• Open Access

The Relationship between Serum Calcium Concentration and Outcome in Horses with Renal Failure Presented to Referral Hospitals

Authors

  • B. LeRoy,

    1. Department of Veterinary Pathology, College of Veterinary Medicine, University of Georgia, Athens, GA
    Current affiliation:
    1. Bristol, WI
    Search for more papers by this author
  • A. Woolums,

    Corresponding author
    1. Department of Large Animal Medicine, College of Veterinary Medicine, University of Georgia, Athens, GA
    • Department of Veterinary Pathology, College of Veterinary Medicine, University of Georgia, Athens, GA
    Search for more papers by this author
  • J. Wass,

    1. Quantum Cat Consultants, Lake Forest, IL
    Search for more papers by this author
  • E. Davis,

    1. Department of Anatomy and Physiology, College of Veterinary Medicine, Kansas State University, Manhattan, KS
    Search for more papers by this author
  • J. Gold,

    1. Department of Large Animal Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, College Station, TX
    Current affiliation:
    1. Mountain Horse Medical Center, Park City, UT
    Search for more papers by this author
  • J.H. Foreman,

    1. Department of Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois at Urbana-Champaign, Urbana-Champaign, IL
    Search for more papers by this author
  • K. Lohmann,

    1. Department of Large Animal Clinical Sciences, Western College of Veterinary Medicine, University of Saskatchewan, Saskatoon, SK
    Search for more papers by this author
  • J. Adams

    1. Department of Small Animal Medicine, College of Veterinary Medicine, University of Georgia, Athens, GA
    Current affiliation:
    1. Hull, GA
    Search for more papers by this author

Corresponding author: A. Woolums, Department of Large Animal Medicine, College of Veterinary Medicine, University of Georgia, Athens, GA 30602; e-mail: awoolums@uga.edu.

Abstract

Background

Hypercalcemia is common in horses with renal failure, but it is not known whether it impacts prognosis.

Hypothesis/Objectives

The primary objective of this study was to determine whether hypercalcemia was associated with decreased likelihood of survival to discharge in horses with renal failure. Secondary objectives were to determine whether hypercalcemia was more common in acute (ARF) or chronic renal failure (CRF), whether feeding alfalfa was associated with hypercalcemia, and whether serum creatinine concentration was associated with survival.

Animals

Medical records of 63 horses presented to referral hospitals for renal failure were evaluated. Cases were classified as ARF or CRF based on historical and clinical findings.

Methods

The distribution of hypocalcemic, normocalcemic, and hypercalcemic cases in the ARF and CRF groups was determined. Mean serum calcium and creatinine concentrations for survivors and nonsurvivors, and for ARF and CRF cases, were compared. Mean serum calcium concentrations for cases fed alfalfa or not fed alfalfa were compared.

Results

Hypercalcemia was significantly more common in CRF than ARF cases. CRF cases fed alfalfa were significantly more likely to be hypercalcemic. There was no significant difference in serum calcium concentration between survivors and nonsurvivors. Serum creatinine concentration was significantly higher in nonsurvivors and in ARF cases.

Conclusions and Clinical Importance

Horses with CRF are more likely to be hypercalcemic than horses with ARF. Hypercalcemia was not associated with outcome in renal failure cases in this study. Additional research on the impact of dietary calcium on long-term well-being in horses with CRF is warranted.

Abbreviations
ARF

acute renal failure

CRF

chronic renal failure

USG

urine specific gravity

Renal failure in horses may result from diseases that specifically target the kidneys or occur secondary to nonrenal pathology that leads to dysfunction of renal glomeruli, tubules, associated vasculature, or some combination of these. Laboratory testing plays an integral role in the identification of renal dysfunction. Most commonly, laboratory confirmation of renal failure is based on serum concentrations of blood urea nitrogen and creatinine above the upper limit of the reference range (ie, azotemia) in conjunction with submaximally concentrated urine based on urine specific gravity (USG).[1] In addition to azotemia and a low USG, laboratory abnormalities commonly present in horses with renal failure include derangements of serum calcium and phosphorous concentrations, with hypercalcemia being a relatively common although inconsistent finding in horses with renal failure.[2-7] Although horses with experimentally induced acute renal failure (ARF) secondary to bilateral nephrectomy or administration of nephrotoxic substances were found to develop hypercalcemia,[8, 9] hypercalcemia in naturally occurring disease appears to be more commonly associated with chronic renal failure (CRF).[6, 7]

