This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Aging, is properly cited. The complete bibliographic information, a link to the original publication on https://aging.jmir.org, as well as this copyright and license information must be included.
The global population of older adults (aged >60 years) is expected to triple to 2 billion by 2050. Proportionate rises in older adults affected by loneliness and social isolation (or social connectedness) are expected. Rapid deployability and social changes have increased the availability of technological devices, creating new opportunities for older adults.
This study aimed to identify, synthesize, and critically appraise the effectiveness of technology interventions improving social connectedness in older adults by assessing the quality of reviews, common observations, and derivable themes.
Following the guidelines of PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses), 4 databases (PsycINFO, PubMed, Embase, and MEDLINE) were searched between February 2020 and March 2022. We identified reviews with adults aged ≥50 years in community and residential settings, reporting outcomes related to the impact of technologies on social disconnectedness with inclusion criteria based on the population, intervention, context, outcomes, and study schema—review-type articles (systematic, meta-analyses, integrative, and scoping)—and with digital interventions included. Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) was used to measure the strength of outcome recommendations including the risk of bias. The reviews covered 326 primary studies with 79,538 participants. Findings were extracted, synthesized, and organized according to emerging themes.
Overall, 972 publications met the initial search criteria, and 24 met our inclusion criteria. Revised Assessment of Multiple Systematic Reviews was used to assess the quality of the analysis. Eligible reviews (3/24, 12%) were excluded because of their low Revised Assessment of Multiple Systematic Reviews scores (<22). The included reviews were dedicated to information and communications technology (ICT; 11/24, 46%), videoconferencing (4/24, 17%), computer or internet training (3/24, 12%), telecare (2/24, 8%), social networking sites (2/24, 8%), and robotics (2/27, 8%). Although technology was found to improve social connectedness, its effectiveness depended on study design and is improved by shorter durations, longer training times, and the facilitation of existing relationships. ICT and videoconferencing showed the best results, followed by computer training. Social networking sites achieved mixed results. Robotics and augmented reality showed promising results but lacked sufficient data for informed conclusions. The overall quality of the studies based on GRADE was medium low to very low.
Technology interventions can improve social connectedness in older adults. The specific effectiveness rates favor ICT and videoconferencing, but with limited evidence, as indicated by low GRADE ratings. Future intervention and study design guidelines should carefully assess the methodological quality of studies and the overall certainty of specific outcome measures. The lack of randomized controlled trials in underlying primary studies (<28%) and suboptimal methodologies limited our findings. Robotics and augmented or virtual reality warrant further research. Low GRADE scores highlight the need for high-quality research in these areas.
PROSPERO CRD42022363475; https://tinyurl.com/mdd6zds
The use of technology to support older adults against feelings of loneliness and social isolation provides novel opportunities that have grown in the field of aging, as technology demonstrates that information and communications technology (ICT) use and training [
Social isolation and loneliness in older adults have been extensively researched. Many studies showed that the prevalence of these problems increases with age. For example, the prevalence of loneliness among young adults, early to middle–aged adults, and late to middle–aged older adults are 39.7%, 43.3%, and 48.2%, respectively [
Socially disconnected older adults are also vulnerable to a range of health disorders, including infection [
Rapidly deployable technologies, along with socioeconomic changes that have reduced the cost of technology, have increased the accessibility of technological devices, creating new opportunities for older adults [
Several reviews have summarized works on technology interventions for older adults experiencing loneliness [
For a standardized systematic report on these reviews, we must assess the quality of the reviews and find common observations and derivable themes. An umbrella review method can provide a focus for areas where there are competing interventions and amalgamate evidence from multiple quantitative and qualitative reviews [
To bridge this gap in the literature, we aimed to explore the findings and limits of current knowledge on the impact of technology interventions on social disconnectedness in older adults. We also emphasize areas requiring further research. In a comprehensive umbrella review, we synthesized the various categories and types of the used technology interventions, discussed their effectiveness and limitations, and finally explored their potential and need for further research. Finally, we amalgamated all the evidence from the umbrella review and used Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) to make recommendations for interventions targeting social connectedness. This review attempts to answer the following questions:
What technology interventions are used to influence social connectedness in older adults?
How effective are these technology interventions in improving social connectedness in older adults, and what aspects make them effective?
