| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Review |
Department of Internal Medicine, College of Medicine, University of Kentucky, Lexington, KY (J.W.A.)
Risk Factor Medical Center, St. Michael’s Hospital and Department of Nutrition, University of Toronto, Toronto, CANADA (C.W.C.K., D.J.A.J.)
Address correspondence to: James W. Anderson, MD, Metabolic Research Group, 1030 S. Broadway St., Suite 5, Lexington, KY 40504-2681. E-mail: jwandersmd{at}aol.com
 |
ABSTRACT |
|---|
 TOPFOOTNOTES ABSTRACT INTRODUCTIONREFERENCES |
|---|
50% within two weeks and these changes are sustained with continued energy restriction. Twelve weeks of energy-restricted diets were associated with these significant decreases: body weight, 9.6%; fasting plasma glucose, 25.7%; serum cholesterol, 9.2%; serum triglycerides, 26.7%; systolic blood pressure, 8.1%; and diastolic blood pressure, 8.6%. Larger weight losses were associated with larger reductions in these values. The reviewed data suggest that US health care providers should endorse the American Heart Association’s and European diabetes associations’ recommendations that diabetic persons achieve and maintain a BMI of 
25 kg/m2. Weight management may be the most important therapeutic task for most obese Type 2 diabetic individuals.
Key words: diabetes, obesity, coronary heart disease, weight loss, diabetes risk
Key teaching points:
• From 60% to 90% of Type 2 diabetes mellitus appears related to obesity or weight gain.
• Obesity of weight gain can increase the risk for developing diabetes by greater than ninetyfold.
• Obesity or weight gain can increase the risk for coronary heart disease in women by sixfold.
• Weight loss of 10% of initial body weight dramatically improves glycemic control and reduces lipid and blood pressure co-morbid risks.
• A weight goal of a body mass index 25 kg/m2 is recommended for persons with Type 2 diabetes mellitus.
• Weight management appears to be the most important therapeutic task for most Type 2 diabetic individuals.
|
INTRODUCTION |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONREFERENCES |
|---|
Obesity and diabetes have become so inseparable that they are like "conjoined twins." Women who were obese at 18 years of age and gain further weight have a risk for developing by diabetes that reaches greater than ninetytwofold the risk for women who were slender at age 18 and did not gain weight [4]. An estimated 60% to 90% of Type 2 diabetes is related to obesity [8–10]. Obesity may be the most impacting risk factor for CHD in non-diabetic individuals [11] and is probably even more important for individuals with diabetes [12,13]. Since CHD is responsible for about 70% to 80% of the deaths in persons with diabetes, addressing this obesity-diabetes connection is vital for adults with diabetes [13,14]. Unfortunately, most pharmacologic approaches for diabetes contribute to weight gain [15], as does intensive therapy [16,17]. This review was stimulated by the possibility that weight management should receive a higher priority in the management of certain obese individuals with diabetes.
To reassess treatment goals and strategies for obese diabetic individuals the following areas were assessed: effect of obesity or weight gain on risk for developing diabetes; impact of obesity or weight gain on risk for coronary heart disease (CHD), the major cause of death in diabetes; influence of diet on risk factors for CHD, diabetes and glycemic control; burden of CHD associated with diabetes; weight loss outcomes for persons with Type 2 diabetes; weight management recommendations.
