Effects of the Dietary Approaches to Stop Hypertension (DASH) Eating Plan on Cardiovascular Risks among Type 2 Diabetic Patients: A Randomized Crossover Clinical Trial

The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters

Citation Azadbakht, Leila, Nafiseh Rashidi Pour Fard, Majid Karimi, Mohammad Hassan Baghaei, Pamela J. Surkan, Majid Rahimi, Ahmad Esmaillzadeh, and Walter C. Willett. 2011. Effects of the Dietary Approaches to Stop Hypertension (DASH) eating plan on cardiovascular risks among type 2 diabetic patients: A randomized crossover clinical trial. Diabetes Care 34(1): 55-57.

Published Version doi:10.2337/dc10-0676

Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:8605323

Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of- use#LAA Clinical Care/Education/Nutrition/Psychosocial Research BRIEF REPORT

Effects of the Dietary Approaches to Stop Hypertension (DASH) Eating Plan on Cardiovascular Risks Among Type 2 Diabetic Patients A randomized crossover clinical trial

1,2 4 LEILA AZADBAKHT, PHD PAMELA J. SURKAN, PHD the Shaheed Motahari Hospital of 3 3 NAFISEH RASHIDI POUR FARD, BSC MAJID RAHIMI, MSC Fooladshahr, Isfahan, during 2009. On 3 1,2 MAJID KARIMI, MD AHMAD ESMAILLZADEH, PHD 3 5,6 the basis of the sample size formula sug- MOHAMMAD HASSAN BAGHAEI, PHD WALTER C. WILLETT, MD, DRPH ϭ ϩ gested for crossover trials (9) n [(Z1Ϫ␣/2 2 ϫ 2 ⌬2 Z1Ϫ␤) S ]/2 , we determined that 21 patients were needed for adequate power. OBJECTIVE — To determine the effects of the Dietary Approaches to Stop Hypertension A diagnosis of type 2 diabetes was (DASH) eating pattern on cardiometabolic risks in type 2 diabetic patients. confirmed if a patient either had a fast- ing plasma glucose Ն126 mg/dl or was RESEARCH DESIGN AND METHODS — A randomized crossover clinical trial was taking oral glucose lowering agents or undertaken in 31 type 2 diabetic patients. For 8 weeks, participants were randomly assigned to insulin (10). Exclusion criteria included a control diet or the DASH eating pattern. any secondary cause of hyperglycemia, RESULTS — After following the DASH eating pattern, body weight (P ϭ 0.007) and waist use of estrogen therapy, untreated hy- circumference (P ϭ 0.002) reduced significantly. Fasting blood glucose levels and A1C de- pothyroidism, smoking, and kidney or creased after adoption of the DASH diet (Ϫ29.4 Ϯ 6.3 mg/dl; P ϭ 0.04 and Ϫ1.7 Ϯ 0.1%; P ϭ liver diseases. Cardiovascular risks such 0.04, respectively). After the DASH diet, the mean change for HDL cholesterol levels was higher as fasting blood glucose, A1C, weight, (4.3 Ϯ 0.9 mg/dl; P ϭ 0.001) and LDL cholesterol was reduced (Ϫ17.2 Ϯ 3.5 mg/dl; P ϭ 0.02). waist circumference, and lipid pro- Additionally, DASH had beneficial effects on systolic (Ϫ13.6 Ϯ 3.5 vs. Ϫ3.1 Ϯ 2.7 mmHg; P ϭ files were the primary outcomes. All 0.02) and diastolic blood pressure (Ϫ9.5 Ϯ 2.6 vs. Ϫ0.7 Ϯ 3.3 mmHg; P ϭ 0.04). participants provided informed written consent. This study was approved by the CONCLUSIONS — Among diabetic patients, the DASH diet had beneficial effects on car- research council and ethics committee of diometabolic risks. the Isfahan University of Medical Sciences Diabetes Care 34:55–57, 2011 (registered in http://www.clinicaltrials.gov; ID number NCT01049321). ardiovascular complications are Effects of the DASH eating pattern in pa- the most frequent problem among tients with metabolic syndrome (4) and hy- Study procedures C type 2 diabetic patients (1). There- pertension (5,6) and other populations We used a randomized crossover de- fore, a therapeutic approach that can (7,8) can be generalized to individuals with sign. After a run-in period of 3 weeks, control cardiometabolic risks might diabetes. patients were randomly assigned to a have beneficial effects for diabetic pa- Therefore, we assessed how the control diet or a DASH diet for 8 weeks. tients (2). DASH eating pattern affects cardiometa- This was followed by a wash-out period Although the Dietary Approaches to bolic risks in type 2 diabetic patients. of 4 weeks. The project dietitian en- Stop Hypertension (DASH) diet was origi- rolled participants and randomly allo- nally developed to prevent or treat high RESEARCH DESIGN AND cated them to groups using random blood pressure (2), it is now recommended METHODS — We enrolled 44 pa- sequencing generated in SPSS at the end as an ideal eating pattern for all adults (3). tients diagnosed with type 2 diabetes at of the run-in period. Because this was a dietary intervention, patients were not ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● blinded. From the 1Food Security Research Center, Isfahan University of Medical Sciences, Isfahan, Iran; the 2Depart- ment of Nutrition, School of Public Health, Isfahan University of Medical Sciences, Isfahan, Iran; 3Shaheed Motahari Hospital, Fooladshahr, Isfahan, Iran; the 4Department of International Health, Johns Hopkins Diets Bloomberg School of Public Health, Baltimore, Maryland; the 5Department of Nutrition, Harvard School We prescribed two diets for each pa- 6 of Public Health, Boston, Massachusetts; and the Department of Epidemiology, Harvard School of Public tient: the control diet and the DASH Health, Boston, Massachusetts. Corresponding author: Leila Azadbakht, [email protected]. diet. The control diet included a macro- Received 8 April 2010 and accepted 3 September 2010. Published ahead of print at http://care.diabetesjournals. nutrient composition of 50–60% car- org on 15 September 2010. DOI: 10.2337/dc10-0676. Clinical trial reg. no. NCT01049321, clinicaltrials.gov. bohydrates, 15–20% protein, Ͻ30% © 2011 by the American Diabetes Association. Readers may use this article as long as the work is properly total fat, and Ͻ5% of caloric intake from cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons. simple sugars (11). This composition org/licenses/by-nc-nd/3.0/ for details. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby was more similar to the Iranian dietary marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. pattern and dietary habits. The DASH

