Taking a somewhat different approach to the treatment of pediatric asthma, a group of relatively recent studies revisited the contribution of psychogenic and family factors in the etiology of pediatric asthma—a focus pursued by researchers with some interest in the 70s, then apparently dropped in favor of an emphasis on asthma as a medical, rather than psychological/relational, problem (see below). The latest of a series of three articles reported that four studies (Feinberg, 1988; Schwartz, 1988; Pennington, 1992; Madrid, Ames, Skolek & Brown, 2000) were conducted in the period between 1988 and the article’s publication in 2000 (Madrid, Ames, Skolek, & Brown). For some reason, the earliest study, by Feinberg, is not cited in this article, but is cited (but not referenced) briefly in the second and described in the third, which summarizes the first three studies before presenting the fourth study.
These studies were apparently conducted to support extensive clinical experience, with the hope of scientifically validating a seemingly efficacious treatment method for an increasingly prevalent and costly illness. The purpose of the studies was to investigate the possibility that some cases of pediatric asthma are the result of a greater than normal number of bonding disruptions “near or at the time of birth”, and that such disturbances and the resulting asthma can be successfully treated using a targeted therapeutic approach. The studies further attempted to show that a considerable measure of success in alleviating asthma symptoms in the children can be obtained by treating only the mother, with additional improvements with treatment of children past the preverbal stage.
The first article of the three (Madrid & Schwartz, 1991) describes the second study in the series, in which 30 mothers of asthmatic children and 30 mothers of well children were interviewed using the Maternal Infant Bonding Survey (M.I.B.S.) in order to discover whether bonding disruptions were more prevalent in the histories of asthmatic children than those of well children. The researchers’ premise was that bonding disruptions such as the physical or emotional separation of mother and infant soon after birth affect the emotional attachment between mother and child and that the absence of this attachment, or “non-bonding”, rather than maternal rejection and/or over-protection, as postulated by previous researchers, is what underlies some cases of pediatric asthma.
The authors claim that they derived the definition of the term “non-bonding” from Klaus and Kennell’s research, referenced in their book Maternal-Infant Bonding (1976). While Klaus, Kennell and Klaus are well-known authorities and advocates of optimal birth in the facilitation of better mother-infant bonding, Madrid and Schwartz appear to have relied for their definitions solely on this book, while apparently disregarding seminal work on attachment by researchers such as Bowlby (1969, 1973, 1988), (Ainsworth (Ainsworth, Blehar, Waters, & Wall, 1978), and Stroufe (Stroufe, Fox, & Pancake, 1983). Such writers distinguish between “bonding”, as a descriptor of mothering behaviors towards the infant, and “attachment”, describing the biologically-driven process by which the infant forms a relationship with its mother, and further distinguish between different styles of attachment: Secure, Avoidant, Resistant/ambivalent (Ainsworth, Blehar, Waters, & Wall, 1978) and Disorganized/disoriented (Main and Solomon, 1990). Unfortunately, Klaus and Kennell did not, themselves, define their terms too clearly: despite the title of their book, they use the words “attachment” and “bonding” interchangeably throughout the book, without discriminating between their application to mother or child.
Madrid and Schwartz (1991) further make the assumption that “non-bonding” exists and, indeed, is commonplace in mother-child pairs in which the child has asthma (86% in this study!). Other writers in the field of attachment research, including Main (Main & Solomon, 1990) believe that it is rare for infants not to attach at all (children in certain orphanages might exemplify those who may not have the opportunity to form an attachment): it is certainly more likely that a mother may have difficulty bonding with her infant. While Madrid and Schwartz state clearly that their study “did not attempt to provide a comprehensive, operational definition of bonding” (p. 354) they write “Mothers who do not bond with their children . . “ in one paragraph, and three paragraphs later report that they used raters “to categorize the children as bonded or non-bonded . . .” (p. 54). Thus they state in the same article, indeed, on the same page, that bonding is something the mother does, something the child does, and something that takes place “between mother and child”! They do not clarify whether it is, in fact, probable that a child could be “non-bonded” as opposed to having a dysfunctional attachment to their mother. Since there has apparently been no study demonstrating that “non-bonding” exists in children predominantly raised by their mothers or another significant caregiver, the study stands on shaky ground in using the term as it does. Credibility would have been enhanced by using terms such as “disrupted bonding” or “dysfunctional attachment” to describe either the experience of mother or child, respectively, or the effects of certain events on the interpersonal process between mother and child. In addition to the above, the article uses additional terms that are not clearly defined (e.g. “soon after birth”).
