Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Its attractiveness among the public has been arbitrated by popular media, including coverage of celebrity drug activities associating the drug with high society2. In these ways, references to cocaine in the media and popular culture portrayed real-world behaviors of cocaine use.
a Cross correlation for cocaine mentions in lyrics leading cocaine use incidence at different lag times. b Cross correlation for cocaine mentions in lyrics leading cocaine deaths at different lag times. Blue lines indicate the 95% confidence interval around cross-correlation value of 0.
Our study explores the ability of song lyrics to signal epidemiological trends in incidence and mortality of cocaine use. Specifically, our results found that an increase in cocaine mentions in song lyrics is associated with increased incidence of cocaine use in the same year and death by cocaine 2 years later. The temporal relationship between cocaine lyrics and incidence of cocaine use and related deaths was confirmed with significant cross-correlation in the same year for cocaine use and 2 years for cocaine overdose mortality. The lead time period of 2 years found in this study between lyrics mentioning cocaine and incidence of cocaine and cocaine overdose mortality is supported by epidemiological evidence that the time between initiation of cocaine use and seeking help for addiction is between 1 and 3 years22,23. Musical trends that depict cocaine in lyrics may be an early signal of a rising interest and use of cocaine in the same year. The lag-time period of 2 years between cocaine lyrics and cocaine mortality may indicate that there is an incubation period of 2 years between addiction and fatality. Treatment intervention is critical in this time period to prevent the onset of drug dependence and death.
Our results also provide insights into the pathways by which media may influence cocaine use behaviors. In addition to being a measurement tool to estimate cocaine epidemiology, media itself can influence public perceptions of drug use and lead to increased drug use at the population level. Therefore, music about cocaine may not only provide an early signal of cocaine-related behaviors, but it could also act as an exposure that encourages the use of cocaine. Our study did not test the potential dynamic and cyclical relationship between cocaine lyrics and patterns of cocaine use. While cocaine mortality did not appear to estimate future cocaine mentions in song lyrics, further investigation into the influence of cocaine use on song lyrics is warranted.
An important limitation with our study is that our results depict associations and cannot confirm causality. Analytical steps were included in the methodology to remove potential spurious correlations but this is not enough to deem our results as causal. Therefore, our findings should be interpreted with caution. Cocaine street price was included in the model to control for economic fluctuations that would impact purchasing behaviors and initiation of drug use25. Spurious correlations were controlled through first-order differentiation. Further tests on the differenced series confirmed autocorrelations were no longer present after first-order differentiation.
In this study, we were only able to conduct analyses at the yearly level and not more granularly. Data derived from Lyrics.com is only available at the annual level. We explored several song lyrics engines including Spotify, Genius Lyrics, Metro lyrics, and Billboard; however, these other platforms either do not allow their data to be available or it is also at the yearly level. We selected Lyrics.com because of its word query capabilities within lyrics that other song lyric engines do not provide. This restricted us to have 18 effective observations from 2000 to 2017 for song lyrics data. However, in our analysis, another limiting factor was the incidence of cocaine data from the National Survey on Drug Use and Health published by the Substance Abuse and Mental Health Services Administration and cocaine-related deaths were obtained from the Center for Disease Control Multiple Causes of Death WONDER database, which only are available aggregated at the yearly level. Therefore, our analysis still would have been limited to this level of granularity. Additionally, since our definition of cocaine-related mortality includes deaths using the ICD-10 code T40.5, it may include underlying causes of death other than cocaine. However, other ICD-10 codes in the mortality dataset do not specify the actual drug involved. While this is a limitation in the use of death certificates for identifying cause-of-death, this is a consistently used measure for cocaine-related deaths. Furthermore, the quality of testing for drug overdose has improved over time27. Thus the rise in deaths by cocaine may be in part due to the more accurate determination of deaths attributed to cocaine.
