Assessment of depression and anxiety in young and old with a question-based computational language approach

Abstract

Middle aged adults experience depression and anxiety differently thanyounger adults. Age may affect life circumstances, depending on accessibility ofsocial connections, jobs, physical health, etc, as these factors influence theprevalence and symptomatology. Depression and anxiety are typically measured usingrating scales; however, recent research suggests that such symptoms can be assessedby open-ended questions that are analysed by question-based computational languageassessments (QCLA). Here, we study middle aged and younger adults’ responsesabout their mental health using open-ended questions and rating scales about theirmental health. We then analyse their responses with computational methods based onnatural language processing (NLP). The results demonstrate that: (1) middle agedadults describe their mental health differently compared to younger adults; (2)where, for example, middle aged adults emphasise depression and loneliness whereasyoung adults list anxiety and financial concerns; (3) different semantic models arewarranted for younger and middle aged adults; (4) compared to young participants,the middle aged participants described their mental health more accurately withwords; (5) middle-aged adults have better mental health than younger adults asmeasured by semantic measures. In conclusion, NLP combined with machine learningmethods may provide new opportunities to identify, model, and describe mental healthin middle aged and younger adults and could possibly be applied to the older adultsin future research. These semantic measures may provide ecological validity and aidthe assessment of mental health.