rss
Veterinary Record 174:554 doi:10.1136/vr.102054
  • Research
  • Paper

Prevalence and types of hyponatraemia, its relationship with hyperglycaemia and mortality in ill pet rabbits

  1. A. Montesinos, MVSc, Licenciado en Veterinaria
  1. Centro Veterinario Los Sauces, Madrid, Spain
  1. E-mail for correspondence: cvsauces{at}cvsauces.com

Abstract

Prevalence of hyponatraemia has not been extensively studied in pet rabbits, and the reference data for calculated plasma tonicity and osmolarity are not available. This retrospective clinical study reports the prevalence of hyponatraemia, hyposmolarity and hypotonicity in ill pet rabbits (n=356). The relationship between sodium and glucose levels was studied (n=134). Mortality rates within seven days associated with different sodium levels were calculated in ill rabbits (n=322). Venous blood samples in lithium heparin were processed using iStat EC8+ cartridges. The 95% RI for plasma sodium, calculated osmolarity and tonicity from 51 healthy pet rabbits were 136–147 mEq/l, 284–312 mOsm/l and 278–302 mOsm/l, respectively. The prevalence of hyponatraemia, hypotonicity and hyposmolarity was 39.0 per cent, 28.7 per cent and 18.0 per cent, respectively. Pseudohyponatraemia was present in 28.1 per cent and true hyponatraemia was present in 71.9 per cent of the cases of hyponatraemia. Sodium levels less than 129 mEq/l were found to be associated with 2.3-fold increase in mortality risk. Plasmatic sodium levels in rabbits decrease in conditions of hyperglycaemia in a similar manner as it occurs in human beings. As hyperglycaemia is quite a common condition in rabbits, simultaneous measurement of plasmatic sodium along with glucose in ill rabbits is advised. Hyponatraemia is a common condition in ill rabbits and, depending on its type (true hyponatraemia or pseudohyponatraemia), of varying clinical relevance. Calculation of plasmatic tonicity is necessary for differentiation of types of hyponatraemia.

Introduction

Sodium is the most abundant extracellular cation in the body and the most responsible for plasmatic osmolality together with other solutes that remain mainly in the extracellular fluid (ECF), such as glucose or urea (Arieff and others 1976, DiBartola 2012). Hyponatraemia represents the most common electrolyte abnormality in hospitalised human beings (Ghali 2008, Laczi 2008). The decrease in sodium levels becomes pathologic when it leads to a state of hypotonicity of the ECF with tendency of free water to move into the intracellular space. The main harmful effects occur in the nervous system due to the risk of cerebral oedema. Symptoms depend on the rate of development and the degree of hyponatraemia, ranging from mild depression to stupor, convulsions, coma and death (Arieff and others 1976, Drobatz and Mandell 2000).

Plasma osmotic activity values published in literature refer to either real total plasma osmolality (measured with an osmometer) or ­calculated plasma osmolarity (using formulas with the patient's ­laboratory data). Several formulas have been proposed, the most used in ­veterinary medicine is (Nelson and others 2009, DiBartola 2012): Formula

Tonicity or effective plasma osmolality is the ability to initiate the movement of water between the extracellular and intracellular compartments, and it depends on the presence of non-permeant solutes (glucose and sodium) (DiBartola 2012). It more accurately describes the risk of water accumulation in the intracellular compartment. Tonicity may be estimated as (Nelson and others 2009, DiBartola 2012): Formula

Hyponatraemia aetiology can be classified based on plasma tonicity, the two main groups are: true hyponatraemia (hypotonic hyponatraemia) and pseudohyponatraemia (isotonic hyponatraemia and hypertonic hyponatraemia). Hypotonic hyponatraemia is usually associated with low total plasma osmolality except in cases of high uraemia or presence of ethanol or mannitol in blood. Pseudohyponatraemia is classified in isotonic (associated to hyperglycaemia, hyperproteinaemia and hyperlipidaemia) and hypertonic (also called ‘redistributive’, mainly associated with hyperglycaemia or mannitol administration) (Weisberg 1989, Milionis and others 2002, Nguyen and others 2007, Nelson and others 2009, Reddy and Mooradian 2009, DiBartola 2012).

