDL-cholesterol level is the direct consequence of the reduction of triglyceride-rich lipoprotein catabolism, due to decreased lipoprotein lipase activity (see above).

 In diabetic ketoacidosis, HDL-cholesterol level is significantly decreased (Weidman et al., 1982). This is a consequence of hypertrigliceridemia observed in this condition. Indeed, the augmented level of plasma triglyceride-rich lipoproteins drives, through CETP, the transfer of triglycerides from triglyceride-rich lipoproteins to HDLs leading to the formation of triglyceride-rich HDL particles. HDLs enriched in triglycerides become very good substrate for hepatic lipase, leading to increase their catabolism and, thus, to decrease plasma HDLcholesterol level. This low HDL-cholesterol condition resolves rapidly after well titrated insulin therapy (Weidman et al., 1982).

#### **4.2 Treated type 1 diabetes**

Patients with treated type 1 diabetes may show quantitative lipid disorders. In a prospective study performed in 895 young subjects with type 1 diabetes, 20.1% had plasma triglycerides above 1.7 mmol/l, 9.6% had LDL-cholesterol above 3.4 mmol/l and 25.9% had non-HDL cholesterol above 3.4 mmol/l (Marcovecchio et al., 2009). It has been shown that abnormal lipid levels, in type 1 diabetes, predict worse cardiovascular outcomes (Soedamah-Muthu et al., 2004). HbA1c has been shown to be independently correlated with LDL-cholesterol, non-HDL cholesterol and triglyceride levels, indicating that these disorders were mostly observed in patients with poor glycemic control (Marcovecchio et al., 2009). In a British follow-up study of 229 children with type 1 diabetes, LDL cholesterol and non-HDL cholesterol values increased with duration of diabetes (Edge et al., 2008). In that study, total cholesterol, triglycerides and non-HDL cholesterol were positively correlated with HbA1c and around 10% of the patients had lipid values outside recommendations (Edge et al., 2008). In a large study performed in 29 979 patients with type 1 diabetes, multivariate analyses showed a significant positive association between HbA1c and total cholesterol (p<0.0001), LDL cholesterol (p<0.0001) and a significant negative association between HbA1c and HDL cholesterol (p<0.0001) (Schwab et al., 2009). In the Diabetes Control and Complications Trial (DCCT), HbA1c correlated positively with total cholesterol, LDLcholesterol and triglycerides at baseline (The DCCT Research Group, 1992). Data from the Coronary Artery Calcification in type 1 diabetes (CACTI) study, which examined 652 patients with type 1 diabetes, have shown, in patients not using hypolipidemic agents, that a higher HbA1c was associated with significantly higher levels of total cholesterol, triglycerides, LDL cholesterol and non-HDL cholesterol (Maahs et al., 2010). In that study, 1% change in HbA1c was associated with an increase of 0.101 mmol/l (4 mg/dl) for total cholesterol, of 0.052 mmol/l (4.5 mg/dl) for triglycerides, of 0.103 mmol/l (4 mg/dl) for LDL cholesterol and of 0.129 mmol/l (5 mg/dl) for non-HDL cholesterol (Maahs et al.,

Lipid Disorders in Type 1 Diabetes 51

subfraction 3 (Winocour et al., 1986). It has also been reported that elevation of HDL in type 1 diabetic patients with good glycemic control was caused by an increase of HDL particles containing only apoA-I (LpA-I) (Kahri et al., 1993). This increase in plasma HDL-cholesterol could be the consequence of the elevated Lipoprotein Lipase/Hepatic Lipase ratio that is observed in patients with well controlled type 1 diabetes (increased Lipoprotein Lipase activity and normal Hepatic Lipase activity) (Kahri et al., 1993). The increased Lipoprotein Lipase activity observed in these patients is likely to be due to peripheral hyperinsulinemia as a consequence of the subcutaneous route of insulin administration (Kahri et al., 1993).

