Separation and confinement anxiety in a golden retriever × standard poodle dog

Abstract

An 8-month-old spayed female golden retriever × standard poodle dog was presented for assessment because of vocalization, destructive behavior, and urination when left home alone and confined in a crate. Previous treatments included multiple medications, graduated departures, decoupling of departure cues, and punishmentbased techniques. Adverse reactions to paroxetine and lorazepam were noted before successful treatment with clomipramine, “safe haven” training, and behavior modification.

Key clinical message:

Separation anxiety is a common behavioral disorder in dogs. It is often treatable but may require changes in the approach to both medication and behavior modification based on patient and client factors.

Separation anxiety is one of the most common reasons dogs are referred to veterinary behaviorists, rehomed, and relinquished to shelters (1). The underlying cause of separation anxiety, or separation-related behaviors, is unknown, with the same clinical signs representing different etiologies among dogs (2,3). DeAssis et al (2) identified 7 principal components representing underlying motivations for clinical signs commonly attributed to separation anxiety, including exit frustration, social panic, elimination, redirected frustration, reactive communication, immediate frustration, and noise sensitivity. Common clinical signs include vocalization, destructive behavior, and house-soiling occurring in the absence, or perceived absence, of the owner (4). Neutered and mixed-breed dogs are overrepresented, along with dogs that have experienced trauma or a change in routine, those obtained from pet stores and shelters or at a young age, and those whose owners have certain cognitive biases and attachment styles (5). There is no breed predisposition (5), although a genetic component is suspected since some puppies display signs of separation anxiety from the first instance of being left alone and do not become accustomed to separation from people. Comorbidities such as noise phobias (6,7), pain (8), and other fears and phobias (9) are common; and pharmaceutical intervention with medications that inhibit reuptake of serotonin were demonstrated to be effective (e.g., 10,11). In Canada, for treatment of separation anxiety, there are currently 2 approved medications: clomipramine (Clomicalm; Elanco, Guelph, Ontario), a tricyclic antidepressant; and fluoxetine (Reconcile; Virbac, Cambridge, Ontario), a selective serotonin reuptake inhibitor (SSRI). This case report describes a dog that did not respond to fluoxetine or paroxetine (another SSRI) but responded positively to clomipramine.

Case description

An 8-month-old spayed female golden retriever × standard poodle (golden doodle) dog was presented for assessment because of vocalization, destructive behavior, and urination while alone. The dog was purchased at 8 wk of age from a breeder and was the only pet of a married couple living on the 3rd floor of an apartment building with neither a balcony nor a yard. The living arrangements at the breeder and early socialization and training techniques were not disclosed at the time of purchase. However, the couple subsequently learned that this puppy’s mother displayed similar behaviors when left alone. The clients started crate-training at 8 wk of age; however, the puppy would vocalize uninterrupted for at least 1 h each time she was crated, even with the clients present.

At 13 wk of age, the dog was left home alone for the first time, for 2 h. She urinated and defecated in the crate during the absence and was vocalizing and panting upon the clients’ return. After this incident, the crate was replaced with an exercise pen and the dog was left home alone for 1 h so the clients could determine if confinement was the primary problem. During this time, the dog escaped from the pen, vocalized, destroyed her bed, and scratched and chewed walls and a door frame. She also urinated on the floor near the exit. Upon their return, the dog jumped on the clients, panting and whining. These incidents were not filmed. At this time (5 mo of age), the couple consulted with their primary veterinarian, who prescribed trazodone [50 mg (3.5 mg/kg), PO, q12h]. There were no notable changes to the dog’s behavior, but she was lethargic. Therefore, trazodone was discontinued and fluoxetine [20 mg (1.1 mg/kg), PO, q24h] was prescribed.

Three weeks after the dog started receiving fluoxetine, a trainer was consulted and a behavior modification program initiated. The trainer instructed the clients to crate the dog and leave the home while the trainer remained in the home. During these sessions, each time the dog started to vocalize, the trainer would bang on top of the crate. The owners were not comfortable with this protocol and terminated the program after the second session. They noted the dog would now struggle and pant when they attempted to crate her.

