American Diabetes Association Guide to Nutrition Therapy for Diabetes. Marion J. Franz

Whole-grain foods contain fiber, vitamins, minerals, phenolic compounds, phytoestrogens, and other unmeasured constituents, which have been shown to lower serum lipids and blood pressure, improve glucose and insulin metabolism and endothelial function, and alleviate oxidative stress and inflammation in the general population (He 2010). In a prospective study of 7,822 women with type 2 diabetes, intakes of whole grain, cereal fiber, and bran were inversely associated with all-cause and CVD mortality during a 26-year follow-up (He 2010). Bran intake had the strongest association, and germ intake, which was also evaluated, was not associated with all-cause or CVD mortality.

      Glycemic index/glycemic load. The glycemic index (GI) measures the relative area under the glucose curve of 50 g digestible carbohydrate compared with 50 g of a standard food, either glucose or bread. The GI index does not measure how rapidly blood glucose levels increase after eating different types of carbohydrate-containing foods, which implies that a high-GI food peaks quickly and a low-GI food peaks later. In a review of studies comparing different types of low- and high-GI foods and glucose, in people without diabetes, glucose peaks occurred consistently at ~30 minutes, regardless of whether the food was categorized as low-, medium-, or high-GI, with a modest difference in glucose peak values between high- and low-GI foods (Brand-Miller 2009). In contrast to what is often stated, low-GI foods did not produce a slower rise in blood glucose, nor did they produce an extended, sustained glucose response.

      The estimated glycemic load of foods, meals, and eating patterns is calculated by multiplying the GI by the amount of carbohydrate in each food and then totaling the values for all foods in a meal or eating pattern. The glycemic load is used most often in research studies, especially in epidemiological studies, but because of the calculations needed, it is not likely a practical approach for individuals to use for planning meals or prandial insulin doses.

      After reviewing 15 studies reporting on the relationship between the GI values of foods/diets and metabolic outcomes, the Acad Nutr Diet EBNPG concluded the following: “Studies comparing high- versus low-GI diets report mixed effects on A1C levels. These studies are complicated by differing definitions of high-GI or low-GI diets or quartiles, as well as possible confounding dietary factors” (Acad Nutr Diet 2008; Franz 2010). The guidelines note that definitions of low- versus high-GI diets range from 38 to 77% for low-GI diets and from 63 to 98% for high-GI diets. Other problems include the variability of GI responses from carbohydrate-containing foods within and among individuals. As with carbohydrate, most individuals with diabetes appear to consume a moderate-GI diet, and it is unknown whether reducing the usual GI by a few units will result in improved glycemic control. Of the 15 studies reviewed, 12 are of short duration (<3 months) with a limited number of subjects. Only three studies were of 1-year duration. After 1 year, one study reported no difference in actual GI between the low-GI and control groups, and two studies reported no differences in A1C between the low-GI and control groups.

      The ADA also noted the conflicting evidence of randomized clinical trials of low- versus high-GI diets and also expressed concern about the variability in responses to specific carbohydrate-containing foods. ADA also noted that most individuals already consume a moderate-GI diet; however, for individuals consuming a high-GI diet, consuming a low-GI diet may result in a modest benefit in postprandial hyperglycemia (ADA 2008).

      Nonnutritive sweeteners and sugar alcohols. Five nonnutritive sweeteners are approved by the U.S. Food and Drug Administration (FDA) as food additives: aspartame, saccharine, acesulfame K, neotame, and sucralose; one other—stevia—is approved as Generally Recognized As Safe (GRAS). The FDA also sets a sweetener Acceptable Daily Intake (ADI), which is the level a person can safely consume on average every day over a lifetime without risk. The ADI is typically 1/100th of the amount of the nonnutritive sweeteners shown to be safe in animal studies. The ADA notes that before being allowed on the market, all nonnutritive sweeteners undergo rigorous scrutiny and are shown to be safe when consumed by the public, including people with diabetes and women during pregnancy (ADA 2008). The Acad Nutr Diet EBNPG note that although nonnutritive sweeteners independently do not effect changes in glycemic responses, some of the products sweetened with nonnutritive sweeteners contain energy and carbohydrate from other foods, and these foods need to be taken into consideration (Acad Nutr Diet 2008; Franz 2010).

