Thursday, July 1, 2010

The role of attachment theory in addiction


Attachment is a complex set of behavioral adaptation that increases survival of a species by having parents take care of infants in their early and vulnerable years. The infant is provided with “mental representation” which guides the infant in relationship to others and affects regulation. Secure infant are free to explore and attach to others in a manner helpful to meeting their own emotional needs. Infants who have not had healthy attachment, often due to early maternal neglect, are prone to be anxious and may develop later psychopathologies. Anxiety is central to affect disregulation due to lack of early attachment. The propensity of substance abuse may be high in those lacking early maternal attachment. Drug use alleviates anxiety by stimulating similar biochemical mechanisms used in healthy attachment. Oxytocin and vasopressin are biological substrates thought to be involved in attachment.


Attachment theory is formulated by John Bowlby and later improved by Mary Ainsworth, an astute and meticulous experimentalist (Panksepp, 1998). Bowlby sought to update Freud’s theory of attachment anchoring it on the science of his day. According to Freudian theory, an infant’s attachment derives from love of sensuous oral gratification. Bowlby thought that feeding habit did not adequately explained attachment of an infant to the mother. Attachment in his new understanding was tied more to biology where the growth of brain occurs in critical periods and stages influenced by the environment. The social environment changes over the stages of infancy and induces the reorganization of brain structure.

This new reformulation of attachment is a radical departure from psychoanalytic thinking since it ascribes profound influence mothers’ have over their infants’ behavior, in contrast to psychoanalysis, where internal psychic structure of the infant only determines the level of attachment Thus, this new theory of attachment is less psychically determined compared to psychoanalysis because it takes environmental influence into consideration. However, it should be noted that the influence the mother holds on her infant was appreciated by later school of psychoanalysts, especially object relation theorist, who diverged from classical Freudians. This paper will explore the neurobiology of attachment especially as it relates to drug addiction. The neuropeptides oxytocin and vasopressin have long been implicated in attachment and will also be reviewed in the paper.

I am interested in studying drug addiction in relation to attachment theory. My interest grew out of my work as a therapist where most of the clients I see are addicted to drugs. Since there is a great deal of time and money spent going through the ritual of acquiring drugs and using drugs, I started to think of addiction as a form of attachment or bonding. The addict is soothed and gratified by taking a substance knowing perfectly well that in the long run this behavior is often met with detrimental end. Hence, my hypothesis is that those who have not been able to attach as infants’ to their caregiver may be prone to addiction since the preferred substance used takes place as a “surrogate mother” offering a semblance of love and security to the addict.

Neurobiological Basis of Attachment

An infant’s attachment to a caregiver results in innumerable befits as a result of optimal adaptation to an environment. Hence, insecure-attachment may have clinical relevance since psychopathology is often an inability to form proper attachment and bonding. Maturation occurs as a result of timed neurodevelopment processes with an intricately coordinated series of contacts between the maternal-social environments (Panksepp, 1998). Hence, the caregiver plays a pivotal role in responding to the need of the infant with careful mediating functions such as soothing, validating, conforming and stimulating – sequential interventions necessary for neural development.

Infancy is an important period of neuronal development and the development is affected not only by internal biological factors but also social mediation. Therefore, the fact that the infant is born with the potential for wiring with outside environmental stimulation is a testament for appreciating the gene-environment interaction. However, this does not necessarily mean that the infant is born, as it were, with a “clean slate”; rather, the infant is born with some pre-wired connections and other postnatal neural connection to occur at later critical stages of neurodevelopment. It is important that postnatal neural connections occur, among other things, from environmental input with appropriate environmental stimuli at appropriate stage of development to determine attachment.

Literature Review

Numerous studies have been conducted to determine the chemistry of pair bonding in animal model. Insel (1997) reviews experiments using voles to study the effect of oxytocin and vasopressin on pair bond formation. Prairie voles were selected because of highly developed social behavior observed in the field and also because they showed similar affiliative response in the laboratory. Prairie voles are also monogamous. In comparison, montane voles of related specie, sharing similar nonsocial behavior with the paiere voles, nonetheless were observed showing little social contact and not having been monogamous.

To test partner preference, researchers placed three cages attached to each other with openings on the side of the cages for a female vole to roam around. The female was placed in the middle cage while two male voles were tethered in their respective cages. The purpose of this experiment is to observe if the female vole would spend time alone or, with her original partner or, a novel stranger. The female was placed in the cage for 24-hours. In a 3-hour preference test, parier voles spent more time in partner’s cage. A similar experiment with a montane vole demonstrated that the female vole spent more time in her cage without any preference to either of the two males.