care.diabetesjournals.org DIABETES CARE, VOLUME 34, NUMBER 1, JANUARY 2011 55 DASH and novel cardiovascular risks diet was rich in fruits, vegetables, whole Table 1—Means of the cardiometabolic variables among type 2 diabetic patients after con- grains, low-fat dairy products, and low sumption of the DASH or control diet in saturated fat, total fat, cholesterol, re- fined grains, and sweets. The amount of Control diet* DASH diet† P‡ sodium intake was 2,400 mg per day (3). Patient adherence was assessed in n 31 31 terms of attendance at monthly visits Weight (kg) and through analysis of the 3-day food Baseline 75.0 Ϯ 1.7 73.4 Ϯ 1.8 0.001 diaries. End of trial 72.9 Ϯ 1.8 68.4 Ϯ 1.7 0.001 Change Ϫ2.0 Ϯ 0.3 Ϫ5.0 Ϯ 0.9 0.006 Measurements Waist circumference (cm) All measurements were taken according Baseline 104.6 Ϯ 1.9 103.4 Ϯ 2.0 0.01 to standard protocols. The laboratory End of trial 102.7 Ϯ 2.0 96.6 Ϯ 1.9 0.001 staff was blinded to the treatment Change Ϫ1.9 Ϯ 0.4 Ϫ6.7 Ϯ 1.2 0.002 status. SBP (mmHg) Baseline 137.4 Ϯ 2.8 134.5 Ϯ 3.8 0.39 Statistical analysis End of trial 134.2 Ϯ 3.1 120.8 Ϯ 3.2 0.001 We used general linear models (paired Change Ϫ3.1 Ϯ 2.7 Ϫ13.6 Ϯ 3.5 0.02 Student t tests) to globally compare DBP (mmHg) means of the all variables at the end of the Baseline 81.9 Ϯ 2.2 81.8 Ϯ 1.7 0.95 two different diet periods and the mean End of trial 81.2 Ϯ 2.9 72.2 Ϯ 2.7 0.01 change for each variable in the two Change Ϫ0.7 Ϯ 3.3 Ϫ9.5 Ϯ 2.6 0.04 groups. Statistical analyses were per- FBG (mg/dl) formed using SPSS for Windows version Baseline 171.8 Ϯ 10.9 160.9 Ϯ 10.1 0.73 13.0 (SPSS, Chicago, IL). End of trial 159.0 Ϯ 8.3 131.5 Ϯ 7.3 0.003 Change Ϫ12.8 Ϯ 6.7 Ϫ29.4 Ϯ 6.3 0.04 RESULTS — Of the 44 participants, 31 A1C (%) type 2 diabetic patients (13 male and 18 Baseline 7.9 Ϯ 1.9 7.7 Ϯ 1.9 0.19 female) completed the entire crossover End of trial 7.4 Ϯ 1.7 6.1 Ϯ 0.5 0.05 study (one patient was diagnosed with Change Ϫ0.5 Ϯ 0.02 Ϫ1.7 Ϯ 0.1 0.04 cancer and one with anemia, and eleven TG (mg/dl) patients did not follow the study Baseline 189.7 Ϯ 19.3 170.7 Ϯ 12.4 0.17 protocol). End of trial 178.7 Ϯ 18.4 185.1 Ϯ 13.8 0.53 Analysis of the 3-day diet self-report Change Ϫ10.9 Ϯ 6.8 Ϫ14.4 Ϯ 10.7 0.79 showed that calorie intake of two the HDL-C (mg/dl) groups was not significantly different Baseline 41.2 Ϯ 1.0 41.2 Ϯ 1.0 0.97 (2,165 Ϯ 29 vs. 2,189 Ϯ 35 Kcal/day in End of trial 42.5 Ϯ 1.0 45.6 Ϯ 1.1 0.001 the control and DASH diets, respectively; Change 1.3 Ϯ 0.7 4.3 Ϯ 0.9 0.001 