MacLean (1985) has proposed that the neural substrates mediating separation distress, maternal care, and play belong to the same paleomammalian portion of the limbic system. According to his theory, these social behavior tendencies are all elaborated within the cingulate cortex and related neural structures. He has argued that the "separation call" or distress vocalization is the mammal's "earliest and most basic" vocalization pattern. More primitive forms of animal life (e.g., reptiles)—lacking a cingulate cortex—do not display evidence of maternal care, separation-distress vocalization, or play. Socially directed vocalization patterns may have originally evolved to maintain close contact between the mammalian mother and her immature offspring. in addition to maternal caregiving and separation distress, play between con-specifics also appeared with the evolution of mammals, perhaps serving to facilitate social harmony among litter mates.
Panksepp (1982) also views distress vocalization as stemming from a primal mammalian emotional system but more specifically originating in those areas of the brain that mediate panic and explosive behavior. in addition to the cingulate gyrus, other brain sites that contain dense concentrations of opioid receptors are implicated in the organization of attachment and separation distress. These areas include the amygdala, dorsome-dial thalamus, hypothalamus, and the central gray area:
It is proposed that sites from which distress vocalizations and explosive agitated behavior can be elicited represent the approximate trajectories of panic circuitry, the major adaptive function of which is to sustain social cohesion among organisms whose survival depends on reciprocity of care-soliciting and care-giving behaviors. (Panksepp, 1982:414)
These various areas of the brain and interconnecting circuits are stimulated by the differential presence or absence of pertinent social stimuli evoking or allaying social distress and panic.
Limbic Opioid Circuitry and the Mediation of Social Comfort and Distress
Both MacLean and Panksepp have noted the existence of a highly concentrated pattern of opiate receptors in the neural circuits believed to mediate social comfort, separation-distress reactions, and various other relevant emotional responses. Panksepp's lab has performed numerous studies demonstrating a direct linkage between brain opioid activity and the elaboration of separation distress, contact comfort, and play (Panksepp et al., 1984). After many years of study investigating imprinting and isolation-induced distress in ducklings, Hoffman (1996) has concluded, along with Panksepp, that social attachment and bonding is probably mediated by opioid receptors activated in the presence of adequate social stimuli. Social bonding and separation distress also appear to be closely related to opioid activity in monkeys. Keverne and associates (1989) at Cambridge University measured significant elevations of beta-endorphins in the cerebral spinal fluid of Ta-lapoin monkeys upon being reunited with conspecifics after a period of isolation. In addition, they have found that social grooming among paired monkeys is probably mediated by an endogenous opioid mechanism. Monkeys exposed to naloxone blockade engaged in more grooming interactions, whereas low doses of morphine reduced such affiliative exchanges. A similar differentiating effect is observed among rhesus monkeys, where cooing (a primate separation-distress vocalization) is decreased by the administration of morphine and increased by naloxone (Kalin and Shel-ton, 1989).
Panksepp and colleagues (1980, 1988) found that low doses of morphine significantly reduce separation-distress vocalizations by puppies, guinea pigs, and chicks. In addition, the researchers demonstrated that socially deprived kennel dogs become more socially responsive and obedient after being administered low doses of morphine and more uncontrollable when given naloxone (Panksepp et al., 1983). Knowles and coworkers (1987) observed that well-socialized adolescent dogs (6 to 8 months old) exhibit increased care-seeking behavior (tail wagging and social contact) under the influence of naloxone, whereas morphine reduces such social behavior. It should be noted that initial efforts to demonstrate a relationship between naloxone and separation-distress vocalization in puppies failed. Although these early efforts failed to show a relationship between naloxone blockade and separation-distress vocalization, the researchers did find a significant relationship between naloxone and other canine social behavior patterns, including separation distress, when an intermittent operant element was involved:
Recently we have measured other care-soliciting behaviors in the dog, and we find that naloxone can facilitate tail-wagging and face-licking. Also, we have recently observed that naloxone facilitates vocalizations in dogs when there is the possibility of a clear operant component. For instance, in several litters of puppies being tested for social motivation, we have observed naloxone-treated animals to vocalize more frequently when they are intermittently prevented from making social contacts. Accordingly, our failure to see a clear facilitation of DVs [distress vocalizations] in puppies following naloxone in a simple separation situation does not constitute a negation of the hypothesis that opioid-blockade should increase care-soliciting behaviors. (Panksepp et al.,
Additional evidence for an opioid mechanism mediating social emotion and attachment has been reported in rodents by D'Am-ato and Pavone (1993), who measured differences in pain thresholds between mice paired with siblings versus controls paired with unrelated mice of a similar age. Following 2 months of separation, reunited sibling mice exhibited a significantly higher pain threshold than controls. The full expression of the analgesic effect following the reunion of siblings took approximately 2 hours. The researchers found that the effect was blocked with the administration of naloxone, confirming the involvement of an opioid mechanism.