Intensive subcutaneous insulin therapy results in normalization of plasma glucose, but at the expense of peripheral hyperinsulinemia, which is likely to modify lipoprotein metabolism (as discussed above). Implantable insulin pumps with intraperitoneal insulin administration mimic the physiologic route of insulin delivery and are likely to restore the normal portal-peripheral insulin gradient. For this reason, several studies have been performed to analyze the modification of lipoprotein metabolism after replacement of subcutaneous insulin therapy by intraperitoneal insulin therapy. Plasma triglycerides have been found increased in one study (Selam et al., 1989) and unchanged in three other studies (Bagdade & Dunn, 1996; Ruotolo et al., 1994; Duvillard et al., 2005). Total cholesterol and apoB were found unchanged (Bagdade & Dunn, 1996; Ruotolo et al., 1994; Duvillard et al., 2005). HDL-cholesterol has been found decreased (Selam et al., 1989) or not modified (Bagdade & Dunn, 1996; Ruotolo et al., 1994; Duvillard et al., 2007). The discrepancies of these studies that may be due to confounding factors such as degree of glycemic control and peripheral insulin levels during subcutaneous insulin therapy. Further studies are needed to clearly evaluate the effect of intraperitoneal insulin administration on lipoprotein

In type 1 diabetic patients with nephropathy and overt albuminuria, elevated plasma levels of total cholesterol, triglycerides and LDL-cholesterol are observed whereas HDLcholesterol is decreased due to a fall in HDL2 (Dullaart ,1995; Taskinen, 1992; Jensen et al., 1987). In the EURODIAB IDDM Complications study, macroalbuminuria was associated with significantly increased plasma triglycerides, cholesterol, LDL-cholesterol and LDL/HDL ratio in both sexes and decreased HDL-cholesterol in women (Mattock et al.,

Some quantitative lipid modifications are also observed in type 1 diabetic patients with microalbuminuria. Microalbuminuric patients compared with normoalbuminuric patients show increased plasma apoB (Jones et al., 1989, Dullaart et al, 1989a; Jay et al., 1991), LDL cholesterol (Jones et al., 1989, Dullaart et al, 1989a) and apoB/apoA1 ratio (Dullaart et al, 1989a; Jay et al., 1991). A positive correlation has been found between urinary albumin excretion rate and plasma apoB and apoB/apoA1 ratio (Dullaart et al, 1989a). In the EURODIAB IDDM Complications study, microalbuminuria was associated with increased plasma triglycerides (Mattock et al., 2001). In a prospective study performed in 895 young subjects with type 1 diabetes, total cholesterol and non-HDL cholesterol were independently related to longitudinal changes in albumin-to-creatinine ratio (Marcovecchio et al., 2009). The mechanisms responsible for these lipoprotein abnormalities in type 1 diabetic patients

**4.2.3 Subcutaneous insulin therapy versus intraperitoneal insulin therapy** 

metabolism.

2001).

**4.2.4 Type 1 diabetes with nephropathy** 

with microalbuminuria remain unclear.

2010). In a recent study, performed in 512 young patients with type 1 diabetes and in 188 healthy age-matched controls, patients with suboptimal control (HbA1c 7.5%) had much more lipid quantitative disorders than patients with optimal control (HbA1c<7.5%) (Guy et al., 2009). All these data suggest that quantitative lipid abnormalities are more frequent, when type 1 diabetes is not well controlled.

In addition, some patients with type 1 diabetes may have insulin resistance, in situation of abdominal obesity and/or family history of type 2 diabetes. Such patients have been shown to have greater dyslipidemia (Purnell et al., 2003). In a recent study performed in 60 young type 1 diabetic patients and 40 adults with type 1 diabetes, it has been shown, using hyperinsulinemic clamp studies, that lower glucose infusion (more insulin resistance) was associated with lower levels of HDL cholesterol in youths with type 1 diabetes and with higher levels of triglycerides and higher triglyceride/HDL ratio in both youths and adults (Maahs et al., 2011). These data indicate that insulin resistance may be an additional factor that could induce quantitative lipid abnormities in some type 1 diabetic patients with a background of insulin resistance (abdominal obesity, family history of type 2 diabetes). In this chapter we will consider only the typical situation of type 1 diabetes without insulin resistance.

#### **4.2.1 Treated type 1 diabetes with poor or suboptimal glycemic control**

In case of poor or suboptimal control, patients with type 1 diabetes may show increased plasma triglyceride levels (Dullaart,1995). This hypertriglyceridemia is due to increased production of VLDL, promoted by elevated circulating free fatty acids secondary to the relative insulin deficiency (Nikkilä & Kekki, 1973).

