**4. Diet, lifestyle and cardiovascular disease**

Eight studies reported an association between macronutrients and CVD in type 1 diabetic patients. Of these eight, six are cross-sectional studies (Toeller M et al. 1999 1,2,3; Helgeson 2006; Øverby NC et al. 2006; Snell-Bergeon JK et al. 2009). Only Strychar et al. (Strychar I et al. 2009) and Georgopoulos et al. (Georgopoulos A et al. 2000) performed a randomized controlled trial. One study reported an association between lifestyle risk factors (including alcohol) and atherosclerosis, which is often the underlying cause of CVD (Bishop et al. 2009). Eight studies reported an association between physical activity and CVD risk factors (Kriska AM et al. 1991; Lehmann R et al. 1997; Fuchsjäger-Mayrl G et al. 2002; Herbst A et al. 2007; Valerio G et al. 2007; Bishop et al. 2009; Trigona B et al. 2010; Seeger JPH et al. 2011), and two studies reported an association with dietary patterns (Gunther ALB et al. 2008; Liese AD et al. 2011). Furthermore no studies were found examining the effect of glycaemic index/glycaemic load on CVD in type 1 diabetic patients.

#### **4.1 Macronutrients**

Data on the relationship between macronutrients and incident CVD is lacking in patients with type 1 diabetes. Limited information on macronutrients is available from crosssectional studies. Main focus was on fat, in particularly saturated fat, and fiber and CVD risk factors were used as a proxy for CVD events.

#### **4.2 Cross-sectional studies on fat and fiber in relation to CVD**

In more detail, one cross-sectional study (Øverby NC et al. 2006) found a higher than recommended percentage of energy intake from fat and saturated fat among type 1 diabetic patients compared with healthy same-age control subjects and a lower than recommended intake of fiber. They conclude that this higher intake of energy from saturated fat and this lower intake of energy from dietary fiber, vegetables and fruits could increase the risk of atherosclerosis, which is often the underlying cause of CVD. Another study (Helgeson 2006) reported a higher than recommended percentage of energy intake from fat and saturated fat among type 1 diabetic patients, but they did not study associations with CVD or CVD risk factors. Another cross-sectional study (Toeller M et al. 1999 2) found similar associations between dietary fiber and CVD. Higher fiber intake had a protective significant effect against CVD in type 1 diabetic women but not in men. In type 1 diabetic men it leads to positive changes of the serum cholesterol pattern (higher HDL, lower LDL, lower ratio total cholesterol:HDL cholesterol). In another study (Toeller M et al. 1999 3) a significant increase in energy adjusted total and LDL-cholesterol levels was associated with higher intakes of total fat, saturated fat and cholesterol. This was associated with a higher prevalence of CVD, although after adjusting for dietary fiber intake, these associations were attenuated. A third study by Toeller et al. (Toeller M et al. 1999 1) found a higher intake of cholesterol, total fat and saturated fat in Eastern Europe compared to Southern or North-Western Europe. They also found more frequent acute and chronic complications (including CVD) in Eastern European people. However, since it was a cross-sectional study they could not conclude if this was due to the high intake of cholesterol, total fat and/or saturated fat. In the CACTI study (Snell-Bergeon JK et al. 2009) found an increased risk of CVD in type 1 diabetic patients eating high amounts of fat and saturated fat. Carbohydrates were negatively correlated with CHD risk factors (higher total cholesterol, LDL cholesterol, obesity, poorer glycaemic control). Furthermore higher intakes of fat and protein were associated with greater odds of coronary artery calcium (CAC), which is a strong predictor for coronary events approximating CVD risk. The opposite was true for carbohydrate intake, higher intake was associated with a reduced odds of CAC.

In conclusion a higher intake of total fat as well as saturated fat is positively correlated with CVD or CVD risk factors (atherosclerosis and CAC in these studies) and a higher intake of carbohydrate is negatively correlated with CVD or CVD risk factors. Furthermore dietary fiber is independently related to a lower risk for CVD in type 1 diabetic women. Since all these studies were cross-sectional, they could only look at the intake of certain nutrients and the prevalence of CVD or CVD risk factors at a certain time point. They could not conclude if these are related to each other and if the nutrients are responsible for the lower or higher prevalence of CVD.

#### **4.3 Randomized controlled trials**

28 Type 1 Diabetes – Complications, Pathogenesis, and Alternative Treatments

Eight studies reported an association between macronutrients and CVD in type 1 diabetic patients. Of these eight, six are cross-sectional studies (Toeller M et al. 1999 1,2,3; Helgeson 2006; Øverby NC et al. 2006; Snell-Bergeon JK et al. 2009). Only Strychar et al. (Strychar I et al. 2009) and Georgopoulos et al. (Georgopoulos A et al. 2000) performed a randomized controlled trial. One study reported an association between lifestyle risk factors (including alcohol) and atherosclerosis, which is often the underlying cause of CVD (Bishop et al. 2009). Eight studies reported an association between physical activity and CVD risk factors (Kriska AM et al. 1991; Lehmann R et al. 1997; Fuchsjäger-Mayrl G et al. 2002; Herbst A et al. 2007; Valerio G et al. 2007; Bishop et al. 2009; Trigona B et al. 2010; Seeger JPH et al. 2011), and two studies reported an association with dietary patterns (Gunther ALB et al. 2008; Liese AD et al. 2011). Furthermore no studies were found examining the effect of glycaemic

Data on the relationship between macronutrients and incident CVD is lacking in patients with type 1 diabetes. Limited information on macronutrients is available from crosssectional studies. Main focus was on fat, in particularly saturated fat, and fiber and CVD

