The Obesity Paradox and Diabetes

Home , Obesity paradox

In Brief The obesity paradox (survival advantage in overweight/obese patients with FROM RESEARCH TO PRACTICE/FOUR OPPORTUNITIES TO PREVENT DIABETES AND CORONARY HEART DISEASE type 2 diabetes) has called into question the importance of weight loss in overweight people with diabetes. A systematic review of weight loss studies with a minimum of 1-year outcomes in people with diabetes reported inconsistent beneficial effects of weight loss on A1C, lipids, and blood pressure. To lower the risk of cardiovascular disease, a better nutrition therapy intervention may be reducing energy intake, which may or may not lead to weight loss, and selecting cardioprotective foods in appropriate portion sizes. However, any nutrition therapy intervention must be based on lifestyle changes the person with diabetes is willing and able to make.

The Obesity Paradox and Diabetes

It is being called the “obesity para- to heavier patients, cardioprotective dox”: research findings that people metabolic effects of increased body Marion J. Franz, MS, RD, CDE with obesity-related illnesses and who fat, and/or beneficial effects of higher are overweight or obese have better metabolic reserves.1 outcomes, including less mortality, A U-shaped association of weight than their normal-weight peers.1 These with mortality is reported in people seemingly contradictory results have with diabetes. In a recent study,4 the also been shown to apply to people records of 106,640 people with type with acute coronary syndrome, stroke, 2 diabetes in Scotland were reviewed and diabetes.2–9 The consequences of and BMI recorded around the time obesity are clear: increased risk for of diagnosis and mortality through- diabetes, high blood pressure, heart out the next ~ 5 years was assessed. disease, stroke, and kidney disease.10 Mortality risk was higher in people So, the question of emerging interest with a BMI of 20 to < 25 kg/m2 and becomes why, then, once the disease in those with a BMI ≥ 35 kg/m2. develops, does being overweight Vascular mortality was also higher appear to be beneficial? for each 5-kg/m2 increase in BMI > Flegal et al.,1 in a systematic review 30 kg/m2 but was lower below this and meta-analysis, reported that, in level. Another study5 assessed the the general public, severe obesity was relationships between BMI and all- associated with an increased risk for cause mortality in African-American death from all causes but that lesser and white men with type 2 diabetes amounts of excess weight either did and observed a significantly higher not increase this risk or were pro- mortality risk (70%) in those with tective. They concluded, “. . . excess a BMI within the normal range mortality in obesity may predomi- (18.5–24.9 kg/m2) than in heavier nantly be due to elevated mortality at subjects, with a higher mortality rate higher body mass index (BMI) levels. in African Americans (95%) than in Overweight was associated with sig- whites (53%). In a study of patients nificantly lower all-cause mortality.” who developed diabetes,6 total, car- They further noted that these results diovascular, and noncardiovascular are consistent with previous findings mortality rates were higher among that have also shown lower mortal- normal-weight than among obese ity among overweight and moderately subjects. This finding was true regard- obese patients. Possible explanations less of diabetes type. Two other include earlier medical care and recent studies, Translating Research aggressive risk factor treatment given into Action for Diabetes7 and the Diabetes Spectrum Volume 26, Number 3, 2013 145 PROactive trial,8 also observed that Weight Loss and Diabetes-Related kg.24 In the trials reporting data col- participants who were of normal Outcomes lected for > 1 year, none of the average weight at the baseline examination A PubMed search was conducted to weight losses were back to baseline by or who lost weight during the trial determine the outcomes from base- the end of the studies. (PROactive) experienced higher mor- line to study completion of nutrition tality than participants who were therapy weight loss interventions in A1C results All of the studies reported the effect overweight or obese. This obesity overweight or obese adults with type 2 of weight loss on 1-year A1C values. paradox (survival advantage in obese diabetes. Research was reviewed from Improvements in A1C were reported patients with type 2 diabetes) was 1 January 2000 to 1 February 2013. Eleven randomized, clinical trials from eight of the weight loss inter- also shown to exist in patients with ventions.15,18,21,22,25 However, one of 9 with a completion rate of ≥ 70% and diabetes and cancer. Similar findings 18 12-month laboratory data reported the trials extending to 18 months had been reported in two earlier stud- reported significant improvements 11,12 were identified and are included in ies. All of the above brings into Table 1.15–27 Because the duration of in A1C at 12 months that were not question the role of weight manage- trials affects their outcomes, and to maintained to 18 months. The MED ment in people with chronic diseases, better compare outcomes among stud- reported the largest improvement in especially in people with diabetes. ies, 1-year outcomes are presented in A1C at 1 year, –1.2%,21 and the ILI the table for all trials. However, also reported the second largest, –0.64%.25 Weight Loss/Management: included are data from two trials Significant improvements in A1C were Prevention Versus Treatment reporting 18-month data,18,19 two with also reported from the use of meal of Diabetes 2-year data,23,24 and two with 4-year replacements15 and one low-fat,21 one The research results described above data.21,25–27 One trial21 did not report high-protein,22 and one high-CHO22 suggest that, perhaps, guidelines rec- the statistical significance of interven- study. At 4 years, the MED and ILI ommending weight loss should apply tion changes from baseline to study reported continued improvements primarily to prevention and not neces- end, and one trial25 reported statistical in A1C, –0.9 and –0.36%, respec- sarily to treatment of chronic diseases, significance for the intervention arm tively.21,27 Perhaps of equal importance including diabetes. Strong evidence only; however, both trials are included is that nonsignificant changes in A1C exists for the benefits of moderate in the intervention summaries. were reported from 11 weight loss 16,17,19,20,23,24 weight loss for the prevention of type Eight of the studies compared interventions at 1 year and 18 2 diabetes.13 The question becomes, varying weight loss interventions (10 from one study at 18 months. what are the benefits from weight loss different interventions),17–24 and three Also of interest, five trials compared weight loss interventions with as glycemic impairments progress from studies compared the intervention to 15,16,25,27 differing macronutrient percentages prediabetes to overt type 2 diabetes? usual care or to a control group. (high-MUFA vs. high-CHO,19 low- The goals of medical nutrition Weight loss interventions imple- mented in 19 study arms included CHO vs. low-fat,20 high-protein therapy (MNT) for individuals with 22,23 meal replacements (2 studies),15,17 indi- vs. high-CHO, and low-CHO diabetes include achieving and main- 24 vidualized food plans (2 studies),16,17 vs. low-fat ). All five reported that taining blood glucose levels in the one study with two group behavioral weight changes did not differ sta- normal range or as close to normal weight management arms,18 low fat tistically between arms, and weight as is safely possible, a lipid and lipo- (3 studies), 20,21,24 high monounsatu- losses ranged from 1.9 to 4.0 kg. protein profile that reduces the risk rated fat (MUFA),19 high carbohydrate Furthermore, eight of the interven- for cardiovascular disease (CVD), and (CHO) (3 studies),19,22,23 low CHO tion arms reported nonsignificant blood pressure levels in the normal (2 studies), 20,24 high protein (2 stud- changes in A1C from baseline to study range or as close to normal as is safely ies), 22,23 Mediterranean-style diet end,19,20,23,24 and only two intervention possible.13 To achieve these goals, (MED),21 and intensive lifestyle inter- arms (in one study) reported signifi- weight loss has been recommended vention (ILI).25–27 Although physical cant but modest changes in A1C.22 for all overweight or obese individu- activity was suggested or encouraged These findings support a conclusion als who have diabetes or are at risk in several studies,16,18,19,20,22 only two that a variety of eating patterns with for diabetes, with the level of evidence studies, those testing MED and ILI, differing macronutrient percentages rated A (clear evidence from well-con- included physical activity recommen- are moderately effective for weight loss ducted, generalizable, randomized, dations and measured and reported but may not improve A1C levels. 21,25,27 15,17,23,24 controlled trials that are adequately adherence. Four studies Weight loss interventions reporting improvements in A1C at 1 year had powered).14 But perhaps the benefits did not mention physical activity. weight losses of 4.8, 4.2, 3.0, 2.7, 2.2, of weight loss in the treatment of type Weight changes and 1.0 kg, with the MED (–6.2 kg, 2 diabetes need to be reexamined. An Weight losses from interventions 7.2%) and ILI (8.4 kg, 8.6%) hav- inconclusive picture emerges from ranged from 1.9 to 8.4 kg at 12 ing larger weight losses. Weight loss review of the benefits of weight loss months; 16 of the interventions15–24 interventions reporting no statistically on A1C, lipids, and blood pressure reported weight losses ranging from significant improvement in A1C at 1 in people with type 2 diabetes. This 2.4 to 4.8 kg. The MED and ILI year had similar weight losses rang- also brings into question the effects interventions reported the largest ing from 1.9 to 4.4 kg. Although there of weight loss on risk factors for the weight loss at 1 year, 6.2 and 8.4 is overlap in the weight loss effect prevention and treatment of CVD in kg, respectively,21,25 and a low-CHO on A1C, it does appear that larger people with diabetes. intervention reported the smallest, 1.9 amounts of weight lost are more likely 146 Diabetes Spectrum Volume 26, Number 3, 2013 to improve A1C; weight losses of ~ style intervention resulting in weight provided by a registered dietitian (or