A consultation with another trainer was scheduled and confinement of any kind was discontinued. The second trainer planned to desensitize the dog to predeparture cues (e.g., picking up keys) and started graduated timed departures with the trainer observing via a remote camera. During these sessions, the dog was observed licking her abdomen. The primary veterinarian increased the fluoxetine dosage to 30 mg (1.5 mg/kg), PO, q24h, and treated a moist dermatitis with a tapering course of trimeprazine tartrate with prednisolone. Licking of the dog’s abdomen appeared to cease after treatment.

After the dose of fluoxetine was increased, the dog’s owners noted that she appeared more relaxed around certain predeparture cues (e.g., keys being moved), but she continued to follow the clients to the front door with each departure, panting and salivating. On camera, she was observed pacing by the door within 60 to 90 s after the door closed and started to vocalize (bark, whine, and howl) within 3 to 4 min after the departure. Nonstop vocalization continued while they clients were away (which, at this stage, never exceeded a period of 8 min). Upon their return, she continued to whine and pant for several minutes. Neither urination nor destructive behavior occurred during these departures, although it was not clear if this was because of the increased fluoxetine dose or the short duration of the absence. After 6 wk, this protocol was discontinued due to cost and the clients’ judgment of minimal improvement. They stopped leaving the dog home alone and enrolled her in day care; however, this was not considered financially sustainable.

The owners next discontinued the fluoxetine and referred the dog for a behavior consultation. During the consultation, the dog appeared relaxed; she settled within a few minutes after the start of the consultation and slept for the duration without alerting to movements from either owner. She had a body condition score of 4/9 and physical, orthopedic, and neurological examinations revealed no significant findings. Complete blood count, blood chemistry, and total T4 had been assessed by the referring veterinarian 4 mo before the appointment, with no significant findings. A urinalysis was recommended; however, since urination in the home had ceased, it was declined.

The diagnosis was separation and confinement anxiety because the dog’s clinical signs (vocalization, panting, pacing, destruction, elimination) only occurred when left alone free in the house or in her crate without people (2,5). She displayed clinical signs (panting, pacing, whining) associated with predeparture cues (4,7) and was highly aroused when the owners returned (4). The dog refused to enter her crate voluntarily, and when confined, would bark and scratch on the tray near the door. As these behaviors were displayed when the dog was confined, regardless of the clients’ presence, confinement anxiety was also diagnosed (7).

Differential diagnoses included hyper-attachment disorders (12,13); noise phobias and storm phobias (6,7,14); frustration (2); isolation distress; territorial aggression; compulsive disorder; generalized anxiety; incomplete housetraining; pain (8); cardiovascular, renal, hormonal, or urinary tract disease; hepatic encephalopathy; or seizures (8,15). Since the dog did not display evidence of hyper-attachment (following the owner around the house), this was an unlikely cause for her behavior (16). Noise or storm phobias were ruled because noises were not detected on video when the dog was home alone, and she did not react to storm events or other noises in a fearful manner. Compulsive disorder, generalized anxiety, and territorial behavior were unlikely, as the dog’s clinical signs occurred exclusively in the clients’ absence and within minutes after their departure. Incomplete housetraining was ruled out since urinary or fecal accidents were no longer observed when the clients discontinued leaving the dog home alone. Pain on physical exam was not detected or observed during or after activity. No changes in separation-related behaviors were observed after a trial of a nonsteroidal anti-inflammatory medication (meloxicam, 0.01 mg/kg q24h) and gabapentin (4.4 mg/kg q12h) after her ovariohysterectomy. Cardiovascular disease was unlikely due to normal thoracic auscultation, and renal or hepatic disease were not indicated on recent blood work. No seizure activity had been observed.

Treatment included management, pharmaceuticals, and behavior modification. The clients were advised to create a “safe space” in the home that was not confining but prevented access to the front door. They were instructed on how to shape relaxation in this space by providing a comfortable bed; using a calming pheromone, Adaptil (Ceva Animal Health) (17), or lavender (18); playing classical music (7,19); and providing a long-lasting treat such as a food-dispensing or puzzle toy, for enrichment (19). The clients were instructed to not give treats or any food outside of mealtime, with the hypothesis that pairing of a high-value food item with the space (and eventually, with departure) would help to countercondition the negative emotional response of being left alone. They were asked to take the dog to this spot daily, leaving her to consume the food but not leaving the apartment together. When each client left individually, the person left at home was instructed to avoid interacting with the dog but not to ignore her. This was to reduce arousal immediately before departure (20). Upon returning home, the clients were instructed to speak to the dog in a calm voice and reward calm behaviors (e.g., all 4 feet on the floor) rather than escalating her arousal by matching her greetings (20,21). They were also advised against the use of positive punishment (21,22).

