Use of a High-Protein Diet in the Management of Feline Diabetes Mellitus*

Transcription

1 Use of a High-Protein Diet in the Management of Feline Diabetes Mellitus* Glenn Frank, DVM, PhD a Wendy Anderson, DVM, PhD b Helene Pazak, DVM, PhD b,c Elizabeth Hodgkins, DVM, JD a,d Joan Ballam, MS b Dottie Laflamme, MS, DVM, PhD b a Heska Corporation 1613 Prospect Parkway Fort Collins, CO b Ralston Purina Company Checkerboard Square St. Louis, MO c Current address: Michigan State University College of Veterinary Medicine East Lansing, MI d Current address: Veterinary Pet Insurance 4175 East La Palma Avenue Anaheim, CA ABSTRACT A study was conducted to evaluate the clinical response of diabetic cats to a high-protein, low-carbohydrate diet. Adult cats with diabetes mellitus of at least 4 months duration were recruited and fed a high-fiber, moderate-fat canned diet for 1 to 2 months during the standardization period. All cats were then transitioned to a high-protein, low-carbohydrate canned diet for a 3-month treatment period. Analyses of treatment effect included hematology, serum biochemistry, fructosamine, lipid profile, and postprandial glucose curves. Cats were also monitored for changes in body weight, appetite, activity level, urinary habits, and insulin requirements. Nine cats completed the study protocol. All cats remained generally stable throughout the treatment period, al- *This study was sponsored by a grant from Ralston Purina Company, St. Louis, MO. though there was a slight overall improvement in activity. Insulin levels were decreased in eight of the nine cats when transitioned from the high-fiber diet to the high-protein diet, and insulin injections were completely stopped in three of the cats. Results of regression analysis indicated that exogenous insulin could be reduced by over 50% with no loss in glucose control, as measured by serum fructosamine. Results of this study support the use of a highprotein, low-carbohydrate diet in the management of cats with diabetes mellitus. INTRODUCTION The fundamental mechanism of the machinery of the mammalian body is to unlock and control the chemical energy in foods and to use that energy to fuel cellular functions. Diabetes mellitus is a disturbance in the workings of this machinery, leading to disruptions in metabo- 238

2 G. Frank, W. Anderson, H. Pazak, E. Hodgkins, J. Ballam, D. Laflamme lism of carbohydrates, proteins, and fats. Many characteristics of feline diabetes are reportedly similar to those of type 2 diabetes in humans, which is associated with pancreatic islet amyloid deposition. 1 The hallmarks of type 2 diabetes in both cats and humans are decreased insulin secretion, especially in response to a glucose load, and insulin resistance. One difference is that most cats with type 2 diabetes require exogenous insulin, while most humans with type 2 diabetes are not insulin dependent. This may occur because diabetes is more advanced in feline patients when it is finally diagnosed. For many years, diets high in complex carbohydrates have been recommended for human diabetic patients. These recommendations have been extrapolated to veterinary patients, including cats. 2,3 However, recent studies suggest that high-protein, low-carbohydrate diets may be beneficial in the management of diabetes. 48 Feeding commercial diets of various nutrient compositions to normal dogs resulted in significantly different postprandial glucose and insulin concentrations. 4 The relative glycemic index and maximum glucose increments were smaller and the time to peak glucose was greater for those fed a high-protein, low-carbohydrate diet versus a high-carbohydrate diet. Similar findings have been reported for both healthy humans and those with noninsulin-dependent diabetes mellitus. 5,6 Ingestion of protein in nondiabetic, healthy humans and well-controlled, noninsulin-dependent diabetics resulted in flattened postprandial glucose curves compared with individuals ingesting an isoglucogenic amount of carbohydrate. 7 Protein ingestion in these patients allowed greater food and calorie intake with a much smaller elevation of blood glucose as compared with intake of carbohydrates. A more recent study in noninsulin-dependent diabetics documented clinical benefits from consumption of a high-protein diet. 8 These patients had improved glycemic control as measured by reduced glycosylated hemoglobin following 8 weeks on a high-protein diet. In the naturally carnivorous cat, replacing carbohydrate calories with protein calories may provide a distinct advantage in diabetes management. The objective of this study was to evaluate the clinical response of diabetic cats to a high-protein, low-carbohydrate diet. MATERIALS AND METHODS Subjects Client-owned adult cats with stable, naturally occurring diabetes mellitus of at least 4 months duration were recruited for the study. All cats were under the care of veterinarians at one of four primary care practices. Indoor cats of any breed or gender that were currently well managed with exogenous insulin were considered eligible for the study. Cats selected were negative for feline leukemia virus and feline immunodeficiency virus. Exclusion criteria included clinically significant dysfunction of any organ other than the endocrine pancreas, such as renal, hepatic, or gastrointestinal dysfunction, as determined by medical history, physical examination, complete blood count (CBC), and biochemical profile. Additionally, cats treated with oral hypoglycemic agents, rather than insulin, were excluded. Experimental Design and Data Collection The trial consisted of a standardization period and a treatment period. All cats were fed a high-fiber, moderate-fat canned diet (Prescription Diet Feline w/d, Hill s Pet Nutrition, Inc., Topeka, KS) for 1 to 2 months during the standardization period, after which all cats were transitioned onto a high-protein, low-carbohydrate canned diet (Purina Veterinary Diets, DM Diabetes Management Formula, Ralston Purina Co., St. Louis, MO) for the 3-239