This umbrella review followed the standardized procedures [
The search strategy involved controlled vocabulary searching; phrase searching; and applying Boolean logic, limits, and filters. A comprehensive systematic search of 4 databases (PsycINFO, PubMed, Embase, and MEDLINE) was conducted between February 2020 and March 2022. The reference lists were also examined for additional reviews. The following search terms were used: “ageing,” “aging,” “older adults,” “reviews,” “2000-22,” and synonyms for “social isolation and loneliness,” “social connectedness,” and “technology interventions.” As an example,
Limiters—Published Date: 20 000 101-20 211 231; Language: English; Publication Type: Academic Journal; English Language; Language: English; Year of Publication: 2000-22; Publication Year: 2000-22; Publication Type: Peer Reviewed Journal; English; Language: English; Exclude Dissertations Search modes—Boolean or Phrase Sort by best Match
The inclusion criteria were formulated using the population, intervention, comparison or context, outcomes, and study schema [
Inclusion and exclusion criteria.
PICOSa framework | Inclusion criteria and reasons | Exclusion criteria and reasons |
P—participants | Persons aged >50 years, who are living in community or residential settings with no major neurocognitive impairments | Participants aged <50 years |
I—interventions | Interventions using any form of information and communications technology, smart communication devices, internet-based communication systems, information systems, video games, technological devices, and robots or technological pathways allowing for social interaction. These interventions must be specifically targeted at impacting or improving social connectedness in older adults | Nontechnology interventions, smart devices for home, or telehealth technologies not designed to impact social connectedness (eg, diabetes-measuring devices) |
C—context | Community settings, independent living, and participants in nursing and care homes | Hospital settings, mental and physical illnesses, and disease or illness-specific cases |
O—outcomes | Quantitative or qualitative outcome data or results focusing on social isolation or loneliness or social connectedness | Reviews lacking descriptions of outcome data |
S—study | Review articles of any type using a systematic, qualitative, or quantitative method, including narrative, quantitative, and qualitative comparative studies. Articles must describe a clear intervention and include qualitative and quantitative comparative studies | Reviews with no technology intervention, no clear outcomes, or no systematic review processes. Reviews earlier than 2005 were not included because technology interventions before this time would not be directly comparable with ones of the present day |
aPICOS: population, intervention, comparison or context, outcomes, and study.
The abstracts and titles of all potentially relevant articles were screened. Full texts were then evaluated, and duplicates were removed. Uncertainties were discussed among the research team members to reach a consensus. Relevant data of the included articles were summarized in tables and checked for accuracy by a second investigator (CH).
The data analysis was based on a thematic synthesis with an inductive, iterative process consisting of 3 main stages: (1) free line-by-line review of the results, synthesis tables, and discussion sections of the included papers; (2) organization of themes into related areas; and (3) the identification, development, and refinement of detailed descriptions of factors that impacted the effectiveness of technology interventions [
The methodological qualities of the reviews were assessed using the Revised Assessment of Multiple Systematic Reviews (R-AMSTAR) [
The overall certainty of the evidence was evaluated using the GRADE method, which analyzes the risk of bias (imprecision, inconsistency, indirectness, and publication bias) and assesses the quality of the included evidence, which we used to make recommendations [
The article elimination process is summarized as a flowchart in
Flowchart of the literature search following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses).
Among the 24 selected articles, 3 (12%) articles with R-AMSTAR scores <22 were excluded because they failed a priori systematic review processes (lacked clarity in scope or purpose, had a priori–defined participant population, had unclear outcomes of interest, lacked clarity on interventions, involved nonspecific subgroup analyses, and lacked meaningful hypotheses). The 21 remaining reviews were of moderate quality, with none meeting all of the R-AMSTAR criteria.
