Patient Perspectives on Loneliness Interventions in Primary Care: A Qualitative Study

Abstract

There is a loneliness epidemic in the US. Embedding interventions within primary care may facilitate access to effective treatments for loneliness. This study characterized experiences of loneliness, strategies used to counteract loneliness, and perspectives on loneliness interventions among primary care patients. Semi-structured interviews were completed by 17 adults (M age = 41.9 years; SD = 18.9) who indicated social isolation on routine screens administered within primary care. Interviews were transcribed and rapid qualitative analysis was performed. Summaries were created and patterns within the data were grouped into themes. Participants spoke of intrapersonal, relational, and situational/contextual causes of loneliness. All participants reported being aware of strategies to reduce loneliness, with the majority having success using one or more strategies. Opinions were mixed regarding loneliness interventions within primary care, with half of the participants expressing positive impressions towards addressing loneliness in this setting and others voicing skepticism regarding whether primary care was the best setting to address loneliness, even among participants with positive overall impressions. Only two participants reported having spoken with their primary care provider about loneliness. Despite the mixed perspectives, participants indicated that strategies to improve coping, psychological states (e.g., amotivation), and enhance their social engagement would improve their loneliness. In conclusion, perspectives on loneliness interventions within primary care were mixed, suggesting that patient psychoeducation and alliance-building may be needed prior to delivering loneliness interventions in this setting. Given the varied identified causes and needs, multifaceted interventions may be needed to ameliorate the personal and public health burden of loneliness.

Introduction

In 2023, US Surgeon General Vivek Murthy declared an “epidemic of loneliness”, citing research indicating rising rates of loneliness (Bruce et al., 2019; Shovestul et al., 2020), with more recent estimates suggesting that 20% of American adults experience daily loneliness (Gallup, 2024). Defined as a state of emotional distress from lacking desired interpersonal relationships (Heinrich & Gullone, 2006), loneliness has increased in prevalence since the start of the COVID-19 pandemic (Ernst et al., 2022; Hwang et al., 2020), with some studies suggesting elevations were sustained after the pandemic (Poštuvan et al., 2024; Rebechi et al., 2024; Schütz & Bilz, 2024). Loneliness and social isolation are associated with adverse health outcomes, including all-cause mortality, functional decline, depression, anxiety, insomnia, suicidal ideation, and suicidal behaviors (Allan et al., 2021; Fassberg et al., 2012; Hawkley & Cacioppo, 2003; Holt-Lunstad et al., 2015; Musich et al., 2015). In this context, the National Academies of Engineering, Sciences, and Medicine advocates for greater health system involvement and innovative research on treatments for loneliness (National Academies of Engineering, Sciences, and Medicine, 2020), echoing a growing inclination within healthcare systems to intervene directly into social determinants of health (Gottlieb et al., 2019).

Primary care has the potential to play an important role in addressing the public health burden of loneliness. Individuals who lack or avoid social relationships typically still seek medical care (National Academies of Engineering, Sciences, and Medicine, 2020). Primary care practices also increasingly identify individuals experiencing loneliness and social isolation as part of routine screening for social determinants of health (DeVoe et al., 2016; Gard et al., 2019; Pinto & Bloch, 2017) and mental health services are becoming more widespread within this setting (Rotenstein et al., 2023). While loneliness is considered a social determinant within the biopsychosocial model of overall health, there are also biological, social, and psychological determinants of loneliness itself. Relevant drivers of loneliness include chronic medical conditions (e.g., by limiting mobility), family and community dynamics (e.g., divorce, absence of community interest groups), and maladaptive cognitions (e.g., poor self-esteem, social anxiety; Cheung et al., 2019; Heu et al., 2021; Megalakaki & Kokou-Kpolou, 2022; Savikko et al., 2005). Thus, whereas healthcare providers have been found to be supportive of loneliness-based interventions (Stefanidou et al., 2021), they have also described feeling powerless in addressing the complex nature of the problem (Jovicic & McPherson, 2020).

