• Japanese dog breeds;
  • Na-K pump;
  • pseudohyperkalemia;
  • red blood cell


  1. Top of page
  2. Abstract
  3. Acknowledgment
  4. References


Canine RBCs are expected to have high sodium and low potassium (LK) concentrations. However, some dogs have a low sodium and high potassium (HK) phenotype due to a Na-K pump in the mature RBC membrane. Awareness of this particularity avoids misinterpretation of hyperkalemia in affected dogs.


The purpose of this study was to compare the CBCs, serum biochemistry profiles, urinalysis data, and electrocardiography data in HK and LK Akita dogs.


In this study, the intra-RBC potassium (KRBC) was measured in 48 healthy adult Akita dogs. The HK group included dogs with KRBC concentration at least 5-fold that of plasma, while the LK dogs served as controls.


The HK phenotype was determined in 10 dogs (21%). With the exception of higher plasma potassium concentrations in the HK group compared with LK controls (6.6 mmol/L vs 4.4 mmol/L, P < .001), there were no other differences in serum biochemistry variables. In hematology, mean HGB concentration (13.5 vs 15.9, P < .001), PCV (42.9 vs 46.7, P = .009), RBC (6.1 vs 7.4, P < .001), and MCHC (31.3 vs 34.0, P < .001) were significantly lower in the HK group, while MCV (70.0 vs 63.4, P < .001) was higher compared with the LK controls. There were no significant differences in urinalysis or electrocardiography data between groups.


HK Akita dogs showed significant differences in several hematologic variables, as well as higher plasma potassium concentration. Therefore, the HK phenotype should be considered in the interpretation of these variables in this breed.

Canine RBCs usually have high sodium and low potassium (LK) concentrations because the membrane Na-K pump is lost during maturation from reticulocytes to mature RBCs.[1] In some dogs, the Na-K pump persists, resulting in low sodium and high potassium (HK) RBCs. This HK- mutation appears to have occurred first in Korean dogs in ancient times and later spread across Japan.[2] To date, the HK phenotype has been identified in 10 of 13 Japanese breeds, and its prevalence in the Akita breed is 26.3%.[3] In addition, there is increased amino acid uptake in HK RBCs, resulting in abnormally increased concentrations of aspartate, glutamate, glutamine, and glutathione.[4] The RBC count and HGB concentration were lower in HK mongrel dogs, whereas the MCV was higher compared with LK dogs from the same family.[4] The HK RBCs half-life was approximately half that of LK RBCs, suggesting that glycolysis occurs at a higher rate in the former.[5] In addition, HK RBCs tend to undergo hemolysis when exposed to oxidative or osmotic stress.[6] This was reported in 2 dogs that presented hemolytic anemia after ingestion of small amounts of onion,[6] and the higher susceptibility of HK dogs to the oxidant action of onions was later established.[7] Certain parasites, such as Babesia gibsoni, multiply better in HK compared with LK RBCs, possibly predisposing HK phenotype dogs to a more severe course of the disease.[8] A greater osmotic fragility of HK RBCs compared with LK cells was reported in another study.[9] Moreover, dogs with the HK phenotype should not be used as blood donors.[10]

Identifying the HK RBC phenotype in dogs, especially in breeds with an established higher prevalence has diagnostic implications. The aim of this study was to characterize and compare the CBC, serum biochemistry variables, urinalysis data and electrocardiographic findings in Akita dogs with HK and LK phenotypes.

This study included clinically healthy adult Akita dogs from 4 breeding kennels in Southern Brazil. These included dogs imported from Japan, Argentina, Italy, and Spain, and their litters born in Brazil. The clinical examination of these dogs included determination of heart and respiratory rates, body temperature, auscultation, abdominal palpation, mucous membrane evaluation, check of capillary refill time, lymph node palpation, and establishment of hydration status. Animals showing any abnormalities detected by physical examination or reported by the owner were excluded from the study. Dogs were recruited with their owner's consent. This study was conducted in accordance with standard ethical and animal welfare guidelines. After determining their intra-RBC potassium concentration (KRBC), the dogs were divided into 2 groups, the HK phenotype group and the LK phenotype controls.

