Dietary composition can influence patterns of regional fat loss

Abstract In this paper we summarise previously published evidence indicating that dietary composition can influence patterns of regional fat loss. This novel observation was made initially in the course of an outpatient study comparing the efficacy of two diets commonly prescribed for the treatment of type 2 diabetes: a high carbohydrate, high fibre, low fat diet, and a 'modified fat' diet containing a high proportion of monounsaturated fat. Both diets are low in saturated fat and therefore expected to be associated with a reduction in the risk of coronary heart disease. However, we were concerned that due to its relatively high fat content and energy density, the modified fat diet would be associated with weight gain when consumed in the free-living situation. In order to address this question, a randomised crossover design study was conducted in which subjects with type 2 diabetes consumed each diet for three months. Both diets were associated with moderate weight loss. The unexpected and significant finding was that the low fat diet was associated with a disproportionate loss of lower body fat, resulting in an adverse change in the ratio of upper to lower body fat. Loss of body fat during the modified fat diet was proportional, with no change in the ratio of upper to lower body fat. Furthermore, the modified fat diet was clearly preferred over the low fat diet. Studies are currently underway to elucidate the underlying mechanisms and gain further understanding of the significance of these findings, [Aust J Nutr Diet 1998;55(4 Suppl):S32-S36].

Keywords: low fat; mono-unsaturated fat; type 2 diabetes; body fat distribution

The dietary recommendations for treatment of type 2 diabetes have changed markedly over the last 70 years ( 1). Early in this period, diets very high in protein and fat and containing very little starch and no sugar were prescribed. These were followed by the 'portion diet', a 'sugar-free' diet which strictly regulated consumption of foods containing complex carbohydrates, but put less emphasis on discouraging intake of high fat foods. Despite evidence that such high saturated fat, low carbohydrate diets increased the risk of macrovascular disease ( 2), it was not until the 1970s, when Trowell published his hypothesis that fibre-rich starch foods protected against type 2 diabetes and other 'diseases of affluence' ( 3), that attitudes towards carbohydrate liberalised. At about the same time, evidence was emerging from the Framingham Study indicating that people with type 2 diabetes were at particularly high risk for cardiovascular disease ( 4).

High carbohydrate, high fibre diets in the treatment of diabetes
Interest in the prescription of high carbohydrate, high fibre diets for the treatment of diabetes began to build once it was established that these diets could improve both glycaemic control and lipid profile ( 5). This approach was consistent with epidemiological evidence that, on a population basis, high carbohydrate (CHO) diets were associated with low diabetes-related mortality and low diabetes prevalence ( 6). Yet the benefits of this type of diet appeared to reside less in the total quantity of carbohydrate, than in the low fat content (particularly saturated fat) and the high level of dietary fibre ( 7-9). Furthermore, while the benefit of fibre-rich high CHO diets, providing 50 to 60% of daily energy as carbohydrate, had been proven in closely controlled clinical settings ( 8, 9), it was less evident that they would be easily followed by people with type 2 diabetes living in the community ( 10). High CHO diets can be bland and unattractive and compliance is frequently poor among people habitually accustomed to a high fat intake ( 10). One Australian study reported that for a group of 212 subjects with type 2 diabetes who had been prescribed a high fibre, high CHO diet, mean dietary fibre intake was not high (25 ± 15 g/day) and most consumed less than 50% energy as carbohydrate ( 10). Such moderate fibre intake may be insufficient to produce any marked beneficial effects on metabolic and lipid control ( 8, 9). The type of dietary fibre is also important ( 7). Indeed, some high carbohydrate low fat diets of moderate fibre content were shown to worsen both blood glucose control and the dyslipidaemia of type 2 diabetes (increasing triacylglycerol levels and decreasing high density lipoprotein (HDL) cholesterol) ( 11).