Continuing the partner preference test, females were separated from their mates for 7 days and were placed in the cage with access to the male cage. As a pretest an intruder was also placed inside the cage of the experimental male. Following 24-hour mating, the experimental male receives yet another intruder male placed in his cage. In the highly affilitive parier vole, aggression was minimal before mating but increased 40-folds within 24 hours and the increase was sustained for several months. In contrast, the same experiment with the montane vole did not show an increase in aggression toward the intruder male.

The experiment Insel (1997) reviews in A Neurobiological Basis of Attachment is elegant but does not prove that oxytocin and vasopressin are attachment neuropeptides per se. For instance, finding ways to index the neuropeptides using whole blood analysis or measuring metabolites of oxytocin and vasopressin in CSF would help in comparing neuropeptide levels (Mofitt, et. al, 1998). For instance, it is an important finding if measured oxytocin correlates with high level of pair bonding across species to determine if oxytocin is a robust indicator for attachment.

Previous behavioral studies have assessed the degree of serotonergic activity in response to a fenfluramine challenge. Brain serotonin and aggression are the best-correlated studies to date. In the study researchers infused serotonin agonist fenflurimine into subjects veins and the fenflurimine would bind to receptors in the hypothalamus and cause the release of the hormone prolactin from the pituitary gland into circulation (Pine et. Al, 1997). This study has shown inverse relationship between fenflurimine induced prolactin and behavioral aggression. Thus, if oxytocin or vasopressin is found to be related to attachment like serotonin and aggressive behavior, then the finding would be considered robust.

Whole blood serotonin rates have been related to violent behavior in men but not in women (Moffitt, et al 1998). The hormone oxytocin has been associated in several medical conditions and can easily be measured in a specific enzyme immunoassay (EIA) ( Alfven, 2004 ). Since it is my hope to demonstrate an association with either oxytocin or Vasopressin and how these neurochemicals relate to attachment in the prairie vole population, it is helpful that there is also a test in which oxytocin EIA can be performed on microtitre plates coated with a rabbit polyclonal antibody which does not significantly cross react with vasopressin.

Assays such as the one available for oxytocin are also available for vasopressin in which a polyclonal antibody to vasopressin is used to competitively bind and demonstrate, in an inversely proportional manner, the concentration of vasopressin in the sample. A variable in this project may be that rodents and humans both release vasopressin in response to physical and emotional stress, and therefore the degree of stress of the sampling on the final numbers must be integrated into the study methodology. .

By observing prairie vole populations for issues of attachment within different animal models, and subsequently measuring for levels of both oxytocin and vasopressin, it is my hope that I will be able to correlate which neurochemical is more closely associated with stronger or lesser attachment on a statistically significant basis. This research then will be a precursor for the study of the neurochemical basis of attachment in human beings.

Preference for a certain original male for mating may not be an entirely accurate measurement for attachment. For instance, the female montain vole does not appear to have a particular preference for the two males she has access to in an experimental setting. It may be the case that she can form attachment to either male but chooses to spend more time in her cage to avoid inciting aggression in a closed perimeter where she can also become a victim. In other words, oxytocin level may have little in the way the montain vole expresses an attachment behavior since behavior is a function of an evolutionarily adaptive strategy.

Perhaps the study would have been more reveling on attachment if it also included mother-infant voles. A montain vole that shows no preference for male in the experimental setting may show preference to her infant compare to a male vole or even of another infant of a different mother. Hence, unlike the prairie vole, attachment may be limited between a mother and an infant in montane vole.

Neuropeptide level may also fluctuate for a reason. Thus, quality of attachment may vary across species because of different evolutionary adaptive functions. For instance the prairie and mountain voles differ in bonding preference because monogamous relationship in the parier vole may decrease aggression thus maximizing survival by cooperative child rearing. Whereas the montain vole had to adapt in an environment that polygamous mating behavior is a successful mating strategy for increasing survival of the specie.

Bales & Carter (2003) investigated postnatal oxytocin by manipulating neonatal parieri vole male. In the experiment neonatal vole males received either an injection of Oxytocin (OT), and injection of Oxytocin antagonist (OTA), Saline or just handling. These male voles were tested for partner preference as adults after 1-hour cohabitation with a nonestrous female.

