Thursday , September 20 2018
Home / Resources / Clinical Gems / ADA/JDRF Type 1 Diabetes Sourcebook, Excerpt #8: Nutrition, Part 4

ADA/JDRF Type 1 Diabetes Sourcebook, Excerpt #8: Nutrition, Part 4

Aug 11, 2013

Anne Peters, MD, and Lori Laffel, MD, MPH, Editors
Jane Lee Chiang, MD, Managing Editor


Marion J. Franz, MS, RD, CDE; Alison B. Evert, MS, RD, CDE; Gail Spiegel, MS, RD, CDE; Carol Brunzell, RD, CDE; Joyce Green Pastors, MS, RD, CDE; Joshua J. Neumiller, PharmD, CDE, CGP, FASCP; Laurie A. Higgins, MS, RD, LDN, CDE; and Mary Ziotas Zacharatos, RD, CDE, LD


There are no adequately controlled studies that link pathogenesis of carbohydrate intolerance to trace elements.51,52 Although animal studies have suggested that deficiencies in many of the trace elements — including zinc, chromium, magnesium, copper, manganese, and vitamin B6 — may lead to glucose intolerance, the evidence is not definitive. Animal laboratory diets, in comparison to human diets, can be easily manipulated. Thus, animal study results should not be extrapolated to humans without validating the findings in human pilot studies.53

In studies with individuals with T1D, trace-metal and water-soluble vitamin urinary losses are increased during uncontrolled hyperglycemia and glycosuria. Therefore, the micronutrient effect may depend on the degree of glucose tolerance. Also, the micro nutrient effect on insulin secretion is biphasic. Low vitamin concentrations may stimulate insulin secretion, and high concentrations may have an inhibitory effect.53….

In human studies, dietary micronutrient amounts are often unknown. To further confuse the micronutrient role and diabetes, serum or tissue content of certain elements — copper, manganese, iron, and selenium — can be higher in people with diabetes than in control subjects without diabetes. On the other hand, serum ascorbic acid (vitamin C), B vitamins, and vitamin D may be lower in individuals with diabetes, whereas vitamins A and E have been reported to be normal or increased.53

Regardless, micronutrients are intimately involved in carbohydrate or glucose metabolism, insulin release, and insulin sensitivity. Unfortunately, this information is frequently extrapolated beyond what the research supports. The ADA recommends that individuals optimize food choices in meal plans to meet RDA and DRI intakes for all micronutrients.11,53


Earlier studies in T1D subjects showed no acute effect on blood glucose levels with moderate alcohol intake with meals. Further studies reported associations with late-onset hypoglycemia.54 Possible causes include inhibition of gluconeogenesis, reduced hypoglycemia awareness due to cerebral effects of alcohol, or impaired counterregulatory responses to hypoglycemia. A study with men with T1D who consumed wine in the evening (0.75 g alcohol/kg body weight; ~20 oz for a 70-kg individual) resulted in hypoglycemia treatment being required after breakfast.55 Growth hormone levels were significantly reduced, but no other differences in insulin or other hormone levels were reported. 55 Similarly, in T1D adults, hypoglycemia (blood glucose 50 mg/dl) resulted in lower peak growth hormone levels compared to placebo; however, this study was also associated with a decrease in insulin sensitivity.56 In a study similar to the men’s wine study, T1D individuals drank either orange juice or vodka with their evening meal. Continuous glucose monitoring data showed that individuals who drank alcohol (0.85 g alcohol/kg body weight) had more than twice as many hypoglycemic episodes throughout the next 24 h than those who drank orange juice.57,58

In T1D subjects, both mild alcohol intoxication and hypoglycemia (blood glucose ~43 mg/dl) were associated with deteriorating reaction time and other cognitive function tests, and total impairment was greater when both were experienced together.59 The authors emphasized that individuals must test blood glucose levels before driving. They should not drive when mildly hypoglycemic, even if asymptomatic.58

Elevated total ketone body concentrations are characteristic of both diabetic ketoacidosis (DKA) and alcoholic ketoacidosis (AKA). However, compared to AKA, DKA is characterized by a higher glucose concentration and lower β-hydroxybutyrate:acetoacetate and lactate:pyruvate ratios. Hormonal profiles are similar with decreased insulin levels and elevated levels of counterregulatory hormones.60 T1D subjects who liberally consumed alcohol at lunchtime had elevated postprandial β-hydroxybutrate levels vs. suppressed levels with placebo.61 The authors propose that binge drinking may increase the risk of significant ketosis, especially with erratic insulin administration, and recommend that patient education materials highlight these potential problems.58

Adolescents, in particular, must be instructed on alcohol and its potential hypoglycemic effects, and on responsible drinking. Adolescents who drive should be instructed on blood glucose monitoring before driving and on carrying a carbohydrate source with them at all times in case hypoglycemia should occur.

