Associations between dietary factors in canned food ... - AVMA Journals

01-07-0173R.qxd

1/8/2002

10:10 AM

Page 163

Associations between dietary factors in canned food and formation of calcium oxalate uroliths in dogs Chalermpol Lekcharoensuk, DVM, MPH; Carl A. Osborne, DVM, PhD; Jody P. Lulich, DVM, PhD; Rosama Pusoonthornthum, DVM, PhD; Claudia A. Kirk, DVM, PhD; Lisa K. Ulrich; Lori A. Koehler; Kathleen A. Carpenter; Laurie L. Swanson

Objective—To identify dietary factors in commercially available canned foods associated with the development of calcium oxalate (CaOx) uroliths in dogs. Animals—117 dogs with CaOx uroliths and 174 dogs without urinary tract disease. Procedure—Case dogs were those that developed CaOx uroliths submitted to the Minnesota Urolith Center for quantitative analysis between 1990 and 1992 while fed a commercially available canned diet. Control dogs were those without urinary tract disease evaluated at the same veterinary hospital just prior to or immediately after each case dog. A content-validated multiple-choice questionnaire was mailed to each owner of case and control dogs with the permission of the primary care veterinarian. Univariate and multivariate logistic regressions for each dietary component were performed to test the hypothesis that a given factor was associated with CaOx urolith formation. Results—Canned foods with the highest amount of protein, fat, calcium, phosphorus, magnesium, sodium, potassium, chloride, or moisture were associated with a decreased risk of CaOx urolith formation, compared with diets with the lowest amounts. In contrast, canned diets with the highest amount of carbohydrate were associated with an increased risk of CaOx urolith formation. Conclusions and Clinical Relevance—Feeding canned diets formulated to contain high amounts of protein, fat, calcium, phosphorus, magnesium, sodium, potassium, chloride, and moisture and a low amount of carbohydrate may minimize the risk of CaOx urolith formation in dogs. (Am J Vet Res 2002;63:163–169)

oxalate (CaOx).1 In 1992, 24% (2,379/10,000) of uroliths submitted were composed of CaOx,2 and by 1999, this percentage had more than quadrupled to 31% (24,267/77,191).3 The increasing diagnosis of CaOx uroliths in dogs presents a challenge to clinicians, because treatment is difficult. Effective medical protocols for dissolution of these uroliths have not been developed, in part because of incomplete knowledge regarding the cause of this disease.4 Although surgery usually provides effective short-term treatment if all uroliths are removed, persistence of underlying causes of urolith formation often results in recurrence. Because a reproducible model for study of CaOx urolithiasis in dogs has not been identified, we have directed our efforts toward evaluation of the epidemiologic features of this naturally occurring disease. Identification and elimination of risk factors associated with calcium oxalate urolith formation in dogs would likely improve urolith detection, treatment, and prevention. Recently, we reported an association between CaOx uroliths and overweight middle-aged (8- to 12year-old) castrated male dogs of several breeds (Miniature and Standard Schnauzers, Lhasa Apso, Yorkshire Terrier, Bichon Frise, Shih Tzu, and Miniature and Toy Poodles).5 Results of epidemiologic studies of CaOx urolithiasis in cats6 and human beings7 indicate that diet-related factors may also influence the risk of CaOx urolithiasis. The purpose of the study reported here was to identify factors in commercially available canned diets associated with the development of CaOx uroliths in dogs. Materials and Methods

I

n 1986 at the Minnesota Urolith Center, 7% (57/839) of uroliths from dogs were composed of calcium

Received Jun 21, 2001. Accepted Sep 20, 2001. From the Minnesota Urolith Center, Department of Small Animal Clinical Sciences, College of Veterinary Medicine, University of Minnesota, St. Paul, MN 55108 (Lekcharoensuk, Osborne, Lulich, Ulrich, Koehler, Carpenter, Swanson); Department of Medicine, Faculty of Veterinary Science, Chulalongkorn University, Bangkok, Thailand 10330 (Pusoonthornthum); and Hill’s Science and Technology Center, 1035 NE 43rd St, Topeka, KS 66617 (Kirk). Dr. Lekcharoensuk’s present address is Department of Medicine, Faculty of Veterinary Medicine, Kasetsart University, Nakhonpathom, Thailand 73140. Supported in part by a grant from Hill’s Science and Technology Center. Address reprint requests to Dr. Osborne. AJVR, Vol 63, No. 2, February 2002

Case selection—Case dogs consisted of a subset of those dogs described in a previous report.5 Briefly, case dogs were those that developed CaOx uroliths submitted to the Minnesota Urolith Center for quantitative analysis between 1990 and 1992 while fed a commercially available canned diet. Only uroliths retrieved from dogs residing in the United States and Canada were included. To enhance the opportunity for staff of veterinary hospitals to submit uroliths, analysis was provided without charge. Uroliths were analyzed by the use of optical crystallography and, when necessary, infrared spectroscopy. Uroliths from all dogs included in this study were composed of at least 70% CaOx. Dogs fed dry, semimoist, or combinations of dry, semimoist, and canned foods were not evaluated. To evaluate the long-term effects of commercially available canned diets, dogs that consumed a primary brand of canned food for < 6 months were also excluded from this study. To minimize confounding effects associated with recent treatment for urinary tract disease, dogs with a 163

