SciELO - Scientific Electronic Library Online

vol.90 número5Normalización escala Early Feeding Skills (EFS) en recién nacidos prematurosBatería para la evaluación del movimiento en niños -2- banda 1. Confiabilidad de la versión en español índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados

  • En proceso de indezaciónCitado por Google
  • No hay articulos similaresSimilares en SciELO
  • En proceso de indezaciónSimilares en Google


Revista chilena de pediatría

versión impresa ISSN 0370-4106

Rev. chil. pediatr. vol.90 no.5 Santiago oct. 2019 


Patterns of attachment in early infancy in normative samples, alternative caregiving systems, and high risk infants

Felipe Lecannelier1  3 

Germán Monje2  3 

Humberto Guajardo1 

1 Faculty of Medical Sciences, University of Santiago, Chile, Santiago of Chile.

2 Faculty of Social Sciences, University Santo Tomas, Puerto Montt, Chile.

3 Psychologist, Chile.



Scientific evidence gathered over the last decades has demonstrated that early attachment is a vital process for the understanding, prevention, and intervention of people's mental and physical health. However, information about the attachment system functioning in Chile is scarce. Objective: To des cribe attachment styles distribution in populations of children under different types of care.

Subjects and Method:

Through a retrospective descriptive method, 714 mother-child pairs (1 to 36 months- old) selected at random and purposefully, were assessed using the Strange Situation Procedure, Atta chment during Stress Scale (ADS), and Student-Teacher Relationship Scale. The samples were collec ted from randomly selected normative families, infants attending JUNJI nurseries and kindergartens, and infants from CONIN centers and who live in prison with their mothers.


The samples from six studies show the spectrum of the attachment system functioning in diverse conditions: 70% secure and 30% insecure, and 51.1% secure and 48.9% insecure in normative samples; 48.5% secure and 51.5% in secure in alternative care; 39.6% secure and 60.4% insecure, and 25% secure and 75% in secure styles in high-risk samples.


the study presents interesting evidence on the atta chment distributions in childhood, which allow reflecting on the uneven Chilean reality with regard to early social and emotional development.

Keywords: Attachment; early infancy; preventive interventions; early risk


The Attachment Theory is considered as the most comprehensive research program in social sciences, due to its number of studies (around 17,000 published researches), the scope throughout the life cycle (from pregnancy to adulthood), and the analysis levels mul tiplicity (form genetic to cultural)1,2. However, the re are many questions related to the behavior of the attachment system in different contexts and cultures. An important line of research focuses on comparing the distribution of four attachment patterns (Secure, Insecure- avoidant, Insecure-ambivalent, and Disor ganized) in different cultures and societies3. In general terms, there has been a tendency to find 2/3 of secure patterns, and the last third distributed in avoidant pat terns (12-15%), ambivalent patterns (15%), and disor ganized patterns (9%). However, cultural differences are evident4: for example, there are discrepancies bet ween Western European countries (towards avoidan ce) and Japan and Israel (towards ambivalence)4. The current conclusion is that although the attachment system has an evolutive-biological root (universal in the species), its cultural influence cannot be forgotten5.

Another related line of research focuses on unders tanding the behavior of the attachment system in the alternative care conditions (e.g. nurseries) and/or high vulnerability environments (e.g. early institutionalization). In the first case, broad evidence has demonstra ted that alternative care can have negative effects in the attachment development (both with alternative care givers and main attachment figures), especially when the stay time in the institutions exceeds four hours per day and the child:caregiver ratio is higher than six6. In the second case, studies in abused7 and early institu tionalized children8 have demonstrated that the pro portion of 2/3 secure attachment and 1/3 insecure and disorganized attachment is inverted, proving that these conditions are an extreme risk factor for the expected development of the child.

However, all these studies were performed in nor thern countries, and there is no complete evidence on the distribution of attachment patterns in normative and high-risk samples, and in alternative care contexts in Latin-American countries. Therefore, the objective of this study is to analyze the differences in attachment styles in different care contexts. Thus, the results allow representing a complete overview of the attachment system behavior in Chilean children, being the only contribution in the region.