In summary, moderate alcohol (one drink/day or less for women and two drinks/day or less for men) consumption appears to have minimal, if any, acute effects on glucose levels and insulin need, but patients must be aware of the occurrence of late-onset hypoglycemia, likely due to alcohol-related growth hormone reduction. Thus, alcohol should be consumed with food.13 Also, patients should repeatedly self-monitor blood glucose levels after drinking alcohol to assess if hypoglycemia treatment is needed. The additive effect of alcohol and hypoglycemia on cognitive function highlights the urgency of avoiding alcohol when planning to drive.58


Carbohydrate counting is a meal-planning method based on the principle that all types of carbohydrate (except fiber) are digested, with the majority being absorbed into the bloodstream as glucose molecules. The total carbohydrate consumed has a greater effect on blood glucose elevations than the specific type consumed. There are two main methods of meal planning: an I:C ratio to adjust prandial insulin for variable carbohydrate intake (physiological insulin regimen) or a consistent carbohydrate meal plan when using a fixed insulin regimen.10

The most widely used method of meal planning for youth with T1D is carbohydrate counting. Rigid meal plans have been replaced with more flexible ones, matching insulin to the child’s nutrition (carbohydrate) intake. For people who have difficulty with carbohydrate counting, simplified, healthy eating meal-planning guidelines are recommended. Which method a child uses will depend on the insulin regimen and the family’s skill level. Often, the youth will start a basal- bolus insulin regimen (multiple daily injections [MDIs]) and then transition to an insulin pump, if desired. Nutritional recommendations will be made based on a child’s age and eating patterns.10

To accurately count carbohydrate amounts, children and their families are taught how to read the nutrition facts on food labels for total carbohydrate grams. Families should measure or weigh foods periodically to reinforce accurate portion sizes, and thus accurate carbohydrate content, so the correct insulin dose can be taken. There are books, websites, and smartphone applications that provide carbohydrate content for unlabeled foods. Families should have easy access to one of these resources to accurately estimate carbohydrates. Some school districts are displaying carbohydrate information for school breakfasts and lunches, facilitating carbohydrate counting; if not displayed, the information is available from the school lunch program.10

Because accurate carbohydrate counting is essential for accurate insulin dosing, researchers have evaluated carbohydrate-counting accuracy in the pediatric population. Research in children, adolescents, and their parents indicates that individuals may not be accurately estimating the carbohydrates. In one study, parents of 4- to 12-year-old children overestimated carbohydrate intake of their children by an average of 120% of the nutrition database calculated intake.62 Another study found that adolescents either significantly over- or underestimated carbohydrate content of 23 of 32 individual foods presented as real foods or food models.63 Lastly, a study conducted in the United Kingdom and Australia found that adolescents estimated carbohydrates within 10–15 g of the actual amount for 73% of meals presented.64 These authors concluded that adolescents carbohydrate count reasonably well, but if accuracy was defined more stringently (within 10 g of the actual amount), then many estimates would have been inaccurate. Additional research is needed to help determine the best strategies for helping children, adolescents, and their families enhance their carbohydrate-counting skills and potentially improve glycemic control.10

Fixed Carbohydrate Meal Plan

Both children and adults using fixed daily insulin doses must use a carbohydrate-counting meal-planning approach or some other method of quantifying carbohydrate intake.10 Alternatives to carbohydrate counting include 1) the plate method, and 2) preplanned menus. Accuracy in portion sizes remains important, and creative education is encouraged to promote accuracy, such as using beverage glasses, plates, and bowls that have lines or patterns that guide serving (portion) sizes.

Food Factors Affecting Glycemic Control

Postprandial hyperglycemia involves more than knowing how to count carbohydrates. Many T1D patients struggle to understand why their blood glucose levels dramatically fluctuate on a daily basis despite eating consistent carbohydrate grams. One explanation may be due to inadequate education on how to accurately dose prandial insulin and quantify carbohydrate intake.65 The CDC reports that only 55.7% of people with diabetes participate in a diabetes self-management education (DSME) class, suggesting that many patients never receive formal instructions on meal planning, such as carbohydrate counting, to enable accurate quantification of carbohydrate intake.66 Consequently, they may either under- or overdose prandial insulin requirements. An accurate prandial insulin dose to actual food (carbohydrate grams) intake is a critical component of basal-bolus insulin therapy.12 Aside from correct carbohydrate counting, several extrinsic and intrinsic variables affect glycemic control. Extrinsic factors, such as macronutrient distribution of the meal, fasting or preprandial blood glucose level, available insulin, antecedent exercise, and degree of insulin resistance may influence the impact of carbohydrates on the postprandial response.13 Additionally, intrinsic variables include type and source of carbohydrate, the physical form of the food (e.g., whole food vs. juice), starch type (e.g., amylopectin vs. amylose), method of food preparation (e.g., baking vs. frying), cooking time and amount of heat and moisture used, degree of processing, and ripeness of food.13 Individuals can use information from self-monitoring of blood glucose (SMBG) and continuous glucose sensors to better learn how both the extrinsic and intrinsic variables affect their glycemic control.12