P ϭ 0.62). The results were the same re- LDL-C (mg/dl) garding the actual protein intake (15 vs. Baseline 114.7 Ϯ 3.5 118.7 Ϯ 3.6 0.23 16%) and total fat intake (28 vs. 29%) as End of trial 111.9 Ϯ 4.1 101.5 Ϯ 3.1 0.02 well as the percentage of the carbohydrate Change Ϫ2.7 Ϯ 4.8 Ϫ17.2 Ϯ 3.5 0.02 intake (57 vs. 55%) in the control and Total cholesterol (mg/dl) DASH diet groups, respectively. These Baseline 213.3 Ϯ 6.0 214.9 Ϯ 5.7 0.80 two diets were different in sodium con- End of trial 205.0 Ϯ 6.6 192.7 Ϯ 4.6 0.03 tent (2,310 vs. 2,996 mg/day in the con- Change Ϫ8.3 Ϯ 6.3 Ϫ22.1 Ϯ 5.7 0.11 trol and DASH diets, respectively). The Data are means Ϯ SE. *The control diet was a designed to control diabetes. The general recommenda- DASH diet had higher amount of calcium tion for macronutrient composition of the diet was 50–60% carbohydrates; 15–20% protein and Ͻ30% (1,299 vs. 912 mg/day), potassium total fat. The amount of simple sugar was less than 5% of calorie intake. †The DASH diet was rich in fruits, vegetables, whole grains, and low-fat dairy products, and low in saturated fat, total fat, choles- (4,399 vs. 3,219 mg/day) and fiber (30 vs. terol, refined grains, and sweets. The amount of sodium intake was 2,400 mg per day. ‡P values are for 26 g/day). In the DASH eating pattern ver- comparisons between the two diet periods (general linear model). DBP, diastolic blood pressure; FBG, sus the control diet, the number of serv- fasting blood glucose; HDL-C, HDL cholesterol; LDL-C, LDL cholesterol; SBP, systolic blood pressure; ings of fruit (5 vs. 3), vegetables (6.8 vs. TG, triglyceride. 4), dairy (3 vs. 2), and whole grains (4.5 vs. 2.5) was higher. effects on type 2 diabetic patients’ cardio- calorie–dense diets for weight loss (12). Effects of the two diets on cardiom- metabolic parameters. Furthermore, the dairy content, which etabolic risks are shown in Table 1, indi- The prescribed caloric intake of both might be related to weight reduction (13), cating a significant reduction in most risk diets was the same, but the calorie density was higher in the DASH diet was higher factors from the DASH diet. of food in the DASH diet was lower than than the control diet. that in the control diet. A long-term– The DASH eating pattern also had a CONCLUSIONS — We found that weight-loss trial over 18 months also in- more beneficial impact on the patient’s the DASH-eating pattern had beneficial dicated beneficial effects of using low- glycemic control. More fiber, phytoestro-

