Infant-Mother Attachment Can Be Improved through Group Intervention: A Preliminary Evaluation in Spain in a Non-Randomized Controlled Trial

The quality of infant-mother attachment has been linked to competence in different domains of child development. Research indicates that early intervention can enhance the quality of infant-mother attachment, though its efficacy in a group format has yet to be evaluated. The current study is aimed at examining the usefulness of a group intervention in enhancing infant-mother attachment. An intervention aimed at addressing aspects such as maternal responsivity, sensitivity and childrearing behaviour was developed by the researchers and experienced psychologists. The intervention spanned a period of 14 months starting from the third quarter of pregnancy. The intervention was evaluated among 24 mothers from the Basque region of Spain. The sample consisted of children of both genders in a similar proportion: 45.8% were boys and 54.2% were girls. The children in this sample were full-term born and did not present symptoms of any serious pre- or postnatal complications. The intervention had a statistically non-significant medium effect. Infants whose mothers had received the intervention showed higher rates of secure attachment compared to children from the control group, as assessed by the Strange Situation observation procedure. A potentially significant confounding variable, maternal attachment, was balanced across the intervention and comparison groups. We can tentatively point out that a group intervention may enhance the quality of infant-mother attachment. Nevertheless, because the study design was not randomized, the results of this study remain preliminary and need replication in a full randomized controlled trial designed study. Keywords: infant-mother attachment, maternal sensitivity, group intervention, infants, Spain. La calidad del apego bebé-madre ha sido asociada a la competencia en diferentes dominios del desarrollo infantil. La investigación muestra que la intervención temprana puede mejorar la calidad del apego infante-madre, aunque su eficacia en formato grupal no ha sido todavía suficientemente evaluada. El presente estudio se dirige a examinar la utilidad de una intervención grupal en la mejora del apego del bebé hacia su madre. Esta intervención ha sido desarrollada por las investigadoras y otras psicólogas expertas, y está dirigida a abordar aspectos tales como la responsividad, la sensibilidad y la conducta de crianza de las madres. La intervención tuvo lugar durante un periodo de 14 meses, comenzando desde el tercer cuatrimestre del embarazo. La intervención fue evaluada en 24 madres del País Vasco (España). La muestra estaba formada por bebés de ambos géneros en proporción similar: 45,8% eran niños y 54,2% eran niñas. Los niños de esta muestra eran nacidos a término y no presentaban ningún síntoma de complicaciones pre- o postnatales serias. La intervención tuvo un efecto medio estadísticamente no significativo. Los bebés cuyas madres habían recibido la intervención mostraron tasas más altas de apego seguro en la Situación Extraña que los bebés de las madres del grupo control. El modelo mental de apego de las madres, como potencial variable extraña, estaba equilibrado en los grupos de intervención y comparación. Podemos señalar de forma tentativa que una intervención grupal puede fomentar el apego seguro del bebé hacia su madre. Sin embargo, debido al carácter no aleatorio del diseño, los resultados de este estudio son preliminares y necesitan una replicación a través de un estudio con un diseño controlado aleatorizado. Palabras clave: apego bebé-madre, sensibilidad materna, intervención grupal, infantes, España.






Publication: The Spanish Journal of Psychology
Author: Torres, Bárbara
Date published: July 1, 2011

The quality of infant-mother attachment has been linked to competence in different domains of child development. Research indicates that children who are securely attached to their mothers have better social-emotional functioning, cognitive development and school outcomes later in life (Barlow & Svanberg, 2009; Sroufe, 2000). Saliency of attachment patterns can be inferred from two interrelated sets of studies. Recent reviews of studies on attachment and psychopathology (DeKlyen & Greenberg, 2008; Dozier, Stovall-McClough, & Albus, 2008; Kobak, Cassidy, Lyons- Ruth, & Ziv, 2006) show that insecure, and disorganized, attachment is associated with greater risk for psychopathology throughout the life span. Insecure attachment is not a form of psychopathology as such but may establish a trajectory that, along with other risk factors, can increase the risk of difficulties later in life (Rutter, 1985). On the other hand, there is also evidence that attachment security can work as a protective factor since, in the previously mentioned reviews, it has been seen to ameliorate exposure to high risk situations such as low socio-economic status, teen parenthood and parental depression.

Given the increased awareness of the value of quality infant-mother attachment, interventions explicitly aimed at enhancing early child-parent attachments have proliferated during the last 20 years. According to Berlin (2007), these interventions vary widely in terms of the participants they serve, the services they provide, and the rigor with which they have been evaluated. These intervention programmes have been applied to "normal" as well as to various atrisk mother and/or children populations, such as depressed mothers, imprisoned mothers, drug abusing mothers, preterm infants, adopted and foster cared children, maltreated children, etc. (Bakermans-Kranenburg, van IJzendoorn, & Juffer, 2003). There is currently sufficient evidence indicating that early intervention enhances the quality of infant-mother attachment (Bakermans-Kranenburg et al., 2003; Berlin, 2007; Berlin, Zeanah, & Lieberman, 2008; Egeland, Weinfield, Bosquet, & Cheng, 2000). In particular, most recent reviews (Berlin, 2007; Berlin et al., 2008) have concentrated on intervention programmes aimed at early attachment security which were published and evaluated through randomized trials. Nevertheless, as important as spending efforts in testing an intervention under strict experimental conditions is the feasibility aspect. This would further improve the value of an intervention, by making it ecologically valid within existing (contextual) conditions. Despite its applied relevance, this question has not been sufficiently addressed in attachment-based interventions.

