GUAR GUM EFFECTS ON BLOOD SERUM LIPIDS AND GLUCOSE CONCENTRATIONS OF WISTAR DIABETIC RATS1

GUAR GUM EFFECTS ON BLOOD SERUM LIPIDS AND GLUCOSE CONCENTRATIONS OF WISTAR DIABETIC RATS¹

Andrea DARIO FRIAS², Valdemiro C. SGARBIERI²,^*

SUMMARY

The effects of guar gum derived from the endosperm of Cyamopsis tetragonoloba (75% soluble fiber, 7.6% insoluble fiber, 2.16% crude protein, 0.78% total lipids, 0.54% ash and 9.55% moisture) on diabetic rats were studied concerning food intake, body weight gain, blood serum cholesterol, triacylglycerols, glucose, LDL-, and HDL-cholesterol concentrations. The effect of gum on indexes of protein absorption and utilization was also investigated. Diets containing 0%, 10% and 20% (w/w) guar gum were fed to diabetic rats for 28 days. In spite of the fact that diabetes elevated blood lipids in all animals, guar gum diet significantly decreased (p <0.05) serum concentrations of cholesterol and triacylglycerols. Furthermore, a concomitant increase in HDL-cholesterol with a substancial elevation of the HDL/LDL cholesterol ratio was found. The most significant result in this assay was the drastic reduction of blood glucose in diabetic rats treated with guar gum diet. The gum promoted a general improvement in the condition of the diabetic rats, in body weight gain and indexes of protein absorption and utilization. The results of this research suggest that guar gum, at concentrations equal to or higher than 10%, should be effective in the treatment of hypercholesterolemia and diabetes, in humans.

Keywords: guar gum, cholesterol, glucose, diabetes.

RESUMO

EFEITOS DA GOMA GUAR SOBRE AS CONCENTRAÇÕES SÉRICAS DE LIPÍDIOS E GLICOSE EM RATOS DIABÉTICOS "WISTAR". Os efeitos da goma guar derivada do endosperma de Cyamopsis tetragonoloba (75% fibra solúvel, 7,6% fibra insolúvel, 2,16% proteína bruta, 0,78% lipídios totais, 0,54% cinza e 9,55% umidade) foram estudados em relação à ingestão de alimento, ganho de peso corporal, concentrações séricas de colesterol, triacilgliceróis, glicose, HDL e LDL-colesterol. O efeito da goma sobre índices de absorção e utilização de proteína também foi investigado. Dietas contendo 0%, 10% e 20% de goma guar foram utilizadas na alimentação de ratos diabéticos por 28 dias. Apesar do diabetes ter elevado os lipídios sangüíneos em todos animais, as dietas com goma guar decresceram significantemente (p <0,05) as concentrações séricas de colesterol e triacilgliceróis. Além disso, foi encontrado um aumento no nível de HDL-colesterol, com uma elevação substancial na relação de HDL/LDL colesterol. O resultado mais significativo deste ensaio, foi a drástica redução da glicose sangüínea nos ratos diabéticos tratados com dietas contendo goma guar. A goma promoveu uma melhora geral nas condições dos animais diabéticos, no ganho de peso corporal e nos índices de absorção e utilização protéica. Os resultados desta pesquisa sugerem que a goma guar, em concentrações igual ou superiores a 10%, podem ser eficazes no tratamento da hipercolesterolemia e diabetes em humanos.

Palavras-chave: goma guar, colesterol, glicose, diabetes.

1 — INTRODUCTION

Dietary fiber has become more and more important since the work of Burkitt et al. [9], Burkitt [10] and Trowell [35], with the hypothesis of a direct relationship between fiber deficient diets and an acceleration in the development of certain chronic and degenerative diseases, which had been prevailing in industrialized countries. A great deal of work has been reported since then showing the benefit of including in the daily diet a fairly high concentration of fiber [16,24,30,32,39].

Dietary fiber has been classified in two fractions, the insoluble fraction containing mainly cellulose, lignin and some hemicelluloses; the soluble fraction comprising mainly pectins, some hemicelluloses and gums. Dietary fiber, particularly the insoluble fraction, exerts a physical action on the intestine, stimulating peristaltic movements, accelerating intestinal food transit, increasing fecal volume and weight and improving fecal consistency. These actions tend to avoid constipation, reduce diverticulosis and the incidence of colon cancer [7,18,27]. On the other hand, the soluble fiber retards intestinal transit and is partially fermented in the large intestine, producing low molecular weight volatile fatty acids which are beneficial to metabolism of intestinal tissues and seem to be related to the cholesterol lowering effect of soluble fibers [15,20,31,36,37].