Data from the 21 reviews were extracted using a piloted, standardized data extraction form that captures and summarizes findings. As both technology interventions and extracted outcome data were heterogeneous, they were deemed inappropriate for a quantitative synthesis using meta-analytic techniques. Instead, a narrative synthesis summarizing the effectiveness of interventions was implemented. Under the methodological considerations of umbrella reviews, the results were reported descriptively in tabular form (
The 21 selected reviews included 16 (76%) systematic reviews (reviews of evidence on a clearly formulated question and the use of systematic and explicit methods to identify, select, and critically appraise the relevant primary research), 2 (10%) integrative reviews (reviews that summarize past empirical or theoretical literature to provide a comprehensive understanding), 2 (10%) scoping reviews (preliminary assessments of the potential size and scope of the available research literature), and 2 (10%) meta-analyses (statistical analyses combining the results of multiple scientific studies). Most of the reviews covered the beneficial impact of technologies on loneliness, whereas others focused on social isolation, connectedness, and quality of life. General ICT was the most commonly applied intervention technology. The publication period was from 2005 to 2022, but 19 of the selected reviews were published within the last 7 years. Of the 21 reviews, 1 (5%) review focused on assistive technology for communication. Overall, 19% (4/21) of reviews focused on general interventions for social connectedness but examined technologies such as general ICT and videoconferencing, and 10% (2/21) of reviews focused on communication technologies for social connectedness in older adults. In all, 38% (8/21) of reviews investigated the impact of general internet and computer technologies on social isolation and loneliness. Of 21 reviews, 1 (5%) review examined the impact of smart technologies on social connectedness, and another (1/21, 5%) study reported the impact of health promotion technologies on social isolation and loneliness. In all, 10% (2/21) of reviews explored the ability of general ICT to improve the quality of life. Of 21 reviews, 1 (5%) review examined interventions to reduce social isolation and loneliness during the COVID-19 pandemic; 2 (10%) reviews focused on the impact of SNS on loneliness, another (1/21, 5%) examined interventions for preventing loneliness in nursing homes, and another (1/21, 5%) evaluated the benefits of telehealth in alleviating loneliness and increasing medication compliance. Here, telehealth was implemented through video health care professional visits to older adults. The 21 reviews covered a total of 326 underlying primary studies on technology interventions. It is worth pointing out that we were not able to confirm the presence of gray literature or studies that looked at technology interventions in the reviews.
The interventions discussed in the reviews were general ICT (11/21, 52%), videoconferencing (4/21, 19%), computer and internet training (3/21, 14%), telecare (2/21, 10%), SNS (2/21, 10%), and robotics (2/21, 10%). The reviews reported mixed results. Positive effects of ICT on loneliness were the most commonly reported, followed by the positive impacts of ICT on social isolation or connectedness. Reviewing data from the underlying primary studies in the reviews, the most effective intervention mode for social connectedness was identified as general ICT, followed by videoconferencing and robotics (
Effectiveness versus ineffectiveness of different intervention modes on social connectedness, identified in the underlying primary studies of the review papers (n=321).
Study intervention | Effective, n (%) | Ineffective, n (%) |
3D or augmented reality (N=1) | 1 (100) | 0 (0) |
Video gaming (N=4) | 3 (75) | 1 (25) |
Videoconferencing (N=14) | 11 (78) | 3 (22) |
Robotics (N=22) | 16 (73) | 6 (27) |
Telecare (N=34) | 22 (65) | 12 (35) |
SNSa (N=61) | 30 (49) | 31 (51) |
Computer training (N=66) | 39 (59) | 27 (41) |
ICTb (N=119) | 86 (72) | 33 (28) |
aSNS: social networking site.
bICT: information and communications technology.
Among the 21 selected reviews, only Choi et al [
Of the 21 studied reviews, 14 (67%) dealt with general ICT (which was a catch-all term defining a diverse set of technological tools and resources used to transmit, store, create, share, or exchange information), 4 (19%) with videoconferencing, 3 (14%) with computer and internet training, 2 (10%) with telecare, 2 (10%) with robotics, 2 (10%) with SNS, 3 (14%) with gaming, and 1 (5%) with 3D augmented reality (AR). Among the primary studies, general ICTs were the most commonly adopted interventions (with 119 studies), followed by computer training, SNS, telecare, and robotics (
Frequencies of intervention categories in the primary studies (N=321).
Primary studies found in review | Frequency, n (%) |
3D or augmented reality | 1 (0.3) |
Video gaming | 4 (1.2) |
Videoconferencing | 14 (4.4) |
Robotics | 22 (6.9) |
Telecare | 34 (10.6) |
SNSa | 61 (19) |
Computer training | 66 (20.6) |
ICTb | 119 (37.1) |
aSNS: social networking site.
bICT: information and communications technology.