According to Berlin (2007), Berlin et al. (2008), prevention and intervention programmes for supporting early attachment security focus on three therapeutic tasks. The first two tasks concern intervention targets: (1) the parent's internal working models, and (2) parenting behaviours (specifically parent's sensitivity and responsivity to infant's needs). The third task concerns the intervention process; it is argued that the parent's relationship with the intervener serves as the engine of therapeutic change. By highlighting the importance of parents' internal working models, the proposed tasks go beyond mere intervention focused on parenting behaviour.

Group-Interventions for the Promotion of Secure Attachment

The degree in which the programmes deal with these therapeutic tasks is diverse, and interventions vary in format and services delivered. Nevertheless, the majority of intervention programmes are based on individualized work with parent(s) or child-parent dyads, mostly through home visitation services. This is the case of some programmes which have received empirical support through randomized trials -e.g., Child-Parent Psychotherapy (Cicchetti, Rogosch, & Toth, 2006); Minding the Baby (Slade et al., 2007, cited in Berlin et al., 2008); Attachment Biobehavioral Catch-Up (Dozier, Libdhiem, & Ackerman, 2007); The Nurse-Family Partnership (Olds, 2007); and the different versions of the Video-Feedback Intervention to Promote Positive Parenting (Juffer, Bakermans-Kranenburg, & van IJzendoorn, 2008). Typically these individualized programmes aim at identifying 'gaps' in the mother-infant interactions and focus on improving interactions on a oneto- one basis, usually supporting this kind of work through video feedback technique. Programmes also differ in terms of the length and intensity of the intervention implemented, but it does not seem possible to have yet a conclusive answer about their relative efficacy (Greenberg, 2007). However, while the efficacy of individualized programmes has been established, the general usefulness of these types of programme is potentially compromised by the heavy financial and personnel demands of the latter (Cunningham, Bremner, & Boyle, 1995).

Compared to individual interventions, group-based ones have some drawbacks, such as the negative potential impact on the group processes of deviant individuals, lower flexibility in the arrangement of the sessions, and a likely greater difficulty for each parent in establishing a positive and trusting relationship with the facilitator. However, if the efficacy of group interventions can be established, the benefits of a group-based intervention are likely to outweigh the aforementioned limitations. Group interventions may potentially offer an economical and efficient approach to reaching a greater number of clients. Together with the lower economic cost of a group-based programme, these kinds of intervention may take advantage of potentially powerful mechanisms that could be missing in individual interventions. Group intervention offers an ideal scenario for the development of socializing techniques, social networks, imitative behaviour, interpersonal learning and group-specific classical therapeutic factors (Yalom, 1985). Since resources available in public institutions are usually limited, it is crucial to examine how effective group-based interventions are in improving attachment security.

A few programmes have tried to deal with the issue of high cost associated with home visitations by combining them with interventions delivered in group format, like the UCLA Family Development Project (Heinicke, Fineman, Ponce, & Guthrie, 2001), the Steps to Effective and Enjoyable Parenting (STEEP) programme (Egeland & Erickson, 2004) or the intervention designed by Brisch, Bechinger, Betzler, and Heinemann (2003). Nevertheless, efforts to promote secure infant-mother attachment which exclusively relies on group-based interventions are particularly rare. In fact, in a meta-analytic study of 70 attachment interventions (Barkermans-Kranenburg et al., 2003), only three studies involved group interventions, although none of these examined the impact on infantattachment security. One exception is the Circle of Security (COS) project (Cooper, Hoffman, Powell, & Marvin, 2007; Marvin, Cooper, Hoffman, & Powell, 2002), which aims at enhancing attachment security in infants and preschool children and delivers its intervention in groups. Highly trained and experienced master's level therapists deliver the intervention (20 weekly sessions) to 6-8 parents at a time, including extensive and careful review of videotaped child-parent interactions for each parent in the group (Berlin et al., 2008). Therefore, this intervention is tailored to each parent-child dyad; hence, it can be said that, while they have group meetings, the intervention itself is individualized which makes it relatively expensive and therefore reduces one of the main benefits of a group intervention. The Right From The Start programme (RFTS, Niccols, 2008) represents a recent effort to promote early infant attachment security through a parents' group. In contrast to COS, the RFTS intervention was evaluated using a randomized design. RFTS uses the Coping Modelling Problem Solving Approach (Cunningham et al., 1995) through an 8-session parent group to enhance caregiver skills in reading infant cues and responding sensitively. This is a learning approach in which individuals actively participate by: (a) identifying some common parenting errors shown through videotaped models, (b) discussing the consequences of these errors, (c) suggesting alternatives to them, (d) and identifying some advantages in the alternative approaches (Niccols, 2008). This intervention shows that a group-based intervention for parents can be as effective as home visiting in improving infant attachment security and maternal sensitivity (Niccols, 2008).

