Sep 25 2020
There has been a loneliness pandemic in the last 20 years, marked by growing rates of opioid use and suicides, increased health care costs, lost productivity, and rising mortality.
Ellen Lee, MD, assistant professor of psychiatry at UC San Diego School of Medicine. Image Credit: UC San Diego Health Sciences.
According to the experts, the ongoing COVID-19 pandemic, with its associated lockdowns and social distancing, has only made things worse.
Precisely evaluating the depth and breadth of societal loneliness is a tedious task, restricted by available tools, like self-reports.
Now in a new proof-of-concept article, recently published online in the American Journal of Geriatric Psychiatry on September 24th, 2020, a team of researcher headed by scientists from the University of California San Diego School of Medicine, has utilized artificial intelligence technologies to study the natural language patterns (NLP) to determine the levels of loneliness in older adults.
Most studies use either a direct question of ‘how often do you feel lonely,’ which can lead to biased responses due to stigma associated with loneliness or the UCLA Loneliness Scale which does not explicitly use the word ‘lonely. For this project, we used natural language processing or NLP, an unbiased quantitative assessment of expressed emotion and sentiment, in concert with the usual loneliness measurement tools.
Ellen Lee, MD, Study Senior Author and Assistant Professor of Psychiatry, University of California San Diego School of Medicine
In the recent past, many studies have recorded increasing rates of loneliness in numerous populations of people, specifically those who are most vulnerable, like older adults. For instance, a UC San Diego study published earlier in 2020 discovered that 85% of residents, who live in a separate senior housing community, had reported moderate to severe levels of loneliness.
The latest study also considered independent senior living residents: 80 participants aged between 66 and 94, with an average age of 83 years.
However, instead of merely asking and recording answers to queries from the UCLA Loneliness Scale, trained study staff also interviewed participants in more unstructured conversations that were investigated using IBM’s NLP-understanding software as well as other machine-learning tools.
NLP and machine learning allow us to systematically examine long interviews from many individuals and explore how subtle speech features like emotions may indicate loneliness. Similar emotion analyses by humans would be open to bias, lack consistency, and require extensive training to standardize.
Varsha Badal, PhD, Study First Author and Postdoctoral Research Fellow, University of California San Diego School of Medicine
Among the findings:
- Lonely people had longer responses in qualitative interviews, and more significantly expressed sadness to direct queries related to loneliness.
- Compared to men, women were more likely to acknowledge feeling lonely at the time of interviews
- Men tend to use more joyful and fearful words in their reactions when compared to women.
According to the authors, the study underlines the differences between research evaluations for loneliness and a person’s subjective experience of loneliness, which could be reconciled by NLP-based tools.
The previous findings show how there might be a “lonely speech” that could be used for detecting loneliness in older adults, enhancing how families and clinicians evaluate and treat loneliness in older adults, particularly during times of social isolation and physical distancing.
According to the authors, the study demonstrates the possibility of utilizing natural language pattern analyses of transcribed speech to better parse and interpret complex emotions, such as loneliness. They stated that qualitative loneliness was predicted by machine-learning models with an accuracy of 94%.
Our IBM-UC San Diego Center is now exploring NLP signatures of loneliness and wisdom, which are inversely linked in older adults. Speech data can be combined with our other assessments of cognition, mobility, sleep, physical activity and mental health to improve our understanding of aging and to help promote successful aging.
Dilip Jeste, MD, Study Co-Author and Senior Associate Dean for Healthy Aging and Senior Care, University of California, San Diego
Jeste is also the co-director of the IBM-UC San Diego Center for Artificial Intelligence for Healthy Living.
The study’s co-authors include Sarah A. Graham, UC San Diego; Colin A. Depp, UC San Diego, and Veterans Affairs San Diego Healthcare System; Kaoru Shinkawa and Yasunori Yamada, IBM Research-Toyko; Ho-Cheol Kim, IBM Research-Almaden, San Jose, California; and Lawrence A. Palinkas from the University of Southern California.
The study was financially supported in part by the IBM Research AI through the AI Horizons Network, the National Institutes of Mental Health (grants T32-MH019934, K23-MH119375-01), the Brain and Behavior Research Foundation, the Veterans Affairs San Diego Healthcare System, and the Stein Institute for Research on Aging.
Journal Reference:
Badal, V. D., et al. (2020) Prediction of Loneliness in Older Adults using Natural Language Processing: Exploring Sex Differences in Speech. American Journal of Geriatric Psychiatry. doi.org/10.1016/j.jagp.2020.09.009.