Type 1 diabetic patients with poor or suboptimal glycemic control show increased LDLcholesterol levels as compared to non–diabetic individuals and type 1 diabetic patients with optimal glycemic control (Dullaart, 1995; Guy et al., 2009). Indeed, in this condition, VLDL production is increased (see above), when catabolism of triglyceride-rich lipoproteins is not importantly decreased, which leads to increase LDL production (Dullaart, 1995).

#### **4.2.2 Treated type 1 diabetes with optimal glycemic control**

In well controlled type 1 diabetes, the lipid profile is totally different than in poorly controlled type 1 diabetes (Dullaart, 1995; Nikkilä & Kekki, 1973).

Plasma triglycerides are normal or slightly decreased (Dullaart, 1995; Nikkilä & Kekki, 1973). This slight decrease in plasma triglycerides may be observed with intense insulin therapy because of increased down control of VLDL production by augmented plasma insulin levels as a consequence of the subcutaneous route of insulin delivery (Dashti & Wolfbauer, 1987; Taskinen, 1992). Furthermore, in patients with well controlled type 1 diabetes, peripheral hyperinsulinemia has been shown to be associated with increased lipoprotein lipase activity that could be an additional factor responsible for decreased plasma triglycerides (Nikkilä et al., 1977).

Plasma LDL-cholesterol level is normal or slightly decreased (Winocour et al., 1986). This slight decrease in plasma LDL-cholesterol may be observed with intense insulin therapy as a consequence of decreased VLDL production by peripheral hyperinsulinemia (see above).

Plasma HDL-cholesterol level is normal or slightly increased in well controlled type 1 diabetic patients (Dullaart, 1995). Some studies have shown an increase in HDL subfraction 2 (Eckel et al., 1981; Kahri et al., 1993), when others have found an increase in HDL

2010). In a recent study, performed in 512 young patients with type 1 diabetes and in 188 healthy age-matched controls, patients with suboptimal control (HbA1c 7.5%) had much more lipid quantitative disorders than patients with optimal control (HbA1c<7.5%) (Guy et al., 2009). All these data suggest that quantitative lipid abnormalities are more frequent,

In addition, some patients with type 1 diabetes may have insulin resistance, in situation of abdominal obesity and/or family history of type 2 diabetes. Such patients have been shown to have greater dyslipidemia (Purnell et al., 2003). In a recent study performed in 60 young type 1 diabetic patients and 40 adults with type 1 diabetes, it has been shown, using hyperinsulinemic clamp studies, that lower glucose infusion (more insulin resistance) was associated with lower levels of HDL cholesterol in youths with type 1 diabetes and with higher levels of triglycerides and higher triglyceride/HDL ratio in both youths and adults (Maahs et al., 2011). These data indicate that insulin resistance may be an additional factor that could induce quantitative lipid abnormities in some type 1 diabetic patients with a background of insulin resistance (abdominal obesity, family history of type 2 diabetes). In this chapter we will consider only the typical situation of type 1 diabetes without insulin

In case of poor or suboptimal control, patients with type 1 diabetes may show increased plasma triglyceride levels (Dullaart,1995). This hypertriglyceridemia is due to increased production of VLDL, promoted by elevated circulating free fatty acids secondary to the

Type 1 diabetic patients with poor or suboptimal glycemic control show increased LDLcholesterol levels as compared to non–diabetic individuals and type 1 diabetic patients with optimal glycemic control (Dullaart, 1995; Guy et al., 2009). Indeed, in this condition, VLDL production is increased (see above), when catabolism of triglyceride-rich lipoproteins is not

In well controlled type 1 diabetes, the lipid profile is totally different than in poorly

Plasma triglycerides are normal or slightly decreased (Dullaart, 1995; Nikkilä & Kekki, 1973). This slight decrease in plasma triglycerides may be observed with intense insulin therapy because of increased down control of VLDL production by augmented plasma insulin levels as a consequence of the subcutaneous route of insulin delivery (Dashti & Wolfbauer, 1987; Taskinen, 1992). Furthermore, in patients with well controlled type 1 diabetes, peripheral hyperinsulinemia has been shown to be associated with increased lipoprotein lipase activity that could be an additional factor responsible for decreased

Plasma LDL-cholesterol level is normal or slightly decreased (Winocour et al., 1986). This slight decrease in plasma LDL-cholesterol may be observed with intense insulin therapy as a consequence of decreased VLDL production by peripheral hyperinsulinemia (see above). Plasma HDL-cholesterol level is normal or slightly increased in well controlled type 1 diabetic patients (Dullaart, 1995). Some studies have shown an increase in HDL subfraction 2 (Eckel et al., 1981; Kahri et al., 1993), when others have found an increase in HDL

**4.2.1 Treated type 1 diabetes with poor or suboptimal glycemic control** 

importantly decreased, which leads to increase LDL production (Dullaart, 1995).