Pharmaceutical therapy with paroxetine was initiated at a dosage of 10 mg (0.4 mg/kg), PO, q24h for 7 d, to minimize adverse effects (23,24); it was then increased to 20 mg (0.8 mg/kg), PO, q24h. As an SSRI, paroxetine blocks the serotonin reuptake transporter, resulting in increased serotonin availability for neurotransmission (25,26). Dosage recommendations for paroxetine in dogs range from 0.5 to 2.0 mg/kg q24h (25,26). Potential adverse effects include inappetence, vomiting, diarrhea, agitation, urinary retention, and dry mouth (26). In addition, lorazepam [1 mg (0.04 mg/kg)] was prescribed with instructions to give this PO 60 min before departures (if both owners were required to leave the house and there was no alternative). Lorazepam, a benzodiazepine, is used to treat anxiety-based disorders in human psychiatry and veterinary medicine (27,28). Benzodiazepines bind to GABA-A receptors to enhance GABAergic neurotransmission (27). Potential adverse effects include vomiting, diarrhea, sedation, ataxia, paradoxical excitement, increased appetite, increased aggression, memory deficits, and chemical dependency (27). Off-label use of both medications was discussed.

At a 2-month follow-up, the owners noted the dog was now vocalizing more when they were present and following the female owner around the home. The couple had chosen their office as the “safe space,” and they had been able to leave the dog with the door closed for up to 30 min while at home. Unfortunately, the owners had continued to leave the home despite being instructed not to do so, citing frustration and compassion fatigue and lack of improvement in the dog’s behavior. Lorazepam was given once, but it was discontinued because the dog was excessively vocal and mobile (would not lie down). The couple had begun interviewing potential new homes for the dog.

Medication was changed to clomipramine at 25 mg (1.1 mg/kg), PO, q12h for 14 d, to minimize side effects, subsequently increasing to 37.5 mg (1.65 mg/kg), PO, q12h. Before starting the clomipramine, paroxetine was decreased in 50% increments per week over the course of 2 wk, then discontinued for 48 h before starting clomipramine. Switching between psychopharmaceutical classes is poorly described in veterinary literature; however, in human psychiatric literature, this is considered an appropriate switch between pharmaceutical classes (29). Clomipramine is a tricyclic antidepressant that blocks reuptake of both serotonin and norepinephrine. It is given at 1.0 to 3.0 mg/kg, PO, q12h, in dogs (26). It is approved as Clomicalm (Novartis) to treat separation anxiety in dogs, with several published studies documenting improvement in various clinical signs of separation anxiety (11,30,31). Potential adverse effects include decreased appetite, lethargy, vomiting, diarrhea, constipation, dry mouth, urinary retention, arrhythmia, and elevated intraocular pressure (25).

Two months after the dog started receiving clomipramine, the clients noted she appeared more relaxed when home alone. She would enter the office on cue, wait for her puzzle toy and, on video, was lying down for 30 min after the clients’ departure. Occasionally she was observed rising, exploring, and sniffing the room. She would bark occasionally, but there was no panting, whining, repeated barking, or elimination. However, the dog’s vocalization when the clients were home (new behavior after starting paroxetine) had not normalized. Therefore, the clomipramine dose was increased to 50 mg (2.1 mg/kg), PO, q12h. One month later, minimal barking was noted when the clients were present, and the dog could be left alone for up to 2 h at a time. On camera, the dog was observed lying down or eating for the full duration of most departures, whereas at other times she would walk around and vocalize once or twice. This behavior was markedly reduced compared to before treatment. The clients could not identify a pattern or trigger for the inconsistency in her behavior — only that it continued to decrease over time. At this visit, the clients were instructed to continue relaxation exercises in the dog’s “safe space” and maintain the dose of clomipramine.

One year after the initial appointment, the dog could be left home alone for 2 to 3 h without observable clinical signs. One of the clients was working from home, so longer-duration absences had not been attempted. The clients had stopped looking to rehome the dog. At 1.5 y after the initial appointment, routine blood work revealed no significant findings. Ideally, blood work would have been repeated within 1 y after starting medication; however, this timing coincided with high rates of COVID-19 at this location, and so the blood work was delayed. The clomipramine dose was maintained due to the couple’s upcoming move. A slow wean-off will be considered if the dog’s behavior remains stable in the new home.

