Structured Opportunities for Connection to Improve Aging-related Loneliness

Background It has been suggested that humans have a basic need to belong.1 Social relationships are fundamental to emotional fulfillment, behavioral adjustment and cognitive function. While social integration is critical to human development across the life span, it is particularly important in later life, a time when social relationships often decrease.2 In recent decades, changes and fragmentation in family structure have increased the number of persons living alone, albeit well recognized that 'being alone' is not synonymous with loneliness.3 Loneliness is defined as a subjective, unwelcome feeling of lack or loss of companionship; a mismatch between quantity and quality of social relationships.4 Loneliness encompasses feelings of isolation, disconnectedness and/or not belonging. It often leads to emotional distress and may be associated with detrimental effects on both mental and physical health outcomes.5 Worldwide, prevalence rates of loneliness range from 2% - 61% across both institutional and community settings. Both loneliness and social isolation have been declared as profound threats to health and well-being.6 Recently, a national survey conducted with over 20,000 participants to quantify loneliness and its correlates, estimated that loneliness affects 33% of the US population over the age of 55. While there is some evidence that rates of loneliness decrease with age,7 the negative effect of loneliness on physical health and cognition validates the importance of identifying older adults at risk of loneliness and providing interventions that may reduce attributable poor health outcomes.

One interventional approach, popularized in the 1970's in the UK and Scandinavia, for addressing loneliness and social isolation is 'social prescribing', a model that enables healthcare practitioners to refer patients to non-clinical services and community-based activities to support well-being.12 Social prescribing employs a community-based, person-centered model in which a health coach tailors the social activity 'prescription' to support individuals in better understanding their own needs and taking action to improve their health and well-being. Often key to implementation of the 'prescription' are Link Workers who suggest personalized plans to patients and follow-up with patients to monitor engagement and outcomes.13 The aim of social prescribing is to equip physicians with alternative interventions that can reduce unnecessary prescriptions and referrals, while empowering patients to take greater responsibility for their health by connecting with available community resources to reduce loneliness and increase social relationships. Despite popularity and perceived potential benefits, the scientific rigor and validated evidence regarding social prescribing is sparse and very few studies have included comparison groups to support an inference of direct causation.14,15 In 2020, Roland et. al.16 concluded that current evidence fails to provide sufficient detail to judge either success or value of social prescribing.

Preliminary Data Given reported variation in prevalence of loneliness in age-specific, US populations, we conducted a preliminary cross-sectional study to estimate the prevalence of loneliness in the population of persons aged ≥ 55 years in the Endeavor Health catchment area. Four data collection sites were identified; patients scheduled for medical outpatient visits, laboratory testing or imaging procedures were approached in the respective waiting rooms and invited to complete a brief survey. The only demographic data collected was each patient's age group, i.e. 55-64, 65-74,75-84, ≥85 years. The selected instrument to measure loneliness was the validated, short form of the DeJong Gierveld Loneliness Scale (DeJong scale), designed to measure overall, emotional and social loneliness. Additionally, a single-item screener question,11 'How often do you feel lonely?' was also included to allow for response comparison to the DeJong scale. Endeavor Health Institutional Review Board non-human subjects research determination as a quality improvement project was obtained.

With a range of 28 to 69 completed surveys per site, a sample size of 204 was reached. Based on a total DeJong score of ≥3 (moderate loneliness), the estimated prevalence of loneliness in this sample was 28.7% (range by site: 18%-43%). Using the single-item screener the prevalence of loneliness was 33.8% based on a response of feeling lonely occasionally or more often.

Study Aims This collaborative study, conducted by Mather Institute and Endeavor Health, seeks to develop and test a social prescription model to reduce loneliness and support well-being among older adults receiving primary care at Endeavor Health in Evanston, Illinois. The objective of this pilot study will be to address the feasibility, uptake, sustainability and impact of primary care-initiated social prescribing among community-dwelling older individuals.

Patient Population Eligible subjects will include patients 55 years of age or older attending an outpatient visit at one selected primary care medical site of Endeavor Health who screen positive for loneliness using the one-item Loneliness Screener and consent to study participation.

Study Design This initial study will be a quasi-experimental pre-post design based on 'intention-to-treat' principles. Each participant will serve as their own control. The primary independent variable will be delivery of the 'social prescription', specifically provision of a recommendation and information regarding social activities for older adults available in the community. The primary dependent variable will be change in loneliness as measured by the six-item DeJong Scale.

