King's College London

Research portal

Improvement for most, but not all: changes in newspaper coverage of mental illness from 2008 to 2019 in England

Research output: Contribution to journalArticle

Original languageEnglish
Article numbere177
Pages (from-to)e177
JournalEpidemiology And Psychiatric Sciences
Early online date5 Nov 2020
Accepted/In press4 Oct 2020
E-pub ahead of print5 Nov 2020
Published5 Nov 2020


King's Authors


AIMS: Time to Change, an anti-stigma programme in England, has worked to reduce stigma relating to mental illness in many facets of life. Newspaper reports are an important factor in shaping public attitudes towards mental illnesses, as well as working as a barometer reflecting public opinion. This study aims to assess the way that coverage of mental health topics and different mental illnesses has changed since 2008. METHOD: Articles covering mental health in 18 different newspapers were retrieved using keyword searches on two randomly chosen days of each month in 2008, 2009, 2010, 2011, 2013, 2014, 2016 and 2019. A content analysis approach using a structured coding framework was used to extract information from the articles. Logistic regression models were used to estimate the change in odds of each hypothesised stigmatising or anti-stigmatising element occurring in 2019 compared to 2008 and 2016 with a Wald test to assess the overall significance of year as a predictor in the model. Further logistic regression models were used to assess the association between the diagnosis that an article was about and the odds that it was stigmatising, and whether this relationship is moderated by year of publication. RESULTS: A total of 6731 articles were analysed, and there was a significant increase in anti-stigmatising articles in 2019 compared to 2008 (OR 3.16 (2.60-3.84), p < 0.001) and 2016 (OR 1.40 (1.16-1.69), p < 0.001). Of the 5142 articles that specified a diagnosis, articles about schizophrenia were 6.37 times more likely to be stigmatising than articles about other diagnoses (OR 6.37 (3.05-13.29) p < 0.001), and there was evidence that the strength of this relationship significantly interacted with the year an article was published (p = 0.010). Articles about depression were significantly less likely to be stigmatising (OR 0.59 (0.69-0.85) p = 0.018) than those about other diagnoses, while there was no difference in coverage of eating disorders v. other diagnoses (OR 1.37 (0.67-2.80) p = 0.386); neither of these relationships showed an interaction with the year of publication. CONCLUSION: Anti-stigma programmes should continue to work with newspapers to improve coverage of mental illness. However, interventions should consider providing specific guidance and promote awareness of rarer mental illnesses, such as schizophrenia, and evaluation should examine whether reductions in stigma extend to people with all mental illness diagnoses.

Download statistics

No data available

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454