A second problem with this study is that no explanation is given for the lack of citations later than the 70s and 80s: all but one of the later ones being those of colleagues from the same facility. In fact, there was apparently little or no relevant research regarding the etiology of pediatric asthma during the 80s and 90s. A recent study (Celano, 2001) stated:
The salience of psychogenic and family factors in asthma faded in the 1980s and 1990s as asthma was redefined as a medical (not a psychosomatic) problem (Mrazek & Klinnert, 1988). Although a small but consistent scientific literature continued to explore the role of psychological or family interaction influences in asthmatic episodes or disease course, the prevailing perspective was that family processes affect morbidity only via asthma management behaviors.
A similar sentence included in the article would have gone far to justify the lack of more recent citations in support of Madrid’s hypothesis.
The selection of subjects for the study is inadequately described so the reader is left unclear as to the recruitment method, selection procedures, or whether the mothers and children in the experimental and control groups were matched for age, race, severity of illness, socioeconomic status, etc. There is also no indication of how children were diagnosed and whether a physician’s letter or report was utilized in this regard. Since one-third of the experimental group were already clients of the Erickson Institute and others were referred by study participants, while control group participants were also referred through personal contacts, there is no pretense that the subjects are unbiased, and this bias is s a major weakness of the study. A study in which recruitment of new subjects was framed around a request that permitted participants to be unbiased as to the study’s hypothesis would have produced far more credible results.
A fourth problem with this study lies in the use of an unvalidated instrument, the M.I.B.S. This was apparently developed and administered by staff of the Erickson Institute, which includes the authors of this study, and thereby reflects their bias. A 19-item checklist, it seems to be neither comprehensive nor objective enough to be administered and scored by untrained personnel, introducing considerable researcher bias into the interview process and the rating of the results, despite the additional use of quantitative analysis. In fact, it is stated that the two raters for the study disagreed on a significant number of cases (12 out of 30) and discussions towards an agreed-upon rating did not involve an unbiased third party, casting further doubt on the accuracy and validity of the findings.
The questions on the M.I.B.S., themselves, while apparently intended to identify factors that might disrupt bonding or attachment, exclude a number of possible events that might contribute to this, or be in themselves causes of pediatric asthma. For example, there are no questions concerning whether the mother or anyone in the household smoked while she was pregnant, there is no question regarding whether the infant was resuscitated, suctioned or intubated at birth, or what the Apgar scores were (a subjective, but universally used measure of a newborn’s condition at one and five minutes after birth, based on five observable factors). There is no question regarding other birth interventions (chemical induction/augmentation, forceps or vacuum extraction, hypoxia, narcotic use, diagnosis at birth of respiratory distress, etc.) that would be traumatic for the infant and therefore could affect the early ability to attach, or whether there were any significant heart-rate decelerations during the delivery. There is no question regarding whether a cesarean section was scheduled or emergent, or whether the child’s umbilical cord was clamped before he/she had taken the first breath. Remarkably, there is also no question about whether or when the mother first breastfed her child, although breastfeeding success is correlated throughout the literature with enhanced bonding. In addition, since childbirth is such an intense and in some cases overwhelming experience, many women might be unable to identify certain complications which affected their baby and could influence its subsequent attachment. There are also only two questions concerning events in the prenatal period: inadequate to identify some competing emotional factors on the part of the mother or infant: for example, there is no question regarding whether the child was planned or wanted or how supported the mother was during the pregnancy. While some women will be very forthcoming in response to a request such as “Please describe any conditions you experienced”, others might require more specific prompting before they are able to identify relevant factors. A more comprehensive form of the survey would undoubtedly reveal other factors that may contribute to or cause pediatric asthma, therefore their omission, and the use of biased interviewers and raters, cast serious doubt on the veracity of the postulated causes of the illness. Thus the study described in this article contains some serious methodological flaws, while offering a tantalizing glimpse of a plausible explanation and a potentially valuable treatment for the problem of pediatric asthma.