Because the plasma tonicity is maintained within specific physiologic limits, sodium plasma levels decrease when glycaemia increases (pseudohyponatraemia). Therefore, the interpretation of natraemia value should take into consideration also the glycaemia value or the plasma tonicity (Hillier and others 1999, Nelson and others 2009, DiBartola 2012). It is essential to differentiate between true hyponatraemia and pseudohyponatraemia. Pseudohyponatraemia (isotonic or hypertonic) itself has no consequences for the patient's health: the clinician must identify it and avoid treating it; the underlying disease should actually be addressed. In human medicine, complications from pseudohyponatraemia have been associated mostly with subsequent medical interventions that were unnecessarily applied, such as diuresis, fluid restrictions and hypertonic saline administration (Illowsky and Laureno 1987, Bern 2006, DiBartola 2012). True hyponatraemia, on the other hand, is a serious clinical condition that must be correctly diagnosed and treated. The assessment of symptoms, degree and, when possible, duration of hyponatraemia is important, as the treatment protocol depends on these factors. Excessive rapid correction of true hyponatraemia leads to neurological complications and should be avoided (Illowsky and Laureno 1987, Reddy and Mooradian 2009).

To date, prevalence of hyponatraemia was not extensively studied in pet rabbits. This paucity is partially due to the fact that the devices for the electrolyte analysis were not readily available for general veterinary practitioners. In the last decade, the availability of easy-to-handle point-of-care blood gas and electrolyte analysers, such as iStat system, that requiere small sample volume makes these kinds of analysis more readily available for the clinicians in human and veterinary medicine (Schneider and others 1997, Steinfelder-Visscher and others 2008, Tausch 2011). The lower limit for the reference intervals of physiologic sodium levels using different methods varies up to 8 mEq/l from one to another literature source (Harcourt-Brown 2002, Meredith and Crossley 2002, Mader 2004, Hernandez-Divers 2005, Tausch 2011, Ardiaca and others 2013). Physiologic plasmatic sodium levels in pet rabbits using iStat handheld analyser were studied recently (Tausch 2011, Ardiaca and others 2013).

To the author's knowledge, the reference data for physiologic plasma tonicity and osmolality in pet rabbits are not available. Osmometers are expensive and not commonly available in pet practices, therefore, the calculated values for plasma osmolarity and tonicity are most frequently used in veterinary medicine. The previously published data for osmometer-determined plasma osmolality in laboratory rabbits (Whitlock and Wheeler 1964, Zornow and others 1987, Gerstberger and others 1992, Frosini and others 2000, Dontas and others 2011) have very limited application in clinical veterinary work with pet rabbit patients. The control groups from experimental studies are not always representative of a normal population in physiological condition, nor are they representative of the pet rabbit population. Subsequently, hyponatraemia most probably remains underdiagnosed and, in some cases, mismanaged in pet rabbits.

The aim of this retrospective clinical study was to assess the prevalence and types of hyponatraemia, as well as prevalence of hyposmolarity and hypotonicity in clinically ill pet rabbits attended in a veterinary centre. Secondarily, the prevalence of normal and high levels of natraemia, osmolarity and tonicity were calculated. For that purpose, the reference ranges for sodium, glucose, blood urea nitrogen (BUN), calculated osmolarity and tonicity using EC8+ cartridges for iStat system were established in healthy pet rabbits. Additionally, we evaluated the relationship between sodium and glucose levels in plasma in conditions of hyperglycaemia. Finally, the influence of dysnatraemia on mortality was assessed calculating the mortality rate associated with different sodium levels.

Materials and methods

Reference intervals: healthy pet rabbits

Venous samples from 51 clinically healthy rabbits (27 females and 24 males) aged from three months to 6.4 years presented for routine health examination, vaccination or elective surgery were gathered. The owners of all rabbits were informed and consented the sampling of their pets. All sampling from animals was part of routine blood work aimed to assess the clinical status of the patient. The study received the approval from an ethics review committee (Comité Ético de Bienestar Animal de GREFA Number 12/0001).