7–8% are needed. loss would reduce rates of heart nutritionist) independently or as part FROM RESEARCH TO PRACTICE/FOUR OPPORTUNITIES TO PREVENT DIABETES AND CORONARY HEART DISEASE disease, stroke, and CVD deaths in of an overall diabetes self-manage- Lipid levels overweight and obese people with ment education (DSME) program. Ten of the trials (11 weight loss inter- type 2 diabetes, a group at increased Although weight loss is sometimes vention arms) measured lipid levels. risk for such events. Half of the 5,145 reported, it is not the primary goal of The most consistently reported posi- people enrolled in the study were the nutrition therapy interventions. tive change was in HDL cholesterol. randomly assigned to receive an ILI, Eleven studies published after 2000 However, all but the ILI and MED and the other half were assigned to reported improvements in A1C with reported nonsignificant changes in a control group that received a gen- independent MNT interventions, and various lipids as a result of weight eral program of diabetes support and seven studies showed improvement loss. HDL was reported for 17 of the education (DSE). Both groups received when MNT was part of DSME.29 interventions; 10 reported positive routine medical care from their own Randomized, clinical studies and changes in HDL,15,17,19,20–22,24,25 and health care providers. Participants observational outcome studies docu- 7 reported nonsignificant changes. randomized to the ILI received meal mented decreases in A1C of ~ 1–2% Triglycerides were also reported for replacements or structured food plans, (range –0.5 to –2.6%), depending on 17 interventions; 6 reported lower- were encouraged to achieve 175 min- the type and duration of diabetes and ing of triglycerides,17,21,22,25 and 11 utes of physical activity per week, and the baseline A1C value. reported nonsignificant changes. attended three to four education/coun- Of interest are the types of nutri- LDL cholesterol was reported from seling sessions per month. At 4 years, tion therapy interventions that are 15 interventions, with only the ILI participants in the ILI group aver- most effective. Interventions for reporting a positive change.25 Total aged a weight loss of 4.7 kg compared people with type 2 diabetes include cholesterol was reported for 16 inter- to 1.1 kg in the control group.27 reduced energy/fat intake, individu- ventions, and only the MED resulted In September 2012, the National alized MNT, portion control and in a positive change.21 Institutes of Health stopped the ILI healthy food choices, and carbo- group early, acting on the recommen- hydrate counting, and for people Blood pressure dation of the study’s data and safety with type 1 diabetes, carbohydrate Eight studies (14 weight loss interven- monitoring board.28 The independent counting and matching insulin to car- tions) reported the effect of weight advisory board found that the ILI did bohydrate intake. Although it is clear loss on blood pressure. Five stud- no harm but was not on a trajectory that there is not one nutrition therapy ies reported positive blood pressure that would result in greater decreases intervention that applies to all people changes,15,19,21,24,25 and three studies in cardiovascular events compared to with diabetes, a consistent theme for reported no changes.20,22,23 the control group. The board recom- individuals with type 2 diabetes is that mended continuing to follow all Look a reduced energy intake, which may Medication-related effects AHEAD participants to identify other or may not lead to substantial weight Weight loss may also have an effect potential longer-term effects of the loss, consistently improves glycemic on dosage of anti-diabetes, lipid, and intervention. control. blood pressure medications. However, The benefits on A1C, HDL choles- An eating pattern designed to both changes in these medications can also terol, triglycerides, and blood pressure lower glucose and improve lipids and confound results of the nutrition were significantly greater in the ILI blood pressure, along with regular therapy intervention. Limited data group compared to the DSE group physical activity, is the cornerstone are available on medications taken after 4 years (weight P < 0.001, A1C of diabetes care.30,31 A cardioprotec- at baseline and whether medication P < 0.001, HDL cholesterol P < 0.001, tive eating pattern provides 25–35% adjustments were made as a result triglycerides P < 0.001, systolic blood of calories from fat, with < 7% of of weight loss. Four trials did not pressure P < 0.001, and diastolic blood calories from saturated and trans report on medication changes;15,18,19,23 pressure P = 0.01), but reductions in fatty acids. The majority of the fat two studies reported no change in LDL cholesterol were greater in the intake is from unsaturated fatty acids. medications in one study arm.17,22 DSE group (P = 0.009) because of Cholesterol intake is ideally < 200 General decreases in medications more aggressive use of medications to mg/day. Evidence indicates that this were reported in seven studies from lower lipid levels in the this group.27 It type of eating pattern can reduce weight loss interventions.16,17,20–22,24,25 is encouraging and important to note total cholesterol by 7–21%, LDL cho- Only one study reported an increase in that both groups had a lower number lesterol by 7–22%, and triglycerides medications and that was in the Look of cardiovascular events compared by 11–31%. 32 In addition, controlling AHEAD trial control group for lipids to previous studies of people with sodium intake to 2,400 mg/day has and is noted in the next section.27 diabetes.27 an approximate systolic blood pressure–lowering range of 2–8 mmHg.31 Look AHEAD (Action for Health in MNT for Diabetes Although the ILI in the Look Diabetes) Trial So, if weight loss is not the complete AHEAD trial resulted in weight loss Because of its size and duration, it is nutrition therapy answer for improve- and improved A1C, some lipids, and important to summarize the Look ments in cardiometabolic outcomes blood pressure, it did not improve AHEAD trial.25–27 The trial, con- for diabetes, what other interventions these risk factors enough to result in ducted in 16 centers in the United have been reported to be beneficial? better CVD protection than standard States and planned to last 11.5 years, A systematic review was conducted to diabetes (education/medical) care. was designed to test whether a life- determine the effectiveness of MNT Equally important, it is not clear how Diabetes Spectrum Volume 26, Number 3, 2013 147