All the reviews reported large numbers and diverse outcome measures of primary studies. Besides constructs of social disconnectedness (loneliness, social support, social contact, number of confidants, social networks, social connectedness scales, social isolation, and social well-being), many studies assessed factors such as quality of life, self-esteem, stress, and depression. Although not directly related to social disconnectedness, these factors may affect or be affected by social disconnectedness and may be useful to include outcome measures alongside social connectedness. A minority of the reviews also reported outcome measures of empowerment.
When analyzing these quantitative primary studies, the reviews commonly applied validated tools, such as the University of California Los Angeles (UCLA) Loneliness Scale (or a modified version) and the De Jong Gierveld Scale [
The definitions and uses of outcome measures differed across the reviews. A total of 62 outcome indicators of social connectedness were used in the primary studies. Most reviews did not report on the lack of intervention effects (including the absence of significance values); moreover, the primary studies adopted a mixture of validated and nonvalidated outcome measures, making such reporting difficult. Consequently, they could not conclude whether the primary studies had validatable statistically significant outcomes.
The social concepts used for determining outcomes varied in range and diversity. In many reviews, the source papers did not define social participation or social isolation but instead evaluated these factors as general or neighboring concepts [
A few of the reviews highlighted that inconsistency and lack of specific definitions hindered the grouping and evaluation of their chosen papers [
Cattan et al [
Ibarra et al [
Gardiner et al [
In conclusion, different definitions and measurements of loneliness, social isolation, and social connectedness have led to diverse findings and wide variations across and within disciplines, defying a coherent picture of the research. Although some of the more recent studies and reviews have addressed this heterogeneity, reliable and succinct findings will remain elusive without further investigations.
Many interventions implemented in the individual papers of the reviews were broadly divisible into group and one-to-one interventions. In general, group interventions were more frequently implemented than one-to-one interventions, although both types were effective [
The imbalance between the group and one-to-one interventions impairs comparisons between the 2 types and conclusions regarding their comparative successes. Nevertheless, some of the reviews pointed out the possible advantages and limitations of these intervention approaches. Poscia et al [
Overall, group interventions appear to improve social disconnectedness, but the insufficient number of one-to-one interventions prevents an objective comparison and firm conclusions of the best interaction type. However, the GRADE assessment of the quality of evidence suggested a very low advantage of group interventions over one-to-one interventions (
Technology interventions that enhance social connectedness include general ICT, video games, robotics, and the Personal Reminder Information Social Management system (a custom-designed experimental SNS for older adults). Less conclusive evidence exists for the beneficial effects of SNS [
Overall, technologies appear to positively affect loneliness, social isolation, and other psychosocial aspects of older adults’ lives. Khosravi et al [
When technologies were intended to strengthen existing connections, their positive impacts on loneliness and social isolation were more consistent [
Choi and Lee [
Individual reviews reported less conclusive outcomes of the overall technology use. The results of Morris et al [
Specifically, the following technology interventions appear to reduce social isolation but lack rigorous statistical support for a positive effect: robotics, telecare, and SNS [
Overall, 86% (18/21) of reviews examined the impact of technology intervention on loneliness. The reviews covered 324 primary studies involving 66,565 participants. Of the 18 reviews, 15 (83%) reported a positive effect of technology on loneliness; the remaining 3 (16%) studies found a 0 or negative effect. From the reviews, it can be concluded that technology interventions exert an overall positive influence on social isolation and loneliness (social disconnectedness), but their effectiveness depends on the design of the study. Longer training times, shorter study durations, and facilitation of existing relationships tended to increase the effectiveness of the intervention. The quality of evidence supporting the effectiveness of technology interventions on social connectedness (GRADE assessment) was moderate to low.