The main objectives of this investigation were to determine the composition of the guar gum used commercially as a food additive in Brazil; to compare, in diabetic rats, the effect of diets containing 0, 10 and 20% guar gum on the blood serum concentrations of glucose, triacylglycerols, total cholesterol, LDL-, and HDL-cholesterol; to determine the effects of these diets on the food intake, body weight gain and indexes of protein absorption and utilization, by diabetic rats.

2 — MATERIALS AND METHODS

2.1 - Test material

Commercial guar gum was obtained from Grinsted of Brazil (Embú-São Paulo, Brazil) and was of a food grade.

2.2 - Chemical characterization

The moisture content, ash, and crude protein (N x 6.25) were determined by the AOAC [3] methods. Total lipids was determined using the Bligh and Dyer [6] method. Soluble and insoluble fiber were quantified by the method of Asp et al. [5].

2.3 - Animals,diabetes induction and diets

Male Wistar rats weighing an average of 239.84 ± 19.49g were used. After 16 h fasting, animals were anesthetized with ether and received through the penis dorsal vein an alloxan injection of 40 mg/Kg body weight, in saline solution [20]. The pH of the saline had been previously adjusted to 4.5 with 0.01 M citrate buffer. Diabetes development was checked daily by determination of glucose concentration in the urine; the animals were considered diabetic after reaching a glycosuria of 300 mg glucose/dL. Five rats were sacrificed at the beginning of the experiment to obtain blood and liver samples at zero time to use as a reference. The other animals were randomly assigned to three dietary groups of 5 rats each. The diets were prepared according to composition shown in Table 1 and the experiment had a total duration of 28 days. The animals were maintained in individual cages with free access to diets and water. Temperature of the assay room was maintained at 22 ± 2º C and alternate periods of 12 h light-dark controlled automatically. Individual body weight and food intake were weekly recorded during the assays.

Blood was collected at the beginning (zero time) and at days 7, 14, 21 and 28 of the experiment. Animals from each diet were made to fast 16 h prior to sampling and blood collected from the tail vein. Serum glucose concentration was determined weekly and the lipidemic indexes were determined only at the beginning and at the end of the assays. Blood samples were kept in a water bath at 37º C for 30 min and then centrifuged (2500 rpm/15min) for the serum collection. Also, at the beginning and at the end of the experiment livers were carefully dissected, weighed and frozen at -20º C before analysis.

2.4 - Biochemical analysis

Total cholesterol [19], triacylglycerols [8], glucose [34], HDL- cholesterol [11] and LDL- cholesterol [14] were determined in the blood serum using kits (CELM, Barueri, São Paulo). The LDL-cholesterol was quantified by the expression:

LDL cholesterol= total cholesterol - triacylglycerols/5 - HDL cholesterol

Liver lipids was extracted by the method of Haug & Hostmark [17] and total cholesterol was determined according to Huang et al. [19].

2.5 - Protein absorption and utilization

From the 7^th to the 14^th day of the experiment the rats were maintained in metabolic cages for collection of feces and urine for nitrogen determination [3]. Nitrogen balance, protein digestibility, biological value and net protein utilization were determined by the Pellett and Young procedures [28].

2.6 - Statistical analysis

A SANEST statistical program was employed for the analysis of variance, utilizing the F test and an entirely randomized design. The results are presented as the arithmetic mean ± SEM for the control and experimental groups. Differences among means for the groups were evaluated using the Tukey test at 5% probability (p £ 0.05).

3 — RESULTS

The proximate percent composition of the guar gum used in this study is shown in Table 2. As expected, the main constituent of the gum appears as soluble fiber (75%) followed by insoluble fiber (7.6%). Total fiber accounted for 82.6% of the material.

The profiles of blood glucose for diabetic rats during the 4-weeks feeding period are illustrated in Figure 1. The basic value of 350mg/dL remained essentially constant for the animals fed the control diet (diet 1) with 0% guar gum. There was a very significant drop (p £ 0.05) of blood glucose concentrations for the rats fed diets with 10% and 20% guar gum, and no significant difference was detected between these two concentrations.