Blood samples were collected by cephalic or jugular venipuncture in EDTA, lithium-heparin, and plain vacuum tubes with gel separator (BD Brasil, São Paulo, Brazil) for a CBC, and analysis of electrolyte and other serum biochemistry variables, respectively. The EDTA, lithium-heparin blood samples, and the urine samples were refrigerated pending analysis, which was performed within 4 h from collection. Blood samples for serum biochemistry analysis were allowed to clot, refrigerated, and centrifuged within 2 h from collection. Harvested sera were stored at −20°C until analyzed, usually within 2 weeks after collection. All dogs were fasted overnight prior to collection and none of the samples showed lipemia and/or hemolysis.

Plasma potassium (lithium-heparin samples) was measured using an indirect ion-selective electrode methodology (Modular P Chemistry Analyzer; Roche/Hitachi, Indianapolis, IN, USA). KRBC was measured as described elsewhere.[10] Briefly, RBCs were separated from plasma, rinsed twice with PBS and lysed in a 1:1 solution with distilled water. The potassium in this solution was measured and then multiplied by 2.[10] Dogs were considered to have a HK phenotype if their KRBC concentration was 5-fold higher than their corresponding plasma potassium concentration. To verify possible interferences of delay in separating cells from plasma in HK samples, potassium was also measured in the serum (centrifuged within 2 h from collection) of the HK dogs and compared with plasma potassium (performed within 4 h from collection). ALT and ALP activities, cholesterol, creatinine, and urea concentrations were measured using a dry-chemistry analyzer (Reflotron; Roche Diagnostica, São Paulo, Brazil). Total magnesium, and total protein and albumin concentrations were determined by spectrophotometry using a wet semi-automated clinical analyzer (Metrolab 1600 DR; Buenos Aires, Argentina) at 37°C.

CBCs were performed using an automated hematologic impedance analyzer (CC 530 CELM; São Paulo, Brazil). PCV was measured manually with a microhematocrit centrifuge. Blood smears were prepared from each sample, air-dried, and stained using a Romanowsky staining solution (Panótico Rápido, Laborclin, São Paulo, SP, Brazil) for morphological evaluation and a differential WBC count.

Urine specific gravity (USG) was measured with a refractometer. Urine was centrifuged at 300g for 5 min, and routine dipstick (Multistix/Clinitek 50 Urine Chemistry Analyzer; Bayer Health Care Diagnostics Division, Leverkusen, Deutschland) and sediment analysis were performed.

Electrocardiogram (ECG) was performed on nonsedated dogs that were manually restrained in right lateral recumbency. All ECGs were recorded using a portable electrocardiograph (C10 Electrocardiograph; TEB, São Paulo, Brazil). Standard 6-lead ECGs (leads I, II, III, aVR, aVL, and aVF) were recorded for 1–2 min. Rhythm analysis (regular rhythm, respiratory sinus arrhythmia, pathological arrhythmias) was performed, and measurements were taken of heart rate (HR) in beats per min (bpm), amplitude and duration of P waves and QRS complexes, P-R segment duration, S-T segment deviation, amplitude of T wave, and duration of QT intervals from lead II from the recordings. The mean electric axis (MEA) was calculated using the algebraic sum of the QRS deflections in lead I and lead III, and the values were plotted in a reference table.[11] Recordings were made at a paper speed of 50 mm/s and amplitude of 1 cm/mV. A 35-Hz filter was used in all cases.

The data distribution pattern was examined using the Shapiro–Wilk test. The CBC, urinalysis, and biochemistry data for the LK and the HK dogs were compared using t- and Mann–Whitney U-tests, for normally and nonnormally distributed parameters. The ECG data were compared using the Fisher exact probability test. A P < .05 was considered significant. All analyses were performed using a statistical software package (Assistat Version 7.6 beta; UFCG, Campina Grande, Brazil).