Modified fat diets and metabolic control in diabetes
An alternative approach, first promulgated by Garg and coworkers, was to replace saturated fat in the diet of patients with type 2 diabetes with mono-unsaturated fat instead of with carbohydrate ( 12). Such high monounsaturated fat (high-MUFA) diets have been shown by a number of groups to be associated with clear cut improvements in glycaemic control and the dyslipidaemia of type 2 diabetes ( 13). Furthermore, the high MUFA diet has been reported to ameliorate a number of other cardiovascular disease risk factors, including insulin sensitivity ( 14), elevated blood pressure ( 15), susceptibility to oxidative modification of LDL ( 16) and secretion of yon Willebrand factor by endothelial cells ( 17).

Despite the strong rationale for recommending the high MUFA diet to improve metabolic control and reduce the risk of macrovascular disease in type 2 diabetes, concerns regarding the risk of weight gain persisted. The relationship between fat intake and adiposity has been well documented in both population studies (reviewed in ( 18)) and dietary interventions. For example, a study by Lissner et a1. ( 19) reported that healthy women gained weight on a diet where 45 to 50% of energy was supplied as fat. but lost weight on a low fat, high CHO diet (15 to 20% of energy from fat). This impact of energy density on energy intake has been confirmed in numerous studies subsequently.

Modified fat diets and energy balance
We addressed the possibility that the high MUFA diet may be associated with weight gain when consumed outside the controlled metabolic ward situation by comparing the impact on body weight of a high CHO and a high MUFA diet in people with type 2 diabetes who were living at home ( 20). In a randomised, crossover design study, 24 subjects followed a high CHO diet and a high MUFA diet, each for a 12-week period. Contrary to expectations, the subjects lost significant weight on both diets: 0.7 ± 0.3 kg (mean ± SD) on the high CHO diet, and 1.3 ± 0.4 kg on the high MUFA diet (both P < 0.05). This is consistent with the data in a clinical setting where energy intake is regulated (reviewed in ( 21)).

The reasons why a high MUFA diet does not result in expected weight gain are as yet unexplained. The body's fat balance is influenced by factors affecting fat intake, fat storage and fat oxidation, and different types of fat may have quite different effects on these processes ( 22). Physical activity also affects energy and therefore fat balance ( 21). It is possible that different types of dietary fat differ in their effects on physical activity. The diabetic subjects in the above study ( 20) reported not only that they preferred the high MUFA diet to the low fat diet, but also that they felt better (more energetic) while following it.

In relation to fat oxidation, evidence from both animal and human studies suggests that different types of fat are handled quite differently. For example, Jones et al. ( 22) reported that when different fatty acids radio labelled with (sup 13)C were given in breakfast meals to men on test diets, oleic acid (C18:1 n-9) was oxidised more rapidly than linoleic acid (C18:2 n-6), which in turn was oxidised more rapidly than stearic acid (C18:0). This and other data suggest that the type of fat influences the rate of fat oxidation and raise the possibility of different types of fat differing in their impact on energy balance. Further studies are indicated to elucidate the role of fat type on the development of obesity.

Dietary fat and body fat distribution
It is not only the deposition of total body fat that is a risk factor for the development of type 2 diabetes, the distribution of this fat regionally is also critical (reviewed in ( 23)). Although the android pattern of fat distribution seen in males and in women with diabetes is strongly associated with the metabolic syndrome, there is a paucity of data on the effect of different diets on the body composition of people with type 2 diabetes. Yet it is important to establish whether it is possible to develop dietary strategies to aid reduction of centralised body fat.

As part of a larger study of dietary composition and energy balance in diabetes ( 20) we studied the effects of high CHO and high MUFA diets on body composition in 16 subjects with type 2 diabetes [six men and ten postmenopausal women, aged 61.8 ± 1.8 years (mean ± SEM)] in a randomised cross-over design study ( 24). After a baseline period of four weeks when subjects followed their habitual diet, subjects were randomised to follow a fibre rich high CHO low fat diet or a high MUFA diet, each for a period of 12 weeks. In the high MUFA diet, consumption of olive oil, nuts and avocado were encouraged. Although use of olive oil was preferred, subjects were also allowed some olive oil based margarine (66.2% oleic acid). Macronutrient intakes on these two diets (as assessed by self-report through seven-day weighed food records) are given in Table 1. On the high MUFA diet subjects consumed more total fat and mono-unsaturated fat and less carbohydrate, dietary fibre and protein than on the high CHO diet (P < 0.005).

