SPECIAL ARTICLE

Social connection as a critical factor for mental and physical health:​

evidence, trends, challenges, and future implications
Julianne Holt-­Lunstad
Departments of Psychology and Neuroscience, Brigham Young University, Provo, UT, USA

Rising concerns about social isolation and loneliness globally have highlighted the need for a greater understanding of their mental and physical health
implications. Robust evidence documents social connection factors as independent predictors of mental and physical health, with some of the strongest
evidence on mortality. Although most data are observational, evidence points to directionality of effects, plausible pathways, and in some cases a causal
link between social connection and later health outcomes. Societal trends across several indicators reveal increasing rates of those who lack social connec-
tion, and a significant portion of the population reporting loneliness. The scientific study on social isolation and loneliness has substantially extended over
the past two decades, particularly since 2020;​however, its relevance to health and mortality remains underappreciated by the public. Despite the breadth
of evidence, several challenges remain, including the need for a common language to reconcile the diverse relevant terms across scientific disciplines, consis-
tent multi-­factorial measurement to assess risk, and effective solutions to prevent and mitigate risk. The urgency for future health is underscored by the po-
­ten­tial­ly longer-­term consequences of the COVID-­19 pandemic, and the role of digital technologies in societal shifts, that could contribute to further declines
in social, mental and physical health. To reverse these trends and meet these challenges, recommendations are offered to more comprehensively address gaps
in our understanding, and to foster social connection and address social isolation and loneliness.

Key words:​Social connection, social isolation, loneliness, mental health, physical health, mortality, public health

(World Psychiatry 2024;​23:​312–332)

In a joint statement published in January 2024, the govern­ In the same year, the South Korean government took a tangible
ments of the US, Japan, Morocco, Sweden, Kenya and Chile high­ step, offering monthly stipends to encourage young socially iso­
lighted “the importance of social connection to the health and lated individuals to reintegrate into society12. Outside govern­
well-­being of individuals, communities and societies”1. This came ments, the World Health Organization (WHO) launched in 2023
at the heels of the COVID-­19 pandemic, a more than three-­year a Commission on Social Connection, a three-­year effort to raise
period in which the global population had to isolate, practice “so­ global awareness and mobilize support in this area13.
cial distancing” and, in many cases, was homebound, all factors These efforts have been prompted by data documenting recent
contributing to reduced social contact. However, while that global increases in social isolation and loneliness, and decreases in social
health crisis helped raise awareness of the importance of this issue, connection globally4. Factors including modernization in society,
scientific evidence was already documenting the significant men­ economic disparities, the introduction of digital technologies,​
tal and physical health implications of declining social connection. shifts in civic engagement, growing political divides and radicali­
Social connection is widely acknowledged to be a fundamental zation, and others, have been examined as potential contributors
human need2,3, linked to higher well-­being, safety, resilience and to this decline in social connection. Whether this is a social reces­
prosperity, and to longer lifespan4. Across social species, research sion, a loneliness epidemic, or a public health crisis, it is clearly a
demonstrates that social connection is one of the strongest pre­ pressing issue.
dictors of survival, both early and later in life, through adaptive This is a critical moment to act and bridge the gaps in our collec­
behavioral and biological mechanisms5,6. The availability and di­ tive knowledge to mitigate adverse outcomes. However, there are
versity of social relationships, interactions and networks are crit­ several challenges to be addressed. Over the years, the relevance of
ical for health and well-­being4,7,8. Therefore, it is imperative to un- social connection to our health has emerged in various disciplines,
­­derstand how new trends involving social connection relate to shifts leading to a complex and potentially confusing evidence base. This
in important societal outcomes such as mental disorders and phys­ calls for a common language to be established. However, in the
ical diseases. process, we risk oversimplifying the issue and falling short of an
Rising global concerns about a “loneliness epidemic” in pub­ adequate response. With increasing public and governmental at­
lic discourse have been accompanied by increased academic re­ tention, this is a critical time to take stock of the strengths and gaps
search and heightened engagement among communities, insti­ in the existing evidence, the challenges to be faced, and the impli­
tutions and governments. These concerns are being reflected in cations for the future.
national and international responses to this “epidemic”. In 2018,
the UK appointed a Minister of Loneliness9, establishing a national
strategy and awareness campaign. Japan followed by appointing a SOCIAL CONNECTION AND MENTAL HEALTH
Minister of Loneliness in 202110. Beginning in 2018, the European
Union has produced several reports on loneliness11. In 2023, the There is a robust evidence base linking social connection to men­
US Surgeon General issued an Advisory and a framework for a tal health outcomes. Social connection plays a vital role in prevent­
national strategy on “our epidemic of loneliness and isolation”4. ing mental health problems, maintaining good men­tal health,

312 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
and aiding in the recovery from both moderate and severe mental and reducing depressive symptoms, and better treatments for de­
health conditions, while isolation and loneliness have been associ­ pression are likely to reduce loneliness.
ated with poorer mental health. Most of this evidence regards de­ The link between social connection and depression has also
pression, with fewer studies considering other mental disorders. been examined among patients in medical settings, suggesting po­
Although most available data are observational and cannot tential spillover effects on other clinical conditions. For example,
demonstrate causality, there are longitudinal studies that provide low social support had a significant positive association with ante­
more robust evidence to indicate directionality, and recent evi­ natal depression, which contributes significantly to maternal physi­
dence using Mendelian randomization to establish causal rela­ cal health21. In a review, 83% of studies found that pregnant women
tionships14. In some cases, associations appear to be bidirectional, with low social support had greater depressive symptoms16.
meaning that there is evidence to suggest that social isolation and The links between social connection and mental health are also
loneliness increase the risk for poorer mental health, as well as evi­ relevant within occupational settings. The strain on employees a­-
dence that poorer mental health increases the risk for isolation and cross sectors, particularly those hit hardest during the COVID-­19
loneliness15. pandemic – such as health care providers, educators, and other
“essential employees” – has brought greater attention to burnout
and other mental health concerns. A meta-­analysis of studies in
Depression health care workers found that a lack of social support significant­
ly contributed to higher risk for acute stress disorder, burnout, an­x­
There is a strong positive association of social isolation and lone­ iety, depression, and post-­traumatic stress disorder22.
liness with depression from youth to older adulthood. Further, high­
er social connectedness is protective towards depressive symptoms
and disorders16. Cognitive health
When looking at adults of all ages, 18 years and older, data from
the US National Health Interview Survey examined the impact of Several meta-­analyses consistently show that stronger social
living alone and the availability of social and emotional support connection – including social networks (e.g., number of social con­­
on depression17. Adults living alone reported significantly higher tacts, frequency of interaction, marital status, living arrangement)
depression than those living with others, and this difference held and social engagement (e.g., attending social groups;​visiting fam­
across several sociodemographic factors. Adults never or rarely re­ ily, friends and neighbors;​engaging in voluntary or paid work, par­
ceiving social and emotional support were twice as likely to report ticipation in cultural or leisure activities) – is associated with better
depression, but adults living alone were still more likely to report cognitive function, but the evidence is less consistent for percep­
depression even compared to adults living with others who did tions of loneliness.
not receive social and emotional support17. For example, a meta-­analysis including over 2.3 million partic­
Importantly, longitudinal evidence suggests that social isola­ ipants showed that living alone, having a smaller social network,
tion and loneliness likely cause or worsen depression over time. having a low frequency of social contact, and having poor social
For example, a systematic review of 32 longitudinal studies from support were risk factors for dementia, while loneliness was not23.
the general population examined whether subjective feelings of However, other meta-­analyses did find that greater loneliness was
loneliness predicted the onset of a new diagnosis of depression18. significantly associated with incident dementia24,25. Conversely,
Studies followed participants from six months to 16 years, with greater social engagement, including a greater number of social
an average follow-­up of 3.5 years. The odds of developing new memberships, number of social contacts, and more social par­
depression in adults were more than double among those who ticipation, may be protective, as these were associated with lower
reported often feeling lonely compared to those rarely or never dementia risk23,26.
feeling lonely. While there were more studies among older adults,
the findings were consistent among younger age groups, includ­
ing university students and new mothers. SOCIAL CONNECTION AND PHYSICAL HEALTH
Using two large datasets – the Psychiatric Genomics Consor­
tium meta-­analysis of major depression (N=142,646)19, and the Robust evidence links social connection, isolation and lone­
Million Veteran Program (N=250,215)20 – to apply a two-­sample liness to an increased incidence of several physical diseases and
Mendelian randomization design, loneliness appeared to cause to earlier death. The strength of this evidence has been acknowl­
incident major depression and depressive symptoms14. These edged in multiple National Academy of Science, Engineering, and
analyses were then reversed using loneliness outcome data from Medicine (NASEM) consensus study reports15,27, scientific state­
the UK biobank. Remarkably, data demonstrated that loneli­ ments by professional associations such as the American Heart
ness causally predicts major depression, but the reverse is also Association28, and the US Surgeon General Advisory issued in
true, with major depression causally predicting loneliness14. This 20234. The evidence can be found in several meta-­analyses and
suggests that loneliness is both a cause and a consequence of ma­ systematic reviews that document the overall effects on physical
jor depression;​thus, public health strategies to reduce loneliness morbidity29-­31, and on disease-­related as well as all-­cause mortal­
may potentially be effective in preventing the onset of depression ity32-­43. There are also meta-­analyses on clinical outcomes such as

World Psychiatry 23:3 - October 2024 313

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
response to vaccine44. Diabetic outcomes may be due to better self-­care among those
This body of evidence led a NASEM consensus study report to who are more socially connected. For example, in a meta-­analysis
conclude that “social isolation is a major public health concern”15. of 28 studies, social support was significantly associated with
This is noteworthy, since the report was published before the better self-­care, particularly glucose monitoring, and was strong­
COVID-­19 pandemic and there has been a significant volume of er among those with type 2 than type 1 diabetes50. Improving dia­
research on this topic from 2020 onward. betic outcomes via social connection can have cascading public
health implications, given that diabetes mellitus often leads to
other health outcomes, including heart disease, kidney failure,
Physical morbidity blindness, amputation and dementia.
There is also evidence to suggest that poor social connection is
There is a rich and growing body of evidence across a variety of associated with worse outcomes among those who are already ill.
physical health outcomes, including major health indicators such For example, heart failure patients who self-­reported high levels
as cardiovascular diseases, stroke and diabetes mellitus. of loneliness had a 68% increased risk of hospitalization, a 57%
Cardiovascular diseases are the leading cause of death globally, higher risk of emergency hospital visits, and a 26% increased risk
accounting for roughly one third of all deaths;​therefore, factors of outpatient visits compared with patients reporting low loneli­
that increase or decrease this risk can have a major global health ness51. In a meta-­analysis of 13 studies on heart failure patients,
impact45. Dozens of studies have found that social isolation and poor social connection was associated with a 55% greater risk of
loneliness significantly influence the risk of cardiovascular and ce­ hospital readmission52. This was consistent across both objective
rebrovascular morbidities15,29. and perceived social isolation, living alone, lack of social support,
The culmination of this evidence resulted in a statement pub­ and poor social network. These data suggest that improving social
lished by the American Heart Association in 2022, acknowledging connection among those who are sick can improve medical out­
this risk from objective and perceived social isolation28. According comes.
to this review of the evidence, there is a clear link of social isola­
tion and loneliness with risk for coronary heart disease and stroke.
Among the evidence, a synthesis of data across 16 independent Mortality
longitudinal studies demonstrates that poor social relationships
(social isolation, poor social support, loneliness) were associated Several reviews of the evidence, including a NASEM scien­
with a 29% increase in the risk of incident coronary heart disease tific consensus study, have concluded that some of the strongest
and a 32% increase in the risk of stroke29. These findings were con­ evidence linking social connection, isolation and loneliness to
sistent across genders. health-­relevant outcomes is that concerning mortality15. Large
Low social connection and loneliness have also been associat­ population-­based epidemiological studies have tracked initial­ly
ed with a greater risk for hypertension. Indeed, data from the Na­ healthy populations over time, for years and often decades, doc­­
tional Social Life, Health and Aging Project suggest that the impact umenting that those who are more socially connected live long­
of social isolation on risk for hypertension exceeds that of clinical er35,38,41,42, while those who experience social deficits (isolation,
factors such as diabetes mellitus, pointing to a “causal role of social loneliness, living alone, poor-­quality relationships) are more likely
connections in reducing hypertension” in older age46. to die earlier, regardless of the cause of death33,36,37,39,40,43. Although
Diabetes mellitus is a leading source of disability, lost produc­ social isolation has been implicated as a risk factor for death by sui­
tivity, mortality, and lower quality of life, affecting nearly half a bil­ cide53, most meta-­analyses on mortality exclude suicide as a cause
lion people worldwide, with a significant global economic burden of death.
on individuals, health care systems, and countries47. Studies have Based on meta-­analytic data, one estimate suggests that the as­
repeatedly shown that social connection (e.g., family support and sociation between social connection and survival may be as high
involvement) can positively influence the management and over­ as 50%42, while isolation is associated with 32% and loneliness
all health of individuals with type 1 and 2 diabetes. Large popula­ with 14% increased risk for earlier death33. While estimates vary to
tion studies also demonstrate the influence of social connection some extent, they may be conservative, given that many reviews
on the incidence of type 2 diabetes. For example, people with and meta-­analyses often exclude studies that focus specifically on
smaller social networks were more likely to have been recently di­ deaths due to unnatural causes such as unintended injuries, vio­
agnosed with type 2 diabetes, to have previously been diagnosed lence or suicide. While there are more studies and stronger effects
with this condition, and to have diabetic complications48,49. on cardiovascular-­related deaths (e.g., myocardial infarction,
However, gender differences have been found along different stroke) and cancer-­related deaths (e.g., leukemia, lymphomas,
indicators of social connection. Low social participation was link­ breast cancer)41, more research is still needed on these, in addition
ed to pre-­diabetes and complications among women but not men, to other disease-­related causes of death.
while living alone increased the likelihood of previously diag­ Over the years, the number of studies, the rigor of methodology,
nosed type 2 diabetes and its complications in men but not in wom­ and the size of samples have all increased substantially, replicating
en48,49. These findings were independent of glycemic control, quali­ the finding that social connection decreases the risk of premature
ty of life, and cardiac risk factors. mortality and providing stronger confidence in this evidence. For