01-07-0173R.qxd

1/8/2002

10:10 AM

Page 164

history of any type of upper or lower urinary tract disease were also excluded. Likewise, dogs consuming diets specifically formulated for treatment of urinary tract disease were excluded. Because the composition of diets fed to immature dogs commonly changes during growth, and because CaOx uroliths adversely affect < 1% of dogs that form CaOx uroliths,8 dogs < 1 year of age were also excluded. Control selection—Control dogs also consisted of a subset of those dogs described in a previous report.5 Briefly, the control population was selected from dogs without urinary tract disease evaluated at the same veterinary hospital just prior to or immediately after each case dog. Control dogs were also fed canned food diets, and exclusion criteria for this group were the same as those for case dogs. Questionnaire design and administration—After CaOx uroliths were received at the Minnesota Urolith Center, a content-validated9 multiple-choice questionnairea designed to collect information about each dog’s signalment, diet (brands, quantities, and types fed and duration of consumption), and medical history (present and previous illnesses, treatments, and therapeutic diets) at the time of urolith detection was mailed to each owner with the permission of the primary care veterinarian. The same questionnaire was also mailed to owners of control dogs. If owners did not respond, they were mailed reminder cards or contacted by telephone. Diet evaluation—The questionnaire allowed owners to

list up to 3 brands of commercial canned foods consumed by their dogs. Owners were asked to specify the quantity of each brand of diet fed to each dog. On the basis of this information, the brand of canned food fed in the largest quantity was designated the primary brand. When equal amounts of 2 or more brands were fed or the amount fed was unknown, the first canned diet listed was designated the primary brand. However, the amount fed listed by the owner was not used to determine the quantity consumed by each dog for data analysis. Instead, to minimize recall bias, the quantity of each canned diet consumed by each dog was estimated on the basis of the daily adult maintenance caloric requirement for dogs. Information regarding the urine acidifying potential (ie, urine pH after feeding diet) and the quantity of components (eg, protein, carbohydrate, fat, fiber, calcium, phosphorus, magnesium, sodium, potassium, chloride, and moisture) in each diet was supplied by the manufacturers. Mean values were typically used for data analysis. However, when manufacturers reported a range of values, the median value was used. The quantity of protein, carbohydrate, fat, and fiber was expressed in g/100 kcal, whereas the quantity of calcium, phosphorus, magnesium, sodium, potassium, and chloride was expressed in milligram per kilocalories. Moisture was expressed as the percentage of water in each diet. Statistical analyses—The quantity of each dietary component fed to case and control dogs was compared by use of a t-test for 2 independent samples.10 Correlations among dietary components were assessed by use of the Pearson cor-

Table 1—Differences in dietary components and urine acidifying potential of canned food diets fed to dogs with (cases) and without (controls) calcium oxalate (CaOx) urolithiasis Controls

Cases

Component

No.

Mean SD

No.

Mean SD

t-test

P value*

Protein (g/100 kcal) CHO (g/100 kcal) Fat (g/100 kcal) Fiber (g/100 kcal) Ca (mg/kcal)† P (mg/kcal)

174 174 174 174 174 174

8.15 2.09 5.44 4.47 5.96 1.54 0.97 1.87 3.39 1.32 2.51 1.04

117 117 117 117 117 117

7.11 2.68 8.16 5.38 5.18 1.71 1.68 3.11 2.63 1.32 1.88 1.03

3.52 –4.52 4.05 –2.22 4.82 5.04

0.001 0.001 0.001 0.028 0.001 0.001

Mg (mg/kcal) Na (mg/kcal) K (mg/kcal) Cl (mg/kcal) Moisture (%) UAP (pH)‡

173 174 173 115 174 38

0.26 0.20 1.62 0.90 2.15 0.57 1.88 0.84 75.98 2.35 6.78 0.27

117 117 117 94 117 52

0.29 0.38 1.27 0.78 1.91 0.78 1.69 0.61 74.81 2.57 6.69 0.32

–0.94 3.43 2.88 1.84 4.04 1.38

0.349 0.001 0.004 0.068 0.001 0.171

*Significance set at P 0.05. †mg/kcalⴜ10 = g/100 kcal. ‡Urine pH after feeding diet. CHO = Carbohydrate. Ca = Calcium. P = Phosphorus. Mg = Magnesium. Na = Sodium. K = Potassium. Cl = Chloride. UAP = Urine acidifying potential.

Table 2—Pearson correlation coefficients between dietary components and urine acidifying potential of canned food diets fed to 117 dogs with and 174 dogs without CaOx urolithiasis Variable

Protein

CHO

Fat

Protein CHO Fat Fiber Ca P Mg Na K Cl Moisture UAP

Fiber

1 –0.46 0.29 0.22 0.66 0.67

— 1 –0.93 0.48 –0.52 –0.54

— — 1 –0.49 0.41 0.45

— — — 1 –0.21 –0.21

0.44 0.63 0.77 0.46 0.76 –0.06*

0.02* –0.40 –0.40 –0.03* –0.50 –0.16*

–0.18 0.33 0.24 –0.07* 0.40 0.13*

0.23 –0.17 0.40 0.04* 0.02* –0.53

Ca

P

Mg

Na

K

Cl

— — — — 1 0.98

— — — — — 1

— — — — — —

— — — — — —

— — — — — —

— — — — — —

— — — — — —

— — — — — —

1 0.10* 0.21 0.25 0.38 –0.06*

— 1 0.35 0.86 0.69 –0.22

— — 1 0.37 0.46 –0.19*

— — — 1 0.11* –0.03*

— — — — 1 0.07*

— — — — — 1

0.08* 0.81 0.35 0.36 0.80 0.40

0.05* 0.80 0.37 0.31 0.79 0.03*

Moisture UAP

*Not significantly (P 0.05) different from 0.

