Catching up with…Suicide Research: July 2021

Spread the science

A recent social media post grabbed my attention. The government of Japan appointed a Minister of Loneliness to address increasing suicide rates. This trend is, in part, driven by rising rates among women. Tetsushi Sakamoto will be responsible for implementing policy-based solutions to help reverse the trend. Japan is not the first country, though, to recognize how social isolation and loneliness spill over into the realm of public health.

Before he was re-appointed Surgeon General, Vivek Murthy wrote a book on the importance of social relationships. He drew from his experience on a nationwide listening tour that was ostensibly focused on health concerns. During this time, he learned about the ways that social connections can help or perpetuate health conditions. Even predating this, in 2018 Great Britain implemented a comprehensive loneliness strategy, which you can read about here.

Digging a little deeper, I did a quick PubTrawlr search for “loneliness.” It’s a vibrant area of research.

Just in the last month, there were five review papers published, which are incredibly diverse in terms of the content area. Reviews on geriatrics, education, neurobiology, and more: there’s a lot out there to dig into. I already read the Hehir et al. article and forwarded it to my local school board (of which, I sadly will not be a member).

Monthly Review

Using the same search criteria as my last post, but focused only on the past 31 days, we pulled all articles from Suicide, Suicide & Life-Threatening Behavior, Archives of Suicide Research, Suicidology Online, and Suicidologi and then searched for any instance of “suicid*”. This yielded 336 articles, which is a lot even if you were to screen them. The word cloud below shows what words are occurring more frequently across these 336 abstracts. Bigger words showed up more often.

To be a little bit more nuanced, the network plot below shows word strings that occur in the abstracts. The size of the circle corresponds to the number of uses of a term, and the thickness of the line refers to the frequency of occurrence between terms. This visualization usually does a fair job of capturing overall themes and methods (like logistic regression).

Review Articles

Because we always like to start with review articles, we identify these specifically. There were 26 articles in the last month. We put these in a .pdf table that you can download at the link below.

Topics and Themes

We then clustered the articles according to the content that they talk about. The figure below shows the frequency of the main topics of articles. The y-axis shows the key, distinguishing terms for each topic. This is what makes each topic unique. The bars show how often that topic appears across the 336 articles. Some notable topics include COVID-19, prisoner health and impact on survivors.

The topic that includes de, el, and la is an artifact of articles being published in a language other than English.

We then looked at relationships between the topics. The correlation plot below highlights relationships that were above 0.17. The decision to go with this specific parameter is arbitrary. We didn’t want the figure to be too crowded but also wanted to show some of the connections. The circles show the number of articles in each topic.

Representative Articles

In the below .pdf table, we show the article that is most representative of each of these twenty-five topics. We also included a summary of each topic that we derived through extractive summarization. It yields somewhat clunky results, but does give an overall gist of the topic. If you want to get a comprehensive view of last month’s published research, but can only focus on a few things, you may want to look at the articles for the largest topics.

Twitter Trends

This month, we also took a look at recent Twitter trends. The most basic Twitter API only allows people to pull the previous seven days of tweets, so this is a more limited snapshot than above. Below are the most frequent words and phrases that occurred in tweets with the hashtag #suicide.

The below figure shows the top tweeters for the hashtag “suicide.” Despite the name, suicidemediabot does not actually appear to be a bot. Generally, Twitter bots repeat a specific action, like retweeting instances of a hashtag. Instead, this handle looks at, “How does the #media cover #suicide, and how can we make it better?” An interesting follow!

Finally, these were the top tweets with #suicide, as defined by the number of likes. Dr. Tedros Ghebreyesus is the director-general of the World Health Organization (WHO). Cst. Ponsioen is a law enforcement officer in Vancouver.

Research Use in Behavioral Health

How do people consume literature? Over the next few days, we’re gathering information from behavioral health researchers and providers to help us understand how people use recent findings. Our goal is to be adapt our methods to get stuff out of the journals and into the hands of people doing the work. Answer the survey below for some cool rewards and/or swag! We’ll be closing the survey in a few days, mainly because our swag supply is currently limited.

For any comments or feedback, don’t hesitate to comment below or email us at gavin [at] pubtrawlr.com


Spread the science

Leave a Reply

Your email address will not be published. Required fields are marked *