314 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
example, longitudinal data from the UK Biobank regard nearly half (loneliness) or the objective (isolation) aspect matter most? The
a million people, reducing the likelihood of random error54. These English Longitudinal Study of Ageing, conducted in nearly 5,400
data demonstrate that social isolation significantly increases risk adults over the age of 50, followed for an average of 5.4 years, found
for earlier all-­cause mortality, overall and consistently across sub­ that loneliness was associated with an increased risk for cardio­
groups (i.e., males and females, young and older, health and un­ vascular disease (coronary heart disease and stroke), but did not
healthy, various ethnicities), even after adjusting for a robust set of find the same outcomes for social isolation56. On the other hand, the
lifestyle, socioeconomic, biological, and health risk factors55. UK Biobank, a large-­scale research effort collecting data on nearly
Several meta-­analyses and systematic reviews have document­ half a million people, followed for an average of 7.1 years, found
ed similar findings across different ways of examining the issue, that both isolation and loneliness were associated with an in­
including social relationships broadly, social networks, social con­ creased risk of acute myocardial infarction and stroke57. However,
tact frequency, marital/partnership status, marriage dissolution, the impact of social isolation remained significant after adjust­ing
social isolation, loneliness, and living alone32,43. While the mag­ for other risk factors, while the effect of loneliness was attenuated.
nitude of the effect varies to some extent across studies and de­ Both isolation and loneliness were significant predictors of cardio­
pending on which aspect of social connection is being examined, vascular outcomes; however, the relative importance seemed to be
the evidence points to the same general conclusion:​indicators of stronger for objective isolation.
greater social connection are associated with reduced risk, while Research is increasingly looking at the relative importance of
indicators of social deficits are associated with greater risk for pre­ isolation and loneliness, and considering multiple outcomes si­
mature mortality. multaneously. Growing evidence suggests that loneliness has a
stronger impact on mental health outcomes, while isolation has
a stronger impact on physical health outcomes31,58. For example,
THE RELATIVE INFLUENCE OF ISOLATION a large national prospective study, examining the effects of social
AND LONELINESS isolation and loneliness on 32 physical, behavioral and mental
health outcomes, demonstrated that both were independent pre­
When predicting the risk of future disease, does the subjective dictors, but isolation had a stronger effect on mortality while lone­

Figure 1 Simplified model of possible direct and indirect, directional and bidirectional, and potentially cyclical pathways by which social con­
nection is associated with morbidity and mortality

World Psychiatry 23:3 - October 2024 315

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
liness had a stronger effect on mental health outcomes58. ness). The relevant mechanisms are both plausible and supported
by evidence. Poorer physical health can also contribute to both
greater isolation or loneliness and poorer mental health, creating
PATHWAYS AMONG SOCIAL AND HEALTH complex bidirectional associations.
FACTORS These associations may also be cyclical. Poor social connec­
tion can dysregulate our physiology and behavior in ways that put
The evidence on the protective effects of being socially con­ us at risk of developing poorer health. Poorer health may reduce
nected and the risk associated with social disconnection is often people’s willingness, ability or access to connect socially, resulting
studied and discussed separately. However, these conditions in­ in greater isolation, which in turn impedes their ability to manage
tersect in meaningful ways. This includes direct and indirect, bidi­ their illness, leading to worse prognoses.
rectional and cyclical, as well as additive and multiplicative effects. We also need to understand the complexity of the factors con­
Much of the evidence to date has focused on establishing the di­ tained within the model and how that can potentially result in ad­
rect and indirect effects. A simplified model of these pathways is ditive and multiplicative effects. For example, co-­occurring defi­
illustrated in Figure 1. cits of social connection (e.g., living alone, small social network,
Among the growing body of literature on social connection and low levels of social support, and loneliness) may contribute to bio­
health, studies often focus on establishing a directional influence logical, psychological and behavioral pathways, potentially mag­
of some aspects of social connection (represented as B in the fig­ nifying the risk to health. Furthermore, like many behavioral and
ure) on various health or mortality outcomes (represented as D). lifestyle risk factors that can influence multiple chronic health con­
Further work has examined the mechanisms (represented as C) ditions, the evidence similarly points to poor social connectedness
that provide plausible psychological, biological and behavioral ex­ leading to greater risk (and greater social connectedness reducing
planations for these effects59-­64. risk) for multiple health conditions. Thus, it is probable that poor
Several reviews and meta-­analyses document the evidence social connection can increase the risk of comorbidities among
point­ing to psychological pathways such as perceived stress60,65;​ physical, mental and cognitive health conditions. This is consistent
be­­havioral pathways such as sleep66, physical activity and smok­ with data from the Health and Retirement Study which demon­
ing67;​and biological factors such as inflammation68. Studies fur­ strate that social isolation was significantly associated with 32 indi­
ther examine the risk factors (represented as A in the figure) that cators of physical, behavioral and psychological health outcomes58.
can potentially compromise one’s social connection.
The associations of primary interest in research have been be­
tween B and D, with B treated as the predictor variable and D as STRENGTHS AND GAPS IN THE EVIDENCE
the outcome variable. Subsequent research has treated C as me­
diator variables and A as risk factors. However, associations are The scientific evidence base for the health relevance of social
likely far more complex. connection is robust, with consistent findings emerging over the
Many factors examined as plausible pathways (represented as past few decades, reinforced across several scientific disciplines
C in Figure 1) are also notable outcomes, often treated as clinical (e.g., epidemiology, neuroscience, sociology, medicine, psychol­
endpoints. For example, social isolation and loneliness have been ogy), and using a variety of methodological approaches (e.g., lon­
linked to poorer nutritional/eating behaviors considered harmful gitudinal, cross-­sectional, experimental).
to health, including low fruit and vegetable intake, and poorer over­­ Several meta-­analyses and systematic reviews document con­
all diet quality69. There is also evidence that those who are socially verging evidence linking social connection, isolation and/or lone­
isolated are less likely to get preventive screenings, such as a mam- liness to psychological, cognitive and physical health. Together,
mo­gram70. these include hundreds of studies with millions of participants.
A synthesis of 122 empirical studies examined the effects of While most data are observational, there is substantial prospective
differences in social connection on medical adherence71. Higher evidence to establish the temporality of effects42, and evidence to
social connectedness, particularly social support, has been linked support a gradient or dose-­response effect46.
to better medical adherence across several physical diseases, es­ There is also experimental evidence in humans and animals to
pecially hypertension72,73 and type 2 diabetes mellitus74,75. Simi­ support a potential causal association. For example, experimen­
larly, other factors such as stress can be both an endpoint and a tally housing animals in isolation versus socially leads to poorer
mechanism by which social connection influences morbidity and outcomes, including the development of tumors, stroke, impaired
mortality. healing, and death5. Animal models have also validated poten­
The directionality, or bidirectionality, of these associations may tial molecular, cellular, immunological and behavioral effects for
be relevant. While those linked to mortality are unidirectional (i.e., human social disconnection3. These experimental studies further
end-­of-­life stops any further influence), nearly all other pathways map causal associations between social perception, neural activ­
may be bidirectional. While there is robust evidence of directional ity, immunological function, and health3.
effects (i.e., those less socially connected are more likely to develop In humans, randomized controlled trials (RCTs) experimental­
poorer health conditions), the reverse can also be true (i.e., poorer ly test the potential benefits of social interventions. For instance,
health also predicts a greater risk for social isolation and loneli­ a meta-­analysis of 106 RCTs found that patients who received

316 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
psychosocial support in addition to treatment as usual had 20% it results in many health outcomes as well (e.g., cardiovascular
increased odds of survival than those in the control group who re­ disease, cancer). Indeed, Bradford Hill and proponents of these
ceived only standard medical treatment76. Although there was vari­ guidelines have noted that meeting all criteria is unnecessary;​
ability across types of support interventions, the findings were con­ rather, the more evidence to support the criteria, the stronger the
sistent across patients being treated for cardiovascular diseases, likelihood of causality77,78. Nonetheless, critiques of the Bradford
cancer and other conditions. Hill guidelines point to the need for more sophisticated analyses.
Drawing causal inferences among factors known to influence Additional promising evidence exists to support potential
health is essential to determine etiology and prevention efforts. causal associations beyond the Bradford Hill criteria. Drawing
However, randomization is not always appropriate in the context causal inferences may be appropriate from sophisticated regres­
of understanding isolation, loneliness, and social connection. Fur­ sion analyses of longitudinal observational data81, applying a data-­
thermore, although the RCT study design is considered the gold integration framework82, and Mendelian randomization83. While
standard for causal inference, it is also criticized because RCTs few studies focusing on indicators of social connection and health
often have homogeneous and small sample sizes due to inclu­ have employed these methods, those that do are supportive14.
sion/exclusion criteria, limiting generalizability to real-­world ap­ Thus, reviews of this evidence have concluded that the cumulative
plication. Thus, additional methods are needed to draw causal evidence supports the likelihood of a causal association between
inferences for public health. While causal inference is challenging better social connection and better health5,78,80.
and much debate exists, several models that provide promising Despite considerable strengths in the evidence, several nota­
support for a causal relationship between social connection and ble gaps remain in our knowledge. Some gaps became glaringly
health have been applied. apparent during the COVID-­19 pandemic, when the scientific
The Bradford Hill guidelines are among the most widely adopt­ community struggled to answer basic questions for the broader
ed criteria for drawing causal inferences among variables unsuit­ public, such as:​How much socializing is needed for health ben­
able for randomization. These guidelines emphasize nine crite­ efits? How soon do adverse mental and physical health conse­
ria:​strength of association, consistency, specificity, temporality, quences emerge when we lack social connection? Is there equiva­
biological gradient, plausibility, coherence, experiment, and anal­ lence between in-­person and remote means of socializing? What
ogy77. Reviews of the evidence on social connection and health can we do to reduce loneliness? Indeed, there are likely many more
have found support for nearly all the Bradford Hill criteria78-­80. The questions for which we do not have adequate or firm answers at
only criterion not met was specificity, indicating that exposure to the moment.
the potential cause (social connection) is associated with multi­ple While there are many strengths in our current body of evidence,
outcomes rather than a particular outcome and no others. How­ gaps in this evidence may become barriers or limit our ability to
ever, smoking also would not meet this criterion for causality, since translate this evidence into practice. To address these gaps more

Table 1 Strengths of the evidence, challenges posed by gaps, and consequent priority needs in research on social connection
Strengths of evidence Challenges Priority needs

Converging evidence across scientific Variability in conceptualization and measurement A multi-­factorial approach is needed.
disciplines
Many validated assessment tools Variability in assessment tools limits comparisons across time, Consistency of assessment to establish prevalence
or different samples. rates and track trends.
Validated instruments may not be generalizable to other Improve or create new measures that are valid, reliable
cultures, settings, and contemporary modes of socializing. and acceptable.
Dose-­response of social connection Most research and attention are on extreme risk and older A focus across the risk trajectory (including
across the lifespan adults. prevention) and across ages is needed.
Converging evidence across social Fewer studies examine multiple components in the same Further evidence of potential independent, additive
connection components sample. and synergistic effects is needed to assess risk more
precisely.
Further evidence is needed on how each factor may
differentially influence different kinds of outcomes.
Evidence on mortality is consistent Fewer studies include or differentiate:​comprehensive health Basic research to fill these gaps is needed.
across causes of death, country of outcomes, low-­and middle-­income countries, marginalized
origin, gender, and health status groups, varying modalities of socializing (e.g., in-­person,
remote, non-­human).
Robust evidence of mortality and Weaker and mixed evidence on effective strategies to Evidence-­based solutions:​rigorous evaluations
objective health consequences mitigate risk (weaker methodologies were employed;​most allowing for strong inference;​interventions
interventions are individually focused;​most interventions across the socio-­ecological model;​prevention and
are targeted at those most severely affected). mitigation of risk earlier on in the risk trajectory.
Less is known about other non-­health outcomes. Evidence on more diverse outcomes (e.g., economic,
civic engagement, education, incarceration).