It is not clear whether the presence of hypercalcemia has any prognostic relevance in horses with renal failure. The purpose of this retrospective study was to evaluate the relationship between serum calcium concentration at presentation and outcome in horses with renal failure presented to 1 of 5 referral hospitals: the University of Georgia Veterinary Teaching Hospital, the Kansas State University Veterinary Medical Teaching Hospital, the Texas A&M University Veterinary Medical Teaching Hospital, the University of Illinois Veterinary Teaching Hospital, and the Veterinary Teaching Hospital at the Western College of Veterinary Medicine at the University of Saskatchewan. The relationship between serum calcium concentration and a history of consuming alfalfa hay was also evaluated, as was the relationship between serum creatinine concentration at presentation and short-term outcome (defined as survival to hospital discharge).

Materials and Methods

Case Selection Criteria

Medical records were searched to identify azotemic horses (defined by increased serum creatinine concentration) with USG < 1.020 for which a serum calcium concentration was also measured. Results recorded for analysis were those that were first measured after admission. Only cases that had not received IV or PO fluids before measurement of USG were included in the analysis. Cases were defined as either acute or chronic based on the following definitions: ARF cases had a recent onset (<1 week) of clinical signs; they also may have had one or more of the following historical, clinical, or laboratory findings: marked pigmenturia associated with hemolytic anemia, evidence of rhabdomyolysis, rapid onset of oliguria, or microscopic evidence of acute renal tubular damage identified by renal biopsy or necropsy. Chronic renal failure cases had a history of weight loss unattributable to any other cause; they also may have had one or more of the following historical, clinical, or laboratory findings: moderate normocytic normochromic anemia, marked proteinuria, or microscopic evidence of severe glomerulonephritis, renal mineralization, or interstitial fibrosis, identified by renal biopsy or necropsy. Outcome was recorded as to whether the horses were discharged alive or not. Data from horses < 6 months of age were excluded from analysis. Whether horses were fed alfalfa before presentation or not was recorded; it was not possible to include the duration of time when alfalfa was fed in the analysis, because that information was not included in most records. Horses were divided into 3 categories designated as hypercalcemic, normocalcemic, or hypocalcemic based on comparison of their serum total calcium concentrations to the reference range at the the clinical pathology laboratory of the hospital where they were evaluated.

Statistical Analysis

Descriptive statistics for clinicopathologic variables measured in all cases (USG, serum calcium concentration, and serum creatinine concentration) and the distribution of each variable in survivors and nonsurvivors and for ARF and CRF cases were determined. Data distributions were tested for normality by the Shapiro–Wilk goodness of fit test, and because the distributions for USG, serum calcium concentration, and serum creatinine concentration were not normally distributed, results for survivors and nonsurvivors were compared by the Wilcoxon/Kruskal–Wallis (rank sums) Test. Values for ARF and CRF cases were compared by the Wilcoxon signed-rank test. Cases were defined as hypocalcemic, normocalcemic, or hypercalcemic based on the reference ranges for the hospitals where horses were presented. Calcium distributions in horses with ARF and CRF were compared by a contingency table, with Pearson chi-squared analysis used to compare distributions in ARF and CRF cases. The relationship between serum calcium concentration and alfalfa hay consumption was evaluated for ARF cases and CRF cases by one-way ANOVA, with nonparametric means comparisons for ARF. All analyses were completed by a statistical analysis software program.1 For all analyses, significance was set at P < .05.

Results

A total of 63 cases met the criteria for inclusion and had medical records that included all data required for the analysis. Thirty-two horses survived to discharge, whereas 31 did not. It was not possible in all cases to determine when nonsurvival was because of euthanasia, poor prognosis, or financial constraints. Twenty-four horses were determined to be in ARF, whereas 39 were determined to be in CRF. The descriptive data for serum total calcium concentration, serum creatinine concentration, and USG for all 63 cases are presented in Table 1, and the means for each value for survivors and nonsurvivors and for cases of ARF or CRF are presented in Table 2.