Meal-Planning Approaches and Tools

Other than carbohydrate counting, meal-planning approaches such as the glycemic index also have been studied. Australian researchers developed a food insulin index, a physiological basis for ranking foods according to insulin demand for 120 single foods.67 They concluded that the relative insulin demand evoked by mixed meals consumed by lean healthy subjects is best predicted by a physiological index (food insulin index) based on integrating insulin responses to isoenergetic portions of single foods. Eating patterns that provoke less insulin secretion may be helpful in managing diabetes. In 2011, another Australian study compared a novel algorithm based on the food insulin index for estimating mealtime insulin dose to carbohydrate counting in T1D adults using CSII.68 The study concluded that, when compared with carbohydrate counting, the food insulin index algorithm improved acute postprandial glycemia in well-controlled T1D subjects. The authors acknowledge that clinical application of these findings is not currently feasible, since the food insulin index does not presently appear on food labels and the food insulin index database includes only ~120 foods.12

Another group collected data on food intake, physical activity, insulin administration, and blood glucose test results in T1D patients using self-administered questionnaires.69 Sixty-four percent of the participants incorrectly estimated their prandial insulin, revealing that optimal prandial insulin dosing is not easy and requires continuous assessment and related education and support, even after a long duration of diabetes.12

Insulin dosing aids such as bolus insulin calculation cards and dosing guides have been developed to reduce potential calculation errors.70–72 Bolus calculators with personalized insulin-dosing algorithms can be programmed in a wide range of devices, such as personal digital assistants (PDAs), smartphone applications, or insulin pumps.12,73,74

The Diabetes Interactive Diary is an automatic carbohydrate and insulin bolus calculator installed on a mobile phone, using patient-physician communication via text messages. When compared with a standard carbohydrate-counting education program, the Diabetes Interactive Diary was as effective as a traditional carbohydrate-counting education program, without an increased hypoglycemia risk.75 Technology has reduced education time while significantly improving treatment satisfaction and several quality-of-life dimensions. Adaptive aids are popular with the tech-savvy but may be useful for those with health literacy and numeracy concerns, such as young children or adults who cannot perform complex mathematical equations required for intensive insulin therapy.76 Technology may allow more people with insulin-requiring diabetes to have access to diabetes self-management tools, education, and support.12

Factors that May Affect Long-Term Adherence to Carbohydrate Counting

Three studies have explored the food and eating practices of T1D subjects who converted to flexible intensive insulin therapy (FIIT) as part of the DAFNE course.5,16,77 Ironically, in efforts to simplify food choices for easier carbohydrate estimation, patients may rely on prepackaged foods, with higher saturated fats and salt, but with nutrition labels, rather than calculate the carbohydrate content for fresh fruits, vegetables, and other unprocessed items that do not have food labels. FIIT participants also expressed anxieties about miscalculating carbohydrate amounts and injecting the wrong dose. This caused participants to eat the same foods repeatedly, limiting intake of new foods or foods with difficult-to-determine carbohydrate content. Some participants intentionally choose low- or no-carbohydrate foods to simplify prandial dose calculations. Despite formal intensive insulin therapy classes, many subjects feared hypoglycemia when matching mealtime insulin to desired food (carbohydrate) intake.5 These data raise factors that need to be addressed during initial and ongoing nutrition therapy. Strategies are needed to successfully sustain this therapy on a daily basis.12,77

One study interviewed DAFNE program participants at 6 weeks and 6 and 12 months on assimilating course principles.16 Subjects initially (6 weeks) felt support from other participants, for example, by sharing experiences. However, after 6 months, subjects valued support from responsive health care professionals that focused on collaborative decision making. The investigators concluded that diabetes educators must clearly communicate to participants that FIIT principles take time (perhaps over 12 months). Support at 6 months appeared to be an important timeframe for subjects, since motivation at this point was lowest for many.12

