1. Introduction
Overweight and obesity are currently associated with more deaths worldwide than underweight, according to a report by the World Health Organization in 2014.1 Obesity, a problem previously prevalent only in rich regions, is now a public health challenge also in low- and middle-income countries. Worldwide, 44% of diabetes can be attributed to overweight and obesity.1 Indeed, diabetes incidence has also increased in a parallel manner to obesity, reaching epidemic proportions. In 2014, the estimated global prevalence of diabetes was 9% among people aged 18 years and older.2
Nutrition is a key component of diabetes management, where it fulfills general (adequate growth and development, weight maintenance) and specific (cardiovascular (CV) protection, glycemic control) purposes. At the same time, eating has strong cultural implications and changes, and limitations to food intake have a great impact on quality of life.3,4 Thus, when giving dietary advice, it is crucial to focus on those recommendations whose benefits are based on strong, clinical evidence.
When the terms ânutrition and diabetesâ are used to start a search in PubMed, they result in 18,920 hits (February 9, 2015). If ânutrition OR dietâ are combined with diabetes, the number increases to 55,575 hits. Nevertheless, when the search is limited to randomized controlled trials, only 1049 and 3686 hits are found, respectively. Thus, although the interest and research on nutrition and diabetes are extensive, the highest level of clinical evidence represents only a minor fraction of the published studies.
The aim of this chapter is to review the relevant, clinical evidence available about the effects of nutritional interventions on the control of type 1 diabetes (T1D) and type 2 diabetes (T2D) and their complications. For this purpose, the most recently published international guidelines on the subject have been considered (Section 3) and a systematic review of randomized controlled trials has been performed (Section 4). In addition, to put present evidence into context, a historical description of nutritional recommendations and their changes in the past century is provided (Section 2).
2. Historical Perspective
Nutritional recommendations for diabetes have changed dramatically in the past century. Before the discovery of insulin, patients with T1D were advised to fast in order to obtain sugar-free urine. Once this was achieved, dietary carbohydrate content was increased by 10 g/day until persistent glycosuria appeared.5 This was done using green, bulky vegetables and, depending on the patient's âtolerance,â also some garden vegetables and sometimes even potatoes and cereal. Fruit generally remained a minor fraction of carbohydrate intake in these patients, preferably used as dessert. The carbohydrate tolerance of each patient was used to design his or her maintenance diet, where the maximum carbohydrate allowance was the highest amount at which the urine remained sugar-free. Protein intake was calculated to consist of 1.0â1.5 g/kg; the rest of the calories were accounted for by fat. The diet was adapted to the severity of diabetes and to the presence/absence of acidosis.5 The progression from a âvegetable day,â containing only 5 g of carbohydrate, is described by Hill as follows: â...carbohydrate 15 g, protein 25 g, fat 150 g. From this, the diet is slowly raised, increasing first the fat, then the protein and lastly the carbohydrate. The fat is never raised above 200 g and the calories seldom above 2200. On this, the patients hold their weight, feel well, and usually remain sugar-free.â6 After the discovery of insulin, the carbohydrate allowance increased progressively, as did the recommended total caloric intake7 (Table 1). In 1933, Elliott Joslin recommended the following: âAt present the diet I give my patients is approximately carbohydrate 140 g, protein 70 g, fat 90 g. Children need much more protein and if they require more calories I am inclined to give these calories equally divided between carbohydrate and fat.â Indeed, advocates for a normal diet for the affected children started their campaigns, based on better nutritional results, fewer acute complications, and last, but not least, better acceptance.8 The 1940s and 1950s witnessed a debate on whether patients with diabetes should be on a controlled or a free diet.9,10 Several studies showed similar results on weight and hypoglycemia, although larger glucose fluctuations were observed with the free diet.11 In the discussion, Forsyth et al. stated âFrom our experience of a group of 50 diabetics given liberal diets and insulin over a period of five years we are satisfied that, if adolescents and obese diabetics are excluded, clinical control, as defined earlier in this paper, can be attained in most patients. However, the degree of hyperglycemia and glycosuria and the daily fluctuation of blood-sugar levels are undoubtedly greater in such patients than in those on controlled diets.â Nevertheless, soon, evidence appeared to support that hyperglycemia was associated with a higher risk of retinopathy and vessel calcification.12,13 Thus, emphasis was put on the degree of glycemic control, while still supporting a relatively âfreeâ diet. According to Forsyth herself: âBy the term âfree dietâ we imply liberty rather than license. Simple instructions are given to ensure the quality of the diet, and regular timing of meals is considered essential. Concentrated carbohydrates, such as table sugar, jam, chocolate, and sweets, are restricted.â The American Diabetes Association (ADA) released its first exchange lists to facilitate constant dietary composition.14 The link between dietary fat and atherosclerosis was recognized and, thus, fat intake was progressively reduced, especially at the expense of saturated fat.15,16
The first oral agents (sulfonylureas and phenformin) were available for the treatment of diabetes from the late 1950s,1...