Subjects and Method

Retrospective descriptive study. For clarification purposes, the six studies are categorized according to 1) Normative population in familiar context (study 1 and 2); 2) Normative population in alternative care (nurseries and kindergartens) (study 3 and 4); 3) Po pulation at extreme risk (early institutionalization and mothers-children in prison) (study 5 and 6).


- Sociodemographic indicators guideline: A general information guideline was created that included name, age, sex, date of birth, and general medical history.

- Strange Situation Procedure (SSP)9: Experimental procedure that consists of the detailed observation of the interactions between the mother and her child, in 8 episodes of 3 minutes, where the mother separates and reunites with her children, with another strange person in the room11. The instru ment is corrected based on a micro-analytical ob servation of the attachment behavior of the child, in relation to 4 scales scored from 1 to 7 points. The scales are: 1) proximity seeking; 2) contact maintaining; 3) avoidance of proximity and con tact, and 4) resistance to contact and comforting. The exhaustive and trained codifications show four attachment patterns: Secure Attachment; Insecure-Avoidant Attachment; Insecure-Ambi valent Attachment; and Disorganized Attachment. The procedure codification was performed by the first author (FL), who has the official internatio nal certification for this purpose (Institute of Child Development, University of Minnesota).

This procedure is considered the Gold Standard of early attachment assessment system (e.g. 341 pu blished studied up to 201010, due to its concurrent validity with other development measurements, predictive (up to adulthood), and multicultural validity11.

- The Massie-Campbell Attachment During Stress (ADS) Scale: This instrument is a modified, adap ted, and validated version of an observation gui deline of attachment quality indicators in the pe diatric care context12, observed in both the main caregiver and the child13. The observation guide line includes six behavioral attachment indicators (gazing, vocalizing, touching A, touching B, hol ding, affect, and proximity) which are organized in a scale ranging from 1 to 5 points, where each score shows a specific description of the type of behavior to be used (e.g. a score of 1 in the ‘gazing’ indi cator of the mother means that ‘she always looks away from child’s face’). Scores close to 1-2 show an insecure-avoidant attachment, and scores clo se to 5 show an insecure-ambivalent attachment. Scores close to 3-4 indicate a secure attachment. Many studies showed validity and reliability le vels appropriate both nationally14 and internatio nally15, and it was used as a standard attachment assessment measure (at 4 and 12 months of age) in the National Child Protection Program “Chile Crece Contigo” (Chile Grows with You), as well as in different studies of high-risk population and the efficiency in early intervention models16. All inte ractions were video recorded and their codification was performed by two experts with more than five years of experience in the use of the scale and the attachment observation.

- The Student-Teacher Relationship Scale (STRS)17: The Student-Teacher Relationship Scale short form (STRS-SF) was validated in Chile by Becar, Cheyre, and Diemer18, showing good results of re liability and factorial validity. It is a Likert scale, aimed at students aged between 4 and 8 years (but it can be used in younger children between 2 and 3 years of age). This scale is used to assess the percep tions that the teachers have about their relationship with a specific student. Therefore, it assesses the emotional relationship variable between teacher- student. It consists of a self-reporting questionnai re for the teacher. It has two sub-scales: Conflict: which measures the degree that he/she perceives his/her relationship with a student as negative and conflictive; Closeness: which measures the degree of affection, warmth, and open communication with a student, and a Total Scale which evaluates the general quality of the relationship.


1) Normative population in a familiar context (Stran ge Situation Procedure): Mothers-Babies (12 to 20 months of age) were selected in both studies (1 and 2). Regarding study 1, the selection process was carried out in primary health care centers (Lo Bernechea) and hospitals (Barros Luco Hospi tal), while in study 2, the selection was carried out through social networks and personal contacts. Se vere psychiatric psychopathology of the mothers, as well as congenital diseases of the child that do not allow an appropriate execution of the experi mental procedures were considered as exclusion criteria, and immigrant mothers were also exclu ded. Mothers and their children were invited to the offices of the Servicio de Psicología Integral of the Universidad del Desarrollo and the Unidad de Psi quiatría Sur of the University of Chile, where they were assessed (double mirror) after a series of ins tructions and signing the informed consent. These evaluations were video recorded for later analysis, which was performed after two months, and then finally analyze the data through the statistical soft ware SPSS Version 19.