In more detail, one cross-sectional study (Øverby NC et al. 2006) found a higher than recommended percentage of energy intake from fat and saturated fat among type 1 diabetic patients compared with healthy same-age control subjects and a lower than recommended intake of fiber. They conclude that this higher intake of energy from saturated fat and this lower intake of energy from dietary fiber, vegetables and fruits could increase the risk of atherosclerosis, which is often the underlying cause of CVD. Another study (Helgeson 2006) reported a higher than recommended percentage of energy intake from fat and saturated fat among type 1 diabetic patients, but they did not study associations with CVD or CVD risk factors. Another cross-sectional study (Toeller M et al. 1999 2) found similar associations between dietary fiber and CVD. Higher fiber intake had a protective significant effect against CVD in type 1 diabetic women but not in men. In type 1 diabetic men it leads to positive changes of the serum cholesterol pattern (higher HDL, lower LDL, lower ratio total cholesterol:HDL cholesterol). In another study (Toeller M et al. 1999 3) a significant increase in energy adjusted total and LDL-cholesterol levels was associated with higher intakes of total fat, saturated fat and cholesterol. This was associated with a higher prevalence of CVD, although after adjusting for dietary fiber intake, these associations were attenuated. A third study by Toeller et al. (Toeller M et al. 1999 1) found a higher intake of cholesterol, total fat and saturated fat in Eastern Europe compared to Southern or North-Western Europe. They also found more frequent acute and chronic complications (including CVD) in Eastern European people. However, since it was a cross-sectional study they could not conclude if this was due to the high intake of cholesterol, total fat and/or saturated fat. In the CACTI study (Snell-Bergeon JK et al. 2009) found an increased risk of CVD in type 1 diabetic patients eating high amounts of fat and saturated fat. Carbohydrates were negatively correlated with CHD risk factors (higher total cholesterol, LDL cholesterol, obesity, poorer glycaemic control). Furthermore higher intakes of fat and protein were associated with greater odds of coronary artery calcium (CAC), which is a strong predictor for coronary

**4. Diet, lifestyle and cardiovascular disease** 

index/glycaemic load on CVD in type 1 diabetic patients.

**4.2 Cross-sectional studies on fat and fiber in relation to CVD** 

risk factors were used as a proxy for CVD events.

**4.1 Macronutrients** 

Two randomized controlled trials reported an association between macronutrients and CVD (**Table 2**), but demonstrated conflicting conclusions. In one trial (Strychar I et al. 2009), the authors concluded that a diet lower in carbohydrate and higher in MUFA might be preferable to a diet higher in carbohydrate and lower in MUFA for type 1 diabetic patients. This was solely based on the positive effect on triglyceride (TG) levels and plasminogen activator inhibitor 1 levels (PAI-1) in the first diet. A significant decrease in PAI-1 was found after 6 months in the lower carbohydrate and higher MUFA diet. In the other diet there was a significant increase after 6 months of follow up. PAI-1 is an inhibitor of fibrinolysis, a process that degrades blood clots. A lower level of PAI-1 means less inhibition and more degradation of blood clots, which means a lower chance of developing atherosclerosis. Also a decrease in TG levels was found after 6 months following the low carbohydrate/high MUFA diet, although this decrease was not significant. In the other diet group there was an increase in TG levels, also this increase was not significant. Furthermore they conclude that the lower carbohydrate/higher MUFA diet was only a proper choice for nonobese individuals with weight control since this diet had induced a weight gain of 2% (1.6 kg) after 6 months. The other trial (Georgopoulos A et al. 2000) found exactly the opposite using a crossover design. They found that a diet high in carbohydrates might be preferable to a diet high in MUFA. Mainly because of the higher atherosclerotic risk due to more and bigger very low-density lipoprotein (VLDL) particles in the last diet. Furthermore the TG levels did not significantly differ between the two diets in this study.

In conclusion, these trials show that the effect of carbohydrate or MUFA on cardiovascular disease risk factors in type 1 diabetic patients is still not elucidated. Although they recommend exactly the opposite (higher intake of MUFA preferable vs. higher intake of carbohydrate preferable) they both found that a high MUFA or a high carbohydrate diet did not affect the TG levels. Their conclusions are based on PAI-1 and VLDL levels, which are not such a good predictors for atherosclerosis (and by extension CVD) as TG levels are. Furthermore none of these randomized controlled trials examined the potential positive effect of dietary fiber on CVD or the potential negative effect of saturated fat found in crosssectional studies.

#### **4.4 Alcohol**

One cross-sectional study (Bishop et al. 2009) reported findings on the association between alcohol and cardiovascular disease. No significant association was found between alcohol consumption (±13.8 drinks/month) and CAC, a marker of coronary artery atherosclerosis (adjusted OR=0.9, 95% CI: 0.8-1.1, p=0.15). The positive effect of moderate alcohol


Diet, Lifestyle and Chronic Complications in Type 1 Diabetic Patients 31

consumption on CVD as in the general population is not confirmed for type 1 diabetic patients in this study. However, this could also be due to the kind of study (cross-sectional) and the fact that markers for CVD were used instead of CVD as endpoint. There are no prospective studies which have addressed the relation between alcohol and CVD in type 1

Of the eight studies that reported an association between physical activity and CVD there were five cross-sectional studies (Kriska AM et al. 1991; Herbst A et al. 2007; Valerio G et al. 2007; Bishop et al. 2009; Trigona B et al. 2010) and three trials (Lehmann R et al. 1997; Fuchsjäger-Mayrl G et al. 2002; Seeger JPH et al. 2011). No prospective cohort studies were