World Psychiatry 23:3 - October 2024 317

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
comprehensively, Table 1 provides an overview of some of the Loneliness trends provide an incomplete picture of the state of
strengths and challenges currently existing in the evidence base, social connection, and we must look at the other ways in which
further pointing to where future efforts may be prioritized. individuals and communities may lack connection. For example,
data from the American Time Use Survey, regarding how Ameri­
cans spend their day, demonstrate that, over the past two decades,
EXAMINATION OF TRENDS Americans have spent more time in isolation and less time with
household and non-­household family members, friends, com­
Examining trends in prevalence rates, awareness and research munity engagement, and companionship87. Although the COV­
on social connection, isolation and loneliness offers valuable ID-­19 pandemic exacerbated these trends, social isolation was
insights into the trajectory of societal dynamics and the evolving increasing, and engagement with family, friends and others (co-­
evidence base. Tracking prevalence rates allows us to understand workers, neighbors, acquaintances) was declining for years prior
the scale of these phenomena, informing translation to applica­ to the pandemic. This is consistent with other trends, such as those
tion and practice. Concurrently, heightened or lack of awareness documenting a decline in social capital and participation in reli­
reflects the perceived importance of the significance and motiva­ gion88,89, and changes in family structure (e.g., decline of extended
tion to act upon social factors for mental and physical health. families, rise of single-­occupancy households)90 – many of which
These trends are both shaped by the evolving landscape of re­ are seen globally.
search and may reflect an uneven knowledge base. Collectively, Contemporary society in much of the world is evolving rapidly,
they illuminate the evolving intersection between societal shifts, likely contributing to our current trends and having important
individual experiences, and the scientific understanding of the implications for the direction of the trends going into the future.
intricate connections between social dynamics and health out­ Rapid shifts that may be relevant to social and population health
comes. Staying attuned to these trends is essential for developing include the increasing aging population, widespread adoption
targeted interventions and policies that effectively address the of remote working, increased automation, economic strain and
challenges posed by social connection, isolation and loneliness in inequity91, migration and mobility, mental health crisis among
contemporary society. youth, rise in xenophobia, civil and political unrest, and environ­
mental crises, all of which may potentially exacerbate trends con­
cerning social connection.
Trends in society These trends of declining social connection, combined with the
evidence on the bidirectional associations with mental and physi­
Societal trends over the past several decades indicate that, as cal morbidities, point to an urgent need to take action. Because mul­
a population, we have become less socially connected and more tiple factors have been contributing to these trends, building over
isolated, and that a high proportion of the population is lonely. decades, simply returning to pre-­pandemic levels of connection or
Based on the available data, loneliness has generally shown reducing time on social media may only bring limited benefits.
little improvement over the last few decades, and may be getting
worse. For example, a massive synthesis of 345 studies on emerg­
ing adults (ages 18-­29), who completed the UCLA Loneliness Trends in scholarly attention
Scale between 1976 and 2019, found that average loneliness levels
linearly increased annually across the 43 years84. Furthermore, a There are also striking trends in the scientific study of the topic.
meta-­analysis of data from 113 countries concluded that a sub­ The surging interest in social isolation and loneliness is reflected
stantial proportion of the population in many countries experi­ in research, as demonstrated by the substantial increase of studies
ences problematic levels of loneliness85. on this topic over recent years, potentially providing greater under­
According to the Gallup Global State of Connection survey, standing and justification for action. Thus, understanding how
near­­­­ly a quarter (24%) of the global population reports feeling lone­liness and isolation have been studied over time may provide
“very lonely” or “fairly lonely”, although there was variability across additional insight.
countries86. Of the 29 countries where at least one third of the pop­ To examine publication trends, we first used the PubMed by
ulation felt lonely, 22 were in Africa, four were in the Middle East, Year search tool. Because of the diverse literature on social, mental
and three in South Asia. This also demonstrates that loneliness is and physical health outcomes, the search was limited to two so­
not just a wealthy Western country issue, and may even be more cial variables (loneliness and social isolation) and two health out­
severe in other areas of the world. However, inconsistent mea­ comes (depression and mortality). We further scanned additional
surement tools and scoring methods have led to vastly different scientific databases (including PsycINFO for depression) using
prevalence estimates. Notably, prevalence rates often favor one the same social and health variables. The searches were limited
indicator (e.g., loneliness) over others, yet indicators may interact to articles published in peer-­reviewed academic journals between
in meaningful ways. Thus, the prevalence of those who lack social 1972 and 2023. The mortality search terms included “social isola­
connection in one or more ways may be far larger than any esti­ tion” OR “loneliness” AND “death and dying” or “mortality” or
mate of a single indicator. “mortality rate” or “mortality risk”. The depression search terms in­

318 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Figure 2 Frequency of loneliness or isolation as search terms in the scientific literature over time. Note that the far-­left column refers to 1972-­
2003, while each of the other columns refers to two years.

cluded “social isolation” OR “loneliness” AND “major depression”. spread use of social media platforms may have played a dual role
PsychINFO also allows narrowing search by methodology:​thus, in awareness. Increased experience of feelings of loneliness asso­
we further used the search parameters “empirical”, “quantitative”, ciated with that use, and the facilitation of awareness campaigns,
“longitudinal”, “prospective”, “retrospective”, and “clinical trial”. discussions and support networks related to health and well-­being
All studies using those search terms were bracketed into time may occur simultaneously92.
periods to determine how many articles populated by our search Government initiatives may have also played a role in greater
terms were published within each period. Studies published in awareness. Countries have recognized the urgency of the issue
the past two decades (2004-­2023) were demarcated into 2-­year and appointed Ministers, formulated policies, and developed
­periods (2004-­2005, 2006-­2007, 2008-­2009, etc.), while studies strategies to address loneliness and isolation, and highlight social
published in 1972-2003 were grouped (31 years). Figure 2 graphi­ connection as a priority. Awareness efforts have also been under­
cally presents the number of studies on isolation and loneliness taken by national and international civil society organizations, co­
over the years. alitions, and networks that have emerged as powerful advocates7.
Data suggest an exponential increase in the scientific study These include the UK Campaign to End Loneliness, the Canadi­
of social isolation and loneliness. Over the past two decades, the an Genwell Project, Australia’s Ending Loneliness Together, and
number of relevant articles has grown, with significant increase the annual Global Loneliness Awareness Week. These collective
since 2020. For example, the number of papers published in each efforts aim to raise awareness, promote community engagement,
subsequent two years since 2020 exceeds the number of studies and foster a culture of connection.
from 1972 to 2003 combined. However, it is unclear whether sci­ Unfortunately, trends in public awareness appear to be lim­
entific interest in other indicators of lacking social connection is ited to only certain outcomes. A large survey of US and UK adults
similarly surging. published in 2018 found that, when the public was asked to rank
various factors contributing to a longer life (e.g., not smoking, exer­
cising, limiting alcohol, maintaining a healthy weight), social con­
Trends in awareness nection was amid these factors, but it was rated among the lowest
in importance, significantly underestimating its impact relative to
Several factors may presumably contribute to greater aware­ effect sizes reported in the scientific literature93.
ness of the importance of social connection and related aspects Due to a variety of factors occurring since that survey was pub­
of lacking connection (i.e., social isolation and loneliness). These lished – i.e., the COVID-­19 pandemic, national awareness cam­
include scientific advancement, social media, government initia­ paigns, and increased prevalence within the population – pub­
tives, the COVID-­19 pandemic, and advocacy. lic perception of the health relevance of social connection was
Significant advances in scientific research over the past few expected to increase. However, 2023 data from the UK and a na­
decades, especially in the last 5-­7 years, may have shed light on tionally representative sample of US adults demonstrate that there
the scale of the problem and provided greater confidence in scien­ has been essentially no change94. Despite increases in public dis­
tific findings. Advancements in social technologies and the wide­ course on social isolation and loneliness, the importance of these

World Psychiatry 23:3 - October 2024 319

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
and other aspects of social connection for health and survival are ties for individuals, communities and societies globally.
still underestimated among the public. Key challenges include developing a common language, iden-
tifying and activating appropriate and effective approaches, and
adapting to societal changes. These challenges are also intercon­
Implications from these trends nected. Developing a common language is essential to understand­
ing the underlying contributors, predicting outcomes, and measur­
Overall, these trends point to a large and potentially increasing ing changes in risk and protection. Understanding these challenges
scale of those lacking social connection, and parallel trends sug­ helps us develop better approaches to preventing and mitigating
gest increasing attention within scholarship on the consequent risk, and adapt these approaches as society evolves.
impact on health. Yet, the public perception of risk does not align
with either the increasing scale or evidence of the magnitude
(effect sizes) for health. This suggests that increased education and A common language
awareness of the health relevance is needed.
Discrepancies between the scientific evidence and public per­ Given the array of terms used in the scientific literature, one po­
ception may have significant implications. First, public percep­ tential barrier to prioritization within health settings is lack of pre­
tion may significantly influence how resources are allocated and cision in terminology. It is clear that we need a common language,
prioritization of various issues within public health agendas95. If but the term “loneliness” may fall short. Loneliness is often used
the public does not perceive social connection and markers of its as a catch-­all term outside academic scientific contexts, but it is
deficit (e.g., loneliness and social isolation) as relevant to health93, defined and measured more narrowly within the scientific litera­
funding and efforts may not be directed towards addressing ture. While definitions of loneliness vary somewhat, there is broad
them adequately, despite their demonstrated impact on health consensus that it is distinct from social isolation7,99.
outcomes4,15. Second, public perception influences individual Loneliness is a subjective, unpleasant feeling based on the dis­
behaviors and societal norms. If social connection is not widely crepancy between one’s desired and actual level of social connec­
recognized as a protective factor, and loneliness and isolation as tion100. It is most often distinguished from social isolation as a sep­
serious health risks, individuals may be less likely to change their arate but related construct7. While isolation and loneliness can co­
own behavior or support others experiencing loneliness or isola­ exist, they differ in meaningful ways. Social isolation is objectively
tion96. This may perpetuate social disconnection and exacerbate being alone, having few relationships or infrequent social contact.
the problem. Thus, social isolation is objective, while loneliness is subjective.
Finally, accurate awareness of the health implications among Although both social isolation and loneliness can be involuntary,
the public may facilitate destigmatizing the issue and promoting isolation may be chosen101. Both are indicators of lacking social
help-­seeking behavior97. When people view loneliness and isola­ connection, but there are many indicators of social connection
tion as a personal rather than a health issue, they may be less in­ and, thus, many indicators of social connection deficits8. Social dis­
clined to seek support and resources to address these challenges. connection and loneliness are not equivalent43, and this has impli­
Aligning public perception with the evidence on the importance of cations for measurement and assessment, intervention, policy, and
social connection is essential to shaping effective policies, nurtur­ more.
ing more connected and supportive communities, and promoting Across scientific disciplines, several constructs have emerged
health. as relevant. Table 2, although not comprehensive, highlights some
of the most widely used terms represented in the research and
identified in the US Surgeon General Advisory4. Pinning down
CHALLENGES definitions is challenging, given that the same term has been used
to refer to different things, while different terms are used to de­
The WHO defines health as “a state of complete physical, men­ scribe the same thing among studies. Some terms, such as social
tal and social well-­being, not merely the absence of disease or in­ capital, lack a clear consensus on definition102,103.
firmity”98. Considering this definition, social well-­being is a critical Why is this important? These terms refer to related but distinct
element of health that has been underappreciated and raises sev­ constructs. Reviews of this evidence find that these measures are
eral challenges that we must address. not highly correlated empirically8,104. Thus, when we only mea­
If physical and mental health are more than the absence of sure one of these, we cannot assume that we are capturing the full
physical or mental illness, we should be taking a similar approach scope of how social factors influence health.
to social health. However, recent attention appears primarily fo­ We need a common language. “Social connection” has been
cused on indicators of social deficits, specifically social isolation offered as an umbrella term to encompass these distinct but re­
and loneliness. Yet, our collective and individual capacity as hu­ lated terms4,8,15,105. From this perspective, the myriad of diverse
mans to think, feel, engage with others, pursue livelihoods, and concepts in the scientific literature can be organized into three
experience fulfillment is intrinsically tied to our health – physical key themes or components:​structure, function and quality. The
health, mental health and social health. The active encourage­ first component, structure, represents the human need to have
ment, safeguarding and recovery of social health are crucial priori­ others in our life and is often measured by the size and variability

