Association of Cigarette Smoking Patterns Over 30 Years With Audiometric Hearing Impairment and Speech-in-Noise Perception

Abstract

The implications of cigarette smoking and smoking cessation for hearing impairment remain unknown. Many studies on this topic have failed to account for attrition among smokers in their findings. To assess the association of cigarette smoking patterns with audiometric and speech-in-noise hearing measures among participants of the Atherosclerosis Risk in Communities Study. This cross-sectional study included participants of the Atherosclerosis Risk in Communities Study from 4 US communities. The analysis includes data from visit 1 (1987-1989) through visit 6 (2016-2017); data were analyzed from March 16 through June 25, 2021. Audiometric hearing and speech-in-noise testing was offered to all participants at visit 6. Participants with incomplete audiometric data or missing data for educational level, body mass index, drinking status, a diabetes or hypertension diagnosis, or occupational noise were excluded. In addition, individuals were excluded if they self-reported as having other than Black or White race and ethnicity, or if they self-reported as having Black race or ethnicity and lived in 2 predominantly White communities. Smoking behavior was classified from visit 1 (1987-1989) to visit 6 (2016-2017) using group-based trajectory modeling based on self-reported smoking status at each clinic visit. Hearing was assessed at visit 6. An audiometric 4-frequency (0.5, 1, 2, 4 kHz) pure-tone average (PTA) was calculated for the better-hearing ear and modeled as a continuous variable. Speech-in-noise perception was assessed via the Quick Speech-in-Noise Test (QuickSIN) and modeled continuously. Attrition during the 30 years of follow-up was addressed by inverse probability of attrition weighting. A total of 3414 participants aged 72 to 94 years (median [IQR] age, 78.8 [76.0-82.9] years; 2032 [59.5%] women) when hearing was measured at visit 6 (2016-2017) were included in the cohort; 766 (22.4%) self-identified as Black and 2648 (77.6%) as White individuals. Study participants were classified into 3 smoking groups based on smoking behavior: never or former smoking at baseline (n = 2911 [85.3%]), quit smoking during the study period (n = 368 [10.8%]), and persistent smoking (n = 135 [4.0%]). In fully adjusted models, persistent smoking vs never or former smoking was associated with an average 2.69 (95% CI, 0.56-4.81) dB higher PTA (worse hearing) and 1.42 (95% CI, -2.29 to -0.56) lower QuickSIN score (worse performance). Associations were stronger when accounting for informative attrition during the study period (3.53 [95% CI, 1.14-5.93] dB higher PTA; 1.46 [95% CI, -2.52 to -0.41] lower QuickSIN scores). Smoking cessation during the study (vs never or former smoking) was not associated with changes in hearing. In this cross-sectional study, persistent smoking was associated with worse audiometric hearing and speech-in-noise perception. Hearing measures among participants who quit smoking during the study period did not differ from those for never or former smokers, indicating that smoking cessation (as opposed to persistent smoking) may have benefits for hearing health.