Development and Evaluations of interventions in Spain

Prevention programmes aimed at enhancing child outcomes are generally rare in Spain. The same would apply to specific interventions aimed at promoting children's wellbeing in this geographical context (Trenado, Pons- Salvador, & Cerezo, 2009). We only know of one programme, the Parent-Child Psychological Support Programme (PAPMI), which has attempted to enhance children's wellbeing (Cerezo, Pons-Salvador, & Trenado, 2008; Cerezo, Trenado, & Pons-Salvador, 2009).

PAPMI is a preventive action focused on babies from 0 to 18 months old, because these children are a group that is particularly vulnerable to parental practices and at high risk of child abuse (Cerezo, Dolz, Pons-Salvador, & Cantero, 1999). The goal of the PAPMI was to improve maternal competence by strengthening synchronous mother-child relationship, in order to prevent conflicts and maladjusted interactions which could precipitate child abuse episodes. This goal was achieved through monthly individualized instructions to the mother about her child's development (Cerezo, Cantero, & Alhambra, 1997; Cerezo et al., 1999). Preliminary results indicated that PAPMI had positive outcomes. Although this was not an attachment-focused programme, it apparently seemed to have indirect consequences over children's security attachment because the more visits the families received, the higher percentage of securely attached children was observed (Cerezo et al., 2008). While these results are promising, and given the importance of security of attachment in children, an intervention focused on attachment-related issues would be needed for the promotion of a healthy socio-emotional development, and using a group format that would imply less costs.

Cantero (2003) summarized the elements for an attachment-based intervention that would promote a secure emotional bond with the caregiver. She described five characteristics. The attachment intervention should: a) aim at increasing parental sensitivity to the infants cues and needs; b) focus on the child-parent relationship instead of isolating features of one or the other; c) take into account the child's role in relation to his/her parents and his/her own development; d) consider parents' mental models of relationship and their own attachment experiences as a possible influence in their involvement with the child; and e) have a professional who provides a security base for the dyad (i.e., is responsive to questions/doubts in times of need). To the best of our knowledge, no preventive intervention with those components has been implemented within our geographical context. There is an urgent need to carry out studies aimed at evaluating the efficacy of different intervention strategies in enhancing the quality of children's attachment.

Moreover, for the success of that intervention, we believe that two contextual factors should be taken into account in its design. First, the public nature of the Spanish Health System allows the whole population to receive service regardless of individuals' income, which makes it the perfect setting for delivering this kind of intervention. Second, the wide coverage of Spanish Health System does not allow for an individualized format, so one dyad-tailored interventions are generally effective (e.g., Dozier et al., 2007; Juffer et al., 2008; Olds, 2007; Svanberg, 2009) but not feasible and sustainable in the long run in our context.

The purpose of this study was to develop and implement a group-intervention aimed at promoting secure attachment of children in Spain. The efficacy of the intervention would be tested by means of the children's attachment outcomes and the feasibility of having a group of mothers adapted to the time and timetable constraints of the intervention over a period of 14 months.

Methods

Study site and participants

The study was carried out with mothers in the Basque region of Spain. The intervention and research team carried out a series of meetings with health workers (i.e. midwives and Health Centre Coordinators) to introduce the study and ask for their cooperation in identifying and recruiting women in their third quarter of pregnancy. Following these meetings, midwives recruited expectant mothers from both public and private health centres. Among the women who were approached, 24 pregnant women in their third quarter agreed to join the study. Having obtained informed consent, the women were allocated to either intervention or comparison group.

Measures

Socio-demographic Questionnaire. A 40-item questionnaire was used to collected data on the sociodemographic background of participants. The questionnaire captured data on various factors including: age, maternal education level, mothers' employment status, family income level, marital status, immigration status, and if the baby was the first child. Additionally, we also asked for data on the degree to which the current pregnancy was planned and/or its desirability.

Adult Attachment Interview (AAI). Given there is strong evidence that parents' state of mind with respect to attachment as assessed by the AAI is one of the strongest predictors of infant attachment status deriving from the Strange Situation (Hesse, 2008; van IJzendoorn, 1995), we evaluated mothers' state of mind in relation to attachment through the AAI. The AAI uses a semi-structured interview to assess individuals' internal working models of attachment (George, Kaplan, & Main, 1985; Hesse, 2008). Data from the AAI can be used to identify 4 models of attachment. However, given the small sample size in this study, individuals are classified according to two categories: secure and insecure. Scoring was done by the first author, who has received training and certification in the coding of AAI.1

Strange Situation (SS). It is a systematized observation procedure for the assessment of infant's attachment style through 8 episodes (Ainsworth, Blehar, Waters, & Wall, 1978).