56 DIABETES CARE, VOLUME 34, NUMBER 1, JANUARY 2011 care.diabetesjournals.org Azadbakht and Associates gen, and isoflavone intake due to higher ceived health insurance from the Isfahan Steel A, Puchau B, Zulet MA, Martinez JA, Mar- fruit and vegetable consumption, along Company and attended the Shaheed Motahari tinez-Gonzalez MA. Hypothesis-oriented with more weight reduction might be re- Hospital of Fooladshahr. food patterns and incidence of hyperten- sponsible for these effects (4). No potential conflicts of interest relevant to sion: 6-year follow-up of the SUN (Segui- The present study suggests the DASH this article were reported. miento Universidad de Navarra) pro- L.A. and A.E. conceptualized and designed spective cohort. Public Health Nutr 2010; diet plan could reduce LDL and increase the study, performed statistical analyses, 13:338–349 HDL cholesterol. Our previous research drafted the manuscript, and interpreted data. 7. Levitan EB, Wolk A, Mittleman MA. Con- on patients with metabolic syndrome also N.R.P.F. participated in data collection and sistency with the DASH diet and inci- indicates a beneficial effect of this type of entry and prescribed diets to the participants. dence of heart failure. Arch Intern Med. diet on lipid profiles (4). There was no M.K., M.H.B., and M.R. participated in data 2009;169:851–857 difference in the serum triglyceride levels collection and took measurements. P.J.S. 8. Liese AD, Nichols M, Sun X, D’Agostino when we compared the effects of the two helped draft the manuscript and edited the En- RB Jr, Haffner SM. Adherence to the diets. However, DASH was compared glish version of the manuscript. W.C.W. DASH Diet is inversely associated with in- with the control diet, which also had ben- helped draft the manuscript and provided cidence of type 2 diabetes: the insulin re- comments contributing to the interpretation sistance atherosclerosis study. Diabetes eficial effects on lowering the serum tri- of results. All authors approved the final Care 2009;32:1434–1436 glyceride level. manuscript for submission. 9. Fleiss JL. The Design and Analysis of Clini- Higher intake of legumes such as soy The authors thank the participants of the cal Experiments. London, John Wiley and in the DASH diet might also be responsi- study for their enthusiastic support. Sons, 1986, p. 263–271 ble for its beneficial effects on metabolic 10. Harris TJ, Cook DG, Wicks PD, Cappuc- parameters (14). The kind of fat con- cio FP. Impact of the new American Dia- sumption in different diets is also impor- References betes Association and World Health tant. Consuming higher amounts of 1. Kalofoutis C, Piperi C, Kalofoutis A, Har- Organisation diagnostic criteria for diabe- nonhydrogenated vegetable oil with the ris F, Phoenix D, Singh J. Type II diabetes tes on subjects from three ethnic groups DASH diet might be related to its more mellitus and cardiovascular risk factors: living in the UK. Nutr Metab Cardiovasc current therapeutic approaches. Exp Clin Dis 2000;10:305–309 favorable effects. Cardiol 2007;12:17–28. 11. Anderson JW. Diabetes mellitus: medical Because nonadherent participants did 2. Vollmer WM, Sacks FM, Ard J, Appel LJ, nutrition therapy. In Modern Nutrition in not participate in all phases of the study, Bray GA, Simons-Morton DG, Conlin PR, Health and Disease. 