Cai YuJia, Zeng HaiFeng, Tao MaoCan
Department of Cosmetic Dermatology, The First Affiliated Hospital of Zhejiang Chinese Medical University, Hangzhou, China.
Zhejiang Provincial Hospital of Chinese Medicine, Hangzhou, China.
J Cosmet Dermatol. 2024 Dec;23(12):4123-4128. doi: 10.1111/jocd.16498. Epub 2024 Aug 13.
Rosacea can be seen in many patients nowadays, and the related causes are complex. Despite a certain association between smoking and rosacea being reported by several studies, the actual causality has not been established for the possible bias and confounders.
We used Mendelian randomization (MR) to evaluate a potential causal effect of smoking on rosacea risk. Statistics on smoking and rosacea were obtained from the FinnGen project and Neale Lab Consortium. The causal association was assessed by multiple methods including inverse variance weighted (IVW), MR Egger, weighted median, and weighted mode. Furthermore, sensitivity analyses were also conducted to address pleiotropy, along with the leave-one-out method.R version 4.2.3 was applied for the analyses.
The IVW estimation revealed that previous smoking has a deleterious effect on rosacea (odds ratio [OR] = 6.7729, 95% confidence interval [CI] = 1.5691-29.2356, p = 0.0104). By contrast, there was no statistically relationship between current smokers and rosacea (OR = 0.6180, 95% CI = 0.0605-6.3094, p = 0.6847). Results were similar in the analysis based on the weighted median method (previous smoking: OR = 8.6297, 95% CI = 1.0131-73.5071, p = 0.0486; current smoking: OR = 0.2896, 95% CI = 0.0106-7.9132, p = 0.4627). The stability of the causal effect estimates was supported by several sensitivity analyses and the leave-one-out method.
Our MR study found support forrosacea risk and previous smoking. Although no evidence was found to increase the risk of rosacea in current smokers, to prevent various diseases associated with smoking, the public should be encouraged to avoid smoking at the very beginning.
如今在许多患者中都可见到酒渣鼻，其相关病因复杂。尽管多项研究报告了吸烟与酒渣鼻之间存在一定关联，但由于可能存在的偏倚和混杂因素，尚未确立实际的因果关系。
我们使用孟德尔随机化（MR）来评估吸烟对酒渣鼻风险的潜在因果效应。吸烟和酒渣鼻的统计数据来自芬兰基因项目和尼尔实验室联盟。通过多种方法评估因果关联，包括逆方差加权（IVW）、MR埃格检验、加权中位数和加权模式。此外，还进行了敏感性分析以解决多效性问题，并采用了留一法。分析使用R版本4.2.3。
IVW估计显示，既往吸烟对酒渣鼻有有害影响（优势比[OR]=6.7729，95%置信区间[CI]=1.5691 - 29.2356，p = 0.0104）。相比之下，当前吸烟者与酒渣鼻之间无统计学关系（OR = 0.6180，95% CI = 0.0605 - 6.3094，p = 0.6847）。基于加权中位数方法的分析结果相似（既往吸烟：OR = 8.6297，95% CI = 1.0131 - 73.5071，p = 0.0486；当前吸烟：OR = 0.2896，95% CI = 0.0106 - 7.9132，p = 0.4627）。因果效应估计的稳定性得到了多项敏感性分析和留一法的支持。
我们的MR研究发现既往吸烟与酒渣鼻风险之间存在关联。尽管未发现当前吸烟者增加酒渣鼻风险的证据，但为预防与吸烟相关的各种疾病，应从一开始就鼓励公众避免吸烟。