Integrated behavioral health providers may be well suited to deliver loneliness interventions within primary care. Integrated care models include collaborative care management (CoCM) and primary care behavioral health (PCBH). In these models, a behavioral health professional such as a clinical social worker or psychologist has regular contact with primary care patients to coordinate medical care or deliver psychosocial interventions for common mental health or health behavior concerns (Archer et al., 2012; Reiter et al., 2018). Given the effectiveness of these models for addressing other mental conditions and their increasing implementation and financial sustainability (Reist et al., 2022), brief loneliness interventions could be delivered as a component of these models.

A meta-analysis of randomized clinical trials of loneliness interventions identified cognitive restructuring as the most effective approach; however, none of the related trials were conducted in primary care (Masi et al., 2011). Prior studies of loneliness interventions in primary care include a qualitative meta-synthesis of social prescribing (Liebmann et al., 2022), a widely-used intervention for loneliness/social isolation originating in the United Kingdom. Social prescribing involves linking patients with activities and support services within the community (Bickerdike et al., 2017; Drinkwater et al., 2019; Morse et al., 2022), for the purpose of improving mental health, mental well-being, general health, or quality of life (Cooper et al., 2022). Social prescribing has been found to contribute to enhanced well-being and an ongoing desire to build social connections (Liebmann et al., 2022). However, patients reported challenges regarding a disconnect between their interests and the prescribed activities, misalignment between the program goal of fostering connections with patient goals that might not include forming new relationships, and the rapid pace at which community participation was recommended (Liebmann et al., 2022).

Given the absence of randomized clinical trial evidence for loneliness interventions in primary care and qualitative data describing potential limitations to social prescribing, we sought to identify patient perspectives towards addressing loneliness in US primary care practices, including a possible focus on cognitive determinants of loneliness and the involvement of integrated behavioral health providers. We examined loneliness characteristics among primary care patients, strategies they used to counteract loneliness, and perspectives on loneliness intervention components. We obtained perspectives on social prescribing and cognitive strategies (Kharicha et al., 2018), as well as their views on, and suggestions for, other components that could be included in loneliness-based interventions (e.g., employing a volunteer to attend social events with patients).

Methods

Study Design

Due to limited information within the literature, we chose an exploratory qualitative design to gain a deeper understanding of factors that contribute to primary care patients’ loneliness and how they feel about their providers addressing it through social prescribing and cognitive strategies. This study was approved by the Institutional Review Board at the University of Michigan.

Sampling and Recruitment

Participants were recruited from two Michigan Medicine primary care health centers. As part of routine practice, all patients at these offices are administered questions assessing social isolation (“How often do you feel isolated from others?”) and suicidal ideation (SI; “How often have you been bothered by thoughts that you would be better off dead, or of hurting yourself?”). Participants were initially eligible for this study if they were at least 18 years old, indicated isolation within the past six months, and endorsed SI within the past year. Suicidal ideation was an inclusion criterion because this work was conducted to inform a parent study whose aim is to reduce loneliness as a suicide prevention strategy. Patients meeting these criteria were mailed or emailed a letter with opt-out instructions, and then contacted via telephone by a member of the study team. After four attempts to contact, it was considered a passive decline and no other attempts were made. Patients who were reached were asked to complete the 3-item UCLA Loneliness Scale (Hughes et al., 2004), with individuals scoring ≥ 5 being eligible for participation, as well as the 6-item Callahan Cognitive Impairment Screener (Callahan et al., 2002), with individuals scoring ≤ 2 being eligible for participation. After completing the first 15 interviews, the SI criterion was removed to purposively sample a more racially diverse sample including those without SI. All participants provided informed consent prior to study enrollment. Participants were compensated $40 for their time.

Data Collection

Eligible participants completed a semi-structured interview (see the Online Supplement) that was created based on published literature (Heu et al., 2021; Liebmann et al., 2022; Masi et al., 2011), content expertise of the study team, and goals of the project (M length = 49.8 minutes, SD = 8.8, range = 32-62 minutes). Open-ended questions focused on how participants experienced loneliness, contributing factors, strategies patients had previously used to reduce loneliness, and perspectives on primary care-based loneliness intervention approaches. Interviews were conducted via Zoom by a trained member of the research staff (JJ). Interviews were audio recorded and transcribed verbatim by members of the research team.