Procedure

During the last quarter of pregnancy, mothers were interviewed using the AAI. A researcher blind to AAI classifications assigned the mothers to the control or intervention group based on 2 variables: a) possibility of attending the group sessions during the fixed timetable (which notably made the two groups unequal in terms of number of participants), and b) their pregnancy status: two women who were close to the expected delivery date (in the following 6 weeks) were placed in the control group. It must be noted that two mothers who would qualify to be in the intervention group due to their initial compatible timetable were finally allocated to the control group because they dropped off at the beginning (second session) of the intervention. Mothers in the control group received a written booklet with general advice about baby caring (appropriate touching, music, infants' sleeping patterns, playing with the baby, early language development, healthy emotional development, etc.) and were contacted by phone twice to evaluate their general health and adaptation to motherhood during the study period. Mothers in the intervention group attended 22 sessions (see Table 1 for details of each session) from the 7th month of pregnancy until the baby was 12 months old. Attendance rate was very high for most of them (50% of mothers attended almost all sessions: 90%), and high in general (all mothers considered, there was a rate of attendance at 80% of sessions, ranging from 61.9% to 100%). The intervention programme was designed by the first author, a psychotherapist with research experience in childhood attachment, along with a team of three other licensed psychologists with experience in group intervention with parents. The intervention was delivered by two psychologists from the same centre in the midwives' room of one of the public health centres of San Sebastian (Spain).

When the children were 16 months old, they were observed in the lab with the Strange Situation procedure. In this study, children were classified according to two categories: secure and insecure. One of the authors (MJC), who was blind to both mothers' AAI classifications and (intervention) group membership, established babies' attachment pattern. MJC has extensive experience in the coding of the SS observation procedure.

Results

Sample Descriptives

Chi-square and Mann-Whitney's U analyses were conducted to check comparability of intervention (n = 8) and control (n = 16) groups. An evaluation of the background variables indicated that the groups only differed in their educational levels and age, but were similar in the other characteristics. Thus, there were no significant differences between the two groups in income level, χ2(1, N = 24) = 0.001, p = 1.00; work status, χ2(1, N = 24) = 0.51, p = 0.48; marital status, χ2(1, N = 24) = 1.09, p = 0.30; immigration status, χ2(1, N = 24) = 0.00, p = 1.00; birth order, χ2(1, N = 24) = 0.38, p = 0.54; and desirability of pregnancy, χ2(1, N = 24)= 0.13, p = 0.72. Moreover, as Table 2 shows, there were no significant differences between mothers' attachment models when AAI classifications were dichotomized into secure versus insecure mind states in relation to attachment , χ2(1, N = 24) = 0.00, p = 1.00.

However, it was observed that there were statistically significant differences in education levels -education level: χ2(1, N = 24) = 9.000; p = .003, with mothers in the control group having higher educational level- and age -Mann- Whitney's U = 43.000; mean rank (intervention) = 13.81; mean rank (control) = 9.88; eta square = .07], mothers in the control group being somewhat older. Although the education level statistically differs across the two groups, the higher rate of college-educated women in the control group than in the intervention group (93.8% vs. 37.5%) would not possibly account for a higher rate of securelyattached children. On the contrary, and following studies that suggest that lower SES is associated with insecure attachment (e.g., van IJzendoorn & Bakermans-Kranenburg, 2010), this situation (as an indicator of SES) would diminish the likelihood of having a higher rate of secure children in the intervention group. Nevertheless, another SES indicator (income level) and mothers' attachment security, both being comparable, would still make the two groups comparable.

Regarding children's characteristics, in the whole sample 45.8% were boys and 54.2% were girls (4 boys and 4 girls in the intervention group). All of them were full-term born and did not evidence any health problem after the birth, nor any noteworthy health problem during the whole period of data collection.

Effects of the intervention on attachment

To check possible intervention effects in children's attachment patterns, control and intervention groups were compared using a chi-square test which turned out to be statistically non-significant: χ2(1, N = 24) = 1.371, p = .242.

However, although the chi-square test did not reach statistical significance, there was the indication of an effect of the intervention on the improvement of security of attachment in children that may be interpreted as being considerable or medium (Rea & Parker, 1992). In samples of small size like ours, the criterion of statistical significance is not the best approach to understand differences because of its dependence on the sample size. In fact, most pilot studies are conducted to determine the power of a subsequent study that would involve randomized trial intervention based on the effect size (Thabane et al., 2010). Therefore, we relied on the relative difference of the two groups as expressed in terms of effect size (phi coefficient). Thus, the phi coefficient was used as an index of the strength of the relationship between group (intervention vs. control) and attachment outcome (insecure vs. secure). We found a statistically non-significant medium effect (phi = 0.24) of the intervention on attachment security, with higher rates of secure attachment in children in the intervention group. The comparison of children's attachment patterns in intervention and control groups showed that 75% (n = 6) of babies in the intervention group were securely attached to their mothers, whereas this was true for 50% (n = 8) of the children in the control group. Although, the small sample size does not allow a reliable statistical analysis of the effect of attendance on intervention effectiveness, a rough inspection shows that the only two insecure children come from mothers who did evidence an attendance at the sessions lower than 80%. This lower attendance in mother may have been related to more problems, which, in turn, would relate to insecure attachment (Cyr, Euser, Bakermans-Kranenburg, & van IJzendoorn, 2010).