Obesity and Weight Change Effects on Diabetes Risk
Obesity is the major risk factor for development of diabetes, accounting for 60% to 90% of the variance [8,9]. Colditz and colleagues [4] from the Nurses’ Health Study assessed the impact of three different factors on risk for developing diabetes in women: (1) BMI at age 18 years, an assessment of adolescent obesity, (2) weight gain after age 18, and (3) attained BMI at 30 to 55 years of age, assessing adolescent obesity as well as weight gain. The diagnosis of diabetes was based on either elevated plasma glucose levels or treatment with oral hypoglycemic agents or insulin. The age-adjusted risk for developing diabetes over 14 years of age increased from a standardized risk of 1.0 for a BMI of <22 kg/m2 at age 18 to a relative risk (RR) of 93.2 for an attained BMI of 
35.0 at 30 to 55 years of age. Knowler and colleagues [10] reported remarkably similar risks for diabetes associated with obesity in Pima Indians who had one or both parents with diabetes. The incidence of diabetes was determined by glucose tolerance test. They noted that the age-adjusted RR for developing diabetes if one or both parents had diabetes was 90.3 for men or women with BMI 40 kg/m2 compared to persons who also had one or both parents with diabetes but had a BMI < 20 kg/m2 (RR = 1). Selected data from these two studies are summarized in Fig. 1 because they demonstrate a remarkable similarity from two diverse populations and also illustrate the potential impact of obesity and weight gain on risk for developing diabetes.
|
50%, lost 6% of their initial body weight and maintained a weight loss of 3.5% at four years. One price of intensive glycemic control to achieve normal hemoglobin A1c values is weight gain. While metformin fosters modest weight loss, several classes of oral antidiabetes agents—sulfonylureas, meglitinides, and thiazolidinediones—as well as insulin promote weight gain [15]. In the DCCT, subjects with Type 1 diabetes receiving intensive treatment gained a mean of 4.75 kg [16]. Similarly, in the UKPDS subjects with Type 2 diabetes receiving intensive treatment gained about 5 kg [17]. While these two studies [15,16] reported that intense glycemic control might reduce risk for CHD, the role of glycemic control in risk for CHD is not established [14]. From our previous report [11], we can estimate that a 5 kg weight gain in men would increase CHD risk by almost 30%; the estimated changes in lipids and blood pressure related to a 5 kg weight gain would increase risk by approximately another 20% [11]. For obese diabetic individuals, it is uncertain whether the benefits of glycemic control related to protection from nephropathy and retinopathy are counterbalanced by a potential and substantial increase in risk for CHD.
Obesity or Weight Gain and Risk for CHD
Obesity or weight gain contributes importantly to the risk for diabetes as well as CHD. In 1983, Hubert and colleagues [26] reported that, after adjustment for coexisting risk factors, obesity was an independent risk factor for CHD. Subsequent studies have confirmed and extended these observations. Manson and colleagues [27] reported that even mild overweight significantly increased risk for CHD in middle-aged women. To assess the effect of BMI on risk for CHD, we [11] recently performed a meta-analysis of ten studies. The RR for CHD for women with an initial BMI of 33 kg/m2 was 3.17 (95% Confidence Intervals [CI] of 2.37 to 3.97) compared to women with an entry BMI of 23 kg/m2. For men, the RR for CHD for the higher BMI was 3.19 (95% CI, 2.62 to 3.76) compared to men with the lower BMI at entry. Thus, individuals with a BMI of 33 kg/m2 have a risk for CHD that is 320% of that for persons with a BMI of 23 kg/m2 [11]. Weight gain after teenage years, independent of initial BMI, is also associated with a substantial risk for CHD. Four studies [11] assessed the risk of weight gain for women after age 18 years old. The variance-adjusted RR was 1.83 (95% CI, -0.14 to 3.80) for women who gained 10–19 kg compared to those with a stable weight [11]. Four studies [11] assessed the risk of weight gain for men after 21 to 25 years of age. The variance-adjusted RR was 1.46 (95% CI, 0.86 to 2.07) for men who gained 10–19 kg compared to those who maintained a stable weight. Data were not available to calculate the CHD risk for obesity at 18 to 25 years of age plus weight gain. Obviously many obese teenagers gain further weight in adult life. Willett et al. [28] estimated this combined risk. Women who have a BMI 23.3 kg/m2 at age 18 and gain 20 kg over 14 years of follow-up have a RR of 6.0 or a risk of 600% compared to women who have a BMI of <23.3 kg/m2 and maintain a stable weight.