As an indicator of compliance the proportion of linoleic acid and oleic acid were determined in plasma cholesteryl ester fractions (Figure t). During the high CHO diet the proportion of linoleic acid in plasma cholesteryl esters fell from 47.2 ± 1.0% to 44.6 ± 1.1% (mean ± SEM) consistent with a dietary reduction of total fat and polyunsaturated fat intake (Figure 1A). Levels of oleic acid in plasma cholesteryl esters fell during the high CHO diet and rose during the high MUFA diet (Figure 1B).

Body composition in this study was assessed by dual energy X-ray absorptiometry (DEXA). In addition to the estimation of total body fat mass (FM) and lean tissue mass (LTM), the distribution of body fat was determined using methods developed by Ley et al. ( 25). Briefly, the upper body was delineated by an upper horizontal border drawn below the chin, by vertical borders lateral to the ribs and by a lower border formed with two oblique lines passed through the hip joints. The lower body encompassed that area below the two lines passed through the hip joints.

The DEXA total body scans indicated that both diets, although they differed in the type and amount of dietary fat, had very similar effects on total body fat (Figure 2).

On both diets, subjects lost significant FM over time (P < 0.05). LTM however, was retained [subjects lost only 0.10 ± 0.27 kg and 0.15 ± 0.23 kg (mean ± SD) of LTM on the high CHO diet and high MUFA diet, respectively]. Despite the similar losses of total body FM, the two diets differed it] their effect on regional body fat distribution. During the high CHO diet, subjects retained upper body fat and lost lower body fat, whereas on the high MUFA diet subjects lost upper body and lower body fat proportionately (Figure 3). Thus when subjects followed the high CHO diet for 12 weeks, the ratio of upper body fat to lower body fat increased, whereas on the high MUFA diet the ratio of upper body fat to lower body fat remained unchanged (Table 2).

These data were analysed by a two-factor repeated measures ANOVA where the type of diet (high MUFA versus high CHO) and time (before and after intervention) were taken as within-subjects factors while sequence (the order in which subjects followed the diets) was taken as a between-subjects factor. Because the study had a randomised cross-over design and was subjected to this analysis; it could be shown that although the results were affected by the sequence in which diets were taken, the type of diet still clearly had a significant effect (P < 0.05). The high CHO diet was associated with a disproportionate loss of lower body fat and was less effective therefore in reducing the centralised body fat distribution characterising type 2 diabetes, This novel finding has been confirmed in a subsequent similar study in a group of 21 women, The adverse change in the ratio of upper body fat to lower body fat is consistent with the tendency of this diet to cause deterioration of glycaemic control and increasing dyslipidaemia ( 11).

The mechanism underlying this unexpected adverse effect of low fat, high carbohydrate diets on patterns of regional fat loss is not understood. It is well established that body fat at different sites is heterogenous, differing in adipocyte size, in the capacity for adipocytes to replicate and differentiate, in adipocyte responsiveness to different hormones and metabolic signals, as well as in patterns of vascularisation and sympathetic innervation (reviewed in ( 23)).

In conclusion, the evidence supporting the use of diets enriched with mono-unsaturated fat to improve metabolic control and reduce the risk of cardiovascular disease in type 2 diabetes is now substantial. Not only have Mediterranean-type diets been shown to reduce a broad range of cardiovascular risk factors (including insulin resistance, blood pressure, the susceptibility of LDL to oxidative modification, diabetic dyslipidaemia, endothelial secretion of yon Willebrand Factor and fibrinolysis), the observation that such diets are not associated with weight gain or an adverse pattern of fat distribution in therapeutic interventions makes them particularly appropriate for the treatment of type 2 diabetes.

This work was supported by grants from the NHMRC (# 950921) and Diabetes Australia.