Discussion

In this study we aimed at developing a preventive intervention based on attachment postulates, and at testing its effectiveness and feasibility in a preliminary study conducted with mothers in Spain. We can tentatively answer affirmatively to the question of whether group intervention can enhance security of the infant attachment. A statistically non-significant medium-size effect of the intervention supports this statement. This is comparable to the medium effect size found in the meta-analysis of attachment interventions where the Cohen d was 0.39 for intervention focuses exclusively on enhancing parental sensitivity, and even higher than the small effect size (Cohen d = 0.19) found for all kinds of intervention carried out in order to promote attachment security in children (Bakermans- Kranenburg, et al., 2003).

There has been a general paucity of data indicating the efficacy of group interventions in the field of attachment. As we mentioned before, in this meta-analysis of 70 attachment interventions, only three studies involved group interventions, but none of these examined the impact on infant-attachment security. This paucity in the use and evaluation of group intervention may have been in part motivated by doubts as to their efficacy. The current result provides impetus for further research in the area. We believe that group interventions offer great potential in promoting securely attached infants, partly due to the remarkable advantages of group interventions above individual ones, such as lower economic cost, less stigmatization of parents, promotion of social support, and parental empowerment through helping others (Cunningham, Davis, Bremner, Dunn, & Rzasa, 1993; Niccols, 2008; Yalom, 1985).

The design of our study differed slightly from the few other group-based interventions we know of. However, we believe that the changes made to our approach significantly increased the potential value of our intervention approach without compromising on quality and outcome. Like in the COS intervention (Cooper et al., 2007; Marvin et al., 2002), we worked with a small group of mothers and a relatively extended intervention. Nevertheless, our intervention may be more feasible for large-scale implementation because it is not so highly tailored to each infant-mother dyad as the COS. On the other hand, both the COS intervention and the one labelled RFTS (Niccols, 2008) use videos-videofeedback in the former case, and videotaped models in the latter- as one of the principal tools for working with parents. However, video technology was not included in our group intervention because that would have significantly increased the cost and also would have made this intervention more difficult to implement in our context, an important goal in our study.

Another important difference between our intervention and other interventions that have been reported earlier is that we started the intervention in the last stage of pregnancy (6 pre-natal sessions), as also happens in the UCLA Family Development Project and the Minding the Baby programme. We did so for two reasons. First, given the crucial role that group cohesiveness plays in group based interventions (Yalom, 1985), we wanted to promote group cohesion prior to childbirth. Although other authors (e.g., Bakermans- Kranenburg, et al., 2003) have proposed a later beginning in interventions (when the baby is 6 months old ), we believe that in group format interventions it is particularly relevant to pay attention to group processes, such as cohesion and peer support, that can enhance the effectiveness of the intervention as early as possible.

Equally important, our intervention aimed at improving not only infant-mother attachment, but also other aspects of child development. Since we wanted to enhance the efficacy of our intervention, we started at the prenatal stage. This would ensure that mothers acquire the right childrearing attitude as soon as the child is born. There is sufficient evidence to indicate that general parenting interventions implemented during the prenatal period have an incremental value compared to those that are implemented several months after childbirth (e.g., Parr & Joyce, 2009; Svanberg, 2009). On the other hand, this earlier start results in our intervention having a relatively high number of sessions in comparison with others, such as the Attachment and Biobehavioral Catch- Up programme (ABC), the Video-Feedback Intervention to promote Positive Parenting (VIPP) or the RFTS group intervention. As we have noted earlier, there is not enough evidence to conclude which kind of intervention is better for which kind of population in order to enhance early infant attachment security. In fact, some reviews (i.e., Egeland et al., 2000, p. 70) suggest that "lengthy, intensive, and carefully timed" services are needed in order to address internal working models and parenting behaviors. Others (Bakermans- Kranenburg et al., 2003, p. 208) concluded that interventions of less broad scope, but which only focus on sensitive maternal behaviour, seem to be quite successful when it comes to improving insensitive parenting and to developing infant attachment security. Nevertheless, halfway between these opposite views, Berlin (2007) and Greenberg (2007) proposed that "less could be more" for some families, whereas "more could be better" for others. Interestingly, as Berlin et al. (2008) noted, it seems that short-term attachment programmes, both individualized-based -e.g., such as ABC and VIPP- and group-based -such as Circle of Security (COS, Cooper et al., 2007) and RFTS-, includes the use of video technology. Finally, we agree with Berlin and associates (2008) in stating that participant relationship with the intervener has a crucial relevance in promoting therapeutic change. Therefore, in our case, we think that both the earlier start and the relatively prolonged duration of the intervention helped to foster group cohesion and a positive work alliance between participants and group facilitators.

Of special interest here is also its suitability for the Spanish context, where all individuals may receive assistance from the public health care system but the resources are limited so as to allow individualized prevention programmes. For that reason, group interventions settled in the regular timetables of the health care providers, such as the one tested in this study, are even more effective. First, they serve their purpose of increasing the likelihood of enhancing the quality of attachment in child-mother dyads. Second, they do so not only in the context of this study, but can also be generalized and may be easily reproduced in everyday settings of the Spanish health care system. Finally, a contextual remarkable aspect of this study is that the validity of the measurement of attachment is ensured. To the best of our knowledge, this is the first study conducted in Spain in which the two "gold standard" assessments of adult and child attachment (Berlin, 2007) are used: the Adult Attachment Interview (AAI, George et al., 1985) as a measure of the mother's mental state in regard to attachment, and the Strange Situation procedure (Ainsworth et al., 1978), to evaluate infant-mother attachment.