FIGURE 1.
Blood serum glucose (mg/dL) concentration of male diabetic Wistar rats fed various diets after 0, 7, 14, 21 and 28 days feeding period.

¹ Mean + standard desviation, 5 animals per diet;

² 1 = control, 0% guar gum; 2 = 10% guar gum e 3 = 20% guar gum;

³ (a,b) Different letters indicate statistically different results (p £ 0.05).

The influence of guar gum (10 and 20%) on total cholesterol, triacylglycerols, HDL-, and LDL-cholesterol of diabetic rats are shown in Figure 2. Blood concentrations were determined at the beginning of the experiment (T0) and at the day 28 (T28), end of the assay. It is noticeable that there was, in general, an increase in the blood concentrations of the various components compared with the starting levels. However, the increase in total cholesterol and triacylglycerols was significantly greater for the control diet (diet 1 with 0% guar gum) than for diets 2 and 3 (10 and 20% guar gum, respectively). These data emphasize the strong hypercholesterolemic and hyperlipidemic effects of diabetes. Looking at the lipoprotein cholesterol data, it becomes evident that diets 2 and 3 increased HDL-cholesterol and decreased LDL-cholesterol significantly (p £ 0.05) in relation to the control diet with 0% guar gum.

FIGURE 2
-Blood serum cholesterol, triacylglycerol, HDL-, and LDL — cholesterol, concentration (mg/dL), of male diabetic Wistar rats.

¹ Mean + standard deviation, 5 animals per diet; determinations were done at the beginning of the assays (T0) and at 28-days feeding (T28);

² 1 = control, 0% guar gum; 2 = 10% guar gum; 3 = 20% guar gum (data obtained at T28);

³ (a,b,c) Different superscript letters indicate statistically different results (p £ 0.05).

Cholesterol determinations in the liver were also done at the beginning (T0) and at the end of the assays (T28). Although in the animals fed diets 1, 2 and 3 the liver cholesterol concentrations were higher (3.0 0.3 mg liver cholesterol/g fresh tissue) than at start T0 (2.5 0.2 mg liver cholesterol/g fresh tissue) these elevations were not statistically significant indicating that guar gum consumption by diabetic rats did not interfere in the liver cholesterol concentration.

The influence of 10 and 20% guar gum in the diet (diets 2 and 3, respectively) on food intake and body weight of diabetic rats is presented in Table 3. Diets 2 and 3 reduced food intake significantly (p £ 0.05). The animals fed the control diet (diet 1) exhibited body weight reduction, whereas the rats fed 10 and 20% guar gum diets presented a small but significant increase in body weight.

The effects of guar gum on apparent digestibility (Da), protein apparent biological value (VBa), net protein utilization (NPU), protein efficiency ratio (PER) and nutritional efficiency ratio (NER) were evaluated and the results are shown in Table 4. The animals fed guar gum diets (diets 2 and 3) showed VBa, NPU, NER and PER higher than animals fed control diet (diet 1), although Da values have been similar for all diets. The data indicate that guar gum consumption promoted an important improvement in the animals diabetes condition.

4 — DISCUSSION

Numerous clinical studies have shown that guar gum intake decreases plasma cholesterol concentrations mainly due to reduction of plasma LDL-cholesterol concentration without affecting HDL-cholesterol levels in normal and diabetic rats and patients with hyperlipidemia [4,32,37]. In this study, in spite of an increase of all lipidemic parameters (total cholesterol, triacylglycerols, HDL-, and LDL-cholesterol) imposed by diabetes, the animals fed guar gum diets showed lower serum cholesterol, triacylglycerols and LDL-cholesterol concentrations, compared to the animals fed 0% guar gum diet. Moreover, the significant increase observed in HDL-cholesterol levels in rats fed guar gum diets showed the beneficial effects provided by guar gum ingestion. Some studies mentioned an inverse relation between HDL-cholesterol level and the risk of cardiovascular disease. The high concentration of plasma HDL-cholesterol can exert an anti-atherogenic effect [12]. The elevated concentrations of HDL-cholesterol promote more efficient removal of cholesterol present in the peripheric tissues, mainly in the arterial wall, transfering it to the liver for catabolism and excretion. On the other hand lower HDL-cholesterol concentration could lead to an excessive accumulation of cholesterol in the tissue.