320 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table 2 Terms commonly found in the scientific literature that are distinct but related (adapted from the US Surgeon General’s Advisory4)
Term Definition

Loneliness A subjective distressing experience that results from perceived isolation or inadequate meaningful connections, where inadequate
refers to the discrepancy or unmet need between an individual’s preferred and actual experience.
Social capital The resources to which individuals and groups have access through their social connections. The term is often used as an umbrella
for both social support and social cohesion.
Social cohesion The sense of solidarity within groups, marked by strong social connections and high levels of social participation, that generates
trust, norms of reciprocity, and a sense of belonging.
Social connectedness The degree to which any individual or population might fall along the continuum of achieving social connection needs.
Social connection A continuum of the size and diversity of one’s social network and roles, the functions that these relationships serve, and their
positive or negative qualities.
Social disconnection Objective or subjective deficits in social connection, including deficits in relationships and roles, their functions and/or quality.
Social infrastructure The programs (such as volunteer organizations, sports groups, religious groups, and member associations), policies (such as public
transportation, housing and education), and physical elements of a community (such as libraries, parks, green spaces, and
playgrounds) that support the development of social connection.
Social isolation Objectively having few social relationships, social roles, group memberships, and infrequent social interaction.
Social negativity The presence of harmful interactions or relationships, rather than the absence of desired social interactions or relationships.
Social networks The individuals and groups a person is connected to and the interconnections among relationships. These “webs of social
connections” provide the structure for various social connection functions to potentially operate.
Social norms The unwritten rules that we follow which serve as a social contract to provide order and predictability in society. The social groups
we belong to provide information and expectations, and constraints on what is acceptable and appropriate behavior. Social norms
reinforce or discourage health-­related and risky behaviors (lifestyle factors, vaccination, substance use).
Social participation A person’s involvement in activities in the community or society that provides interaction with others.
Social support The perceived or actual availability of informational, tangible and emotional resources from others, commonly one’s social network.
Solitude A state of aloneness by choice that does not involve feeling lonely.

of relationships within a network, being part of groups, and regu­ butions of the structure, function and quality of social connection.
lar social interactions. It is the foundation upon which the other Data across multiple scientific disciplines have linked various
components of social connection are built. The second compo­ social connection indicators to health outcomes8. Strong struc­
nent, function, recognizes that these connections serve essential ture, function and quality of social connection may be considered
functions or purposes. Namely, connections can be relied upon optimal for health. On the contrary, when all three are low, this
for support to meet various needs and goals. Functions are often would be associated with high to severe risk. However, there may
measured by the interchange of support that is received or per­ be unevenness in the extent to which any individual experiences
ceived to be available, which can be emotional, informational or the three components of social connection. The descriptions in
tangible, and can help us navigate life’s challenges. The bulk of the Table 3 help illustrate the disaggregation across these components
studies within the scientific literature have primarily examined and their relation to various risk profiles. Nonetheless, there is
indicators of these structural or functional components. How­ likely further complexity of risk, given that many indicators within
ever, a growing body of research is assessing and recognizing the each component of social connection are on a continuum and
importance of the quality of social relationships, networks and may have synergistic effects. For example, longitudinal data from
interactions. Thus, the third component, quality, refers to our con­ nearly half a million people, followed for an average of 12.6 years,
nections’ positive and negative aspects. High quality is often mea­ demonstrated that low levels on both structural and functional
sured by the level of satisfaction or intimacy, whereas low quality indicators of social connection resulted in a significantly higher
includes social negativity such as conflict, strain or ambivalence. risk for cardiovascular disease mortality (hazard ratio, HR=1.63),
While Figure 3 is helpful in identifying these core conceptual compared to low levels on structural (HR=1.27) or functional
themes, individual measurement approaches may overlap to some (HR=1.17) components alone54.
degree between social connection components. Furthermore, spe­ Conceptually, loneliness may represent the signal or symptom
cific assessment tools may appear to align clearly within one com­ of unmet social needs. However, loneliness does not represent low
ponent, but contain items that overlap with other components104. levels across all three social connection components. Compari­
Generally, high levels of each of these components have been sons demonstrate these distinctions. For example, meta-­analyses
linked to better health and lower levels of poorer health. To more that establish the effect size for the aggregate measures of social
comprehensively understand underlying causes, predict out­ connection on mortality were significantly larger than the effect
comes, and measure risk, we need to consider the distinct contri­ size for loneliness43,105. Thus, loneliness is not the same as lacking

World Psychiatry 23:3 - October 2024 321

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Figure 3 Social connection as a multi-­factorial umbrella term encompassing the structural, functional and quality aspects represented in the
scientific literature (adapted from Holt-­Lunstad8)

social connection. social contact with others. Furthermore, if an individual is low on

one component, we may be missing potential protection associ­
ated with high levels on the other components. Thus, one’s overall
Measurement social risk profile may be incomplete because of the limited scope
of assessments.
Another challenge related to the need for a common language
is the lack of consensus on measurement104. The most widely
used measurement tools have helped to build a robust evidence Effective intervention and prevention strategies
base, but may have limitations when applied to other settings.
For instance, most measurement tools were developed in West­ The next major challenge is reducing risk through effective in­
ern countries, prior to the widespread shift to digital and remote tervention and prevention strategies. Social connection is com­
means of socializing. Measurement not only needs to be predic­ plex, with various factors contributing to its increase or decrease,
tive of the outcomes of interest, but must also be feasible to use. directly and indirectly8. Generally, social connection occurs natu­
Notably, what is feasible may differ in different contexts, such as rally among individuals and within communities. However, when
research, clinical settings, population surveillance, and evaluation it does not, intervention becomes necessary to reduce risk. Direct
of the effectiveness of interventions. Currently there is no measure actions, programs or initiatives can be implemented to increase
that is multi-­factorial, validated, and feasible or adapted to be­ social connection or decrease forms of social disconnection inten­
come feasible across settings. tionally.
Given the multi-­factorial conceptualization of social connec­ Key challenges include:​a) the capacity to develop and evalu­
tion7,80, a considerable challenge is developing a feasible multi-­ ate intervention strategies;​b) the difficulties to understand what
factorial measure. Not all social connection components are typ­ works best for whom in what context;​and c) the limited scope of
ically assessed, because this would take too much time. Due to existing strategies, and the need to ensure the full scope of social
time and space constraints, assessments in medical settings and connection across the socio-­ecological model, sectors of society,
population surveillance may only assess one indicator;​however, and life course.
this approach will likely result in risk assessment errors. For ex­
ample, if an individual is assessed on an indicator of the structural
component of social connection (e.g., frequency of social contact) Developing and evaluating interventions
and found to have high levels, we may assume that this person is
at low risk. However, this assessment may miss low levels on the The evidence supporting the positive effects of social connec­
other two components (e.g., low social support, poor quality re­ tion is far more robust and methodologically rigorous than the
lationships), which may put the individual at risk. Similarly, we evidence supporting the effectiveness of interventions aimed at
might assume that this person is at low risk if scores are low on creating it when it is not occurring naturally, or at reducing social
an assessment of loneliness, yet the person may have little or no disconnection. However, this challenge (i.e., more substantial evi­

322 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table 3 Conceptualization of potential risk to mental and physical health according to distribution across the level of social connection com-
ponents
Risk level Structure Function Quality Description

Optimal-­low risk High High High Large and varied social network, with regular social contact with people who can be relied upon
for support and assistance when needed. These include deep and meaningful relationships
characterized by caring and compassionate interactions.
Low-­moderate risk High High Low Large and varied network, with regular contact among people who can be counted upon for
support. However, these relationships are strained and/or lack depth, and interactions are void
of caring or compassion.
High Low High Large and varied network, with regular social contact with meaningful and high-­quality relation­
ships. However, these are not able or available to provide support or assistance when needed.
Low High High Small social network and infrequent contact. However, the limited social contact is among those
who can be relied upon for support, perhaps strangers or volunteers. Nonetheless, it is caring
and compassionate.
Moderate-­severe risk High Low Low Large and varied social network, and regular contact with others. However, they cannot be relied
upon for support. These are strained relationships and interactions, with a lack of caring and
compassion.
Low High Low Small social network and limited social contact with others. Support is available and provided by
others, perhaps by strangers or volunteers;​however, it lacks depth, is accompanied by strain, or
lacks caring and compassion.
Low Low High Small social network and limited social contact with others. It is not possible to rely upon others
for support. However, the limited social contact is caring and compassionate.
Severe risk Low Low Low Small social network and little social contact. There is no one to rely upon. What little social
contact does occur is strained or lacks caring and compassion.

dence of health risks compared to treatments to eliminate these caregiver, professional, volunteer);​modality (e.g., in-­person, phone,
risks) is common among many health issues. The National Insti­ virtually);​sub-­population group (e.g., older adults, children, dis­
tutes of Health estimate that therapeutics for any condition take, abled, university students, veterans, new parents), and many other
on average, 10-­15 years to develop, because 95% of new therapeu­ characteristics.
tics fail106. With the increased urgency to address the crisis of social Interventions also vary in their timing and duration (e.g., once
isolation, loneliness and social disconnection, we cannot take a or repeated, hours to years);​their outcomes (e.g., social, health,
“something is better than nothing” attitude, assuming that all ap­ performance);​their target (e.g., general population, high-­risk
proaches will be helpful. populations);​and goals (e.g., prevention, mitigation, treatment).
Rigorous evaluations are needed. However, the resources and Effectiveness may depend on the specific characteristics of the tar­
capacity to develop and evaluate interventions are limited – par­ geted population, the type and intensity of the intervention, and
ticularly for interventions conducted outside academic institu­ its length15. This variation creates a considerable complexity. We
tions. Rigorous methodologies are often not utilized, resulting in highlight here the interventions with the most promising body of
a low-­quality body of evidence107,108. To strengthen this evidence, evidence.
the Multiphase Optimization Strategy (MOST) approach has been
utilized for other health issues and could similarly be applied to
this area109. The MOST framework is an iterative implementation Loneliness interventions
method that uses empirical information about component effects
within real-­world constraints to develop, evaluate and optimize There is now a sizable body of research examining interven­
interventions110. tions focused specifically on reducing loneliness. Systematic re­
views and meta-­analyses generally find that these interventions
are associated with significantly reduced loneliness and improved
Understanding what works for whom in what context social support. For example, an umbrella review of 211 studies,
including seven different types of interventions, examined their
There is a growing body of evidence examining the effective­ effectiveness in reducing loneliness116. They were befriending
ness of interventions, including multiple meta-­analyses and re­ programs, technological interventions, meditation/mindfulness,
views of the evidence108,111-­116. Interventions vary in terms of their animal therapy or robopets, social cognitive training, social skills
social connection focus (e.g., loneliness, social isolation, school training, and social support. Of these intervention types, social
connectedness, social skills, social support, neighborhood cohe­ support, social cognitive training, and meditation/mindfulness
sion);​setting (e.g., home, clinic, community, school, whole of so­ significantly decreased loneliness.
ciety policies);​delivery (e.g., self-­directed, peer group, family or Among loneliness interventions designed to target specific

