Introduction

Loneliness is a distressing psychological experience caused by a personal perception of discrepancies between one’s desired social connections and one’s actual level of connectedness.1–3 A survey was conducted by gathering data from a diverse group of 75 891 individuals aged 50 years or older, representing different countries. Based on the findings, the prevalence of loneliness was 25.32% in the USA, 17.55% in England and ranged from 5.12% to 20.15% across European countries.4 5 The older adult population globally faces considerable risks from feelings of loneliness. Research indicates that being alone can negatively affect a person’s mental and physical health, leading to conditions such as high blood pressure, reduced cognitive function, Alzheimer’s disease, heart problems, obesity and increased feelings of anxiety and depression.6 7 Furthermore, loneliness is associated with a 25%–30% higher mortality rate over a span of 7 years based on a meta-analysis.6 Therefore, early intervention is crucial to address the pressing issue of loneliness among older adults, ultimately enhancing their quality of life and overall well-being.8

Loneliness is a negative subjective experience often confused with social isolation in many studies. Social isolation refers to an individual’s position in a social network, which is an objective lack of social connection or social engagement.9 Socially isolated persons are not necessarily lonely, and lonely persons are not necessarily socially isolated in an objective sense.10 Where a person lands on the subjective continuum depends on their relationship standards.11 Although there is a close relationship between loneliness and social isolation, the concept and risk factors are different. Therefore, different intervention strategies must be developed to address these two issues. This study focuses on loneliness because of its high incidence among older adults in the community.

Loneliness can be addressed through a variety of interventions, such as social support, exercise programmes, psychological support, volunteering, skill development interventions and health management.12–16 Despite the lack of consensus, investigating strategies aimed at diminishing loneliness among older adults is vital.11 17 Social support implies that individuals can rely on and trust those around them and receive substantial support and assistance from them.18 19 Such support encompasses various forms of assistance, including financial aid, practical help and guided counselling techniques, all of which contribute to improving the coping mechanisms of older individuals in the face of diverse challenges and adversities.20 Furthermore, the augmentation of social support among older adults can be achieved through relationship feedback and positive social interactions, ultimately mitigating their feelings of loneliness.21 Diverse entities, such as family, friends, community members and volunteers, are able to facilitate these interactions. By establishing strong connections with older individuals, these individuals can provide companionship, entertainment and assistance, thereby facilitating their integration into the communal fabric.22 Therefore, social support plays an important role in reducing loneliness among older adults living in a community.

Barrera and Ainlay have proposed a rational classification system of social support, which includes material assistance, behavioural assistance, intimate interaction, directive coaching, feedback and positive social interaction.23 Except for the single intervention in the classification system, multicomponent and multifactorial interventions have also been explored by researchers. Multicomponent intervention means that everyone receives the same fixed combination of interventions.23–25 Multifactorial intervention means that people receive multiple interventions, although the combination of these interventions is tailored to the individual based on an individual assessment.11 20 26 27

Wild et al have verified that positive social interaction can alleviate loneliness after 6 months of intervention.28 Other studies have also shown that providing behavioural assistance to older adults significantly improves social interactions and decreases loneliness.29–32 Furthermore, many studies have combined various types of social support to alleviate loneliness, for example, a multicomponent intervention improved social self-efficacy by decreasing loneliness by combining behavioural assistance, guided support and positive social interaction.33–35 Although some trials have reported significant results, many studies have methodological limitations and large heterogeneity. The present research primarily focuses on the evaluation of various interventions aimed at mitigating loneliness among older adults. These interventions encompass exercise programmes, social support, technological interventions and animal-assisted interventions.29 36–44 To date, no systematic review has been undertaken regarding the effectiveness of different kinds of social support (eg, material assistance, behavioural assistance, intimate interaction, directive coaching, feedback, positive social interactions) on loneliness among older adults in a community. The existence of this knowledge gap not only impedes the efficient resolution of loneliness among older adults but also presents obstacles to the long-term viability of social support interventions. Therefore, a robust systematic review of the effectiveness of different types of social support on loneliness is necessary.

Gaining insight into the significance of various forms of social assistance for diverse older individuals will enable us to effectively address their requirements and offer suitable aid and resources. Additionally, studying the effects of different types of social support on loneliness in older people can promote social integration and improve their quality of life and well-being.34 This will also offer guidance to community-based organisations and governmental departments.45 This study aims to investigate the relative effectiveness of different types of social support interventions or interventions that include a social support component in addressing loneliness in older adults. The research question is as follows: What is the comparative effectiveness of different types of social support for loneliness among older adults in a community?

Methods and analysis

We followed the reporting guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis for Protocols 2015 (PRISMA-P).46 Online supplemental file 1 contains the comprehensive checklist for PRISMA. The protocol for this MA has been duly registered in PROSPERO under the registration number CRD42020226523. The research commenced on 1 October 2022 and is anticipated to conclude by 31 October 2023.