3 TABLE 1. Nutrient Profiles* of Diets Fed to Diabetic Cats Percent (As Fed) Diet Percent (Dry Matter) Nutrient High-Fiber High-Protein High-Fiber High-Protein Protein Carbohydrate Crude fiber Fat *Nutrient profiles as published in each manufacturer s product guide. Prescription Diet Feline w/d, Hill s Pet Nutrition, Inc., Topeka, KS. Purina Veterinary Diets TM DM Diabetes Management Formula TM, Ralston Purina Co., St. Louis, MO. month treatment period. The assigned diets were to be fed exclusively, although up to 10% of intake from treats or other sources was permitted without excluding the patient from the study. Dietary nutrient profiles are listed in Table 1. Prior to being entered into the study, each cat was given a physical examination; body weight was recorded; and venous blood was drawn for CBC, serum biochemistry, lipid profile (total cholesterol, high-density lipoproteins [HDLs], and triglycerides), and fructosamine analysis. The assays were performed at a single laboratory and were repeated at the end of each period and at the midpoint of the treatment period. A 6- to 8-hour postprandial glucose curve also was completed at entry and after each trial period. Glucose curves were evaluated in different manners between clinics, with some using test strips and others using inclinic glucose monitors for analysis. However, the methods were consistent over time, so it was possible to evaluate changes within each cat. Cats were monitored closely during the treatment period and insulin dosage was adjusted as needed, based on the clinical judgment of the attending veterinarian. Cat owners maintained a diary to record food intake; insulin administration; any changes in activity, drinking, or excretory habits; or other significant observations. Statistical Analysis Analysis of variance for repeated measures on each cat was used to determine differences among test periods for each of the quantitative response variables. Linear contrasts were used to compare values from midpoint to the end of the test period; to compare the pooled treatment period values (protein diet) with the standardization values (fiber diet); and to compare pooled values from the standardization period and the treatment period with the values at entry. When the statistical assumptions of the parametric analysis of variance were violated, a nonparametric Friedman s test was used. Linear regression was used to assess the effect of changes in insulin dose on serum fructosamine and glucose concentrations. Mean values are reported for all parametric analyses, and median values are reported for those analyzed by nonparametric methods. Significant differences for all tests were determined by a P value of less than.05. RESULTS Eighteen cats were entered into the study. Of these, adequate data were collected for nine. One 240