The studies were approved by the ethics commit tee of the School of Psychology of the Universidad del Desarrollo and financed by the National Com mission for Scientific and Technological Research (CONICYT).

2) Normative population in alternative care (nur series and kindergartens) (study 3 and 4): Both samples were selected from the National Kinder gartens Board (JUNJI) institutions from the IV, VI, VIII, and Metropolitan regions as part of a study to assess the effectiveness of the Attachment Promotion Program and the Social-Emotional Learning financed by the Ministry of Social De velopment and Family. The age of the children ranged from 1 to 26 months (nurseries) and from 12 to 53 months (kindergartens). The selection of both samples was purposively made by JUNJI authorities, therefore, there were no inclusion or exclusion criteria. The video recordings were per formed in the context of the nurseries and kinder gartens, and the codification was carried out by trained expert evaluators. The study was approved by the ethics committee of the Universidad del Desarrollo and of the JUNJI.

3) Population at extreme risk (early institutionaliza tion and mothers-children in prison) (study 5 y 6): The first sample was selected from care centers of the Corporación para la Nutrición Infantil (CO NIN), which shelters children from 0 to 4 years with different chronic pathologies and/or psycho social vulnerability, with ages ranging from 0 to 33 months. Children from centers of six cities were assessed (Arica, Valparaiso, Temuco, Valdivia, Santiago, Los Angeles). The video recordings were made in the centers, and the codification was done by trained expert evaluators. The second sample was selected from 11 prison facilities (Arica, Iquique, Antofagasta, Copiapó, La Serena, Valparaíso, Rancagua, Concepción, Valdivia, Santiago, and Los Angeles), and consisted of mothers that live with their children (0 to 23 months). The dyads were as sessed in the prison facilities, the video recordings were performed in the centers, and the codification was carried out by trained expert evaluators. Both studies were approved by the ethics committee of the Universidad del Desarrollo and Gendarmería de Chile (GENCHI), and financed by the Ministry of Social Development and Family (Chile), as research to measure the effectiveness of the Attachment Pro motion and Respectful Care Program.


(Table 1) summarizes the sample size, the age (ave rage and standard deviation), and children’s sex of the six studies.

Table 1 Age (in months) and Standard Deviation of Children and Sample Size. 


The results are divided into two parts: The distri bution description of the attachment styles in the six studies/samples, and the analysis of the correlation between the attachment and the age and sex.

In relation to the two studies with normative sam ples, it was observed that the proportion of secure atta chment was 70% and the insecure and disorganized attachment was 30% (see Table 2, study 1). In study 2 (Table 3), it was observed 51.5% in the secure attach ment and 48.5% in the insecure and disorganized atta chments. Likewise, the proportion of avoidant attach ments increased from 10% to 30%. The ambivalent and disorganized proportion was stable.

Table 2 Distribution of Attachment Classification in Study 1. 

Table 3 Distribution of Attachment Classification in Study 2. 

Regarding the studies in alternative care samples (study 3 and 4), 49% in the secure attachment and 51% in the insecure and disorganized attachment were observed (Table 4). The high proportions of avoidant attachment stand out as well as the low percentage of ambivalent patterns.

Table 4 Distribution of Attachment Classification in Study 3. 

In the study conducted in kindergartens, where a different methodology was used, it was found that the educators perceive a high conflict with 35% of their students and 56% of low closeness with them (Table 5).

Table 5 Level of Closeness and Conflict in Study 4. 

Finally, in the studies in high-risk samples (study 5 and 6, see Table 6 and Table 7), it was found an inversion in the normative pattern of attachment distribution (2/3 secure and 1/3 insecure and disorganized). In the case of the samples from the CONIN centers of institutio nalized children, a 24.5% risk was observed (tendency towards disorganization), and 26.4% in the avoidant pattern (the total proportion of insecure was 60.3%). In the case of mothers and children in prison sample, 52.5% of high-risk attachments and only 25% of secu re attachments were found. Both results show a highly severe development in children younger than two years of age, and therefore alerting in both high vulnerability conditions.