164

AJVR, Vol 63, No. 2, February 2002

01-07-0173R.qxd

1/8/2002

10:10 AM

Page 165

relation method.11 On the basis of guidelines recommended by Newton et al,12 we interpreted correlation coefficients (r) as follows: r = 0 to 0.29 or 0 to –0.29, no correlation; r = 0.3 to 0.69 or –0.3 to –0.69, weak correlation; and r = 0.7 to 1.0 or –0.7 to –1.0, strong correlation. To calculate crude odds ratios (OR) and 95% confidence intervals (CI), dietary components fed to case and control dogs were grouped into quartiles on the basis of amount and analyzed as categorical variables by use of univariate logistic regression, using the logarithmic approximation technique (Woolf method).13 In this context, the lowest quartile for each dietary component was used as a basis for comparison between groups. Because it was impractical for participating veterinarians to match breed, age, sex, neutering status, and body condition of each case with a corresponding control, the influence of these nondietary factors on each dietary component on CaOx urolith formation was assessed by use of multivariate logistic regression. Odds ratios and 95% CI for each dietary component adjusted for breed, age, sex, neutering status, and body condition were calculated by including these covariates in the multivariate logistic regression model. Additionally, OR were assessed by use of χ2 analysis to examine linear Table 3—Odds ratios (OR) and 95% confidence intervals (CI) for risk associated with amounts of dietary components or urine acidifying potential of canned foods, expressed in quartiles, and development of CaOx urolithiasis in dogs* Univariate analysis Component† Protein (g/100 kcal) 1.48–6.25 6.26–8.52 8.53–9.14 9.15–14.22 P for trend CHO (g/100 kcal) 0.13–2.50 2.51–3.55 3.56–10.21 10.22–18.22 P for trend Fat (g/100 kcal) 2.84–4.11 4.12–6.31 6.32–6.89 6.90–8.26 P for trend Fiber (g/100 kcal) 0.08–0.23 0.24–0.31 0.32–1.28 1.29–10.91 P for trend Moisture (%) 70.0–73.2 73.3–76.1 76.2–77.8 77.9–82.0 P for trend UAP (pH) 6.10–6.59 6.60–6.74 6.75–6.99 7.00–7.40 P for trend

Multivariate analysis‡

Median

OR

95% CI

OR

95% CI

4.52 7.50 8.53 9.32 NA

1.00 0.32 0.24 0.27 0.001

Reference 0.16–0.65 0.11–0.52 0.14–0.51 NA

1.00 0.43 0.14 0.21 0.001

Reference 0.17–0.98 0.05–0.36 0.10–0.47 NA

2.50 3.55 9.52 13.19 NA

1.00 0.96 0.83 3.48 0.001

Reference 0.39–2.35 0.37–1.83 1.50–8.07 NA

1.00 0.49 0.80 2.62 0.001

Reference 0.17–1.40 0.32–1.97 1.001–6.88 NA

3.04 5.29 6.32 6.90 NA

1.00 0.53 0.27 0.28 0.001

Reference 0.27–1.05 0.12–0.61 0.14–0.55 NA

1.00 0.58 0.17 0.25 0.002

Reference 0.26–1.31 0.06–0.44 0.11–0.57 NA

0.22 0.31 0.47 2.35 NA

1.00 0.41 0.71 0.67 0.958

Reference 0.19–0.91 0.33–1.49 0.32–1.42 NA

1.00 0.31 0.71 0.41 0.104

Reference 0.12–0.79 0.30–1.69 0.17–0.99 NA

72.4 74.0 77.0 77.9 NA

1.00 0.71 0.22 0.28 0.001

Reference 0.43–0.97 0.09–0.24 0.16–0.39 NA

1.00 0.43 0.09 0.16 0.001

Reference 0.19–0.97 0.03–0.24 0.07–0.39 NA

6.25 6.60 6.99 7.00 NA

1.00 0.26 0.27 0.37 0.343

Reference 0.07–1.03 0.04–2.01 0.10–1.30 NA

1.00 0.29 0.10 0.34 0.304

Reference 0.05–1.51 0.01–1.16 0.08–1.48 NA

*Data were obtained from 174 dogs without urinary tract disease and 117 dogs with calcium oxalate urolithiasis. †The range of values for each quartile are listed. ‡Adjusted OR for purebred (purebred/mixed), age ( 8 years old/ 8 years old), sex (male/female), reproductive status (neutered/sexually intact), and overweight (yes/no). NA = Not applicable.


trends across quartiles of dietary components.14 Multivariate logistic regression analysis incorporating multiple dietary components adjusted for the aforementioned covariates was not evaluated because of multicollinearity between dietary components and insufficient sample size. Odds ratio estimations were considered significant if the 95% CI for the OR did not include 1.0.15 On the basis of recommendations of Lilienfeld et al,16 we classified significant associations that yielded OR between 1.1 and 1.9, or between 0.5 and 0.9, as weak associations. We interpreted all significant risk (OR > 2) and protective (OR < 0.5) factors as clinically (biologically) significant. However, such relationships would require further experimental or prospective study to support causality.

