Diabetes Research and Clinical Practice 65 (2004) 235–241

Beneﬁcial effect of low carbohydrate in low calorie diets onvisceral fat reduction in type 2 diabetic patients with obesity

Yoh Miyashita

, Nobukiyo Koide

, Masaki Ohtsuka

, Hiroshi Ozaki

, Yoshiaki Itoh

,Tomokazu Oyama

, Takako Uetake

, Kiyoko Ariga

, Kohji Shirai

∗

Center of Diabetes, Endocrine and Metabolism, Sakura Hospital, School of Medicine, Toho University,564-1 Shimoshizu, Sakura-City, Chiba 285-0841, Japan

Department of Nutritione, Sakura Hospital, School of Medicine, Toho University, Chiba, Japan

Received 15 March 2003; received in revised form 9 December 2003; accepted 29 January 2004

Abstract

The adequate composition of carbohydrate and fat in low calorie diets for type 2 diabetes mellitus patients with obesity is notfullyestablished.Theaimofthisstudywastoinvestigatetheeffectsoflowcarbohydratedietonglucoseandlipidmetabolism,es-peciallyonvisceralfataccumulation,andcomparingthatofahighcarbohydratediet.Obesesubjectswithtype2diabetesmellituswere randomly assigned to take a low calorie and low carbohydrate diet (

11, 1000kcal per day, protein:carbohydrate:fat

25:40:35) or a low calorie and high carbohydrate diet (

11, 1000kcal per day, protein:carbohydrate:fat

25:65:10) for 4weeks. Similar decreases in body weight and serum glucose levels were observed in both groups. Fasting serum insulin levelswerereducedinthelowcarbohydratedietgroupcomparedtothehighcarbohydratedietgroup(

−

30%versus

−

10%,

P <

05).Total serum cholesterol and triglyceride levels decreased in both groups, but were not signiﬁcantly different from each other.High-density lipoprotein–cholesterol (HDL–C) increased in the low carbohydrate diet group but not in the high carbohydratediet group (

15% versus 0%,

P <

01). There was a larger decrease in visceral fat area measured by computed tomographyin the low carbohydrate diet group compared to the high carbohydrate diet group (

−

40cm

versus

−

10cm

P <

05). Theratio of visceral fat area to subcutaneous fat area did not change in the high carbohydrate diet group (from 0.70 to 0.68), but itdecreasedsigniﬁcantlyinthelowcarbohydratedietgroup(from0.69to0.47,

P <

05).Theseresultssuggestthat,whenrestrictdiet was made isocaloric, a low calorie/low carbohydrate diet might be more effective treatment for a reduction of visceral fat,improved insulin sensitivity and increased in HDL–C levels than low calorie/high carbohydrate diet in obese subjects with type2 diabetes mellitus.© 2004 Elsevier Ireland Ltd. All rights reserved.

Keywords:

Low carbohydrate diet; Low calorie diet; Type 2 diabetes mellitus; Obesity; Visceral fat

∗

Corresponding author. Tel.:

81-43-462-8811;fax:

81-43-489-9770.

E-mail address:

kshirai@kb3.so-net.ne.jp (K. Shirai).

1. Introduction

The distribution of body fat is known to be im-portant in the development of obesity-associatedmetabolic disorders. In particular, accumulation of visceral fat is related to the development of insulin

236

Y. Miyashita et al. / Diabetes Research and Clinical Practice 65 (2004) 235–241

resistance[1,2]and to a high incidence of coronaryheart disease[3,4].A low calorie diet is recognized to be a fundamen-tal therapy for type 2 diabetic patients with obesityto improve insulin sensitivity by weight reduction. Aprinciple of this diet is to provide an adequate amountof proteins, vitamins and minerals, and to restrict car-bohydrates and fats as energy components. There havebeen several reports concerning the role of nutrientcomposition of the diet for weight reduction in obe-sity[5–7].For example, Rabast et al.[5]reported that low carbohydrate diet is beneficial for weight reduc-tion, which is associated with low insulin levels and asuppression of appetite by ketone body formation. Onthe other hand, Yong and Werner et al. reported thatlow carbohydrate diet is not always superior to highcarbohydrate diet for body weight reduction[6,7]. Asa result, there is no definite evidence for appropriateratio of carbohydrate in low calorie diet.Recently, in rats given a normal calorie diet, Zava-roniandReavenetal.[8]reportedthatadietcomposedof 60% carbohydrates raised glucose levels and de-creased insulin sensitivity in rats. Furthermore, Coul-ston and co-workers[9]reported that high carbohy-drate diet increased serum glucose, insulin and triglyc-eride levels, and decreased HDL–C in diabetic pa-tients.We have previously reported that low calorie dietcomposed of low carbohydrate is preferable for an im-provement of glucose and lipoprotein metabolism intype 2 diabetes mellitus patients with obesity[10].In this paper, to clarify whether high or low carbohydratediet is more beneficial for type 2 diabetes mellitus pa-tients with obesity, the effects of both diets on visceralfat accumulation as well as parameters of glucose andlipid metabolism were studied.