The second article (Madrid & Pennington, 2000) appears from its structure and headings to be more a brief review of the literature than the description of a research study. Its purpose seems to be to repair the deficit in the previous article of failing to build an adequate case from the literature for its hypothesis before describing the research conducted by its authors to support the introduction of the treatment protocol developed in conjunction with their studies. A single case example from the Erickson Institute’s clinical experience is described, which apparently gave rise to the notion that treating the mother for bonding disruptions resulted in reduction or cessation of symptoms in the child. Pennington’s own dissertation study is awarded a mere paragraph, while the article’s purpose seems to be to provide support for the hypothesis of a relationship between disruptions or failures in bonding and pediatric asthma, and to call for the examination of a targeted treatment protocol to repair disrupted bonding. The article continues to rely on Klaus and his associates to define bonding—this time as a “biological, psychological, and emotional . . . intricate dance” (p. 280) between mother and child during a period of impressionability shortly after birth. However, since Klaus and associates are explicitly describing the mother’s experience of the events and the relationship with her infant, little attention is paid to the infant’s part in this dance either in their writing, or in this article and the associated research. Such a focus would justify the bias towards the use of the term “bonding” rather than “attachment”, however this article continues the mistake of the previous one in continuing to allude to “non-bonded babies” (p. 284). In this case, they are described as “often colicky and may dislike being cuddled” as well as “difficult to please,” “fidgety,” and “not really comfortable with people” (pp. 284-285) when older. No criteria are described for identifying these behaviors, nor are other causes of such behaviors ruled out. While these behaviors may be associated with non-bonding, there are many perinatal factors other than bonding disruptions that can contribute to their formation.
In describing previous research the authors state that “it has become widely accepted that separation of mother and child is good for neither” but cite no studies as examples. This optimistic perspective is far from today’s reality and some degree of mother-infant separation is still routine practice in many hospitals throughout the country. It would take very little effort, unfortunately, to find hospitals at which studies similar to those they describe as having been conducted 20 years ago could still be done. Like the previous article, this one does not address whether adequate bonding can take place later.
This article also represents a missed opportunity to provide a more thorough description of Feinberg’s research, for which neither the number of participants in the study nor the degree of significance of the results was provided. It might also have offered the details of Pennington’s study, which, though apparently interesting in its findings of four specific “non-bonding” events found to occur significantly more often in the histories of asthmatic children, and correlating with Klaus and Kennell’s findings, is given short shrift.
What is surprising is that, with their reliance upon Klaus and Kennell’s work for their fundamental premise and the basic elements of their survey, Madrid and Pennington did not appear to have reviewed the updated version of their book (Klaus, Kennell & Klaus, 1995), in which the information provided on bonding is much expanded as a result of, the authors say, 13 years of NICDH grant funding.
This article does assist the reader in understanding the findings of the M.I.B.S. by reporting that in Schwartz’s study (1988) “twenty percent of well children had two or more non-bonding events in their mother’s histories. For asthmatic children, the corresponding figure was 70%.” (p. 287). Unfortunately, since the raters were biased, the identification of these non-bonding events appears to be purely subjective.
In this article, two new terms are introduced: “intrinsic asthma” and “responding asthma”, in conjunction with the summary of Schwartz’s results. Unfortunately, neither term is defined, nor sources cited, leaving the reader wondering, since there was found to be no difference in incidence of non-bonding events between the two, why the terms were introduced in this brief summary in the first place!