This section explores the findings of general ICT interventions reported in the reviews. General ICT is an umbrella term for generic technology devices, services, applications, and internet platforms [
Many of the reviewed studies found that ICT interventions not only significantly reduce loneliness but also exert a positive impact on other aspects of social isolation, providing social support and connectedness, communication with family and friends, and ICT-accessible information sources [
Damant et al [
Only 2 reviews provided a homogenous meta-analysis. Both reviews reported positive impacts of general ICTs on social disconnectedness. In total, these reviews included 119 primary studies: 86 reporting a positive impact on social isolation or loneliness and 33 reporting unclear results or no impact. The studies agreed that increasing the frequency of general ICT use enhances social connectedness, improving the ease with which older adults can interact and maintain contact with others, thus reinforcing social connections with friends and family. The evidence that frequent ICT use facilitates the creation of new relationships or contacts is much weaker, further supporting, in part, the conclusions of Damant et al [
Together, these results suggest that general ICT can facilitate established connections and might supplement or replace older communication methods. Its role in establishing new connections is uncertain. Our results suggest that when considering ICT interventions (at least for older adults), it is important to distinguish between their ability to maintain relationships, potential ability to deepen relationships, and inability to help create new relationships. The GRADE strength of the ICT category, although only moderate, was the highest among the categories because a large number of primary studies, including RCTs, were reviewed in this category, and there was consensus and clarity on the outcome measures.
Although SNS is a subcategory of ICT, it warrants its own heading because 33% (7/21) of reviews discussed separate finding on SNS. The reviews gave mixed results. Whereas some studies supported the use of SNS in reducing loneliness, a sizable number showed no impact or even an increase in loneliness after SNS use [
These findings may partly depend on the type of SNS, as different types of SNS support different features. For example, Facebook may promote socialization more effectively than YouTube, whereas YouTube may better facilitate explicit knowledge acquisition and information transfer than Facebook. Ibarra et al [
On the downside, SNS use raises several concerns: privacy, lack of perceived usefulness, and possibly demographic factors [
Overall, 61 primary studies examining SNS were found in the reviews: 31 reporting positive impacts of SNS on social isolation and loneliness and 30 reporting unclear or no impacts of SNS. Therefore, the effectiveness of SNS is inconclusive. The results suggest that older users can obtain support, acquire knowledge, and maintain their existing relationships through SNS. In terms of combating social disconnectedness and establishing new relationships, SNSs are less effective and can be detrimental at times. However, the effectiveness of SNS in developing new relationships, fostering and maintaining existing ones, and acquiring knowledge and support has not been explored in depth, and the idiosyncrasies of SNSs must be unraveled in further research. The strength of evidence (GRADE assessment) of the reviews in this category is low because of indirectness, missing information, and publication bias.
Overall, videoconferencing appeared to exert a positive impact on loneliness and social connectedness. The visual aspect of this intervention seemed especially appealing to older adults [
Gardiner et al [
When used in health support, videoconferencing yields mixed results. The intervention often decreases the loneliness and social isolation of residents in care and nursing homes, but a few studies have found no difference from the baseline [
Overall, 14 primary studies in this subcategory were found in the reviews. Of these studies, 11 reported a positive impact on social isolation or loneliness. Owing to reviews such as by Schuster and Hunter [
Among the studied reviews, only Ibarra et al [
In total, 13 reviews evaluated the impact of computer and internet training on various guises. All reviews found a positive impact of these interventions on social connectedness and loneliness [
Mixed results were also obtained for this category. Baker et al [
Unusually, among the reviews, Williams et al [
As some reviews did not differentiate between the impacts of training and subsequent use, any assumptions would be dubious. Morris et al [
Telecare was among the less frequent interventions in the review studies, but when included, it appeared to reduce social isolation and loneliness [
Although none of the authors described the key features of successful telecare interventions, an emergent theme from successful primary studies was a high frequency of contacts. Interventions designed for regular and frequent contact were apparently more successful than interventions delivered on demand (eg, when a resident needed clinical attention). Overall, 34 primary studies in the analyzed reviews covered this category. The impact of telecare on social connectedness was inconclusive, and uncertainty was further increased by the poor reporting of the results. Consequently, the GRADE strength of evidence in this area was very low.
Robotics is a cutting-edge field and was mentioned in only 6 reviews. Some studies found that a pet robot provides the same level of benefit as animal-assisted therapy, which is known to reduce loneliness and social isolation [
Khosravi et al [
Khosravi and Ghapanchi [
Although these reviews indicate that social connectedness can be increased through robotics, this category is still new, and further studies on AI conversational agents and other robotic interventions are required. Therefore, the GRADE strength of evidence in this category is moderate to low.
