Breastfeeding and Respiratory Infections in the First 6 Months of Life: A Case Control Study.

Viral respiratory tract infections (VRI) are a major reason for hospitalization in children younger than 5 years. A case control study was conducted to investigate the potential role of breastfeeding in protecting children <1 year of age from VRI. Patients admitted for a respiratory tract infections routinely underwent a nasopharyngeal aspirate, which was tested with an RT-PCR for 14 respiratory viruses. Hospitalized infants positive for viruses were enrolled as cases; healthy controls were enrolled among patients admitted for ultrasound hip screening. The effect of breastfeeding on pertussis was investigated through multivariable analysis. We enrolled a total of 496 patients: 238 cases and 258 healthy controls. Among cases, eighty-six patients (36.1%) had a rinovirus, 78 (32.8%) an RSV, 22 (9.2%) an adenovirus, and 37 (15.5%) a coinfections with multiple viruses. The number of households was significantly higher in cases (mean in cases 4.5; mean 3.7 in controls, < 0.001) and the proportion of infants having siblings (79% in cases vs. 43% in controls, < 0.001). Proportion of smoking mothers was higher in cases than in controls (21.4 vs. 10.1%, = 0.001). Among cases 44.5% were exclusively breastfed at symptoms onset vs. 48.8% of healthy controls. According to the multivariable analysis, being exclusively breastfed at symptom onset was associated with a higher risk of viral respiratory infection (3.7; 95% CI 1.64-8.41), however a longer breastfeeding duration was protective (OR 0.98; 95% CI 0.97-0.99). Also having at least one sibling was associated to a higher risk (OR 3.6; 95% CI 2.14-5.92) as well as having a smoking mother (OR 2.6; 95% CI 1.33-4.89). Breastfeeding remains a mainstay of prevention for numerous diseases and its protective role increases with duration. However, being breastfed when mothers carry a respiratory infection may increase the risk of transmission, acting as a proxy for closer contacts. In future studies, potential confounding variables as pattern of contacts with other individuals, should be taken into account.