Results One thousand seven hundred seventeen questionnaires were sent to owners of dogs with CaOx uroliths, and 2,852 questionnaires were sent to owners of dogs without urinary tract disease. Response rate was 85% (1,454/1,717) for owners of case dogs and 75% (2,147/2,852) for owners of control dogs. On the basis of exclusion criteria and availability of dietary component information, information for 117 dogs with CaOx uroliths and 174 dogs without urinary tract diseases was included for statistical analyses. Table 4—Odds ratios and 95% CI determined for the risk associated with amount of minerals and electrolytes in canned foods, expressed in quartiles, and development of CaOx urolithiasis in dogs* Univariate analysis Component† Ca (mg/kcal) 0.50–1.69 1.70–3.19 3.20–4.39 4.40–5.40 P for trend P (mg/kcal) 0.30–1.39 1.40–2.19 2.20–3.09 3.10–4.40 P for trend Mg (mg/kcal) 0.03–0.18 0.19–0.21 0.22–0.31 0.32–2.53 P for trend Na (mg/kcal) 0.20–0.79 0.80–1.19 1.20–2.49 2.50–5.10 P for trend K (mg/kcal) 0.60–1.59 1.60–1.99 2.00–2.39 2.40–4.20 P for trend Cl (mg/kcal) 0.80–1.49 1.50–1.59 1.60–1.69 1.70–4.90 P for trend

Median

OR

95% CI

Multivariate analysis‡ OR

95% CI

1.60 2.10 3.50 5.10 NA

1.00 0.33 0.20 0.17 0.001

Reference 0.16–0.69 0.09–0.43 0.08–0.37 NA

1.00 0.34 0.15 0.15 0.001

Reference 0.14–0.82 0.06–0.38 0.06–0.39 NA

1.10 1.70 2.50 3.90 NA

1.00 0.40 0.16 0.15 0.001

Reference 0.19–0.82 0.07–0.36 0.07–0.34 NA

1.00 0.43 0.09 0.12 0.001

Reference 0.18–1.03 0.03–0.26 0.05–0.33 NA

0.13 0.19 0.25 0.35 NA

1.00 0.33 0.79 0.44 0.088

Reference 0.17–0.67 0.39–1.59 0.24–0.83 NA

1.00 0.28 1.06 0.31 0.005

Reference 0.12–0.65 0.46–2.45 0.14–0.66 NA

0.70 1.00 1.30 2.50 NA

1.00 0.35 0.44 0.21 0.001

Reference 0.17–0.73 0.22–0.90 0.10–0.44 NA

1.00 0.43 0.31 0.18 0.001

Reference 0.18–1.02 0.13–0.74 0.07–0.45 NA

1.20 1.99 2.39 2.60 NA

1.00 0.33 0.18 0.23 0.001

Reference 0.14–0.76 0.08–0.39 0.11–0.47 NA

1.00 0.55 0.14 0.19 0.001

Reference 0.20–1.48 0.05–0.38 0.08–0.45 NA

1.20 1.50 1.69 2.00 NA

1.00 0.52 0.33 0.29 0.003

Reference 0.21–1.34 0.12–0.86 0.13–0.67 NA

1.00 0.62 0.32 0.24 0.001

Reference 0.21–1.81 0.10–0.96 0.09–0.63 NA

See Table 3 for key.

165

01-07-0173R.qxd

1/8/2002

10:10 AM

Page 166

Compared with controls, case dogs were fed diets containing either less protein, fat, calcium, phosphorus, sodium, potassium, or moisture or more carbohydrate or fiber (Table 1). However, these differences may have been confounded by correlations between dietary components (Table 2). Results of univariate logistic regression analyses for crude OR of dietary components expressed as quartiles (P for trend < 0.05) indicated that dogs fed diets with low quantities of protein, fat, calcium, phosphorus, sodium, potassium, chloride, or moisture or a high quantity of carbohydrate had an increased risk for developing CaOx uroliths (Table 3 and 4). We observed significant differences in breed, age, sex, neutering status, and body condition between case and control dogs (Table 5). Therefore, we performed multivariate logistic regression analysis adjusted for these covariates (Table 3 and 4). Dogs fed diets with either low quantities of protein, fat, calcium, phosphorus, magnesium, sodium, potassium, chloride, or moisture and high quantities of carbohydrate had an increased risk of developing CaOx uroliths (P for trend < 0.05). The magnitude of association between dogs with CaOx uroliths and various dietary components Table 5—Patient and demographic characteristics of 117 dogs with CaOx uroliths (cases) and 174 dogs without urinary tract disease (controls) Characteristics

No. of cases

No. of controls

OR

95% CI

93 24

126 48

1.5 1

0.8–2.6 Reference group

86 31

88 86

2.7 1


91 26

76 98

4.5 1


97 20

127 47

1.8 1


65 52

46 128

3.5 1


Breed Pure Mixed Age 8y 8y Sex Male Female Reproductive status Neutered Sexually intact Body condition Overweight Not overweight

Table 6—Summary of the magnitude of associations between dietary factors in canned food and development of calcium oxalate uroliths in dogs Factor (highest quartile)*

Risk

Odds ratio

Protein CHO Fat Fiber Ca P

4.8 times less likely 2.6 times more likely 4.0 times less likely 4.0 times less likely 6.7 times less likely 8.3 times less likely

0.21† 2.62† 0.25† 0.41 0.15† 0.12†

Mg Na K Cl Moisture UAP

3.2 times less likely 5.6 times less likely 5.3 times less likely 4.2 times less likely 6.3 times less likely NS

0.31† 0.18† 0.19† 0.24† 0.16† NS

*Comparison of the highest quartile with the lowest quartile. †Significant (P for trend 0.05) linear association. NS = Not significant. See Table 1 for key.