This study included 48 adult Akita dogs, 10 of which (21%) had an HK phenotype and 38 (79%) an LK phenotype. The HK group included 6 intact males and 4 intact females, while the LK group included 14 intact males and 22 intact females, with no group difference in the proportion of genders (P = .29). The median ages of the HK and LK groups were 3 years (range 1–6) and 3 years (range 1–5), respectively, with no difference between the 2 groups. All animals showed ideal body condition scores (total body weight was not considered).

Akita dogs with the HK phenotype were found in all of the 4 kennels included in the study. Six different lineages could be recognized that were not related up to their fourth generation (great-grandparents).

Mean plasma potassium concentration in the HK group was 6.6 mmol/L (SD 1.0; reference interval [RI] 4.37–5.35 mmol/L),[12] and KRBC was 34.4 mmol/L (SD 9.8), while in the LK group, mean plasma potassium concentration was 4.4 mmol/L (SD 0.4) and mean KRBC was 2.4 mmol/L (SD 0.6). Both plasma and KRBC potassium concentrations differed significantly between the HK and LK groups (P < .001). Potassium was lower in serum than in plasma in the HK group (6.1 ± 0.9 mmol/L vs 6.6 ± 1.0 mmol/L), but the difference was not significant (P = .11). No other serum biochemistry differences between groups were noted.

The HGB concentration, PCV, RBC count, and MCHC were significantly lower (Table 1) in HK dogs compared with LK dogs, and were often at their lower RI. Mean MCHC was below RI. Mean MCV in the HK dogs was significantly higher compared with the LK dogs, but within RI.[13] There were no differences between the HK and LK groups in mean total and differential WBC counts.

Table 1. Group means of CBC variables from high potassium (HK) and low potassium phenotype (LK) Akita dogs.
 HK DogsLK Dogs P RI[13]
  1. Data are means ± standard deviation.

RBC (× 106/μL)6.1 ± 0.67.4 ± 0.7< .0015.5–8.5
HGB (g/dL)13.5 ± 1.615.9 ± 1.2< .00112–18
PCV (%)42.9 ± 4.146.7 ± 3.2.00937–55
MCV (fL)70.0 ± 3.863.4 ± 3.5< .00160–77
MCHC (%)31.3 ± 1.434.0 ± 1.0< .00132–36

The urinalysis data showed no difference between HK and LK dogs. Three LK dogs had USG lower than 1.015, but no other alterations in urinalysis, CBC, or biochemical profile relative to the urinary system were observed in these dogs.

No rhythmic abnormalities or important alterations were observed in the ECG of both groups. ECG values in Lead II are presented in Table 2. No significant difference was observed between HK and LK dogs for normal or values outside the RI.

Table 2. Group medians of electrocardiographic variables recorded from high potassium (HK) and low potassium phenotype (LK) Akita dogs.
 HK PhenotypeLK PhenotypeReference Interval for Medium-Sized Dogs[11]
Median and IQR% of Dogs Outside RIMedian and IQR% of Dogs Outside RIP-Value
  1. Data are medians (interquartile ranges).

  2. a

    Six dogs had a biphasic T wave, and 32 dogs had a positive T wave.