World Psychiatry 23:3 - October 2024 323

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
age-­based sub-­populations, there are meta-­analyses of evidence social, emotional and practical needs. An integrative review of the
for those focused on young people, university students, and older evidence found that social prescribing has generally favorable
adults. A meta-­analysis of 39 studies (including 25 RCTs) focused effects in reducing social isolation and loneliness. However, the
on loneliness in children and adolescents found that it could be quality of the evidence was mixed and generally weak method­
reduced, with no significant differences between various types of ologically118. The interventions were diverse and heterogeneous in
interventions115. A systematic review of 37 interventions among design and implementation, relied on self-­report, and often lacked
university students found that bringing students together for an adequate controls.
activity or to socialize, in-­person or virtually, helped reduce feel­ While social prescribing is a promising approach gaining pop­
ings of loneliness117. Meditation/mindfulness benefited those ularity, further research is needed, including RCTs and meta-­anal­
who preferred not to join groups. Other reviews identify several yses, as multiple other systematic reviews provide a weak or mix­ed
effective interventions for reducing loneliness and increasing so­ ­picture122-­124. More robust evidence is needed to understand how
cial connection in older adults, including social support groups, strong the effects are for individuals, sub-­populations and com­mu­
technology-­based interventions, and community-­based activi­ n­ities regarding loneliness, isolation and social connectedness, and
ties118,119. to identify the most effective approaches for different populations.
Overall, based on the current evidence, no intervention type
(e.g., changing maladaptive social cognitions, enhancing social
skills, providing psychoeducation, supporting socialization, in­ Technology-­based or virtual interventions
creasing opportunities for social interaction) seems to be superior
to the others. The majority of this evidence has been classified as Technology-­based or virtual interventions – such as online so­
low to critically-­low quality116. cial networking, video conferencing, messaging apps, and virtual
companions or pets – are implemented with the aim of reducing
social isolation or loneliness among specific populations. System­
Interventions in clinical settings atic reviews of the evidence found that technology-­based inter­
ventions were effective in reducing loneliness among older adults
Given the robust evidence of the medical relevance of social and individuals with mental health issues111,125,126.
c­ on­nection, addressing isolation and loneliness in clinical settings The WHO has developed an evidence and gap map for tech­­
among patients may improve health outcomes. Early evidence nology-­based interventions for reducing social isolation and
pointed to greater survival among cancer patients who participat­ lone­­liness among older adults127. This includes 200 studies and
ed in social support groups along with standard treatment com­ 97 systematic reviews. Most interventions utilized video confer­
pared to standard treatment alone120. Since then, various types of encing and calls, though assistive robots and virtual pets were also
programs have been developed to help support patients across common.
different medical conditions, but with mixed outcomes. None­ The effectiveness of digital interventions may vary depending
theless, when the body of the evidence was examined as a whole on the specific population and the type of technology used. Caution
via meta-­analysis, including 106 RCTs, medical patients random­ should be used, given that some studies found no effectiveness
ized to receive some type of psychosocial support intervention in and, in some cases, negative outcomes. For example, data from the
addition to standard medical treatment had a 20% increased sur­ National Social Life, Health and Aging Project found that, despite
vival, and 29% increased survival time compared to patients who increases in remote modes of contact with others, individuals still
only received standard treatment76. experienced loneliness, depression and decrease in happiness128.
While there was considerable variability in the effects among While some technology-­based interventions may be promis­
the interventions, on average, providing patients with psychoso­ ing, not all effectively reduce social isolation or loneliness. More
cial support was as effective in increasing survival as many stan­ research is needed to fully understand their effectiveness, for
dard medical interventions, including smoking cessation and which groups, and how they can be optimally implemented.
lifestyle interventions. Thus, not only do high levels of naturally
occurring social connection increase one’s lifespan, but providing
interventions to support patients in medical settings also seems to School connectedness
increase survival. This evidence suggests, consistent with NASEM
recommendations, that addressing the social needs of patients There is strong evidence that interventions aimed at increasing
by integrating this component into existing treatment within the school connectedness, or the feeling of belonging and engagement
health care system may be a promising approach15,121. within the school community, can positively impact student out­
comes, from academic achievement to reduced suicidality129-­131.
In one review, classroom management approaches were associ­
Social prescribing ated with improved school connectedness among students, in­
cluding teacher caring and support, peer connection and support,
Social prescribing involves referring patients outside the med­ student autonomy and empowerment, management of classroom
ical setting to community-­based services and activities to address social dynamics, teacher expectations, and behavior manage­

324 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ment132. broadly or these groups.
Research has shown that students who feel more connected to One perspective is that prioritization should be given to the
their school are more likely to attend class regularly, have higher most vulnerable populations and addressing their needs. By fo­
grades and test scores, are less likely to engage in risky behaviors cusing on specific populations, such as marginalized or underserv­
(e.g., substance abuse, violence), and have better health133-­136. ed communities, efforts can be directed toward reducing health dis­
Classroom practices that build strong, supportive and trusting parities and ensuring that resources reach those who need them
relationships help reduce patterns that inappropriately exclude the most145. Many sources recommend tailoring interventions to
some groups of kids132. address specific needs, barriers and enablers within these groups,
in order to increase the effectiveness of these interventions15,146.
This approach may be a more efficient use of limited resources,
Policy funding and personnel. Moreover, it is easier to measure the im­
pact of targeted approaches, as they are narrowly defined and
There is growing interest in the role of policy as an intervention, address a smaller population. However, identifying and targeting
with many calls to enact pro-­social policies, or policies to address people “at-­risk” may potentially pathologize and stigmatize such
isolation and loneliness4,137-­139. Policies are explicit guidelines groups further and place the burden of change on the individual.
which provide a framework for decision-­making;​are enforced by Thus, it has been argued that we should focus on the factors that
groups, organizations or governments;​and can directly or indi­ put people at risk instead of group membership147.
rectly impact social connection. Similar to the Health in All Policy Another perspective is that we should focus efforts across the
approach that recognizes the health implications across sectors population to have larger shifts, rather than just targeting a small
(e.g., education, employment, health, nutrition, housing, transpor­ portion of the population77. Broad approaches can lead to sys­
tation)140, a “Social in All Policy” approach should recognize the temic changes in policy, environment, and societal norms, laying
social relevance of policies across sectors80,138. the foundation for long-­term health improvements. Implementing
Policies can directly influence social contact (e.g., policies on broad interventions might also benefit from economies of scale,
visitation or family member involvement in medical care), or can reducing the cost per individual reached compared to targeted in­
focus on changing other kinds of outcomes (e.g., economic, en­ terventions.
vironmental) that substantially influence social connection (e.g., Both targeted and broad approaches are necessary, starting
policies on neighborhood zoning, bussing routes, remote work). with broad measures to address general issues, while using target­
Reviews of existing policies cover social and emotional learning ed interventions to address specific needs within the population.
curricula in schools130,141;​state-­level farmer wellness programs142;​ However, targeted approaches should be focused on the factors
expansion of telehealth services to provide mental health services associated with risk (e.g., marginalization) rather than group
in schools143;​and workplace policies that include shorter total membership, to avoid further stigmatization. Universal approach­
work hours and earlier end of the workday, enabling workers to es may help prevent social disconnection, whereas more targeted
attend to family responsibilities and achieve greater work-­life har­ approaches may be needed for those who are already isolated,
mony144. There is existing US legislation, including the Older Amer­ lonely, or socially disconnected in other ways for prolonged peri­
icans Act of 1965, which was amended in 2020, to address social ods or at severe levels. A hybrid strategy can leverage the strengths
­is­olation and loneliness. of both approaches to maximize public health outcomes.
Many policies are being introduced with the intent to facilitate
social connectedness. However, given the scale and magnitude of
public health implications, they need to be evaluated for effective­ Limited scope of existing approaches
ness like any other intervention.
Despite the growing body of research focused on interventions,
the scope of solutions is limited in several ways. The Systemic ap­
Targeted vs. broad approaches proach Of Cross-­sector Integration and Action across the Lifespan
(SOCIAL) framework points to gaps and opportunities in solu­
Another major challenge is whether to focus solutions on peo­ tions across the socio-­ecological model, sectors of society, the life
ple most severely affected or broadly on the population. When so­ course, and prevention80.
cial connection needs are not met, the mental and physical health Evidence points to underlying root causes across the socio-­eco­
consequences are broadly found across age and other demograph­ logical model (e.g., individual, interpersonal, community, institu­
ics. However, isolation and loneliness are unequally distributed a­- tion, society)8, yet most interventions are being deployed at the in­
cross the population. Groups that experience marginalization – i.e., dividual level148. A scoping review of interventions for older adults,
lesbian, gay, bisexual, transgender, queer and/or questioning (LG­ including evidence from 30 countries, found that the majority of
BTQ+) people, racial minorities, migrants, those with disabilities91 interventions only measured loneliness, and only three societal-­
– and life circumstances that may or may not co-­occur with aging level interventions were found149.
(e.g., functional or sensory impairments) are disproportionally af­ The health care sector, including both clinical and communi­
fected. Thus, a significant challenge is whether to focus efforts more ty health settings, is most often the target of interventions and pro­

World Psychiatry 23:3 - October 2024 325

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
grams. However, we need to expand our approaches across sec­ cal thinking, problem-­solving), self-­identity (e.g., self-­confidence,
tors to engage the whole of society. No one sector of society is likely self-­esteem), ability to navigate cultural norms and values, and
to be able to address this issue. The “Social in All Policy” app­roach overall physical health155.
138
recognizes the health and social implications across sectors and Early life experiences with caregivers, peers, schools and com­
“systematically takes into account the health implications of deci­ munities are all key contributors to the early social environment
sions, seeks synergies, and avoids harmful health impacts in order that ensures longer-­term well-­being and survival156-­160. However,
to improve population health and health equity”150. the pandemic severely limited socialization for roughly three years,
Social connection is vital at every stage of life, yet most solutions during this critical developmental period, for infants, young chil­
are focused later in life111,151. The evidence of a dose-­response ef­ dren and adolescents. Since it is well documented that early social
fect of social connection on biomarkers of health across stages of experiences significantly predict later social, mental and physical
life46, and the importance of early social environment5, highlights health136,161,162, the long-­term health implications of the pandemic
the need for efforts to address this issue across life. Social isola­ may be disproportionately borne for an entire generation.
tion during childhood, for instance, is associated with increased There may also be longer-­term consequences among adults,
cardiovascular risk factors in adulthood, such as increased blood due to widespread behavioral adaptation that may be sustained
glucose levels, high blood pressure, and obesity152. over time. The massive adoption and implementation of tools to
As is the case with most health issues, primary, secondary and cope with reduced social contact (e.g., remote work, contactless
tertiary prevention approaches are needed to address social iso­ delivery services, streaming entertainment services, telehealth,
lation and loneliness. Unfortunately, few interventions focus on automation) came with significant advantages, including in­
helping the society become more socially connected. Prevention creased flexibility, autonomy, convenience, safety, and in some
efforts may have many longer-­term benefits, such as avoiding cases cost-­effectiveness163-­167.
costly interventions later, reducing disease burden, and improving These advantages have led to preferences that may limit our so­
quality of life153. cial contact – particularly with co-­workers and weak-­ties. Reduc­
Efforts to gather and synthesize data, and to identify evidence tions in social contact with both may be critical factors for future
gaps, are underway. These and similar efforts aim to help create health, given the significance of workplace relationships168,169 and
centralized resources to single out evidence-­based interventions the evidence on the importance of weak-­ties170,171. While these
effective in reducing social isolation and loneliness, or increasing preferences are obviously not universal, a large portion of the pop­
social connectedness. However, without sustained funding, there ulation values such conveniences. Even if they are not preferred,
will be difficulties to evaluate the evidence supporting their effec­ they are often incentivized by lower costs166. For example, roughly
tiveness. half of patients preferred in-­person visits and half preferred a vid­
eo visit, but 23.5% switched to a video visit if the cost was lower172.
What was once initiated or scaled to help us cope with isolation, is
IMPLICATIONS FOR THE FUTURE OF MENTAL now what may be reinforcing isolation, with potential long-­term
AND PHYSICAL HEALTH implications for exacerbating existing levels of social disconnec­
tion and corresponding health consequences.
The world is beginning to recognize the vital importance of so­ The widespread behavioral adaptation to spending more time
cial connection to the health and welfare of countries. Consider­ alone, or not leaving the house, may contribute to a societal shift
ing the trends that have led to concerns of a public health crisis that normalizes social isolation. This is increasingly being depicted
of social disconnection, we must proactively evaluate the long-­ in cultural narratives of a “social-­battery” that is drained by social­
term implications if these conditions do not improve or perhaps izing, and the benefits of “self-­care”, “me-­time”, and solitude. While
continue to worsen. Two of the most pressing concerns that have there is evidence of some benefits of solitude173, the evidence is
the potential to worsen trends are the unknown long-­term con­ quite heterogeneous, and benefits appear limited to short-­term
sequences of the COVID-­19 pandemic and the rapidly evolving bouts of solitude, not chronic time alone174-­175. There is instead – as
technological landscape of society. we have seen – robust empirical evidence of the harmful effects of
social isolation on mental and physical health outcomes, and in­
creased risk for premature mortality. If time spent alone is praised
Long-­term implications of the COVID-­19 pandemic and encouraged, while the risks of social isolation are diminished
within public discourse, the consequences to health are likely to
During the COVID-­19 pandemic, social developmental pro­ be magnified.
cesses were significantly disrupted, with potentially critical long-­ The pervasive experience of social isolation and loneliness dur­
term health implications. Socialization during early life plays a cru­ ing the pandemic also fueled self-­proclaimed “experts” and influ­
cial role in shaping a child’s development and long-­term health154. encers who pushed common-­sense approaches to wellness, and
Early socialization provides the foundation for healthy relation­ in some cases misinformation176-­177. Coupled with a growing dis­
ships (e.g., learning to communicate, cooperate, negotiate, share), trust in institutions, including science, this may lead to confusion
emotional well-­being (e.g., learning empathy, expressing emo­ on what is credible. When local community organizations and the
tions), cognitive development (e.g., information processing, criti­ general public are skeptical or distrust science, government, and