Table 6 Distribution of Attachment Classification in Study 5. 

Table 7 Distribution of Attachment Classification in Study 6. 

The objective of the second type of analysis was to confirm the hypothesis that attachment patterns are not related to the age and sex of the child20. Through a series of correlations, no significant relations between the attachment and these types of variables were found in none of the studies.


The outcomes of the six studies show interesting discussions and conclusions for the unequal realities of childhood in Chile (and probably in Latin-America). Regarding the two normative samples, two main as pects should be discussed. First, it was confirmed the westernized trend of 2/3 secure attachment and 1/3 insecure and disorganized attachment4. The current discussion of whether this distributive pattern is so mething rather universal or specific to each culture3,5 raises the necessary reflection on whether Chile is part of the individualist-western societies or these distribu tions follow biological-universalist criteria21. There fore, a series of multicultural studies conducted in 24 countries on temperament, 16 on mental health, and 16 in parenting would confirm the first hypothesis22-24, thus, Chile would be reaching an adaptation difficulty pattern, which is a typical characteristic of individua listic and industrialized societies. Secondly, and related to the previous point, the increase of avoidant attach ment styles over the last 10 years stands out. And again, it is worth discussing if this distribution is related to this tendency towards individualism that our coun try seems to be experiencing. For example, it has been noted that avoidant styles are usually associated with internalizing mental health difficulties (depression, anxiety, shyness, etc.), as well as related to emotional over-control, and the multicultural results, where our country has participated, showed a 16-20% prevalence of internalizing disorders in preschool Chilean chil dren, which is three times higher than the global prevalence21.

Likewise, studies in Chilean nurseries and kin dergartens continue to confirm this Chilean trend towards avoidance/internalization. The outcomes in nurseries are especially worrying since there is a 37.1% of avoidant attachments (which added to the 12.4% of risk attachments, makes 47.5% of insecure attach ment before 12 months of age). During the first year, the theory indicates that the attachment system is in its formative stage, and our avoidant attachment rate (and low ambivalent attachments) leads to the con clusion that children enter nurseries with a tendency towards avoidance or they develop it in such places as a result from the care received or both. After analyzing the type of care given by the early-childhood educators under a micro-observation scale, the second hypothe sis was confirmed, in other words, the educators’ care is rather focused on the basic needs of the child, but not so on their socioemotional needs. In the educatio nal context of kindergartens, it was observed that the early-childhood educators report a high conflict with 35% of their students, and 44% of low closeness. Both results are also worrying, considering that evidence has shown that high conflict levels predict a negative edu cational climate and a low academic performance25. In relation to this, it could be analyzed if the teaching style and the expectations of the educators pressure a type of conflictive relationship when children express their stress. This has been confirmed in two findings whe re after interviewing the educators, it was found that a high percentage have expectations that a model child is ‘self-controlled, well behaved, does not cry, does not bother, and stays calm”26. In addition, in this study we observed that around 40-60% of the educators have beliefs contrary to what could be called ‘respectful rai sing’27. It is important to highlight that both studies confirmed the hypothesis of an avoidant/internalizing pattern in Chilean society.

Finally, the samples at extreme risk confirmed the devastating effects of adverse environments on early attachment development. The study of the children from CONIN (early institutionalization) confirmed the international results, where only 1/3 show a secure attachment and 2/3 show an insecure or disorganized attachment8. The difference and fundamental contri bution of this investigation are that, unlike the majo rity, the results were obtained before the first year of life and with direct-contact educators (DCEs). Both as pects highlight the double fact that the earlier the ins titutionalization, the higher is the risk, and that, unless the DCEs are specialized in more emotional care, it is not possible to create a minimum desirable environ ment for the adequate development of the child8. The national study of mothers and their children in prison makes an important contribution to the research re lated to attachment and prison. On the one hand, the 52.5% of risk attachment level found is not only very concerning but constitutes one of the highest propor tions found in high-risk studies. This compels reflec tion on the care conditions of the mothers (emotional rather than physical). On the other hand, it has been demonstrated that the stay of the child with his/her mother is a protective factor against recidivism28 and that an adequate intervention on respectful care can reduce the attachment risk effects and the socioemotional development of the children28. Therefore, these results showed the highly harmful effects of the prison context and the urgency of implementing intervention programs that are continuous, organized, and effective in these contexts.

