Placebo-Controlled Clomipramine Trial for the Treatment of Feather Picking Disorder in Cockatoos

Introduction

Feather picking disorder, also called behavioral or psychogenic feather picking, is a common condition among captive psittacine birds and may be one of the most challenging problems facing avian veterinarians.^1–4 Feather picking disorder has been described as an exaggeration of normal preening behavior that occurs when the usual components of feather maintenance become pathological.⁵ Feather removal, feather chewing, or chewing of soft tissue are among the abnormal behaviors associated with feather picking disorder.⁶ Feather damage or loss is limited to areas the bird can reach (body and neck); feathers on the head remain normal.⁷

The consequences of feather picking may be strictly aesthetic. However, chronic feather replacement and loss of normal insulation may weaken the affected bird, making it more susceptible to becoming chilled or to disease. In addition, substantial blood loss may result from broken blood feathers and soft-tissue bleeding.⁷⁸

Feather abnormalities, feather loss, and removal of feathers may be caused by medical or behavioral conditions.³⁹ Medical and nutritional causes of feather removal or loss must be explored before a diagnosis of feather picking disorder can be made. Sick birds have been observed to remove feathers, and many diseases have been implicated as contributing factors for excessive feather removal.³⁸ The role of allergies in feather picking has been difficult to document, but interpretation of intradermal skin tests for selected allergens may soon be possible.¹⁰¹¹ All birds with feather loss require a thorough history and physical examination, with appropriate diagnostic testing. A diagnosis of feather picking disorder is made when a physical cause for the problem behavior is not identified or when treatment of medical conditions fails to resolve the feather picking problem.^1–4

Because parrots housed in a variety of captive environments engage in stereotypic behaviors, feather picking, and self-mutilation, the importance of environmental enrichment for avian species has been emphasized.¹² Feather picking disorder may result from management conditions that do not allow the bird to engage in species-typical behaviors or do not provide appropriate target stimuli for these behaviors.¹³ Other suggested causes of feather picking include boredom, frustration, stress, sexual behavior, early weaning, attention-seeking behavior, overcrowding, separation anxiety, and changes in routines.^1–5 With chronic feather picking, the precipitating cause may no longer be present. Many affected birds are fed proper diets, provided with stimulating toys and appropriate housing, exposed to fresh air and sunlight, and appear to have good relationships with family members.¹⁷⁸ Therefore, alternative mechanisms for the disorder need to be explored.

Feather picking disorder has been compared to human disorders involving compulsive and impulsive behaviors. Bordnick, et al., noted similarities between feather picking in birds and trichotillomania in humans.¹⁴ Trichotillomania, an impulse-control disorder, is characterized by removal of hair that results in alopecia. Hair twirling, chewing or mouthing the hair, trichophagia, and skin picking may also occur.¹⁵ Obsessive-compulsive disorder is characterized by obsessions, intrusive thoughts or images, compulsions, and repetitive behaviors performed in an attempt to prevent or reduce anxiety. Hand washing is an example of a commonly reported compulsive behavior in humans, which may share similarities with repetitive grooming behaviors in animals, including feather picking in birds.¹⁶¹⁷

Successful clinical outcomes using clomipramine therapy for compulsive grooming disorders have been demonstrated in nonhuman species. In a retrospective study, five domestic cats (Felis catus) with psychogenic alopecia responded positively to treatment with clomipramine.¹⁸ Seksel and Lindeman reported that cats with compulsive disorders responded to clomipramine within the first 4 weeks of treatment.¹⁹ Dogs (Canis lupus familiaris) with acral lick dermatitis improved significantly when treated with clomipramine for 5 weeks.²⁰ Hewson, et al., examined the efficacy of clomipramine for the treatment of dogs with various compulsive behaviors in a double-blind, placebo-controlled, crossover study. Based on owner assessments, they found that treatment with clomipramine for 4 weeks was effective at reducing symptoms, but it was not curative.²¹

If feather picking disorder is behaviorally analogous to compulsive grooming disorders in other species, then similar neurochemical mechanisms may exist that could result in comparable responses to pharmacological agents. While the exact biological basis is unknown, the responses of some cases of obsessive-compulsive disorder and trichotillomania to serotonergic medications provide evidence that serotonin dysfunction is involved in mediating these behaviors.¹⁶ Serotonergic agents, such as clomipramine, are the most commonly prescribed pharmacological treatment for compulsive disorders.^22–25 Clomipramine inhibits the reuptake of serotonin from the extraneuronal space into the presynaptic axon terminal, resulting in an increase in synaptic activity. The increase in monoamine neurotransmitters that results from reuptake inhibition is believed to eventually cause down-regulation of postsynaptic serotonin receptors, which coincides with clinical improvement.²⁶