0.001) 0.001) 0.01) 0.001) 0.01) 3.0 6.2 < < 11.6 1.9 9.4 1.4 0.2 0.2 5.6 5.6 < = 0.001) < = 0.002) 0.3 1.9 0.4 = 0.004) < 15.6 12.6 2.6 22.6 1.0 4.2 = 0.003) ± ± = 0.012) ± 13.2 ± ± ± ± ± ± ± ± ± ± ± ± 1.0 P ± P ± ± P P P 0.0001) 0.001) 0.001) P ± P P P ± P < < <

4 ( 0.4 (all NS) (all 0.4 0.5 (all NS) (all 0.5 6 ( 8 ( 4.0 3.8 0.3 0.1 0.2 6 ( 2.5 (2.8–2.2) 2.5 .9 1.8 1.5 2.3 5 ( 5.1 5 ( 3.0 2.9 (3.2–2.7) 2.9 2.9 0.5 0.8 NS) (all 0.7 1.0 0.2 6.8 P P P 2.8 2.0 3.0 (3.4–2.5) 8 ( 8.8 2.0 11 ( 1.0 10 ( 1.8 1 (NS) 1 5.3 (5.8–4.96) 5.0 5.1 2.5 2 (NS) 2 9 ( 2.0 ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ 1.0 ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↑ ↓ (both (all NS) (all (both (both Blood Pressure (mmHg) 1. SBP 1. DBP 1. 2. SBP 2. DBP 1. SBP 1. DBP 1. 2. SPB 2. DBP 1. SBP 1. DBP 1. 2. SBP 2. DBP 1. SBP 1. 1. DBP 1. 2. SBP 2. DBP 1. DBP 1. 2. SBP 2. DBP SBP 1. 1. SBP 1. 2. DBP 1. DBP 1. 2. SBP 1.SBP DBP 1. 2. SBP 2. DBP 1. SBP 1. 1. DBP 1. 2. SBP 2. SBP 2. DBP 1. SBP 1. DBP 1. 2. SBP 2. SBP 2. BP 1. SBP 1. DBP 1. 1. DBP 1. 2. SBP 2. DBP Not reported 1. SBP 1. No significant changes from 12 months 1. DBP 1. 2. SBP 2. DBP 1. SBP 1. Not reported Not reported

