Neonatal Blood Levels of Growth Factors and Pleiotropic Cytokines in Association with Autistic Disorder: A Danish Population-Based Case-Control Study

Background: Both agnostic and targeted biomarker investigations have been conducted for associations with autism, commonly targeting the inflammatory pathway hypothesized to contribute to autism pathogenesis. Cytokines, components of immune system function and investigated as markers of prenatal infectious or inflammatory risk for autism, can function also as growth factors during development and in the mature organism. Growth factors act as signaling molecules between cells, stimulating cell growth, proliferation, differentiation and maturation and are important for regulating cellular processes. There are many large families of growth factors, such as the neurotrophins which include brain-derived neurotrophic factor (BDNF) investigated in relation to autism.

Objectives: Conduct a population-based case-control study of autistic disorder (AD) in association with neonatal levels of growth factors and pleiotropic cytokines measured in peripheral blood samples as indicators of abnormal or disrupted developmental and homeostatic processes, including immune activation, in the perinatal period.

Methods: The study cohort was comprised of all singleton Danish births from 1 January 1997 through 31 December 2003 identified in the Danish Medical Birth Register. Diagnoses of AD and perinatal and parent characteristics were obtained from the Psychiatric Central Research Register and Medical Birth Register. Study resources restricted the final sample to 550 randomly selected AD cases, matched in proportion on birth year to an existing control sample of 880 randomly selected cohort births. The biomarker panel was derived from a literature review for growth factors and pleiotropic cytokines reported to potentially affect neurodevelopment, especially autism. Based on multiplexing feasibility, a final panel of 19 biomarkers was developed. Measurements were performed on neonatal dried blood spots archived in the Danish Neonatal Screening Biobank. Case-control median biomarker levels, overall and by various strata, were compared using Tobit regression on the logarithm of the biomarker level and logistic regression was used to assess case-control differences in dichotomous variables defined on the basis of the 20th or 80th centile cut-offs of the control group distribution for each biomarker. Recursive partitioning was used in a combined analysis of all biomarkers to detect biomarker combinations which best discriminated cases and controls.

Results: Compared to control children, children with AD were more likely to have neonatal brain-derived neurotrophic factor levels below the 20^th centile (adjusted odds ratio 1.56, 95% confidence interval 1.13-2.16) and IL-3 levels above the 80^th centile (adjusted odds ratio 1.58, 95% confidence interval 1.12-2.23) and less likely to have leptin levels above the 80^th centile (adjusted odds ratio 0.50, 95% confidence interval 0.35-0.70).  From recursive partitioning, the combination of leptin levels below the 24^th control centile and ciliary neurotrophic factor levels above the 8^th control centile level discriminated cases from controls with 32% more cases than expected.

Conclusions: Small, but possibly discriminatory, shifts in neonatal peripheral blood levels of pleiotropic neurotrophic, neuroendocrine and immunomodulatory factors were observed to be associated with AD. Although these findings may reflect altered immune modulation in newborns later diagnosed with AD, alterations in multiple developmental networks may be implicated.