**4.2.2 Treated type 1 diabetes with optimal glycemic control** 

controlled type 1 diabetes (Dullaart, 1995; Nikkilä & Kekki, 1973).

when type 1 diabetes is not well controlled.

relative insulin deficiency (Nikkilä & Kekki, 1973).

plasma triglycerides (Nikkilä et al., 1977).

resistance.

subfraction 3 (Winocour et al., 1986). It has also been reported that elevation of HDL in type 1 diabetic patients with good glycemic control was caused by an increase of HDL particles containing only apoA-I (LpA-I) (Kahri et al., 1993). This increase in plasma HDL-cholesterol could be the consequence of the elevated Lipoprotein Lipase/Hepatic Lipase ratio that is observed in patients with well controlled type 1 diabetes (increased Lipoprotein Lipase activity and normal Hepatic Lipase activity) (Kahri et al., 1993). The increased Lipoprotein Lipase activity observed in these patients is likely to be due to peripheral hyperinsulinemia as a consequence of the subcutaneous route of insulin administration (Kahri et al., 1993).

#### **4.2.3 Subcutaneous insulin therapy versus intraperitoneal insulin therapy**

Intensive subcutaneous insulin therapy results in normalization of plasma glucose, but at the expense of peripheral hyperinsulinemia, which is likely to modify lipoprotein metabolism (as discussed above). Implantable insulin pumps with intraperitoneal insulin administration mimic the physiologic route of insulin delivery and are likely to restore the normal portal-peripheral insulin gradient. For this reason, several studies have been performed to analyze the modification of lipoprotein metabolism after replacement of subcutaneous insulin therapy by intraperitoneal insulin therapy. Plasma triglycerides have been found increased in one study (Selam et al., 1989) and unchanged in three other studies (Bagdade & Dunn, 1996; Ruotolo et al., 1994; Duvillard et al., 2005). Total cholesterol and apoB were found unchanged (Bagdade & Dunn, 1996; Ruotolo et al., 1994; Duvillard et al., 2005). HDL-cholesterol has been found decreased (Selam et al., 1989) or not modified (Bagdade & Dunn, 1996; Ruotolo et al., 1994; Duvillard et al., 2007). The discrepancies of these studies that may be due to confounding factors such as degree of glycemic control and peripheral insulin levels during subcutaneous insulin therapy. Further studies are needed to clearly evaluate the effect of intraperitoneal insulin administration on lipoprotein metabolism.

#### **4.2.4 Type 1 diabetes with nephropathy**

In type 1 diabetic patients with nephropathy and overt albuminuria, elevated plasma levels of total cholesterol, triglycerides and LDL-cholesterol are observed whereas HDLcholesterol is decreased due to a fall in HDL2 (Dullaart ,1995; Taskinen, 1992; Jensen et al., 1987). In the EURODIAB IDDM Complications study, macroalbuminuria was associated with significantly increased plasma triglycerides, cholesterol, LDL-cholesterol and LDL/HDL ratio in both sexes and decreased HDL-cholesterol in women (Mattock et al., 2001).

Some quantitative lipid modifications are also observed in type 1 diabetic patients with microalbuminuria. Microalbuminuric patients compared with normoalbuminuric patients show increased plasma apoB (Jones et al., 1989, Dullaart et al, 1989a; Jay et al., 1991), LDL cholesterol (Jones et al., 1989, Dullaart et al, 1989a) and apoB/apoA1 ratio (Dullaart et al, 1989a; Jay et al., 1991). A positive correlation has been found between urinary albumin excretion rate and plasma apoB and apoB/apoA1 ratio (Dullaart et al, 1989a). In the EURODIAB IDDM Complications study, microalbuminuria was associated with increased plasma triglycerides (Mattock et al., 2001). In a prospective study performed in 895 young subjects with type 1 diabetes, total cholesterol and non-HDL cholesterol were independently related to longitudinal changes in albumin-to-creatinine ratio (Marcovecchio et al., 2009). The mechanisms responsible for these lipoprotein abnormalities in type 1 diabetic patients with microalbuminuria remain unclear.