Ethical Responsibilities

Human Beings and animals protection: Disclosure the authors state that the procedures were followed ac cording to the Declaration of Helsinki and the World Medical Association regarding human experimenta tion developed for the medical community.

Data confidentiality: The authors state that they have followed the protocols of their Center and Local regu lations on the publication of patient data.

Rights to privacy and informed consent: The authors have obtained the informed consent of the patients and/or subjects referred to in the article. This document is in the possession of the correspondence author.

Conflicts of Interest: Authors declare no conflict of interest regarding the present study.

Financial Disclosure: Comisión de Nacional de Investigación Científica y Tecnológica (CONICYT), Proyecto Fondecyt N° 1140265 y Ministerio del Desarrollo Social.


1. Cassidy J, Shaver PhR. (ed). Handbook of attachment, Third Edition: Theory, Research, and clinical applications. 3nd Edition. New York: The Guilford Press, 2018. [ Links ]

2. Lecannelier F. Apego e Intersubjetividad: El legado de los vínculos tempranos en el desarrollo humano y la salud mental. Segunda Parte. La teoría del Apego. Santiago de Chile: Editorial LOM, 2009. [ Links ]

3. Mesman, J, Van Ijzendoorn, M, Sagi-Schwartz, A. Cross-cultural pattern of attachment: Universal and contextual dimensions. En Cassidy J, Shaver PhR, editors, Handbook of attachment, Second Edition: Theory, Research, and clinical applications. 3nd Edition. New York: The Guilford Press; 2018, p. 827-55. [ Links ]

4. Van Ijzendoorn, MH., Zagi-Schwartz, A. Cross- cultural patterns of attachment. Universal and contextual dimensions. En Cassidy J, Shaver PhR, editores, Handbook of Attachment. Theory, Research and Clinical Applications. New York: Guilford Press; 2008, p. 880-905. [ Links ]

5. Lecannelier F. La Teoría del Apego: Una mirada actualizada y la propuesta de nuevos caminos de exploración. Aperturas Psicoanaliticas, 2018; 58, rescatado en [ Links ]

6. Belsky J. Early child care and early child development: Majors findings of the NICHD Study of early child care. Eur J Dev Psych. 2006; 3: 35-78. [ Links ]

7. Van Ijzendoorn MH, Schuengel C, Bakermans-Kranenburg M. Disorganized attachment in early childhood: Meta analysis of precursors, concomitants, and sequelae. Dev Psychopathol. 2009;11:225-50. [ Links ]

8. Bakermans-Kranenburg MJ, Steele H, Zeanah Ch, Muhamedrahimov RJ, Vorria P, Dobrova-Krol NA, Steele M, van Ijzendoorn MH, Juffer F, Gunnar MR. Attachment and emotional development in institutional care: Characteristics and catch-up. Monogr. Soc Res Child Dev. 2011;76:62-91. [ Links ]

9. Ainsworth MDS, Blehar MC, Waters E, Wall S. Patterns of Attachment: A Psychological Study of the Strange Situation. Hillsdale, NJ: Erlbaum, 1978. [ Links ]

10. Simonelli A, De Palo F, Parolin M, Moretti M. Interactive behaviors and attachment patterns in the strange situation procedure: A validation of the Ainsworth model. Psychol Beh Sci. 2014;3:75-84. [ Links ]

11. Solomon J, George C. The measurement of attachment security and related constructs in infancy and early childhood. En Cassidy J, Shaver PhR, editors, Handbook of attachment, Second Edition: Theory, Research, and clinical applications. 3nd Edition. New York: The Guilford Press; 2018, p. 366-98. [ Links ]

12. Massie HN and Campbell BK. The Massie-Campbell scale of mother-infant attachment indicators during stress. In Call J, Galenson E, and Tyson R editors. Frontiers of infant psychiatry. New York: Basic Books;1983. pp 394-412. [ Links ]

13. Lecannelier F, Zamora C. Escala de Apego durante Stress (ADS-III). Tercera versión: Manual de administración, corrección y guía de conductas. Santiago: Universidad del Desarrollo, 2013. [ Links ]