13.3

1.0,

0.001) 0.001) 0.2, ↓ ↓ < < ± 3.7 = 0.017), 4.6, TG TG 4.6, 2.0 0.1, = 0.020), ↑ 1.0 P P 8.9 (NS) 8.9 17.7 (NS) 17.7 126 ± P 26.7, 7.7 (NS), (NS), 7.7 24.6, ↑ 0.8, ± 0.8, TG 0.8, ↑ 0.1, ↓ ↓ P ± 3.1, ± 274.0 ↓ ± 6.6 (NS), (NS), 6.6 3.4 ↑ ± ↑ 28 (all NS); NS); (all 28 ± ± ± ↑ ↓ ↓ 3.9 5, TG ↓ TG 5, 1.4 7.0 0.0001) ↑ 1.5, TG 1.5, TG ↑ 0.3 41.6 ( 7.0, TG TG 7.0, ↑ 8.9 (NS) 8.9 1.0 26.6 ( 26.6 (NS) 26.6 11.6 ( = 0.002) 1.0 3.5 (NS) 3.5 0.9 (NS) 0.9 14.2 5.1 < 1 ( all NS); ↑ 3.5 ↓ 0.01) 1.5 (NS), HDL ↓ ↓ 11.6 ( ↓ ↓ 1.9 (NS), HDL1.9 ↑ ↓ ↓ ↓ 7.7 (NS), HDL 7.7 ↑ ↓ ↓ ↓ ↑ P < ↑ ↓ 3.9 (NS), HDL ↓ ↓ P = 0.038) ↓

↓ P 19.8 (22.11–23.21) P 25.6 (27.9–23.2) (27.9–23.2) 25.6 6.1, HDL6.1, ↓ ↓ 8.8, TG 3.1, HDL 3.1, 2.3, HDL 6.2 ( 5 ( ↓ ↑ 0.001), TG TG 0.001), ↓ ↓ 3, HDL ↑ ↑ 3. TG ↓ TG 3. 28 ( = 0.012) < = 0.050), TG ↓ = 0.02), LDL = 0.02), 0.01) ↓ ↓ 5.2, TG 5.7, TG TG 5.7, P 0.6, HDL 28.63, HDL P < 2.0 (all = 0.02), LDL = 0.02), P 24.6 Lipids (mg/dl) 1.8 14.7, HDL 14.7, P ± ± 8.9 ± ± ± 7.9, HDL 7.9, 6.6, HDL 27.1, LDL 27.1, ± 41.7, LDL 41.7, 29.4, LDL 29.4, 29.2, LDL ± 7.7, HDL 7.7, ± P P ± ± ± ± ± ± ± ± 40.0 = 0.024), TG TG = 0.024), = 0.008), TG = 0.008), TG = 0.029), TG TG = 0.029), 50.5 0.4 (NS), TG 0.4 2.3 (all NS) 2.3 (all 2.3 ( 8.9 ( 0.7 (NS), TG 0.7 0.3 1.9 5 ( 4.3 ( 5.7 9.2 (10.0–8.4), HDL (10.0–8.4), 9.2 8.8 (12.1–10.4), HDL 8.8 (12.1–10.4), 5.2 ± ± ± ↑ ↓ ↑ ↑ ↑ ↑ ↑ ↑ ↑ P P P P ↓ ↓ 0.4 (NS), LDL0.4 2, LDL 6.2 ↓ ↓ 5.0 5.8 5.3, LDL 3.9 2.7, LDL 2.7, 1.0 1.8 (all NS) (all 1.8 30.3 12.8 ( 11.6 (NS), LDL 11.6 3.9 24.8 9.7 (all NS) (all 9.7 5.8 (NS),5.8 LDL 3.9 (NS), LDL 3.9, LDL 3.9, 9.2 ( 9.7 14.6 7.7 (NS), LDL 7.7 5. LDL5. 6.2 15.1 10.7, LDL 10.7, ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↑ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↑ ↑ 3.1 ( 3.1 ( 3.1 ( 39.0 4.3 ( 19.4 (3.43–3.91), TG TG (3.43–3.91), TG (1.73–2.22), TG TG ↑ HDL TG HDL HDL TG ↑ TG ↑ ↑ HDL TG ↓ ↓ (all NS); HDL NS); (all HDL HDL (all NS); TG TG NS); (all HDL NS) (all HDL NS) (all HDL (all NS); HDL NS); (all 1. LDL1. 2. LDL 2. LDL 1. LDL1. 2. TC 2. TC 2. TC 1. TC TC 1. 1. TC TC 1. 2. TC 2. 0.0 TC (NS), LDL 1. TC TC 1. TC 1. 1. TC TC 1. 1. TC TC 1. 2. TC 2. TC 2. TC TC, LDL, HDL, TG TC 1. (both groups NS changes) TC 1. 1. TC TC 1. No significant changes from 12 months 1. TC TC 1. 2. TC 2. TC Not reported 2. TC 1. TC TC 1.

1.4 1.02 1.52 0.89

1.30 0.4 0.6

0.01 ± ± ± 0.1 0.1 ± 0.6 1.0 0.1 ± ± ± ± ± ± ± ± ± 0.0001) 0.001) 0.001) 0.001) 0.001) 0.0001) 0.0001) < < < < < < <

0.09 0.14 0.28 0.5 0.20 (NS) 0.20 0.09 (NS) 0.09 0.6 0.15 (NS) 0.15 0.1 0.4 0.2 0.1 (NS) 0.1 0.1 (NS) 0.1 (NS) 0.2 0.24 0.36 0.64 0.2 (NS) 0.2 0.23 0.9 0.18 (NS) 0.18 0.1 (NS) 0.1 0.02 0.30 (NS) 0.30 0.6 0.24 A1C (%) 0.1 (NS) 0.1 0.1 (NS) 0.1 1.2 0.2 P P P P P P P ↓ ↓ ↑ ↓ ↑ ↑ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ (0.4–0.33) (0.13–0.06) ( ( ( (NS) (NS) (NS) (NS) ( ( ( ( (NS) ↓ 1. 2. 1. 2. 1. 2. 1. 2. 1. 2. 0.0 (NS)1. 2. 1. 2. 1. 2. 1. 2. 1. 2. 1. 2. 1. 2. No significant change from 12 months 1. 2. 1. 2. 1. 2. 1. 2. No change 1 Year Outcomes 1 Year Also Reported) > 0.01) 2.2) 0.6) 0.01) < ↑ = 0.038) < ↓ = 0.005) = 0.005) P P P P