The mechanisms involved in the circulating and tissue cholesterol reduction are not yet clearly established. A few theories try to explain the involvement of fibers in the mechanism. The involvement of the soluble fiber fraction with an unstirred water layer present along luminal surface of mucosa intestinal cells, increases the thickness of that layer forming a physical barrier to the cholesterol absorption [13, 23]; the binding or sequestration of bile acids in the small intestine making these molecules unavailable to form the intraluminal micelles needed for fat and cholesterol absorption, which may thus be reduced [25], and the formation of short chain fatty acids (acetate, propionate and butyrate) resulting from the colonic fermentation of dietary fiber that can induce changes in glucose and fat metabolism [30]. These are the principal hypothesis investigated.

The results reported in this paper on the effect of guar gum on blood serum glucose concentration agree with those of Smith & Holm [33], Ray et al. [29] and Watters & Blaisdell [40] who have also demonstrated a decrease in post-prandial glicemia in diabetic individuals receiving pectin or guar gum in the diet. However, there are studies wherein no reduction was observed [38]. These conflicting results among workers can probably be explained by different fiber concentrations utilized or by the different modes of fiber administration. It has been shown that fiber needs to be well mixed with the food that is to be ingested, to permit its maximum efficacy [22].

It is also possible that the action of soluble polysaccharides in decreasing post-prandial hyperglicemia is related to their viscosity. The diets rich in soluble fiber provide an increase of intestinal content viscosity, since these fibers are molecules that hold water and have the property of forming coloidal gels. This decreases the contact of food with the intestinal mucosa and the enzymatic digestion rate, consequently reducing the intestinal absorption of monosaccharides and disaccharides [21].

The guar gum diets utilized in this investigation promoted an important improvement in the physiological conditions of the diabetic rats, mainly because the hyperglicemia in these animals was controlled throughout the assay. This improvement extended to body weight gain. Although the body weight gain was not high, it could be considered important since the diabetes condition usually induces catabolic processes that lead to body weight reduction, due to the accelerated catabolism of protein, carbohydrate and lipids. The guar gum minimized these catabolic processes providing a higher nitrogen retention. Areas [4] also observed these benefits in diabetic rats fed diets with 10 and 25% of orange pulp. In our study, the lower protein utilization by the animals in the control group could be a reflex of a high reserve energy mobilization induced by the diabetes condition, leading to the significant weight loss observed.

5 — CONCLUSION

On the basis of all data presented the following conclusions can be drawn from the present investigation: guar gum in the diet (10% and 20% concentration) decreased blood glucose significantly, reduced total cholesterol and triacylglycerol concentrations and improved the HDL/LDL cholesterol ratio by increasing the level of HDL-cholesterol and decreasing LDL-cholesterol; it also reduced food intake drastically but increased body weight gain or maintenance. This study confirmed the beneficial effects of guar gum intake in improving the condition of animals with experimentally induced diabetes.

6 — REFERENCES

[1] AMERICAN INSTITUTE OF NUTRITION. Report of the American Institute of Nutrition Ad Hoc Committee on Standards for Nutritional Studies. J. Nutr., Bethesda, v. 107, n. 7, p. 1340-1348, 1977.

[2] AMERICAN INSTITUTE OF NUTRITION. Ad Hoc Commitee on Standards for Nutrition Studies. Second report. J. Nutr., Bethesda, v. 110, n. 8, p. 1726, 1980.

[3] AOAC. Official Methods of Analysis. 12^th ed. Washington, DC: Association of Official Analytical Chemists,1975.

[4] AREAS, M.A. Estudo dos efeitos da polpa de laranja sobre parâmetros fisiológicos, nutricionais, bioquímicos e morfológicos em ratos normais e diabéticos. Tese de doutorado, Faculdade de Engenharia de Alimentos, Universidade Estadual de Campinas, 1994.

[5] ASP, N.G.; CLAES, G.J.; HALLMER, H. SILJESTRON, M. Rapid enzymatic assay of insoluble and soluble dietary fiber. J. Agric. Food Chem., Washington, v. 31, n. 3, p. 476-482, 1983.

