Department of Dermatology, Ningbo First Hospital, Zhejiang University, Ningbo, China.
Department of Dermatology, Ningbo First Hospital, Zhejiang University, Ningbo, China,
Dermatology. 2020;236(4):345-360. doi: 10.1159/000503781. Epub 2019 Nov 6.
The effect of breastfeeding on atopic dermatitis (AD) remains controversial. To determine the association -between breastfeeding and AD, we conducted an updated meta-analysis of prospective cohort studies.
A comprehensive search of PubMed, EMBASE, MEDLINE and Cochrane Library was conducted. Studies meeting the predetermined criteria were evaluated by 2 authors independently. The pooled relative risk (RR) adjusted for confounders with its 95% CI was calculated by a random-effects model. Heterogeneity was explored by subgroup analysis and meta-regression.
A total of 27 studies were included for meta-analysis. The pooled estimates for the effect of total and exclusive breastfeeding on AD were 1.01 (95% CI 0.93-1.10) and 0.99 (95% CI 0.88-1.11), respectively. Heterogeneity was substantial across studies (total: p < 0.01 or I2 = 65.2%; exclusive: p < 0.01 or I2 = 72.3%). There was a weak evidence for a protective effect of breastfeeding against AD in cohorts with atopic heredity (total: RR 0.85, 95% CI 0.74-0.98; exclusive: RR 0.83, 95% CI 0.70-0.97). In cohorts without atopic heredity, the effect shifted to the risk side when limited to exclusive breastfeeding (RR 1.19, 95% CI 1.02-1.40) while it dropped towards null when limited to total breastfeeding (RR 1.11, 95% CI 0.94-1.31).
There is no association between AD and breastfeeding, regardless of total or exclusive breastfeeding patterns. There is some evidence for a protective function of exclusive and total breastfeeding in a cohort with atopic heredity. The effect shifts to the risk side in cohorts without atopic heredity. However, these findings should be interpreted with caution because heterogeneity is evident.
母乳喂养对特应性皮炎（AD）的影响仍存在争议。为了确定母乳喂养与 AD 之间的关联，我们对前瞻性队列研究进行了更新的荟萃分析。
对 PubMed、EMBASE、MEDLINE 和 Cochrane Library 进行了全面检索。由 2 名作者独立评估符合预定标准的研究。使用随机效应模型计算经混杂因素调整后的合并相对风险（RR）及其 95%置信区间（CI）。通过亚组分析和 meta 回归探索异质性。
共有 27 项研究纳入荟萃分析。母乳喂养对 AD 的总效应和纯母乳喂养的效应估计值分别为 1.01（95%CI 0.93-1.10）和 0.99（95%CI 0.88-1.11）。研究之间存在显著的异质性（总：p<0.01 或 I2=65.2%；纯：p<0.01 或 I2=72.3%）。在具有特应性遗传的队列中，母乳喂养对 AD 具有弱的保护作用（总：RR 0.85，95%CI 0.74-0.98；纯：RR 0.83，95%CI 0.70-0.97）。在没有特应性遗传的队列中，当仅限于纯母乳喂养时，效应转移到风险侧（RR 1.19，95%CI 1.02-1.40），而当仅限于总母乳喂养时，效应趋近于零（RR 1.11，95%CI 0.94-1.31）。
无论总母乳喂养或纯母乳喂养模式如何，AD 与母乳喂养之间均无关联。在具有特应性遗传的队列中，纯母乳喂养和总母乳喂养具有一定的保护作用。在没有特应性遗传的队列中，效应转移到风险侧。然而，由于存在明显的异质性，这些发现应谨慎解释。