0.001) 0.001) 0.001) 0.001) 0.001) 0.001) 0.001) 0.001) < < = 0.002) < < < = 0.020) < < < 1.0 to to 1.0 4.1 to to 4.1 0.8 ( 1.9 0.6 P 3.2 4.8 1.9 0.6 0.8 ( 0.6 ( 0.8 0.2 6.9 5.8 ( P 2.0 5.4 3.8 0.6 0.2 4.8 ( 0.6 P ↑ P P P P P P P ↓ ± P ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.0001) 0.0001) < <

Weight LossWeight (kg) 4.2 4.8 0.7 2.4 ( 4.7 4.4 0.6 ( (Five Trials With Trials (Five 3.0 ( 3.9 ( 3.2 3.1 3.5 3.8 6.2 2.9 ( 2.2 ( 2.4 8.7 4.0 1.1 1.0 3.9 ( 3.8 3.2 ( 3.1 3.0 2.2 ( 2.0 ( 2.4 ( 2.7 1.9 ( 1.7 P P ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ( ( 1. 2. 2. 1. 1. 2. 1. 1. 2. 1. 1. 2. 2. 2. 1. 2. 2. 1. 2. 1. 1. 1. 2. 2. No significant change from 12 months 1. 2. 2. 1. 1. 2. 1. 2. -

- -

Interventions intervention (ILI) support/ education (control) carbohydrate carbohydrate carbohydrate style diet (MED) management (individualized plan) food replacement plan food carbohydrate plan (meal replacements) energy(reduced intake) ioral weight management ioral weight manage saturated fat carbohydrate ment plus motivational interviewing 1. Intensive1. lifestyle 2. Diabetes 1. High1. protein 2. High High1. protein 2. High 1. Low 2. Low fat 1. Mediterranean-1. 2. Low fat 1. Case Case 1. 2. Usual care 1. Soy-based meal 2. Individualized 1. Low 2. Low fat 1. Prepared1. meal 2. Usual care 1. Group behav 2. behav Group High1. monoun 2. High Table 1. Diabetes 1. Weight Outcomes Loss Trials: Table of Interventions Compared at 1 Year Baseline to Values

of n [ n ( Subjects Enrolled completers, completers]) percentage of 5,145 (4,815, 5,145 4 at 94% years) 5,145 (4,959, 1 at 96% year) 419 (294, (294, 419 24 at 70% months) 61 (47, 77% (47, 61 months) 24 at 108 (99, 92% 92% (99, 108 1 year) at 419 (310, 12 at 74% months) 89% (54, 61 12 months) at 215 (195, (195, 215 4 at 91% years) 215 (195, (195, 215 1 at 91% year) 144 (115, 80%) 104 (82, 79%) 81%) (85, 105 57 67% (38, at 18-month extension) 217 (202, 18 at 90% months) 119 (92, 77%) (195, 217 12 at 90% months) 77% (95, 124 12 months) at 15 18 16 ** 17 * 25–27 22 23 20 21 24 19 Study *Statistical significance from baseline not reported. **Statistical significance from baseline only reported for 1-year ILI. diastolicDBP, blood pressure; HDL, HDL cholesterol; LDL, LDL cholesterol; NS, not significant; physical PA, activity; systolic triglycerides. SBP, TG, blood pressure; cholesterol; TC, total Look AHEAD Research Group Guldbrand Guldbrand et al. Larsen et al. Krebs et al. Esposito Esposito et al. Li et al. Wolf et al. Wolf Davis et al. Brehm et al. West et al.West Metz et al.

148 Diabetes Spectrum Volume 26, Number 3, 2013 FROM RESEARCH TO PRACTICE/FOUR OPPORTUNITIES TO PREVENT DIABETES AND CORONARY HEART DISEASE

4 ( 0.4 (all NS) (all 0.4 0.5 (all NS) (all 0.5 6 ( 8 ( 4.0 3.8 0.3 0.1 0.2 6 ( 2.5 (2.8–2.2) 2.5 .9 1.8 1.5 2.3 5 ( 5.1 5 ( 3.0 2.9 (3.2–2.7) 2.9 2.9 0.5 0.8 NS) (all 0.7 1.0 0.2 6.8 P P P 2.8 2.0 3.0 (3.4–2.5) 8 ( 8.8 2.0 11 ( 1.0 10 ( 1.8 1 (NS) 1 5.3 (5.8–4.96) 5.0 5.1 2.5 2 (NS) 2 9 ( 2.0 ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ 1.0 ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↑ ↓ (both (all NS) (all (both (both Blood Pressure (mmHg) 1. SBP 1. DBP 1. 2. SBP 2. DBP 1. SBP 1. DBP 1. 2. SPB 2. DBP 1. SBP 1. DBP 1. 2. SBP 2. DBP 1. SBP 1. DBP 1. 2. SBP 2. DBP 2. DBP 1. SBP 1. DBP 1. 2. SBP 2. DBP SBP 1. 1. DBP 1. 2. SBP 1.SBP DBP 1. 2. SBP 2. DBP 1. SBP 1. 1. DBP 1. 2. SBP 2. SBP 2. DBP 1. SBP 1. DBP 1. 2. SBP 2. SBP 2. BP 1. SBP 1. DBP 1. 1. DBP 1. 2. SBP 2. DBP Not reported 1. SBP 1. No significant changes from 12 months 1. DBP 1. 2. SBP 2. DBP 1. SBP 1. Not reported Not reported