The researchers report in their result, with the exception of those voles treated with saline, they did not find a significant difference in social contact with the rest of the experimental group. However, when they studied side-to-side contact in minutes, voles injected with OT were found to have preference for a familiar female vole than a stranger during a 3-hour test. There may be at least two explanations for the observed phenomenon. The male vole may have a better chance copulating with the familiar female than a stranger since the male vole has had prior contact and presumably had invested courtship time; or the female may recognize the scent of the male and as a result emit pheromones to attract the male vole to come to her. Thus, in this scenario the role of oxytocin may be incidental.

An experiment with two familiar voles to a particular male may shed light to the question of partner preference vis a vie oxytocin. What would be the male preference of two familiar females where one of the female had social relations and the other sexual relations with the male vole? If the preference is strongly inclined to the female with past copulation history, then oxytocin may be suspected to have a role specific to sexual pair bond formation as supposed to familiarity or kinship pair bond formation.

The handled group also showed a similar preference, (but statistically nonsignificant) for familiar female partner. However, there are a number of issues to consider in this research. For instance, despite injecting a group of male voles with oxytocin antagonist, there social contacts were not different from the group of voles injected with oxytocin. Had the OTA group been social isolates, the role oxytocin plays in pair bond formation would have been significant. However, the social contact of the OTA group is just as extensive as those with OT injection. This experimental observation suggests that there maybe other mechanisms for pair pond formation in addition to oxytocin.

One way of studying the role Oxytocin plays in pair bond formation is by having two separate groups of pairer voles, one experimental group having numerous parair voles interacting in a social setting and, another group of voles isolated individually for a period of time. Then, compare oxytocin level between these two groups. Then reverse the experiment where the socially interacting voles are removed and placed in isolation. And those isolated are placed in a social setting. Again, measure the oxytocin level to determine differences on oxytocin level.

If the result is high oxytocin level with high pair bonding and low oxytocin level with low pair bonding, and if the result is statistically significant; then oxytocin level maybe said to correlate with pair bonding for that particular specie under study. If the result can be replicated with other species, oxytocin maybe sine qua non attachment neuropeptide. It is probably the case that since behavior is a complex social sand biological phenomenon, the underlying biological substrates may also be complex.

An important experiment may also involve placing a newborn parier vole in isolation for a period of time and compare the level of oxytocin between the isolated prairie voles with a controlled praier infant vole nurtured by its mother. If oxytocin is involved in attachment and pair bond formation, what is the effect on attachment removing an infant parier vole at critical stages of growth? Can the infant parier vole resume healthy attachment if joined back with the mother? Are there differences in the level or quality of attachment depending on, when, or how long, a parier infant vole is isolated (separated) from the caregiver or from the community parir voles?

There may be confounding factors such as fluctuating oxytocin level due to circadian rhythm, fluctuating oxytocin during, for instance, mating season or after having given birth. Oxytocin level might also be affected by stress, unfamiliar social environment, and lack of movement, all of which are difficult to control in an experimental setting. Could it be that attachment or pair bonding has different mechanisms in male and female. A male parier vole may be monogamous because of high level of testosterone whereby monogamy allows continual access to a certain female. Thus, oxytocin maybe correlated with testosterone or testosterone may affect the level of oxytocin. Thus, testosterone may also be important is determining in pair bonding.

Addiction and Attachment

Bonding is a dialectical experience whereby shared needs are best met between two people, initially between a caregiver and an infant. As the infant grows, attachment to the father and siblings also become pronounced and later, attachment would include peers and other people as well. The binary response of gratification and distress of an infant would give way to complex interaction over time. It is in the context of attachment that the infant then, the child learns to mediate the whole gamut of affective response.

Human survival very much follows a utilitarian philosophy of minimizing pain and maximizing pleasure. Thus, early childhood emotional neglect or abuse may cause the person as a teenager or as an adult problems forming healthy attachment. Inability to attach is manifested by numerous failures and hardship an addict encounters over time. To live a successful life the addict would have to abide by the rules of the society and help stave off ostracism and punishment for using drugs. For the addict inability to attach means inability to calm or soothes himself under stress. Inability to attach is an inability to self-regulate emotion learned in early infancy. Thus, the addict would resort to drugs as a way of soothing self from uncomfortable feelings even when the larger society highly disapproves of such behavior.

