BackgroundFood allergy (FA) in infants has become a common disease worldwide. There are many controversies surrounding the relationships among levels of cord blood 25-hydroxy vitamin D3 [25(OH)D3], total immunoglobulin E (IgE), and FA. MethodsIn this study, we recruited pregnant women in the third trimester undergoing obstetric examination in Chongqing City, Western China. Healthy full-term singleton births between May to August 2018 and November 2018 to January 2019 were included in the summer-birth and winter-birth cohorts, respectively. Questionnaires on vitamin D status in pregnancy and family allergies were used to investigate the pregnant women. The levels of <12 ng/mL, 12~20 ng/mL, and >20 ng/mL 25(OH)D3 in cord blood detected by liquid chromatography tandem mass spectrometry were considered deficient, insufficient, and sufficient, respectively. The electrochemiluminescence method was used to detect the total lgE levels in cord blood, classified into low-IgE (<0.35IU/mL) and high-IgE (≥0.35IU/mL) levels, respectively. Within postnatal 6 months, allergic symptoms in infants were investigated using questionnaire during the infants’ monthly physical examinations. Suspected cases of FA underwent a history inquiry, skin prick test, food elimination test, and open-food challenge for diagnosis of FA. Multivariate logistic regression was used to analyze the risk factors of FA in infants. ResultsIn this study, we recruited 741 pairs of pregnant women and infants, including 343 infants in the summer-birth cohort and 398 infants in the winter-birth cohort. The incidence of FA within postnatal 6 months was 6.88%, showing significantly higher incidence of FA in the winter-birth cohorts than in the summer-birth cohorts (10.3% vs. 2.9%, χ2 = 15.682, P = 0.000). Among the 741 infants, 47.1%, 27.5%, and 13.8% of infants had deficient, insufficient, and sufficient 25(OH)D3, respectively, in the cord blood; 81.5% and 18.5% of infants had total low-IgE and total high-IgE levels, respectively, in the cord blood. No significant correlation was found between the 25(OH)D3 and IgE levels (r = −0.038, P = 0.300). Logistic regression analysis showed that winter birth [odds ratio (OR) 95% confidence interval (CI): 4.292 (2.003~8.359)] compared with infants in summer birth group, and sufficient (>20 ng/mL) 25(OH)D3 levels in cord blood [OR (95% CI): 2.355 (1.129~4.911) compared with infants in the deficient group (<12 ng/mL) and 3.782 (1.680~8.514) compared with infants in the insufficient group (12~20 ng/mL)] were independent risk factors for FA in infants within postnatal 6 months. ConclusionsWinter birth and sufficient 25(OH)D3 levels in infant cord blood were independent risk factors for FA in infants. 25(OH)D3 and total IgE levels in cord blood cannot be used as predictors of FA in early infancy.