Impact of Diet on CHD Risk Factors and Diabetes
Dietary alterations have a major impact on risk factors for CHD. The Adult Treatment Panel III (ATP III) [29] recommends a multifaceted lifestyle approach to reducing risk for CHD including these: obtain 50% to 60% of energy from carbohydrates, emphasizing whole grains, cereals, fruits and vegetables; reduce saturated fat to <7% of energy intake and cholesterol to <200 mg/day; include 20 to 30 grams of fiber daily with 10 to 25 grams of soluble fiber intake; reduce weight; and increase physical activity. The ATP III recognizes overweight and obesity as major, underlying risk factors for CHD and recommends weight reduction and physical activity. This ATP III raised the risk level for persons with diabetes without CHD to a risk level equivalent to presence of CHD. The American Heart Association [30] recommendations for persons with diabetes include these guidelines: reduce energy intake and increase physical activity to achieve and maintain a healthy body weight (defined as a BMI 25 kg/m2); reduce intake of total dietary fat to <30% of energy; and decrease intake of saturated fat to <7% of energy.
The nutrient composition of the diet has crucial role for achieving and maintaining a desirable body weight (i.e., BMI 25 kg/m2) and reducing risk for CHD. We recently compared the CHD risk of several popular diets if consumed at a weight-maintaining level for an intermediate-time period compared to consumption of a typical U.S. diet [31]. These calculations suggest that popular low-carbohydrate, high-fat diets would increase serum cholesterol values by almost 30% while a very low-fat, vegetarian diet would decrease serum cholesterol by 30%. While there are concerns that high-carbohydrate diets may increase fasting serum triglyceride values, in fact, high-carbohydrate, high-fiber diets decrease fasting serum triglycerides values slightly [32,33]. These diets—rich in potassium, magnesium and other minerals—should also decrease blood pressure levels significantly [34]. Thus, high-carbohydrate, high-fiber, low-fat diets rich in vegetables, fruits and whole grains would improve a number of risk factors for CHD [31,34]. These low-fat diets also are effective for long-term maintenance of weight loss [35]. While individuals are receiving intensive counseling by dietitians and active support from primary health care providers, instead of recommending an atherogenic high-fat diet aimed at rapid weight loss it seems prudent to provide instruction in a lifelong, health-promoting diet [34].
Weight loss also significantly decreases risk factors for CHD. Effects of weight loss on several CHD risk factors were recently calculated [11]. A 10 kg reduction in body weight would be expected to produce these changes in risk factors for CHD: fasting serum cholesterol values, -9.9%; LDL-cholesterol, -6.8%; fasting serum triglycerides, -19.3%; HDL-cholesterol, +1.5%; systolic blood pressure, -4.9%; and diastolic blood pressure, -3.8% [11].
Weight loss also dramatically improves glycemic control [36,37] and can produce a remission in diabetes [38,39]. Even before significant weight loss occurs, energy restriction decreases plasma glucose values [36,37] and enhances insulin sensitivity [40]. A decrease in dietary fat intake also enhances insulin sensitivity [41]. In addition to contributing to a desirable weight, dietary components appear to have an important role in retardation of the development of diabetes [19,42,43]. High fat intakes, especially high intakes of saturated fat, seem to increase risk for developing Type 2 diabetes [19,42–44]. Protective dietary practices include generous intakes of these foods: low glycemic index foods [35,42,45], foods rich in cereal fiber and whole grains [42,46] and foods low in saturated fats [22,42]. In general, intake of a "healthy diet" rich in vegetables, fruits and whole grains appears to have significant protective effects [19,42,46]. Dietary practices such as these are likely to slow the evolution of a diabetic diathesis or impaired glucose tolerance into a more florid diabetes state [19].