People with insulin-requiring diabetes may also diligently perform dose calculations using their individualized algorithms when beginning intensive insulin therapy.73 However, adherence to the ongoing determination of the prandial insulin dose may become relaxed as the individual with diabetes gains familiarity with the self-adjustment of the insulin. As time passes, there may be the tendency to begin to approximate premeal doses by titrating insulin based on the standard or usual carbohydrate content of the meal. In addition, many people with insulin-requiring diabetes may actually be hesitant to take on the responsibility of increasing or decreasing their insulin doses on the basis of their carbohydrate intake and premeal blood glucose level.12,73


Disordered Eating Behaviors, Eating Disorders, and Other Age-Related Concerns

Body image and weight-management issues in T1D adolescents and young adults with T1D may lead to eating disorders and disordered eating behaviors (see chapter 8, Psychosocial Issues in Type 1 Diabetes). It is unclear if there is an increased prevalence of diagnosable eating disorders and disordered eating behaviors in T1D patients compared with the general population. Some studies show a higher rate in T1D patients, while others have found the same or lower rates.78 Estimates in T1D adolescent and young adult females range from 3.8 to 27.5% for patients classified as bulimic or having binge eating disorder. When insulin omission is considered purging, the estimate is as high as 38-40%. The presence of eating disorders has been associated with increases in retinopathy, neuropathy, transient lipid abnormalities, hospitalizations for diabetic ketoacidosis, and poor short-term metabolic control. Adolescents with diabetes should be screened regularly for signs of potential eating issues and concerns with weight and body image as well as insulin omission. After screening, issues can be addressed and action can be taken to prevent the development of an eating disorder, which is very complicated to treat especially when it coexists with T1D. Warning signs that suggest an eating disorder in adolescents include inadequate weight gain or growth, significant weight loss without illness, suboptimal overall glycemic control, and recurrent diabetic ketoacidosis. If signs of disordered eating and weight or body image concerns are present, they need to be addressed; referral to a dietitian and psychotherapist or psychologist is recommended. If a patient is at high risk of an eating disorder, the patient should be referred to an eating disorder program for an assessment and treatment, if necessary.10

Finally, adolescents with diabetes may experiment with alternative eating patterns, such as vegetarianism or nutritional supplement use. Practical information on these topics will enable adolescents to make wise choices for their health.10


Celiac disease (CD) is an important entity to consider since those with T1D are also at increased risk for developing CD. Individuals diagnosed with both CD and T1D should seek the care of an RD familiar with the nutritional management of both entities. The RD should also provide comprehensive support and education about gluten-free diets (GFD).

The Gluten-Free Diet

Since nutritional deficiencies have been reported with long-term GFD, comprehensive nutrition assessments must be done to ensure adequate nutrient intake.79,80 A GFD can be extremely challenging, since ongoing monitoring of ingredients in foods and food processing are intricate parts of nutrition interventions.81

For newly diagnosed children and adults with CD, studies report that adherence to a gluten-free eating pattern results in significant improvements in serum Hb, iron, zinc, and calcium, as a result of intestinal healing and improved absorption. However, adherence to the gluten-free eating pattern may result in a diet that is high in fat and low in carbohydrates, fiber, iron, folate, niacin, vitamin B12, calcium, phosphorus, and zinc. A small number of adult studies show a trend toward weight gain after diagnosis.81,82

Several studies report that patients with CD (treated and untreated) are more likely to experience gastrointestinal symptoms such as diarrhea, constipation, abdominal pain and bloating, nausea or vomiting, reduced gut motility, and delayed gastric emptying than healthy control subjects. However, long-term adherence to a GFD has been shown to reduce the prevalence of these symptoms.81,82

Implementing the Gluten-Free Diet

A GFD can be more expensive than a normal diet and requires extensive, repeated counseling and RD instruction. Patients with CD must be meticulous label readers and knowledgeable about food processing, preparation, and handling practices to avoid cross-contamination with gluten-containing grains. As little as 10 mg gluten (1/50th a slice of bread) can cause significant mucosal inflammation in some individuals. Therefore, it is recommended that contaminating gluten should be kept to <50 mg/day in the treatment of CD.83 Gluten-free grains, seeds, and flours may inadvertently become contaminated; therefore, food manufacturers are urged to test gluten-free products for contamination using validated testing methods.84 The FDA has proposed that gluten-free products contain <20 parts/million gluten using validated testing methods. Unfortunately, this testing would be voluntary. Prescription and over-the-counter medications, vitamins, minerals and supplements, and nonfood items such as play dough may also contain gluten.81

Studies evaluating the impact of gluten-free diets on T1D have shown conflicting results, with some showing no change in A1C levels in either children or adults,85–92 while others have shown improvements in A1C.93–97 One study noted a worse A1C in children.98 This indicates that with education and support, patients can successfully implement the CD and T1D food plans to maintain or improve A1C levels.81