Data Analysis

Data were analyzed using a rapid analytic approach (Hamilton, 2013; Vindrola-Padros & Johnson, 2020). A summary template was developed using domains from the interview guide (see the Online Supplement). Each interview transcript was reviewed independently by two members of the research team (SG, JJ, AL, PP, KS) to create detailed summaries for each participant. The team members then met to compare and combine summaries during a series of team meetings. This process continued until all transcripts were reviewed. The summaries were then transferred, by domain, to a master matrix to compare responses across participants and facilitate theme identification. Themes were identified by systematically reviewing and grouping related domains to form broader themes that reflected the core concepts emerging from the data and informed the research question. The complete list of themes and examples are presented in the Online Supplement. Additionally, findings were identified that will be used to refine the parent study’s interventions for a subsequent randomized controlled trial (clinicaltrials.gov ID: NCT06656975).

Results

Sample Characteristics

Participants were 17 adults (see Table 1 for sample characteristics). Fifteen were recruited from a larger sample of 96 patients who responded positively to both the isolation item from 01/01/2023-06/30/2023 and the SI item from 06/30/2023-07/01/2023. After eliminating the SI criterion, 2 more participants were recruited from an additional sample of 10 patients who responded positive to the isolation item only during the same time period. Including these two participants did not substantively change the findings. Therefore, data from all 17 participants are reported herein.

Table 1.

Characteristics of the study sample (N = 17).

Characteristic	Mean (N)	Standard deviation (%)
Age	41.8	17.2
Sex
Female	12	70.6%
Male	5	29.4%
Race
White	14	82.4%
Black	2	11.8%
Asian	1	5.9%
Ethnicity
Hispanic	2	11.8%
Non-Hispanic	15	88.2%
UCLA Loneliness Scale	6.8	1.4
PHQ-9 Item 9	1.3	1.0
Callahan Cognitive Screener	0.1	0.2

Note. PHQ-9 = Patient Health Questionnaire-9.

Loneliness Characteristics

Participants spoke of intrapersonal, relational, and situational/contextual causes of loneliness. Intrapersonal causes of loneliness are internal factors within an individual that contribute to feelings of isolation. The most common intrapersonal causes were low self-esteem (n = 8; 47.1%), reactions to unmet expectations (n = 6; 35.3%), social anxiety (n = 5; 29.4%), body image (n = 5; 29.4%), trauma history (n = 4; 23.5%), health issues (n = 3; 17.6%), and social skills deficits (n = 3; 17.6%). Relational causes of loneliness are factors within a person's relationships that contribute to feelings of loneliness for the participants. Common relational causes were unfulfilling relationships (n = 9; 52.9%), death of a loved one (n = 8; 47.1%), other relationship loss (not due to mortality; n = 6; 35.3%), and challenges meeting others’ expectations (n = 3; 17.6%). Situational causes of loneliness are feelings of isolation that arise from life circumstances or changes. The most common situational/contextual causes were spending time alone (n = 7; 41.2%), the COVID-19 pandemic (n = 6; 35.3%), and life transitions (e.g., moving out of state, starting a new job; n = 4; 23.5%). Of the 17 participants, 11 (64.7%) endorsed a history of SI. Eight of those participants (72.7%) identified loneliness as a contributor to their SI.

“[Loneliness] was one of the reasons why I've learned that I have to get out of the house on my days off. I couldn't stand being by myself because I'd be lost in thought. I would go into dark places where I start overthinking like, “There's no purpose of me being here” and “Everybody would be better if I wasn’t here.”

(ID#11845)

Participants also identified multiple protective factors against loneliness that were situational in nature (i.e., outside of the individual’s control). Impromptu social interactions (n = 4; 23.5%), such as spontaneous outreach, was a commonly identified protective factor. Meaningful conversations, vs. surface level conversations (n = 4; 23.5%), and spending time in physical proximity to others were also identified as being protective (n = 4; 23.5%).

"I would definitely say just any social interaction that goes beyond a certain threshold of intimacy [is beneficial]."