166

was greatest for phosphorus, followed by calcium, moisture, sodium, potassium, protein, chloride, fat, magnesium, and carbohydrate (Table 6). Discussion Results of our study indicate that diets with high amounts of protein, fat, calcium, phosphorus, magnesium, sodium, potassium, chloride, or moisture were associated with a decreased risk for CaOx urolith formation in dogs. In contrast, diets with a high carbohydrate content were associated with an increased risk for CaOx urolith formation. We did not observe a significant association between CaOx urolith formation and dietary fiber content. Moreover, our results did not reveal a significant association between diets formulated to maximize urine acidity and an increased risk for CaOx urolith formation. However, this latter result should be interpreted with caution, because sample size was small. It is generally believed that consumption of highprotein diets increases the risk of CaOx urolith formation in humans17 and dogs8 by promoting acidosis and subsequent hypercalciuria. Therefore, our observation that dogs fed diets high in protein (9.2 to 14.2 g/100 kcal) were 4.8 times less likely to develop CaOx uroliths than dogs fed diets low in protein (1.5 to 6.3 g/100 kcal) was unexpected. One plausible mechanism that may explain, at least in part, why CaOx uroliths formed less often in dogs fed diets high in animal protein is that such diets contain relatively high quantities of potassium.18 Also, the high-protein diets in our study contained more potassium (r = 0.77) than the low-protein diets. Results of our study revealed that dogs fed diets high in potassium were 5.3 times less likely to develop CaOx uroliths, compared with dogs fed diets low in potassium. The high-protein diets in our study also contained more moisture (r = 0.76) than the low-protein diets. Thus, consumption of high-protein diets could be expected to enhance renal excretion of water, thereby reducing the concentration of mineral in urine. Consumption of a high-protein diet (13.7 g/100 kcal) increases water consumption, urine volume, and urinary phosphorus excretion in healthy cats but does not result in an increase in urinary calcium excretion.19 Increased urinary phosphorus excretion could conceivably reduce the risk for CaOx urolithiasis by increasing the urine concentration of pyrophosphate (a CaOx crystal inhibitor)20 and reducing production of calcitriol by the kidney.21 In our study, dogs fed diets high in carbohydrate (10.2 to 18.6 g/100 kcal) were 2.6 times more likely to develop CaOx uroliths than dogs fed diets low in carbohydrate (0.1 to 2.5 g/100 kcal). Consumption of high-carbohydrate diets by humans may increase the risk of CaOx urolithiasis by increasing calcium excretion.22 In healthy dogs23 and humans,24 transient hypercalciuria has been observed in response to dietary carbohydrate. Postprandial increases in plasma insulin concentration, which in turn impairs proximal renal tubular reabsorption of calcium, are suggested as a plausible explanation of this phenomenon. However, further studies are required to investigate the role of dietary carbohydrate in CaOx urolithiasis in dogs. AJVR, Vol 63, No. 2, February 2002

01-07-0173R.qxd

1/8/2002

10:10 AM

Page 167

In our study, dogs fed diets high in fat (6.9 to 8.3 g/100 kcal) were 4 times less likely to develop CaOx uroliths than dogs fed diets low in fat (2.8 to 4.1 g/100 kcal). One plausible explanation is that diets high in fat were low in carbohydrate (r = –0.93). However, rats fed high-fat diets were at an increased risk for CaOx urolithiasis. It was postulated that this was a result of increased intestinal absorption and renal excretion of oxalic acid as a result of formation of insoluble nonabsorbable fatty acid-calcium salts in the intestinal lumen. We did not observe a significant association between dietary fiber content and CaOx urolith formation. However, consumption of rice and soy bran by healthy people significantly reduces urinary calcium excretion.25,26 This effect is attributed, at least in part, to a reduction in intestinal absorption of calcium as a result of rapid movement of ingesta through the gastrointestinal tract.27 Further, the rate of CaOx urolith formation decreased in humans that consumed large quantities of wheat bran.27,b However, results of another study26 of healthy humans suggest that the physical and chemical properties of different types of bran in diets have different effects on urinary excretion of calcium and oxalic acid.26 Our study was not designed to identify quantities of specific types of fiber fed to case and control dogs. For decades, the prevailing consensus was that restriction of dietary calcium would reduce urinary calcium excretion and, therefore, formation of CaOx uroliths. However, results of recent epidemiologic studies28,29 in humans reveal that restriction of dietary calcium is actually a risk factor for CaOx urolith formation. Humans who consumed higher quantities of dietary calcium had a reduced risk for development of CaOx urolithiasis, compared with those who consumed low quantities. It was postulated that consumption of a diet high in calcium increases formation of nonabsorbable CaOx in the intestinal lumen, which results in reduced urinary excretion of oxalic acid. Results of the present study support these findings, in that dogs fed diets high in calcium (4.4 to 5.4 mg/kcal) were 6.7 times less likely to develop CaOx uroliths than dogs fed diets low in calcium (0.5 to 1.7 mg/kcal). Oral consumption of phosphorus has also been hypothesized to reduce the risk of CaOx urolithiasis. In humans, the hypocalciuric effect of dietary phosphorus is well established.30 In fact, humans with CaOx uroliths are often given neutral phosphate supplements to reduce urinary calcium excretion, CaOx crystalluria, and recurrence of CaOx uroliths.31,32 In addition, salts of orthophosphate enhance urinary excretion of pyrophosphates and citrate.33,34 Pyrophosphates and citrate are CaOx-crystallization inhibitors. In contrast, diets deficient in phosphorus may stimulate calcitriol production, which in turn promotes intestinal absorption of calcium and phosphorus.32 Also, diets deficient in phosphorus may enhance intestinal absorption of calcium that has not combined with phosphorus to form an insoluble salt.8 Our results are consistent with these observations in that dogs fed diets high in phosphorus (3.1 to 4.4 mg/kcal) were 8.3 times less likely to develop CaOx uroliths than dogs fed diets low in phosphorus (0.3 to 1.4 mg/kcal). AJVR, Vol 63, No. 2, February 2002

