Mild polycythemia associated with compulsive disorder in dogs1

Introduction 

“Compulsive disorder (CD)” in animals is defined as repetitive, exaggerated, or sustained behaviors that are observed outside of the original context (Hewson and Luescher, 1996). Compulsive disorder can develop in many species, including dogs, cats, birds, and horses. Typical clinical signs in dogs are excessive licking, possibly developing into an acral lick dermatitis/granuloma; tail-chasing behavior (frequently seen in german shepherd dogs); circling or spinning behavior (frequently seen in bull terriers); pacing; chasing imaginary prey; licking air or objects; light or shadow chasing; and fixating on objects. Cats are often presented for wool sucking and psychogenic alopecia. Typical behavior seen in birds includes feather picking. Horses are often represented to crib or weave (Overall 1997, Luescher 1998, Luescher 2003, Virga 1999, Landsberg et al 2003).

CD in dogs appears similar to human obsessive compulsive disorder (OCD) in its behavioral presentation and response to medication, and in (breed-related) genetic predisposition (Goldberger and Rapoport 1991, Rapoport et al 1992). The prevalence of OCD in human beings is about 2% to 3% (Nakajima and Nakajima 1994, Hollander 1997, Sasson et al 1997). There have been no in-depth studies conducted to determine the prevalence of CD in dogs or cats. However, one study at Ontario Veterinary College reported that CD in dogs accounts for 6% of all canine behavior cases diagnosed at that clinic (Luescher et al., 1991).

The etiology of CD is not known, but it is thought to be an anxiety disorder. CD is believed to develop from behaviors that are caused by conflict or frustration (Luescher, 2003). Conflict behavior results from multiple opposing and similarly strong motivations (for example, satisfying hunger and avoiding danger) (Manning 1967, Wood-Gush 1988). Frustration refers to a situation in which an animal is motivated to perform a behavior but is prevented from doing so (for example, being caged but motivated to exercise) (Luescher, 2003).

We conducted a clinical trial study to evaluate efficacy of one selective serotonin reuptake inhibitor (antidepressive medicine used commonly to treat human OCD) for canine CD treatment. As part of the trial, the health status of all subject dogs with CD was evaluated by physical examination, complete blood count (CBC), serum chemistry panel, and urinalysis. While reviewing all the dogs’ blood examination results, the investigators observed an interesting trend in many dogs with CD. Almost 25% of CD dogs’ hematocrit values were relatively high compared to the reference range at the Purdue University Veterinary Teaching Hospital (VTH) where blood examination was performed. Statistical comparison was made if dogs diagnosed with CD had relatively high hematocrit values compared to the reference range that is considered as the value of the clinically normal dogs at Purdue University VTH.

Mild elevations in hematocrit value in human psychiatric disorder patients have been discussed previously (Mazzoli and Mendels 1967, Benitone and Kling 1969, Mathew and Wilson 1986, Eschet et al 1988, Murray and Hodgson 1991, Walsh 1991, Mazzoli and Bewnazzi 1992, Haghighat et al 1996, Haghighat et al 1996, Lowe 2001). The previous studies with human psychiatric patients reported that mild elevation in hematocrit value was observed, but the definite physiological mechanism behind the findings was not found. On the other hand, some polycythemia cases in humans have been associated with behavioral and mental changes because of a change in blood flow to the brain. This study reports and discusses hematocrit elevation and its possible relationship to CD symptoms in dogs.

Results 

CD dog group had significantly (P value < .05) higher values in hematocrit, hemoglobin (Hb), and mean corpuscular volume (MCV), along with a significant decrease in white blood cell count (WBC) and mean corpuscular hemoglobin concentration (MCHC) when compared to the reference ranges (Table 2 and Fig.). The univariate analysis of variance showed that gender and neuter status were associated with hematocrit (P = .01). A post hoc Tukey test revealed that neutered males had significantly higher hematocrit than intact males (P = .03), but no other significant differences were observed with other pairs of sex status. The percentage of dogs with CD that had a higher value than the midpoint of the reference range was 85.7% for hematocrit, and 24% of CD dogs had hematocrit values higher than the reference range (Fig. 1). Eighty-nine percent of the CD dogs had Hb values higher than the midpoint of the reference range, as was the case for 70% of the CD dogs for red blood cell count (RBC). The Mann-Whitney test comparing hematocrit values in the group of dogs with the 6 categories of CD showed no significant difference.