(ID#12487)

Strategies to Reduce Loneliness

All participants reported being aware of strategies to reduce loneliness, with the majority having success using one or more strategies (n = 13, 76.5%). Distraction (n = 11; 64.7%), contacting close family members and/or friends (n = 11; 64.7%), joining groups (n = 9; 52.9%), psychotherapy (n = 5; 29.4%), exercise (n = 3; 17.6%), meeting new people (n = 3; 17.6%), volunteering (n = 3; 17.6%), and spending time outside of the home (n = 3; 17.6%) were the most described strategies. Participants described trying multiple strategies to determine the most effective approach. Only 4 participants (23.5%) stated that their chosen strategies yielded little to no success and/or they have not been able to implement a strategy.

“Well, it's hard. I do all the things. I reach out to friends or family. I'll call my mom; I'll go out on a run. I'll go do something physical. I'll try and distract myself in a variety of ways. I also will just try working on the thing that's stressing me out and see if I work on it and get it done, then maybe that will alleviate the stress and therefore also the loneliness. I haven't found anything that works really well, and definitely not something that works every time.”

(ID#10470)

Despite using a variety of strategies, only two participants (11.8%) reported discussing loneliness with a primary care provider, both of whom had previously received psychotherapy for loneliness or a related issue (e.g., social anxiety). However, 11 participants (64.7%) had discussed other mental health issues with their primary care provider, such as depression, anxiety, and grief.

“I have been diagnosed with depression. They put me on medication. So, I was taking this medication, for I don't know, roughly 4 years, and I feel like it just wasn't doing anything for either depression or loneliness, and so I just quit taking it… But no health providers ever mentioned anything about loneliness.”

(ID#11484)

Perspectives on Loneliness Intervention Components in Primary Care

Opinions were mixed regarding loneliness treatment within the primary care setting. Nine participants (52.9%) expressed positive impressions towards addressing loneliness within primary care. However, there was skepticism regarding whether primary care was the best setting for loneliness interventions endorsed by eight participants (47.1%), even among participants who had positive impressions of such interventions.

“Yeah, that would be helpful because I struggle greatly with like routine and schedule. I feel like I’m all over the place when in reality, I'm literally just like right here in my bed. It would probably help with some of the overwhelming feelings for sure.”

(ID#11976)

“I'll be honest, the first thing I think of when I think of loneliness is not my primary care provider. When I heard about the study, I was like, “oh, that's interesting, I never thought of going to my doctor about this”.”

(ID#10470)

Participants reported several barriers that had prevented them from discussing loneliness with a primary care provider. Seven participants (41.2%) expressed that they do not perceive loneliness to be a medical condition or to be addressed in a healthcare setting. Other barriers reported by participants included perceptions that primary care providers would not be adequately trained to address mental health and/or loneliness-related concerns, beliefs that the only approach to help loneliness is “helping yourself”, and feeling judged by staff when conducting loneliness screenings.

“You know, [talking with my primary care team about loneliness] is not something I've ever actually considered… It's not like my doctor's gonna hook me up with a new best friend or something.”

(ID#11484)

“They're not trained to [address loneliness], and that's a problem… I mean, I've been in some family practices now where they, if you come in for your annual physical, they'll throw a piece of paper at you to ask you mental health questions. And they never really delved into those questions. It's like somebody told them that they should be doing this, so they do it.”

(ID#11447)

“After I fill out the questionnaire and they look over it, I feel like they're judging me… I notice more so pity or more so them looking at me like they're concerned… And that normally puts me in a weird funk after going to the doctor.”

(ID#11845)

When asked about content areas most important for loneliness interventions, participants described a range of topics, with the most common being coping skills (n = 8; 47.1%), ways to get involved with groups/activities (n = 4; 23.5%), time management (n = 2; 11.8%), social skills (n = 2; 11.8%), motivation enhancement (n = 2; 11.8%), accountability (n = 2; 11.8%), and self-esteem (n = 2; 11.8%). Participants stated that help with logistics and accountability (n = 10; 58.8%), planning (n = 7; 41.2%), identifying ways to decrease the effort required to engage in activities (e.g., via warm handoffs, having a volunteer facilitate connections; n = 5; 29.4%), and scheduled check-ins (n = 2; 11.8%) would be helpful to incorporate within loneliness interventions.

Most participants (n = 11; 64.7%) were interested in online connections to reduce loneliness. Online groups were perceived to facilitate connection with individuals with shared interests or experiences.