326 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
each other, this may increase social disconnection and stifle the “loneliness epidemic”, with 24/7 access to emotional support for
development and acceptance of effective interventions and pro­ all, and increased automation will free up more time for leisure.
grams to reduce isolation and loneliness. From a dystopian perspective, AI will replace human interaction,
and diminish trust in others due to blurred lines between fact and
fantasy. Moreover, several jobs will no longer require humans, re­
Technological advancements sulting in a lack of meaning and purpose in life, and perhaps even
the risk of a downfall of humanity altogether185.
The rapid evolution of digital technologies has already demon­ Some of the potential short-­term benefits that already have
strated co-­occurring shifts in socializing. Much has been written some limited evidence involve AI-­powered virtual companions or
about the mental and physical health implications of the use of chatbots that can engage in conversation and provide immediate
social media92,178-­183. Similar rapid developments and widespread emotional support186, and 24/7 access to mental health support
adoption are occurring with artificial intelligence (AI) and large as well as increased accessibility among those experiencing bar­
language models (LLM) tools, that have the potential to similarly riers (e.g., language, privacy concerns, social anxiety) to in-­person
result in both benefits and harms, but to an exponentially larger therapy. However, increased isolation may occur if there is over-­
extent. reliance on AI interaction and emotional support and forming
The long-­term consequences of AI are yet unknown, but pre­ attachments to AI companions187,188. Generative AI may also mag­
dictions often fall into either a utopian or dystopian outlook for nify our own biases, leading to information echo chambers that
the future, both of which have implications for social health and further isolate us from others189,190.
policy184. From a utopian perspective, AI will be the cure to the Generative AI is neither inherently good or bad for health and

Table 4 Recommendations for national strategies to foster social connection and address social isolation and loneliness

Policy and strategy Make social connection a priority in policy agendas of governments and other organizations.
Establish a national strategy and leadership at all levels to track, advance and coordinate policies and programs across agencies
or units.
Assemble an inter-agency, cross-­sector coalition to assess and address social implications across all policies and programs.
Establish a centralized resource or database for evidence-­based interventions and policies.
Integration within the Prioritize social connection in prevention and integration into treatment in clinical settings.
health system Assess and track risk within the electronic medical records.
Adequate training, resources and support for health care providers.
Healthy digital Establish greater transparency and cooperation to independently evaluate drivers of connection and disconnection.
environments Increase accessibility (access, affordability, knowledge) to digital tools and environments with demonstrated benefits.
Establish safeguards (laws, regulations, guidelines, autonomy) to reduce risk associated with harmful elements.
Evidence, evaluation, Creation of a global social connection index to allow for comparisons across nations.
measurement Establish consistent national measure of social connection, for population surveillance at a national level.
Establish a national research and policy center/institute to coordinate cross-­sector collaboration in research.
Establish Grand Challenges in Social Connection Research, and funding to sustain efforts to address them.
Education and awareness Establish public-­facing national awareness campaigns, ensuring accurate and inclusive messages based on high-­quality evidence.
Establish National Health Guidelines for Social Connection (similar to dietary guidelines).
Include social connection in public-­facing health educational resources (websites) of major health organizations.
Integrate social connection into formal health education curriculum across all educational settings (primary, secondary, post-­
secondary, higher education, continuing learning, advanced and continuing education for health professionals).
Establish age-­appropriate formal education curriculum and practices to foster social connection skills.
Norms and culture Media, arts and entertainment, local and national leaders, and others in positions of influence, can model positive behaviors
that facilitate connection (e.g., respect, openness, responsiveness, kindness, support)
Create routines, habits and programs that reinforce regular social connection within formal (workplace, education) and
informal (neighborhoods, recreation and leisure) settings.
Strengthen norms, incentives and opportunities to create a culture of service.
Establish coalitions and networks to coordinate efforts and share best practices.
Infrastructure Design physical places and spaces to foster socializing (e.g., public, commercial, recreational, religious). Design should
consider features of accessibility and inclusiveness across ages, abilities, and economic circumstances.
Evaluate existing infrastructure to identify barriers to social connection. Redesign, reduce or eliminate features of infrastructure
that are barriers.
Create pro-­social policies, and evaluate existing policies for barriers relevant to infrastructure (e.g., zoning laws, investing in
public transportation, housing and desegregation).
Reform policies to allow for the use of existing underutilized public spaces (e.g., schools during nights and weekends, churches
on weekdays, commercial buildings during off hours) for community social events and gatherings.
Develop programs, services and resources (e.g., recreation, volunteer programs, senior centers, community gardens) to support
more connected communities.

World Psychiatry 23:3 - October 2024 327

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
humanity. Our current decisions and actions will starkly impact Recent surges in the scientific study of social isolation and lone­
the trajectory of our future, extending across all sectors of soci­ liness have replicated and expanded upon earlier findings, provid­
ety184. ing larger sample sizes, more rigorous methodologies, and greater
confidence. However, studies have also shown that the public un­
derestimates the relevance of social connection for health relative
Recommendations to reverse trends to what has been documented in the scientific literature93,94. Since
awareness is a critical step to behavior change96, education efforts
Several countries are beginning to take steps to promote social should be prioritized as part of health promotion.
connection, and the global COVID-­19 pandemic crystalized and There has been a proliferation of interventions with promising
accelerated the urgency to act and to coordinate efforts. Table 4 results, most often improvements in loneliness. These interven­
provides a set of recommendations for national strategies to foster tions vary widely in their approaches, foci, modalities and features;​
social connection and address social isolation and loneliness. yet no one approach appears superior to others151. There are also
These recommendations align with those made by the National important limitations worth noting. Most interventions are indi­
Academy of Sciences, Engineering, and Medicine15,27,92, expert vidually focused, and attention to prevention or early interven­
consensus documents and reports, the US Surgeon General Ad­ tion is limited148. Furthermore, most reviews and meta-­analyses
visory4, the WHO191, the US Centers for Disease Control and Pre­ of the evidence only examine the effectiveness of interventions on
vention192, the American Heart Association28, and national organ- changing social outcomes (e.g., loneliness), with fewer also evalu­
izations or groups of states (e.g., European Union, UK, Australia, ating the effectiveness on changing health outcomes. Overall, the
Japan)193-­195. proposed interventions lack the level of scientific rigor of the evi­
dence that supports their need.
Existing trends in social disconnection and declining health are
CONCLUSIONS likely to persist if social factors continue to get relegated as periph­
eral to health, and interventions are only aimed at people most
In recent decades, we have witnessed a progressive decline in severely affected. In the presence of growing trends of distrust
the social connectedness of individuals and communities at the in institutions, including science, identifying and implementing
global level. At the same time, scientific evidence has been credi­ effective solutions may be challenging. Furthermore, long-­term
bly demonstrating a significant causal effect of lack of social con­ implications from the COVID-­19 pandemic and evolving digital
nection on leading physical and mental health indicators, such technologies point to potential worsening of existing trends in so­
as cardiovascular disease, stroke, depression and dementia80. In cial disconnection.
some cases, these associations are bidirectional, cyclically rein­ Looking to the future, the trajectory of social, mental and physi­
forcing poorer social connection and worse health. The strongest cal health declines is unknown, but may be accelerated. Global
evidence documents an independent directional influence of so­ scale reductions in social contact and subsequent behavioral ad­
cial connection indicators on risk for disease-­related and all-­cause aptations may reinforce sustaining social isolation or have delayed
mortality, adjusting for a robust set of demographic, lifestyle, bio­ downstream effects. Among infants, young children and adoles­
logical, and health relevant factors15,54,55. Furthermore, evidence cents, the limited social exposure at critical developmental stages
points to several plausible biological, behavioral and psychologi­ may result in longer-­term health consequences into adulthood.
cal mechanisms through which these associations of social con­ Across ages, behavioral adaptations through tools and mecha­
nection with morbidity and mortality may occur. The WHO now nisms meant to cope with isolation (e.g., remote work, streaming
recognizes social connection as a global public health priority1,13. entertainment, telehealth, contactless delivery) may instead sus­
Despite significant strengths, the body of research evidence is tain reduced social contact. Further developments in digital tech­
complex and uneven, generating several challenges. We need a nologies, such as AI, have the potential to both help and exacer­
common language to describe and measure the multiple indica­ bate the problem.
tors of social connection and its deficits. Despite the use of “lone­ Despite challenges, there is sufficient scientific evidence to
liness” as a catch-­all term, this construct is distinct from other re­ prompt action. Importantly, themes have emerged prompting rec­
lated ones (e.g., social isolation). There is convergence of evidence ommendations for individuals, communities and countries. Prior­
of the health relevance across indicators of social connection, itizing these recommendations will be critical for reversing trends
or lack thereof. Nonetheless, the relative effect sizes vary in their of social isolation and loneliness, and advancing social connection
magnitude. Social isolation appears to be a stronger predictor of to positively influence the health and well-­being of individuals and
physical health outcomes, while loneliness is a stronger predictor society at large.
of mental health outcomes58. Further, the influence of poor social
connection across its multiple components appears to be much
stronger than that of only one component42,54. Explicit recognition REFERENCES
of the separate contribution of the structure, function and quality 1. WHO Executive Board and WHO Commission on Social Connection. The im­
components of social connection is needed in measurement, risk portance of social connection to human health and well-­being. https:​//​apps.​
assessment, and health promotion. who.​int/​gb/​state​ments/​​EB154/​​PDF/​USA-​Group-​of-​count​ries-​21.​pdf.