One study (Kriska AM et al. 1991) examined the relationship between physical activity and the occurrence of CVD in type 1 diabetic patients. They found the lowest occurrence of CVD in people being 4-7 hours a week physical active (sports and leisure physical activity). The other four cross-sectional studies examined an association between physical activity and CVD risk factors. They all found a positive association. Another two studies (Herbst A et al. 2007; Valerio G et al. 2007) found that increased frequency of regular physical activity was associated with lower TG levels. One of these (Herbst A et al. 2007) found besides the positive association with TG levels also a positive significant association between regular physical activity and HDL cholesterol levels. Another study (Trigona B et al. 2010) found that 60 min/day of moderate-to-vigorous physical activity was associated with an enhanced endothelial function in type 1 diabetic patients. Impaired endothelial function is considered as an early sign of atherosclerosis, which is often the underlying cause of CVD. And finally, (Bishop et al. 2009) a significant inverse association between physical activity and CAC, a

In conclusion all these studies found a beneficial effect of physical activity on cardiovascular risk factors. However, since all these studies were cross-sectional, they could only look at physical activity and the prevalence of CVD or CVD risk factors at a certain time point. They could not conclude if these are related to each other and if physical activity was responsible

The three trials reporting an association between physical activity and cardiovascular disease risk factors (**Table 3**) were consistent in their conclusions. They all emphasize an important role for physical activity in type 1 diabetic patients. Two studies (Fuchsjäger-Mayrl G et al. 2002; Seeger JPH et al. 2011) examined the association between physical activity and brachial artery flow-mediated dilation (FMD). Endothelial dysfunction is reflected by an impaired FMD response and is an early sign of atherosclerosis. An increase in FMD was found in type 1 diabetic patients following an exercise training program (endurance sports; on average 2 times a week 60 minutes, **Table 3**). In both trials this increase was significant (p=0.038 and p=0.040 respectively). Two studies (Lehmann R et al. 1997; Fuchsjäger-Mayrl G et al. 2002) examined the impact of physical activity on lipid related cardiovascular risk factors (LDL cholesterol, HDL cholesterol and TG). They both found a decrease in LDL cholesterol levels in the training group, but only in one of these (Lehmann R et al. 1997) this decrease was significant (p=0.02). An additional effect was reported in one of these studies (Lehmann R et al. 1997) with a significant increase in HDL cholesterol levels (p=0.03) in the training group. No effect of physical activity on TG levels

found. The studies will be discussed in the following paragraphs by study design.

marker of coronary artery atherosclerosis, was demonstrated.

for the lower prevalence of CVD.

diabetic patients.

**4.5 Physical activity** 

consumption on CVD as in the general population is not confirmed for type 1 diabetic patients in this study. However, this could also be due to the kind of study (cross-sectional) and the fact that markers for CVD were used instead of CVD as endpoint. There are no prospective studies which have addressed the relation between alcohol and CVD in type 1 diabetic patients.

#### **4.5 Physical activity**

30 Type 1 Diabetes – Complications, Pathogenesis, and Alternative Treatments

\* 54-57% CH, 27-30% total fat (10% MUFA)

\*\* 43-46% CH, 37-40% total fat (20% MUFA)

\*\*\* 40% total fatty acids (25% MUFA, 6% PUFA, 9% saturated), 45% CH and 15% protein

\*\*\*\* 24% total fatty acids (9% MUFA, 6% PUFA, 9% saturated), 61% CH and 15% protein

PAI-1: plasminogen activator inhibitor 1; TG: triglycerides; VLDL: very low-density lipoprotein; SD: standard deviation; SEM: standard error of

mean; CH: carbohydrate; MUFA: monounsaturated fatty acid; HCLF: high carbohydrate, low fat; LCHF: low carbohydrate, high fat

Table 2. Randomized controlled trials; diet and cardiovascular disease

Of the eight studies that reported an association between physical activity and CVD there were five cross-sectional studies (Kriska AM et al. 1991; Herbst A et al. 2007; Valerio G et al. 2007; Bishop et al. 2009; Trigona B et al. 2010) and three trials (Lehmann R et al. 1997; Fuchsjäger-Mayrl G et al. 2002; Seeger JPH et al. 2011). No prospective cohort studies were found. The studies will be discussed in the following paragraphs by study design.

One study (Kriska AM et al. 1991) examined the relationship between physical activity and the occurrence of CVD in type 1 diabetic patients. They found the lowest occurrence of CVD in people being 4-7 hours a week physical active (sports and leisure physical activity). The other four cross-sectional studies examined an association between physical activity and CVD risk factors. They all found a positive association. Another two studies (Herbst A et al. 2007; Valerio G et al. 2007) found that increased frequency of regular physical activity was associated with lower TG levels. One of these (Herbst A et al. 2007) found besides the positive association with TG levels also a positive significant association between regular physical activity and HDL cholesterol levels. Another study (Trigona B et al. 2010) found that 60 min/day of moderate-to-vigorous physical activity was associated with an enhanced endothelial function in type 1 diabetic patients. Impaired endothelial function is considered as an early sign of atherosclerosis, which is often the underlying cause of CVD. And finally, (Bishop et al. 2009) a significant inverse association between physical activity and CAC, a marker of coronary artery atherosclerosis, was demonstrated.

In conclusion all these studies found a beneficial effect of physical activity on cardiovascular risk factors. However, since all these studies were cross-sectional, they could only look at physical activity and the prevalence of CVD or CVD risk factors at a certain time point. They could not conclude if these are related to each other and if physical activity was responsible for the lower prevalence of CVD.