“[Online connections] significantly make me less lonely because they're in tons of different time zones so there's always someone around to talk to. Also, because I found them through my interests instead of just getting along with them well in person, I can talk to them about the things I'm super interested in.”

(ID#11777)

However, other participants raised concerns with online relationships (n = 7; 41.2%). They described negative experiences and expressed difficulty trusting individuals they met online.

"Those didn't really work just because one, I've got a lot of things on my plate already, so it's hard to find the time to do the first option. And then the second option is like, not everybody who's looking for friends on the Internet is necessarily a person you want to be your friend."

(ID#12487)

Participants identified barriers that would interfere with their ability to engage in loneliness interventions. The most common barriers were logistical (n = 15; 88.2%), motivation (n = 8; 47.1%), lack of information (n = 5; 29.4%), and anxiety or low self-esteem (n = 5; 29.4%). One participant also expressed concerns about their primary care physician being aware of their loneliness, preferring to talk with behavioral health care professionals instead, stating that this concern might prevent them from engaging in a loneliness intervention in this setting.

“There's a lot of things I don't want my general physician to know that is really none of her business. It's not necessarily affecting my physical health, you know, but like my mental side.”

(ID#11609)

When asked about ways to overcome those barriers, participants identified making concrete plans with a provider (n = 10; 58.8%), engaging a volunteer to attend social gatherings (n = 8; 47.1%), involving friends and family (n = 8; 47.1%), and discussing strategies for managing anxiety (n = 5; 29.4%) as potentially facilitating their engagement with loneliness intervention content. However, other participants expressed concerns about receiving assistance due to potential implications regarding their inability to complete tasks independently (n = 2; 11.8%).

“I think [making plans with a provider] would be helpful to most people, but I'm not sure it'd be - I’d have to look and see because I'm in sort of a weird spot where I know what my problems are. I know kind of what I need to do. It's just a matter of actually doing them that is the problem with me.”

(ID#11316)

"I think [having a volunteer’s assistance from primary care] would either go one of two ways. I feel like it would either work really well, because I would need that facilitator, or it would make me feel more isolated or alone, I suppose, because of the fact that I needed someone else to act as a middleman. It would almost feel like I'm failing at doing it myself, so I need someone else to step in and do it for me."

(ID#12487)

Discussion

Causes for loneliness were heterogeneous in this sample, as consistent with prior research (Achterbergh et al., 2020; Cohen-Mansfield et al., 2016; Coll-Planas et al., 2021; Hemberg et al., 2022; Heu et al., 2021). In keeping with a cross-cultural examination of loneliness antecedents (Heu et al., 2021), participants identified intrapersonal, relational, and situational factors spanning biopsychosocial domains (e.g., gastrointestinal distress, physical disability, mental health issues, death of loved ones). Relative to prior work, body image concerns and trauma history were more consistently endorsed as a contributor to loneliness in this sample. The COVID-19 pandemic also emerged as a consistent situational cause of loneliness, as consistent with meta-analytic findings suggesting increased loneliness during the COVID-19 pandemic (Ernst et al., 2022). Given that loneliness causes were heterogeneous and multi-dimensional, intervention approaches may benefit from exploring each patient’s individual contributors to loneliness and utilizing flexible techniques capable of addressing diverse causes.

This qualitative study revealed mixed perspectives regarding loneliness interventions in primary care. Despite most participants having tried a range of strategies on their own, only two had discussed loneliness with the primary care provider. Although most participants thought integrating loneliness interventions into primary care could be beneficial, others expressed concerns. Specifically, participants stated that they did not perceive loneliness to be a medical issue and questioned whether primary care offices had the resources or training to adequately address loneliness. This finding is somewhat consistent with prior work showing that older adults in England did not perceive primary care-based loneliness services to be desirable (Kharicha et al., 2017). However, perspectives were somewhat more favorable in this sample of community-dwelling adults in the United States. Taken together, findings suggest that although many primary care patients think loneliness interventions within primary care could be beneficial, others perceive that loneliness would be better addressed in other settings. To overcome patient skepticism, loneliness interventions may require up-front alliance-building, education regarding the benefits of reducing loneliness, and information regarding why the service is provided in primary care, including education on integrated behavioral health models.