A drug is an inanimate substance and does not interact with the addict the same way a mother interacts with her infant, even though, the addict is also attached to the drug. Healthy attachment has to be interactive with another person. Thus, the addict’s attachment is that of an inferior kind or, at best artificial.

Anxiety plays a pivotal role in lack of attachment. Children whose parents do not consistently respond to their signals and emotional needs may lead to distrust that others will be there for them; as a result of this mistrust, some children may isolate themselves while others become extremely dependent.

Attachment theory does not conclusively explain drug addiction. It is at best a working hypothesis, which requires a great deal of research in this area. Sociological reasons for addiction vis a vie environmental influence and access to drugs poses a serious challenge in distilling addictive behavior best explained just by attachment theory. Moreover, individual variability in terms of affect tolerance and affect discharge may also play a role regardless of psychosocial factors.

The most notable way of measuring attachment is using Adult Attachment Interview. The premise in this psychological instrument is to assess adult’s internal working. During the interview parents are asked to describe and evaluate their childhood memories of attachment experiences and important relationships. The instrument is thought to be valid because studies had shown that attachment is highly stable over the first two years of life (Owen, et al, 1984).

The shortcoming of the assumption is that attachment is stable over the first two years of life and the fact that attachment stability has been standardized observing infants in a middle class environment. It is conceivable that environmental factors such as divorce in the family or the birth of other siblings may affect attachment behavior. In addition, this instrument needs to be tested in longitudinal studies of not only infants born in different social classes, but also in different cultures. The varying cultural practices of child rearing may affect attachment styles. Hence, the instrument may only be valid at best to a middle class family.


There maybe a sound evolutionary reason for attachment as an optimal survival mechanism. In order to have emotional and physical needs met, an individual would need to be able to reciprocate in kind. These are subtle negotiations that have been perfected over an animals’ evolutionary history since cooperative behavior increases the chance of survival and continuation of a species. It is in the context of attachment to another person first, and to other people later, that an individual learns moral proscriptions of the tribe. An individual also learns suitable methods of affect regulation such as shame and aggression. What is morally sanctioned behavior are often attributes that the tribe finds desirable because it helps to protect the integrity of the tribe as a group. In evolutionary terms, behaviors that may compromise the integrity of the tribe, i.e. maladaptive for survival, brings forth certain punishment to the rule- breaker. Adaptive behaviors are passed on over the generation as part of the cultural milieu.

Attachment theory makes strong argument for the important role relationship in infancy and childhood has in later life. This makes sense because during infancy and childhood neural connections are made much faster and the brain seems to have greater plasticity. However, there are situations where early neglect has not made a negative impact on a person judging from observable maladaptive behavior. Not all infants who lacked early attachment suffer from psychopathologies.

Individual variability also plays a role on how much infants seek or need attachment. There are also degrees to attachment. For instance, what is the minimum level of attachment needed to prevent the propensity for drug abuse? Attachment is definitely needed for proper growth and adaptation, but can attachment be quantified? Do all infant need the same level of attachment? It is not very likely that attachment, as a complex behavior would be explained entirely by oxytocin. Attachment behavior is probably a result of environmental, biological and sociological factors.


Alven, Gosta. (2004) Plasma Oxytocin in children with recurrent abdominal pain. Journal of Pediatric Gastroenterology & Nutrition. 38(5):513-517, May 2004.

Bales, K. L. & Carter, C. S. (2003). Developmental Exposure to Oxytocin Facilitates Partner Preferences In Male Prairie Voles. Behavioral Neuroscience, 4: 854-859.

Insel, T. R. (1997). A Neurobiological Basis of Social Attachment. American Journal of Psychiatry, 154: 726-735.

Moffitt, E. T., Brammer, G. L., Caspi, A., Fawcett, J. P., Raleigh, M., Yuwiler, A., & Silva, P. (1998). Society of Biological Psychiatry, 43: 446-457.

Owen, M. T., Easterbrooks, M.A., Chase, L., & Goldberg, W.A. (1984). The relation between maternal employment status and the stability of attachment to mother and father. Child Development, 55, 1894-1901.

Panksepp, J. (1998). Affective Neuroscience. Oxford University Press, NY.

Pine, D.S., J.D. Coplan, et al (1997). Neuroendocrine response to fenfluramine challenge in boys. Archives of General Psychiatry 54: 839-846

No comments:

Post a Comment