According to Khosravi et al [
Similar to robotics, 3D environments have been newly introduced as a loneliness-reduction intervention technique and are rarely reported. Khosravi et al [
There were few reviews that examined the usability of technology and its impact on the effectiveness of interventions. Some reviews identified a link between usability and acceptance of technology; more accessible devices were distinctly more likely to be embraced by users than less accessible devices [
However, systematic reviews typically neglect the human-computer interaction components of intervention technology. Moreover, standardized measures of usability (eg, the System Usability Scale) for intervention studies have not been defined [
Overall, the reported studies showed that whether technology can reduce loneliness depends on its usability. An intervention perceived as difficult to use by older adults cannot be effective. This aspect must be further investigated to improve the success of technology interventions.
Owing to a lack of evidence, the GRADE confidence in the effect of usability on the success of intervention technologies is very low.
On the basis of the results,
We have also summarized the key recommendations for study design targeting social isolation, connectedness, and loneliness in
Summary of key recommendations for technology interventions.
Category | Key recommendations | Certainty of evidence |
General ICTa |
Simple technology interventions can be more successful than complex ones. Usability is a potentially important outcome. ICT is not recommended for increasing either the quantity or quality of communications or helping to establish new relationships. It is recommended for maintaining and enhancing existing relationships and access to services (such as health-related services). |
Moderate |
SNSb |
SNS is not recommended as an intervention for loneliness and isolation as SNS use has often been shown to worsen loneliness. SNS is useful in knowledge and support acquisition scenarios, which can themselves reduce loneliness. Research shows that SNSs are generally more successful in these scenarios than in making new connections. Privacy is an important concern among older adults and needs to be considered when designing an intervention. Usability is potentially a very important theme and needs to be factored into the study design. |
Low |
Videoconferencing |
Videoconferencing reduces loneliness by providing social support and improving the existing conditions in health care–type situations. Financial investment (eg, cost of computer hardware) needs to be considered when planning a videoconferencing intervention. |
Moderate low |
MIMc |
MIM is recommended for rapid deployment as it is easy to use, and applications such as WhatsApp additionally allow the sharing of pictures, which can improve social connectedness. MIM can replace email, but designers must be wary because any lack of responses can increase the perception of loneliness. |
Very low |
Computer and internet training |
Longer training periods are recommended with shorter-duration studies (as highlighted above) as they have been the most effective. For reducing loneliness, group-based training is more effective than one-to-one training. The study design should reflect whether the training or use of the intervention causes reduction in loneliness. RCTsd are particularly important in the study design as they determine precise effect sizes. |
Low |
Telecare |
Frequency of contact combined with telecare solution influences the success of an intervention. Interventions designed for regular frequent contact are more successful than interventions delivered on-demand; for example, when a resident needs clinical attention. Videoconferencing groups such as group counseling can help to reduce feelings of anxiety, isolation, and loneliness and provide emotional and social support; however, designers must understand that some participants do not immediately feel at ease with others, especially in a group setting. |
Very low |
Robotics |
Pet robots can provide the same advantages as animal-assisted therapy in reducing loneliness and social isolation; study designs can mimic previous studies in this area. Conversational agents provide companionship through social interaction, enabling older adults to connect with family members and friends (social presence). These agents can be effective and are recommended for intervention studies. RCTs are recommended in the study design of robotic interactions, especially as this area is understudied. |
Moderate low |
Gaming |
Video gaming devices such as Wii, which capture natural physical activities, are recommended as they reduce loneliness and provide better social interactions than typical video games. |
Very low |
3D and augmented reality |
Too few of the existing studies provide robust recommendations, and further longitudinal and cross-sectional RCT studies are needed in this area. |
Very low |
aICT: information and communications technology.
bSNS: social networking site.
cMIM: mobile instant messaging.
dRCT: randomized controlled trial.
Summary of key recommendations for technology interventions.
