Predicting Adult Physical Illness from Infant Attachment:

pred adult illness childArticle Written by Jennifer Puig, Michelle M. Englund, Jeffry A. Simpson, and W. Andrew Collins
Health Psychology (2013) Vol 32(4) April 1, 2013
US National Library of Medicine

A Prospective Longitudinal Study



Recent epidemiological and longitudinal studies indicate that attachment relationships may be a significant predictor of physical health in adulthood. This study is among the few to prospectively link attachment classifications assessed in infancy to physical health outcomes thirty years later in adulthood, controlling for various health-related confounds.


Participants were 163 individuals involved in a 32-year longitudinal study of risk and adaptation who have been followed since birth. Attachment classifications were assessed at ages 12 and 18 months using the Ainsworth Strange Situation Procedure. Stability of attachment security was derived from these assessments. At age 32, participants completed a questionnaire asking about the presence of or treatment for current physical illnesses.


Binary logistic regression analyses controlling for health-related confounds at age 32 indicated that individuals who were insecurely attached (i.e., anxious-resistant or anxious-avoidant) during infancy were more likely to report an inflammation-based illness in adulthood than those classified as securely attached during infancy. There also was a trend whereby individuals classified as anxious-resistant reported more non-specific symptoms in adulthood than those classified as either anxious-avoidant or secure. Individuals who were continuously insecure during infancy were more likely to report all types of physical illness in adulthood.


These findings reveal the lasting effect of early interpersonal relationships on physical health and suggests that infancy may be a fruitful point for prevention efforts. The widespread influence that attachment has on endogenous and exogenous health-related processes may make it particularly potent in the prevention of later physical health problems.

Clinical and developmental psychologists have long assumed that early life experiences hold a privileged place in influencing later life outcomes (Sameroff, 2010; Sroufe, 1997). Although this framework is often associated with developmental precursors to psychopathology, it has been extended to the etiology of chronic diseases (e.g., Filetti et al., 1998; Shonkoff, Boyce, & McEwen, 2009). Researchers in medicine and psychology have suggested that developmental principles should be applied to the study of physical illnesses (e.g., Barker, Gluckman, Godfry, Harding, Owens, & Robinson, 1993; Boyce, 1985), but a lack of longitudinal data has prevented this framework from being tested (for an exception, see Friedman & Martin, 2011; Moffitt et al., 2010). The current longitudinal study was guided by a developmental approach (see Sroufe, Egeland, Carlson, & Collins, 2005) and examines how early parent-child attachment relationships may impact adult physical health outcomes 30 years later.

The links between interpersonal functioning and health outcomes have long been established. Several large-scale epidemiological studies have revealed that not having social relationships is a significant risk factor for poor health, similar in magnitude to smoking, obesity, and chronic physical inactivity (House, Landis, & Umberson, 1988). Other findings (e.g., Christakis & Fowler, 2007) indicate that obesity spreads through social networks across geographical and interpersonal distance up to three `degrees of separation' from target participants. Furthermore, the quality of close relationships, especially marital relationships, affects immune functioning, neuroendocrine functioning, and reactivity to stressful events, rendering individuals vulnerable to various diseases (Coan, Schaefer, & Davidson, 2006; Kiecolt-Glaser, Glaser, Cacioppo, & Malarkay 1998). These and other studies indicate the significant impact that relationships can and do have on the health and well being of adults.

Researchers have also focused on the role of early adverse experiences in laying the foundations for adult physical illness. For example, the `fetal origins hypothesis' states that maternal health and nutrition in the prenatal period send signals to the fetus about the relative harshness of world in which he or she will be born (Barker, 1995). Supporting this hypothesis, individuals with low birth weight (<5.5 lbs.) are more susceptible to diabetes and heart disease as adults than are individuals with typical birth weights (5.5–8.5 lbs; Barker et al., 1993; Roseboom et al., 2000). Other studies have found associations between adult health and adverse events in childhood and adolescence. The Adverse Childhood Experiences (ACE) Study retrospectively assessed abuse and household dysfunction in adults who were receiving a standard medical evaluation. This study found a linear association between exposure to abuse and dysfunction before age 18 and the likelihood of having a physical illness in adulthood (Felitti et al., 1998).

Viewed together, these studies document associations between both early adverse experiences and health and the quality of relationships and health. However, little research has examined the links between the quality of early relationships and physical health in adulthood. Despite the fact that relationship functioning is an emergent process shaped by an individual's prior history of relationships (Ainsworth, 1989), virtually all of the existing research examining relationship functioning and health outcomes involves concurrent measures taken at one time-point. Furthermore, longitudinal studies of social functioning and health (e.g., Christakis & Fowler, 2007; House et al., 1988) only examine time periods in adulthood. Incorporating prospective assessments of relationship functioning from early in development may give us an unbiased understanding of how the quality of relationships in infancy influence adult health.