4 G. Frank, W. Anderson, H. Pazak, E. Hodgkins, J. Ballam, D. Laflamme TABLE 2. Changes in Selected Serum Biochemistry and Hematology in Cats with Diabetes Mellitus* on High-Protein, Low-Fat Diet Treatment End of End of Period Treatment Reference Variable Entry Stabilization P value Midpoint P value Period P value Range Alkaline phosphatase 58.0 (4.3) 30.0 (4.3) NS 44.0 (5.1) (4.3) NS 6106 (U/L) Anion gap (meq/l) 24.3 (0.5) 25.7 (0.5) (0.6) NS 27.0 (0.5) NS 1027 Urea nitrogen (mg/dl) 28.0 (1.0) 27.2 (1.0) (1.2) (1.0) NS 1732 Phosphorus (mg/dl) 3.8 (0.2) 4.2 (0.2) (0.2) NS 4.1 (0.2) NS Total bilirubin (mg/dl) 0.1 (0.04) 0.2 (0.04) NS 0.2 (0.04) NS 0.2 (0.04) Cholesterol (mg/dl) (16.2) (16.2) (19.0) (16.2) NS High-density (14.4) (14.4) (16.9) (14.4) NS NA lipoprotein (mg/dl) Nonesterified fatty 0.8 (0.1) 0.6 (0.1) NS 1.1 (0.2) NS 1.3 (0.1).05 NA acids (meq/l) Hemoglobin (g/dl) 13.7 (0.2) 13.4 (0.2) NS 12.7 (0.3) NS 13.6 (0.3) Hematocrit (%) 42.2 (0.8) 41.9 (0.8) NS 38.7 (1.0) NS 41.8 (0.9) Mean corpuscular 32.6 (0.2) 32.1 (0.2) (0.2) (0.2) NS 3036 hemoglobin concentration (g/dl) Mean corpuscular 45.4 (0.5) 45.7 (0.5) NS 44.9 (0.6) (0.5) volume (fl) Red blood cells 9.3 (0.2) 9.2 (0.2) NS 8.6 (0.2) NS 9.6 (0.2) ( 10 6 /µl) *Includes only variables that changed with treatment and reached or approached statistical significance; values are means (± SEM) unless otherwise indicated. Entry versus end of stabilization period. End of stabilization versus treatment period midpoint. Treatment period midpoint versus end of treatment period. Median value (± SEM) reported due to non-normal distribution. NA = not available; NS = not significant. cat was withdrawn due to difficulties stabilizing after a dietary change, and eight others were eliminated for incomplete data or poor client compliance (2); dietary indiscretion (3); or euthanasia for reasons unrelated to the study (3). Few differences were observed for serum biochemistry or hematology variables. Only those that approached or achieved statistical significance are shown in Table 2. Serum urea nitrogen increased significantly during the treatment period, but serum creatinine was virtually unchanged (median ranged from 1.6 to mg/dl throughout the study). Serum cholesterol and HDL levels decreased while cats were consuming the high-protein diet, despite greater fat intake, yet the HDL: total cholesterol ratio actually increased. Serum triglycerides were slightly decreased by the high-protein diet, but the reduction was not significant. The clients recorded perceptions of appetite, activity, urination, and overall health showed a slight overall improvement in activity and all cats remained generally stable throughout the 241

5 TABLE 3. Owner Assessment of Clinical Responses and Changes in Measured Insulin Values from End of Stabilization Period to End of Treatment Period for Diabetic Cats on High-Protein, Low-Carbohydrate Diet Change in Clinical Response* % Decrease Cat Appetite Activity Urination Overall Health in Insulin NR *0 = no change; +1 = improved; 1 = worsened. NR = not recorded. treatment period (Table 3). Body weight also remained stable throughout the study, with the median ranging from 5.7 kg at entry to 5.5 kg at the end of the stabilization period to 5.6 kg at the end of treatment period. Exogenous insulin usage decreased slightly during the stabilization period; however, this did not reach statistical significance (Table 4). Both total dose and dosage rate of insulin decreased significantly (P <.01) from the end of the standardization period to the end of the treatment period (Table 4). Response to dietary change occurred quickly as demonstrated by the lack of differences (P >.10) in both the insulin dose and the dosage rate between the midpoint and end of the treatment period. The daily dose of insulin was decreased in eight of the nine cats, with injections completely stopped in three of these. While serum fructosamine increased in the three animals that were taken off insulin, overall well-being and body weight remained stable or improved for two of these three cats during the 3-month treatment period. Exogenous insulin could be reduced by over 50% in cats fed the high-protein diet with no increase in serum fructosamine (Figure 1). A significant linear relationship (P <.01) was found for the change in serum fructosamine as a function of change in insulin dose. Exogenous insulin was not reduced in one cat, and this cat s serum glucose and fructosamine returned to normal levels while on the high-protein diet and insulin. The effect of insulin change on serum glucose was not statistically significant, most likely due to the high variability and relatively small sample size (Figure 2). Likewise, changes in mean blood glucose area under the curve (AUCBG) were not significant (P >.10). Mean AUCBG was mg/dl/min at the start of the stabilization period, mg/dl/min at the end of the stabilization period, and mg/dl/min at the end of treatment. DISCUSSION A primary goal of diabetes therapy is the elimination of owner-observable signs that occur secondary to hyperglycemia and glycosuria. Avoidance of insulin injections is highly desirable by some pet owners. Glucose control, as determined by the owner s assessment of be- 242

