Sometimes Ghosts Stay Because They’re Loved: The Science of Connection and What It Means for How We Heal
“Sometimes ghosts stay because they’re loved.”
— Sarah Addison Allen, Other Birds
In Sarah Addison Allen’s novel Other Birds, a young woman arrives alone at a small apartment complex and discovers that the building holds more than tenants. There are ghosts. Not frightening ones. Quiet presences who linger because they are still tethered to someone living, still connected to the place where love once held them. The residents, each carrying grief or dislocation or some unnamed wound, begin to form a family that nobody planned. They heal not through resolution but through proximity. Through the slow accumulation of being known.
It is a story about what happens when isolated people stop being isolated. And it is, without naming it, a story about what medicine is only beginning to measure.
Allen’s ghosts are a metaphor worth taking seriously. Not because the dead literally stay, but because connection does. The relationships we carry, the ones we lose, the ones we rebuild, all of them leave a physiological signature. They shape immunity, inflammation, cardiovascular risk, recovery from trauma, and the capacity to survive grief. The science here is no longer speculative. It is among the most robust findings in modern public health.
And it raises a question the healthcare system has been slow to ask: if belonging is this protective, why does so little of clinical care account for it?
Loneliness Is Not a Feeling. It Is a Clinical Risk Factor.
The U.S. Surgeon General’s 2023 advisory did not frame loneliness as an emotional inconvenience. It named social isolation as an urgent public health crisis, one with mortality risk comparable to smoking fifteen cigarettes a day (USDHHS, 2023; Holt‑Lunstad et al., 2015 [foundational]). A major meta‑analysis found that lack of social connection is associated with a 29% increased risk of heart disease and a 32% increased risk of stroke (Valtorta et al., 2016). The American Heart Association followed with a scientific statement urging clinicians to assess social isolation alongside traditional cardiovascular risk factors (Cené et al., 2022).
These are not soft numbers. Over half of Americans report regular loneliness (USDHHS, 2023). Among older adults, social isolation is associated with billions of dollars in additional healthcare costs each year.
The body does not distinguish between physical threat and social absence. When connection is missing, the brain registers danger. Cortisol rises. Inflammation increases. Blood pressure climbs. Over time, loneliness alters gene expression: genes involved in inflammation become more active while antiviral defenses become less so (Cole et al., 2015; Hawkley & Cacioppo, 2010 [foundational]). The neuroscientist John Cacioppo described loneliness as analogous to hunger, a biological signal that something essential is missing (Cacioppo & Cacioppo, 2014 [foundational]).
This is not weakness. It is physiology. The body was designed for connection, and when connection is absent, the body keeps score.
Trauma, Grief, and the Architecture of Repair
Allen’s characters do not heal because someone fixes them. They heal because someone stays. That distinction matters clinically.
In trauma research, social support is consistently one of the strongest protective factors against post‑traumatic stress. A landmark meta‑analysis found that the presence of support after traumatic events was among the top predictors of better psychological outcomes (Ozer et al., 2003 [foundational]). Trauma‑informed care, as codified by SAMHSA, builds on this principle: that safety in relationships is often the precondition for recovery, not the afterthought (SAMHSA, 2014 [foundational]).
Grief operates on similar terrain. When people feel supported after a loss, they tend to experience less severe grief and are less likely to develop depression or prolonged bereavement (Levi‑Belz & Lev‑Ari, 2019). The risk of mortality rises sharply in the months following the loss of a spouse, particularly when the surviving partner lacks a support network (Holt‑Lunstad et al., 2015). Grief is not meant to be carried alone. The research is clear on this. But the healthcare system rarely asks who is carrying it with you.
In Other Birds, the ghosts do not resolve anything. They witness. They remain present. And their presence changes the quality of what the living can bear. This is not mysticism. It is, in clinical language, relational continuity: the sustained presence of someone who sees the whole picture and does not leave when things become complicated.
What Changes When Someone Stays
Medicine can prescribe a medication. It can schedule a follow‑up. It can refer to a specialist. What it struggles to provide is the sustained, coherent presence that holds care together between those moments.
Families often tell us they do not need more information. They need someone who already knows the story. Someone who can translate what the specialist said into what happens Tuesday morning. Someone who notices when a referral falls through or when a caregiver stops mentioning how they are doing. The evidence supports this instinct. Patients with strong social support show better surgical recovery, better treatment adherence, and better chronic disease management (Holt‑Lunstad et al., 2015). Peer support programs have been shown to improve psychological wellbeing in cancer patients (Le et al., 2024) and reduce rehospitalization in mental health care (Mental Health America, 2019). Connection is not supplementary to care. It is the context in which care succeeds or fails.