Clinical study: ill pet rabbits

The clinical study was based on 356 samples collected from ill rabbits (162 females and 194 males) aged from one month to nine years attended in a veterinary centre. All the animals were critically ill patients that had undergone blood-gas analysis on admission to the same veterinary centre. Rabbits with mild illness without perceivable alterations in their appetite, hydration and mental status were not included in this study as they are not candidates for blood-gas analysis.

The clinical study with ill pet rabbits comprised four parts:

  1. Prevalence of natraemia, tonicity and osmolarity deviations

  2. Prevalence of different types of hyponatraemia

  3. Association between natraemia and glycaemia

  4. Association between natraemia and outcome

For all the rabbits, healthy and ill, sex and age were recorded. Animals were divided in four groups by age (rabbits under four months of age; rabbits between four months and one year; ­rabbits between one year and four years and rabbits older than four years).

Sampling and analysis technique

In all cases, blood was collected from the lateral saphenous vein. Syringes pretreated with lithium heparin were used as previously described (Ardiaca and others 2013). All samples were processed immediately using direct ion-selective electrodes method with a handheld blood-gas analyser (iStat-1, Abbott Diagnostics, Illinois, USA) and disposable cartridges (iStat EC8+) allowing simultaneous sodium, glucose and BUN determinations. The cartridges were tempered at room temperature prior to the procedure. Using the collected data, the total plasma osmolarity and tonicity were calculated. Calculated plasma osmolarity (Posm) was determined as (Nelson and others 2009, DiBartola 2012): Formula

Calculated plasma tonicity (Ton) was estimated as (Nelson and others 2009, DiBartola 2012): Formula

Coefficients of variation (CV) for sodium, glucose, BUN, calculated osmolarity and tonicity were calculated. For this purpose, samples from five rabbits were repeatedly analysed five times each.

The iStat analyser reports BUN levels up to 140 mg/dl (50 mmol/l). Higher levels are reported as higher than 140 mg/dl (μ140 mg/dl)). In the case of the samples from ill rabbits with BUN levels higher than 140 mg/dl, a BUN value of 140 mg/dl was assigned to calculate the osmolarity.

Statistical analysis

All data were checked for normal distribution (D'Agostino-Pearson test) and outliers (Tukey method) (Tukey 1977). The reference intervals for sodium, glucose and BUN levels and for calculated tonicity and osmolarity in clinically healthy rabbits were calculated using a robust method following the American Society for Veterinary Clinical Pathology (ASVCP) guidelines. Differences in sodium, BUN and glucose levels and calculated tonicity and osmolarity among sexes and groups of ages were studied using Kruskal-Wallis test. Mann-Whitney U test was used to compare results from healthy and ill animals. The prevalence of low, normal and high levels of natraemia, osmolarity and tonicity, as well as the prevalence of hypotonic, isotonic and hypertonic hyponatraemia with corresponding 95% CI were calculated. Rank correlation analysis with calculation of Spearman's r and/or Kendall's tau coefficients and correspondent 95% CI was performed to evaluate the degree of association between sodium and glucose levels in healthy and ill animals with hyperglycaemia. A regression model with an equation was calculated. χ2 test was performed and relative risk calculated in order to evaluate differences between mortality rates at different sodium levels. All statistical analysis was performed by means of MedCalc Software and following CLSI (Clinical and Laboratory Standards Institute) guidelines.

Results

Reference intervals: healthy pet rabbits

Results for sodium, glucose and BUN levels and calculated plasma osmolarity and tonicity from 51 healthy pet rabbits were processed, and corresponding reference ranges were obtained. Data for sodium, BUN, calculated tonicity and osmolarity followed normal distribution (D'Agostino Pearson test; Pμ0.05) and no far-out values were identified among these data. Among the data for glucose level, one far-out value (251 mg/dl) was identified and withdrawn from the reference interval calculation. Once this value was withdrawn, data for glucose levels followed a normal distribution (D'Agostino Pearson, P=0.0708). Results are summarised in Table 1.

TABLE 1:

Sodium, glucose, BUN, calculated (calc) tonicity and calculated osmolarity in healthy pet rabbits

The mean CV for sodium, glucose, BUN, calculated osmolarity and calculated tonicity using the method described above were 0.53 per cent, 1.41 per cent, 0.95 per cent, 0.51 per cent and 0.53 per cent, respectively.

