Demographic Disparities Among Emerging Adults Experiencing Loneliness

Introduction

Loneliness and social isolation have emerged as issues of epidemic proportions in the United States.^1–3 Loneliness, defined as the perceived lack or inadequacy of personal relationships increases risk of health conditions such as cardiovascular disease,⁴ hypertension,⁵ diabetes,⁶ and cognitive decline.⁷ Other studies indicate an association between loneliness and the development of mental health issues, including depression, anxiety, and suicide ideation.^8,9 While discussion about loneliness tends to focus on older adults, a large body of evidence suggests it has been on the rise for decades among emerging adults.¹⁰ Multiple studies suggest this trend was exacerbated by the COVID-19 pandemic, and the increasing feelings of loneliness were associated with increases in anxiety and depression.^2,11,12 The overall prevalence of mental illness comorbidities among young people has increased to an alarming level and represents a significant proportion of the global burden of disease.¹³

Given the broad impacts loneliness can have throughout the life course, and its relationship to other chronic physical and mental health issues, it is critical to study and plan how to intervene to promote social connectedness.^1,13,14 High-quality relationships and broad social networks serve as crucial protective factors and contribute to overall well-being.¹⁴ Researchers acknowledge the role primary care can play in prevention, identifying and addressing loneliness.³ A gap remains, however, in our understanding of the differing degrees to which loneliness may be experienced across demographic groups of emerging adults. The objective of this study was twofold: to explore disparities in rates of loneliness among a sample of emerging adults in primary care, post-pandemic, and to examine which factors might best predict experiences of loneliness. This insight can be useful to aide researchers and clinicians in developing culturally appropriate interventions and designing strategies that promote effective implementation.¹⁵

Methods

Utilizing a patient list from a Washington State–based primary-care system’s electronic health record (EHR), participants were invited to complete a 19-item, cross-sectional survey. Invitations were sent via text message during the recruitment phase of a pilot intervention trial between August and December 2023. The patient list was comprised of emerging adults ages 18–25 who had at least one primary care visit in the previous year documented in the EHR. Unless they opted out, eligible participants were texted two times. This was sufficient to reach the target intervention enrollment for the pilot.

The survey collected validated demographic items—age, race and ethnicity, sexual orientation and gender identity (SOGI), relationship status, and living situation—self-reported healthcare utilization, and the 3-item University of California Los Angeles (UCLA) loneliness score, which measures perceived lack of companionship, feeling left out, and feeling isolated from others. These loneliness scores range from 3–9, and individuals who score a 6 or greater are considered lonely. The scale was initially validated for phone use in 2004,¹⁶ with newer studies validating its use broadly across various populations.^17,18 Descriptive summary statistics were generated, and between-group differences were determined with Fisher’s exact tests for categorical variables and Mann-Whitney U tests for ordinal variables in the R statistical package. A hierarchical logistical regression model produced odds ratios for experienced loneliness on the UCLA loneliness score threshold. Variables for the model were prioritized based on their significance (P-value) in the bivariate analyses.

This study was approved by the University of Washington Institutional Review Board (STUDY00018180). A partial waiver of HIPAA authorization was issued to allow the research team to contact participants for recruitment purposes. Participants received e-consent information just before the survey, and a waiver of documentation of informed consent was granted by the Institutional Review Board for the survey procedures.

Results

243 participants responded to the survey out of 2199 contacted, resulting in a response rate of 11.1%, which is in the range of other EHR-based recruitment studies in primary care.^19–21 The mean age of the sample was 21.9 (SD = 2.33), and 136 individuals (56%) reported experiencing loneliness. A majority of participants were women (69.1%), heterosexual (56.8%), and single (61.7%). Just under half (45.7%) identified as White, and most participants did not live alone (82.3%). A comparison of the survey responders and nonresponders suggests that the sample was somewhat overrepresented by those assigned female at birth, participants who are White, and those at the older end of the age spectrum. Demographic characteristics for the sample and nonrespondents are summarized in Table 1.

In bivariate analyses presented in Table 2, no association with reported loneliness was observed for variables of sex assigned at birth, race/ethnicity, or living alone. Women and nonbinary individuals were significantly lonelier than men (P = 0.03), sexual and gender minority (SGM) individuals (lesbian, gay, bisexual, asexual, transgender, Two-Spirit, queer, intersex, and/or any other gender or sexual identity that is characterized by nonbinary constructs of sexual orientation, gender and/or sex) were lonelier than the heterosexual group (P < 0.01), and single individuals were lonelier than those in a relationship (P = 0.02). Individuals experiencing loneliness were not more likely to visit urgent care, the emergency room, or a primary care clinician, but they were more likely to visit a behavioral health provider (P = 0.01).

In the logistic regression (R² = 0.11), sexual and gender minorities had 2.01 greater odds of loneliness compared to heterosexuals (95% CI = 1.09–3.75; P = 0.03), and individuals in a relationship had 0.37 the odds of feeling lonely compared to single individuals (95% CI = 0.20–0.66; P < 0.01). These results are summarized in Table 3.