It was valuable to find here (p. 287) a sentence stating clearly that “all non-bonded children do not have asthma and that all asthmatic children are not non-bonded.” Belatedly, this sets a context for the body of work by these authors. However, in the conclusion of the article, an additional study (Mrazek, Klinnert, Mrazek & Macey, 1991) is cited (pp. 287-288), apparently to support the argument for the relationship between “failures in bonding” and pediatric asthma, which introduces the term “genetically predisposed asthmatic children” as those studied with their mothers by these researchers. However, whether or not asthma can be transmitted genetically is never addressed here or elsewhere in the article, leaving an additional unanswered question on the table, and detracting from the case the authors attempt to build for further study of the phenomenon of interest.
It is stated at the beginning of the third article (Madrid, Ames, Skolek & Brown, 2000), that its purpose is to formalize the examination of the research hypothesis and the presentation of findings, thus attempting to remedy the shortcomings of the previous two articles and the studies described therein. It follows the standard format for journal publication of the findings of research in which the authors conducted a small pilot study to test the effectiveness of a treatment approach in repairing bonding disruptions between mothers and their asthmatic children. Since many of the citations in this article are of the authors’ colleagues, and these colleagues’ citations are not explicitly referenced, it is still unclear how many non-related writers have addressed the issue of strained mother-child relationships. There is also still no explanation provided for the lack of citations (other than those of colleagues) from the 80s and 90s.
The six subjects for the study were those who completed the study from an initial pool of 19 who volunteered to participate in response to announcements in schools and health centers in a named specific geographical area. Mothers who volunteered were sent a research package that included the M.I.B.S., a Mother’s Report: a nine-item questionnaire adapted from a similar document already in use and requesting objective data on the child’s asthmatic condition, and a Child’s Report, another nine-item questionnaire designed to permit the child to evaluate his or her own breathing under specific conditions. Mother’s report and Child’s report were both administered three times: before the study began, two weeks after completion of the mother’s treatment, and one month after the child’s treatment was completed. In two cases the children in question were too young to respond to this questionnaire. A Clinical Scoring Check List, developed by the researchers who provided the original of the Mother’s Report, was used with the Mother’s Report “for some presentations of the data in order to determine the changes in the severity of asthma.” (p. 95) Treatment was conducted first on the mother, and then on the child if he/she was old enough to participate. It would appear that the methodology for this study was designed to provide somewhat more objective measures that could also be cross-checked by comparing mother and child’s reports.
Subject selection still appeared problematic, in that “two mothers did not wish to participate because they did not think the hypothesis for this study was valid,” while a third “was not comfortable with the theoretical assumptions.” (p. 93). Properly framed, the invitation to participate should ideally have attracted participants regardless of whether they agreed with the hypothesis or not, although issues concerning birth, bonding, and parenting can be extremely sensitive for some parents, who might have resistance to certain concepts no matter how neutral their presentation. A larger number of subjects would have made for a study with greater reliability, but possibly would also have exceeded the allotted research budget.
Despite the addition of some more objective measures, the continued use of the M.I.B.S., an unvalidated and rudimentary instrument, requiring the expertise of the Senior Investigator, Madrid, to rate participants bonded or non-bonded, renders the substance of the study highly subjective due to overwhelming investigator bias, while permitting no adequate way of ruling out other possible causes of pediatric asthma in the child subjects.
The study is handicapped by the absence of a control group. In this case, such a group should have responded to the questionnaires but received some form of education session(s), rather than the full treatment process, to rule out any improvements as a result of simply receiving researcher attention. On the other hand, the detailed description of the treatment process, clearly calling for considerable expertise in the techniques used, and clearly tailored to the history of each individual mother-child dyad, conveys the potential potency of the procedure employed, while highlighting the fact that therapy itself is not something easily standardized or quantified and may therefore be more amenable to qualitative, rather than quantitative, research. A purely quantitative approach might have called for a standardized protocol, or script, for the treatment itself, and while this method might have accomplished some of the treatment goals, it is likely that results would be less convincing with the absence of the art of therapy itself, and the experience and convictions of the therapist conducting the sessions.