328 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2. Leary MR, Baumeister R. The need to belong. Psychol Bull 1995;​117:​497-529. 30. Huang J, Xiao L, Zhao H et al. Living alone increases the risk of developing type
3. Cacioppo JT, Cacioppo S, Cole SW et al. Loneliness across phylogeny and a call 2 diabetes mellitus:​a systematic review and meta-­analysis based on longitudi­
for comparative studies and animal models. Perspect Psychol Sci 2015;​10:​​202- nal studies. Prim Care Diabetes 2024;​18:​1-6.
12. 31. Park C, Majeed A, Gill H et al. The effect of loneliness on distinct health out­­
4. Office of the US Surgeon General. Our epidemic of loneliness and isolation:​the comes:​a comprehensive review and meta-­analysis. Psychiatry Res 2020;​294:​​
US Surgeon General’s Advisory on the healing effects of social connection and 113514.
community. www.​hhs.​gov. 32. Zhao Y, Guyatt G, Gao Y et al. Living alone and all-­cause mortality in com­
5. Snyder-­Mackler N, Burger JR, Gaydosh L et al. Social determinants of health munity-­dwelling adults:​a systematic review and meta-­analysis. eClinicalMed­
and survival in humans and other animals. Science 2020;​368:​eaax9553. icine 2022;​54:​101677.
6. Sapolsky RM. Social status and health in humans and other animals. Annu Rev 33. Wang F, Gao Y, Han Z et al. A systematic review and meta-­analysis of 90 cohort
Anthropol 2004;​33:​393-418. studies of social isolation, loneliness and mortality. Nat Hum Behav 2023;​7:​
7. Badcock JC, Holt-­Lunstad J, Garcia E et al. Position Statement:​Addressing so­ 1307-19.
cial isolation and loneliness and the power of human connection. Global Ini­ 34. Wang Y, Jiao Y, Nie J et al. Sex differences in the association between marital
tiative on Loneliness and Connection, 2022. status and the risk of cardiovascular, cancer, and all-­cause mortality:​a system­
8. Holt-­Lunstad J. Why social relationships are important for physical health:​a atic review and meta-­analysis of 7,881,040 individuals. Glob Health Res Policy
systems approach to understanding and modifying risk and protection. Annu 2020;​5:​4.
Rev Psychol 2018;​69:​437-58. 35. Shor E, Roelfs DJ, Yogev T. The strength of family ties:​a meta-­analysis and
9. Walker P. May appoints minister to tackle loneliness issues raised by Jo Cox. meta-­regression of self-­reported social support and mortality. Social Networks
The Guardian, January 16, 2018. 2013;​35:​626-38.
10. Kodama S. Japan appoints ‘minister of loneliness’ to help people home alone. 36. Shor E, Roelfs DJ, Curreli M et al. Widowhood and mortality:​a meta-­analysis
Nikkei Asia, February 13, 2021. and meta-­regression. Demography 2012;​49:​575-606.
11. European Commission. Loneliness publications. https:​//​joint-​resea​rch-​cen­ 37. Shor E, Roelfs DJ, Bugyi P et al. Meta-­analysis of marital dissolution and mor­
tre.​ec.​europa.​eu/​scien​tific-​activ​ities-z/​lonel​iness/​​lonel​iness-​publi​catio​ns_​en. tality:​reevaluating the intersection of gender and age. Soc Sci Med 2012;​75:​46-
12. Kim H, Ng KW. Hikikomori:​why S Korea is paying young recluses to leave 59.
home. https:​//​www.​bbc.​com/​news/​world-​asia-​65554271. 38. Shor E, Roelfs DJ. Social contact frequency and all-­cause mortality:​a meta-­
13. World Health Organization. WHO Commission on Social Connection. https:​//​ analysis and meta-­regression. Soc Sci Med 2015;​128:​76-86.
www.​who.​int/​groups/​commi​ssion-​on-​social-​conne​ction​. 39. Roelfs DJ, Shor E, Kalish R et al. The rising relative risk of mortality for singles:​
14. Sbarra DA, Ramadan FA, Choi KW et al. Loneliness and depression:​bidirec­ meta-­analysis and meta-­regression. Am J Epidemiol 2011;​174:​379-89.
tional Mendelian randomization analyses using data from three large genome-­ 40. Rico-­Uribe LA, Caballero FF, Martin-­Maria N et al. Association of loneliness
wide association studies. Mol Psychiatry 2023;​28:​4594-601. with all-­cause mortality:​a meta-­analysis. PLoS One 2018;​13:​e0190033.
15. National Academies of Sciences, Engineering, and Medicine. Social isolation 41. Pinquart M, Duberstein PR. Associations of social networks with cancer mor­
and loneliness in older adults:​opportunities for the health care system. Wash­ tality:​a meta-­analysis. Crit Rev Oncol Hematol 2010;​75:​122-37.
ington:​National Academies Press, 2020. 42. Holt-­Lunstad J, Smith TB, Layton JB. Social relationships and mortality risk:​a
16. Wickramaratne PJ, Yangchen T, Lepow L et al. Social connectedness as a deter­ meta-­analytic review. PLoS Med 2010;​7:​e1000316.
minant of mental health:​a scoping review. PLoS One 2022;​17:​e0275004. 43. Holt-­Lunstad J, Smith TB, Baker M et al. Loneliness and social isolation as risk
17. Mykyta L. Living alone and feelings of depression among adults age 18 and factors for mortality:​a meta-­analytic review. Perspect Psychol Sci 2015;​​10:​227-
older. Natl Health Stat Report 2024;​199:​1-11. 37.
18. Mann F, Wang J, Pearce E et al. Loneliness and the onset of new mental health 44. Uchino BN, Landvatter J, Zee K et al. Social support and antibody responses to
problems in the general population. Soc Psychiatry Psychiatr Epidemiol 2022;​​ vaccination:​a meta-­analysis. Ann Behav Med 2020;​54:​567-74.
57:​2161-78. 45. World Health Organization. Cardiovascular diseases. www.​who.​int.
19. Wray NR, Ripke S, Mattheisen M et al. Genome-­wide association analyses 46. Yang YC, Boen C, Gerken K et al. Social relationships and physiological deter­
iden­­tify 44 risk variants and refine the genetic architecture of major depression. minants of longevity across the human life span. Proc Natl Acad Sci USA 2016;​
Nat Genet 2018;​50:​668-81. 113:​578-83.
20. Levey DF, Stein MB, Wendt FR et al. Bi-­ancestral depression GWAS in the Mil­ 47. World Health Organization. Diabetes. www.​who.​int.
lion Veteran Program and meta-­analysis in >1.2 million individuals highlight 48. Brinkhues S, Dukers-­Muijrers N, Hoebe C et al. Socially isolated individuals are
new therapeutic directions. Nat Neurosci 2021;​24:​954-63. more prone to have newly diagnosed and prevalent type 2 diabetes mellitus –
21. Bedaso A, Adams J, Peng W et al. The relationship between social support and the Maastricht study. BMC Public Health 2017;​17:​955.
mental health problems during pregnancy:​a systematic review and meta-­ 49. Brinkhues S, Dukers-­Muijrers N, Hoebe C et al. Social network characteristics
analysis. Reprod Health 2021;​18:​162. are associated with type 2 diabetes complications:​the Maastricht study. Diabe­
22. Serrano-­Ripoll MJ, Meneses-­Echavez JF, Ricci-­Cabello I et al. Impact of viral tes Care 2018;​41:​1654-62.
epidemic outbreaks on mental health of healthcare workers:​a rapid systema­t­ 50. Song Y, Nam S, Park S et al. The impact of social support on self-­care of patients
ic review and meta-­analysis. J Affect Disord 2020;​277:​347-57. with diabetes:​what is the effect of diabetes type? Systematic review and meta-­
23. Penninkilampi R, Casey A-­N, Singh MF et al. The association between social analysis. Diabetes Educ 2017;​43:​396-412.
engagement, loneliness, and risk of dementia:​a systematic review and meta-­ 51. Manemann SM, Chamberlain AM, Roger VL et al. Perceived social isolation
analysis. J Alzheimers Dis 2018;​66:​1619-33. and outcomes in patients with heart failure. J Am Heart Assoc 2018;​7:​e008069.
24. Lara E, Martin-­Maria N, De la Torre-­Luque A et al. Does loneliness contribute 52. Heidari Gorji MA, Fatahian A, Farsavian A. The impact of perceived and ob­
to mild cognitive impairment and dementia? A systematic review and meta-­ jective social isolation on hospital readmission in patients with heart failure:​a
analysis of longitudinal studies. Ageing Res Rev 2019;​52:​7-16. systematic review and meta-­analysis of observational studies. Gen Hosp Psy­
25. Kuiper JS, Zuidersma M, Oude Voshaar RC et al. Social relationships and risk of chiatry 2019;​60:​27-36.
dementia:​a systematic review and meta-­analysis of longitudinal cohort stud­ 53. Motillon-­Toudic C, Walter M, Séguin M et al. Social isolation and suicide risk:​
ies. Ageing Res Rev 2015;​22:​39-57. literature review and perspectives. Eur Psychiatry 2022;​65:​e65.
26. Duffner LA, Deckers K, Cadar D et al. The role of cognitive and social leisure 54. Foster HME, Gill JMR, Mair FS et al. Social connection and mortality in UK
activities in dementia risk:​assessing longitudinal associations of modifiable Biobank:​a prospective cohort analysis. BMC Med 2023;​21:​384.
and non-­modifiable risk factors. Epidemiol Psychiatr Sci 2022;​31:​e5. 55. Elovainio M, Hakulinen C, Pulkki-­Raback L et al. Contribution of risk factors to
27. National Academies of Sciences, Engineering, and Medicine. Enhancing com­ excess mortality in isolated and lonely individuals:​an analysis of data from the
munity resilience through social capital and connectedness:​stronger together! UK Biobank cohort study. Lancet Public Health 2017;​2:​e260-6.
Washington:​National Academies Press, 2021. 56. Valtorta NK, Kanaan M, Gilbody S et al. Loneliness, social isolation and risk of
28. Cené CW, Beckie TM, Sims M et al. Effects of objective and perceived social cardiovascular disease in the English Longitudinal Study of Ageing. Eur J Prev
isolation on cardiovascular and brain health:​a scientific statement from the Cardiol 2018;​25:​1387-96.
American Heart Association. J Am Heart Assoc 2022;​11:​e026493. 57. Hakulinen C, Pulkki-­Raback L, Virtanen M et al. Social isolation and loneliness
29. Valtorta NK, Kanaan M, Gilbody S et al. Loneliness and social isolation as risk as risk factors for myocardial infarction, stroke and mortality:​UK Biobank co­
factors for coronary heart disease and stroke:​systematic review and meta-­ hort study of 479 054 men and women. Heart 2018;​104:​1536-42.
analysis of longitudinal observational studies. Heart 2016;​102:​1009-16. 58. Hong JH, Nakamura JS, Berkman LF et al. Are loneliness and social isolation

World Psychiatry 23:3 - October 2024 329

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
equal threats to health and well-­being? An outcome-­wide longitudinal ap­ com/​analy​tics/​509675/​state-​of-​social-​conne​ctions.​aspx.
proach. SSM Popul Health 2023;​23:​101459. 87. Kannan VD, Veazie PJ. US trends in social isolation, social engagement, and
59. Uchino BN. Social support and health:​a review of physiological processes po­ companionship – nationally and by age, sex, race/ethnicity, family income,
tentially underlying links to disease outcomes. J Behav Med 2006;​29:​377-87. and work hours, 2003-­2020. SSM Popul Health 2022;​21:​101331.
60. Hostinar CE, Sullivan RM, Gunnar MR. Psychobiological mechanisms under­ 88. Koscielniak M, Bojanowska A, Gasiorowska A. Religiosity decline in Europe:​
lying the social buffering of the hypothalamic-­pituitary-­adrenocortical axis:​a age, generation, and the mediating role of shifting human values. J Relig Health
review of animal models and human studies across development. Psychol Bull 2024;​63:​1091-116.
2014;​140:​256-82. 89. Höllinger F, Makula L. Religiosity in the major religious cultures of the world.
61. Gunnar MR, Hostinar CE. The social buffering of the hypothalamic-­pituitary-­ Int J Sociol 2021;​51:​345-59.
adrenocortical axis in humans:​developmental and experiential determinants. 90. Smock PJ, Schwartz CR. The demography of families:​a review of patterns and
Soc Neurosci 2015;​10:​479-88. change. J Marriage Fam 2020;​82:​9-34.
62. Cole SW. Human social genomics. PLoS Genet 2014;​10:​e1004601. 91. Barreto M, Qualter P, Doyle D. Loneliness inequalities evidence review. www.​
63. Eisenberger NI, Moieni M, Inagaki TK et al. In sickness and in health:​the co-­ wcpp.​org.​uk.
regulation of inflammation and social behavior. Neuropsychopharmacology 92. National Academies of Sciences, Engineering, and Medicine. Social media and
2017;​42:​242-53. adolescent health. Washington:​National Academies Press, 2024.
64. Uchino BN, Bowen K, Kent de Grey R et al. Social support and physical health:​ 93. Haslam SA, McMahon C, Cruwys T et al. Social cure, what social cure? The pro­­­­
models, mechanisms, and opportunities. In:​Fisher EB, Cameron LD, Chris­ pensity to underestimate the importance of social factors for health. Soc Sci
tensen AJ et al (eds). Principles and concepts of behavioral medicine:​a global Med 2018;​198:​14-21.
handbook. New York:​Springer, 2018:​341-72. 94. Holt-­Lunstad J, Proctor AS. Blind spots in health perception:​underestimation
65. Cohen S, McKay G. Social support, stress and the buffering hypothesis:​a theo­ of social connection as a determinant of health. Submitted for publication.
retical analysis. In:​Taylor SE, Singer JE, Baum A (eds). Handbook of psychology 95. Melfi CA, Drake BG, Tierney WM. The role of public opinion in drug resource
and health, Vol. 4. London:​Routledge, 2020:​253-67. allocation decisions. Pharmacoeconomics 1996;​9:​106-12.
66. Kent de Grey RG, Uchino BN, Trettevik R et al. Social support and sleep:​a 96. Rosenstock IM. The health belief model and preventive health behavior.
meta-­analysis. Health Psychol 2018;​37:​787-98. Health Educ Monogr 1974;​2:​354-86.
67. Shankar A, McMunn A, Banks J et al. Loneliness, social isolation, and behavior­ 97. Kerr NA, Stanley TB. Revisiting the social stigma of loneliness. Pers Individ Dif­
al and biological health indicators in older adults. Health Psychol 2011;​30:​​377. fer 2021;​171:​110482.
68. Uchino BN, Trettevik R, Kent de Grey RG et al. Social support, social integra­ 98. World Health Organization. Constitution. www.​who.​int.
tion, and inflammatory cytokines:​a meta-­analysis. Health Psychol 2018;​37:​​ 99. Prohaska T, Burholt V, Burns A et al. Consensus statement:​loneliness in older
462-71. adults, the 21st century social determinant of health? BMJ Open 2020;​​10:​
69. Hanna K, Cross J, Nicholls A et al. The association between loneliness or social e034967.
isolation and food and eating behaviours:​a scoping review. Appetite 2023;​191:​​ 100. Perlman D, Peplau LA. Toward a social psychology of loneliness. In:​Gilmour R,
107051. Duck S (eds). Personal relationships. London:​Academic Press, 1981:​31-56.
70. Miller J, King J, Trivers K. Vital signs:​mammography use and association with 101. O’Sullivan R, Burns A, Leavey G et al. Impact of the COVID-­19 pandemic on
social determinants of health and health-­related social needs among women – loneliness and social isolation:​a multi-­country study. Int J Environ Res Public
United States, 2022. MMWR Morb Mortal Wkly Rep 2024;​73:​351-7. Health 2021;​18:​9982.
71. DiMatteo MR. Social support and patient adherence to medical treatment:​a 102. Serageldin I, Grootaert C. Defining social capital:​an integrating view 1. In:​
meta-­analysis. Health Psychol 2004;​23:​207-18. Wiesner E (ed). Evaluation and development. New York:​Routledge, 2017:​201-
72. Magrin ME, D’Addario M, Greco A et al. Social support and adherence to treat­ 17.
ment in hypertensive patients:​a meta-­analysis. Ann Behav Med 2015;​49:​​307- 103. Moore S, Kawachi I. Twenty years of social capital and health research:​a glos­
18. sary. J Epidemiol Community Health 2017;​71:​513-7.
73. Shahin W, Kennedy GA, Stupans I. The association between social support 104. Valtorta NK, Kanaan M, Gilbody S et al. Loneliness, social isolation and social
and medication adherence in patients with hypertension:​a systematic review. relationships:​what are we measuring? A novel framework for classifying and
Pharm Pract 2021;​19:​2300. comparing tools. BMJ Open 2016;​6:​e010799.
74. Adisa R, Olajide OO, Fakeye TO. Social support, treatment adherence and out­ 105. Holt-­Lunstad J, Robles TF, Sbarra DA. Advancing social connection as a public
come among hypertensive and type 2 diabetes patients in ambulatory care set­ health priority in the United States. Am Psychol 2017;​72:​517-30.
tings in Southwestern Nigeria. Ghana Med J 2017;​51:​64-77. 106. US National Center for Advancing Translational Sciences. New therapeutic
75. Miller TA, Dimatteo MR. Importance of family/social support and impact on uses. https:​//​ncats.​nih.​gov/​resea​rch/​resea​rch-​activ​ities/​​ntu.
adherence to diabetic therapy. Diabetes Metab Syndr Obes 2013;​6:​421-6. 107. Lasgaard M, Løvschall C, Qualter P et al. Are loneliness interventions effec­
76. Smith TB, Workman C, Andrews C et al. Effects of psychosocial support inter­ tive in reducing loneliness? A meta-­analytic review of 128 studies. Eur J Public
ventions on survival in inpatient and outpatient healthcare settings:​a meta-­ Health 2022;​32(Suppl. 3):​ckac129.266.
analysis of 106 randomized controlled trials. PLoS Med 2021;​18:​e1003595. 108. Beckers A, Buecker S, Casabianca E et al. Effectiveness of interventions tackling
77. Hill AB. The environment and disease:​association or causation? Proc R Soc loneliness. Luxembourg:​Publications Office of the European Union, 2022.
Med 1965;​​58:​295-300. 109. Collins LM, Baker TB, Mermelstein RJ et al. The multiphase optimization strat­
78. Howick J, Kelly P, Kelly M. Establishing a causal link between social relation­ egy for engineering effective tobacco use interventions. Ann Behav Med 2011;​​
ships and health using the Bradford Hill guidelines. SSM Popul Health 2019;​8:​ 41:​208-26.
100402. 110. Collins LM, Murphy SA, Nair VN et al. A strategy for optimizing and evaluating
79. Holt-­Lunstad J. The major health implications of social connection. Curr Dir behavioral interventions. Ann Behav Med 2005;​30:​65-73.
Psychol Sci 2021;​30:​251-9. 111. Shah SGS, Nogueras D, van Woerden HC et al. Evaluation of the effectiveness
80. Holt-­Lunstad J. Social connection as a public health issue:​the evidence and of digital technology interventions to reduce loneliness in older adults:​system­
a systemic framework for prioritizing the “social” in social determinants of atic review and meta-­analysis. J Med Internet Res 2021;​23:​e24712.
health. Annu Rev Public Health 2022;​43:​193-213. 112. McElfresh JJ, Skiba MB, Segrin CG et al. Interventions for loneliness among
81. VanderWeele TJ. Can sophisticated study designs with regression analyses of adult cancer survivors:​a systematic review and meta-­analysis. J Psychosoc
observational data provide causal inferences? JAMA Psychiatry 2021;​78:​244-6. Oncol 2021;​39:​509-33.
82. Shi X, Pan Z, Miao W. Data integration in causal inference. Wiley Interdiscip 113. Hickin N, Kall A, Shafran R et al. The effectiveness of psychological interven­
Rev Comput Stat 2023;​15:​e1581. tions for loneliness:​a systematic review and meta-­analysis. Clin Psychol Rev
83. Sanderson E, Glymour MM, Holmes MV et al. Mendelian randomization. Nat 2021;​88:​102066.
Rev Methods Primers 2022;​2:​6. 114. Choi M, Kong S, Jung D. Computer and internet interventions for loneliness
84. Buecker S, Mund M, Chwastek S et al. Is loneliness in emerging adults increas­ and depression in older adults:​a meta-­analysis. Healthc Inform Res 2012;​18:​
ing over time? A preregistered cross-­temporal meta-­analysis and systematic 191-8.
review. Psychol Bull 2021;​147:​787-805. 115. Eccles AM, Qualter P. Alleviating loneliness in young people – a meta-­analysis
85. Surkalim DL, Luo M, Eres R et al. The prevalence of loneliness across 113 coun­ of interventions. Child Adolesc Ment Health 2021;​26:​17-33.
tries:​systematic review and meta-­analysis. BMJ 2022;​376:​e067068. 116. Veronese N, Galvano D, D’Antiga F et al. Interventions for reducing loneliness:​
86. Gallup & Meta. The global state of social connections. https:​//​www.​gallup.​ an umbrella review of intervention studies. Health Soc Care Community 2021;​