[6] BLIGH, E.G. & DYER, W.J. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol., Ottawa, v. 37, n. 8, p. 911-917, 1959.

[7] BRODRIBB, J.N. Dietary fiber in diverticular disease of the colon. In: SPILLER, G.A. & KAY, R.M. (eds.). Medical Aspects of Dietary Fiber. New York: Plenum Press, 1980.

[8] BUCOLO, G. & DAVID, H. Quantitative determination of serum triglycerides by the use of enzymes. Clin. Chem., New York, v. 19, n. 5, p. 476-482, 1973.

[9] BURKITT, D.P.; WALKER A.R.P.; PAINTER, N.S. Effect of dietary fibre on stools and transit-times, and its role in the causation of disease. Lancet, London, v. 2, p.1408-1411, 1972.

[10] BURKITT, D.P. Some diseases characteristic of modern western societies. Br. Med. J., London, v.1, p. 274-278, 1973.

[11] COOPER, C. Manual of Laboratory Operations. Lipid Research Clinics Program. Lipid and Lipoprotein Analysis, National Heart and Lung Institute, NIH (USA),1974.

[12] CAREW, T.E.; HAYES, S.B.; KOSCHINSKY, T.; STEINBERG, D. Mechanism by which high density lipoprotein may slow the atherogenic process. Lancet, London, v. 1, p. 1315-1317, 1976.

[13] CASPARY, W.F.; ELSENHANS, B.; SUFKE, U.; PLOK, M.; BLUME, R.; LEMBCKE, B.; CREUTZFELDT, W. Effect of dietary fiber on absorption and motility. In: Greidanus, T.J.B. van W. (ed). Frontiers of Hormone Research. Washington, DC: National Academy Press, p. 202-217, 1980.

[14] FRIEDEWALD, W.T.; LEVY, R.I. ; FREDRICKSON, D.S. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of preparative ultracentrifuge. Clin. Chem., New York, v. 18, p. 499-502, 1972.

[15] GLORE, S.R.; TREECK, D.V.; KNEHANS, A.W.; GUILD, M. Soluble fiber and serum lipids: a literature review. J. Am. Dietetic Assoc., Chicago, v. 94, n. 4, p. 425-436, 1994.

[16] HABER, G.B.; HEATON, K.W.; MURPHY, D. et al. Depletion and disruption of dietary fibre. Effects on satiety, plasma-glucose and serum insulin. Lancet, London, v. 2, p. 679-682, 1977.

[17] HAUG, A. & HOSTMARK, A.T. Lipoprotein lipases, lipoproteins and tissue lipids in rats fed fish oil or coconut oil. J. Nutr., Bethesda, v. 117, n. 6, p. 1011-1017, 1987.

[18] HELLER, S.N.; HACKLER, L.R.; RIVERS, J.M. Dietary fiber: the effect of particle size of wheat bran on colonic function in young adult men. Am. J. Clin. Nutr., New York, v. 33, p.1734-1744,1980.

[19] HUANG, T.C.; CHEN, C.P.; WEFLER, V.; RAFTERY, A. A stable reagent for the Liebermann-Burchard reaction. Anal. Chem., Washington, v. 33, p. 1405-1407, 1961.

[20] HUGHES, J.S. Potential contribution of dry bean dietary fiber to health. Food. Technol., Chicago, v. 9, p. 122-126, 1991.

[21] INK, S.L. & HURT, D. Nutrition implications of gums. Food Technol., Chicago, v. 41, n. 1, p. 77-82, 1987.

[22] JENKINS, D.J.A.; NINEHAM, R.; CRADDOCK, C.; Mc FEELY, P.C.; DANALDSON, K.; LEIGT, T.; SNOOK, J. Fibre and diabetes. Lancet, London, v. 1, p. 434-435, 1979.

[23] JENKINS, D.J.A. Slow release carbohydrate and the treatment of diabetes. Proc. Nutr. Soc., London, v. 40, p. 227-235, 1981.

[24] KRITCHEVSKY, D.; TEPPER, A.S.; SATCHITHANANDAM, S.; CASSIDY, M.; VAHOUNY, G. Dietary fiber supplements: effects on serum and liver lipids and liver phospholipids composition in rats. Lipids, Chicago, v. 23, p. 318-321, 1988.