The advantages of adopting a developmental theory linking social functioning with physical illness were first noted by Boyce (1985), who proposed that the primary attachment relationship between caregiver and child should be foundational in affecting the quality of social functioning, which in turn should affect physical health during childhood (see also Bowlby, 1969). This early relationship serves as a secure base from which infants can explore the world as well as a safe haven when they are distressed, and it reflects the quality of care that infants have received during the first years of life (Ainsworth, Blehar, Waters, & Wall, 1978). According to attachment theory, the quality of early care that children receive is internalized and then shapes their social functioning in adulthood (Bowlby, 1973). Individuals who are securely attached during infancy have a history of receiving sensitive and appropriate care, which increases their confidence in their ability to both provide and receive sufficient care later in life (Waters, Merrick, Treboux, Crowell, & Albersheim, 2000). Individuals who are insecurely attached during infancy, in contrast, have received sub-optimal care (i.e., either inconsistent care or rejection from caregivers). As a result, individuals with insecure histories have greater difficulty giving and/or receiving care in their adult relationships (Simpson & Rholes, 2012). Research suggests that infant attachment relationships are associated with aspects of health in childhood that may be linked to health across the lifespan (Anderson & Whitaker, 2011).

Observations of the connection between early attachment relationships and physical health, however, are rare in the literature. The findings, most of which are based on self-reported romantic attachment styles in adulthood, suggest that securely attached adults report more salubrious and fewer deleterious health behaviors than insecurely attached adults. These findings are stronger for individuals involved in romantic relationships (Scharfe & Eldredge, 2001). Self-reported attachment styles are also associated with physical illness (McWilliams & Baily, 2010). Adults who report being insecurely attached also report more illnesses than securely attached individuals. Moreover, adults who have an anxious attachment style report more cardiovascular-related illnesses. Those who have an avoidant style report more pain conditions.

The current longitudinal study contributes to the extant literature by prospectively examining the links between the quality of attachment relationships during the first two years of life and various health problems in adulthood. We hypothesized that individuals who were classified as insecurely attached in infancy (either anxious-resistant or anxious-avoidant) would report more physical health problems at age 32 compared to individuals who were classified as securely attached in infancy. Previous research has found that adult insecure attachment styles uniquely predict categories of physical illness (McWilliams & Baily, 2010). We tested whether insecure infant attachment classifications differentially predict inflammation-related illnesses and non-specific somatic complaints. Individuals classified as anxious-avoidant tend to suppress emotions and are prone to exaggerated inflammatory responses when exposed to interpersonal stressors (Gouin et al., 2008; Sroufe & Waters, 1977). Based on these findings, we predicted that infants classified as anxious-avoidant would later report more inflammation-related illnesses in adulthood. Infants classified as anxious-resistant are more likely to develop anxiety disorders in adolescence (Warren, Houston, Egeland, & Sroufe, 1997). Because anxiety disorders often include somatic symptoms such as muscle tension, headache, and upset stomach (American Psychiatric Association, 2000), we predicted that infants classified as anxious-resistant would be more likely to report non-pathognomonic symptoms (hereafter referred to as “non-specific symptoms”) in adulthood. Finally, we examined how the continuity of attachment security assessed at two time-points during infancy (at 12 and 18 months) predicted physical health outcomes in adulthood. We hypothesized that individuals who were classified as secure at both time-points would report the fewest health problems in adulthood.



Participants were drawn from a longitudinal study of risk and adaptation (see Sroufe et al., 2005, for a description). A low-income sample of women receiving free health care from public health clinics in a mid-western city between 1975–1977 were recruited during their third trimester of pregnancy. The current participants are the first-born children of the original participants; all were born into low-SES, high-risk environments. Two hundred and twelve participants were assessed at 12 months, 197 were assessed at 18 months, and 163 were assessed at age 32. Reasons for attrition include loss of contact with participant, moving out of state, and declining to participate. Eighty were male (49.1%) and 83 were female (50.9%). The racial composition was 64.5% white, 11.0% African-American, 18.7% mixed race, and 5.8% undetermined due to missing father information (See Table 1 for complete demographics).

Source Article
Article Written by Jennifer Puig, Michelle M. Englund, Jeffry A. Simpson, and W. Andrew Collins
Health Psychology (2013) Vol 32(4) April 1, 2013
US National Library of Medicine


© 2019 Light Your Sparkle, LTD
All Rights Reserved
website build | design: InCreation