职业性累积噪声暴露时长与听力损失和血脂谱的纵向关联：一项9年队列研究

文献信息

中文摘要

背景： 职业性噪声暴露会导致听觉损伤，并且与全身代谢异常密切相关。长期噪声暴露对血脂纵向变化的影响仍不清楚。
目的： 本纵向研究考察了男性工人基线累积噪声暴露时长与血清甘油三酯（TG）和总胆固醇（TC）的关联，并进一步检验双耳高频平均听阈（BHFTA）——噪声性耳蜗损伤的客观标志物——是否改变了血脂的纵向趋势。
方法： 在这项回顾性纵向队列研究中，431名男性工人被随访9年。职业性噪声暴露根据中国国家标准GBZ/T 224-2010定义为暴露于8小时等效连续A计权声级至少80 dB(A)（Lex, 8h），并操作为从工厂行政记录中确定的噪声暴露岗位的累积就业时长。采用线性混合效应模型分析随访时间、基线累积噪声暴露时长（均值中心化、时间不变）及其交互作用对自然对数转换的TG和TC的影响，并调整年龄、体质指数、降脂药物使用、吸烟和饮酒。在代表性暴露时长（3、10和20年）处估计简单斜率。由于无法直接测量个体生物学噪声剂量，在敏感性分析中使用BHFTA作为累积生物学剂量和个体易感性的客观标志物，并在BHFTA值为均值±1个标准差（SD）处估计简单斜率。
结果： 基线累积噪声暴露时长与血清TG水平呈显著正相关，每增加一年暴露对应TG水平约高2.0%（β = 0.0203/年，基于ln(TG)，95% CI：0.0029-0.0377，P = 0.022），而其与TC的关联未达到统计学显著性（P = 0.077）。随访时间与基线暴露时长的交互作用方向为负，即基线暴露时间较长的工人血脂年增长速率趋于较小；这在TG上达到临界显著性（β = -0.0013，95% CI：-0.0027-0.0001，P = 0.066），在TC上无统计学显著性（P = 0.479）。估计的年TG增幅在基线暴露时长3、10和20年之间呈现单调的描述性梯度（分别为每年5.35%、4.39%和3.03%）。在敏感性分析中，BHFTA与随访时间对两个结局均呈正交互作用（TG：0.0023，95% CI：0.0013-0.0034；TC：0.0005，95% CI：0.0001-0.0009）；随着BHFTA从均值-1 SD增至均值+1 SD（14.3-24.9 dB），TG的估计年增幅从2.07%升至4.61%，TC从0.73%升至1.28%。
结论： 基线累积噪声暴露时长与TG水平升高相关，而随时间减弱的血脂增长速率可能部分反映了健康工人幸存者效应。相比之下，BHFTA可能比行政暴露记录更敏感地揭示慢性噪声暴露的代谢后果。听力保护和代谢监测应整合到职业健康监测中。

英文摘要

BACKGROUND: Occupational noise exposure causes auditory damage and is also closely associated with systemic metabolic abnormalities. The impact of long-term noise exposure on longitudinal lipid changes remains unclear.
OBJECTIVE: This longitudinal study examined the association of baseline cumulative noise exposure duration with serum triglycerides (TG) and total cholesterol (TC) in male workers, and further examined whether binaural high-frequency average hearing threshold (BHFTA), an objective marker of noise-induced cochlear damage, modified longitudinal lipid trends.
METHODS: In this retrospective longitudinal cohort study, 431 male workers were followed for 9 years. Occupational noise exposure was defined according to the Chinese national standard GBZ/T 224-2010 as exposure to an 8-hour equivalent continuous A-weighted sound level of at least 80 dB(A) (Lex, 8h), and was operationalized as the cumulative duration of employment in noise-exposed positions ascertained from plant administrative records. Linear mixed-effects models were used to analyze the effects of follow-up time, baseline cumulative noise exposure duration (mean-centered, time-invariant), and their interaction on natural log-transformed TG and TC, adjusted for age, body mass index, lipid-lowering medication use, smoking, and alcohol consumption. Simple slopes were estimated at representative exposure durations (3, 10, and 20 years). Because individual biological noise dose could not be directly measured, BHFTA was used as an objective marker of accumulated biological dose and individual susceptibility in sensitivity analyses, with simple slopes estimated at BHFTA values of mean ± 1 standard deviation (SD).
RESULTS: Baseline cumulative noise exposure duration was significantly and positively associated with serum TG levels, with each additional year of exposure corresponding to an approximately 2.0% higher TG level (β = 0.0203 per year on ln(TG), 95% CI: 0.0029-0.0377, P = 0.022), whereas its association with TC did not reach statistical significance (P = 0.077). The follow-up time and baseline exposure duration interaction was negative in direction, that is, the annual rate of lipid increase tended to be smaller among workers with longer baseline exposure; this reached borderline significance for TG (β = -0.0013, 95% CI: -0.0027-0.0001, P = 0.066) and statistical non-significance for TC (P = 0.479). The estimated annual TG increase displayed a monotonic descriptive gradient across baseline exposure duration of 3, 10, and 20 years (5.35%, 4.39%, and 3.03% per year, respectively). In sensitivity analyses, BHFTA interacted positively with follow-up time for both outcomes (TG: 0.0023, 95% CI: 0.0013-0.0034; TC: 0.0005, 95% CI: 0.0001-0.0009); the estimated annual increase rose from 2.07 to 4.61% for TG and from 0.73 to 1.28% for TC as BHFTA increased from mean - 1 SD to mean + 1 SD (14.3-24.9 dB).
CONCLUSIONS: Baseline cumulative noise exposure duration was associated with elevated TG levels, and the attenuated rate of lipid increase over time may partly reflect a healthy worker survivor effect. In contrast, BHFTA may be more sensitive than administrative exposure records in revealing the metabolic consequences of chronic noise exposure. Hearing conservation and metabolic monitoring should be integrated into occupational health surveillance.