Lipid Disorders in Type 1 Diabetes 53

occurs within LDL in parallel with plasma hyperglycemia. It has been shown that apoB glycation reduces significantly LDL binding to the B/E receptor even when apoB glycation is moderate (Witztum et al., 1982; Steinbrecher et al, 1984). Furthermore, glycated LDLs are preferentialy taken up by macrophages through the scavenger receptor, leading to the formation of foam cells in the arterial wall. In patients with type 1 diabetes, advanced glycation end products-modified LDL have been shown to be positively associated with

Moreover, patients with type 1 diabetes may show an increased oxidation of LDL which is promoted by glycemic excursions (de Castro et al., 2005). Increased urinary excretion of malondialdehyde, reflecting enhanced lipid peroxidation, has been reported in patients with type 1 diabetes (Hoeldtke et al., 2009). Oxidative modification of LDL results in rapid uptake by macrophages, leading to foam cell formation. Oxidized LDLs produce chemotactic effects on monocytes by increasing the synthesis of adhesion molecules, such as ICAM-1 (intercellular adhesion Molecule 1) by endothelial cells. Oxidized LDLs stimulate the formation by macrophages of cytokines, such as TNF or IL1, which amplify the inflammatory atherosclerotic process. It has recently been shown that oxidized LDL particles were significantly associated with progression and increased levels of IMT in type

HDL particles from patients with type 1 diabetes are often enriched in triglycerides (Dullaart, 1995; Bagdade et al., 1991a). This modification has been attributed to increased cholesteryl ester transfer between lipoproteins (Bagdade et al., 1991a). In HDL particles from patients with type 1 diabetes, sphingomyelin/lecithin ratio within the peripheral layer is augmented, which may increase HDL rigidity (Bagdade & Subbaiah, 1989). These alterations are not totally reversed after achievement of optimal glycemic control (Bagdade et al., 1991b). ApoA-I within HDL is glycated in patients with type 1 diabetes, which may impair the HDL-mediated reverse cholesterol pathway. Indeed, it has been shown that HDL particles containing glycated apoA-I were less effective to promote cholesterol efflux from

In addition to their role in the reverse cholesterol pathway, HDLs have anti-oxidative, antiinflammatory, anti-thrombotic and vasorelaxant properties, potentially anti-atherogenic (Link et al., 2007). Some of these properties have been shown to be reduced in patients with type 1 diabetes. Indeed, a significant reduction of the activity of paraoxonase, an antioxidative enzyme associated with HDLs, is observed in patients with type 1 diabetes (Boemi et al., 2001; Ferretti et al., 2004). As a consequence, HDLs from patients with type 1 diabetes protect less efficiently erythrocyte membranes and LDL particles against oxidative damage than HDLs from normal individuals (Boemi et al., 2001; Ferretti et al., 2004). Furthermore, using rabbit aorta rings, it has been shown that HDL from patients with type 1 diabetes are no more able to prevent the endothelium dependent vasoconstriction induced by oxidized

LDL, whereas HDL from normal individuals can prevent it (Perségol et al., 2007).

In some studies, an increased cholesteryl ester transfer between lipoproteins (Bagdade et al., 1991a; Bagdade et al., 1994) or an augmented activity of CETP (Colhoun et al., 2001) have been found in normolipidemic patients with type 1 diabetes. In some other studies, increased CETP activity has been reported only in type 1 diabetic patients that smoke or

increased intima media thickness (IMT) (Lopes-Virella et al., 2011).

1 diabetes (de Castro et al., 2005).

the cells (Fievet et al., 1992).

**5.4 Lipid transfer proteins** 

**5.3 HDLs** 

Moreover, serum lipids have been shown to be associated with the progression of nephropathy in type 1 diabetes. In a prospective study performed in 152 patients with type 1 diabetes followed for 8-9 years, LDL-cholesterol was an independent factor associated with progression of nephropathy (Thomas et al., 2006).