Table 1. Consumption of macronutrients in 16 subjects with type 2 diabetes on high carbohydrate (CHO) and high mono-unsaturated fatty acid (MUFA) diets (24)
Legend for Chart:

A - High CHO diet (mean ± SEM)
B - High MUFA diet (mean ± SEM)


Protein (% energy)

25 ± 0.7 22 ± 0.4(a)

Total fat (% energy)

23 ± 1.2 35 ± 1.0(a)

Saturated fat

9 ± 0.5 10 ± 0.6

Mono-unsaturated fat

9 ± 0.6 20 ± 0.6(a)

Polyunsaturated fat

4 ± 0.3 5 ± 0.1

Carbohydrate (% energy)

49 ± 1.2 40 ± 1.0(a)

Fibre (g/day)

31.2 ± 1.2 26.4 ± 1.3(a)

Energy (MJ)

6.5 ± 0.6 6.6 ± 0.5

(a) Significant difference between high CHO and high
MUFA diets, P < 0.005.
Table 2. Change in the ratio of upper-body fat to lower-body fat in 16 subjects with type 2 diabetes who followed high carbohydrate (CHO) or high mono-unsaturated fatty acid (MUFA) diets for 12 weeks (24)
Legend for Chart:

A - Ratio before diet (mean ± SEM)
B - Ratio after diet (mean ± SEM)


High CHO diet

2.00 ± 0.12 2.21 ± 0.10

High MUFA diet

2.29 ± 0.13 2.27 ± 0.12(a)

(a) Significant difference between high CHO and high MUFA
diet by two-way repeated measures ANOVA, P < 0.05.
GRAPH: Figure 1. Percentage of A. linoleic acid and B, oleic acid in serum cholesteryl esters in 16 subjects with type 2 diabetes before (solid histogram) and after (open histogram) high cholesterol (CHO) and high mono-unsaturated fatty acid (MUFA) diets. Each diet was followed for 12 weeks. Vertical bars indicate the SEM. Adapted from ( 24).

GRAPH: Figure 2. Loss of total body fat mass by dual energy X-ray absorptiometry in 16 subjects with type 2 diabetes after 12 weeks on a high cholesterol (CHO) diet (solid histogram) or a high mono-unsaturated fatty acid (MUFA) diet (open histogram), Vertical bars indicate the SEM. Adapted from ( 24).

GRAPH: Figure 3. Change in A. upper body fat or B. lower body fat in 16 subjects with type 2 diabetes who followed a high cholesterol (CHO) (solid histogram) or a high mono-unsaturated fatty acid (MUFA) (open histogram) diet for 12 weeks. Vertical bars indicate the SEM. Adapted from ( 24).

(1.) Nutrition recommendations and principles for people with diabetes mellitus. J Am Diet Assoc 1994;94:504-6.

(2.) Albrink MJ, Lavietes PH, Man EB. Vascular disease and serum lipids in diabetes mellitus: observations over thirty years (1931-1961). Ann Intern Med 1963;58:305-23.

(3.) Trowell HC. Dietary-fiber hypothesis of the etiology of diabetes mellitus. Diabetes 1975;24:762-5.

(4.) Garcia MJ, McNamara PM, Gordon T. Kannel WB. Morbidity and mortality in diabetics in the Framingham population. Diabetes 1974;23:105-11.

(5.) Anderson JW, Midgley WR, Wedman B. Fiber and diabetes. Diabetes Care 1979;2:369-77.

(6.) West KM, Kalbfleisch JM. Influence of nutritional factors on the incidence of diabetes. Diabetes 1971;20:99-108.

(7.) O'Dea K, Walker K. Diet in the aetiology of diabetes. In: Flatt PR, Ionnides C, editors. Drugs, diet and disease. Volume 2. Hefts: Prentice Hall; 1995. p.121-52.

(8.) Riccardi G. Rivellese A, Pacioni D, Genovese S, Mastranzo P, Mancini M. Separate influence of dietary carbohydrate and fibre on the metabolic control in diabetes. Diabetologia 1984;26:116-21.

(9.) O'Dea K, Traianedes K. Ireland P. Niall M, Sadler J, Hopper J, et al. The effects of diet differing in fat, carbohydrate, and fibre on carbohydrate and lipid metabolism in type It diabetes. J Am Diet Assoc 1989;86: 1976-86.

(10.) Campbell LV. Barth R, Gosper J. Unsatisfactory nutritional parameters in non-insulin-dependent diabetes mellitus. Med J Aust 1989;151:146-50.

