Substance Abuse on Twitter

Reading Time: 2 minutes

Against the backdrop of lockdowns, quarantines, and working from home, many of us are turning to social media to make our voices heard. One popular topic is, unfortunately, due to the same stressors, drug abuse, and there is no shortage of commentary on the subject. According to the National Center for Health Statistics, around 70,000 Americans die from drug overdoses in each of the past four years.

Of the drugs mentioned, the most important from the perspective of the total healthcare burden is alcohol. More people consume alcohol than any other substance, including cannabis, cocaine, and heroin. According to the CDC, six people on average die of alcohol poisoning each day, of which over 75% are aged 35-64, and of which over 75% are male. Alcohol dependence, perhaps not surprisingly, was attributed as a factor in over a quarter of alcohol poisoning deaths.

These trends have not shown any signs of abating in light of the coronavirus pandemic since many people are drinking more often and by themselves than in the past. Although alcohol is the most prevalent drug of abuse, over 800,000 Americans used heroin as of 2018 based on a survey of over 70,000 respondents aged 12 and over. Since this is an essential health issue that affects many people, and since many of the deaths are preventable, we thought to examine this critical problem using mentions among over 400,000 tweets of drug abuse topics including drug and alcohol names.



In order to determine the states speaking about alcohol the most on Twitter, collected 405,589 tweets mentioning alcohol and drug types for the past 30 days ( 5/25/20-6/25/20)

We a random sampling of tweets in the United States that contained the words: Beer, Cider, Wine, Whiskey, Brandy, Vodka, Rum, Gin, Tequila, Absinthe, Liqueur Cocaine, DMT, Heroin, Ketamine, LSD, Marijuana, MDMA, Meth, Methamphetamine, PCP and Opium.

We collected the data by using textual analysis. The data we collected included, date, time, location, city, state, country, tweet text, images, photos, user IDs, number of likes, number of retweets, and comments.

We then determined each state’s weighted average based on the population.


Want to use our study on your website?

Please feel free! All that we ask is that you include a link back to this page so your readers can view the full study.