The three trials reporting an association between physical activity and cardiovascular disease risk factors (**Table 3**) were consistent in their conclusions. They all emphasize an important role for physical activity in type 1 diabetic patients. Two studies (Fuchsjäger-Mayrl G et al. 2002; Seeger JPH et al. 2011) examined the association between physical activity and brachial artery flow-mediated dilation (FMD). Endothelial dysfunction is reflected by an impaired FMD response and is an early sign of atherosclerosis. An increase in FMD was found in type 1 diabetic patients following an exercise training program (endurance sports; on average 2 times a week 60 minutes, **Table 3**). In both trials this increase was significant (p=0.038 and p=0.040 respectively). Two studies (Lehmann R et al. 1997; Fuchsjäger-Mayrl G et al. 2002) examined the impact of physical activity on lipid related cardiovascular risk factors (LDL cholesterol, HDL cholesterol and TG). They both found a decrease in LDL cholesterol levels in the training group, but only in one of these (Lehmann R et al. 1997) this decrease was significant (p=0.02). An additional effect was reported in one of these studies (Lehmann R et al. 1997) with a significant increase in HDL cholesterol levels (p=0.03) in the training group. No effect of physical activity on TG levels


Diet, Lifestyle and Chronic Complications in Type 1 Diabetic Patients 33

was found in both studies. Furthermore all three studies (Lehmann R et al. 1997; Fuchsjäger-Mayrl G et al. 2002; Seeger JPH et al. 2011) assessed physical fitness by VO2max (peak oxygen uptake). They all found a positive significant association between physical activity and

In conclusion the three trials show that physical activity improves physical fitness as well as endothelial function in type 1 diabetic patients. A positive effect on lipid related

Two cross-sectional studies reported an association between dietary patterns, in this case the 'Dietary Approaches to Stop Hypertension' (DASH) diet, and CVD risk factors (Gunther ALB et al. 2008; Liese AD et al. 2011). No cross-sectional or prospective studies were found examining the effect of a Mediterranean diet or a Western diet on CVD in type 1 diabetic

One study (Gunther ALB et al. 2008) reported an association between adherence to the DASH diet and hypertension in type 1 diabetic patients. They found that a higher adherence to this diet amongst type 1 diabetic patients was inversely related to hypertension (OR=0.6, 95% CI: 0.4-0.9, p=0.007). They did not investigate a possible association between the DASH diet and CVD, but used hypertension as the main risk factor for CVD. Another study (Liese AD et al. 2011) reported a possible association between the DASH diet and other CVD risk factors (total cholesterol, LDL cholesterol, HDL cholesterol, TG, LDL particle density, apolipoprotein B, body mass index (BMI), waist circumference, and adipocytokines) than blood pressure. A significant and inverse association between the DASH diet and LDL/HDL ratio was found. An estimated 0.07 lower LDL/HDL ratio was found in the highest adherence group compared with the lowest adherence group. No significant association was found between LDL particle density, BMI, waist circumference,

In conclusion a positive effect of adherence to the DASH diet on hypertension and LDL/HDL ratio, which are important risk factors for CVD, was found. Unfortunately there

Overall, fiber and saturated fat intake play an important role in type 1 diabetic patients, with a beneficial and detrimental effect on the chronic complications respectively. Many researchers have shown the inappropriate intake of these nutrients in patients with type 1 diabetes. A protein restriction diet helped reduce micro/macro albuminuria in known type 1 diabetic patients with nephropathy, however, the compliance was low. Also moderate alcohol intake and physical activity may have beneficial effects in type 1 diabetic patients. Most of the findings are consistent with the guidelines for type 1 diabetic patients (**Table 4**). The main limitations are the lack of prospective studies on diet and lifestyle in type 1 diabetics, lack of randomized controlled trials and the limited number of studies on dietary cholesterol, protein, carbohydrates, fat, fiber and no cardiovascular morbidity data. The available studies, with their limitations, all indicate that diet and lifestyle play an important role in preventing chronic complications of type 1 diabetes. To put the findings in the literature in perspective, current nutritional recommendations are evaluated in the

were no studies found examining the effect of dietary patterns on CVD events.

**5. Current recommendations on diet and lifestyle in patients with type 1** 

VO2max. The relation between physical fitness and CVD was not examined.

**4.6 Dietary patterns** 

adipocytokines, or TG and the DASH diet.

**diabetes put in perspective** 

patients.

cardiovascular risk factors was only found in one study (Lehmann R et al. 1997).

was found in both studies. Furthermore all three studies (Lehmann R et al. 1997; Fuchsjäger-Mayrl G et al. 2002; Seeger JPH et al. 2011) assessed physical fitness by VO2max (peak oxygen uptake). They all found a positive significant association between physical activity and VO2max. The relation between physical fitness and CVD was not examined.

In conclusion the three trials show that physical activity improves physical fitness as well as endothelial function in type 1 diabetic patients. A positive effect on lipid related cardiovascular risk factors was only found in one study (Lehmann R et al. 1997).