Table 1. Descriptive data for serum total calcium concentration, serum creatinine concentration, and USG (urine specific gravity) for the 63 cases evaluated to determine the relationship between serum calcium concentration and survival in horses with renal failure.
 First QuartileMedianThird QuartileRange (Min–Max)
Serum total  calcium  concentration  (mg/dL)11.312.314.26.9–21.6
Serum  creatinine  concentration  (mg/dL)3.46.29.81.6–24.7
USG1.0091.0121.0151.005–1.018
Table 2. Serum total calcium concentration, serum creatinine concentration, and urine specific gravity (USG) for horses surviving to discharge (“survival”) or not surviving to discharge (“nonsurvival”) and for horses with acute (ARF) or chronic renal failure (CRF).
Outcome or DiagnosisSerum Total Calcium Concentration (mg/dL) mean (SD)Serum Creatinine Concentration (mg/dL) mean (SD)USG mean (SD)
  1. SD, standard deviation.

  2. Within an outcome or diagnosis, different superscripts indicate that mean serum creatinine concentration was significantly higher for nonsurvivors than for survivors (P = .01); that mean serum calcium concentration was significantly lower for ARF cases than for CRF cases (P < .0001); and that mean serum creatinine concentration was significantly higher for ARF cases than for CRF cases (P = .02).

Survival  (n = 32)12.6 (2.9)6.1 (5.4)a1.012 (0.0)
Nonsurvival  (n = 31)13.1 (2.9)9.4 (5.4)b1.013 (0.0)
ARF (n = 24)11.4 (2.9)a8.7 (6.1)a1.012 (0.0)
CRF (n = 39)13.7 (2.6)b7.1 (5.3)b1.012 (0.0)

There was no significant difference in serum calcium concentration or USG between surviving and nonsurviving horses. In contrast, serum creatinine concentration was significantly higher in nonsurviving horses (P = .01; Table 2). Cases with ARF had significantly lower serum calcium concentration (P < .0001; Table 2) and significantly higher serum creatinine concentration (P = .02; Table 2) than cases with CRF.

Although serum calcium concentration was not significantly associated with outcome, comparison of the proportion of ARF and CRF cases that were hypocalcemic, normocalcemic, or hypercalcemic identified a significant difference in distribution of serum calcium concentrations for each diagnosis (Table 3, P = .001 for Pearson chi-squared analysis). Cases of ARF were significantly more likely to have low or normal serum calcium concentration: 29.2% of ARF cases were hypocalcemic, whereas 58.3% of cases were normocalcemic, and only 3 cases (12.5%) were hypercalcemic. In contrast, CRF cases were significantly more likely to have normal or high serum calcium concentration: 56.4% of CRF cases were hypercalcemic, 35.9% were normocalcemic, and 7.7% were hypocalcemic.

Table 3. Distribution of serum calcium status for 63 cases of acute (ARF) or chronic renal failure (CRF) presented to referral hospitals. Calcium status was defined as hypercalcemic, hypocalcemic, or normocalcemic based on individual hospital laboratory reference ranges.
DiagnosisHypercalcemia n (% of total)Hypocalcemia n (% of total)Normocalcemia n (% of total)Total
  1. Different superscripts for ARF and CRF indicate significantly different distribution of serum calcium status as measured by Pearson chi-squared analysis (P = .001). n = number of cases.