On the basis of several lines of evidence, magnesium has been hypothesized to inhibit the formation of CaOx uroliths. For example, results of in vitro studies revealed that addition of magnesium to synthetic and human urine reduced CaOx supersaturation by combining with oxalic acid.35,36 In addition, supplemental magnesium has been reported to prevent CaOx urolithiasis in humans,37 although the benefits have not been clearly substantiated by results of controlled clinical trials. Our results are consistent with this hypothesis in that dogs fed diets high in magnesium (0.32 to 2.53 mg/kcal) were 3.2 times less likely to develop CaOx uroliths than dogs fed diets low in magnesium (0.03 to 0.19 mg/kcal). However, these observations should be interpreted cautiously, because consumption of excessive magnesium may also be a risk factor for CaOx urolith formation by inducing hypercalciuria. Oral administration of magnesium oxide to humans with CaOx uroliths is associated with an increase in urinary calcium excretion.38 Additionally, when 6 healthy dogs consumed a diet containing 2.5 mg of magnesium/kcal, urinary calcium excretion was 5 times greater than that observed when the same dogs consumed the same diet containing only 0.2 mg of magnesium/kcal.c Unexpectedly, our results indicated that diets high in sodium (2.5 to 5.1 mg/kcal) were associated with a decreased risk for CaOx urolith formation. Dogs fed high-sodium diets were 5.6 times less likely to develop CaOx uroliths than dogs fed low-sodium diets (0.2 to 0.8 mg/kcal). The sodium content in many adult canine maintenance diets recommended in the United States has been approximated to range between 0.4 and 1 mg/kcal.39 The association that we observed was unexpected, because results of previous studies of healthy adult humans28,40,41 and healthy adult dogsc were interpreted to indicate that high dietary sodium consumption increased the risk for CaOx urolith formation by promoting hypercalciuria. However, results of a recent study42 designed to assess the short-term effect of consumption of high- (4 mg of sodium/kcal) and low- (0.4 mg of sodium/kcal) sodium diets on urine calcium concentration in healthy adult dogs emphasized the need to be cautious about formulating generalizations regarding the effects of dietary sodium on urinary calcium concentration. In that study, urine concentration of sodium was significantly greater in dogs that consumed a high-sodium diet. However, urinary concentrations of calcium and oxalic acid were not significantly affected by dietary sodium concentration.42 A plausible unifying explanation of these apparently conflicting results is that although increased dietary sodium intake does increase urinary calcium excretion, it may decrease urine calcium concentration by increasing urine volume. Conceivably, oxalic acid concentrations could also be reduced. Even if the total quantity of urinary calcium excretion per day increased, CaOx uroliths would not be expected to form in urine unless that urine is oversaturated with calcium and oxalic acid.43 Appropriately designed studies evaluating dogs with CaOx urolithiasis are required to test the hypothesis that consumption of a high quantity of dietary sodium results in increased urinary 167

01-07-0173R.qxd

1/8/2002

10:10 AM

Page 168

calcium excretion but reduced urine calcium and oxalic acid concentration. Our results also indicated that diets high in potassium (2.4 to 4.2 mg/kcal) were associated with a decreased risk for CaOx urolith formation. Dogs fed high-potassium diets were 5.3 times less likely to develop CaOx uroliths than dogs fed low-potassium diets (0.6 to 1.6 mg/kcal). A reduced risk of CaOx urolith formation associated with dietary potassium content may be related to potassium-induced alterations in urinary calcium excretion. Results of a study40 of healthy adult humans indicated that reducing dietary potassium intake by substituting KCl with NaCl or KHCO3 with NaHCO3 is accompanied by increased urinary calcium excretion. Additionally, in another study44 of healthy adult humans, potassium supplementation reduced calcium excretion. If a similar effect occurs in dogs, it would provide a plausible explanation as to why dogs fed a high-potassium diet were at a decreased risk of developing CaOx urolithiasis, compared with those fed a low-potassium diet. In our study, dogs fed diets high in chloride (1.7 to 4.9 mg/kcal) were 4.2 times less likely to develop CaOx uroliths than dogs fed diets low in chloride (0.8 to 1.5 mg/kcal). Because several different types of chloride salts (ie, sodium, potassium, calcium, choline [vitamin B4], and pyridoxine [vitamin B6]) are commonly used in many commercial diets,45associations between chloride and CaOx urolithiasis may be confounded. For example, in our study the quantity of dietary chloride was directly correlated to the quantity of dietary sodium (r = 0.86), and dietary sodium content was associated with a decreased risk for CaOx urolithiasis. Our results revealed that dogs fed diets high in moisture (78 to 82%) were 6.3 times less likely to develop CaOx uroliths than dogs fed diets low in moisture (70 to 73%). This finding is consistent with observations that, in dogs, increased water consumption associated with increased excretion of a larger volume of less concentrated urine is an effective strategy to minimize formation of uroliths if urinary mineral concentration is decreased.4 Results of several studies46-48 of clinically normal cats also revealed that consumption of high-moisture diets is associated with production of a greater volume of less concentrated urine, compared with consumption of low-moisture diets.46-48 The association between acidemia, aciduria, and CaOx urolithiasis may be attributable in part to the fact that acidemia promotes mobilization of carbonate and phosphate from bone to buffer hydrogen ion.49-51 Concomitant mobilization of bone calcium stores may result in hypercalciuria. In addition, induction of metabolic acidosis in dogs, humans, and rats results in hypocitraturia.52 Hypocitraturia may increase the risk of CaOx urolithiasis, because citrate is an inhibitor of CaOx crystal formation. Our results did not support the hypothesis that diets designed to maximize urine acidity increase the risk of CaOx urolithiasis in dogs. However, this unexpected finding may be related to the lack of information regarding the urine acidifying potential of most (201/291; 69%) of the diets evaluated in our study. Because our study was not designed to identify 168

