7578409652

%

of the number of independent variablcs to be tested, only main cffccts werc initially fittcd; relcvant interaction items werc subseąuently entered into the resulting eguation to obtain a fmal model. Independent variables tested for inclusion in the model were (I) noise-induccd hearing loss (yes = 1. no = O), (2) W-22 MAX (speech discrimination in noise) (<60 = l_t >60 = 0), (3) family history of hypertension (yes = 1, no = 0). (4) alcohoi consumption (grams of ethanol), (5) body mass iudcx (kg/m²), (6) noisy hobbies (yes = 1. no = 0), (7) age in years, and (8) cigarette consumption (number of packs per day). The finał model (sce Table 5) included throe independent variabłes—body mass index, family history of hypertension, and W-22 MAX—which to-gether cxplained approximately 7% of the variancc in the data.

The data set was subseąuently stratified into the two age groups. When the procedurę was applied to these data, it was found that body mass index was the only variable in the finał model for the younger (55 to 63 years) age group, whercas the finał model for the older (age 64+ years) age group included family history of hypertension, W-22 MAX, and body mass index. The latter model, when compared with the former, explaincd about three times the variancc in the data (/e, 4% v 13%), indicating that we were morę successful in pre-dicting hypertension status among the older subjeets.

Finally, the above analysis repeated the logistic regression only, without the W-22 MAX variable. This

TABLE 4

DistnOutłon ot Sodal Support Scores¹ for Noise-lnduoed and Non-Noise-

loduced Hearing Loss Groups

Age 56-63 y,    Age 64-69 y.

Noise-induced Hearing Non-Noise-Induced Loss    Hearing Loss

Mean (SD) Mean (SO) Mean (SO) Mean (SO)

Appraisal support 24.93(6.3) 25 9(7.1) 27.1 (7.5) 25.8(8.3) Betongingness    22.4    (6.6)    22.9(7.8)    23.5(7.4)    21.5(6.2)

Self-estccm    24.7    (5.8)    24.7(7.0)    24.5(7.3)    23.7(6.7)

Tangibie support    18.8    (6.5)    18.6(8.2)    18.9(7.3)    16.9(5.6)

Low sodal anxiety    22.2    (6.4)    20.9(6.8)    21.4(5.9)    24.1 (7.0)

was done to determine whetlier or not noise-induced hearing loss would onter into the finał model in the absence of W-22 MAX. When the logistic model was fit for all subjeets (see Table 5), noise-induced hearing loss did not enter the finał equation, which consisted of only two variables—body mass indcx and family history of hypertension. When the data were stratified according to age group, noise-induced hearing loss did enter the eguation for the older (age 64 f years) age group. The resulting eąuation for the older age group contained two independent variables. family history of hypertension and noise-induced hearing loss, which together exp!aincd about 9% of the variancc.

Discussion

The aims of the present study were to characterize morę finely the audiometric profile of a group of highly noise-exposed workers at a metal assembly plant and to determine whether severe noise-induced hearing loss was related to the prevalence of hypertension in this group of workers. The definition of severe noise-induced hearing loss was a puretone threshold of >65 dBA at 3, 4. or 6 kHz in either ear (within 20 dBA of each other) and is considcred to be a biologie marker of noise damage. In addition to using this definition of noise-induced hearing loss, an additional index of communi-cative difficulty, Hearing Performance Inventory scoros, was also utilized. This inventory measures five areas of hearing performance and was used as an indi-cator of interference with lifestyle or interpersonal com-munication abilities. A marginally significant relation-ship was noted between sevcre noise-induced hearing loss and the prevalence of hypertension among older men. There were nonsignificant correlations between the HPI and the morę objective puretone audiometric measures of hearing loss. No relationship was found between the total hearing performance inventory scores or HPI subscores and high blood pressure.

Anothcr measure of noise-induced hearing loss and cominunication in everyday life was the speech discrimination score in noise (W-22 MAX). This is an indicator of both the psychosocial impairment and biologie damage to the hearing mechanism. Referred to in this

TABLE 5

Muttipi© Logistic Regression of the Presence of Hypertension in Retired Maki Workers and S*gnif*cant Risk Factors with and withoul W-22 MAX*

Group	Variable	0		SE		>	c^1		P
		With W-22 MAX	Without W-22 MAX	With W-22 MAX	Without W-22 MAX	With W-22 MAX	Withoi/t W-22 MAX	With W-22 MAX	Without W-22 MAX
AU subjeets (n -	BMł	0.12338	0.11438	0.0375	0.0364	10.82	9.90	.001	.002
245)	FHX	0.94741	0.85124	0.2944	0 2846	10.36	8.94	.001	.003
	W-22 MAX	0.58225		0.2852		4.17		.041
Age group 55-63	BMI	0.12123		0.0487		6.19		.013
(n - 132)
Age group 64+	FHX	1.58481	1.57904	0.4639	0.4408	11.67	12.83	.001	.000
(n — 113)	W-22 MAX	1.14020		0.4580		6.20		.013
	BMI	0.11138		0.0580		3.68		.055
	NIHLf		0.90994		0.4681		3.78		.052

• Speech discrimination in noise/better ear.

t NIHL. noise-induced hearing loss; ł65 dB loss at 3.0. 4.0, or 6.0 kHz.

Journal of Occupational Medicine/Volume 32 No. 8/August 1990

695

1

Sodal support: low score = high sodal support; high score = low social support. Sodal anxiety: Iow score = high anxiety.