Study selection criteria

Types of studies to be included: To be eligible for inclusion in this systematic review, studies must satisfy the following requirements: (a) language is either English or Chinese; (b) randomised controlled trials (RCTs), including but not limited to parallel, cross-over and cluster randomised control trials; (c) validated measures of loneliness must be used for baseline assessment, or follow-up assessment, or both; (d) interventions for the control group can be no intervention, usual care or placebo/sham intervention; and (e) published types are peer-reviewed journal articles, book chapters, conference proceedings or other sources. The exclusion criteria are (a) meta-analyses, reviews, case reports, qualitative studies, comments, editorial letters, notes or conference abstracts; (b) no full text and (c) studies reporting duplicated data.

Types of participants: The inclusion criteria for participants in this study are (a) individuals who are aged 60 years or older (depending on what the original authors reported) and (b) who live in a community. If the mean or median age is not explicitly stated, studies may still be considered for inclusion if they present an age range that aligns with the predetermined cut-off established by our review team.

Types of interventions: Six types of single interventions will be included in this study,47 namely, (a) material assistance: providing financial and other tangible support, for example, giving financial assistance; (b) behavioural assistance: sharing tasks and providing help through practical actions, for example, providing practical labour support; (c) intimate interaction: traditional non-directive counselling behaviours, such as listening, comforting and understanding; (d) directive coaching: providing advice, information or guidance to help individuals improve their performance; (e) feedback: providing feedback on an individual’s behaviour, thoughts or emotions and (f) positive social interactions: engaging in social interactions for fun and relaxation. Multicomponent and multifactorial interventions will also be included in the analysis. Interventions are classified into single interventions with one component; multiple interventions with more than one component, but the intervention was the same for all participants; and multifactorial interventions with more than one component, and the intervention was modified for every participant personally. We compared the interventions for alleviating loneliness with no additional treatment (routine care) or with other types of intervention.

Types of outcomes: Loneliness can be described as the discomfort felt when the way one interacts with others does not align with one’s anticipated outcomes.3 The University of California, Los Angeles（UCLA） Loneliness Scale and the De Jong Grieved Loneliness Scale were validated instruments to measure loneliness and will be used in this study.48 49

Search strategy

We will search bibliographic electronic databases starting from their creation to September 2022. A combination of keywords will be used to search EMBASE, PubMed, Cochrane Library, PsycINFO, Scopus, Web of Science, China National Knowledge Infrastructure Library, China Science and Technology Journal Database (Weipu), WanFang Database, and China Biology Medicine Disc. Box 1 displays terms from MeSH or Emtree terminology, including social support, psychosocial support systems, loneliness, social exclusion, social exclusion, ostracism, social alienation, aged, older adults, agenarians and centenarian, using Boolean operators. The search will encompass grey literature sourced from the databases, irrespective of publication type. Additionally, supplementary grey literature will be ascertained through targeted exploration of health organisations’ websites, agencies catering to older adults and pertinent advocacy groups such as the Campaign for Loneliness.

Box 1 Draft literature search for PubMed

#1 (“Social Support”[(Mesh])) OR “Psychosocial Support Systems”[(Mesh])

#2 social support* [Title/Abstract] OR Social Care [Title/Abstract] OR Psychosocial Support System* [Title/Abstract] OR Psychosocial Support* [Title/Abstract] OR Psychological Support System* [Title/Abstract]

#3 #1 OR #2

#4 “Loneliness”[(Mesh])

#5 Social Isolation [Title/Abstract] OR Loneliness [Title/Abstract] OR Social Exclusion [Title/Abstract] OR Social Exclusions [Title/Abstract] OR Ostracism [Title/Abstract] OR Social Alienation [Title/Abstract]

#6 #4 OR #5

#7 “Aged, 80 and over”[(Mesh]) OR “Aged”[(Mesh])

#8 ag* [Title/Abstract] OR Elderly [Title/Abstract] OR Oldest Old [Title/Abstract] OR agenarians [Title/Abstract] OR Nonagenarian [Title/Abstract] OR Octogenarians [Title/Abstract] OR Octogenarian [Title/Abstract] OR Centenarians [Title/Abstract] OR Centenarian [Title/Abstract] OR old people [Title/Abstract] OR old person* [Title/Abstract] OR old adult* [Title/Abstract] older person* [Title/Abstract]