      Reduced-calorie sweeteners approved by the FDA include sugar alcohols (polyols) such as erythritol, isomalt, lactitol, maltitol, mannitol, sorbitol, xylitol, tagatose, and hydrogenated starch hydrolysates. They produce lower postprandial glucose responses than sucrose or glucose and contain on average about 2 calories/g. There is no evidence that the amount of sugar alcohols likely to be consumed will reduce glycemia, energy intake, or weight. Although safe to use, they may cause diarrhea, especially in children (ADA 2008).

      Carbohydrate Summary

      Carbohydrates eaten and available insulin are the primary determinants of postprandial glucose levels. Foods containing carbohydrate—fruits, vegetables, whole grains, legumes, and low-fat dairy foods—are important components of a healthy eating pattern. For people with diabetes, these foods should be included in appropriate amounts and portion sizes in their food/meal plan. In addition, nutrient-dense foods are recommended. Nutrient-dense foods are foods and beverages that have not been “diluted” with the addition of added solid fats and added sugars. Monitoring carbohydrates, whether by carbohydrate counting, choices, or experience-based estimation, remains a key strategy in achieving glycemic control (ADA 2012). Although some individuals may note improvements in postprandial glucose responses with the use of the GI/glycemic load, the concept of the GI/glycemic load adds an additional level of complexity to nutrition therapy recommendations and is perhaps best used for fine-tuning postprandial responses after first focusing on total carbohydrate.

      The Acad Nutr Diet EBNPG recommend the following: “In persons receiving either medical nutrition therapy alone, glucose-lowering medications, or fixed insulin doses, meal and snack carbohydrate should be consistently distributed throughout the day on a day-to-day basis.” Nutrition therapy for people with type 2 diabetes is discussed in Chapter 6. The Acad Nutr Diet EBNPG also recommend the following: “In persons with type 1 (or type 2) diabetes who adjust their mealtime insulin doses or who are on insulin pump therapy, insulin doses should be adjusted to match carbohydrate intake (insulin-to-carbohydrate ratios). This can be accomplished by comprehensive nutrition education and counseling on interpretation of blood glucose patterns, nutrition-related medication adjustment, and collaboration with the health care team” (Acad Nutr Diet 2008; Franz 2010). Nutrition therapy for people using insulin therapy is discussed in Chapters 5, 7, and 21.

PROTEIN AND DIABETES NUTRITION THERAPY

      In people with type 1 or type 2 diabetes with normal renal function, both the Acad Nutr Diet EBNPG and the ADA currently have not found adequate evidence to support recommending a change in the usual protein intake of 15–20% of total daily energy intake (Acad Nutr Diet 2008; ADA 2008). Exceptions for change in protein intake are in individuals who consume excessive amounts of protein foods high in saturated fatty acids, in people who have a protein intake less than the RDA of 0.8 g good-quality protein/kg body weight/day (on average ~10% of energy intake), or in patients with diabetic nephropathy.

      In people with type 2 diabetes, ingestion of protein results in acute insulin and glucagon responses with minimal, if any, postprandial glucose or lipid responses (Acad Nutr Diet 2008; Papakonstantinou 2010a). Studies lasting 5–12 weeks comparing high-protein diets to lower-protein diets showed no differences in longer-term insulin response despite the acute insulin response.

      Studies in people with type 1 diabetes and protein intake are limited. In a study in which a standard lunch (450 kcal) was compared with a protein-added (+200 kcal) lunch, the early glucose response was similar, but the late glucose response (2–5 h) was slightly increased and required 3–4 units of additional insulin. However, the total insulin requirement over the 5 h was not increased (Peters 1993). Large amounts of protein appear to have the potential to modestly increase postprandial