#### **4.6 Dietary patterns**

32 Type 1 Diabetes – Complications, Pathogenesis, and Alternative Treatments

\* two times a week: first day: 30 min running exercise (intervals) and 30 min group-based activities such as ball games, short relay races, running

techniques and stretching; second day: individual exercise session at home involved 30 min of interval running and a 10-min warm-up and

\*\* first 2 weeks: two times a week 1 hour stationary cycling, during the remaining study period three times a week 1 hour stationary cycling

VO2max: peak oxygen uptake; FMD: flow mediated dilation; LDL: low-density lipoprotein; HDL: high-density lipoprotein; TrGR; training group;

Table 3. Randomized controlled trials; physical activity and cardiovascular disease risk factors

\*\*\* 135 min per week endurance sports (biking, long-distance running, or hiking)

C: control group

cooling down (including stretching)

Two cross-sectional studies reported an association between dietary patterns, in this case the 'Dietary Approaches to Stop Hypertension' (DASH) diet, and CVD risk factors (Gunther ALB et al. 2008; Liese AD et al. 2011). No cross-sectional or prospective studies were found examining the effect of a Mediterranean diet or a Western diet on CVD in type 1 diabetic patients.

One study (Gunther ALB et al. 2008) reported an association between adherence to the DASH diet and hypertension in type 1 diabetic patients. They found that a higher adherence to this diet amongst type 1 diabetic patients was inversely related to hypertension (OR=0.6, 95% CI: 0.4-0.9, p=0.007). They did not investigate a possible association between the DASH diet and CVD, but used hypertension as the main risk factor for CVD. Another study (Liese AD et al. 2011) reported a possible association between the DASH diet and other CVD risk factors (total cholesterol, LDL cholesterol, HDL cholesterol, TG, LDL particle density, apolipoprotein B, body mass index (BMI), waist circumference, and adipocytokines) than blood pressure. A significant and inverse association between the DASH diet and LDL/HDL ratio was found. An estimated 0.07 lower LDL/HDL ratio was found in the highest adherence group compared with the lowest adherence group. No significant association was found between LDL particle density, BMI, waist circumference, adipocytokines, or TG and the DASH diet.

In conclusion a positive effect of adherence to the DASH diet on hypertension and LDL/HDL ratio, which are important risk factors for CVD, was found. Unfortunately there were no studies found examining the effect of dietary patterns on CVD events.

### **5. Current recommendations on diet and lifestyle in patients with type 1 diabetes put in perspective**

Overall, fiber and saturated fat intake play an important role in type 1 diabetic patients, with a beneficial and detrimental effect on the chronic complications respectively. Many researchers have shown the inappropriate intake of these nutrients in patients with type 1 diabetes. A protein restriction diet helped reduce micro/macro albuminuria in known type 1 diabetic patients with nephropathy, however, the compliance was low. Also moderate alcohol intake and physical activity may have beneficial effects in type 1 diabetic patients. Most of the findings are consistent with the guidelines for type 1 diabetic patients (**Table 4**).

The main limitations are the lack of prospective studies on diet and lifestyle in type 1 diabetics, lack of randomized controlled trials and the limited number of studies on dietary cholesterol, protein, carbohydrates, fat, fiber and no cardiovascular morbidity data. The available studies, with their limitations, all indicate that diet and lifestyle play an important role in preventing chronic complications of type 1 diabetes. To put the findings in the literature in perspective, current nutritional recommendations are evaluated in the


\*\*\* Evidence grade C-E: evidence obtained from expert committee reports or opinions and/or clinical experiences of respected authorities

Diet, Lifestyle and Chronic Complications in Type 1 Diabetic Patients 35

following paragraphs at a macronutrient level. **Table 4** summarizes the nutritional recommendations as well as the lifestyle recommendations for type 1 and type 2 diabetic patients. These recommendations are for all diabetic patients in general, based in the

The 'Diabetes and Nutrition Study Group of the European Association for the Study of Diabetes' (DNSG EASD) guidelines for persons with type 1 and type 2 diabetes (**Table 4**) recommend that the most appropriate intake of carbohydrates consists of vegetables, legumes, fruits, wholegrain foods and naturally occurring foods rich in fiber. The fiber

Complications Study showed an inverse association between fiber and LDL cholesterol and a positive association between fiber and HDL cholesterol. In addition dietary fiber was

found with a fiber intake of approximately 8.1 g/1000 kcal, which is below the recommended intake. The average fiber intake in type 1 diabetic patients is 8.1 g/1000 kcal, but the recommended intake is 20 g/1000 kcal. Recommendation was only achieved in 0.4% of the type 1 diabetic population (Toeller M et al. 1996). Data from the EURODIAB Prospective Complication Study on fiber intake measured at baseline by 3-day food diaries and presented by each center is given in F**igure 1**. As seen in this figure, even the 10 g/1000 kcal recommended fiber intake by the 'American Diabetes Association' (ADA) was hardly achieved by type 1 diabetic patients. Only Finnish type 1 diabetic patients achieved the ADA fiber recommendation of 10 g/1000 kcal F**igure 1**). Keeping in mind that these samples are clinic based and not population based and that these figures may not exactly reflect the current nutritional intake, however it gives an indication of the status on fiber intake. Although positive effects were already found on CVD with a fiber intake of 8.1 g/1000 kcal, we assume that effects could be probably even higher when recommended levels of fiber intake are reached. Unfortunately, this positive effect of fiber on CVD and CVD risk factors was only found in cross-sectional studies. This makes it very difficult to distinguish cause and effect. Further research in prospective studies or randomized controlled trials is needed

DNSG EASD do not recommend a low carbohydrate diet for type 1 and type 2 diabetic patients (**Table 4**). A low carbohydrate diet does not produce beneficial health effect. It is more acceptable to avoid too much foods high in fast available carbohydrates, foods high in fat and cholesterol. An earlier quote (Helgeson 2006) expressed this precisely: 'families of adolescents with diabetes may be more concerned that the sugar in candy is going to translate into high blood glucose levels today than that the fat in potato chips will translate