Despite the mixed perspectives regarding loneliness interventions in primary care, participants suggested multiple strategies to include in loneliness interventions. Broadly, participants endorsed strategies to improve psychological states (e.g., poor coping, social skills, anxiety, low self-esteem) and to enhance their engagement (e.g., ways to facilitate involvement in activities) with the optimal strategy being context dependent (i.e., whether the person is alone vs. around others; Kharicha et al., 2018, 2021; Vasileiou et al., 2019). These findings therefore support loneliness interventions that involve coping with internal processes, such as Cognitive Behavioral Therapy (Ashrafioun et al., 2024), as well as interventions focused on increasing social engagement, such as Social Prescribing (Drinkwater et al., 2019; Morse et al., 2022). Mental health professionals working in primary care, such as behavioral care managers or behavioral health consultants, are accustomed to treating patients using cognitive and behavioral approaches (Milano et al., 2022; Turner, 2017). Therefore, these findings could be used to develop brief loneliness interventions for delivery within CoCM or PCBH models, overcoming limitations of solely behavioral-based approaches (Liebmann et al., 2022), such as the widely-adopted Social Prescribing.

Perspectives on involving support persons, such as family members or friends, in loneliness treatment was largely favorable. However, perspectives were mixed regarding whether volunteers from the primary care clinic would facilitate an individual’s engagement with community activities. In general, participants thought that volunteers could be helpful for some patients but should be an optional component.

Prior studies have identified mixed perspectives on internet-based relationships among adults experiencing loneliness (Dworschak et al., 2024; Kharicha et al., 2021). In this sample, most participants were interested in online connections. Online groups were perceived to expand the availability of persons with shared interests or life experiences, thus enabling deeper connections. However, a couple of participants raised concerns about the possibility of patients falling prey to exploitation while engaging in online communication. Others had previous negative experiences attempting to form connections with individuals they met online. Therefore, psychoeducation on online boundary setting and communication may be warranted for loneliness interventions that provide online options.

This study is limited in that the sample was relatively small and from a single regional health system; a larger and more diverse sample may have yielded different themes or expressed different perspectives. Researcher biases may have influenced the interpretation of qualitative findings; however, the risk was mitigated by the diverse training backgrounds (i.e., two psychologists, a social worker, a physician, a research assistant) of members of the analysis team. This study focused on perspectives treating loneliness within primary care, which may differ from perspectives on other mental health conditions within this setting. Future work should systematically identify additional factors that contribute to comfort discussing loneliness in this setting, given that a high proportion of mental health care is already delivered within primary care settings (Rotenstein et al., 2023; Sporinova et al., 2019). Candidate constructs that could be relevant for loneliness include prior experiences with mental health treatment, stigma, perceived inefficacy of treatment, lack of psychoeducation, and the background/approach of healthcare professionals, given the association between these constructs and comfort discussing mental health generally in primary care (Bleyel et al., 2020; Corry & Leavey, 2017; Kyanko et al., 2022; O’Loughlin et al., 2022; Tunks et al., 2023). Despite these limitations, this study provided data on how patients perceive a primary care-based loneliness intervention, particularly with respect to anticipating interest in and concerns about such an intervention.

Conclusion

Primary care patient perspectives on loneliness interventions were mixed, with most participants expressing that such interventions would be beneficial, whereas others expressing skepticism about whether primary care was the optimal setting to treat loneliness, suggesting a need for education and alliance-building as a foundation for future intervention development and implementation. Regardless of the setting, participants identified several intrapersonal contributors to loneliness potentially addressable though cognitive behavioral or social prescribing interventions. Given the varied causes and needs identified by participants experiencing loneliness, multifaceted interventions within and outside the healthcare system are likely needed to ameliorate the personal and public health burden of loneliness.

Public Significance Statement.

There is an epidemic of loneliness, with about half of American adults experiencing loneliness. This study assessed the perspectives of primary care patients on intervention components that could be used to develop new treatments for loneliness in primary care settings. Participants identified specific features that would be helpful for reducing their sense of loneliness, such as strategies to improve coping and facilitating social engagement.