6 G. Frank, W. Anderson, H. Pazak, E. Hodgkins, J. Ballam, D. Laflamme TABLE 4. Insulin, Serum Glucose, and Fructosamine Levels in Diabetic Cats Fed Different Diets Insulin Insulin Serum Fructosamine Serum Glucose (IU/day) (IU/kg body weight) (µmol/l) (mg/dl) Stabil Treat Treat Stabil Treat Treat Stabil Treat Stabil Treat Cat Entry * End Mid End Entry * End Mid End Entry * End End Entry * End End ND ND ND Median ** ** 0.52 Mean Reference range *Entry = entry into the study. Stabil End = end of stabilization period, after consuming the high-fiber diet. Treat Mid = midpoint of treatment period, after consuming the high-protein diet. Treat End = end of treatment period, after consuming the high-protein diet. Difference between values at entry and end of stabilization approached statistical significance (P =.10). ** Values at end of stabilization period and midpoint of treatment period differ significantly (P <.01). Values at end of stabilization and period treatment periods differ significantly (P <.01). ND = not determined. 243

7 60 80 Percent Change in Serum Fructosamine Percent Change in Serum Glucose Percent Change in Insulin Dose Percent Change in Insulin Dose Figure 1. Correlation illustrating the effect of changing insulin dose on serum fructosamine concentration in diabetic cats. The changes reflect differences between the end of the standardization period and the end of the treatment period; y = X; r 2 = 0.81; P <.001 havior, food and water intake, and urination as well as the veterinarian s evaluation of body weight and mean serum fructosamine, was maintained in the cats in this study despite an overall reduction in insulin use. Clients were generally pleased with the health status, physical appearance, and overall response of their cats to the test diet. Insulin injections were ceased completely for three of the cats. Despite a resultant increase in serum fructosamine that indicated a reduction in glucose control, the attitude, behavior, and overall well-being of these three cats remained acceptable. For those clients unwilling or unable to administer daily insulin injections, this degree of control may be adequate to prolong an acceptable quality of life in a diabetic cat and provide a reasonable alternative to euthanasia. Greater reliance was placed on the in-home parameters and serum fructosamine level than in-hospital glucose curves to assess response to treatment, which is consistent with the recommendation of others. 9,10 Glucose curves may be Figure 2. Correlation illustrating the effect of changing insulin dose on serum glucose concentration in diabetic cats. The changes reflect differences between the end of the standardization period and the end of the treatment period; y = X; r 2 = 0.19; P =.24 significantly altered by the pet s level of stress and can be difficult to interpret. Serum fructosamine may be a better indicator of long-term glucose control. Fructosamine is formed from the glycation, or binding, of glucose to amino acid constituents of albumin and other serum proteins. This reaction is dependent on the glucose concentration and is irreversible. Thus, the fructosamine concentration depends on the extent of hyperglycemia and the half-life of the protein. Fructosamine provides an indication of glucose control over the previous 2 to 3 weeks. In this study, consumption of the high-protein diet, in contrast to the high-fiber diet, permitted insulin doses to be reduced by more than 50% with no increase in serum fructosamine. Serum fructosamine concentrations less than 450 µg/l are consistent with moderate to good long-term control of hyperglycemia. 10 Although these levels were not achieved during the 3- month treatment period, the clients were generally pleased with the clinical response in their cats. In addition, 6 to 9 months of follow-up 244