The DNSG EASD guidelines for dietary fat for persons with type 1 and type 2 diabetes recommend a saturated and trans-unsaturated fatty acid consumption of <10% of the total energy intake (<8% if LDL cholesterol is elevated). Total fat intake should not exceed 35% of total energy and dietary cholesterol should be <300 mg/day (T**able 4**). Saturated fat is an important risk factor for diabetic nephropathy, diabetic retinopathy as well as CVD (Riley

≈ 20 g/1000 kcal/day. Cross-sectional data of the EURODIAB

2). This effect was already

majority of cases on evidence from type 2 diabetic patients.

inversely and significantly related to CVD (Toeller M et al. 1999

**5.1 Carbohydrates** 

intake should be ideally

to ascertain the role of fiber in CVD.

into cardiovascular disease in 10 years'.

**5.2 Fat** 

following paragraphs at a macronutrient level. **Table 4** summarizes the nutritional recommendations as well as the lifestyle recommendations for type 1 and type 2 diabetic patients. These recommendations are for all diabetic patients in general, based in the majority of cases on evidence from type 2 diabetic patients.

#### **5.1 Carbohydrates**

34 Type 1 Diabetes – Complications, Pathogenesis, and Alternative Treatments

1 obtained from DNSG EASD: Diabetes and Nutrition Study Group of the European Association for the Study of Diabetes(Toeller M July 2010);

2obtained from ADA: American Diabetic Association(American Diabetes Association (ADA) 2011)

\* Evidence grade A: evidence obtained from meta-analyses of randomized controlled trials or at least one randomized controlled trial

\*\* Evidence grade B: evidence obtained from at least one well designed and controlled study without randomization, well-designed quasi-

\*\*\* Evidence grade C-E: evidence obtained from expert committee reports or opinions and/or clinical experiences of respected authorities

Table 4. Nutritional recommendations for persons with type 1 and type 2 diabetes

experimental or non-experimental descriptive study

The 'Diabetes and Nutrition Study Group of the European Association for the Study of Diabetes' (DNSG EASD) guidelines for persons with type 1 and type 2 diabetes (**Table 4**) recommend that the most appropriate intake of carbohydrates consists of vegetables, legumes, fruits, wholegrain foods and naturally occurring foods rich in fiber. The fiber intake should be ideally ≈ 20 g/1000 kcal/day. Cross-sectional data of the EURODIAB Complications Study showed an inverse association between fiber and LDL cholesterol and a positive association between fiber and HDL cholesterol. In addition dietary fiber was inversely and significantly related to CVD (Toeller M et al. 1999 2). This effect was already found with a fiber intake of approximately 8.1 g/1000 kcal, which is below the recommended intake. The average fiber intake in type 1 diabetic patients is 8.1 g/1000 kcal, but the recommended intake is 20 g/1000 kcal. Recommendation was only achieved in 0.4% of the type 1 diabetic population (Toeller M et al. 1996). Data from the EURODIAB Prospective Complication Study on fiber intake measured at baseline by 3-day food diaries and presented by each center is given in F**igure 1**. As seen in this figure, even the 10 g/1000 kcal recommended fiber intake by the 'American Diabetes Association' (ADA) was hardly achieved by type 1 diabetic patients. Only Finnish type 1 diabetic patients achieved the ADA fiber recommendation of 10 g/1000 kcal F**igure 1**). Keeping in mind that these samples are clinic based and not population based and that these figures may not exactly reflect the current nutritional intake, however it gives an indication of the status on fiber intake. Although positive effects were already found on CVD with a fiber intake of 8.1 g/1000 kcal, we assume that effects could be probably even higher when recommended levels of fiber intake are reached. Unfortunately, this positive effect of fiber on CVD and CVD risk factors was only found in cross-sectional studies. This makes it very difficult to distinguish cause and effect. Further research in prospective studies or randomized controlled trials is needed to ascertain the role of fiber in CVD.

DNSG EASD do not recommend a low carbohydrate diet for type 1 and type 2 diabetic patients (**Table 4**). A low carbohydrate diet does not produce beneficial health effect. It is more acceptable to avoid too much foods high in fast available carbohydrates, foods high in fat and cholesterol. An earlier quote (Helgeson 2006) expressed this precisely: 'families of adolescents with diabetes may be more concerned that the sugar in candy is going to translate into high blood glucose levels today than that the fat in potato chips will translate into cardiovascular disease in 10 years'.

#### **5.2 Fat**

The DNSG EASD guidelines for dietary fat for persons with type 1 and type 2 diabetes recommend a saturated and trans-unsaturated fatty acid consumption of <10% of the total energy intake (<8% if LDL cholesterol is elevated). Total fat intake should not exceed 35% of total energy and dietary cholesterol should be <300 mg/day (T**able 4**). Saturated fat is an important risk factor for diabetic nephropathy, diabetic retinopathy as well as CVD (Riley

Diet, Lifestyle and Chronic Complications in Type 1 Diabetic Patients 37

only a few studies examining the effect of MUFA or PUFA on chronic complications in type 1 diabetic patients. A positive association was found between MUFA and retinopathy (Cundiff DK& Nigg CR 2005) but no association was found between MUFA and PUFA and microalbuminuria (Riley MD& Dwyer T 1998). These conclusions are based on posthoc analyses and a cross-sectional study respectively and should therefore be interpreted carefully. Also the conclusion of Strychar et al. (Strychar I et al. 2009) to recommend a diet higher in MUFA and lower in carbohydrate for nonobese type 1 diabetic individuals to reduce CVD risk factors is doubtful. Their conclusion is based on PAI-1 and VLDL levels, which are not such a good predictors for atherosclerosis (and by extension CVD) as TG levels are. And a high MUFA diet did not alter TG levels. Furthermore, the small study population of 30 subjects limits the power of their conclusions. In order to make accurate recommendations concerning MUFA and PUFA intake for type 1 diabetic patients more

research with more participants (preferably in a prospective study) is needed.

