SRCD 2019 Biennial Conference
March 21-23, 2019
Title: Close Relationships in Adolescence and Adulthood as Predictors of Inflammatory Markers in Midlife
Session: Close Relationships and Physical Health: A Lifespan Developmental Perspective
Authors: Jessica Stern, Katherine Ehrlich, Jason Jones, Emma Adam, Jacquelynne Eccles, Jude Cassidy
Abstract: Evidence suggests that high-quality social relationships may reduce risk for chronic disease (Robles et al., 2014). Little research, however, has examined how quality of attachment during specific developmental periods contributes to health risk. The present study examined attachment security in early adolescence, early adulthood, and midlife as predictors of two indicators of systemic inflammation: C-reactive protein (CRP) and interleukin (IL)-6. We focus specifically on African Americans, a population underrepresented in attachment research, yet overrepresented in prevalence of inflammation-related health problems (Yusuf et al., 2001).
In a secondary data analysis of the Maryland Adolescent Development in Context Study (MADICS), we examined indicators of attachment security in a well-characterized sample of 59 African Americans, followed from age 12 to 32. At age 12 and 20, participants reported their perceived ability to depend on their parents in times of need, using a Likert scale. During a home visit at age 32, participants completed a widely used measure of adult attachment style, the Experiences in Close Relationships scale (ECR; Brennan, Clark, & Shaver, 1998); attachment anxiety and avoidance were reverse-scored and averaged to create an index of adult attachment security. After a 20-minute rest period, participants provided blood samples from which CRP and IL-6 were measured using high-sensitivity chemiluminescence; height and weight were also measured.
Initial analyses indicated that the ability to depend on parents declined from age 12 to 20, t=2.49, p=.017; these two measurements were positively correlated, r=.34, p=.024, but neither was significantly related to security on the ECR at 32. Following standard practice, CRP values >10mg/L and IL-6 values >3SD above the mean were treated as outliers and excluded from analyses (Pearson et al., 2003). Hierarchical regressions predicting CRP and IL-6 were run, with sex and BMI entered as a priori covariates in Step 1 and each indicator of attachment security entered in Step 2. To improve normality of regression residuals, CRP and IL-6 values were log-transformed. Controlling for sex and BMI, attachment security at age 12 predicted lower CRP at age 32, B= -.29, p=.009. Attachment security in adulthood (age 20 and 32), however, did not significantly predict CRP. Attachment security at age 12 also predicted lower IL-6 levels at age 32, B= -.31, p=.019. This effect was not significant for security at age 20; however, security at age 32 predicted lower concurrent IL-6, B= -.24, p=.050. When security scores at age 12 and 32 were entered together, only security at age 12 was a significant predictor of each inflammatory marker.
Findings point to a role for attachment security in predicting lower levels of inflammation in midlife among African Americans. Importantly, security in early adolescence was the most robust predictor of both CRP and IL-6, over and above security assessed concurrently. This suggests that the ability to depend on parents in early adolescence, when caregivers are still important sources of support and stress regulation, may contribute to the developmental origins of physical health and disease.