No significant differences were found among sexes or groups of age in sodium, glucose, BUN, calculated osmolarity or calculated tonicity in healthy pet rabbits (Kruskal-Wallis test, Pμ0.05). No linear relationship between sodium and glucose concentration was observed in healthy animals (R==0.05, P=0.7435).

Clinical study: ill pet rabbits

Data obtained for sodium, glucose and BUN levels and calculated osmolarity and tonicity from ill animals are summarised in Table 2.

TABLE 2:

Sodium, glucose, BUN, calculated (calc) tonicity and calculated osmolarity in clinically ill pet rabbits

No significant differences attributable to sex or age group were found in sodium, glucose, BUN, calculated osmolarity and tonicity values among ill animals (Kruskal-Wallis test, Pμ0.05).

Statistical analysis shows that the differences between sodium, glucose and BUN values in healthy and ill animals were highly significant (Mann-Whitney U test, P<0.001). Calculated tonicity values also were significantly different (Mann-Whitney U test, P=0.0101). Calculated osmolarity was not significantly different between ill and healthy rabbits (Mann-Whitney U test, P=0.9881). The data obtained for sodium, glucose, BUN, calculated tonicity and osmolarity presented wider ranges in ill rabbits. The mean and median values for glucose and BUN concentrations were higher in ill rabbits. The mean and median values for sodium and calculated tonicity were lower in ill rabbits.

Prevalence of natraemia, tonicity and osmolarity deviations

The prevalence of different levels of natraemia, osmolarity and tonicity and corresponding 95% CI were calculated and are displayed in Table 3. Twenty-four samples presented BUN levels higher than 140 mg/dl, all but one were either normo- or hyperosmolar. The one hyposmolar sample with BUNμ140 mg/dl was severely hyponatraemic ([Na]=109 mmol/l).

TABLE 3:

Prevalence (%) of alterations of natraemia, calculated tonicity and calculated osmolarity in ill rabbits (n=356)

Prevalence of different types of hyponatraemia

The prevalence of different types of hyponatraemia is summarised in Table 4. Hypertonic hyponatraemia was not observed. Nor were observed hypo- or isotonic hypernatraemia. The only two (0.6 per cent) normonatraemic samples that presented hypotonicity were hypoglycaemic. Ten (2.8 per cent) normonatraemic samples presented hypertonicity, nine of them were hyperglycaemic. All the hypernatraemic samples were hypertonic (Table 4).

TABLE 4:

Prevalence (%) of different types of hyponatraemia, normonatraemia and hypernatraemia according to calculated tonicity in ill rabbits (n=356)

Association between natraemia and glycaemia

Changes in natraemia values in conditions of hyperglycaemia ([Glu]μ185 mg/d/l) were studied in the 134 hyperglycaemic samples from ill rabbits. The rank correlation analysis provided Spearman's r coefficient of =0.505 (95% CI =0.621 to =0.367); P<0.0001) and Kendall's tau coefficient of =0.360 (95% CI =0.464 to =0.246); P<0.0001). When only 85 isotonic hyperglycaemic samples were considered, Spearman's r coefficient was =0.672 (95% CI =0.774 to =0.536); P<0.0001) and Kendall's tau coefficient was =0.495 (95% CI =0.600 to =0.366); P<0.0001) (Fig 1).

FIG 1:

Scatter diagram of plasmatic sodium versus glucose concentrations in hyperglycaemic samples from pet rabbits (n=134). Regression line (continuous) and 95% CI curves (dashed lines) are depicted for isotonic samples (circles; n=85)

Association between natraemia and glycaemia in 85 samples with hyperglycaemia and normal calculated tonicity better fit the following regression model (R=0.77; P<0.0001) (Fig 1): Formula

The 95% CI for the intercept and for the slope values were 144.1 to 147.7 and (=0.034 to =0.024), respectively.