Table 2. Hematological data comparing the Purdue University VTH clinical pathology reference ranges with those for 63 dogs with compulsive disorders

	Variables	Reference Range (Purdue University)			CD group (N=63)		P value (One-tail T-test)
		Reference range	Mean	Standard deviation	Mean	Standard deviation
	TP (g/dl)	6.0–8.0	7.0	0.5	7.1	0.6	0.391
	PCV (%)	37–55	46.0	4.5	51.4	4.9	<0.0005⁎
	Hb (g/dl)	12–18	15.0	1.5	17.0	1.8	<0.0005⁎
	MCV (fl)	60–75	67.5	3.75	70.3	9.3	0.02⁎
	RBC(x10^6ul)	5.5–8.5	7.0	0.75	7.1	1.1	0.27
	MCHC(g/dl)	32–36	34.0	1.0	33.3	2.0	0.013⁎
	WBC (ul)	6–17(x103)	11.5	3.0	8.4	3.0	<0.0005⁎
	Neutrophil Band	0–0.3(x103)	0.15	0.08	0.23	0.25	0.326
	Neutrophil Segmented	3–12(x103)	7.50	2.25	5.70	2.47	<0.0005⁎
	Lymphocyte	1–5(x103)	3.00	1.00	1.67	0.88	<0.0005⁎
	Monocyte	0.15–13(x103)	0.75	0.30	0.41	0.28	<0.0005⁎
	Eosinophil	0.1–1.2(x103)	0.68	0.29	0.49	0.38	<0.0005⁎

⁎ Difference was significant between the CD group and normal range data (P < 0.05). Mean and standard deviation for Purdue reference range was calculated from the reference range.

View Large Image Download to PowerPoint

Figure 1. Hematocrit (HCT) value comparison with normal Gaussian distribution and 63 CD dogs compared to normal distribution model made by Excel using the Purdue VTH Laboratory’s Normal Value. Upper graph from 63 CD dogs and lower graph is a normal Gaussian distribution graph made from Purdue VTH reference range.

Discussion 

The CD dogs’ hematocrit value distribution was shifted to the right, and 15 out of 63 dogs (24%) had polycythemia (Fig.). Polycythemia was defined as an increase in hematocrit and hemoglobin concentration or a RBC count of a blood sample above the reference range (Tvedten and Weiss, 1999). Polycythemia can be categorized, based on the underlying physiological mechanism, into absolute polycythemia and relative polycythemia. Absolute polycythemia is an absolute increase in the total erythroid mass of the body caused by increased bone marrow production of RBCs. Absolute polycythemia can be subdivided into primary absolute polycythemia (polycythemia vera) and secondary absolute polycythemia (lack of oxygen carried in the body, or an increase in the level of erythropoietins resulting from physiological changes in the body such as neoplasia of the kidney or aorta) (Erslev, 1995). Relative polycythemia is an increase in hematocrit without an absolute increase in the total body red cell mass and is caused by a decrease in plasma volume or splenic contractions (ie, a change in the number of red blood cells relative to plasma volume) (Tvedten, 1999). In human medicine, there is a subgroup in polycythemia that is called “apparent polycythemia.” Usually, patients have elevated hematocrit with normal red cell volume and plasma. The condition is reported to be associated with hypertension, stress, obesity, and/or being a heavy smoker (Erslev, 1995).

When the sample CBC results came back with elevated hematocrit values, we checked whether the dog was dehydrated by checking other values so the dogs were not with conditions of hyperproteinemia (Table 2), as well as hyperalbuminemia, hypernatremia, and hyperchloridemia. The investigator also called the owner of the dog to ask if the dog was fasted or deprived from water intake before visiting the veterinarian for the blood draw. No dogs were fasted or water restricted prior to their vet visit. To diagnose whether the dogs had absolute polycythemia (increase in erythrocytes), the dog’s living condition and health status at the time of blood drawing were checked. Only one dog was recruited from the state of Colorado and had chronic exposure to high altitudes, but the other 62 dogs were not from mountainous areas (Table 1). According to the veterinarians who conducted the physical examinations, none of the dogs had clinical signs of any other physical abnormality. To summarize, the dogs did not have evidence of dehydration and were unlikely to have any systemic diseases that might have caused the dogs to have absolute polycythemia.

These facts led us to construct two hypotheses to explain why dogs with CD had relatively high hematocrit values and almost one quarter of the CD dog population in the study had polycythemia. The first is the stress hypothesis, which states that stress, conflict, and frustration lead to epinephrine secretion and compulsive behavior, with subsequent splenic contraction resulting in polycythemia. The second is the primary polycythemia hypothesis, which states that polycythemia results from primary pathologic processes such as cardiovascular disease, or elevation of erythropoietin resulting from kidney problems including neoplasia or bone marrow disease.

Conclusion 

The study found that a large proportion of CD dogs observed have mild elevations in hematocrit. This finding warrants further investigation to determine if it is a cause or an effect of CD in dogs.