No Conclusive Link Between Indoor Fuel Use and Atopic Dermatitis: A Systematic Review and Meta-Analysis with Rigorous Publication Bias Correction

Jing LangJianxun RenYaqin LiHuanhuan QuGaoyong DongYaning LiJinhe WangNa Lang^*

• Xiyuan Hospital of cacms, 北京市, China

Background: Household air pollution (HAP), primarily from the combustion of indoor fuels for cooking and heating, represents a major global public health challenge. Concurrently, the prevalence of atopic dermatitis is rising worldwide. However, epidemiological studies examining the association between indoor fuel use and the risk of AD have produced inconsistent findings. This study aimed to systematically evaluate the existing evidence and rigorously assess the potential impact of publication bias on this association. Methods: A systematic search of PubMed, Embase, Web of Science, and the Cochrane Library was conducted to identify observational studies assessing the association between indoor fuel use and AD risk. Odds ratios (ORs) and their 95% confidence intervals (CIs) were pooled using a random-effects model. Heterogeneity was evaluated using the I2 statistic and Cochran's Q test. Publication bias was comprehensively assessed using funnel plot visualization and multiple statistical methods, including Egger's, Begg's, Peters', Deeks', Macaskill's, and Tang's tests. Sensitivity analyses and bias-correction procedures, including the trim-and-fill method and parametric selection models, were performed. Subgroup analyses were conducted to explore potential sources of heterogeneity. Results: This analysis included 10 studies, comprising 21 independent effect sizes. The initial pooled analysis revealed a weak but statistically significant positive association between indoor fuel use and AD risk (pooled OR=1.158, 95% CI [1.051, 1.276], p=0.0031), with moderate between-study heterogeneity (I2=43.0%, p=0.0195). However, multiple statistical tests indicated a significant risk of publication bias (e.g., Egger's test, p=0.0006). The trim-and-fill method yielded an adjusted OR of 1.040 (95% CI [0.930, 1.164]), and parametric selection models produced similarly null results (e.g., standard selection model adjusted OR=1.000, 95% CI [0.992, 1.008]). Subgroup analyses suggested the association was primarily driven by studies on solid fuels and cooking-only use, but tests for subgroup differences were not statistically significant (p>0.05). Conclusion: After rigorous adjustment for publication bias, the available published evidence does not support the conclusion that indoor fuel use is an independent risk factor for atopic dermatitis. The weak association observed in the literature appears to be a statistical artifact, potentially influenced by systemic publication bias.