Huang Hesen, Zhou Wenkao, Chen Kaiqin, Du Yu, Lin Wei, Ye Yixian
Department of Otolaryngology, The Seventh Affiliated Hospital, Sun Yat-sen University, Shenzhen, China.
Department of Emergency Medicine, Xiang'an Hospital, Xiamen University, Xiamen, China.
Tob Induc Dis. 2025 Sep 17;23. doi: 10.18332/tid/208812. eCollection 2025.
Using data from the National Health and Nutrition Examination Survey (NHANES), this study aimed to assess the association between smoking and hearing loss and explore its heterogeneity across gender and age groups.
This secondary dataset analysis used a cross-sectional design and included NHANES data from 2005-2012 and 2015-2018. The study population consisted of adults without hearing-related disorders. Hearing loss was assessed by pure tone audiometry (PTA) and included low-frequency (LFHL), speech-frequency (SFHL) and high-frequency hearing loss (HFHL), expressed as continuous and categorical variables, respectively. Linear and logistic regression models were used to analyze the association between hearing loss and the exposure variable smoking status.
A total of 4217 adult subjects were included. It was found that smokers had a higher prevalence of LFHL, SFHL and HFHL than non-smokers (LFHL: 15.62% vs 8.51%, SFHL: 23.22% vs 12.98%, HFHL: 53.48% vs 36.95%). In males, in Models 1 (unadjusted), 2 and 3 (adjusted), there were statistically significant differences (p<0.05) in LFHL (β=4.24; 95% CI: 3.32-5.17; β=1.65; 95% CI: 0.80-2.49; β=1.52; 95% CI: 0.66-2.38) in SFHL (β=5.63; 95% CI: 4.56-6.70; β=1.95; 95% CI: 1.05-2.84; β=1.62; 95% CI: 0.72-2.52) and in HFHL (β=10.20; 95% CI: 8.21-12.19; β=2.85; 95% CI: 1.33-4.37; β=2.19; 95% CI: 0.69-3.70) between smokers and non-smokers, for continuous variables of hearing loss. In male hearing loss, categorical variables also showed statistically significant differences between smokers and non-smokers (p<0.05). In the middle-aged group, compared with non-smokers, logistic regression of smoking with all three types of hearing loss showed statistically significant differences (p<0.05) in Models 1, 2 and 3.
There was a significant association between smoking and hearing loss, with maybe more significant associations with all three types of hearing loss in male smokers and a significant relationship between smoking and hearing loss in the middle-aged group.
本研究利用美国国家健康与营养检查调查（NHANES）的数据，旨在评估吸烟与听力损失之间的关联，并探讨其在性别和年龄组间的异质性。
这项二次数据集分析采用横断面设计，纳入了2005 - 2012年以及2015 - 2018年的NHANES数据。研究人群为无听力相关疾病的成年人。通过纯音听力测定（PTA）评估听力损失，包括低频听力损失（LFHL）、言语频率听力损失（SFHL）和高频听力损失（HFHL），分别表示为连续变量和分类变量。采用线性和逻辑回归模型分析听力损失与暴露变量吸烟状况之间的关联。
共纳入4217名成年受试者。发现吸烟者的LFHL、SFHL和HFHL患病率高于非吸烟者（LFHL：15.62%对8.51%，SFHL：23.22%对12.98%，HFHL：53.48%对36.95%）。在男性中，对于听力损失的连续变量，在模型1（未调整）、模型2和模型3（调整）中，吸烟者与非吸烟者在LFHL（β = 4.24；95%置信区间：3.32 - 5.17；β = 1.65；95%置信区间：0.80 - 2.49；β = 1.52；95%置信区间：0.66 - 2.38）、SFHL（β = 5.63；95%置信区间：4.56 - 6.70；β = 1.95；95%置信区间：1.05 - 2.84；β = 1.62；95%置信区间：0.72 - 2.52）和HFHL（β = 10.20；95%置信区间：8.21 - 12.19；β = 2.85；95%置信区间：1.33 - 4.37；β = 2.19；95%置信区间：0.69 - 3.70）方面存在统计学显著差异（p < 0.05）。在男性听力损失方面，分类变量在吸烟者与非吸烟者之间也显示出统计学显著差异（p < 0.05）。在中年组中，与非吸烟者相比，吸烟与所有三种类型听力损失的逻辑回归在模型1、模型2和模型3中均显示出统计学显著差异（p < 0.05）。
吸烟与听力损失之间存在显著关联，男性吸烟者中吸烟与所有三种类型听力损失的关联可能更显著，且中年组中吸烟与听力损失之间存在显著关系。