Association between natraemia and outcome

Outcome within seven days after initial presentation was studied in rabbits with different sodium levels. Fourteen animals with unknown outcome, and twenty rabbits whose deaths were surgery-related, were withdrawn from the analysis. The 322 rabbits with known outcome were divided in four groups according to natraemia:

  1. Severe hyponatraemia ([Na]<129): n=44

  2. Mild hyponatraemia (129≤[Na]<136): n=83

  3. Normonatraemia (136≤[Na]<148): n=178

  4. Hypernatraemia (148≤[Na]): n=17

Relative risk of mortality in hyponatraemic and hypernatraemic groups compared to normonatraemic group were calculated. Results are summarised in Table 5. Mortality rate within seven days after initial presentation was significantly higher in critically ill rabbits admitted to intensive care unit with sodium levels under 129 mEq/l (χ2 test; P<0.0001) when compared with other groups. For the rabbits with severe hyponatraemia, the risk of mortality was 2.3 times higher than for normonatraemic rabbits. Three of these 44 severely hyponatraemic rabbits had pseudohyponatraemia and the rest ­presented hypotonic (true) hyponatraemia. All the survivors among the 44 severely hyponatraemic rabbits were truly hyponatraemic.

TABLE 5:

Mortality rates (%) at different natraemia levels in ill rabbits (n=322)

Discussion

Based on the results of this study, 136–147 mEq/l appears to be a physiological range of sodium in pet rabbits when measured in whole blood in lithium heparin with the EC8+ cartridges for iStat-1 analyser. Literature references report normal sodium levels of 138–150 mEq/l (Gillet 1994), 131–155 mEq/l (Mader 2004, Washington and Van Hoosier 2011), 138–155 mEq/l (Hernandez-Divers 2005) and 134–150 mEq/l (Meredith and Crossley 2002) in rabbits. This means the lower limit proposed by different authors can vary up to 8 mEq/l from one to another literature source.

Numerous data published for rabbits report normal glycaemia and plasmatic BUN limits respectively as 75–150 mg/dl and 15–30 mg/dl (Hernandez-Divers 2005); 108–160 mg/dl and 9–23 mg/dl (Fudge 2000); 75–155 mg/dl and 13–29 mg/dl (Mader 2004); 76–140 mg/dl and 17–24 mg/dl (Harcourt-Brown 2002) and 74–148 mg/dl and 5–25 mg/dl (Suckow and others 2002). Glucose and BUN levels found in healthy rabbits in this study are mainly consistent with these previously published results. It is well known that stress of any kind, including illness, can cause the plasma glucose levels to rise. Fear and stress from handling is possible in healthy rabbits and may be responsible for glucose levels up to 270 mg/dl (Dontas and others 2011, Harcourt-Brown and Harcourt-Brown 2012). Each individual from a pet rabbit population presented to a particular veterinary clinic is subject to stress and possible mild dehydration during transportation. This could be the cause of the mild hyperglycaemia and azotaemia in some of the healthy rabbits in this study. These changes appear to have no effect, however, on the sodium levels and tonicity in healthy individuals in this study.

The values for calculated osmolarity and tonicity obtained in healthy pet rabbits in this study are consistent with those obtained in previous studies of real plasma osmolality in laboratory rabbits. Some of these osmometer-determined osmolality values are 290–300 mOsm/kg (Zornow and others 1987), 290–312 mOsm/kg (Marcus and others 1970), 283–293 mOsm/kg (Gerstberger and others 1992), 305 mOsm/kg (Whitlock and Wheeler 1964), 296–301 mOsm/kg (Whitlock and Wheeler 1964, Marcus and others 1970, Zornow and others 1987, Gerstberger and others 1992, Frosini and others 2000).