Heart Rate (bpm)125 (114–135) 120 (115–136) .870–160
P Wave Duration (second)0.04 (0.04–0.04)100.04 (0.04–0.04)211≤ 0.04
P Wave Amplitude (mv)0.15 (0.15–0.25) 0.2 (0.15–0.2) .5≤ 0.4
PR Interval (second)0.1 (0.1–0.12)100.11 (0.1–0.12)–0.13
QRS Duration (second)0.05 (0.047–0.05) 0.05 (0.05–0.05)15.8.2Maximum 0.06
R Wave Amplitude (mv)0.8 (0.7–1.1) 0.7 (0.5–1.0) .1Maximum 3.0
Q Amplitude (mv)0.1 (0–0.5)100.1 (0–0.2)15.8.3 
QT Interval (second)0.19 (0.18–0.2) 0.19 (0.18–0.2) .50.15–0.25
T (mv)a0.2 (0.14–0.3) 0.2 (0.15–0.25) .8Negative, positive or biphasic. ≤ 25% R wave
MEA (degrees)80 (71–90) 67 (59–86)10.5.240–100

A previous study investigating the incidence of HK dogs in Japan and East Asia found that the HK phenotype has spread across Japan and Korea, but was not present in Taiwan, Indonesia, Mongolia, or Russia.[2] In addition, the HK phenotype has been described as an autosomally recessive inherited trait.[2] The HK phenotype found in our study confirmed the presence of the HK genotype in Akita dogs imported from Japan and other countries, and emphasizes the importance of making practicing veterinarians aware of this condition.

As pseudohyperkalemia in HK dogs can occur if separation of serum from clotted blood is delayed, or in cases of hemolysis, it is essential that hyperkalemia is interpreted properly in such dogs.[10, 14] Therefore, the serum from all Japanese dog breeds and their crosses should be separated immediately after collection. In the present study, potassium was measured in refrigerated plasma samples within 4 h to avoid early hemolysis and loss of KRBC potassium to the plasma. We stipulated a maximum amount of time of 4 h between collection and centrifugation due to the logistics of collection, transportation, and processing of the samples. Other authors reported no difference in plasma potassium in HK Akita dogs up to 8 h after collection,[14] although in a different study, a significant difference in plasma potassium concentration of 3 Akita dogs was shown after 4 h in contact with RBCs.[10] In our study, in spite of the absence of statistically significant differences observed in potassium concentrations in serum and plasma separated from cells within 2 and 4 h, respectively, there was a tendency for increased plasma potassium concentrations. Therefore, based on our findings and also reports in the literature, it appears that plasma can be left on top of the RBC layer for up to 4 h after collection for the measurement of potassium, but, for more accurate results, an earlier separation is encouraged. In both plasma and serum, potassium concentrations in HK dogs were significantly higher than in LK dogs, and they were always outside the RI.

All RBC parameters were significantly different between HK and LK dogs, corroborating the findings of other studies.[4, 9] Studies with HK phenotype dogs have found that HGB, PCV, RBC counts, and MCHC values can be lower in these animals.[4, 9] These values are often at the lower end of, or slightly below, published RI[13] similar to what we found in this study. Likewise, mean MCV values were higher in the HK dogs than in LK dogs in our study. These mild increases may be more significant in some Asian dog breeds, which normally tend to have a lower MCV.[9] It has been suggested that an increase in intracellular water could account for the lower MCHC and higher MCV in these dogs, given the normal range of the other variables.[4, 9] This may be due to the increased osmotic fragility of RBCs as intracellular sodium concentrations are reportedly low.[4] Determination of RBC sodium for future studies in these breeds may further clarify the mechanism responsible for the alterations in RBC morphology. Additionally, although no differences were found in the biochemical profiles of HK and LK Akita dogs in our study, further studies should consider including a more extensive serum chemistry profile in these animals.

Expected electrocardiographic alterations associated with hyperkalemia include prolonged QRS complexes and PR intervals, peaked narrow T waves, ST segment depression, prolongation of the QT interval, bradycardia and other arrhythmias, such as sinus tachycardia, ventricular premature complexes, and atrioventricular dissociation.[15] Such alterations may be seen with plasma potassium concentrations > 5.5 mmol/L.[11] Despite the higher plasma potassium concentrations above the RI in HK dogs compared with LK dogs, no ECG abnormalities were found in this study. However, altered ECG data due to hyperkalemia in case of hemolysis should be anticipated.