Fig. 2. Mean saturated fatty acid intake in 1102 individuals with type 1 diabetes across

Europe (Toeller M, Soedamah-Muthu 2011)

MD& Dwyer T 1998; Toeller M et al. 1999 3; Cárdenas C et al. 2004; Cundiff DK& Nigg CR 2005). The recommended intake is <10% of the total energy intake which was only achieved by a small minority (14%) (Toeller M et al. 1996). Data from the EURODIAB Prospective Complication Study on saturated fatty acid intake measured at baseline by 3-day food diaries and presented by each center is given in **Figure 2**. The even lower saturated fatty acid recommendation of <7% total energy of the ADA was not achieved by any of the centers (**Figure 2**). All centers indicated in F**igure 2** exceed the recommendation of <7% saturated fat of the total energy intake. Type 1 diabetic patients from Italy had the lowest intake of saturated fatty acids, but this intake was still too high (F**igure 2**). Again, keeping in mind that these samples are clinic based and not population based and that these figures may not exactly reflect the current nutritional intake.

Fig. 1. Mean fiber intake in 1102 individuals with type 1 diabetes across Europa (Toeller M, Soedamah-Muthu 2011)

Furthermore the DNSG EASD guidelines recommend oils rich in MUFA (10-20% total energy) and that PUFA should not exceed 10% of total energy intake (T**able 4**). There were

MD& Dwyer T 1998; Toeller M et al. 1999 3; Cárdenas C et al. 2004; Cundiff DK& Nigg CR 2005). The recommended intake is <10% of the total energy intake which was only achieved by a small minority (14%) (Toeller M et al. 1996). Data from the EURODIAB Prospective Complication Study on saturated fatty acid intake measured at baseline by 3-day food diaries and presented by each center is given in **Figure 2**. The even lower saturated fatty acid recommendation of <7% total energy of the ADA was not achieved by any of the centers (**Figure 2**). All centers indicated in F**igure 2** exceed the recommendation of <7% saturated fat of the total energy intake. Type 1 diabetic patients from Italy had the lowest intake of saturated fatty acids, but this intake was still too high (F**igure 2**). Again, keeping in mind that these samples are clinic based and not population based and that these figures

Fig. 1. Mean fiber intake in 1102 individuals with type 1 diabetes across Europa (Toeller M,

Furthermore the DNSG EASD guidelines recommend oils rich in MUFA (10-20% total energy) and that PUFA should not exceed 10% of total energy intake (T**able 4**). There were

may not exactly reflect the current nutritional intake.

Soedamah-Muthu 2011)

only a few studies examining the effect of MUFA or PUFA on chronic complications in type 1 diabetic patients. A positive association was found between MUFA and retinopathy (Cundiff DK& Nigg CR 2005) but no association was found between MUFA and PUFA and microalbuminuria (Riley MD& Dwyer T 1998). These conclusions are based on posthoc analyses and a cross-sectional study respectively and should therefore be interpreted carefully. Also the conclusion of Strychar et al. (Strychar I et al. 2009) to recommend a diet higher in MUFA and lower in carbohydrate for nonobese type 1 diabetic individuals to reduce CVD risk factors is doubtful. Their conclusion is based on PAI-1 and VLDL levels, which are not such a good predictors for atherosclerosis (and by extension CVD) as TG levels are. And a high MUFA diet did not alter TG levels. Furthermore, the small study population of 30 subjects limits the power of their conclusions. In order to make accurate recommendations concerning MUFA and PUFA intake for type 1 diabetic patients more research with more participants (preferably in a prospective study) is needed.

Fig. 2. Mean saturated fatty acid intake in 1102 individuals with type 1 diabetes across Europe (Toeller M, Soedamah-Muthu 2011)

Diet, Lifestyle and Chronic Complications in Type 1 Diabetic Patients 39

(Beulens et al. 2008). Because of this favourable effects we expect a beneficial effect on CVD, however to date no association was found between alcohol and CVD in type 1 diabetes patients (Bishop et al. 2009). In this cross-sectional study, markers for CVD were used instead of CVD as endpoint. Also the association between alcohol and diabetic nephropathy and diabetic retinopathy was only observed in cross-sectional studies. So the current recommendations for alcohol are confirmed by research in type 1 diabetes, but only based on cross-sectional studies, and especially for the association between alcohol and CVD in

There are no specific guidelines concerning physical activity for type 1 diabetic patients. The guidelines mentioned in **Table 4** are only for type 2 diabetic patients. However, it was shown that the guidelines for type 2 diabetic patients are also applicable for type 1 diabetic patients. Several randomized controlled trials (**Table 3**) showed that physical activity (endurance sports; on average 2 times a week 60 minutes) improves physical fitness as well as endothelial function in type 1 diabetic patients (Lehmann R et al. 1997; Fuchsjäger-Mayrl G et al. 2002; Seeger JPH et al. 2011). Especially the improvement in endothelial function is important since endothelial dysfunction is an early sign of atherosclerosis, which is often the underlying cause of CVD. Also a positive effect on lipid related cardiovascular risk factors was found in one study (Lehmann R et al. 1997). However, also this conclusion should be interpreted carefully. Although the evidence is gained from randomized controlled trials, the conditions of these trials are really disappointing. They had a maximum sample size of 23 people, and a minimum sample size of only 9 people. The follow-up period was relatively short, up to four months. The studies of Lehman et al. (Lehmann R et al. 1997) and Seeger et al. (Seeger JPH et al. 2011) not even used a control group. Furthermore CVD risk factors were used instead of CVD as endpoint. So the studies are in agreement with the guidelines but more research in better performed randomized controlled trials is needed to

confirm this positive effect of physical activity on CVD in type 1 diabetic patients.

role in preventing longterm complications of type 1 diabetes.

from http://www.diabetes.org/diabetes-basics/.