Not surprisingly, separation distress and panic manifest themselves behaviorally like symptoms of withdrawal from narcotics.
In addition to the aforementioned subcortical circuits, cortical systems are probably also involved in the regulation of separation distress, enabling dogs to cope with separation without experiencing excessive worry or panic. The amount of separation distress expressed by a dog appears to depend on the additive effects of social loss together with the relative novelty of the situation (time and place of separation) in which isolation occurs. For example, many adult dogs exhibit their first dramatic episodes of separation distress (excessive barking or howling, destructive chewing, or elimination problems) only after the family moves into a new home. Other dogs will remain relatively quiescent as long as they are confined in a familiar part of the house and the owner keeps a regular schedule. However, if they are confined in an unsocialized area (e.g., the basement or garage) or if the owner leaves early or returns late, they may become overly anxious and panic. The most intense separation-distress reactions appear to occur when a dog is left alone in an unfamiliar place. This general observation has been experimentally demonstrated with puppies by J. P. Scott and his associates (1973).
The central issue being raised here is whether two converging neural circuits (subcortical and cortical) might contribute modu-latory influences over separation distress. Specifically, the hypothesis being advanced is whether two complementary circuits are in volved: a circuit that is responsive to the loss of socially significant stimuli, and a second one that is activated by contextual considerations like location (familiar/unfamiliar) and schedule (predictable/unpredictable). An analogous situation occurs in the classical conditioning of fear. During the conditioning of fear involving an acoustic CS, the auditory signal (e.g., tone) generated in the ear is directed via thalamic auditory relays to the amygdala, where it is associated with the fear-eliciting US (e.g., shock). The fearful association between the tone and shock is thereby learned and permanently stored as an emotional memory connecting the CS with the US. However, the animal must also learn in what contexts the fear-eliciting CS is really threatening. Such contextual learning depends on the additional involvement of a complex hippocampal-cortical circuit, which results in the production of consciously accessible memories defining the exact situations (time and place) in which the CS predicts an actual threat (LeDoux, 1994). Under conditions of chronic stress, these various contextualizing functions of the hippocampus may be disrupted. A great deal of evidence suggests that excessive and chronic stress produces degenerative effects on hippocampal regulatory functions (McEwen, 1992). On the other hand, as noted above, these same conditions of stress appear to augment amyg-daloidal functions, potentiating emotional learning and responsivity associated with conditioned fear. In the case of separation-distressed dogs, it would seem reasonable to suppose that hippocampal functions may also undergo a similar progressive deterioration as the result of chronic stress associated with the disorder. The neural degenerative effects of stress may help to explain why separation-reactive dogs fail to adjust to the effects of chronic separation distress. It may also provide a possible clue for the higher incidence of other fears and phobias (especially fear of thunder) presenting with separation anxiety.
In the case of separation distress, traumatic experiences associated with the loss of significant social stimuli may be stored as inaccessible emotional memories (manifesting as a persistent and unmodifiable dread of being alone). Along similar lines of contextual learning in the case of fear, the context or situation in which separation occurs may also serve to modulate significantly the amount of anguish and distress expressed via consciously accessible memories of past experiences with separation and the participation of higher cortical coping mechanisms and control. Context may be defined here in terms of both spatial as well as temporal parameters, that is, referring to the place where the dog is confined, as well as the owner's schedule of departures and returns. in general, one might predict that as contextual familiarity and regularity (place and schedule) increase, the magnitude of separation distress should decrease. This seems to be precisely what occurs when dogs are successfully treated for separation-related problems. Whether such a regulatory coping circuit exists is not definitively known, but i would be surprised to discover that it did not.
