Parental and household smoking and the increased risk of bronchitis, bronchiolitis and other lower respiratory infections in infancy: systematic review and meta-analysis

Abstract

Background: Passive smoke exposure increases the risk of lower respiratory infection (LRI) in infants, but the extensive literature on this association has not been systematically reviewed for nearly ten years. The aim of this paper is to provide an updated systematic review and meta-analysis of studies of the association between passive smoking and LRI, and with diagnostic subcategories including bronchiolitis, in infants aged two years and under.

Methods: We searched MEDLINE and EMBASE (to November 2010), reference lists from publications and abstracts from major conference proceedings to identify all relevant publications. Random effect pooled odds ratios (OR) with 95% confidence intervals (CI) were estimated.

Results: We identified 60 studies suitable for inclusion in the meta-analysis. Smoking by either parent or other household members significantly increased the risk of LRI; odds ratios (OR) were 1.22 (95% CI 1.10 to 1.35) for paternal smoking, 1.62 (95% CI 1.38 to 1.89) if both parents smoked, and 1.54 (95% CI 1.40 to 1.69) for any household member smoking. Pre-natal maternal smoking (OR 1.24, 95% CI 1.11 to 1.38) had a weaker effect than post-natal smoking (OR 1.58, 95% CI 1.45 to 1.73). The strongest effect was on bronchiolitis, where the risk of any household smoking was increased by an OR of 2.51 (95% CI 1.96 to 3.21).

Conclusions: Passive smoking in the family home is a major influence on the risk of LRI in infants, and especially on bronchiolitis. Risk is particularly strong in relation to post-natal maternal smoking. Strategies to prevent passive smoke exposure in young children are an urgent public and child health priority.

Figure 1

Flow diagram of included and excluded studies.

Figure 2

Funnel plot for household passive smoke exposure against lower respiratory infection. Plot shows the standard error of the odds ratio versus odds ratio for each study (random effects model). Vertical dotted lines indicate pooled effect estimate; and dots, individual studies.

Figure 3

Relationship between passive smoke exposure by any household member and the risk of lower respiratory infection (LRI) in infancy using a meta-analysis of comparative epidemiologic studies (Data are presented as odds ratios sub-grouped by the definition of LRI outcome). Squares denote the odds ratio (OR) for a single study with horizontal lines denoting 95% confidence intervals. The centre of the diamond denotes the pooled OR and the corners the 95% confidence intervals. An OR > 1 indicates a higher risk of the outcome in those exposed to passive smoke.

Figure 4

Relationship between passive smoke exposure by both parents and the risk of lower respiratory infection (LRI) in infancy using a meta-analysis of comparative epidemiologic studies (Data are presented as odds ratios sub-grouped by the definition of LRI outcome). Squares denote the odds ratio (OR) for a single study with horizontal lines denoting 95% confidence intervals. The centre of the diamond denotes the pooled OR and the corners the 95% confidence intervals. An OR > 1 indicates a higher risk of the outcome in those exposed to passive smoke.

Figure 5

Relationship between passive smoke exposure by maternal smoking after birth and the risk of lower respiratory infection (LRI) in infancy using a meta-analysis of comparative epidemiologic studies (Data are presented as odds ratios sub-grouped by the definition of LRI outcome). Squares denote the odds ratio (OR) for a single study with horizontal lines denoting 95% confidence intervals. The centre of the diamond denotes the pooled OR and the corners the 95% confidence intervals. An OR > 1 indicates a higher risk of the outcome in those exposed to passive smoke.