Studies in animals and humans have begun to provide clues to understanding the underlying neurobiology of pathological grooming. Thus far, animal models have provided the most compelling genetic, lesion, and neurochemistry data, while data in humans have come primarily from a limited number of neuroimaging studies.
Serotonin may play a role in pathological grooming. In animal models such as acral-lick dermatitis in canines, compulsive behaviors are reduced with SRIs8, 65
. Response to SRIs has also been reported for avian psychogenic feather plucking66, 67
. Similarly, SRIs often reduce compulsive cleaning and washing behaviors in OCD and compulsive skin picking and grooming behaviors in BDD55
. The reduction of compulsive behaviors with medications that enhance serotonin neurotransmission in animals or humans suggests, but does not prove, that serotonin dysregulation plays a role in their pathophysiology68, 69
Dopamine agonists can cause stereotyped behaviors in animals and can exacerbate OCD symptoms and tics70
. Medications that block dopamine can reduce grooming behaviors in animals71
and may be effective in OCD72
. There is also recent evidence of glutamatergic involvement in pathological grooming behaviors in animals, including the possibility that serotonergic medications may mediate their effects on pathological grooming via glutamate systems2
. The neuropeptide oxytocin may also have a role in grooming, as Marroni et al (2007) found that rats exhibited increased grooming behaviors when oxytocin was injected into the central nucleus of the amygdala74
. Consistent with this finding, grooming behaviors may also serve the purpose of enhancing bonding and establishing and maintaining social relationships75, 76
Mouse models have provided early evidence for genetic contributors to pathological grooming. Greer and Capecchi (2002) produced a mutant mouse strain deficient in expression of the Hoxb8
homeodomain gene. These mice exhibited excessive grooming behavior, to the point of hair loss and skin lesions, and excessively groomed cagemates77
. They found that Hoxb8
is normally expressed in multiple areas of the brain, including but not limited to the caudate-putamen (areas relevant to OCD) and cerebellum. The authors proposed this mutant mouse strain as an animal model for trichotillomania, although it may prove relevant to other human grooming conditions.
As mentioned earlier, Welch et al. (2007) used genetically engineered mice to produce a behavioral phenotype of increased anxiety-like and compulsive grooming behaviors that resulted in hair loss and skin lesions2
. The mouse line contained a deletion in the gene encoding for SAPAP3, a protein highly expressed in the striatum that regulates the trafficking and targeting of neurotransmitter receptors and signaling molecules at postsynaptic excitatory synapses. Subsequently, Züchner et al. (2009) examined the human orthologue of SAPAP3 in three clinical groups: trichotillomania, trichotillomania with OCD, and OCD without trichotillomania, and in healthy controls78
. They found heterozygous variants in 4.2% of the trichotillomania/OCD patients but only 1.1% of healthy controls and concluded that SAPAP3 may increase susceptibility for OCD and trichotillomania. These findings need to be replicated. In the first association study of SAPAP3 in humans, Bienvenu et al. (2008) conducted a family-based association analysis of patients with OCD or grooming conditions (onychophagia, skin picking, and/or trichotillomania)79
and found that four of the six single nucleotide polymorphisms (SNPs) were associated with at least one grooming condition, and all three haplotypes were associated with at least one grooming condition. None of the SNPs or haplotypes was associated with OCD specifically. However, as this sample was selected based on an OCD diagnosis as part of a larger study, most of the individuals with grooming conditions also had OCD, and about one third of those with OCD had a grooming condition. The authors therefore conclude that although SAPAP3 seems to be associated more with grooming conditions than OCD, they cannot rule out the possibility that the association is specifically with a subtype of OCD with comorbid grooming conditions. These intriguing findings, too, require replication.
Neuroanatomically, the basal ganglia and limbic systems appear to be involved in grooming behaviors in animals. Several animal studies have implicated these regions in rodents2, 16
. Circuits within the neostriatum, in particular, may be important in balancing sensory-driven and central pattern-generating circuits16
The neurobiology of OCD, in large part elucidated from neuroimaging studies, implicates similar neuroanatomic areas – specifically, cortico-striatal-thalamic circuit dysfunction81
. This circuit is thought to be responsible for habit learning, routine performance of these habits, and acquisition of stereotyped behaviors82
. The orbitofrontal cortex and subregions of the anterior cingulate are considered major paralimbic inputs to these circuits. A model of dysfunction in this circuit involves over-activity in the direct, relative to the indirect, pathway, setting up a loop of recurrent, difficult-to-control thoughts and behavioral sequences81
, a similar phenomenon could explain the recurrent obsessive thoughts and repetitive behaviors that often involve grooming. There have only been a few neuroimaging studies in BDD. One morphometric MRI study demonstrated greater total white matter compared to controls and a leftward shift in caudate asymmetry83
, suggesting striatal involvement in BDD. A more recent morphometric MRI study, however, did not find significant abnormalities in these regions84
. A functional magnetic resonance imaging (fMRI) study of visual processing in BDD demonstrated greater activity relative to healthy controls in fronto-striatal systems including the orbitofrontal cortex and caudate when viewing their own face85
. In addition, activity in these regions varied with BDD symptom severity. This suggests that a pathophysiological process similar to that in OCD may be operating in BDD when viewing their face, and may reflect obsessive thoughts and urges to engage in repetitive grooming or other compulsive behaviors.
There have likewise been only a few neuroimaging studies in trichotillomania. O’Sullivan et al. (1997) found smaller volume of the left putamen relative to controls86
, and another morphometric MRI study found smaller cerebellar volumes, with an inverse correlation between a primary sensorimotor subregion and symptom severity. These findings, along with a positron emission tomography (PET) study showing increased cerebellar metabolism87
, suggests abnormalities of the cerebellum, which plays a role in complex, coordinated motor sequences. Chamberlain et al (2008) found increased grey matter densities in fronto-striatal and limbic systems, relative to controls88
. However, a functional MRI study probing striatal function with the implicit learning task did not find evidence of abnormal striatal activation89
Future studies of disorders involving pathological grooming are needed to clarify their underlying neurobiology. Neuroimaging studies should pay particular attention to the striatal and cerebellar systems, given animal studies indicating their involvement in grooming behaviors2, 16
. Genes and genetic polymorphisms that may be relevant to animal models of pathological grooming may be particularly promising avenues of exploration in studies of human conditions involving pathological grooming. The efficacy of SRIs and dopamine antagonists for animal grooming suggests they are worth studying in those human disorders for which little treatment research has been done (e.g., ORS). If future studies substantiate a role for the glutamate system and oxytocin in pathological grooming in animals, investigation of medications that modulate these systems may be warranted.