Acute renal failurea3 (12.5)7 (29.2)14 (58.3)24
Chronic renal failureb22 (56.4)3 (7.7)14 (35.9)39
Total25 (39.7)10 (15.9)28 (44.4)63

The relationship between alfalfa consumption and serum calcium concentration was evaluated for ARF and CRF cases. For ARF cases, there was no significant difference in serum calcium concentration for horses with a history of alfalfa consumption versus horses with no history of alfalfa consumption, although the power of the study to detect this difference was low. Mean (± standard deviation, SD) serum calcium concentration for ARF cases fed alfalfa was 12.5 ± 2.7, whereas the mean ± SD for ARF cases not fed alfalfa was 10.8 ± 2.7 (P = .13). However, for CRF cases, serum calcium concentration was significantly higher for horses with a history of alfalfa consumption than for those horses with no history of alfalfa consumption. Mean ± SD serum calcium concentration for CRF cases fed alfalfa was 14.7 ± 2.5, whereas the mean ± SD for CRF cases not fed alfalfa was 12.9 ± 2.4 (P = .03).

Discussion

Renal failure leads to abnormalities of calcium and phosphorus metabolism in many species by multiple mechanisms. In general, the serum calcium concentration in animals suffering from renal failure can be affected by 4 major factors[10]: (1) the ability of the renal tubular epithelium to retain or release calcium in appropriate response to circulating concentrations of parathyroid hormone (PTH) and the active form of vitamin D, 1,25-dihydroxy vitamin D3 (1,25(OH)2D3); (2) increased PTH secretion as a response to hyperphosphatemia-mediated decreases in ionized calcium concentration (renal secondary hyperparathyroidism); (3) decreased concentrations of 1,25(OH)2D3 as a result of diminished renal 1α-hydroxylase activity; and (4) dietary calcium intake. A 5th mechanism that has not been evaluated in horses is estrogen-mediated uptake of calcium in the renal tubule.[10]

In humans and most domestic species, renal failure most commonly leads to hyperphosphatemia and hypocalcemia, with hyperphosphatemia attributed to decreased glomerular filtration rate and resultant decreased excretion of phosphorus, and hypocalcemia attributed to the mass action effect associated with hyperphosphatemia and decreased activation of 1,25(OH)2D3. In a retrospective study of 8 human patients with ARF and hypercalcemia, in all cases, hypercalcemia could be attributed to other concurrent conditions, including neoplasia and vitamin D intoxication.[11]

In contrast to humans and other species, hypercalcemia with or without hypophosphatemia has frequently been described in horses with naturally occurring or experimentally induced renal failure.[2-5, 7, 12] These and other reports have led to a commonly held clinical impression that the horse in renal failure has a unique species-based predisposition to hypercalcemia, possibly because of intestinal or renal calcium homeostatic mechanisms which are peculiar to horses. Reports suggest that hypercalcemia is more likely in horses with CRF,[2, 4, 7] but hypercalcemia occurred rapidly in 5 ponies with experimentally induced ARF because of bilateral nephrectomy,[8] and was transiently present in 3 of 4 ponies with experimentally induced ARF because of mercuric chloride and potassium dichromate administration.[9, 13] In contrast, hypercalcemia was not present in any of the 6 horses evaluated for ARF resulting from hemodynamic causes.[14] Differences of opinion regarding whether CRF is more likely than ARF to induce hypercalcemia in horses[4, 12] may have been possible because of a lack of reliable evidence. Most studies have examined small numbers of cases, or did not use objective criteria for case selection, which may have biased results.[4, 5, 7-9] However, in a large retrospective survey of equine CRF cases at multiple institutions, hypercalcemia was present in 67% of 99 horses with CRF.[6] We studied the relationship between serum calcium concentration status and diagnosis in a relatively large number of cases that were selected in an unbiased manner. Our findings support the clinical impression that hypercalcemia is more common in horses with CRF than in horses with ARF. Anecdotally, hypercalcemia in CRF has been attributed to high dietary calcium intake, with serum calcium concentrations reportedly decreasing into the reference range when affected horses are switched to diets containing lower amounts of calcium.[4, 15] This study confirmed that serum calcium concentration was significantly higher in CRF cases with a history of alfalfa hay consumption when compared with CRF cases with no history of alfalfa hay consumption. Although hypercalcemia did not impact survival in this study, it may be prudent to formulate diets for horses with CRF with the minimum amount of calcium necessary; additional research to determine whether dietary calcium intake impacts long-term survival of horses with CRF is warranted. In contrast, serum calcium concentration was not significantly related to alfalfa hay consumption in horses with ARF, although this may have been attributable to low study power to detect a difference.