13.3

1.0,

↓ P 19.8 (22.11–23.21) P 25.6 (27.9–23.2) (27.9–23.2) 25.6 6.1, HDL6.1, ↓ ↓ 8.8, TG 3.1, HDL 3.1, 2.3, HDL 6.2 ( 5 ( ↓ ↑ 0.001), TG TG 0.001), ↓ ↓ 3, HDL ↑ ↑ 3. TG ↓ TG 3. 28 ( = 0.012) < = 0.050), TG ↓ = 0.02), LDL = 0.02), 0.01) ↓ ↓ 5.2, TG 5.7, TG TG 5.7, P 0.6, HDL 28.63, HDL P < 2.0 (all = 0.02), LDL = 0.02), P 24.6 Lipids (mg/dl) 1.8 14.7, HDL 14.7, P ± ± 8.9 ± ± ± 7.9, HDL 7.9, 6.6, HDL 27.1, LDL 27.1, ± 41.7, LDL 41.7, 29.4, LDL 29.4, 29.2, LDL ± 7.7, HDL 7.7, ± P P ± ± ± ± ± ± ± ± 40.0 = 0.024), TG TG = 0.024), = 0.008), TG = 0.008), TG = 0.029), TG TG = 0.029), 50.5 0.4 (NS), TG 0.4 2.3 (all NS) 2.3 (all 2.3 ( 8.9 ( 0.7 (NS), TG 0.7 0.3 1.9 5 ( 4.3 ( 5.7 9.2 (10.0–8.4), HDL (10.0–8.4), 9.2 8.8 (12.1–10.4), HDL 8.8 (12.1–10.4), 5.2 ± ± ± ↑ ↓ ↑ ↑ ↑ ↑ ↑ ↑ ↑ P P P P ↓ ↓ 0.4 (NS), LDL0.4 2, LDL 6.2 ↓ ↓ 5.0 5.8 5.3, LDL 3.9 2.7, LDL 2.7, 1.0 1.8 (all NS) (all 1.8 30.3 12.8 ( 11.6 (NS), LDL 11.6 3.9 24.8 9.7 (all NS) (all 9.7 5.8 (NS),5.8 LDL 3.9 (NS), LDL 3.9, LDL 3.9, 9.2 ( 9.7 14.6 7.7 (NS), LDL 7.7 5. LDL5. 6.2 15.1 10.7, LDL 10.7, ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↑ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↑ ↑ 3.1 ( 3.1 ( 3.1 ( 39.0 4.3 ( 19.4 (3.43–3.91), TG TG (3.43–3.91), TG (1.73–2.22), TG TG ↑ HDL TG HDL HDL ↑ TG ↑ HDL TG ↑ TG ↓ ↓ (all NS); HDL NS); (all HDL HDL (all NS); TG TG NS); (all HDL NS) (all HDL NS) (all HDL (all NS); HDL NS); (all 1. LDL1. 2. LDL 2. LDL 1. LDL1. 2. TC 2. TC 2. TC 1. TC TC 1. 1. TC TC 1. 2. TC 2. 0.0 TC (NS), LDL 1. TC TC 1. TC 1. 1. TC TC 1. 1. TC TC 1. 2. TC 2. TC 2. TC TC, LDL, HDL, TG TC 1. (both groups NS changes) TC 1. 1. TC TC 1. No significant changes from 12 months 1. TC TC 1. 2. TC 2. TC Not reported 2. TC 1. TC TC 1.

1.4 1.02 1.52 0.89

1.30 0.4 0.6

0.01 ± ± ± 0.1 0.1 ± 0.6 1.0 0.1 ± ± ± ± ± ± ± ± ± 0.0001) 0.001) 0.001) 0.001) 0.001) 0.0001) 0.0001) < < < < < < <

0.09 0.14 0.28 0.5 0.20 (NS) 0.20 0.09 (NS) 0.09 0.6 0.15 (NS) 0.15 0.1 0.4 0.2 0.2 (NS) 0.2 0.1 (NS) 0.1 (NS) 0.1 0.24 0.36 0.64 0.2 (NS) 0.2 0.23 0.9 0.18 (NS) 0.18 0.1 (NS) 0.1 0.02 0.30 (NS) 0.30 0.6 0.24 A1C (%) 0.1 (NS) 0.1 0.1 (NS) 0.1 1.2 0.2 P P P P P P P ↓ ↓ ↑ ↓ ↑ ↑ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ (0.4–0.33) (0.13–0.06) ( ( ( (NS) (NS) (NS) (NS) ( ( ( ( (NS) ↓ 1. 2. 1. 2. 1. 0.0 (NS)1. 2. 1. 2. 1. 2. 1. 2. 1. 2. 1. 2. 1. 2. 1. 2. 1. 2. 1. 2. No significant change from 12 months 1. 2. 1. 2. 1. 2. 1. 2. No change 1 Year Outcomes 1 Year Also Reported) > 0.01) 2.2) 0.6) 0.01) < ↑ = 0.038) < ↓ = 0.005) = 0.005) P P P P

Weight LossWeight (kg) 4.2 4.8 0.7 2.4 ( 4.7 4.4 0.6 ( (Five Trials With Trials (Five 3.0 ( 3.9 ( 3.2 3.1 3.5 3.8 6.2 2.9 ( 2.2 ( 2.4 8.7 4.0 1.1 1.0 3.9 ( 3.8 3.2 ( 3.1 3.0 2.0 ( 2.2 ( 2.4 ( 2.7 1.9 ( 1.7 P P ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ( ( 1. 2. 2. 1. 1. 2. 1. 1. 2. 1. 2. 1. 2. 2. 1. 2. 2. 1. 2. 1. 1. 1. 2. 2. No significant change from 12 months 1. 2. 2. 1. 1. 2. 1. 2. -

- -

of n [ n ( Subjects Enrolled completers, completers]) percentage of 5,145 (4,815, 5,145 4 at 94% years) 5,145 (4,959, 1 at 96% year) 61 (47, 77% (47, 61 months) 24 at 419 (294, (294, 419 24 at 70% months) 108 (99, 92% 92% (99, 108 1 year) at 419 (310, 12 at 74% months) 89% (54, 61 12 months) at 215 (195, (195, 215 4 at 91% years) 215 (195, (195, 215 1 at 91% year) 144 (115, 80%) 104 (82, 79%) 81%) (85, 105 57 67% (38, at 18-month extension) 217 (202, 18 at 90% months) 119 (92, 77%) (195, 217 12 at 90% months) 77% (95, 124 12 months) at 15 18 16 ** 17 * 25–27 22 23 20 21 24 19 Study *Statistical significance from baseline not reported. **Statistical significance from baseline only reported for 1-year ILI. diastolicDBP, blood pressure; HDL, HDL cholesterol; LDL, LDL cholesterol; NS, not significant; physical PA, activity; systolic triglycerides. SBP, TG, blood pressure; cholesterol; TC, total Look AHEAD Research Group Larsen et al. Guldbrand Guldbrand et al. Krebs et al. Esposito Esposito et al. Li et al. Wolf et al. Wolf Davis et al. Brehm et al. West et al.West Metz et al.