#9 #7 OR #8

#10 #3 AND #6 AND #9

Data extraction

Two reviewers will independently screen the titles and/or abstracts found in the electronic databases using the search strategy to assess potential eligibility. We will obtain the complete texts of articles considered potentially eligible by each reviewer. Duplicate articles will be automatically filtered out by Endnote V.20, and the two reviewers will further remove duplicates at the title and/or abstract screening stage before assessing the eligibility of each full article. Any conflict will be resolved by engaging in conversation or, if necessary, with the assistance of a third reviewer serving as an arbitrator. Two reviewers will independently conduct data extraction. To ensure uniformity of the extraction form, two reviewers will randomly select studies to test the pilot form. The extracted information will include the following: (a) details about the study (such as author, year, country, duration of follow-up and total number of individuals assigned to the intervention and control groups); (b) participant attributes (eg, number of participants, eligibility criteria, age and gender) and (c) details about the intervention or exposure (eg, type, frequency, length, content, comparison with a control group, delivery format and information about the provider). The type of intervention will be categorised as single, multicomponent or multifactorial. The components of each type of intervention will be briefly described; (d) details of the instruments, including assessment tools and validation information and (e) outcomes involving mean or average alteration; SD; details of SD calculation, such as SE or CIs; pertinent effect size and p values.

Risk of bias assessment

The JBI Critical Appraisal Checklist for RCTs will be used to assess the methodological quality of the included trials. The JBI scale evaluates internal validity using 13 items. The evaluator shall make a judgement of ‘yes’, ‘no’, ‘unclear’ or ‘inapplicable’ for each evaluation item, and finally decide whether to include, exclude or obtain further information after group discussion.

Statistical analysis

Statistical synthesis of multiple studies: First, the results will be descriptively condensed, presenting details about the characteristics of the research, patient attributes, potential for prejudice and frequency of outcomes among the RCTs that were incorporated. To measure the effect sizes, we will use common quantitative indicators, including the risk ratio, OR and standardised mean difference (SMD). To determine the effect size for each study, it is recommended to employ a standardised method such as computing the mean difference between the intervention and control groups and dividing it by the combined SD. In instances where studies solely present follow-up means, effect sizes will be computed using suitable formulas, such as Cohen’s d or Hedges’ g, considering the data’s specific characteristics.50–52 The inclusion of Hedges correction for effect sizes will be considered if indications of potential bias arising from small sample sizes are observed. Forest plots will be employed to display a statistical overview of relative risks, ORs, weighted mean differences and SMDs, along with a statistical amalgamation of the meta-analysis. A statistical method, such as a random-effects or fixed-effects model, will be used to combine the effect sizes of each study. This generates an overall estimate of the treatment effect, along with a CI, which indicates the precision of the estimate.53

Analysis of subgroups: Subgroup analyses will be conducted based on different participant demographic characteristics, types of social support interventions and assessment tools, if possible. The outcomes of these analyses will be contingent on the inclusion of a sufficient number of studies and the availability of suitable outcomes.

Assessment of bias and heterogeneity: Two researchers will independently evaluate the research quality and potential bias in the selected studies using the approved instruments. Any discrepancies or disputes between the reviewers will be resolved through discussion, consensus or arbitration, with the involvement of a third reviewer. We will evaluate the diversity of study results/impact magnitudes among studies by employing Cochran’s Q test with a significance level of ρ<0.05.44 54 To evaluate the level of diversity, we will employ the I2 metric, which quantifies the proportion of variability in the magnitude of impact across research studies. A heterogeneity value of 25% or less will be regarded as indicating low heterogeneity, while a value of 75% will be considered indicative of high heterogeneity or variance among studies.55 56 If there is significant diversity (I2>50%),57 we will investigate the potential for performing stratified meta-analyses and random-effect meta-regression to ascertain whether the methodological approach or the clinical characteristics of the studies included in the meta-analysis were associated with the magnitude of the effect.58 We will implement either the random-effects or fixed-effects model and present the magnitude of random effects along with the level of heterogeneity observed in the model. To evaluate publication bias in the studies included in our systematic review, we will use two different methods: a graphical approach using funnel plots and a statistical approach using the Egger’s test.59

Patient and public involvement

As this study is a systematic review, there will be no direct participation from patients or the general population. However, we will solicit feedback from significant stakeholder groups (such as older adults in the community and relevant associations of service providers) to determine the most efficient methods for disseminating the results of our study.

Ethics and dissemination

Ethical permission is not necessary for this research.

The results of this extensive examination have the potential to improve the welfare of older adults and those responsible for their care. If this comprehensive analysis establishes that interventions targeting social assistance have a significant impact on the feeling of loneliness in later life, it will provide older individuals and their caretakers, along with other interested parties, such as healthcare practitioners, with a promising approach to alleviate loneliness in old age and enhance the overall welfare of older adults. Consequently, the findings of this research will be communicated to older individuals and their caretakers, community members and healthcare practitioners to guarantee a beneficial influence on the health outcomes of older individuals.

The results of this extensive investigation will be disseminated to the public through publications in reputable academic journals, presentations at conferences and communication with policy-makers, healthcare professionals and administrators at local, national and international levels.