Cardiovascular Disease and Diabetes
CHD and other atherosclerotic cardiovascular diseases (CVD) are the predominant cause of death for individuals with Type 1 as well as Type 2 diabetes. An estimated 70–80 percent of diabetic individuals die from CVD, indicating that this must be a major target in management of this condition [14,35,47,48]. Diabetes, per se, is a significant risk factor for coronary heart disease, especially in young and middle-aged women. While a number of studies provide RR assessments, the recent report of Gu and colleagues [49] is representative. For US residents, the diabetic man is at threefold greater risk and the diabetic woman is at 4.6-fold greater risk for death from CHD compared to non-diabetic individuals. Consequently, diabetes, per se, appears to be a greater risk factor for CHD than hypercholesterolemia, hypertension, or cigarette smoking [47,48].
Weight Loss Outcomes in Type 2 Diabetes
Because of the challenges in weight management for diabetic individuals, earlier studies over the past 30 years utilized total fasting or very-low-energy diets (VLED) over short periods of time. Greenfield et al. [50] reported that a ten-day total fast with weight loss of 5.1% of initial body weight decreased fasting plasma glucose values of Type 2 diabetic subjects by 64.5% (from 17.2 mmol/L to 6.1 mmol/L). We performed a systematic search of the literature using Medline and the ancestry approach [51] and identified ten studies of obese Type 2 diabetic subjects who were treated for four to six weeks with VLED. These studies provided weight and glucose values at multiple time points with estimates of variance. Statistical analyses were performed with SAS (SAS Online DOC, Version 8, SAS Institute, Cary, NC, 1999). Tests of homogeneity indicated that fixed-effects meta-analyses were associated with significant heterogeneity. Therefore, we performed meta-analyses of all data using random-effects models as previously described [51,52]. In Fig. 2 we summarize the response from nine studies [53–61] using VLEDs with 192 obese subjects with Type 2 diabetes. Over six weeks, subjects lost 9.6% of initial body weight. Fasting plasma glucose values decreased dramatically to 50% of initial values after two weeks and remained at less than half of the initial values for the six weeks of treatment. Although experimental protocols differed, e.g., oral antidiabetes agents or insulin were usually withdrawn for 3–14 days before initiating diets, there was a remarkable consistency in the response of diabetic individuals to energy restriction. These dramatic responses point clearly to excessive energy intake as a major factor in hyperglycemia of most obese Type 2 diabetic persons [62].
|
30 years [63]; Henry and Gumbiner [36] have reviewed their extensive experience. These VLEDs require medical supervision and are intensive and expensive [40]. Consequently, many health professionals now use LEDs providing between 800–1200 kcal/day in their clinical practice. To assess intermediate-term effects of weight loss on glycemic control and CHD risk factors of diabetic individuals, we systematically reviewed the literature over the past 30 years using Medline and the ancestry approach [51]. We analyzed all identified studies using these criteria: obese diabetic subjects with BMI >30 kg/m2 or >120% of ideal body weight (criteria used in some earlier studies); adults 18 years old; supervised use of energy-restricted diets; achievement of mean weight losses of 5% of initial body weight at 12 weeks; providing data at two time intervals at 6 weeks of intervention; providing data and variance for body weight and fasting plasma glucose at multiple times; and not using anti-obesity agents. After tabulating data we performed meta-analyses using a random-effects model as previously described [51].
From these studies the association between weight loss and fasting plasma glucose values were examined over a 48-week period (Table 1). Eleven reports provided data from 13 study groups [54,64–73]. These studies provided data for 376 obese diabetic subjects at various time intervals. Subjects lost weight steadily over 16 weeks to a weight that was 14.7% below their initial (baseline) weight. Over the next 32 weeks they regained 3 kg. Fasting plasma glucose values decreased to 30.5% below initial values at 16 weeks, remained at this level for about eight weeks, and then increased as subjects gained weight. Although study protocols and medication regimens differed across studies, there was a strong association between weight loss and improved glycemia irrespective of the study protocol.
|
25%. Wing and colleagues [37,65] also noted that greater weight loss was associated with greater improvement in glycemia.