Discussion

This dog suffered from both confinement anxiety and separation anxiety. Confinement anxiety can be misdiagnosed as separation anxiety if the dog is only confined when home alone; however, this dog experienced anxiety in the crate while people were present, before ever being left alone. Since she was subsequently left alone while confined, her confinement anxiety may have contributed to her separation anxiety. Furthermore, since her dam also demonstrated separation-related behaviors, this dog may have developed the same clinical signs regardless of the confinement, although the latter may still have been an epigenetic contributor. During initial treatment, the dog was punished for vocalizing in the crate, which is contraindicated as treatment for behavior disorders (32) due to the potential to induce fear and anxiety. The use of punishment in this case may have exacerbated both the dog’s confinement anxiety and separation anxiety. When the dog vocalized in the crate with the first trainer, it is presumed she was vocalizing because of confinement, not separation, since she could attend day care or be with people other than the clients without any signs of anxiety. If this was the case, this dog’s separation anxiety could be accurately termed isolation distress; however, the literature does not clearly differentiate isolation distress from separation anxiety in dogs.

Pain can contribute to behavior in myriad ways. Behavior changes may be the first sign of a behavior problem, including separation anxiety (8). Radiographs of hips, stifles, and elbows were discussed to rule out congenital orthopedic or other pain-inducing conditions but were declined. Gabapentin and meloxicam were given after ovariohysterectomy but did not decrease the separation-related behaviors. However, the dose of gabapentin was low and unlikely to be efficacious for either pain or anxiety (33).

Minimal behavior modification was recommended in this case because the clients were overextending themselves with an intensive program and experiencing compassion fatigue, which is common for owners of dogs experiencing separation anxiety (34). The behavior-modification plan was simply to teach the dog that the office, whether she was alone there or with people, was a safe place and an area where she would receive high-value food and toys regardless of whether the clients were leaving the home. Whereas many behavior-modification programs for separation anxiety rely on unpredictability (e.g., picking up keys does not necessarily mean imminent departure), this strategy relies on providing predictability and routine with respect to the anxiety-inducing event (7,35).

Several medications were trialed before an effective medication and dose range was achieved. For treatment of depression in humans, psychiatrists often trial antidepressant choice and dose (36). Brain chemistry can influence effectiveness of these medications. Tricyclic antidepressants such as clomipramine inhibit norepinephrine reuptake inhibition and have antihistamine properties. A tricyclic antidepressant can be a better choice in some anxiety patients due to the broader spectrum of neurotransmitters targeted compared to SSRIs, which are more selective for serotonin reuptake inhibition (36). Because of this, paroxetine was specifically chosen over other SSRIs due to its effect on norepinephrine reuptake as well as some mild anticholinergic effects (36). It was hypothesized this anticholinergic effect might help reduce urinary accidents and salivation occurring during separation, though this has not been confirmed in the canine literature. Instead, the anticholinergic effects likely produced the onset of vocalization in situations where the dog had not previously vocalized. This vocalization resolved eventually after changing medications; however, vocalization when the clients were absent was the last separation-related behavior to resolve. It was hypothesized that the vocalization would persist, as reported in some studies (37). In addition, dogs without separation-related behaviors do move around the home vocalizing occasionally (38). Alternatively, it is possible that separation anxiety was not in remission and vocalization only resolved because the dog learned that it was not effective for achieving safety — in this case, the return of a person to the home — and so she “gave up.” This phenomenon is referred to as “learned helplessness,” and has been described in both humans and animals, including dogs (39). However, this dog was observed resting, eating, and exploring her environment, which is more consistent with a reduction of anxiety than an experience of learned helplessness.

No situational, or short-acting, medication was used in this case. Trazodone had been trialed initially, and although the dose was low, did not improve separation-related behavior and resulted in sedation. Trazodone has been used in dogs to aid in treatment of a variety of stressful or fear-inducing events, including separation anxiety (40). The dose of trazodone given to this dog was towards the lower end of the range of effective doses. Lorazepam was also trialed; however, the dog became extremely agitated. Paradoxical excitement from lorazepam was suspected and it was discontinued.

The criteria for discontinuing medication for separation anxiety are patient-dependent. As changes in routine are associated with separation anxiety in dogs (5) and these clients were moving, it was decided not to discontinue clomipramine in this case.