Song audio features are associated with the presence of alcohol references in popular music lyrics

Abstract

Abstract Aims Preliminary evidence suggests that exposure to alcohol-related content in popular music is associated with drinking behaviour, especially amongst young people. Guided by Uses and Gratifications Theory, which posits that audiences select music to satisfy psychological and social needs, alcohol references, and specific audio features may co-occur as part of shared emotional and social experiences. This study assessed whether song audio features are associated with alcohol references in lyrics and whether they could support population-level surveillance of alcohol-related media exposure. Methods We analyzed 6110 Billboard Top 100 songs released between 1959 and 2020, extracting Spotify audio features for each track. Logistic regression models were used to examine associations between audio features and the presence of alcohol references in lyrics. Results Alcohol references were found in 16.1% of songs, with prevalence increasing significantly over time. Songs with higher danceability (OR = 1.35, P < .001), speechiness (OR = 1.39, P < .001), liveness (OR = 1.09, P = .021), positive valence (OR = 1.11, P = .041) and major mode (OR = 1.18, P = .042) more likely to contain alcohol references. Higher acousticness (OR = 0.82, P < .001) and instrumentalness (OR = 0.85, P = .015) were associated with lower odds. Loudness, tempo, and energy were not significantly associated. Conclusions Audio features may help identify songs more likely to contain alcohol references in music and may have potential utility for monitoring alcohol-related messaging across large volumes of popular music. Such tools could complement broader public health efforts to reduce alcohol exposure, especially amongst young people.