Beware! Your Twitter Language Style May Reveal Hidden Risks of Coronary Heart Disease

The use of social media in healthcare continues to evolve, while providers are still working to determine how best to integrate it into healthcare institutions and patient care.

According to the latest psychological research, Twitter and the language used on the platform may serve as effective predictors of coronary artery disease. By analyzing negative emotions reflected in tweets, such as tension, fatigue, and anger, researchers have identified psychological factors as significant indicators, suggesting that users with a more positive tweeting style are at a lower risk of developing the disease.

According to a study conducted by the University of Pennsylvania, “linguistic patterns reflecting negative social relationships, social disengagement, and negative emotions (particularly anger) have emerged as risk factors for disease.” After controlling for the effects of income and education, the associations between most of these factors and disease prevalence remained highly significant.

This study demonstrates that tweets can significantly outperform models incorporating demographic, socioeconomic, and health risk factors—including smoking, diabetes, hypertension, and obesity—in predicting disease.

The research report stated, “It is feasible to capture psychosocial characteristics through social media, and these characteristics serve as strong indicators of community-level cardiovascular disease mortality.”



Researchers analyzed tweets from over 1,300 counties covering 88% of the U.S. population and found that users who posted tweets containing words such as “hate” or profanities were more likely to suffer from heart disease.

This study highlights the promise of leveraging social media data and represents a significant step toward integrating mental health into overall societal well-being.

Research indicates, “We have long believed that the mental health status of a community is crucial to physical health, yet measuring mental health remains highly challenging. By using Twitter as an observational window to study the collective mental state of a community, we can provide a useful tool for epidemiology and for evaluating the effectiveness of public health interventions.”

To this end, if social media is successfully leveraged, it “will represent a new frontier in psychological research.” Data collected from billions of internet users will play an extremely important role in the formulation and modification of public health initiatives, especially when combined with real-world outcomes.

Many startups and vendors are attempting to monetize social media and healthcare. Doximity is one example, positioning itself as a LinkedIn-style interactive platform designed specifically for physicians. Meanwhile, eRounds is connecting physicians who are willing to share their professional expertise in less formal settings.

However, to date, few startups have successfully applied psychology in practical settings, even though it is widely recognized that psychology plays a significant role in mental health and substantially impacts physical health. According to researchers, coronary artery disease is one of the leading causes of death worldwide, and its characteristics align closely with the findings of this study.

“People have long believed that psychological states influence coronary heart disease,” said Margaret Kern, an assistant professor at the University of Melbourne who made significant contributions to this study. “For example, through biological mechanisms, ‘hostility’ and ‘depression’ have both been linked to heart disease at the individual level. However, negative emotions can also trigger behavioral and social responses; under the influence of negative emotions, people are more prone to disordered eating or social withdrawal, thereby indirectly contributing to heart disease.”

Heart disease is a common cause of premature death, and public health officials meticulously verify the number of deaths attributed to it. They also collect rigorous data on health risk factors, such as the prevalence of smoking, obesity, hypertension, and physical inactivity. These data are available at the county level across the United States; therefore, the research team aimed to align this physical epidemiological investigation with its internet-based counterpart from their Twitter study.

The results of this study are consistent with existing sociological research, indicating that the composite characteristics of a community can predict health outcomes more accurately than any individual’s self-report.

This study was sponsored by the Robert Wood Johnson Foundation and the Templeton Religion Trust, and led by Johannes Eichstaedt, a graduate student in the Department of Psychology at the University of Pennsylvania’s School of Arts and Sciences.

(For the latest updates on digital health startups, please follow VCBeat’s WeChat official account: vcbeat. We also welcome you to engage with us on topics of interest, contact us via WeChat, and share your startup projects or related research insights.)