Discussion

To our knowledge, this is one of the first studies to explore loneliness rates among emerging adults in primary care with an intersectional lens. Multiple studies have used the UCLA 3-item loneliness scale to explore the overall prevalence of loneliness in primary care.^22–26 Our data suggest that loneliness persists at stubbornly elevated levels for emerging adults, even post-pandemic. Though more extensive inquiry is needed, it appears that this cohort of emerging adults is already more likely to be engaged with behavioral health providers, implying that loneliness may not be prioritized or adequately addressed in existing services. Evidence is emerging to support interventions like loneliness-focused cognitive behavioral therapy and social prescribing.²⁷ That said, more intentional targeting of loneliness in intervention development and robust outcome evaluation are essential to advance a nascent research agenda.²⁸

Our findings mirror other analyses in adult cohorts more broadly that have found SGM individuals twice as likely as non-SGM individuals to report loneliness in the past year.^29,30 While the proportion of participants in this sample may seem surprising, more recent studies suggest that 28% of Generation Z identify as a sexual and gender minority nationwide.³¹ It also makes sense that this proportion would be higher in a progressive, urban center like the Seattle metropolitan region. On the other hand, our findings parallel some evidence that racial and ethnic minority emerging adults do not universally experience higher rates of loneliness.^29,30 One hypothesis about this difference is that racial and ethnic minorities often have parent and family to foster their ethno-racial socialization, which has been shown to improve psychosocial outcomes.³² SGM individuals, on the other hand, are far less likely to receive the corollary from their family and must intentionally seek out this type of social support, often starting in adolescence or emerging adulthood.³³ Our results suggest that SGM populations could benefit from targeted interventions, which may also have impacts on other mental health disparities these communities face. Future work is needed to understand the specific needs and preferences of these SGM emerging adult populations to develop and adapt appropriate services.

With over half of young adults seen in primary care experiencing loneliness, primary-care clinicians could play an essential role in screening for loneliness. This study demonstrates the preliminary feasibility of leveraging the brief, 3-item UCLA loneliness score for this purpose. Primary-care physicians could follow-up in a brief conversation and, if indicated, refer their patients to interventions to promote more social connectedness. In addition to intervention effectiveness studies, important implementation questions remain about how to do this at scale: What dose of an intervention is adequate? Should clinicians work with individuals one-on-one or in groups? Would interventions take place in the clinic or virtually? How can these services be billed? Questions such as these require not only the engagement of patients, but also of primary care clinicians, managers, and administrators.

The nature of this study limits the impact of these findings in several ways. Given the association between loneliness and other commonly treated issues in primary care (eg, depression and anxiety), this cohort may experience more loneliness than the entire population of emerging adults. As an eligibility screening survey for a small pilot trial, the research team excluded variables worth further exploration, including health status and reason for primary care visit. In addition, while it is well-understood that socioeconomic status is related to loneliness, these variables were also excluded given the measurement difficulties inherent to a period of considerable transition.³⁴ Future work with larger sample sizes should be guided by a theoretical framework, and should more comprehensively probe potential risk and protective factors.

The analysis of representativeness suggests there is some response bias toward older, White participants assigned female at birth. This aligns with other studies in this health system, but the results also underline some challenges with comparing EHR demographic data from the nonresponders to the sample. SOGI variables were, in large part, not documented in the EHR, and race/ethnicity and sex questions were asked differently from the EHR in our survey to align with best practices. Unfortunately, studying the experience of sexual and gender minorities is hampered given that these SOGI identity characteristics are often inconsistently defined and collected in the EHR.³⁵ This will be even more difficult with the challenges to SGM data collection from the US federal government in recent months. More intentional stratification of key SOGI variables can help address this in follow-up studies.

Finally, this cross-sectional snapshot does not provide information on whether the loneliness experienced by these emerging adults is chronic or episodic, nor can causation be inferred from the observed association. More sophisticated longitudinal studies of emerging adult cohorts are needed to better understand loneliness and how it changes throughout this important phase of the life course. Intersectional, qualitative analysis could show how loneliness and social support influence identity development for both SGM populations and racial and ethnic minorities, which could, in turn, inform intervention adaptations to address disparities.

Conclusion

The findings from this survey support the assertion that it is feasible to identify those struggling with loneliness and refer individuals to treatments or community-based resources in primary care. Still, robust studies are needed to establish intervention effectiveness and test implementation strategies that will support integrating new services into primary care. Specific attention is needed to address the disparity that SGM individuals face in their experiences of loneliness and help build and reinforce community connectedness. Researchers should build strategic partnerships and secure meaningful involvement from communities experiencing disparities to leverage local knowledge, practices, and networks to address loneliness.¹⁵ It is already clear that loneliness has far-reaching impacts on multiple facets of physical and mental health, and evidence about how to address this issue is emerging. Guided by a philosophy of whole-person care, family physicians are uniquely positioned to address patient loneliness and isolation.^3,36