In addition to the HPA system readying the body for emergency action, another important CRF-mediated circuit within the brain itself modulates emotionally stressful states resulting from the distress and panic associated with social separation. Panksepp and colleagues (1988) describe an experiment in which CRF was intraventricularly injected into the brains of young chicks. The chicks exhibited pronounced distress vocalizations for 6 hours, even though they were in the presence of social stimuli that normally inhibited such reactivity. Within the CRF brain system, NE counterbalances and restrains CRF activity. Under conditions of prolonged stress, NE is depleted, resulting in the disruption of homeostatic balance between NE and CRF. Some CRF projections terminate in the area of the locus coeruleus and, perhaps, under conditions of chronic stress, CRF may exhaust the production of NE or, in conjunction with a parallel neuromodulatory system (e.g., the opioid system), impede efficient NE production. In addition to CRF-mediated activation of the locus coeruleus, CRF projections innervate the dorsal raphe.
it is known, for example, that endogenous opioids exercise a strong inhibitory restraint over NE-producing neurons (McGaugh, 1990). Under conditions of stressful regulatory imbalance, CRF-facilitated influences may prevail over the mood-enhancing influences of NE. Lowered levels of NE are associated with depression, and not unexpectedly many dogs suffering chronic separation distress also develop signs of depression. In addition to CRF circuits, other neuroendocrine (prolactin and oxytoxin) circuits may also play important roles in the modulation and expression of separation distress (Panksepp,
Strong evidence suggests that early stressful experiences produce lasting changes in the CRF stress-mediating systems of the brain. Relatively brief doses of separation distress produced by periodically removing rat pups from their mothers before they reach 21 days of age produce long-term changes in the rat's brain (Nemeroff, 1998). These early exposures to stress appear to alter permanently the CRF gene expression and, consequently, the rat's stress management system. These changes include the elevation of central CRF and proliferation of CRF receptor density, thereby intensifying the animal's response to CRF throughout its life. In addition to CRF system changes, early stress exposure elevates stress-induced ACTH secretion as well as plasma cortisol levels. Interestingly, the SSRI paroxetine (Paxil) appears to return CRF levels effectively to normal while adjusting the animal's increased receptor sensitivity to more normal levels, as well. In addition, the medication produces an overall reduction of undesirable fearful and anxious behavior. These palliative effects produced by paroxetine are apparently drug dependent. When treatment was discontinued, the earlier CRF levels, receptor sensitivity, and associated stress-mediated behavior returned to pretreatment levels. These findings suggest the possibility that early and repeated or traumatic exposure to separation may incline dogs to become overly responsive to stress-eliciting experiences, perhaps predisposing dogs to develop a variety of fear-related behavior problems and problematical separation anxiety as adults.
Finally, recent studies by Price and colleagues (1998) suggest that stress-related CRF system activation appears to exert a direct in hibitory influence over serotonin production in the dorsal raphe. This CRF-mediated restraint over serotonin production might obviously affect remote areas of the brain dependent on serotonin activity originating in the brain stem. The intimate link between CRF and serotonin output may help to explain the aforementioned stabilizing and serotonin-enhancing effects of paroxetine via CRF system regulation (or normalization). The evidence suggests that paroxetine might be a useful alternative for the management of stress-related behavior problems in dogs; however, currently it is not commonly employed by veterinary behaviorists (Overall, 1997; Dodman and Shuster, 1998). Considering the potential benefits, and the apparent lack of mitigating adverse side effects, perhaps some exploratory clinical trials with the drug should be carried out and evaluated.
Clearly, a possibility exists that some functional dysregulation of the HPA system plays a role in the expression of adult separation-distress problems. Persons suffering depressive disorders frequently exhibit HPA dysregula-tion of cortisol production. To determine the presence of such dysregulation, depressed patients are administered an oral dose of dex-amethasone (a synthetic cortisol), and plasma cortisol levels are measured at various times during the day. In persons exhibiting normal-functioning HPA system regulation, cortisol levels are suppressed; in persons exhibiting HPA system dysregulation, however, plasma cortisol levels are not suppressed. Some evidence indicates that children exhibiting severe separation anxiety show an abnormal response to the dexamethasone-suppression test (Livingston, 1991). Perhaps adult dogs exhibiting chronic or severe separation anxiety may suffer a similar dysfunction of HPA activity. The dexamethasone-suppression test might offer a diagnostic method for isolating such dogs from other dogs presenting with a more psychogenic etiology and symptomatology. Further, an abnormal dexamethasone-suppression test result appears to be moderately predictive of a positive response to antidepressant medications in human patients (Risch and Janowsky, 1986).
Was this article helpful?