Diabetes Spectrum Volume 26, Number 3, 2013 149 the intervention package might be obesity using standard body mass index cat- 14American Diabetes Association: Standards delivered in a real-world medical set- egories: systematic review and meta-analysis. of medical care in diabetes—2013 (Position JAMA 309:71–82, 2013 Statement). Diabetes Care 36 (Suppl. 1):S11– ting. It involved weekly sessions for S66, 2013 the first 6 months, sessions three times 2Angerås O, Albertsson P, Karason K, 15 per month for the next 6 months, and, Råmanddal T, Matejka G, James S, Lagervist Metz JA, Stern JS, Kris-Etherton P, Reusser B, Rosengren A, Omerovis E: Evidence for ME, Morris CD, Hatton DC, Haynes B, for years 2–4, at least monthly contact obesity paradox in patients with acute coro- Resnick LM, Pi-Sunyer X, Clark S, Chester by telephone or e-mail, as well as a nary syndrome: a report from the Swedish L, McMahon M, Snyder GW, McCarron DA: variety of ancillary group classes in Coronary Angiography and Angioplastry A randomized trial of improved weight loss between contacts.27 Registry. Eur Heart J 34:345–353, 2013 with a prepared meal plan in overweight and obese patients. Arch Intern Med 160:2150– In contrast, the MED (plus physi- 3Doehner W, Schenkel J, Anke S, Springer 2158, 2000 cal activity) intervention, which also J, Audebert H: Overweight and obesity are improved A1C, lipids, and blood pres- associated with improved survival, functional 16Wolf AM, Conaway MR, Crowther JQ, sure, involved monthly sessions for outcomes, and stroke recurrence after acute Hazen KY, Nadler JL, Oneida B, Bovbjerg stroke or transient ischaemic attack: observa- VE: Translating lifestyle intervention to prac- the first year and bimonthly sessions tions from the TEMPiS trial. Eur Heart J tice in obese patients with type 2 diabetes. thereafter. An interesting side note is 34:268–277, 2013 Diabetes Care 27:1570–1576, 2004 that, although the authors described 4Logue J, Walker JJ, Leese G, Lindsay R, 17Li Z, Hong K, Saltsman P, DeShields S, the Mediterranean-style eating plan McKnight J, Morris A, Philip S, Wild S, Bellman M, Thames G, Liu Y, Wang H-J, as low carbohydrate, the actual intent Sattar N, on behalf of the Scottish Diabetes Elashoff R, Heber D: Long-term efficacy of was to have the carbohydrate con- Research Network Epidemiology Group: soy-based meal replacement vs an individual- tent be < 50% of the reduced daily Association between BMI measured within ized diet plan in obese type II DM patients: a year after diagnosis of type 2 diabetes and relative effects on weight loss, metabolic energy, with added fat being 30–50 g mortality. Diabetes Care 36:887–893, 2013 parameters, and C-reactive protein. Eur Jr of olive oil.21 Reported intake at 1 year 5 Clin Nutr 59:411–418, 2005 included an average carbohydrate Kokkinos P, Myers J, Faselies C, Doumas M, Kheifbek R, Nylen E: BMI-mortality 18West DS, DiLillo V, Bursac Z, Gore SA, intake of 43% of calories, protein paradox and fitness in African American Greene PG: Motivational interviewing 28%, polyunsaturated fats 29%, and and Caucasian men with type 2 diabetes. improves weight loss in women with type 2 saturated fats 10%. This could better Diabetes Care 35:1021–1027, 2012 diabetes. Diabetes Care 30:1081–1087, 2007 be described as a moderate-carbohy- 6Carnethon MR, De Chavez PJ, Biggs ML 19Brehm BJ, Lattin BL, Summer SS, Boback drate eating pattern, which is typical Lewis CE, Pankow JS, Bertoni AG, Golden JA, Gilchrist GM, Jandacek RJ, D’Alessio of people with type 2 diabetes.33 SH, Liu K, Mukamal KJ, Campbell-Jenkins DA: One-year comparison of a high-mono- In summary, the most common B, Dyer AR: Association of weight status with unsaturated fat diet with a high-carbohydrate mortality in adults with incident diabetes. diet in type 2 diabetes. Diabetes Care nutrition advice given to people with JAMA 308:581–590, 2012 32:215–220, 2009 type 2 diabetes involves weight loss. 7 20 But weight loss is atypically substan- McEwen LN, Kim C, Karter AJ, Haan Davis NJ, Tomuta N, Schechter C, Isasi MN, Ghosh D, Llantz PM, Mangione CM, CR, Segal-Isaacson CJ, Stein D, Zonszein tial (< 5%) over the long term and, Thompson TJ, Herman WH: Risk factors JZ, Wylie-Rosett J: Comparative study of the even if successful, may not result in for mortality among patients with diabetes. effects of a 1-year dietary intervention of a cardiometabolic outcomes as strong Diabetes Care 30:1736–1741, 2007 low-carbohydrate diet versus a low-fat diet on weight and glycemic control in type 2 diabe- as clinicians and individuals with dia- 8Doehner W, Erdman E, Cairns R, Clark tes. Diabetes Care 32:1147–1152, 2009 betes would like to see. More realistic AL, Dormandy JA, Ferrannini E, Anker SD: and helpful nutrition advice for over- Inverse relation of body weight and weight 21Esposito K, Maiorino MI, Ciotola M, Di weight individuals with diabetes may change with mortality and morbidity in Paol C, Scognamiglio P, Gicchino M, Petrizzo patients with type 2 diabetes and cardio- M, Saccomanno F, Beneduce F, Ceriello A, be to pay less attention to the scale and vascular co-morbidity: an analysis of the Giufliano D: Effects of a Mediterranean-style concentrate more on eating smaller PROactive study population. Int J Cardiol diet on the need for antihyperglycemic drug portions while choosing healthy 162:20–26, 2012 therapy in patients with newly diagnosed type 2 diabetes: a randomized trial. Ann Intern foods such as fruits, vegetables, whole 9Tseng CH: Obesity paradox: differential Med 151:306–314, 2009 grains, legumes, low-fat dairy prod- effects on cancer and noncancer mortality ucts, lean meats, and unsaturated fats in patients with type 2 diabetes mellitus. 