Category | Key recommendations | Certainty of evidence |
Group vs one-to-one |
Studies should be designed as group-based interventions, as they appear to better facilitate social connectedness than one-to-one interventions. |
Very low |
Effectiveness of technology interventions |
Certain types of technologies (information and communications technology and videoconferencing) are particularly suitable as interventions for social isolation and loneliness. For best results, studies should be designed to strengthen existing bonds, especially the connections between family members (eg, grandchildren). |
Moderate low |
Frequency of use |
Frequency of use is encouraged (greater use increases the effect size). |
Very low |
Training |
Training, especially in the use of technology, is encouraged as it improves the success of the study. |
Very low |
Duration |
Shorter-duration studies are recommended (shorter studies achieve better results than longer-duration studies). |
Low |
Outcome measures |
The impact of intervention is stronger on social isolation than on loneliness, and studies should be designed to look further on how to impact loneliness. Use of standardized measures such as University of California Los Angeles Loneliness Scale and the Lubben Social Network Scale is recommended. |
Low |
Mechanisms |
Mechanisms by which interventions reduce social isolation through the design of studies, including the gaining of social support, engagement in activities of interest, the making of new connections, and search for new information, should be clearly defined at the outset. |
Very low |
Usability |
Intervention studies should adopt standard measures of usability (eg, System Usability Scale) because the adoption of technology by older adults largely depends on learnability and perceived difficulty of use. These barriers often prevent technology from reducing loneliness in older adults. |
Very low |
This umbrella review, as highlighted in the analyzed reviews, found that different studies adopted a vast diversity of outcome measures and nonstandard definitions of loneliness and isolation [
An umbrella review following the JBI methodology [
The designs and qualities of the reviewed primary studies varied widely. Several reviews included RCTs and pilot, qualitative, and quantitative studies. In addition, the studies reviewed by Khosravi et al [
The findings of many underlying primary studies in the reviews were compromised by poor study designs, leading to conflicting information. For example, when reviewing the effects of computer and internet training on loneliness, Chen and Schulz [
The reviewers generally agreed on the effectiveness of group-based interventions. Reviews examining the designs of the reviewed studies noted group-based interventions yielded positive effects on social disconnectedness [
The reviews varied in scope, from assessments of the effectiveness of interventions, such as videoconferencing, to overviews of studies published in the field. The inclusion criteria and quality assessments of the primary studies also differed among the reviews, diminishing confidence in their findings. Our study confirmed a low quality of evidence in this field, whereas improved technology interventions for older adults are increasingly demanded by both policymakers and health professionals. Although the existing guidelines can encourage standardization of systematic reviews, these guidelines were largely ignored by researchers; accordingly, the strength of the reviews is diminished, which in turn led to the quality of evidence GRADE scores also being generally low.
The scope of the reviews varied from a specific focus on the effectiveness of a targeted intervention (such as computer training) to an overview of the published studies in the field. The inclusion criteria for the primary studies and their quality assessment depended on tools used for rating rigor and bias. Such variations cast doubt on the conclusions of these reviews. This review confirms the lack of high-quality evidence in the field and highlights the failure to adhere to the existing guidelines. Standardization of systematic review reporting is expected to strengthen confidence in the review conclusions.
Unlike their younger counterparts, older adults often lack the skills, functional capacity, and accessibility to adopt digital technology [
To improve the quality of results, interventions should be tailored to match the specific needs of older adults, and sufficient training should be provided for using the interventions. This tailoring requires the involvement or participation of participants in training in a variety of formats [
Our umbrella review is one of the few works that have looked at technology interventions for social connectedness for loneliness, following a well-established systematic approach such as the JBI umbrella review method. In examining other works, we came across reviews that focused on interventions generally [
Most of the reviews demonstrated a need for stronger evidence on the effectiveness of technology interventions that reduce loneliness. Weak methodologies have limited the ability of reviews to establish conclusive remarks on their effectiveness [
The present review may also have been biased by accepting only English-language publications. However, many of the shortcomings and limitations of this umbrella review stem from the underlying problems of the primary papers included in the reviews. Among the common shortcomings were small-scale implementations with small sample sizes, low levels of evidence, and short periods of assessment.