In our study, the HK phenotype occurred in approximately 20% of Akita dogs, similar to what has been described in the literature.[2, 3] Furthermore, as some of the HK dogs included in this study had parents imported from different countries, it can be assumed that the HK phenotype is distributed worldwide.


  1. Top of page
  2. Abstract
  3. Acknowledgment
  4. References

The authors gratefully acknowledge the Waltham Foundation for financial support.

Disclosure: The authors have indicated that they have no affiliations or financial involvement with any organization or entity with a financial interest in, or in financial competition with, the subject matter or materials discussed in this article.


  1. Top of page
  2. Abstract
  3. Acknowledgment
  4. References
  • 1
    Kirk RG, Andrews SB, Lee P. The correlation of composition and morphology during the high to low potassium transition in single erythropoietic cells. J Membr Biol. 1983;76:281286.
  • 2
    Fujise H, Higa K, Nakayama T, Wada K, Ochiai H, Tanabe Y. Incidence of dogs possessing red blood cells with high K in Japan and East Asia. J Vet Med Sci. 1997;59:495497.
  • 3
    Tanabe Y. Phylogenetic studies of dogs with emphasis on Japanese and Asian breeds. Proc Japan Acad. 2006;82:375387.
  • 4
    Maede Y, Inaba M, Taniguchi N. Increase of Na-K-ATPase activity, glutamate, and aspartate uptake in dog erythrocytes associated with hereditary high accumulation of GSH, glutamate, glutamine, and aspartate. Blood. 1983;61:493499.
  • 5
    Maede Y, Inaba M. Energy metabolism in canine erythrocytes associated with inherited high Na+- and K+-stimulated adenosine triphosphatase activity. Am J Vet Res. 1987;48:114118.
  • 6
    Maede Y. High concentration of blood glutathione in dogs with acute hemolytic anemia. Nihon Juigaku Zasshi. 1977;39:187189.
  • 7
    Yamoto O, Maede Y. Susceptibility to onion-induced hemolysis in dogs with hereditary high erythrocyte reduced glutathione and potassium concentrations. Am J Vet Res. 1992;53:134137.
  • 8
    Yamasaki M, Takada A, Yamato O, Maede Y. Inhibition of Na,K-ATPase activity reduces Babesia gibsoni infection of canine erythrocytes with inherited high K, low Na concentrations. J Parasitol. 2005;91:12871292.
  • 9
    Battison A. Apparent pseudohyperkalemia in a Chinese Shar Pei dog. Vet Clin Pathol. 2007;36:8993.
  • 10
    Degen M. Pseudohyperkalemia in Akitas. J Am Vet Med Assoc. 1987;190:541543.
  • 11
    Tilley LP (Ed). Essentials of Canine and Feline Electrocardiography: Interpretation and Treatment. 3rd ed. Philadelphia, PA: Lea &Febiger; 1992:470p.
  • 12
    Kaneko JJ, Harvey JW, Bruss ML. Appendix IX: blood analyte reference values in small and some laboratory animals. In: Kaneko JJ, Harvey JW, Bruss ML, eds. Clinical Biochemistry of Domestic Animals. 6th ed. San Diego, CA: Academic Press; 2008:889895.
  • 13
    Rizzi T, Meinkoth JH, Clinkenbeard KD. Normal hematology of the dog. In: Weiss DJ, Wardrop KJ, eds. Schalm's Veterinary Hematology. 6th ed. Ames, IA: Wiley Blackwell; 2010:799810.
  • 14
    Rich LJ, Bernreuter DC, Cowell RL. Elevated serum potassium associated with delayed separation of serum from clotted blood in dogs of the akita breed. Vet Clin Pathol. 1986;15:1214.
  • 15
    Tag TL, Day TK. Electrocardiographic assessment of hyperkalemia in dogs and cats. J Vet Emerg Crit Care. 2008;18:6167.