Ramsay and Grindlinger used clomipramine (1.0 mg/kg per os [PO] q 24 hours) to treat feather picking in 11 psittacine birds in an open trial. Three of the 11 birds improved, and side effects were rare.²⁷ A Congo African grey parrot (Psittacus erythacus) was also treated for feather picking and self-mutilation with clomipramine (4 mg/kg PO q 12 hours), which resulted in an increase in appetite and paradoxical anxiety.²⁸ The dose was ultimately increased (9.5 mg/kg PO q 12 hours), and buspirone (0.5 mg/kg PO q 24 hours) was eventually added. Feather regrowth occurred, and feather picking was adequately controlled 3 months after the final dosage adjustment.

The goal of the present study was to assess the efficacy of clomipramine in the treatment of feather picking in companion cockatoos (Cacatua spp.) using a double-blind, placebo-controlled study. There are currently no placebo-controlled studies documenting the effects of psychoactive agents on feather picking disease in any psittacine species. A further goal of this study was to obtain objective measures of feather-picking behavior. The behaviors associated with feather picking disorder have not been described in detail. By recording and analyzing the specific behaviors of feather-picking birds, this study provided the framework for a complete ethogram, including antecedents to picking, varied behaviors involved in removal of feathers, and disposition of feathers after removal.

Materials and Methods

Privately maintained birds were recruited for the study through area veterinary practices. Cockatoos with a history of feather picking that had no known medical conditions and had failed to respond to environmental or behavioral modifications were included in the study. Only singly housed birds were selected in order to exclude the possibility that another bird was responsible for the feather damage. Prior to inclusion, all birds were given diphenhydramine (2 mg/kg PO q 12 hours) by their caregivers for a minimum of 14 days to help rule out feather picking secondary to a type I hypersensitivity reaction.³⁸ Institutional Animal Care and Use Committee approval was obtained for the procedures used in this study.

All potential subjects were screened for medical conditions before inclusion in the study. On physical examination, all birds were apparently healthy, with abnormalities limited to the feathers and skin. All birds had negative test results for psittacine beak and feather disease DNA probes, had negative serum antibody titers for Chlamydophila, and all had been previously vaccinated against polyomavirus. Complete blood counts (CBC) and serum biochemical profiles were evaluated on each bird at the beginning and the conclusion of the medication trial. Blood samples were collected via right jugular venipuncture and submitted to an in-house laboratory.

Lead and zinc levels were measured for all birds prior to inclusion. Samples for zinc toxicology evaluation were collected in plastic tubes to avoid contact of the sample with the syringe plunger during venipuncture. These samples were centrifuged immediately upon clotting, and serum was dispensed into a second plastic tube.²⁹ Whole blood in ethylenediaminetetraacetic acid (EDTA) was used for lead analyses. Samples for lead and zinc analyses were packaged with an ice pack and shipped overnight to an outside laboratory^a for evaluation.

Since trauma and stress are common causes of mature heterophilia, birds with mild leukocytosis were included if the leukocytosis occurred in the absence of other clinical signs, and if the physical examination, radiographs, and all other laboratory values were normal. Indicators of systemic infection, such as marked heterophilia, toxic heterophils, or the presence of immature heterophils (i.e., band cells), disqualified the bird until the problem was resolved. The serum biochemical profiles measured uric acid, bile acid, albumin, lactate dehydrogenase, aspartate aminotransferase, and glucose. Acceptable lead levels were those <0.02 parts per million (ppm).^a For zinc, values <3.5 ppm were considered acceptable.³⁰ This maximum value was slightly higher than the reference range (1.25 to 2.29) determined by Osofsky, et al., for Hispaniolan Amazon parrots (Amazona ventralis).³¹ Reference ranges have not been established for most avian species, but according to Fudge and Speer (2001), healthy cockatoos may have higher plasma zinc levels than other avian species.²⁹ Full-body radiographs (ventrodorsal and lateral views) were also examined at the beginning of the study. Only one bird (case no. 3) required sedation during the diagnostic tests, and isoflurane was administered via facemask.

The birds’ body weights were recorded in grams at each visit using a gram scale.^b Fecal Gram stains^c and direct fecal examinations were performed at each visit. Multiple direct fecal wet mounts (two to three per visit) were examined throughout the study in order to maximize the likelihood of detecting Giardia cysts, a postulated medical cause of feather picking.⁸¹⁰

Assignment to either a treatment or placebo group was randomized by species. For each group of four birds of the same species, two would be randomly assigned to the treatment group and two to the placebo group. This process balanced the treatment and placebo groups by species. Researchers and caregivers were blinded as to the agents administered.

