Furthermore, one subjectlacks all audiometric data and 173 participants show HTLsof 95 dB HL at one or more frequencies in both ears.In addition, 357 subjects show HTLs of 95 dB in one earand hearing threshold levels of 90 dB HL or better at allfrequencies in the other ear. For these subjects, only thelatter ear is preserved in the dataset.Data are excluded for 447 workers with insufficientnoise exposure data; they miss either information on jobtitle (n = 19) or duration of employment (n = 428).Finally, the 1,958 currently exposed workers that reportedprior employment in construction are excluded from theinternal control group.The excluded participants do not differ significantlyfrom the included subjects, except for younger age(-3.3 ± 0.5 years) and shorter employment duration(-6.0 ± 2.9 years). However, age-corrected hearing lossis similar in both groups (p = 0.908).The study population thus comprises 27,644 men and54,931 ears.Data analysisAll statistical analyses are performed using SPSS forwindows software, version 15.0. Binaural average thresh-olds are computed for each test frequency and for allsubjects. If threshold levels of only one ear are available,these are regarded as the binaural thresholds and are usedfor further analyses.Audiogram data usually have a positively skewed dis-tribution. However, the tested sample is assumed to belarge enough to approach a normal distribution and para-metric tests are used (Dawson-Saunders and Trapp 1994).The mean binaural hearing threshold levels of exposedworkers are compared to age-matched. ISO-standard val-ues using a paired Student’s t test, and to HTLs of the non-exposed control group using an independent Student’st test.In order to compare hearing thresholds of the noise-exposed workers to those of controls and to NIHL pre-dictions by ISO, HTLs of each participant are corrected forage effects by subtraction of the age-matched median HTLpredicted by annex A of ISO-1999.This ISO model assumes that noise-induced permanentthreshold shift (NIPTS) and age-related hearing loss(ARHL) are additive, according to the following empiricalformula:HTL ¼ ARHL þ NIPTS  ðARHL   NIPTSÞ=120The correction term (ARHL * NIPTS)/120 starts to modifythe result significantly when NIPTS ? ARHL is more thanapproximately 40 dB HL. To avoid underestimation ofNIPTS in this study, this correction term was taken intoaccount in calculating the age-corrected thresholds formeasured HTLs exceeding 40 dB HL.To simplify the results, hearing loss is also evaluatedusing pure-tone averages calculated for 1, 2 and 4 kHz(PTA1,2,4) and for the noise-sensitive frequencies 3, 4 and6 kHz (PTA3,4,6). These parameters are used in multiplelinear regression analyses, to investigate the dependence ofhearing threshold levels on noise intensity and exposuretime. Since there is an important dependence between ageand hearing loss, age is also considered as an explanatory variable. The possible statistically significant interaction ofnoise intensity and noise exposure time is tested by addinga product term in regression analyses.In addition, multiple linear regression analysis is usedfor the analysis of combined action of different parameterson PTA3,4,6 values. Modelling proceeded in several steps.First, bivariate relationships of the covariates with PTA3,4,6are checked by simple linear regression. All analyses areadjusted for age by including age as a covariate. Most ofthe categorical variables are dichotomous, and others areconverted into dummy variables before inclusion into theanalysis. Variables are retained for further modelling if theage-adjusted p value of the inpidual testing was\0.10.Second, a multiple linear regression model is created usingthe selected set of potential predictive variables. Relevantvariables are selected using a backward stepwise elimina-tion procedure, with p\0.05 for inclusion and p\0.10for exclusion.The use of hearing protection devices reduces noiseexposure, which may lead to overestimation of exposurelevels and attenuation of the exposure–response relation-ship (Sbihi et al. 2010). To reduce the effects of hearingprotection, some analyses are adjusted for reported HPDuse by performing stratified analyses for the subgroups ofHPD users and non-users.The level for statistical significance is taken as p\0.01for all analyses.ResultsGeneral population characteristicsThe total population of 27,644 men is pided into a largegroup of noise-exposed employees (n = 24,670) and aninternal non-exposed control group (n = 1,016). Theexposed group is slightly older than that of the controlgroup (average age 44.3 and 40.9 years, respectively, seeTable 2). Noise-exposed workers are significantly longeremployed in both the construction industry and their cur-rent occupation than controls. Mean employment differ-ences are 12.4 and 6.7 years, respectively. More than halfof the exposed workers have always been employed in thecurrent job (55.5%). Of the exposed employees, 75.5%claim to use hearing protection, 22.1% have complaints ofworsened hearing and 39.1% are bothered by noise duringwork. Smoking status, alcohol intake and blood pressure donot differ between the groups.Hearing threshold levelsTo examine the hearing ability of the employees, medianhearing threshold levels of the noise-exposed workers arecompared to median HTLs of the non-exposed controls andto age-matched thresholds reported in annex A of the ISO-1999 standard (Fig. 1). All curves show the well-known deterioration of hearingwith age, which is most prominent in the high frequencyregion. Both the exposed workers and the internal controlsshow significantly poorer hearing threshold levels relativeto the ISO predicted values, across the complete range oftest frequencies. 建筑业由噪音引起的听力损失英文文献和中文翻译(3):http://www.751com.cn/fanyi/lunwen_53357.html