underlying mechanisms by which various dietary components, singly or in combination, promote or prevent CaOx urolith formation, associations that we identified should not be interpreted as proof of a cause and effect relationship. However, these associations may used to help formulate hypotheses for future prospective studies designed to evaluate factors that may reduce the risk of CaOx urolithiasis in dogs. For example, we hypothesize that canned diets formulated to contain high protein, fat, calcium, phosphorus, magnesium, sodium, potassium, chloride, and moisture concentrations and a low carbohydrate concentration may minimize the risk of development of CaOx uroliths. However, before food manufacturers adopt these hypotheses, they must be investigated further by appropriately designed clinical studies of dogs with CaOx urolithiasis. Such studies should include methods to evaluate potential interactions of multiple dietary components. In addition, to minimize confounding effects, such controlled studies should be designed to match dogs with and without CaOx urolithiasis on the basis of breed, age, sex, reproductive status, and body condition. Finally, such studies should incorporate strategies to minimize the recall bias of owners regarding types and quantities of foods fed to each dog. a

Questionnaire available from Dr. Jody P. Lulich, Department of Small Animal Clinical Sciences, College of Veterinary Medicine, University of Minnesota, St Paul, Minn. b Jarrar K, Graef V, Guttmann W. The use of wheat bran in the calcium oxalate metaphylaxis and the decrease of calcium excretion (abstr). Urol Res 1984;12:42. c Lulich JP. Canine calcium oxalate urolithiasis: etiology, pathophysiology, and therapy. PhD thesis, Department of Small Animal Clinical Sciences, University of Minnesota, St Paul, Minn, 1991.

References 1. Osborne CA, Clinton CW, Bamman LK, et al. Prevalence of canine uroliths: Minnesota Urolith Center. Vet Clin North Am Small Anim Pract 1986;16:27–44. 2. Osborne CA. The management of recurrent urolithiasis in patients with a history of calcium oxalate and struvite urolith formation. Partners In Pract 1992;1:2–3, 6. 3. Osborne CA, Lulich JP, Polzin DJ, et al. Analysis of 77,000 canine uroliths: Perspectives from the Minnesota Urolith Center. Vet Clin North Am Small Anim Pract 1999;29:17–38. 4. Lulich JP, Osborne CA, Lekcharoensuk C, et al. Canine calcium oxalate urolithiasis: case-based applications of therapeutic principles. Vet Clin North Am Small Anim Pract 1999;29:123–139. 5. Lekcharoensuk C, Lulich JP, Osborne CA, et al. Patient and environmental factors associated with calcium oxalate urolithiasis in dogs. J Am Vet Med Assoc 2000;217:515–519. 6. Kirk CA, Ling GV, Franti CE, et al. Evaluation of factors associated with development of calcium oxalate urolithiasis in cats. J Am Vet Med Assoc 1995;207:1429–1434. 7. Hesse A, Siener R. Current aspects of epidemiology and nutrition in urinary stone disease. World J Urol 1997;15:165–171. 8. Lulich JP, Osborne CA, Thumchai R, et al. Epidemiology of canine calcium oxalate uroliths: identifying risk factors. Vet Clin North Am Small Anim Pract 1999;29:113–122. 9. Portney LG, Watkins MP. Foundations of clinical research: applications to practice. Norwalk, Conn: Appleton & Lange, 1993; 69–86. 10. Snedecor GW, Cochran WG. Comparison of two samples. In: Statistical methods. 8th ed. Ames, Iowa: Iowa State University Press, 1980;83–106. 11. Snedecor GW, Cochran WG. Correlation. In: Statistical methods. 8th ed. Ames, Iowa: Iowa State University Press, 1980; 177–195. AJVR, Vol 63, No. 2, February 2002

01-07-0173R.qxd

1/8/2002

10:10 AM

Page 169

12. Newton RR, Rudestam KE. Your statistical consultant: answers to your data analysis questions. Thousand Oaks, Calif: Sage Publications Inc, 1999;247–275. 13. Woolf B. On estimating the relation between blood group and disease. Ann Hum Genet 1955;19:251–253. 14. Mantel N. Chi-square tests with one degree of freedom: extensions of the Mantel-Haenszel procedure. J Am Stat Assoc 1963; 58:690–700. 15. Fletcher RH, Fletcher SW, Wagner EH. Clinical epidemiology: the essentials. 3rd ed. Baltimore: The Williams & Wilkins Co, 1996;186–207. 16. Lilienfeld DE, Stolley PD. Foundations of epidemiology. 3rd ed. New York: Oxford University Press, 1994;198–225. 17. Heilberg IP. Update on dietary recommendations and medical treatment of renal stone disease. Nephrol Dial Transplant 2000;15:117–123. 18. Lewis LD, Morris ML Jr, Hand MS. Small animal clinical nutrition III. Topeka, Kan: Mark Morris Associates, 1987;1–28. 19. Funaba M, Hashimoto M, Yamanaka C, et al. Effects of a high-protein diet on mineral metabolism and struvite activity product in clinically normal cats. Am J Vet Res 1996;57:1726–1732. 20. Grases F, Conte A, Genestar C, et al. Inhibitors of calcium oxalate crystallization and urolithiasis. Urol Int 1992;48:409–414. 21. Swenson CL, Weisbrode SE, Nagode LA, et al. Age-related differences in phosphonoformate-induced bone toxicity in cats. Calcif Tissue Int 1991;48:353–361. 22. Leman J Jr, Piering WF, Lennon EJ. Possible role of carbohydrate-induced calciuria in calcium oxalate kidney-stone formation. N Engl J Med 1969;280:232–237. 23. DeFronzo RA, Goldberg M, Agus ZS. The effects of glucose and insulin on renal electrolyte transport. J Clin Invest 1976;58:83–90. 24. DeFronzo RA, Cooke CR, Andres R, et al. The effect of insulin on renal handling of sodium, potassium, calcium, and phosphate in man. J Clin Invest 1975;55:845–855. 25. Ohkawa T, Ebisuno S, Kitagawa M, et al. Rice bran treatment for hypercalciuric patients with urinary calcium calculous disease. J Urol 1983;129:1009–1011. 26. Jahnen A, Heynck H, Gertz B, et al. Dietary fiber: the effectiveness of a high bran intake in reducing renal calcium excretion. Urol Res 1992;20:3–6. 27. Shah PJR, Williams G, Green NA. Idiopathic hypercalciuria: its control with unprocessed bran. Br J Urol 1980;52:426–429. 28. Curhan GC, Willet WC, Speizer FE, et al. Comparison of dietary calcium with supplemental calcium and other nutrients as factors affecting the risk for kidney stones in women. Ann Intern Med 1997;126:497–504. 29. Curhan GC, Willett WC, Rimm EB, et al. A prospective study of dietary calcium and other nutrients and the risk of symptomatic kidney stones. N Engl J Med 1993;328:833–838. 30. Spencer H, Kramer L, Osis D, et al. Effect of phosphorus on the absorption of calcium and on calcium balance in man. J Nutr 1978;108:447–457. 31. Burdette DC, Thomas WC Jr, Finlayson B. Urinary supersaturation with calcium oxalate before and during orthophosphate therapy. J Urol 1976;115:418–422. 32. Pak CYC. Pathophysiology of calcium nephrolithiasis. In: Seldin DW, Giebisch G, eds. The kidney: physiology and pathophysiology. Vol 2. New York: Raven Press, 1992;2461–2480.