Recruitment \& Data Collection

Upon registration at the outpatient site, patients 55 years or older will be asked by the office staff to complete a loneliness screen. To assure efficiency and privacy, this question will be provided on a paper form with the 5 possible responses. The medical practice staff will discuss the benefits of social interaction on health, inform these patients of the Social Prescribing study, and inquire about their interest in possible participation. For those interested in learning more about the study, office staff will share patient contact information with the Study Navigator. The Navigator will reach out to the interested patient via telephone to explain the study in more detail, and a copy of the informed consent document will be sent via email or postal mail to the patient. In a second telephone call, further discussion will occur, and the informed consent process will be completed. Following consent, the navigator will ask the patient to complete the:

• 6-item DeJong Loneliness Scale, which will be considered the patient's baseline measure;
• Baseline survey (demographics, overall health status (maybe PHQ-9), social interaction measure (TBD))

The Study Navigator will discuss the alignment of subject interests with available social activities. Emphasis will be placed on in-person, weekly activity offerings in order to optimize socialization and support development of social connections and friendships. However, virtual options will provide opportunities for social interactions particularly for patients with transportation, mobility or social anxiety challenges. For patients with computer access but limited technology skills, Mather staff have developed a written resource and a video for using virtual platforms which has proven to be effective for older adults; this instructional material is supplemented by a 'help desk' contact at Mather Institute. Specific activities to be included in the study selections are programs offered through local community-based organizations and services such as Evanston Public Library, the Evanston Park District/Levy Senior Center and Mather Institute. All suggested activities will be free of charge to enrolled patients.

The Study Navigator will recontact the subject to document their activity plan. Patients will be able to contact the Study Navigator for questions, concerns, and assistance with registration, if needed. Subjects will be asked to submit a bi-weekly log that documents the activities in which they participated.

During follow-up contacts, the patient will be asked to complete the DeJong 6-item loneliness scale and other relevant measures collected at baseline. All surveys can be conducted by phone within approximately 15 minutes. At month 6, patient perceptions, satisfaction and perceived acceptability of activities will also be documented.

Statistical Considerations For the 56 patients categorized as lonely in the prevalence pilot study, the mean DeJong Gierveld Loneliness score (DeJong score) was 4.16 ± 1.12. In order to detect a reduction in scores of 1.0 with an estimated standard deviation of the paired changes of 1.81, a sample of 28 pairs would be required to achieve a power of 80% with a two-sided alpha level of 0.05. The Cohen's D for this expected change is 0.55, considered a moderate effect size. In order to account for sample attrition that may approach 50%, we will plan to recruit a minimum of 50 subjects.

Study Outcomes

This study will be analyzed based on the 'intention to treat' principle, that is all consented patients will be considered in the outcome regardless of participation in the social prescribing intervention. Social Prescription adherence will be assessed across intervention uptake strata, defined below:

• Full Intervention Uptake - participation in at least two in-person activities per week for most (80%) weeks (with absence due to illness, weather or transportation problem waived) throughout the six-month follow-up period;
• No Uptake - no participation in any planned social-prescription activities,
• Low Uptake - less than 2 hours per week for 80% of weeks over the timeframe (3 or 6 months);
• High Uptake - 3 or more hours per week for 80% of weeks in each timeframe.

The primary outcome will be the change in De Jong Loneliness score from baseline to 6-month assessment. If the patient does not complete all follow-up assessments of loneliness, the 'last observation carried forward' method will be used for determining the change in patient's loneliness score from baseline.

Both parametric and non-parametric methods for statistical analysis will be used as appropriate given distribution of variables. Descriptive continuous measures will be presented using means and 95% confidence interval; frequencies and percentages will be used to describe categorical variables. Analysis of the primary outcome will be based on a paired t-test or a Wilcoxon matched-pairs test if necessary. In all cases a two-sided alpha level of <0.05 will be considered statistically significant.

Limitations/Hurdles Given the logistical challenges of implementing the proposed intervention, we will begin with a single-site pilot study to refine workflows, identify unexpected challenges for study participants related to transportation, scheduling, registration processes, etc. Additionally, communications with community-based organizations regarding participant attendance verification need to be finalized.