Another recurring limitation was the inconsistent definitions of social concepts. Social concepts such as loneliness, social isolation, and social connectedness were formally defined, but the authors did not use these definitions consistently; instead, they were often used interchangeably, inherently confounding measurements of these outcomes. The reviews were generally heterogeneous in focus (eg, addressing loneliness and depression) and discussed various interventions and syntheses of outcomes (eg, meta-analyses, qualitative reviews, and mixed methods). Accordingly, the present review interchanges the terms
The methodological limitations of the reviewed studies impaired the internal validity and usefulness of the reviews for technical and policy decision-making, as highlighted by the reviewers [
The reviewed quantitative studies collected their data with questionnaires using scales developed for the study purpose. The reliability and validity of these nonstandardized scales are difficult to evaluate. Most reviews pointed out the suboptimal methodological quality of studies in this field, particularly the scarcity of RCTs (<28% of studies) and the dominance of quasi-experimental studies, which challenge the delivery of robust conclusions.
Therefore, the results of this review should be interpreted with caution.
Various technology interventions in different formats offer many ways to engage older adults. However, usability was rarely discussed in the reviews and was not assessed as an outcome measure. Although the existing guidelines encourage the standardization of systematic reviews, they have not been followed with the required rigor. Equally, the underlying primary studies of the reviews failed to address causation in a rigorous study design, and their heterogeneity limited their generalizability. It appears that there is a need for more studies on the multidimensional impact of technology on social connectedness, along with the assessment of other measures that may be interacting with technology use (such as educational attainment, psychological resilience, and age-friendliness of environments). Robotics is a relatively new technology that has emerged to be promising, but there are very few studies in this domain. Research on mobile technology interventions for social isolation is also encouraged as mobile phone technology provides opportunities for increasing the uptake of technology interventions targeting loneliness in older adults. Our results on the grading of evidence revealed that the strength of evidence was generally low to very low, indicating that the efficacy of the interventions is unclear and that more rigorous research is needed.
Our review provides insights into strategies to reduce loneliness and isolation for older adults using technology interventions, with implications for future research, policy, and practice. Attention to social connections needs to be incorporated into existing preventative efforts for chronic diseases in older adults. Chronic illnesses develop slowly over decades. Since social connectedness is known to impact multiple mechanistic pathways in both the development and progression of disease, it warrants attention in primary, secondary, and tertiary prevention efforts. Given the lower economic costs of technology interventions for individuals, families, employers, and the broader health care system, we urge health care and health policy professionals to prioritize the investigation of technology interventions for social connections in prevention efforts.
This umbrella review consolidates the state-of-the-art knowledge on the types of technology interventions that influence social connectedness in older adults and their effectiveness. The data were collected from the last 2 decades. Technology purportedly enables long-distance interactions, allowing older adults to become socially connected, obtain support, expand their social networks, and strengthen their existing ties. Some important themes that would improve the effectiveness of technical interventions for older adults emerged from the literature, namely group interventions, short-duration training and study programs, the use of general ICT, and videoconferencing. These implementations are more effective for maintaining existing connections than for building new ones. Certain technologies, such as robotics (including virtual pets), AI-based conversational agents, and MIMs, show promising potential but have been underexplored.
We attempted to determine which technology interventions can effectively improve social connectedness. The following conclusions emerged from our study. Reports on the effectiveness of computer and internet training on loneliness and social isolation provided mixed and inconclusive results. General ICT and internet-mediated communications were shown to reduce loneliness and social isolation in most studies, although the results apparently depend on the frequency of use and the time frame of the study, with shorter studies being more successful than longer ones. ICT interventions help socially isolated older adults through a range of mechanisms, including gaining social support, providing connections to the outside world, introducing new friends, and boosting self-confidence. All of these mechanisms must be studied hand in hand to gain a complete understanding of these processes. Finally, in our GRADE evaluation, most of the evidence was rated as moderate low to very low, reflecting methodological issues, the small number of RCTs, diverse outcome measures and definitions, and mixed results. Such low scores highlight the need for high-quality research in this area.
Search terms used in database.
Revised Assessment of Multiple Systematic Reviews quality ratings.
Characteristics of the included reviews.
Grading of Recommendations, Assessment, Development, and Evaluations—certainty of evidence.
artificial intelligence
augmented reality
Grading of Recommendations, Assessment, Development, and Evaluations
information and communications technology
Joanna Briggs Institute
mobile instant messaging
Preferred Reporting Items for Systematic Reviews and Meta-Analyses
Revised Assessment of Multiple Systematic Reviews
randomized controlled trial
social networking site
University of California Los Angeles
None declared.