10th ed. Shils ME, we could not use intention-to-treat anal- Svetkey LP, Erlinger TP, Moore TJ, Shike M, Ross AC, Caballero B, Cousins ysis. Dietary intake in the present study Karanja N, DASH-Sodium Trial Collabo- RJ, Eds. Philadelphia, Lippincott Wil- was self-reported, and patients were given rative Research Group. Effects of diet and liams and Wilkins, 2006, p. 1051–1053. recommendations to follow a particular sodium intake on blood pressure: sub- 12. Flood A, Mitchell N, Jaeb M, Finch EA, diet (rather than receiving prepared group analysis of the DASH-sodium trial. Laqua PS, Welsh EM, Hotop A, Langer SL, foods), likely resulting in possible imper- Ann Intern Med 2001;135:1019–1028 Levy RL, Jeffery RW. Energy density and fect adherence to the diets. The Omni- 3. Buse JB, Ginsberg HN, Bakris GL, Clark weight change in a long-term weight-loss NG, Costa F, Eckel R, Fonseca V, Gerstein trial. Int J Behav Nutr Phys Act 2009;6:57 Heart (Optimal Macro-nutrient Intake HC, Grundy S, Nesto RW, Pignone MP, 13. Zemel MB, Thompson W, Milstead A, Heart) study (15) has expanded the ma- Plutzky J, Porte D, Redberg R, Stitzel KF, Morris K, Campbell P. Calcium and dairy cronutrient variability of the DASH di- Stone NJ, American Heart Association, acceleration of weight and fat loss during etary pattern, which will be interesting to American Diabetes Association. Primary energy restriction in obese adults. Obes explore in future studies. prevention of cardiovascular diseases in Res 2004;12:582–590. The DASH eating pattern may play an people with diabetes mellitus: a scientific 14. Azadbakht L, Atabak S, Esmaillzadeh A. important role in managing cardiometa- statement from the American Heart Asso- Soy protein intake, cardiorenal indices, bolic risks among type 2 diabetic patients. ciation and the American Diabetes Asso- and C-reactive protein in type 2 diabetes Longer-term studies are needed to assess ciation. Circulation 2007;115:114–126 with nephropathy: a longitudinal ran- the sustainability of these effects. 4. Azadbakht L, Mirmiran P, Esmaillzadeh domized clinical trial. Diabetes Care A, Azizi T, Azizi F. Beneficial effects of a 2008;31:648–654 Dietary Approaches to Stop Hypertension 15. Appel LJ, Sacks FM, Carey VJ, Obarzanek eating plan on features of the metabolic E, Swain JF, Miller ER 3rd, Conlin PR, Er- Acknowledgments— This study was sup- syndrome. Diabetes Care 2005;28:2823– linger TP, Rosner BA, Laranjo NM, Charles- ported by the Isfahan University of Medical 2831 ton J, McCarron P, Bishop LM, OmniHeart Sciences (primary sponsor). The facilities for 5. Forman JP, Stampfer MJ, Curhan GC. Collaborative Research Group. Effects of conducting the biochemical experiments and Diet and lifestyle risk factors associated protein, monounsaturated fat, and carbo- sample recruitment were provided by Shahid with incident hypertension in women. hydrate intake on blood pressure and serum Motahari Hospital of Fooladshahr, Isfahan JAMA 2009;302:401–411. lipids: results of the OmniHeart random- Steel Company, Isfahan. All participants re- 6. Toledo E, de A Carmona-Torre F, Alonso ized trial. JAMA 2005;294:2455–2464

care.diabetesjournals.org DIABETES CARE, VOLUME 34, NUMBER 1, JANUARY 2011 57