14. Acuña E, Maturana P, Muñoz A. Adaptación y validación de la Escala de Apego durante Estrés (ADS) en díadas madre-bebé chilenas. Tesis para optar al grado de magister en Psicologia clínica. Facultad de Psicología, Universidad del Desarrollo, 2011. [ Links ]

15. Carcamo R, van Ijzendoorn M, Vermeer HJ, van der Veer R. The validity of the Massie-Campbell Attachment during Stress Scale (ADS). J Child Fam Stud, DOI 10.1007/s10826-013-9728-z. [ Links ]

16. Lecannelier F, Silva J, Morales R, Melo R, Hoffmann M. Effects of an intervention to promote socio- emotional development in terms of attachment security: a study in early institutionalization in Chile. Infant Ment Health J. 2014;35:151-9. [ Links ]

17. Pianta RC. STRS, Student-Teacher Relationship Scale, Professional Manual. Lutz: Psychological Assessment Resources, 2001. [ Links ]

18. Becar C, Cheyre A, Diemer M. Adaptación y validación de una escala para evaluar la relación afectiva entre profesores y alumnos entre 4 y 8 años de edad: Escala de Relación Afectiva Profesor-Alumno (Student-Teacher Relationship Scale Short Form, STRS-SF). Tesis para optar al grado de magister en Estrategias en Salud Mental Infanti. Facultad de Psicología, Universidad del Desarrollo, 2011. [ Links ]

19. Lecannelier F, Kimelman M, González L, Núñez C., Hoffmann M. Evaluación de patrones de apego en infantes durante su segundo año en dos centros de atención de Santiago de Chile. Rev Argent Clín Psicol. 2008; XVII: 197-207. [ Links ]

20. De Wolff MS, van Ijzendoorn MH. Sensitivity and attachment: A meta analysis on parental antecedents of infant attachment. Child Dev. 1997;68:571-91. [ Links ]

21. Krassner AM, Gartstein MA, Park C, Dragan W, Lecannelier F, Putnam S. East-West collectivist-individualist: A cross-cultural examination of temperament in toddlers from Chile, Poland South Korea, and the US. Eur J Dev Psych. 2017;14:449-64. [ Links ]

22. Rescorla N, Achenbach TM, Ivanova MY, et al. International comparisons of behavioral and emotional problems in preschool children: Parent's report from 24 societies. J Clin Child Adolesc Psychol. 2012; 40:456-67. [ Links ]

23. Rescorla N, Achenbach TM, Ivanova MY, et al. Behavioral/emotional problems of preschoolers: Caregiver/teacher reports from 15 societies. J Emot Behab Disord. 2012; 20: 68-81. [ Links ]

24. Gartstein M, Putnam S. Toddlers, parents, and culture. Findings from the Joint Effort Toddler Temperament Consortium. Routledge, 2018. [ Links ]

25. Pianta RC. Enhancing relationship between children and teachers. Washington, DC: American Psychological Association, 1999. [ Links ]

26. Garrido AM, Weitzel CP. Significaciones en educadoras y auxilizares de párvulo respecto del aprendizaje socioafectivo en la educación parvularia. Tesis para optar al grado de magister en Estrategias en Salud Mental Infantil. Facultad de Psicología, Universidad del Desarrollo, 2011. [ Links ]

27. Lecannelier F, Jorquera L, Silva J. Diseño metodológico para el fortalecimiento del apego seguro, desarrollo socioemocional, y crianza respetuosa en la intervención de niños y niñas salas de cuna y jardines infantiles: Los Modelos AM.AR- Cuidadores y A.M.A.R.-Educadores. Ministerio de Desarrollo Social, Gobierno de Chile, 2014. [ Links ]

28. Murray J, Murray L. Parental incarceration, attachment and psychopathology. Attach Hum Dev. 2010;12:289-309. [ Links ]

Received: January 08, 2019; Accepted: May 13, 2019

Correspondence: Felipe Lecannelier. E-mail:

Creative Commons License Este es un artículo publicado en acceso abierto bajo una licencia Creative Commons