|
Weight Management Considerations
Obesity is difficult to manage in diabetic or non-diabetic individuals. Diabetic individuals may have more difficulty losing weight than non-diabetic individuals because of genetic or metabolic differences, fear of hypoglycemia, antidiabetes medications, other medications, limited physical activity or diet fatigue. Perceived dietary restrictions, such as limitation of carbohydrate intake, may limit therapeutic choices for diabetic individuals. For example, health care professionals may be reluctant to recommend meal replacements, widely used by non-diabetic individuals [79]. Certain centrally acting pharmacologic agents are relatively contraindicated in diabetic individuals because of their cardiovascular effects [80]. Weight loss recommendations for non-diabetic individuals have not been clearly translated into guidelines for diabetic individuals. Expert panels have not reached a consensus about desirable weights for diabetic individuals. Some expert panels recommend that diabetic individuals attain desirable body weights (i.e., non-overweight weights or BMIs of 25 kg/m2) [81–83] as does the American Heart Association [30]. Other expert panels, including the American Diabetes Association, recommend attaining a reasonable weight (i.e., a weight loss of 5% to 7% of current weight irrespective of initial weight) [84–86]. Finally, there is a therapeutic nihilism among health professionals growing out of frustration with prior failures [5]. All of these factors contribute to lack of strong mandates to aggressively treat obesity in persons with diabetes.
The Evidence Report for Treatment of Obesity in Adults [87] outlines these approaches to treatment of obesity: dietary therapy, physical activity, behavioral therapy, combined therapy (to include the first three modalities), pharmacotherapy and surgery. Most individuals with diabetes receive nutrition counseling from a dietitian or nutrition counselor. Many are instructed in low-energy-diets (LEDs) and encouraged to increase their physical activity. These diets are moderately successful for non-diabetic individuals with weight losses at one year averaging about 7.2% of initial body weight [88]. Less information is available for weights after one year for diabetic individuals. The data that we analyzed (Table 2) suggest that structured interventions with LEDs or VLEDs enable diabetic individuals to maintain weight losses of 10% of initial body weight at one year. Weight losses in this range offer significant benefits—i.e., reductions in fasting plasma glucose values of 25% and significant reductions in serum triglycerides—but they often do not lead to "desirable" or non-obese weights. Clearly, optimal cardiovascular-protective effects appear related to achieving non-obese weights.
Our observations suggest that individuals with diabetes should be counseled by their health care providers, especially physicians and dietitians, to achieve non-obese body weights (i.e., BMI 25 kg/m2) [30,81,82]. This suggestion reflects consideration of achieving desirable glycemic control, serum lipoprotein values and blood pressure as well as reducing risk for CHD. Health professionals should ensure that diabetic individuals understand specific health disadvantages of being obese and that losing weight may be the most important thing they can do to preserve health and longevity. When diabetic individuals appreciate the vital importance of weight loss, they may ask for help in becoming successful in weight management. For persons with BMI values of 25–30 kg/m2, dietetic counseling and a structured LED may be the most cost-effective approach. For persons with BMI values >30 kg/m2, medically supervised LEDs may be the most effective approach [88,89]. Intensive medically supervised LEDs or VLEDs are associated with significantly larger weight losses and about threefold better maintenance of weight loss at five years than less intensive weight loss interventions [88,90]. Higher carbohydrate, higher fiber diets, lower fat diets provide glycemic, lipidemic and weight management advantages for long-term management of obese individuals with Type 2 diabetes [91,92].
Conclusions
Obesity and weight gain are major contributors to development of diabetes and CHD. The risk of diabetes for obese adults can be more than ninetyfold the risk for slender adults. Obese women can have up to a sixfold higher risk for CHD than slender women. Small increases in physical activity and sustained weight losses of 5% of initial body weight can reduce risk for developing diabetes by 58%. Intensive weight loss programs enable obese diabetic individuals to lose 7% to 10% of initial body weights and reduce fasting plasma glucose values by >25% with significant improvements in serum lipoproteins and blood pressure. Optimal diets for improvement of glycemia, lipidemia, and weight and for maintenance of these improvements appear to be higher carbohydrate, higher fiber, lower fat diets. This review leads us to conclude that for many obese diabetic individuals, an emphasis on weight management may be the most important therapeutic task.
