Background: Various lines of evidence suggest that antenatal factors are important in determining susceptibility to atopy and asthma. One possible mechanism is cytokines, production of which in the placenta is high throughout gestation and which protect placental integrity via control of local immunological homoeostasis. We investigated antenatal cytokine concentrations in a prospective birth cohort, intensively monitored for atopy and asthma outcomes at age 6 years.
Methods: Cryopreserved cord-blood serum samples from 407 children were assayed for interleukins 4, 5, 6, 10, 12, and 13, interferon gamma, and tumour necrosis factor alpha (TNFalpha). Associations between family, antenatal, and perinatal factors, cord-blood cytokine concentrations, and atopy or asthma outcomes were analysed by logistic regression. Causal effects of cytokines on outcomes were estimated by propensity scores based on family, antenatal, and perinatal factors.
Findings: Detectable cord-blood concentrations of interleukin 4 and interferon gamma were each associated with lower risk of physician-diagnosed asthma (adjusted odds ratios 0.60 [95% CI 0.37-0.99] and 0.60 [0.37-0.97] respectively), current asthma (0.59 [0.33-1.00] and 0.39 [0.22-0.71]), and current wheeze (0.55 [0.32-0.93] and 0.52 [0.31-0.90]) and atopy (sensitisation to some inhalant allergens) outcomes at 6 years. High concentrations of TNFalpha were associated with lower risk of atopy but not with asthma risk. These associations were broadly unaltered by propensity-score adjustment. Maternal smoking was associated with higher risk of both wheeze at 6 years and lower concentrations of interleukin 4 and interferon gamma in cord blood.
Interpretation: The mechanism underlying attenuated T-helper-1/T-helper-2 cytokine production in high-risk children also apparently operates in control of cytokine production in the fetoplacental unit. The finding that this mechanism is dysregulated by maternal smoking suggests it is a target for antenatal environmental factors relevant to asthma aetiology.