A diet high in fiber, low in saturated fat, moderate in protein intake with moderate alcohol consumption as well as physical activity can be recommended for type 1 diabetic patients to prevent complications. Inspite of the lack of large robust prospective studies, using the available evidence, we can conclude that diet as well as lifestyle could play an important

American Diabetes Association (ADA). (2010). "Diabetes Basics." Retrieved 24 March, 2011,

American Diabetes Association (ADA). (2011). "Recommendations Summary Diabetes

http://www.adaevidencelibrary.com/template.cfm?template=guide\_summary&k

Mellitus (DM): Physical Activity." Retrieved 19 April, 2011, from

type 1 diabetic patients more research is needed.

**5.4 Physical activity** 

**6. Conclusion** 

**7. References** 

ey=2062.

The recommendation of the DNSG EASD to consume 2-3 servings of oily fish/week and plant sources of n-3 fatty acids (**Table 4**) is consistent with the findings in studies specific for type 1 diabetes. The prospective cohort study of Lee et al. (Lee CC et al. 2010) found that dietary n-3 PUFAs (eicosapentaenoic acid and docosahexaenoic acid) are inversely associated with the degree but not with the incidence of albuminuria in type 1 diabetes. A hypothesis is that n-3 PUFAs decrease urinary AER via anti-inflammatory mechanisms. It decreases lipopolysaccharide-induced nuclear factor-kB (NF-kB ) activation and monocyte chemoattractant protein (MCP)-1 expression in human renal tubular cells (Lee CC et al. 2010). Further prospective studies and randomized controlled trials are needed to confirm this hypothesis.

#### **5.3 Protein**

With regards to protein, the DNSG EASD guidelines recommend an intake of 0.8 g/kg normal body weight in patients with type 1 diabetes and established nephropathy. There are no firm recommendations regarding protein intake for type 1 diabetic patients with incipient nephropathy. An intake of 10-20% of total energy is recommended for patients with no evidence of nephropathy (T**able 4**). The recommendation for protein intake is most important for patients with diabetic nephropathy. The guideline of a restricted protein diet which contains 0.8 g/kg normal body weight for type 1 diabetic patients with established nephropathy was demonstrated by previous research. Several randomized controlled trials showed that protein normalization (protein intake of approximately 0.8 g/kg/day, T**able 1**) had a positive significant effect on albuminuria, although no effect on GFR was found (Brouhard BH& LaGrone L 1990; Zeller K et al. 1991; Dullaart RP et al. 1993; Raal FJ et al. 1994; Hansen HP et al. 1999; Hansen HP et al. 2002). Even a relative risk of 0.23 (95% CI: 0.07-0.72) was found for ESRD in patients assigned to a low protein diet compared with patients assigned to a usual protein diet (Hansen HP et al. 2002). A hypothesis is that excessive protein intake causes renal vasodilatation and glomerular excessive perfusion leading to a raised glomerular transcapillary hydraulic pressure gradient ending in proteinuria and glomerular damage, conversely, will prevent kidney damage (Percheron C et al. 1995). So, indeed protein restriction is beneficial for type 1 diabetic patients with established nephropathy. However, we have to mention that although this beneficial effect of a restricted protein intake was found in randomized controlled trials, the sample size of these trials were really small (maximum of 82 people). Furthermore, we have to consider the feasibility of a protein intake of 0.8 g/kg/day. Percheron et al. (Percheron C et al. 1995) showed that even with this intake the compliance is poor. Further studies with a larger sample size are needed to find a cutoff point for protein intake which would still have a positive effect on diabetic nephropathy and its feasibility.

#### **Alcohol**

The DNSG EASD guidelines for alcohol for persons with type 1 and type 2 diabetes recommend a moderate use up to 10 g/day for women and up to 20 g/day for men (**Table 4**). In prior studies, moderate alcohol consumers (30-70 g alcohol per week) had a lower risk of diabetic nephropathy (OR=0.36, 95% CI: 0.18-0.71) and diabetic retinopathy (OR=0.60, 95% CI: 0.37-0.99) in patients with type 1 diabetes (Beulens et al. 2008). Alcohol has favourable effects on HDL-cholesterol, inflammation and inhibition of platelet aggregation (Beulens et al. 2008). Because of this favourable effects we expect a beneficial effect on CVD, however to date no association was found between alcohol and CVD in type 1 diabetes patients (Bishop et al. 2009). In this cross-sectional study, markers for CVD were used instead of CVD as endpoint. Also the association between alcohol and diabetic nephropathy and diabetic retinopathy was only observed in cross-sectional studies. So the current recommendations for alcohol are confirmed by research in type 1 diabetes, but only based on cross-sectional studies, and especially for the association between alcohol and CVD in type 1 diabetic patients more research is needed.