A steady clinical companion operates in this space. Not replacing the physician. Not adding another appointment to an already crowded calendar. Holding the continuity that the system does not structurally provide: tracking what was ordered and whether it happened, translating the plan into the patient’s actual life, staying present across the transitions where most things go wrong.
This is structural, not sentimental. The gap between visits is not a design oversight. It is how ambulatory care is built. And for patients with complex needs, that gap is where confusion accumulates, where medications go unreconciled, where caregivers absorb clinical labor no one acknowledged they were carrying.
Someone has to hold the thread. When no one does, the story fragments. And fragmented stories lead to fragmented care.
Some Healing Only Happens in Community
Allen gives us ghosts who stay because love made them reluctant to leave. In clinical life, what stays is less poetic but no less real: the memory of being seen, the relief of not having to explain everything again, the quiet difference it makes when someone already knows what you are going through. The science of connection is not new. But its integration into care remains incomplete. We screen for blood pressure and cholesterol. We do not routinely screen for loneliness. We build discharge plans around medications and appointments. We rarely ask who will be there when the patient gets home.
Some healing only happens in community. Some continuity only holds when someone chooses to stay.
What would change if we treated connection with the same seriousness we treat a lab value?
References
Allen, S. A. (2022). Other birds: A novel. St. Martin’s Press.
Cacioppo, J. T., & Cacioppo, S. (2014). Evolutionary mechanisms for loneliness. Cognition and Emotion, 28(1), 3–21. https://doi.org/10.1080/02699931.2013.837379 [Foundational]
Cené, C. W., et al. (2022). Social isolation and loneliness and cardiovascular and brain health: A scientific statement from the American Heart Association. Journal of the American Heart Association, 11(16), e026493. https://doi.org/10.1161/JAHA.122.026493
Cole, S. W., Capitanio, J. P., Chun, K., Arevalo, J. M. G., Ma, J., & Cacioppo, J. T. (2015). Myeloid differentiation architecture of leukocyte transcriptome dynamics in perceived social isolation. Proceedings of the National Academy of Sciences, 112(49), 15142–15147. https://doi.org/10.1073/pnas.1514249112
Hawkley, L. C., & Cacioppo, J. T. (2010). Loneliness matters: A theoretical and empirical review of consequences and mechanisms. Annals of Behavioral Medicine, 40(2), 218–227. https://doi.org/10.1007/s12160-010-9210-8 [Foundational]
Holt‑Lunstad, J., Smith, T. B., Baker, M., Harris, T., & Stephenson, D. (2015). Loneliness and social isolation as risk factors for mortality: A meta‑analytic review. Perspectives on Psychological Science, 10(2), 227–237. https://doi.org/10.1177/1745691614568352 [Foundational]
Le, P. T., Taylor, C., Bich, N. N., et al. (2024). Stronger together: A peer support intervention to improve psychological well‑being among cancer patients in Vietnam. JCO Global Oncology, 10, e2400031. https://doi.org/10.1200/GO.24.00031
Levi‑Belz, Y., & Lev‑Ari, L. (2019). Is there anybody out there? Attachment style and interpersonal facilitators as protective factors against complicated grief among suicide‑loss survivors. Journal of Nervous and Mental Disease, 207(3), 131–136. https://doi.org/10.1097/NMD.0000000000000940
Mental Health America. (2019). Evidence for peer support [White paper]. https://mhanational.org/sites/default/files/Evidence%20for%20Peer%20Support%20May%202019.pdf
Ozer, E. J., Best, S. R., Lipsey, T. L., & Weiss, D. S. (2003). Predictors of posttraumatic stress disorder and symptoms in adults: A meta‑analysis. Psychological Bulletin, 129(1), 52–73. https://doi.org/10.1037/0033-2909.129.1.52 [Foundational]
Substance Abuse and Mental Health Services Administration. (2014). SAMHSA’s concept of trauma and guidance for a trauma‑informed approach (HHS Publication No. SMA 14‑4884). U.S. Department of Health and Human Services. [Foundational]
U.S. Department of Health and Human Services. (2023). Our epidemic of loneliness and isolation: The U.S. Surgeon General’s advisory on the healing effects of social connection and community. Office of the Surgeon General.
Valtorta, N. K., Kanaan, M., Gilbody, S., Ronzi, S., & Hanratty, B. (2016). Loneliness and social isolation as risk factors for coronary heart disease and stroke: Systematic review and meta‑analysis of longitudinal observational studies. Heart, 102(13), 1009–1016. https://doi.org/10.1136/heartjnl-2015-308790
WHO Commission on Social Connection. (2025). Social connection global report: Restoring social relationships to improve health and wellbeing worldwide. World Health Organization.