Clomipramine and placebo tablets were provided by Novartis Animal Health.^d Clomipramine tablets contained 20 mg of the active ingredient, clomipramine HCl, and placebo tablets contained no active ingredient, but identical inert ingredients. Both tablets were then compounded into 4 mg/mL solutions using raspberry syrup^e and 2% carboxymethyl cellulose as a suspending agent, making them visually indistinguishable. Each bird was given 3 mg/kg of the clomipramine (PO q 12 hours) or an equivalent volume of placebo based on body weight. Medication was administered by the caregiver via syringe or mixed with food.

Caregivers were advised to monitor their birds for side effects resulting from administration of clomipramine. Since there have been no controlled studies using clomipramine in these species, potential side effects were extrapolated from avian case reports and results in mammalian species.^18–212728 All owners signed a consent form prior to onset of the medication, and they were given a standard list of instructions.

Videotaping of each bird was conducted in a laboratory where there was no exposure to chemicals or other birds. The dimensions of the laboratory were approximately 6.3 m × 2.3 m. Each bird was housed in a 97 × 91 × 61-cm acrylic cage.^f The enclosure contained two plastic feeding dishes and two untreated oak dowel perches that were 91 cm in length and 2.5 cm in diameter. For each videotaping session, the caregiver provided one or two of the bird’s favorite toys and the formulated diet that was typically fed to the bird. A super-VHS videocassette recorder was used to videotape the bird.^g The video recorder was attached to a vertical interval time code generator with translator,^h a programmable timer,ⁱ and an 8-mm video camera mounted on a tripod.

Diaries were provided to record daily activities and observations in the home environment during the trial. All caregivers gave medications twice daily, maintained a smoke-free and chemical-free environment, provided toys and daily attention for their birds, and maintained a natural photoperiod. Each caregiver also provided a detailed behavioral history using a comprehensive questionnaire. Progress assessments were collected on a weekly basis, and side effects were recorded. Each caregiver was asked to rate the bird’s improvement based on the subjective changes in feather-picking behavior and feather regrowth. In particular, self-mutilation, feather removal, photoperiod, administration of study medication, and daily interactions with the bird were recorded. All birds were examined by a single clinician (Wilson) at the beginning and the conclusion of the trial. Details of feather and skin conditions were recorded, including types of feathers that were damaged or removed, region of the body affected, approximate percentage of affected feathers, and a subjective rating of improvement.

Video data was collected in the laboratory before starting medication (baseline), at 3 weeks, and at the conclusion of the trial (6 weeks). At each of the three videotaping sessions, the birds were kept for a minimum of 42 hours. Videotaping was done between 8:00 AM and 6:00 PM for two consecutive days. A timing device was set to record for 10 continuous minutes of each hour throughout the day, providing a total of 200 minutes of data for each session. Lights in the laboratory were turned on at 8:00 AM and turned off at sunset or no later than 8:00 PM, providing a minimum of 12 hours of darkness.

The following behaviors were recorded as events: removal of feather (i.e., bird removed feather or portion of feather either at the base or at any location on the feather) and manipulation of feather (i.e., bird manipulated detached feather or portion of feather using beak or claws). Preening was recorded as the duration of time that the bird’s beak was in direct contact with feathers or skin, accompanied by movement of the beak or tongue. Other activities were recorded as follows: nonpreening stereotypy (i.e., a behavioral pattern with no apparent or immediate purpose that was repeated three or more times with minimal variation and did not include preening behavior); preening stereotypy (i.e., a behavior pattern with no apparent or immediate purpose that was repeated at least three times with minimal variation and included preening as a component of the behavioral sequence); and scratching (i.e., any movement of a bird’s claw while in contact with any part of the body).

Results

A total of 20 birds were included in the study. The type of birds included Moluccan cockatoos (Cacatua moluccensis; n=7), umbrella cockatoos (Cacatua alba; n=7), Goffin’s cockatoos (Cacatua goffini; n=4), a citron-crested cockatoo (Cacatua sulphurea citrinocristata; n=1), and a medium sulphur-crested cockatoo (Cacatua galerita eleonora; n=1). Birds ranged from approximately 1 year to >35 years of age. Eleven male and nine female cockatoos participated in the study. Duration of feather picking ranged from 2 to 192 months (mean±standard deviation [SD], 42.3±43.1 months). Eleven birds had no history of self-inflicted soft-tissue trauma, limiting damage to feathers (picking), while nine birds had damaged soft tissue in addition to feathers (mutilation) [Table 1]. Four of the birds were parent-raised birds, and two of the birds had previously been used for breeding purposes. Nine birds received the placebo, and 11 birds received clomipramine.

