THE EFFECTS OF LONG-TERM CIGARETTE SMOKING ON SELECTED VOICE CHARACTERISTICS

The effects of long-term cigarette smoking were examined in a series of phonatory tasks. Subjects participated in an oral reading task, a spontaneous speech task, and the sustained productions of the vowels /i,a,u/ at their modal, highest and lowest pitch levels. Acoustic measures were derived through the use of speech analysis computer programs developed by Y. Horii at Purdue University. In this process, the voice signal to be analyzed, played through a tape recorder, undergoes an A/D conversion at an effective rate of 10,000 times per second (40,000 times per second for jitter and shimmer analysis), then fo measures are derived by using an automatic fo analysis program to process the quantified voice data. The major findings of this investigation may be summarized as follows: (1) Fo for the male smokers was significantly lower than the fo values for the male non-smokers in both oral reading and spontaneous speech. The same trend was evident in the fo analysis for the female subjects but it did not achieve statistical significance. (2) There was a trend for fo to decrease with increases in age in all groups studied. The trend was much stronger for the smokers than for the non-smokers. The combination of age related changes plus cigarette smoking create intrinsic laryngeal structure changes which reduce fo more than the aging process alone. (3) A significant difference was found between the male smokers and the male non-smokers in vocal SPL. The smokers had a vocal intensity which was 5dB less than that of the non-smokers. (4) There was a differential effect of aging on intensity levels for the groups studied. The non-smokers showed a trend for intensity to decrease with increases in age, but the smokers showed an opposite trend with their vocal SPL increasing with increases in age. Smoking scores also showed a positive correlation with intensity level. (5) No statistically significant differences were found between groups on the variable of percent jitter. (6) Jitter sign changes were significantly greater for the female smokers than for the female non-smokers. This result may indicate that the laryngeal feedback system is working harder to maintain the fine adjustments necessary to maintain on-going phonation. (7) No significant differences were found between groups on the variable of shimmer. (8) Shimmer sign change analysis revealed that female smokers exhibit greater sign changes than the female non-smokers. (9) Discriminant function analysis showed that when vowel and trial effects are held constant, maximal separation of the groups could be achieved on the basis of 20 variables