2. Subjects and methods

2.1. Subjects

Twenty-two obese subjects with type 2 diabetesmellitus patients were recruited for the study (16males, 6 females, mean body mass index 27kg/m

).The mean age was 52

0 years old. They wereall hospitalized and took no medication. All of subjects gave informed consent.

2.2. Diet composition

The ratio of the low carbohydrate diet compositionsof protein, carbohydrate and fat was 25:39:35 (%), andin high carbohydrate diet, the ratio was 26:62:10 (%).The ratio of saturated fatty acids, monounsaturatedfatty acids and polyunsaturated fatty acids were 3:4:3in both diets. Palmitic acid constituted about 70% of the saturated fatty acids, oleic acid about 95% of themonounsaturated fatty acids, and a linoleic acid about70% of the polyunsaturated fatty acids. The contentof dietary ﬁbers was about 18g per day in the lowcarbohydrate diet and about 20g per day in the highcarbohydrate diet.

2.3. Program of diet therapy

The subjects were initially given a 3 day low calo-rie diet composed of high carbohydrate (1000kcal perday, P:F:C

26:10:62), after which they were dividedinto two random groups. One group of subjects wastreated with a low calorie diet composed of low car-bohydrate (

low carbohydrate diet group,

11),and the other was treated with a low calorie diet com-posed of high carbohydrate (

high carbohydrate dietgroup,

11). The clinical proﬁle of the subjectsat the start of the different diet therapies is shown inTable 1.There were no signiﬁcant differences betweenthe two groups. The subjects were treated for 4 weekswith these diets. During this study, all patients werewithout medications and treated with exercise therapy(walking, 30min

2 times per day).

2.4. Blood sampling

Blood samples were taken in the morning after 12hof fasting. Serum was obtained within 1h and sampleswere used for measuring of blood sugar (FBS), insulin(basal IRI), serum total cholesterol (TC), triglyceride(TO) and HDL–C.

2.5. Measurement of body weight and total body fat

Body weight and total body fat were measured inthe morning after 12h of fasting. The percent bodyfat (%) and lean body mass (LBM) were measured byimpedance method using Serkoimpemeter (SIF-93 1,Serko, Japan).

Y. Miyashita et al. / Diabetes Research and Clinical Practice 65 (2004) 235–241

237Table 1Clinical proﬁles of low carbohydrate diet group or high carbohydrate diet group before treatmentHigh carbohydrate group (

11) Low carbohydrate group (

11) SigniﬁcanceBody mass index 27

2 27

4 None

ratio

0.71

0.8 0.69

0.7 NoneHbAlc (%) 9.8

2 10.2

2 NoneFasting blood sugar (mg/dl) 200

50 207

36 NoneBasal IRI (



U/ml) 5.9

1.0 6.2

0.9 NoneTotal cholesterol (mg/dl) 193

48 199

35 NoneTriglyceride (mg/dl) 173

60 175

89 NoneHDL–C (mg/dl) 39

16 38

10 None

The ratio of visceral fat area/subcutaneous fat area.

2.6. Measurement of adipose tissue distribution

To measure visceral and subcutaneous fat mass,computed axial tomography was performed[1].The scan was performed at the position of the navel. Vis-ceral fat area was measured by drawing a line withinthemusclewallsurroundingtheabdominalcavity.Thesubcutaneous fat was calculated by subtracting theamount of visceral fat from the total fat area. Fromthese values, a ratio of visceral fat area to subcuta-neous fat area (

ratio) and an area of visceral fatwas obtained.

2.7. Statistical analysis

Statistical analysis was performed using Student’s

-test and paired

-test.

values less than 0.05 wereconsidered signiﬁcant.

Fig. 1. Changes in body weight, body fat (%) and lean body mass during low calorie diets in low carbohydrate diet group (



) and highcarbohydrate diet group (



). (A–C) Show changes of body weight, body fat (%), and lean body mass, respectively. Data are shown asmean

3. Results

3.1. Change in the body weight, fat (%) and LBM

In the low carbohydrate and the high carbohydratediet groups, the body weight decreased from 73 to64kg, and from 71 to 64kg, respectively, after lowcalorie diet for 4 weeks (Fig. 1A). During these lowcaloriediettherapies,fat(%)decreasedinbothgroups,with no change in LBM in either group (Fig. 1B).These changes in body weight, body fat (%) and LBMwere not signiﬁcantly different between the low andhigh carbohydrate diet groups (Fig. 1A and B).

3.2. Change in FBS and basal IRI