While the greater improvement was shown consistently following treatment of the mother, further improvement was shown in several cases after treating the child. It is possible that the child may make some significant contribution to the bonding disruption, in concert with the development of the asthma, yet virtually no attention is given to this possibility, just as the study makes no attempt to relate its terminology to that of the body of research concerning disordered infant attachment. The authors quote the statements of mothers who feel unbonded to their infants, yet repeatedly ascribe “bonded” or “nonbonded” to the infants themselves. In addition, the study does not explicitly address the distinction and relationship between maternal bonding and infant attachment, nor define the limitations of the study to the mother’s part in bonding disruptions.
While mention is made of a delayed improvement in one child, there was apparently no specific longer-term follow-up incorporated into the study, which might have had considerable value in indicating whether the effects of treatment in alleviating asthma symptoms were enduring, while being reasonable simple to implement using the same measures as were used during the study.
Statistically, the number of subjects in this study is too small for the results to have great significance, although the use of the Chi-square analysis helps to indicate whether the study’s findings have statistical validity and it is therefore worth replicating. The article’s concluding discussion asserts, justifiably, that the study indicated that the researchers’ Maternal-Infant Bonding therapy alleviates children’s asthma symptoms by improving mother-child bonding. An additional hypothesis emerged from the study: that “impaired bonding may be easier to remedy at a younger age” (p. 110), particularly with regard to the developmental tasks of adolescence, which would be in direct conflict with efforts to strengthen the mother’s connection to her child. The limitations of the study are addressed explicitly here, and a larger study, with “stronger objective measures” is called for (p. 111). The authors also emphasize the importance, however, of the subjective experiences of the participants, stressing the value of a quick and effective solution to the suffering pediatric asthma imposes on the families who experience it.
While there were major deficiencies in the three articles’ reviews of the literature, some serious methodological flaws in the earlier studies and significant limitations in the later one, and poorly defined and inaccurate terms on which the studies’ hypotheses were founded, the cumulative outcomes of these researchers’ attempts nevertheless present an intriguing approach to the resolution of some cases of pediatric asthma, and represent the all too rare efforts of clinicians to document the potentially valuable outcomes of their clinical explorations and experience.
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the Strange Situation. Hillsdale, N.J.: Erlbaum.
Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York: Basic Books.
Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation and anger. New York: Basic Books.
Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London: Routledge.
Celano, M. P. (2001). Family systems treatment for pediatric asthma: Back to the future. Families, Systems & Health: The Journal of Collaborative Family HealthCare, l (19) (Issue 3), 285-290.
Klaus, M., & Kennell, J. (1976). Maternal-infant bonding. St. Louis: Mosby.
Klaus, M., Kennell, J., & Klaus, P. (1995). Bonding: Building the foundations of secure attachment and independence. Reading, MS: Addison-Wesley.
Madrid, A. & Schwartz, M. (1991). Maternal-infant bonding and pediatric asthma: An initial investigation. Pre- and Perinatal Psychology, 5 (4), 347-358.
Madrid, A. & Pennington, D. (2000). Maternal-infant bonding and asthma. Journal of Prenatal and Perinatal Psychology and Health, 14 (3-4), 279-289.
Madrid, A., Ames, R., Skolek, S. & Brown, G. (2000). Does Maternal-infant bonding therapy improve breathing in asthmatic children? Journal of Prenatal and Perinatal Psychology and Health, 15(2), 90-117.
Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorganized/disoriented during the Ainsworth Strange Situation. In M. T. Greenberg, D, Cicchetti, & E. M. Cummings (Eds.), Attachment in the preschool years: Theory, research, and intervention, 121-160. Chicago: University of Chicago Press.
Mrazek, D., Klinnert, M., Mrazek, P., & Macey, T. (1991). Early asthma onset: Consideration of parenting issues. Journal of the American Academy of Child and Adolescent Psychiatry, 30, 2.
Stroufe, L. A., Fox, N. E., & Pancake, V. R. (1983). Attachment and dependency in the developmental perspective. Child Development 54, 1615-1627.