CGPR: How would you define loneliness, and how does it differ from social isolation?
Dr. Joo: I think of social isolation as the objective measure, the number of relationships someone has, or the frequency of their interactions. Loneliness is the subjective, internal experience: distress about the quality or quantity of one’s relationships, or dissatisfaction with what a person has. The two are closely related and often used interchangeably in conversation, but they don’t always move together. Someone can have a large network and still feel lonely, or have very few contacts and not feel lonely at all. That distinction matters clinically, because addressing isolation by adding contacts doesn’t automatically resolve loneliness, and vice versa.

CGPR: How common is loneliness in older adults, and has that changed over time?
Dr. Joo: It’s a big number. Pooled international data estimate that in high-income countries, about 28% of older adults are lonely—roughly one in four (Chawla K et al, PLoS 2021;16(7):e0255088). In the US, the University of Michigan National Poll on Healthy Aging found that among adults aged 50–80, lack of companionship ran about 34% in 2018, spiked to 41% early in the pandemic, and settled back to 33% by 2024. Social isolation tracked at 26% in 2018, up to 56% in 2020, then down to 29% (Malani PN et al, JAMA 2025;333(3):254–257). Interestingly, younger and midlife adults may be lonelier than older adults overall, probably from juggling careers, kids, and aging parents—the “sandwich generation.”

CGPR: What connects loneliness to psychiatric illness, particularly depression, in late life?
Dr. Joo: The two share overlapping pathways, which isn’t surprising given how much their symptoms overlap. Biologically, loneliness functions as a chronic stressor. It activates the HPA axis and the sympathetic nervous system, and depression involves the same stress-response dysregulation plus chronic inflammation. An acute, transient bout of loneliness probably isn’t a major risk; it’s the chronic form, lasting months to years, that does physiological damage. Cognitively and behaviorally, loneliness breeds maladaptive social cognitions, like hypervigilance to social threat and interpreting neutral cues as rejection. This drives withdrawal and avoidance, a vicious cycle that also lowers mood. I’d add environment too: Neighborhood resources, socioeconomic factors like poverty and discrimination, and adverse life events all shape both loneliness and depression risk. It’s genuinely bidirectional. Loneliness worsens mood, and low mood drives further withdrawal.

“Avoid interventions that just place a person next to other people without addressing the underlying cognitions or building skills.” - Jin Hui Joo, MD

CGPR: You mentioned the link between loneliness and cognitive decline. What does the evidence show?
Dr. Joo: The correlation is strong, and it’s gotten stronger over the past few years. Several large cohort studies show a meaningfully increased risk of all-cause dementia in people who are lonely (Qiao L et al, Front Hum Neurosci 2022;16:899814). The effect looks similar, maybe even larger, for Alzheimer’s disease specifically. Vascular dementia is a different story; there’s no clear association. For cognitive impairment that hasn’t progressed to dementia, the risk is lower and less consistent. There’s even evidence from intervention trials that modifying loneliness through social engagement can influence that risk (Baptista C et al, Aging Ment Health 2024;28(12):1767–1776). The open question is: Does loneliness increase cognitive impairment? Or does early cognitive impairment—through shame, withdrawal, and difficulty engaging socially—increase loneliness? In practice, I assume it runs both ways, and I don’t wait to resolve the chicken-and-egg question before addressing either one.

CGPR: How does social isolation factor into your suicide risk assessment?
Dr. Joo: Loneliness, social isolation, and depression are each independent risk factors for suicide. In my clinic, depression is usually what brings the patient in, so I assess severity first. The PHQ-9 built into our electronic medical record (EMR) includes a suicide item that helps. But I go beyond the scale and ask probing questions about the patient’s social world, including quality of relationships, sense of belonging, and feeling like a burden, which is common in older adults. I also ask about hopelessness, since depressive symptoms and hopelessness may mediate the relationship between social isolation and suicide risk, and it isn’t always captured on standard scales. If a patient truly has no one—no confidant, no support—that raises my concern substantially and changes how urgently I act.

CGPR: Are there validated instruments you’d recommend for systematically screening for loneliness?
Dr. Joo: None are built into most EMRs yet, but several are used in research and are short enough for clinical use. The UCLA Loneliness Scale (20 items) is the most established; a 3-item version is increasingly used. The de Jong Gierveld scale (6 items) captures both social and emotional loneliness. For objective social network size, the Lubben Social Network Scale is common (Editor’s note: For more on these tools, see the table “Loneliness Screening Tools”). The barrier is that a clinician has to seek these out and administer them independently, as there’s no prompt in the workflow. Sometimes the fastest approach is simply asking directly: “Are you lonely?” Some surveys use exactly that question. You do have to be attentive to underreporting and stigma, but when asked within a good relationship and with a soft approach, a direct question can work well.