Despite these obvious advantages, our study has some limitations. First, the lack of randomization in the study design clearly limits the degree to which we can generalize about this result. Logistical grounds (i.e., schedule restrictions imposed by the professionals who delivered the intervention) were the main reasons for not having carried out a randomized intervention. While this is an obvious limitation from a research perspective, it is an added marker of feasibility from an applied point of view. We think that this feasibility perspective should be prominent, given that whichever intervention is proposed can only be useful if it can be efficiently fitted into the current health system. Furthermore, and in line with Thabane and associates (2010), we believe that feasibility per se is an important goal of pilot studies, which pave the way for future randomized studies without compromising unjustified amounts of money in preliminary stages.

Second, the study results are based on a small sample size, which limits our power to pick small differences and also to generalize in other populations. However, given the promising results from this study, both in terms of attachment outcomes and of adherence to the programme, apart from using randomization, future efforts should involve larger samples to obtain a more definite answer on the issues discussed in this paper. If this intervention is found feasible for large-scale implementation, the results would have a significant impact in ensuring accessibility of early preventive intervention for a majority of mothers.

An aspect to further explore in the future as regards testing the design of the intervention, is to adapt it to other types of population. Ours involved a 'relatively low risk' population, despite the fact that infancy is considered as a particularly vulnerable period for child abuse (Cerezo et al., 1999). Several aspects of the study design may make it unsuitable for implementation in a 'high risk population'. A notable aspect of the design likely to be challenging for a high risk population is its set format, which requires fixed timetables during an extended period of time (in our case 22 sessions). It has been observed that high-risk mothers have a high drop-out rate when involved in long interventions (Berlin, 2007), and are less likely to conform to a very structured programme. Therefore, future efforts need to evaluate the feasibility of this intervention in a 'high risk population' and to explore how it could be shortened without losing relevant intervention targets.

In conclusion, our study has suggested that a successful, easy-to-implement, and inexpensive group intervention of a preventive nature could be useful in enhancing the quality of infant-mother attachment. Given that promoting security in attachment is a necessary condition for the optimal development of the child, future efforts should focus on refining and implementing such programmes on a larger scale and in more specific populations.

1 The first author was trained in the Adult Attachment Interview (AAI) by the certified trainer Dr. Sonia Gojman at the training Institute held in Barcelona (Spain) during July of 2005, and was certified as a reliable coder by M. Main and E. Hesse in June of 2007.

References

Ainsworth, M. D., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: Assessed in the Strange Situation and at home. Hillsdale, NJ: Erlbaum.

Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., & Juffer, F. (2003). Less is more: Meta-analyses of sensitivity and attachment interventions in early childhood. Psychological Bulletin, 129, 195-215. doi:10.1037/0033-2909.129.2.195

Barlow, J., & Svanberg, P. O. (2009). Keeping the baby in mind. In J. Barlow & P. O. Svanberg (Eds.), Keeping the baby in mind: Infant mental health in practice (pp. 1-14). New York, NY: Routledge/Taylor & Francis Group.

Berlin, L. J. (2007). Interventions to enhance early attachments. In L. J. Berlin, Y. Ziv, L. Amaya-Jackson, & M. T. Greenberg (Eds.), Enhancing early attachments. Theory, research, intervention, and policy (pp. 3-33). New York, NY: The Guildford Press.

Berlin, L. J., Zeanah, C. H., & Lieberman, A. F. (2008). Prevention and intervention programs for supporting early attachment security. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment. Theory, research, and clinical applications (2nd ed.), (pp. 745-761). New York, NY: The Guilford Press.

Brisch, K. H., Bechinger, D., Betzler, S., & Heinemann, H. (2003). Early preventive attachment-oriented psychotherapeutic intervention program with parents of a very low birthweight premature infants. Attachment and Human Development, 5, 120-135. doi:10.1080/1461673031000108504

Cantero, M. J. (2003). Intervención temprana en el desarrollo afectivo [Early intervention in the affective development]. In A. Gómez-Artiga, P. Viguer, & M. J. Cantero (Eds.), Intervención temprana: Desarrollo óptimo de 0 a 6 años [Early intervention: Optimal development from 0 to 6 years] (pp. 175-203). Madrid, Spain: Pirámide.

Cerezo, M. A., Cantero, M. J., & Alhambra, M. S. (1997). La mejora de la competencia parental en la prevención del maltrato [The improvement of parental competence in the prevention of maltreatment]. Infancia y Aprendizaje, 80, 99-110. doi:10.1174/021037097761396199

Cerezo, M. A., Dolz, L., Pons-Salvador, G., & Cantero, M. J. (1999). Prevención de maltrato de infantes: evaluación del impacto de un programa en el desarrollo de los niños [Prevention of infant maltreatment: Evaluation of the impact of a programme in child development]. Anales de Psicología, 15, 239-250.