The hypercalcemia that occurs in some horses with renal failure is speculated to be related to intestinal or renal calcium homeostatic mechanisms peculiar to horses. Horses have indeed been found to differ from other species in some aspects of calcium regulation. Compared with ruminants consuming the same amount of dietary calcium, horses absorb more calcium from the intestine and excrete more calcium in the urine.[16] Serum ionized calcium concentrations are higher in horses than in humans and many other species, and the calcium set-point (the serum ionized calcium concentration at which PTH secretion is stimulated) is higher in horses than has been reported for humans or dogs.[17] In normal horses and ponies, plasma concentrations of inactive vitamin D (25(OH)D3) as well as active vitamin D (1,25(OH)2D3) were found to be low, relative to concentrations reported in swine and humans.[18] In addition, activity of 1α-hydroxylase, which converts 25(OH)D3 to 1,25(OH)2D3, could not be detected in equine kidneys,[18] leading investigators to speculate that vitamin D-mediated control of serum calcium concentration must differ in the horse when compared with other species. Whether these differences in equine calcium homeostasis are related to the occurrence of hypercalcemia in equine renal failure is currently not known. As horses excrete a relatively large amount of calcium through their kidneys, it has been proposed that hypercalcemia in horses with CRF is simply the result of ongoing intestinal absorption with impaired renal tubular excretion.[15]

The role of PTH in the genesis of hypercalcemia in equine renal failure has been investigated.[3, 9, 19] In 6 horses with naturally occurring renal disease, serum PTH concentration was measured by immunoassay for the carboxy terminal (inactive) fragment. In 3 horses with CRF, hypercalcemia, and hypophosphatemia, serum immunoreactive PTH (iPTH) was below the level of detection for the assay. The remaining 3 horses had normal serum calcium concentrations and normal or high serum iPTH. The results of this study suggested that increased concentrations of PTH were not responsible for the hypercalcemia in horses with CRF.[3] However, in ponies with experimentally induced ARF, immunoassay for the carboxy terminal fragment indicated high or normal concentrations of serum iPTH, when serum calcium concentrations were increased in 3 of 4 ponies. These investigators speculated that decreased renal clearance of PTH was the cause of increased iPTH concentrations, and that this may have contributed to the transient hypercalcemia observed.[9] A 3rd study that validated an immunoradiometric assay for the whole equine PTH (wPTH) found very low blood concentrations of wPTH in horses with CRF, suggesting that CRF does not cause hyperparathyroidism in horses.[19]

Identification of variables that predict prognosis, and that can be measured at initial hospital presentation of renal failure patients, could be useful for clinicians helping clients to decide whether to pursue therapy or not. Although very high concentrations of serum creatinine have been anecdotally associated with poor prognosis by some clinicians, to our knowledge no evidence-based guidelines have been published that link serum creatinine concentration to prognosis. Given the common, but not invariable occurrence of hypercalcemia in horses with renal failure, and the possibility that persistent hypercalcemia could lead to renal tissue mineralization, it was considered logical to evaluate whether serum calcium concentrations were associated with outcome in horses with renal failure or not. In the cases evaluated here, there was no significant difference in serum calcium concentration between survivors and nonsurvivors; this may have been caused by a true lack of relationship between serum calcium concentration and outcome in horses with renal failure, or it may be that the number of cases included resulted in insufficient power to identify a relationship. In contrast, in the cases reported here, serum creatinine concentration was significantly higher in nonsurvivors. In addition, mean serum calcium concentration was significantly higher in horses with CRF than ARF, and the distribution of cases defined as normocalcemic, hypercalcemic, or hypocalcemic based on the first total serum calcium concentration measured at presentation was significantly different for ARF cases when compared with CRF cases. Cases of ARF were more likely to be normocalcemic or hypocalcemic, whereas CRF cases were more likely to be normocalcemic or hypercalcemic.

Acknowledgments

Special thanks to Ms Shay Bush and Mr Jeff Duke for assistance with laboratory database interrogation.

Note

  1. 1

    JMP 9.0; SAS Institute Inc, Cary, NC

Ancillary