22Larsen RN, Mann NJ, Maclean E, Shaw in appropriate portion sizes. Atherosclerosis 226:186–192, 2013 JE: The effect of a high-protein, low-carbohydrate diets in the treatment of type 2 diabetes: To prevent and treat CVD, nutri- 10Centers for Disease Control and Prevention: a 12 month randomized controlled trial. Overweight and obesity: causes and conse- tion therapy for diabetes, instead of Diabetologia 54:731–740, 2011 focusing on weight loss, should focus quences. Available from http://www.cdc.gov/ 23 on 1) nutrition interventions shown obesity/adult/causes/index.html. Accessed 10 Krebs JD, Elley CR, Parry-Strong A, Lunt February 2013 H, Drury PL, Bell DA, Robinson E, Moyes to improve metabolic outcomes (gly- SA, Mann JI: The Diabetes Excess Weight 11Klein R, Klein BE, Moss SE: Is obesity cemia, lipids, and blood pressure), Loss (DEWL) Trial: a randomized controlled related to microvascular and macrovascular trial of high-protein versus high-carbohy- 2) prioritizing goals for individuals, complications in diabetes? The Wisconsin dtate diets over 2 years in type 2 diabetes. 3) negotiating lifestyle changes indi- Epidemiologic Study of Diabetic Retinopathy. Diabetologia 55:905–914, 2012 viduals are willing and able to make, Arch Intern Med 157:650–656, 1997 24Guldbrand H, Dizdar B, Bunjaku B, and 4) assisting patients in choosing 12Ross C, Langer RD, Barrett-Connor E: Lindström T, Bachrach-Lindström M, appropriate portion sizes of foods Given diabetes, is fat better than thin? Fredrikson M, Fredrikson M, Östgren CJ, Diabetes Care 20:650–652, 1997 shown to have health benefits. Nystrom FH: In type 2 diabetes, randomiza- 13American Diabetes Association: Nutrition tion to advice to follow a low-carbohydrate recommendations and interventions for dia- diet transiently improves glycaemic control References betes: a position statement of the American compared with advice to follow a low-fat diet 1Flegal K, Kit B, Graubard B: Association Diabetes Association. Diabetes Care 31 producing similar weight loss. Diabetologia for all-cause mortality with overweight and (Suppl. 1):S61–S78, 2008 55:2118–2127, 2012 150 Diabetes Spectrum Volume 26, Number 3, 2013 25Look AHEAD Research Group: Reduction Available from http://www.nih.gov/news/ Va., American Diabetes Association, 2012, in weight and cardiovascular disease risk health/oct2012/niddk-19.htm. Accessed 13 p. 295–306 FROM RESEARCH TO PRACTICE/FOUR OPPORTUNITIES TO PREVENT DIABETES AND CORONARY HEART DISEASE factors in individuals with type 2 diabetes: February 2013 32Academy of Nutrition and Dietetics: one-year results of the Look AHEAD trial. 29 Disorders of lipid metabolism evidence-based Diabetes Care 30:1374–1383, 2007 Pastors JG, Franz MJ: Effectiveness of medical nutrition therapy in diabetes. In American nutrition practice guidelines update, 2010. 26 Wing RR, Lang W, Wadden TA, Safford M, Diabetes Association Guide to Nutrition Available from http://www.adaevidence Knowler WC, Bertoni AG, Hill JO, Brancati Therapy for Diabetes. 2nd ed. Franz MJ, library.com/topic.ctm?cat=4528. Accessed 20 February 2013 FL, Peters A, Wagenknecht L, The Look Evert AB, Eds. Alexandria, Va., American AHEAD Research Group: Benefits of modest Diabetes Association, 2012, p. 1–18 33Vitolins MZ, Anderson AM, Delahanty weight loss in improving cardiovascular risk 30 L, Raynor H, Miller GD, Mobley C, Reeves factors in overweight and obese individuals Karmally E, Zimmerman JS: Nutrition R, Yamamoto M, Champagne C, Wing with type 2 diabetes. Diabetes Care 34:1481– therapy for diabetes and lipid disorders. RR, Mayer-Davis E, and the Look AHEAD 1486, 2011 In American Diabetes Association Guide Research Group: Action for Health in 27Look AHEAD Research Group: Long-term to Nutrition Therapy for Diabetes. 2nd Diabetes (Look AHEAD) Trial: baseline eval- effects of a lifestyle intervention on weight ed. Franz MJ, Evert AB, Eds. Alexandria uation of selected nutrients and food groups. and cardiovascular risk factors in individuals Va., American Diabetes Association, 2012, J Am Diet Assoc 109:1367–1375, 2009 with type 2 diabetes mellitus: four-year results p. 265–294 of the Look AHEAD trial. Arch Intern Med 31Aebersold K, Ostrovsky N, Wylie-Rosett J: 170:1566–1575, 2010 Nutrition therapy for diabetes and hyper- Marion J. Franz, MS, RD, CDE, 28National Institute of Diabetes and Digestive tension. In American Diabetes Association is a nutrition/health consultant at and Kidney Diseases: Weight loss does not Guide to Nutrition Therapy for Diabetes. Nutrition Concepts by Franz, Inc., lower heart disease risk from type 2 diabetes. 2nd ed. Franz MJ, Evert AB, Eds. Alexandria in Minneapolis, Minn.

Diabetes Spectrum Volume 26, Number 3, 2013 151