Cerezo, M. A., Pons-Salvador, G., & Trenado, R. (2008, September). Child abuse potential and quality of mother-child attachment in the context of a community based prevention program. Paper presented at the XVIIth ISPCAN International Congress on Child Abuse and Neglect. Hong Kong, China.

Cerezo, M. A., Trenado, R., & Pons-Salvador, G. (2009). Prevención del maltrato infantil. El apoyo a los padres y sus relaciones con los hijos y el fomento de las buenas prácticas [Prevention of child maltreatment. Support for parents and their relations with children and the promotion of good practice]. In B. Gómez-Bengoechea (Ed.), Violencia intrafamiliar: hacia unas relaciones familiares sin violencia [Intra-family violence: Toward family relations without violence] (pp.61-90). Madrid, Spain: Instituto Universitario de la Familia de la Universidad Pontificia de Comillas.

Cicchetti, D., Rogosch, F. A., & Toth, S. L. (2006). Fostering secure attachment in infants in maltreating families through preventive interventions. Development and Psychopathology, 18, 623-649. doi:10.1017/S0954579406060329

Cooper, G., Hoffman, K., Powell, B., & Marvin, R. S. (2007). The Circle of Security intervention: Differential diagnosis and differential treatment. In L. J. Berlin, Y. Ziv, L. Amaya- Jackson, & M. T. Greenberg (Eds.), Enhancing early attachments. Theory, research, intervention, and policy (pp. 125-151). New York, NY: The Guilford Press.

Cunningham, C. E., Bremner, R., & Boyle, M. (1995). Large group community-based parenting programs for families of preschoolers at risk for disruptive behavior disorders: Utilization, cost effectiveness, and outcome. Journal of Child Psychology and Psychiatry, 35, 1141-1159. doi:10.1111/j.1469- 7610.1995.tb01362.x

Cunningham, C. E., Davis, J. R., Bremner, R., Dunn, K. W., & Rzasa, T. (1993). Coping modeling problem solving versus mastery modeling: Effects on adherence, in-session process, and skill acquisition in a residential parent training program. Journal of Consulting and Clinical Psychology, 61, 871-877. doi:10.1037/0022-006X.61.5.871

Cyr, C., Euser, E. M., Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2010). Attachment security and disorganization in maltreating and high-risk families: A series of meta-analyses. Development and Psychopathology, 22, 87- 108. doi:10.1017/S0954579409990289

DeKlyen, M., & Greenberg, M. T. (2008). Attachment and psychopathology in childhood. In J. Cassidy & P.R. Shaver (Eds.), Handbook of attachment. Theory, research, and clinical applications (2nd ed.) (pp. 637-665). New York, NY: The Guilford Press.

Dozier, M., Stovall-McClough, K. C., & Albus, K. E. (2008). Attachment and psychopathology in adulthood. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment. Theory, research, and clinical applications (2nd ed.) (pp. 718-744). New York, NY: The Guilford Press.

Dozier, M., Libdhiem, O., & Ackerman, J. P. (2007). Attachment and biobehavioral catch-up: An intervention targeting empirically identified needs of foster infants. In L. J. Berlin, Y. Ziv, L. Amaya-Jackson, & M. T. Greenberg (Eds.), Enhancing early attachments. Theory, research, intervention, and policy (pp. 178-194). New York, NY: The Guilford Press.

Egeland, B., & Erickson, M. F. (2004). Lessons from STEEP: linking theory, research, and practice for the well-being of infants and parents. In A. Sameroff, S. C. McDonough, & K. L. Rosenblum (Eds.), Treating parent-infant relationship problems: Strategies for intervention (pp. 213-242). New York, NY: The Guilford Press.

Egeland, B., Weinfield, N. S., Bosquet, M., & Cheng, V. K. (2000). Remembering, repeating, and working through: Lessons from attachment-based intervention. In J. Osofsky & H. E. Fitzgerald (Eds.), WAIMH handbook of infant mental health. (Vol. 4; pp. 35-89). New York, NY: Wiley.

George, C., Kaplan, N., & Main, M. (1985). The Adult Attachment Interview. Unpublished manuscript. Berkeley, CA: University of California .

Greenberg, M. T. (2007). Enhancing early attachments: Synthesis and recommendations, for research, practice, and policy. In L. J. Berlin, Y. Ziv, L. Amaya-Jackson, & M. T. Greenberg (Eds.), Enhancing early attachments. Theroy, research, intervention, and policy (pp.327-343). New York, NY: The Guilford Press.

Heinicke, C. M., Fineman, N., Ponce, V. A., & Guthrie, D. (2001). Relationship based intervention with at-risk mothers: Outcome in the second year of life. Infant Mental Health Journal, 22, 431-462. doi:10.1002/imhj.1010

Hesse, E. (2008). The Adult Attachment Interview. Protocol, method of analysis, and empirical studies. In J. Cassidy & P.R. Shaver (Eds.), Handbook of attachment. Theory, research, and clinical applications (2nd ed.) (pp. 552-598). New York, NY: The Guilford Press.

