597 - Early Life Inflammation and Long Term Cognitive Outcomes in Bangladeshi Children

Publication Number: 3578.597

Rae Dong, Hassenfeld Children's Hospital at NYU Langone, New York, NY, United States; Charles a. Nelson, Boston Children's Hospital, brookline, MA, United States; Anne Arnett, Boston Children's Hospital, Boston, MA, United States; Georgios Sideridis, Boston Children's Hospital, Weymouth, MA, United States; Rashidul Haque, icddr,b, Dhaka, Dhaka, Dhaka, Bangladesh; William Petri, University of Virginia School of Medicine, Charlottesville, VA, United States; Fahmida Tofail, icddr,b, Dhaka, Dhaka, Bangladesh

Background: Chronic inflammation has been associated with adverse outcomes in linear growth and cognitive development. The early release of cytokines shapes the nature of inflammatory responses, and these responses can be beneficial or detrimental. The contribution of cumulative, long term exposure to inflammatory conditions on later cognitive outcomes remains less clear.

Objective: To examine the association between early childhood inflammatory exposure and cognitive developmental outcomes in a cohort of low-income children living in Bangladesh.

Design/Methods: This cohort examines a population at particularly high risk for chronic and recurrent infections, including diarrheal and febrile illnesses. Infants were enrolled at birth and followed prospectively. Inflammatory markers (cytokines) and cognitive measures were collected at regular intervals, up to age 7 years. A large panel of cytokines was summarized using exploratory and confirmatory factor analysis to reduce spurious results. The longitudinal associations between inflammation and IQ were analyzed with autoregressive cross lagged models.

Results: Exploratory and confirmatory factor analysis supported two inflammatory marker groups characterized by cytokines predominantly involved in viral and injury response (Factor 1) and type-2 immunity (Factor 2). Cytokine factors at 26, 36 and 60 months were modeled with full-scale IQ scores at 48, 60, and 84 months. Higher Factor 1 inflammation at 36 months was significantly associated with lower cognitive scores at 60 months. Stability of inflammatory factors varied across time points, while IQ was stable over age.

Conclusion(s): We present initial data suggesting that subsets of inflammatory cytokines may covary in early childhood. The effects of inflammation on cognition depends on age and the specific constellation of cytokine functioning. A better understanding of immune and inflammatory correlates of long-term neurodevelopment outcomes may enable early detection, prognostication, and more targeted therapies to address developmental delay in at risk groups.

Figure 1: Violin plot analysis of cytokines over time
Cytokine concentrations are indicated un pg/ml on the y axis. Time in months are indicated on the x axis. Mean is indicated by the solid line at each time point.

Table 1: Descriptive statistics of study population with cytokines collected
Descriptive statistics indicate that our study population was about 55% male, and tended to be below average for height weight and BMI for age across all ages studied. Overall, this study population experienced a high level of febrile illness, diarrheal illness and repeated courses of antibiotics.

Figure 2: Unadjusted and Adjusted Path Models
Using 2 factors defined using exploratory and confirmatory factor analysis, models of each factor of cytokines at 26 months, 36 months and 60 months were run against the full-scale IQ scores at 48 months, 60 months, and 84 months. For the factor 1 unadjusted model, the autoregressive path for cognitive scores was found to be significant and earlier cognitive scores were significantly associated with later cognitive scores. The cross lagged path for inflammation at 36 months to cognitive score at 60 months was found to be significant. For every 1 standard deviation increase in inflammation at 36 months, there was a 0.111 decrease in cognitive scores. There were no significant cross lagged paths found for the factor 2 unadjusted model. Ferritin levels at 18 weeks were included as a covariate in the adjusted factor 1 model. Ferritin was not found to be significantly associated with cognitive scores. After including all possible paths and adjusting for multiple comparisons, in factor 1 the cross lagged path for inflammation at 36 months to cognitive score at 60 months remained significant with a p-value of 0.01. For every 1 standard deviation increase in inflammation at 36 months, there was a 0.116 point decrease in cognitive scores.