33. Coe FL, Favus MJ. Disorders of stone formation. In: Brenner BM, Rector FC, eds. The kidney. 2nd ed. Philadelphia: WB Saunders Co, 1981;1950–2007. 34. Smith LH. Medical treatment of idiopathic calcium urolithiasis. Kidney 1983;16:9–15. 35. Kohri K, Garside J, Blacklock NJ. The role of magnesium in calcium oxalate urolithiasis. Br J Urol 1988;61:107–115. 36. Lieske JC, Coe FL. Urinary inhibitors and renal stone formation. In: Coe Fl, Favus MJ, Pak CYC, et al, eds. Kidney stones: medical and surgical management. Philadelphia: Lippincott-Raven, 1996;65–113. 37. Johansson G, Backman U, Danielson BG, et al. Biochemical and clinical effects of the prophylactic treatment of renal calcium stones with magnesium hydroxide. J Urol 1980;124:770–774. 38. Fetner CD, Barilla DE, Townsend J, et al. Effects of magnesium oxide on the crystallization of calcium salts in urine in patients with recurrent nephrolithiasis. J Urol 1978;120:399–401. 39. Debraekeleer J, Gross KL, Zicker SC. Normal dogs. In: Hand MS, Thatcher CD, Remillard RL, et al, eds. Small animal clinical nutrition. Marceline, Mo: Walsworth Publishing Co, 2000;213–260. 40. Caldas AE, Gray RW, Lemann J. The simultaneous measurement of vitamin D metabolites in plasma: studies in healthy adults and in patients with calcium nephrolithiasis. J Lab Clin Med 1978;91:840–849. 41. Brannan PG, Morawski S, Pak CYC et al. Selective jejunal hyperabsorption of calcium in absorptive hypercalciuria. Am J Med 1979;66:425–428. 42. Stevenson AE, Wrigglesworth DJ, Markwell PJ. Dietary sodium chloride, urinary calcium and urinary oxalate in healthy adult dogs, in Proceedings. 9th Int Symp Urolithiasis 2000;794–796. 43. Bartges JW, Osborne CA, Lulich JP, et al. Methods for evaluating treatment of uroliths. Vet Clin North Am Small Anim Pract 1999;29:45–57. 44. Lemann J Jr, Pleuss JA, Gray RW, et al. Potassium administration increases and potassium deprivation reduces urinary calcium excretion in healthy adults. Kidney Int 1991;39:973–983. 45. Gross KL, Wedekind KJ, Cowell CS, et al. Nutrients. In: Hand MS, Thatcher CD, Remillard RL, et al, eds. Small animal clinical nutrition. Marceline, Mo: Walsworth Publishing Co, 2000; 21–107. 46. Palmore WP, Gaskin JM, Nielson JT. Effect of diet on feline urine. Lab Anim Sci 1978;28:551–558. 47. Anderson RS. Fluid balance and diet, in Proceedings. 7th Annu Kal Kan Symp Treat Small Anim Dis 1983;19–25. 48. Markwell PJ. Nutrition and aspects of feline lower urinary tract disease. J Small Anim Pract 1993;34:157–162. 49. Fettman MJ, Coble JM, Hamar DW, et al. Effect of dietary phosphoric acid supplementation on acid-base balance and mineral and bone metabolism in adult cats. Am J Vet Res 1992;53:2125–2135. 50. Dow SW, Fettman MJ, Smith KR, et al. Effects of dietary acidification and potassium depletion on acid-base balance, mineral metabolism and renal function in adult cats. J Nutr 1990;120:569–578. 51. Ching SV, Norrdin RW, Fettman MJ, et al. Trabecular bone remodeling and bone mineral density in the adult cat during chronic dietary acidification with ammonium chloride. J Bone Miner Res 1990;5:547–556. 52. Simpson DP. Citrate excretion: a window on renal metabolism. Am J Physiol 1983;244:F223–F234.

169