It is important to keep in mind that literary sources of reference ranges for sodium, BUN and glucose values frequently do not provide any information about the type of sample (blood, serum or plasma) or the method used for the analysis. Mild variations between different laboratories are common. When using published references, it is important to know which method has been employed to determine the sodium levels as it may influence the results. Flame photometry, or indirect ion-selective electrodes methods, can provide lower sodium values compared to those measured with the direct ion-selective method when hyperlipidaemia or hyperproteinaemia are present (isotonic pseudohyponatraemia) (DiBartola 2012). iStat analyser uses a direct ion-selective method and is less affected by these artefacts. It also offers the advantage of easy handling, small sample size (0.1 ml of whole blood) and quick results (less than two minutes). Recently 95%RI for sodium levels in venous blood from rabbits were reported to be 133–149 mEq/l (with median value of 143 mEq/l) using blood-gas syringes and GC8+ cartridges for iStat (Tausch 2011). Different protocol of blood collection can be responsible for the slightly lower sodium levels compared to our study, as the type and amount of anticoagulant can influence the result (Weisberg 1989, Milionis and others 2002, Nguyen and others 2007, Nelson and others 2009, Reddy and Mooradian 2009, DiBartola 2012). In our study, the CV obtained for sodium, glucose, BUN and calculated osmolarity and tonicity are consistent with results published by Tausch (2011), supporting that iStat system is reliable in rabbits. The impairment of statistics of BUN levels and calculated osmolarity, due to the fact that the iStat analyser only provides BUN levels up to 140 mg/dl, apparently had little influence on the main goals of the present study, and appears to have scarce clinical relevance.

Based on the results of this study, hyponatraemia and hyperglycaemia appear to be common in ill pet rabbits. True hyponatraemia (hypotonic hyponatraemia) and pseudohyponatraemia (isotonic hyponatraemia) can be encountered in a clinical environment. Plasmatic sodium levels in rabbits decrease in conditions of hyperglycaemia in a similar manner as it occurs in human beings. In this study, the equation of the regression model for the negative linear correlation between sodium and glucose levels hyperglycaemic samples from ill rabbits was: [Na]=145.9=0.0288×[Glu]. A similar formula was obtained in an experimental study in human beings: [Na]=142.64=0.024×[Glu] (Hillier and others 1999). This fact suggests that similar chemical mechanisms are responsible for these changes in rabbits and human beings. A drop of 2.9 mEq/l of sodium concentration would be expected for every 100 mg/dl increase in glucose concentration in rabbits. However, further research is necessary to find out if a correction factor can be used. Until more data are obtained, the calculation of plasmatic tonicity is the best available tool for classification of hyponatraemia. To the contrary, no comparable linear correlation between sodium and glucose values was found in healthy individuals. This circumstance is supporting the known fact that glucose is acting basically as a permeant solute in presence and normal function of insulin (Hillier and others 1999).

Several experimental studies proved that natraemia variations are elicited by changes in glycaemia (Hillier and others 1999). Further research may reveal if the reverse is also true, that is, if glucose levels may rise compensatory to hyponatraemia due to gastric stasis, digestive obstruction, diarrhoea, liver disease, congestive heart failure, septicaemia, nephrotic syndrome, pancreatitis, diuretic administration or other conditions. Harcourt-Brown and Harcourt-Brown (2012) reported poor prognosis in severely hyperglycaemic rabbits and correlation between digestive obstruction and severe hyperglycaemia. Minding the association between plasmatic sodium and glucose, this finding is consistent with our results on association between hyponatraemia and mortality.

Clinicians attending pet rabbits must mind that hyponatraemia is a common condition and, when severe, constitutes a prominently increased risk of mortality. Simultaneous determination of sodium and glucose concentrations, and calculation of the plasmatic tonicity, using the formula described in this paper, is important when ­evaluating a clinically ill rabbit patient. Hyponatraemia, with ­simultaneous ­hypotonicity, represents a true hyponatraemic status that needs treatment. The calculated osmolarity has a lower diagnostic value. Several guidelines for correct diagnosis, classification and choice of treatment protocol are available and should be consulted by the clinician. Factors, such as patient's volemia status, duration and severity of hyponatraemia and presence or absence of symptoms, must be considered (Illowsky and Laureno 1987, Reddy and Mooradian 2009, DiBartola 2012).

Footnotes

  • Provenance not commissioned; externally peer reviewed

  • Accepted February 6, 2014.
  • Published Online First 6 March 2014

References

Sign up for job alerts

Vet Record Careers features a comprehensive database of vet jobs.
Browse for your next job, or set personalised job alerts so you hear about vacancies as soon as they go online.
Find out more at Vet Record Careers

Free Sample Issue

This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of the Veterinary Record.
View free sample issue >>