8 G. Frank, W. Anderson, H. Pazak, E. Hodgkins, J. Ballam, D. Laflamme data were available for seven of the cats in the study, five of which achieved serum fructosamine levels below 450 µg/l (data not shown). High-protein diets have been used with good success in other studies. 4,7,8 When commercial diets of various nutrient compositions were fed to normal dogs, differences could be detected in postprandial glucose and insulin concentrations. 4 The relative glycemic index and maximum glucose increment were smaller and the time to peak glucose greater for a highprotein, low-carbohydrate diet compared with a high-carbohydrate diet. 4 The area under the insulin curve was significantly reduced in healthy cats fed a high-protein diet compared with a lower-protein diet (M. Hoenig, unpublished observations, 2000). In normal humans and well-controlled noninsulin-dependent diabetics, ingestion of protein resulted in flattened postprandial glucose curves compared with individuals ingesting an isoglucogenic amount of carbohydrate. 7 A more recent study documented clinical benefits from consumption of a high-protein diet, in which noninsulin-dependent diabetic patients had improved glycemic control as measured by reduced glycosylated hemoglobin following 8 weeks on the high-protein diet. 8 However, in other studies, increased protein intake was associated with an increased glucose response in insulin-dependent diabetics. 6,11 This may reflect a difference between type 1 and type 2 diabetes. However, even if this theory is correct, since diabetes in most cats is thought to be similar to human type 2 diabetes 1, most might be expected to respond to a high-protein diet. Such an effect has been demonstrated in cats. 12 Likewise, in this study, cats fed the high-protein diet were able to maintain glucose control despite decreased insulin use. The recognition that protein delays the time to peak glucose may be of particular importance in diabetic cats. Cats rely on the enzyme hexokinase to help clear glucose from the blood by trapping the glucose intracellularly. 13 Hexokinase has a reduced affinity for glucose, in contrast with glucokinase, the enzyme used by dogs and humans. As a result, normal cats tend to take longer to clear a glucose load. 14 Because maximal tolerance without glycosuria depends on the rate at which the body is able to remove glucose from the bloodstream, the time element assumes an importance at least equal to that of total yield of glucose. 7 If glucose enters the bloodstream at a rate similar to the capacity for cellular uptake, the maximum or peak concentration will be reduced and the likelihood of glycosuria, with associated polyuria and other complications, will be reduced. This effect was demonstrated in human diabetes, 7 where glucosuria occurred within 4 hours following consumption of various carbohydrate sources but not after an isoglucogenic amount of lean beef. While the specific effect of time was not evaluated in this study, the ability to reduce exogenous insulin without compromising glucose control is consistent with that effect. In conclusion, the results of this study showed that consumption of a high-protein, low-carbohydrate diet, rather than a high-fiber diet, allowed insulin to be reduced by 50% without compromising glucose control in diabetic cats. These data and results from other studies support the use of a high-protein, lowcarbohydrate diet in the management of cats with diabetes mellitus. REFERENCES 1. Lutz A, Rand JS: Pathogenesis of feline diabetes mellitus. Vet Clin North Am Small Anim Pract 25: , Nelson RW: Dietary management of diabetes mellitus. J Small Anim Pract 33:213217, Ihle SL: Nutritional therapy for diabetes mellitus. Vet Clin North Am Small Anim Pract 25:585597, Nguyen P, Dumon H, Buttin P, Martin L, Gouro AS: Composition of meal influences changes in postpran- 245

9 dial incremental glucose and insulin in healthy dogs. J Nutr 124(Suppl):2707S2711S, Franz MJ: Protein: Metabolism and effect on blood glucose levels. Diabetes Educ 23:643651, Peters AL, Davidson MB: Protein and fat effects on glucose responses and insulin requirements in subjects with insulin-dependent diabetes mellitus. Am J Clin Nutr 58:555560, Conn JW, Newburgh LH: The glycemic response to isoglucogenic quantities of protein and carbohydrate. J Clin Invest 15:665671, Gutierrez M, Akhavan M, Jovanovic L, Peterson CM: Utility of a short-term 25% carbohydrate diet on improving glycemic control in type 2 diabetes mellitus. J Am Coll Nutr 17:595600, Crenshaw KL: CVT update: Monitoring treatment of diabetes mellitus in dogs and cats. In: Bonagura JD, ed. Kirk s Current Veterinary Therapy XIII, Small Animal Practice. Philadelphia: WB Saunders Co; 2000: Greco DS. Treatment of non-insulin-dependent diabetes mellitus in cats using oral hypoglycemic agents. In: Bonagura JD, ed. Kirk s Current Veterinary Therapy XIII, Small Animal Practice. Philadelphia: WB Saunders Co; 2000: Linn T, Geyer R, Prassek S, Laube H: Effect of dietary protein intake on insulin secretion and glucose metabolism in insulin-dependent diabetes mellitus. J Clin Endrocrinol Metab 81: , Mazzaferro EM, Greco DS, Turner SJ: Treatment of feline diabetes mellitus with a high protein diet and acarbose [abstract]. J Vet Intern Med 14:345, Washizu T, Tanaka A, Sako T, Washizu M, Arai T: Comparison of the activities of enzymes related to glycolysis and gluconeogenesis in the liver of dogs and cats. Res Vet Sci 67:205206, Hoenig M: Insulin, the lynchpin of metabolism. Compend Contin Educ Pract Vet 22(Suppl):7680,