We are developing a conversational app to help disengaged adults struggling with loneliness, Seniors struggling with loss of purpose, self-value and hope can talk through their challenges and re-build social skills and improve mental health. This app uses AI bots trained by humans to initiate conversations and extend loneliness interventions benefits.
Introduction
There has been a significant rise in Loneliness research and awareness which has highlighted the widespread impact it has on our mental and physical health [1]. Research shows that loneliness can lead to depression, diabetes, cardiovascular diseases, dementia, self-harm and early mortality [2]. This has major implication for workplace productivity and healthcare expenses.
Mental Health Impacts
Depression & Anxiety
Cognitive Decline
Physical Health Impacts
Cardiovascular Issues
Immune System
Metabolic Health
Inflammation
Background - Humans need social connections to thrive. It has been proven time and again through research that socially active and connected individuals are generally happier, have fewer health issues and live longer.
Over the course digital revolution and spread of messaging and social media apps, physical and in-person interactions have reduced. The social fabric has been further eroded in the last two years due to COVID restrictions and need for people to isolate.
Gone are the days when colleagues, friends, neighbors, could easily and without pretense catch up to share their joys and sorrows, consult, advise and find resolutions for everyday stress, loneliness, work fatigue etc. This has resulted in socially inadept folks resorting to professional help. Despite the pandemic, digital first approach, our need to find a listening ear has remained unchanged. If anything, it has gone up.
Unfortunately, in current super charged atmosphere it's not easy to start a conversation with anyone. If you can carefully navigate all political, personal and other contentious issues, you are likely to be faced with privacy and availability concerns.
Research is also focused on Loneliness variants (e.g., situational, emotional or social). and challenges with commonly practiced unidimensional interventions (focusing on one therapy like cognitive behavioral therapy [CBT] We now have evidence that interventions using these techniques have limited impact as do not take individual’s core loneliness cause into consideration (one-size-fits-all fallacy).
Safeguards – Our app comes with pre-built guardrails to prevent any harmful content from AI. This also ensure no PHI, bias and negative feedback Our solution also has templates and built-in AI personas. This enables AI to wear multiple hats (like friend, trainer, cheer leader etc.) This enables app to customize to changing user needs.
We will also test the emergency and safety guardrails during the pilot. If user reports harmful intent via a combination or words or sentences, app will capture the instance details, escalate to care-take or emergency contact and connect to 988. If the harm intent is not explicit but implied, app will prompt the user with related questions to clarify their intent.
The duration of this study will be 2 months. We will need participants to chat with the app at least once a day.
Pilot outcome - We are targeting loneliness reductions (SMD scores > 0.5, with CI 95%) 6-9 months post intervention. We are also targeting liveliness score increase in users (tbd)
References
[1] https://reporter.nih.gov/matchmaker/gA8MzKyPQUmdZhU80qP__A/projects - number of grants on loneliness, mental health, isolation and related projects has gone up from low 100s to over 200 between 2023 and 2025
[2] https://www.cdc.gov/social-connectedness/risk-factors/index.html - loneliness and health impact
[3] https://www.nature.com/articles/s44220-024-00382-3 - Walsh, B.E., Rottenberg, J. & Schlauch, R.C. Why loneliness requires a multidimensional approach: a critical narrative review. Nat. Mental Health 3, 175–184 (2025)
[4] https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2024.1427605/full - meta analysis
[5] https://www.apa.org/pubs/journals/releases/amp-amp0001578.pdf
[6] https://agmr.hapres.com/htmls/AGMR_1255_Detail.html Akhter-Khan SC, Au R. Why loneliness interventions are unsuccessful: A call for precision health. Adv Geriatr Med Res. 2020;
[7] https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-023-17097-2 - Morrish, N., Choudhury, S. & Medina-Lara, A. What works in interventions targeting loneliness: a systematic review of intervention characteristics. BMC Public Health 23, 2214 (2023)
[8] https://www.sciencedirect.com/science/article/pii/S1525861024005322 - meta analysis
[9] https://pmc.ncbi.nlm.nih.gov/articles/PMC11676195/#app1-nursrep-14-00281 Zaharia G, Ibáñez-Del Valle V, Cauli O, Corchón S. The Long-Lasting Effect of Multidisciplinary Interventions for Emotional and Social Loneliness in Older Community-Dwelling Individuals: A Systematic Review. Nurse Rep. 2024 Dec 6;14(4):3847-3863.
#lonely, #alone, #sad, #feelinglonely, #depression, #solitude, #mentalhealth, #introvert,