Duration of preening during the initial videotaping period for all 20 birds ranged from 2.1 to 32.0 minutes per hour (mean±standard error [SE], 12.2±1.7 minutes per hour). During the same period, birds averaged 1.5 feather removals per hour and 0.5 feather manipulations per hour. Stereotypies involving preening that were measured during the initial videotaping occurred at an average rate of 46.4±29.0 seconds per hour. Total stereotypy duration for these baseline sessions ranged from 0 to 15.4 minutes each hour (mean±SE, 2.6±0.9 minutes per hour). Baseline values were summarized for birds that damaged feathers (pickers) and for birds that mutilated soft tissue (mutilators) [Table 2]. A posthoc multivariate ANOVA of the baseline behaviors for mutilating birds (n=9) and feather-picking birds (n=11) showed no significant differences between the two populations (Wilks’ λ=0.731; P=0.29).

No adverse events were reported during the study period. Treatment with clomipramine had no significant effects on laboratory values or body weights measured at time 0 and at 6 weeks based on multivariate tests (Wilks’ λ=0.28; P=0.441).

Based on caregiver assessments, eight of the nine birds that received placebo remained unchanged at the end of trial. According to the subjective assessments of the caregivers and the clinician, one placebo-treated bird (case no. 9) was substantially improved at 6 weeks, but not at 3 weeks. Eight of 11 birds treated with clomipramine were considered improved by 6 weeks. Clinical signs had resolved in one bird, three birds had substantially improved, and four had improved somewhat. One bird that received clomipramine (case no. 11) was worse at 6 weeks, according to both the caregiver and the clinician. This bird had mutilated through soft tissue into an air sac. Several prior occurrences of self-mutilation of this area were noted before beginning the clomipramine trial. The remaining two clomipramine-treated birds were unchanged at the conclusion of the trial. Results of Mann-Whitney U tests comparing improvement scores for placebo and clomipramine-treated birds were significant at 6 weeks (U=24.5; z= −2.047; P=0.021) and at 3 weeks (U=27.5; z= −1.903; P=0.028). Clomipramine-treated birds had significantly greater improvement at 6 weeks than at 3 weeks (z= −1.897; P=0.029). Of the eight clomipramine-treated birds that were improved at 6 weeks, seven of them began to improve at 3 weeks. There was significant agreement between the independent assessments of the caregivers and the clinician for overall improvement at 6 weeks (Kendall’s τ, b=0.493; P=0.005).

Testing within-subject variables (e.g., preening duration, feather removal, feather manipulation, and preening stereotypy duration) with a double multivariate ANOVA for repeated measures at time 0, 3 weeks, and 6 weeks showed that clomipramine treatment had no significant effect on these behaviors (Wilks’ λ=0.389; P=0.118) [Table 3]. Birds that received clomipramine (n=11) had been feather picking for a mean of 35.9±30.0 months. There was no significant correlation between duration of feather picking and reduction in time spent preening from baseline to 6 weeks for the clomipraminetreated birds (r=0.314; P=0.174). There was also no significant correlation between duration of feather picking and caregiver assessment of improvement at 6 weeks, using a nonparametric Spearman’s correlation (ρ= −0.226; P=0.25).

For birds that received clomipramine, six were pickers and five were mutilators. A Mann-Whitney U test comparing disease type (picking vs. mutilation) with caregiver assessments of improvement at 6 weeks was not significant (U=10.5; z= −0.851; P=0.20). Multivariate ANOVA for repeated measures using disease type (picking vs. mutilation) as the between-subject factor and within-subject variables (e.g., preening duration, feather removal, feather manipulation, and preening stereotypy duration) at baseline, 3 weeks, and 6 weeks was not significant in the clomipramine-treated birds (Wilks’ λ=0.442; P=0.903).