From the Expert Q&A
“Loneliness and Cognitive Decline in Older Adults”
by Jin Hui Joo, MD
The Carlat Geriatric Psychiatry Report, Volume 5, Number 7 & 8
October/November/December 2026
www.thecarlatreport.com

CGPR: Certain groups carry higher loneliness risk, like the recently bereaved, caregivers, LGBTQ+ older adults, and patients who are homebound or rurally located. Does your approach change by group?
Dr. Joo: There’s no one-size-fits-all; understanding the individual patient’s context and preferences always comes first. But a few patterns are useful. For the recently bereaved, there’s often an acute window shortly after the loss where a clinician can act by connecting the patient with a bereavement group, or mobilizing family and friends who may not realize the person needs extra support. Caregivers are chronically lonely because their social world narrows around caregiving; I look to support groups through organizations like the Alzheimer’s Association, though some patients want that referral and some don’t. LGBTQ+ older adults often carry strained family relationships and prior negative health care experiences, so I look for LGBTQ-specific aging programs rather than a generic senior center. For homebound and rural patients, technology-delivered interventions can work well, provided they’re adapted to the patient’s actual context rather than transplanted from an urban program.

CGPR: What does the intervention literature show about what actually reduces loneliness?
Dr. Joo: It’s mixed, honestly. Engineering social connection is hard. The clearer pattern is that interventions that build psychological skills, like addressing negative social cognitions or improving social skills, outperform passive contact-based programs that simply connect a person with others, such as most befriending programs (Lasgaard M et al, Am Psychol 2026;81(1):36–52). Those simple contact interventions can help, but they tend to have a smaller effect, and you have to think about whether the engagement is sustainable and resourced. Digital delivery is fine, and the evidence supports it, as long as there’s still an active psychological component behind it rather than passive connection alone. My bottom-line advice is to avoid interventions that just place a person next to other people without addressing the underlying cognitions or building skills.

CGPR: Do you have particular resources you turn to when referring patients?
Dr. Joo: A single generic referral to social work often isn’t enough; older adults usually need more sustained engagement to follow through. I look for community organizations focused specifically on loneliness. Friendship Works, for instance, provides companionship and is well known in my area. I also connect patients to aging service access points (ASAPs), which are organized by geographic catchment area; Ethos serves part of Boston, and similar organizations exist for other areas. AgeStrong is another resource with a range of activities. What I think matters most, more than any specific program, is having an in-person facilitator or connector who can help bridge the referral. Information alone, even when it’s online and available, rarely gets an isolated older adult to engage.

CGPR: What about pets? Do they meaningfully reduce loneliness?
Dr. Joo: The evidence is mixed but leans positive (Kretzler B et al, Soc Psychiatry Psychiatr Epidemiol 2022;57(10):1935–1957). Clinically, I’ve seen it work well. Some patients describe a pet as their primary source of companionship. I had one patient whose only real social connection was her dog; when it died, she spiraled into a serious depression, though she eventually built new friendships with other older women in her community. I’d weigh pet ownership the same way I weigh any intervention. Does it fit the person’s preferences and capacity? Mobility limitations or even mild cognitive impairment can make pet care genuinely difficult, so it’s not a universal recommendation, but for the right patient it can be a meaningful source of connection.

CGPR: How do you structure care when loneliness itself is the primary driver, rather than major depression?
Dr. Joo: It’s a slower process than treating a depressive episode directly. I start with patient education, and I repeat it across visits, not just once. I help the patient understand why this matters for their health. Then it’s collaborative problem-solving: what resources exist, what they want, who’s already in their network. Some patients activate quickly and start seeking out connections themselves. Others need more structured work like incremental goal setting, accountability, and some cognitive-behavioral reframing of the maladaptive social cognitions I mentioned earlier. I’d also add something we don’t usually think of as an intervention, which is simply being a psychiatrist who makes the visit itself feel connected and safe. That relationship, even in a short appointment, is doing real work.

CGPR: Any final thoughts?
Dr. Joo: I think most clinicians understand loneliness matters, but they feel limited in what they can do about it in a short visit. The message I’d want them to hear is that you don’t need a referral to make a difference. Sometimes it’s establishing real connection within your 10 minutes, or even just a phone call. Patients are often grateful just knowing you took the time to call, even briefly, because they know you’re busy. I’d also emphasize the magnitude of the health risk itself. The links to dementia, depression, suicide, and health care utilization are substantial enough that loneliness deserves the same attention we give other major risk factors, not just a passing mention.

CGPR: Thank you for your time, Dr. Joo.