Juffer, F., Bakermans-Kranenburg, M. J., & van IJzendoorn, M., H. (Eds.), (2008). Promoting positive parenting: An attachmentbased intervention. New York, NY: Taylor & Francis Group/Lawrence Erlbaum Associates.

Kobak, R., Cassidy, J., Lyons-Ruth, K., & Ziv, Y. (2006). Attachment, stress, and psychopathology: A developmental pathways model. In D. Cicchetti & D. J. Cohen (Eds.), Developmental psychopathology. Volume 1, Theory and Method (2nd edition) (pp. 333-369). New York, NY: John Wiley and Sons.

Marvin, R. S., Cooper, G., Hoffman, K., & Powell, B. (2002). The Circle of Security project: Attachment-based intervention with caregiver-preschool child dyads. Attachment and Human Development, 1, 107-124. doi:10.1080/14616730252982491

Niccols, A. (2008). "Right from the Start": Randomized trial comparing an attachment group intervention to supportive home visiting. Journal of Child Psychology and Psychiatry, 49, 754-764. doi:10.1111/j.1469-7610.2008.01888.x

Olds, D. L. (2007). The nurse-family partnership: Foundations in attachment theory and epidemiology. In L. J. Berlin, Y. Ziv, L. Amaya-Jackson, & M. T. Greenberg (Eds.), Enhancing early attachment. Theory, research, intervention, and policy (pp. 217-249). New York, NY: The Guilford Press.

Parr, M., & Joyce, C. (2009). "First steps in parenting": Developing nurturing parenting skills in mothers and fathers in pregnancy and the postnatal period. In J. Barlow & P. O. Svanberg (Eds.), Keeping the baby in mind: Infant mental health in practice (pp. 52-62). New York, NY: Routledge/Taylor & Francis Group.

Rea, L. M., & Parker, R. A. (1992). Designing and conducting survey research. San Francisco, CA: Jossey-Bass.

Rutter, M. (1985). Resilience in the face of adversity: Protective factors and resistance to psychiatric disorder. British Journal of Psychiatry, 147, 598-611. doi:10.1192/bjp.147.6.598

Sroufe, L. A. (2000). Early relationships and the development of children. Infant Mental Health Journal, 21, 67-74. doi:10.1002/(SICI)1097-0355(20001/04)21

Svanberg, P. O. (2009). Promoting a secure attachment through early screening and interventions: A partnership approach. In J. Barlow & P. O. Svanberg (Eds.). Keeping the baby in mind: Infant mental health in practice (pp. 100-114). New York, NY: Routledge/Taylor & Francis Group.

Thabane, L., Ma, J., Chu, R., Cheng, J., Ismaila, A., Rios, L. P., ... Goldsmith, C. H. (2020). A tutorial on pilot studies: The what, why and how. BMC Medical Research Methodology, 10, 1. doi:10.1186/1471-2288-10-1

Trenado, R., Pons-Salvador, G., & Cerezo, M. A. (2009). Protecting the families: Support and assistance for families. Papeles del Psicólogo, 30, 24-32.

van IJzendoorn, M.H. (1995). Adult attachment representations, parental responsiveness, and infant attachment: A meta-analysis on the predictive validity of the Adult Attachment Interview. Psychological Bulletin, 117, 387-403. doi:10.1037/0033- 2909.117.3.387

van IJzendoorn, M. H., & Bakermans-Kranenburg, M. J. (2010). Invariance of adult attachment across gender, age, culture, and socioeconomic status? Journal of Social and Personal Relationships, 27, 200-208. doi:10.1177/0265407509360908

Yalom, I. D. (1985). The theory and practice of group psychotherapy (3rd ed.). New York, NY: Basic Books.

Received May 5, 2010

Revision received December 19, 2010

Accepted January 17, 2011

Author affiliation:

Bárbara Torres1, Itziar Alonso-Arbiol1, María José Cantero2, and Amina Abubakar3

1Universidad del País Vasco (Spain)

2Universidad de Valencia (Spain)

3Tilburg University (The Netherlands)

This research was supported by two grants from Vicerrectorado de Investigación de la Universidad del País Vasco (UPV05/111; GIU08/09), by two mobility grants from the Spanish Ministerio de Eduación to the second author (JC2008-00012; JC2009-00225), and by the financial support of Ayuntamiento de San Sebastian to carry out the intervention. The intervention programme was designed by the first author, and by C. Achúcarro, C. Herce, and A. Rivero (Centro LAUKA de Estudios e Intervenciones Psicológicas, S.L., de San Sebastián). A. Iglesias and R. Izaguirre (also from the same centre) implemented the programme under C. Achúcarro's surpervision. We would like to thank these professionals for their involvement and collaboration in this work. We would also like to thank the families who took part in this study, and the health workers who facilitated participants' recruitment.

Correspondence concerning this article should be addressed to: Bárbara Torres, Facultad de Psicología, Universidad del País Vasco, 20018 San Sebastián (Spain). Phone: +34-943018317. E-mail: barbara.torres@ehu.es

The use of this website is subject to the following Terms of Use