|
FOOTNOTES |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONREFERENCES |
|---|
Received July 11, 2002. Accepted March 21, 2003.
|
REFERENCES |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. A. Kothare, H. Linnebjerg, Y. Isaka, K. Uenaka, A. Yamamura, K. P. Yeo, A. de la Pena, C. H. Teng, K. Mace, M. Fineman, et al. Pharmacokinetics, Pharmacodynamics, Tolerability, and Safety of Exenatide in Japanese Patients With Type 2 Diabetes Mellitus J. Clin. Pharmacol., December 1, 2008; 48(12): 1389 - 1399. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. C. Feldstein, G. A. Nichols, D. H. Smith, V. J. Stevens, K. Bachman, A. G. Rosales, and N. Perrin Weight Change in Diabetes and Glycemic and Blood Pressure Control Diabetes Care, October 1, 2008; 31(10): 1960 - 1965. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. F Van Gaal, S. W Gutkin, and M. A Nauck Exploiting the antidiabetic properties of incretins to treat type 2 diabetes mellitus: glucagon-like peptide 1 receptor agonists or insulin for patients with inadequate glycemic control? Eur. J. Endocrinol., June 1, 2008; 158(6): 773 - 784. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Janiszewski, T. J. Saunders, and R. Ross Themed Review: Lifestyle Treatment of the Metabolic Syndrome American Journal of Lifestyle Medicine, April 1, 2008; 2(2): 99 - 108. [Abstract] [PDF]
|
|
|
|
 |
|
 F. Brighenti Dietary Fructans and Serum Triacylglycerols: A Meta-Analysis of Randomized Controlled Trials J. Nutr., November 1, 2007; 137(11): 2552S - 2556S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Hollander Anti-Diabetes and Anti-Obesity Medications: Effects on Weight in People With Diabetes Diabetes Spectr, July 1, 2007; 20(3): 159 - 165. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Carver Insulin Treatment and the Problem of Weight Gain in Type 2 Diabetes The Diabetes Educator, November 1, 2006; 32(6): 910 - 917. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W Anderson, S. M Schwartz, J. Hauptman, M. Boldrin, M. Rossi, V. Bansal, and C. A Hale Low-Dose Orlistat Effects on Body Weight of Mildly to Moderately Overweight Individuals: A 16 Week, Double-Blind, Placebo-Controlled Trial Ann. Pharmacother., October 1, 2006; 40(10): 1717 - 1723. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Bonkowski, R. W. Pamenter, J. S. Rocha, M. M. Masternak, J. A. Panici, and A. Bartke Long-lived growth hormone receptor knockout mice show a delay in age-related changes of body composition and bone characteristics. J. Gerontol. A Biol. Sci. Med. Sci., June 1, 2006; 61(6): 562 - 567. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. A. Bazzano, M. Serdula, and S. Liu Prevention of Type 2 Diabetes by Diet and Lifestyle Modification J. Am. Coll. Nutr., October 1, 2005; 24(5): 310 - 319. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Prevalence of Overweight and Obesity Among Adults With Diagnosed Diabetes--United States, 1988-1994 and 1999-2002 JAMA, February 2, 2005; 293(5): 546 - 547. [Full Text] [PDF]
|
|
|
|
 |
|
 R. Khan Weight gain and insulin therapy The British Journal of Diabetes & Vascular Disease, July 1, 2004; 4(4): 264 - 267. [Abstract] [PDF]
|
|
|
|
|
|
J. W. Anderson, K. M. Randles, C. W. C. Kendall, and D. J. A. Jenkins Carbohydrate and Fiber Recommendations for Individuals with Diabetes: A Quantitative Assessment and Meta-Analysis of the Evidence J. Am. Coll. Nutr., February 1, 2004; 23(1): 5 - 17. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||