Lastly, it is possible that mortality by cocaine could occur in the same year of initiation, which was not seen in our results. However, addiction to cocaine is often characterized by repeated use that changes brain and psychological function that promote transitions to problematic patterns of use28. Thus, the accumulation of conditional use of cocaine overtime that leads to mortality by cocaine is likely to occur after years of addictive use of cocaine.
Data on incidence of cocaine use from 2002 to 2017, defined as the nationwide number of new cocaine initiates aged 12 and older in the past year, was obtained from Table 7.28A of the 2017 National Survey on Drug Use and Health published by the Substance Abuse and Mental Health Services Administration3. National incidence was calculated by dividing initiation counts by midyear population estimates obtained from the United States Census Bureau34.
To examine the association of cocaine mentions in lyrics with incidence of cocaine and mortality, we used a first-order finite distributed lag model with unrestricted coefficients. In this model, the first differences of the time series of national counts of incidence of cocaine and mortality were regressed on first differences of the times series of cocaine mentions in lyrics earlier in time27,40,41,42,43.
To provide further validation of our lag-time results, we performed cross-correlation tests between lyrics and cocaine use and lyrics and deaths at varying lags to evaluate time-delayed associations. Cross-correlation analyzes the relationship between two signals and calculates the correlation coefficient at designated lags (displacements) between the two series. The correlation coefficient is maximal at the time point that the two series correlate most closely and the coefficient reveals the strength of the correlation. Cross-correlation analyses were conducted in both directions to examine whether the data supported cocaine epidemiology preceding lyrics or lyrics preceding cocaine epidemiology.
The data that support the finding from this study are publicly available from The Substance Abuse and Mental Health Services Administration and The Centers for Disease Control and Lyrics.com. Aggregated datasets are available on reasonable request.
Y.H. thanks Malcolm Mitchell for the introduction to understanding the deeper meaning of music lyrics. J.S.B. and Y.H. report funding by the National Library of Medicine of the National Institutes of Health (R01LM010812).
Y.H. secured funding for this study, conceived the idea, designed the study and analyses, and wrote the manuscript. A.Z. conducted the analysis and performed the computations, and helped write the manuscript. J.S.B. helped secure funding for this study, assisted in the design, and supervised the project and supported the interpretation of the study findings. All authors contributed substantially to the acquisition, analysis, interpretation, critical revision, and are accountable for all accuracy and integrity of the work.
Simply put, it was love at first trailer viewing. The NYLON team was enraptured by our first glimpses of Cocaine Bear, a film that looked too insane to exist, let alone be based off a true story. A movie about a bear that becomes so dangerously addicted to cocaine that it turns her (yes, her, as Cocaine Bear is a figurative and literal mother) from a relatively harmless forest-dwelling creature into a bloodthirsty murderer, has all the makings of a camp classic. The dark comedy knows exactly what it is: an unapologetic good time.
Moving on from the thoroughly captivating premise, Cocaine Bear is rounded out by amusing comedic performances, a soundtrack of perfectly placed '80s bangers, and genuinely freaky jump scares with gory ends. Were there plot holes, like, say, why the titular character's appetite was ravenous despite consuming several pounds of cocaine? Sure, but if there's any film to practice the art of suspending disbelief, it's Cocaine Bear.
Utilizing the perks of our jobs, the NYLON editorial team saw an early screening of Cocaine Bear in a little New York neighborhood called Times Square. And when in Rome, we did as Romans do and debriefed our thoughts on the film at Margaritaville. Read our roundtable, below, and watch Cocaine Bear in (a preferably packed) theater, now.
LH: The children were very knowledgeable about cocaine. I had a visceral reaction when they ate a heaping teaspoon of coke, and I think the entire theater did, too. I also screamed at every single jump scare.
AV: I feel like the point cannot be driven home strongly enough that you have to go see this in a packed theater. And don't bring too many friends. It's probably funnier with strangers. I was laughing at some of the comments from people behind me.
7fc3f7cf58