Estimating the incidence of cocaine use and mortality with music lyrics about cocaine

Cocaine use and mortality have surged in the United States over the past five years, following a period of decline. This resurgence is particularly noteworthy given cocaine's historical association with high social status and its glamorization in popular culture. Historically, cocaine use has been characterized by its prevalence in social settings, often perceived as a symbol of affluence. This perception, fueled by media portrayals and celebrity associations, has likely contributed to its appeal and downplayed the inherent dangers. The study hypothesizes that the recent increase in cocaine mentions in popular music reflects and potentially influences real-world trends in cocaine use and mortality. The researchers aim to empirically examine this relationship, providing insights into the recent rise in cocaine-related issues and the potential impact of contemporary music trends on population-level health behaviors. The study acknowledges prior research linking drug mentions in popular media to substance use for other drugs like smoking, alcohol, and cannabis, but notes a lack of empirical investigation specific to cocaine and its effect size on use behavior. Existing studies are also criticized for limitations in the scope of songs analyzed and reliance on ranked lists like the Billboard Top 100, potentially excluding a significant number of relevant songs.

The introduction extensively reviews existing literature on the relationship between media portrayals of substance use and subsequent behaviors. While studies have explored the link between media and substance use for other drugs (e.g., smoking, alcohol, cannabis), a gap exists in research specifically examining cocaine. The authors highlight the limitations of previous studies, such as their descriptive nature, limited scope of songs analyzed (often restricted to Billboard top charts), and lack of quantification of the effect size of drug exposure on behavior. The review emphasizes cocaine's unique social context and its portrayal in popular media as a status symbol. It sets the stage for this study by arguing that, given cocaine’s fashionable reputation and social patterns of use, descriptions of it in music are potentially an indication of increased initiation and use. The absence of previous empirical studies examining the epidemiological trends between popular media and cocaine use further justifies the current research.

This study used a mixed-methods approach combining quantitative data analysis with qualitative aspects in the consideration of slang terms. The researchers collected data from three main sources: 1. **National incidence of cocaine use:** Data on the incidence of cocaine use from 2002 to 2017 were obtained from the 2017 National Survey on Drug Use and Health (NSDUH) published by the Substance Abuse and Mental Health Services Administration (SAMHSA). National incidence was calculated by dividing initiation counts by midyear population estimates from the U.S. Census Bureau. 2. **National drug overdose deaths involving cocaine:** Yearly counts of cocaine-related deaths (2000–2017) were obtained from the Centers for Disease Control and Prevention's Multiple Causes of Death WONDER database, using ICD-10 code T40.5. These counts were divided by midyear population estimates. 3. **Cocaine mentions in song lyrics:** The researchers compiled a list of cocaine slang terms, employing a rigorous process of review by two independent researchers. This list was used to programmatically query Lyrics.com for yearly counts of songs mentioning these terms. To account for changes in the overall number of songs produced over time, the counts of cocaine mentions were transformed into a ratio of relative prevalence by dividing by mentions of the word "love", which was determined to be a consistent, dominant theme across decades. **Cocaine street price:** Data on cocaine street price (2000–2016), adjusted for purity and inflation, were obtained from the United Nations Office on Drugs and Crime. **Analysis:** The researchers used a first-order finite distributed lag model with unrestricted coefficients to analyze the association between cocaine mentions in lyrics and incidence/mortality. First-order differencing was used to make the time series data stationary. Ordinary least squares were used to fit the models. The models included cocaine price as a control variable to account for economic factors that might impact purchasing behavior and drug use initiation. Cross-correlation analyses were also conducted to validate the lag-time results. Sensitivity analyses were performed by removing the term “8-ball” and incorporating control models for codeine and heroin lyrics to assess the specificity of the findings related to cocaine.

A total of 6923 unique mentions of cocaine from 5955 unique songs and 2418 unique artists were identified in the Lyrics.com data. Cocaine mentions in song lyrics were stable from 2000 to 2010, then increased by 190% from 2010 to 2017, mirroring a similar surge in cocaine overdose deaths (200% increase from 2012 to 2017). The analysis revealed that a 0.01 increase in relative mentions of cocaine in song lyrics (relative to mentions of "love") was associated with: * An 11.8% increase in the incidence of cocaine use in the same year. * A 15.5% increase in cocaine mortality with a two-year lag. These associations remained robust after adjusting for street price. Cross-correlation analyses confirmed a significant correlation between cocaine lyrics and cocaine use incidence in the same year (r=0.5713, p<0.05), and between cocaine lyrics and mortality two years later (r=0.5747, p<0.05). Sensitivity analyses excluding the term "8-ball" did not materially alter the results. No significant lag times were observed for control models using codeine and heroin lyrics. Based on the model, projections for 2018-2020 indicated continued increases in cocaine use incidence and mortality. The study also noted a shift towards new slang terms for cocaine in recent song lyrics, potentially reflecting changes in user demographics and drug market dynamics. The increase in slang terms suggests the emergence of a new generation of cocaine users.

The findings suggest a strong association between increased mentions of cocaine in song lyrics and subsequent increases in cocaine use and mortality. The two-year lag time observed between lyrics and mortality aligns with epidemiological evidence regarding the time between cocaine initiation and the development of addiction and potentially fatal consequences. The results highlight the potential for popular music to serve as both a signal and potentially a driver of cocaine-related behaviors. While the study cannot definitively establish causality, the observed temporal relationships and the robustness of the findings after controlling for confounding factors offer compelling evidence supporting the association. The shift toward new slang terms for cocaine in recent lyrics highlights the need to continuously monitor language used in popular music to effectively track and understand evolving drug use patterns. The observed lack of association between codeine and heroin lyrics and cocaine epidemiology strengthens the argument for specificity in the cocaine-lyrics relationship.

This study provides initial epidemiological evidence supporting a link between cocaine mentions in popular music lyrics and the incidence of cocaine use and mortality. The two-year lag between lyrical mentions and mortality highlights a potential window for intervention. The shift to new slang terms suggests a changing user demographic. While causality cannot be definitively confirmed, the study underscores the value of monitoring trends in popular music lyrics as an early warning system for changes in cocaine use patterns. Future research should investigate the dynamic interplay between lyrical content and cocaine-related behaviors, focusing on how these messages shape perceptions and behaviors toward drug use. The potential influence of artists' lyrics on the drug epidemic affecting youth warrants further consideration.

The study acknowledges several limitations. First, the findings depict associations, not causality. While efforts were made to control for potential confounding factors, the possibility of unmeasured confounders cannot be entirely ruled out. Second, there is a potential for the incomplete capture of all cocaine slang terms, though sensitivity analyses indicated the robustness of the findings. Third, the analysis is limited to yearly data due to data availability constraints from both Lyrics.com and epidemiological sources. Finally, the definition of cocaine-related mortality may include deaths with underlying causes other than cocaine, which limits the study's ability to definitively ascertain the direct impact of cocaine. These limitations should be considered when interpreting the results.