[25] KRITCHEVSKY, D. Dietary fibre. Ann. Rev. Nutr., Palo Alto, v. 8, p. 301-328, 1988.

[26] LAZAROW, A. & PALAY, S.L. The production and course of alloxan diabetes in the rat. J. Lab. Clin. Med., St. Louis, v. 31, p. 1004-1015, 1946.

[27] NATIONAL ACADEMY OF SCIENCES. Diet, nutrition and cancer. Washington,DC: National Academy Press, 1982.

[28] PELLETT, P.L. & YOUNG, V.R. Nutritional evaluation of protein foods, Tokyo: The United Nations University, 1980.

[29] RAY, T.K.; MANSELL, K.M.; KNIGHT, L.C.; MALMUD, L.S.; OWEN, O. E.; BODEN, G. Long-term effects of dietary fiber on glucose tolerance and gastric emptying in noinsulin dependent diabetic patients. Am. J. Clin. Nutr., Bethesda, v. 37, n. 3, p. 376-381, 1983.

[30] ROBERFROID, M. Dietary fiber, inulin and oligofructose: a review comparing their physiological effects. Crit. Rev. Food Sci. Nutr., Boca Raton, v. 33, n. 2, p. 103- 148, 1993.

[31] SCHEPPACH, W.; WIGGINS, H.S.; HALLIDAY, D.; SELF, R.; HOWARD, J.; BRANCH, W.J.; SCHREZENMEIR, J.; CUMMINGS, J.H. Effect of gut-derived acetate on glucose turnover in man. Clin. Sci., Colchester (Essex), v. 75, p. 363-366, 1988.

[32] SCHNEEMAN, B.O. & GALLAHER, D. Effects of dietary fiber on digestive enzymes. In: SPILLER, G.A.(ed). Handbook of Dietary Fiber in Human Nutrition. Boca Raton:CRC Press, p.305-312, 1986.

[33] SMITH, U. & HOLM, G. Effect of a modified guar gum preparation on glucose and lipid levels in diabetics and healthy volunteers. Atherosclerosis, Amsterdam, v. 4, n. 5, p. 1-10, 1982.

[34] TRINDER, P. Determination of glucose in blood using glucose oxidase with an alternative oxygen acceptor. Anal. Clin. Biochem., New York, v. 6, n. 24, p. 24-27, 1969.

[35] TROWELL, H. Dietary fibre, ischaemic heart disease and diabetes mellitus. Proc. Nutr. Soc., London, v. 32, p.151-157, 1973.

[36] TURNER, P.R.; TUOMILEHTO, J.; HAPPONEN, P.; LAVILLE, A.E.; SHAIKH, M.; LEWIS, B. Metabolic studies on the hypolipidaemic effect of guar gum. Atherosclerosis, Amsterdam, v. 81, n. 2, p.145-150, 1990.

[37] UBEROI, S.K.; VADHERA, S.; SONI, G.L. Role of dietary fibre from pulses and cereals as hypocholesterolemic and hypolipidemic agent. J. Food Sci. Technol., Oxford, v. 29, n. 5, p. 281-283, 1992.

[38] UUSITUPA, M.; SIITONEN, O.; SAVOLAINEN, K.; SILVASTI, M.; PENTTILA, I.; PARVIAINEN, M. Metabolic and nutritional effects of long-term use of guar gum in the treatment of noninsulin-dependent diabetes of poor metabolic control. Am. J. Clin. Nutr., Bethesda, v. 49, n. 2, p. 345-351, 1989.

[39] WALKER. A.R.P. Does the dietary fiber hypothesis really "work"? Cereal Foods World, St. Paul, v. 38, n.3, p.128-134, 1993.

[40] WATTERS, K. & BLAISDELL, P. Reduction of glycaemic and lipid levels in db/db diabetic mice by Psyllium plant fibre. Diabetes, New York, v. 38, p. 1528-1531, 1989.

² Institute of Food Technology,Av.Brasil, 2880, Postal code 13073-001, Campinas — São Paulo, Brazil. E-mail: [e-mail].

* To whom correspondence should be addressed.

¹

Recebido para publicação em 02/03/98. Aceito para publicação em 27/07/98.

Publication Dates

• Publication in this collection
  16 Dec 1998
• Date of issue
  May 1998

History

• Received
  02 Mar 1998
• Accepted
  27 July 1998