Discussion

This study represents the first reported placebo-controlled clomipramine trial for feather picking disease in a psittacine species. Independent assessments by caregivers and a single clinician indicated a significant improvement (i.e., refeathering) with clomipramine treatment compared to placebo at 3 and 6 weeks. These results support the hypothesis that signs of feather picking would improve using a serotonin-selective medication that is commonly used for related disorders in other species. Approximately 64% of birds treated with clomipramine improved, but only 36% were judged to have substantial improvement. These findings were similar to initial results of clomipramine trials in humans afflicted with trichotillomania. Relapses were common in trichotillomania patients after 8 to 12 weeks of treatment, and long-term improvement was difficult to maintain.²⁵

Responses of feather-picking birds to clomipramine treatment needs to be assessed beyond 6 weeks in order to evaluate long-term safety and efficacy. Birds that responded began to show noticeable improvement 3 weeks after starting clomipramine, and some birds continued to improve between 3 and 6 weeks. This delayed response was expected based on the current theories of antidepressant actions on postsynaptic receptors.²⁶ The influence of other neurotransmitter systems on feather picking also must be evaluated using placebo-controlled trials. Dopamine antagonists and opioid receptor antagonists have been used successfully to treat individual cases of feather picking disease.^32–34

With respect to laboratory preening measures, there were no statistically significant differences between the clomipramine and placebo groups. Duration of preening episodes and stereotypic behaviors may not provide accurate measures of improvement in feather picking, however. Comparison of birds with feather picking and normal birds may potentially identify measurable differences in behaviors that could then be used to assess improvement in birds being treated for feather picking disease. It is also possible that behaviors in the laboratory were not representative of the birds’ normal behaviors in the home environment. Unfortunately home videotaping was not practical for this project. Additional studies are necessary to determine how to better measure preening behaviors in a controlled environment. Habituation to the laboratory environment was assumed because all birds engaged in a broad spectrum of apparently normal activities during their stay (e.g., ate, chewed objects, talked, preened, vocalized, and played with toys). A study that measures physiological parameters indicative of stress over an extended laboratory stay might provide a more accurate measure of habituation and the length of time that is required on average to achieve habituation to a new environment. Many caregivers were reluctant to leave their birds even for a short time; so requiring the birds to stay for an extended period would have adversely affected the recruitment of subjects. The relatively small sample size of the study, combined with pronounced individual variation in preening and stereotypic behaviors, may also have influenced the results and reduced statistical power. Multiple species participated in this trial, and species differences could account for some of the variation encountered.

It is interesting to note that one bird’s (case no. 9) condition resolved with the placebo. Posthoc inquiry revealed that an employee who had been responsible for the bird’s care during the day had left just before the trial began. The owner had subsequently taken over all care for the bird. No other changes were identified that could have accounted for the cessation in feather picking. Posthoc statistical analysis with this bird omitted did not result in significant differences in preening measures between groups.

Duration of signs did not statistically correlate with response to clomipramine treatment. Based on this finding, chronicity may not be a good indicator of potential response to treatment. However, the shortest duration of feather picking for birds in this study was 2 months (case no. 10), and this bird’s signs were the only ones to completely resolve during the trial. The bird with the most chronic condition in the clomipramine treatment group (case no. 14) did not improve. A larger sample size may be necessary to demonstrate the effect of duration of signs on treatment response.

Feather picking disorder, as defined clinically, can involve a variety of abnormal self-directed behaviors, including feather damage, feather removal, and soft-tissue damage. While there was no significant difference in response to clomipramine between birds that mutilated soft tissue and those that damaged only feathers, it is still possible that these are distinct conditions and should be evaluated separately. Self-injurious behaviors in humans, including such varied conditions as trichotillomania, skin picking, head banging, lip chewing, and onychophagia, are not treated as single entities.³⁵³⁶

A higher dosage of clomipramine may be required to resolve some feather-picking problems. Dosages for psychoactive medications are not well established for psittacine birds and are often extrapolated from mammalian dosages. Species variations in pharmacologic half-life values of drugs have been ascribed to differences in metabolic pathway rates.³⁷ Pharmacokinetic data for psychoactive medications is lacking for psittacine species, making accurate dosage determination difficult. However, results of this trial suggest that clomipramine, at doses higher than those typically used in mammals, may be useful in the treatment regime for feather picking disease in some species of psittacine birds.

Conclusion

This study provides some evidence that serotonergic agents may be useful adjunctive therapies for chronic feather picking disorder in cockatoo species. However, environmental and behavioral modifications must always be addressed with any behavior problem, and the absence of adverse events and laboratory changes during this trial does not ensure that clomipramine is safe in all psittacine birds. Clomipramine should be used with caution until larger trials have been performed, more species have been evaluated, and toxicity data is obtained.

Acknowledgments

The authors thank Gary Burton, Tony Hughey, and Petra Harden for technical assistance during the study, and Novartis Animal Health for providing clomipramine and placebo tablets.