(11.) Coulston AM, Hollenbeck CB, Swislocki ALM, Chen YI, Reaven GM. Deleterious metabolic effects of high carbohydrate, sucrose-containing diets in patients with non-insulin dependent diabetes. Am J Med 1987;82:213-20.

(12.) Garg A, Bonanome A, Grundy SM. Zhang ZJ, Unger RH. Comparison of a high-carbohydrate diet with a high-monounsaturated-fat diet in patients with non-insulin-dependent diabetes mellitus. N Engl J Med 1988;319:829-34.

(13.) Campbell LV, Marmot PE, Dyer JA, Borkman M. Storlien LH. The high-monounsaturated tat diet as a practical alternative for NIDDM. Diabetes Care 1994:17:177-82.

(14.) Parillo M, Rivellese AA, Ciardullo AV, Capaldo B, Giacco A, Genovese S, et al. A high-monounsaturated-fat/low-carbohydrate diet improves peripheral insulin sensitivity in non-insulin-dependent diabetic patients. Metabolism 1992;41:1373-8.

(15.) Rasmussen OW, Thomsen C, Hansen KW, Vesterlund M, Winther E, Hermansen K. Effects on blood pressure, glucose, and lipid levels of a high monounsaturated fat diet compared with a high-carbohydrate diet in NIDDM subjects. Diabetes Care 1993:16:1565-71.

(16.) Dimitriadus E, Griffin M, Collins P, Johnson A, Owens D, Tomkin GH. Lipoprotein composition in NIDDM: effects of dietary oleic acid on the composition, oxidisability and function of low and high density lipoproteins. Diabetologia 1996;39:667-76.

(17.) Thomsen C, Rasmussen OW, Ingerslev J, Hermansen K. Plasma levels of yon Willebrand factor in non-insulin dependent diabetes mellitus are influenced by dietary monounsaturated fatty acids, Thromb Res 1995;77:347-56.

(18.) Doucet E, Tremblay A. Food intake, energy balance and body weight control. Eur J Clin Nutr 1997;51:846-55.

(19.) Lissner L, Levitsky DA, Strupp BJ, Kalkwarf HJ, Roe DA. Dietary fat and the regulation of energy intake in human subjects. Ant J Clin Nutr 1987;46:886-92.

(20.) Walker KZ, O'Dea K, Nicholson GC, Muir JG, Dietary composition, body weight and NIDDM. Comparison of high-fiber, high-carbohydrate, and modified fat diets. Diabetes Care 1995;18:401-3.

(21.) Shah M, Garg A. High fat and high-carbohydrate diets and energy balance. Diabetes Care 1996;19:1142-52.

(22.) Jones PJ, Pencharz PB, Clandinin MT. Whole body oxidation of dietary fatty acids: implications for energy utilization, Am J Clin Nutr 1985;42:769-77.

(23.) Kissebah AH, Krakower GR, Regional adiposity and morbidity. Physiol Rev 1994;74:761-811.

(24.) Walker KZ, O'Dea K, Johnson L, Sinclair AJ, Piers LS. Nicholson GC, et al. Body fat distribution and non-insulin dependent diabetes: comparison of a fiber-rich high carbohydrate, low-fat (23%) diet and a 35% fat diet high in monounsaturated fat. Am J Clin Nutr 1996;63:254-50.

(25.) Ley CJ, Lees B, Stevenson JC. Sex- and menopause-associated changes in body-fat distribution. Am J Clin Nutr 1992;55:950-4.


By Kerin O'Dea and Karen Z. Walker

K. O'Dea, PhD, Professor, Unit of Nutrition and Preventive Medicine, Monash University, Clayton, Victoria. Correspondence: K. O'Dea, Unit of Nutrition and Preventive Medicine, Department of Epidemiology and Preventive Medicine, Monash University, M onash Medical Centre, 246 Clayton Rd, Clayton, Vic 3168

K. Z. Walker. PhD, MND, Research Fellow, Unit of Nutrition and Preventive Medicine, Monash University, Clayton, Victoria

Share this with your friends