330 World Psychiatry 23:3 - October 2024

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
29:​e89-96. for the post-­COVID era. Council of State Governments, 2022.
117. Ellard OB, Dennison C, Tuomainen H. Interventions addressing loneliness 145. Lim MH, Eres R, Vasan S. Understanding loneliness in the twenty-­first century:​
amongst university students:​a systematic review. Child Adolesc Ment Health an update on correlates, risk factors, and potential solutions. Soc Psychiatry
2023;​28:​512-23. Psychiatr Epidemiol 2020;​55:​793-810.
118. Gardiner C, Geldenhuys G, Gott M. Interventions to reduce social isolation and 146. Akhter-­Khan SC, Au R. Why loneliness interventions are unsuccessful:​a call for
loneliness among older people:​an integrative review. Health Soc Care Com­ precision health. Adv Geriatr Med Res 2020;​2:​e200016.
munity 2018;​26:​147-57. 147. Barreto M, Doyle DM, Qualter P. Changing the narrative:​loneliness as a social
119. O’Rourke HM, Collins L, Sidani S. Interventions to address social connected­ justice issue. Polit Psychol 2024;​45:​157-81.
ness and loneliness for older adults:​a scoping review. BMC Geriatr 2018;​18:​​ 148. Welch V, Ghogomu E, Dowling S et al. In-­person interventions to reduce social
214. isolation and loneliness:​an evidence and gap map. Campbell Syst Rev 2023;​19:​
120. Spiegel D, Bloom JR, Kraemer HC et al. Effect of psychosocial treatment on sur­ e1340.
vival of patients with metastatic breast cancer. Lancet 1989;​2:​888-91. 149. Meehan DE, Grunseit A, Condie J et al. Social-­ecological factors influencing
121. Holt-­Lunstad J, Perissinotto C. Social isolation and loneliness as medical is­ loneliness and social isolation in older people:​a scoping review. BMC Geriatr
sues. N Engl J Med 2023;​388:​193-5. 2023;​23:​726.
122. Reinhardt GY, Vidovic D, Hammerton C. Understanding loneliness:​a system­ 150. World Health Organization. Health in all policies:​Helsinki statement. Frame­
atic review of the impact of social prescribing initiatives on loneliness. Perspect work for country action. https:​//​iris.​who.​int.
Public Health 2021;​141:​204-13. 151. Nurminen M, Thoma-Otremba A, Casabianca E. Mapping of loneliness inter­
123. Costa A, Sousa CJ, Seabra PRC et al. Effectiveness of social prescribing pro­ ventions in the EU. European Commission, 2023.
grams in the primary health-­care context:​a systematic literature review. Sus­ 152. Caspi A, Harrington H, Moffitt TE et al. Socially isolated children 20 years later:​
tainability 2021;​13:​2731. risk of cardiovascular disease. Arch Pediatr Adolesc Med 2006;​160:​805-11.
124. Liz B, Alison B, Paul MW et al. Social prescribing:​less rhetoric and more reality. 153. Levine S, Malone E, Lekiachvili A et al. Health care industry insights:​why the
A systematic review of the evidence. BMJ Open 2017;​7:​e013384. use of preventive services is still low. Prev Chronic Dis 2019;​16:​E30.
125. Ibarra F, Baez M, Cernuzzi L et al. A systematic review on technology-­sup­ 154. Bianco-­Miotto T, Craig J, Gasser Y et al. Epigenetics and DOHaD:​from basics to
ported interventions to improve old-­age social wellbeing:​loneliness, social birth and beyond. J Dev Orig Health Dis 2017;​8:​513-9.
isolation, and connectedness. J Healthc Eng 2020;​2020:​2036842. 155. Hildreth JR, Vickers MH, Buklijas T et al. Understanding the importance of the
126. Wister A, Fyffe I, O’Dea E. Technological interventions for loneliness and social early-­life period for adult health:​a systematic review. J Dev Orig Health Dis
isolation among older adults:​a scoping review protocol. Syst Rev 2021;​​10:​217. 2023;​14:​166-74.
127. Welch V, Ghogomu ET, Barbeau VI et al. Digital interventions to reduce social 156. Huang CX, Halfon N, Sastry N et al. Positive childhood experiences and adult
isolation and loneliness in older adults:​an evidence and gap map. Campbell health outcomes. Pediatrics 2023;​152:​e2022060951.
Syst Rev 2023;​19:​e1369. 157. Abbott-­Chapman J, Gall S, Ollington N et al. The association between child­
128. Hawkley LC, Finch LE, Kotwal AA et al. Can remote social contact replace in-­ hood school engagement and attainment and adult education and health out­
person contact to protect mental health among older adults? J Am Geriatr Soc comes:​preliminary findings from an interdisciplinary research project using
2021;​69:​3063-5. longitudinal Australian cohort data. Presented at the Australian Association
129. Marraccini ME, Brier ZM. School connectedness and suicidal thoughts and for Research in Education (AARE) International Research Conference, Hobart,
behaviors:​a systematic meta-­analysis. Sch Psychol Q 2017;​32:​5-21. November 2011.
130. Corcoran RP, Cheung ACK, Kim E et al. Effective universal school-­based social 158. Abbott-­Chapman J, Martin K, Ollington N et al. The longitudinal association
and emotional learning programs for improving academic achievement:​a sys­ of childhood school engagement with adult educational and occupational
tematic review and meta-­analysis of 50 years of research. Educ Res Rev 2018;​ achievement:​findings from an Australian national study. Br Educ Res J 2014;​
25:​56-72. 40:​102-20.
131. Allen K, Kern ML, Vella-­Brodrick D et al. What schools need to know about 159. Bethell C, Jones J, Gombojav N et al. Positive childhood experiences and adult
fostering school belonging:​a meta-­analysis. Educ Psychol Rev 2018;​30:​1-34. mental and relational health in a statewide sample:​associations across adverse
132. Wilkins NJ, Verlenden JMV, Szucs LE et al. Classroom management and facili­ childhood experiences levels. JAMA Pediatr 2019;​173:​e193007.
tation approaches that promote school connectedness. J Sch Health 2023;​93:​ 160. Fletcher HK, Gallichan DJ. An overview of attachment theory:​Bowlby and be­
582-93. yond. In:​Fletcher HK, Flood A, Hare DJ (eds). Attachment in intellectual and
133. Wigelsworth M, Verity L, Mason C et al. Social and emotional learning in pri­ developmental disability:​a clinician’s guide to practice and research. New
mary schools:​a review of the current state of evidence. Br J Educ Psychol 2022;​​ York:​Wiley Blackwell, 2016:​8-32.
92:​898-924. 161. Hughes K, Bellis MA, Hardcastle KA et al. The effect of multiple adverse child­
134. Weare K, Nind M. Mental health promotion and problem prevention in hood experiences on health:​a systematic review and meta-­analysis. Lancet
schools:​what does the evidence say? Health Promot Int 2011;​26(Suppl.1):​i29- Public Health 2017;​2:​e356-66.
69. 162. Hanson MA, Gluckman P. Early developmental conditioning of later health
135. Weatherson KA, O’Neill M, Lau EY et al. The protective effects of school con­ and disease:​physiology or pathophysiology? Physiol Rev 2014;​94:​1027-76.
nectedness on substance use and physical activity. J Adolesc Health 2018;​63:​ 163. Chen Y, Cortes P, Kosar G et al. The impact of COVID-­19 on workers’ expecta­
724-31. tions and preferences for remote work. American Economic Association (AEA)
136. Steiner RJ, Sheremenko G, Lesesne C et al. Adolescent connectedness and Papers and Proceedings 2023;​113:​556-61.
adult health outcomes. Pediatrics 2019;​144:​e20183766. 164. Jiang Y, Lai P-­L, Yang C-­C et al. Exploring the factors that drive consumers to
137. Hawkley L. Public policy and the reduction and prevention of loneliness and use contactless delivery services in the context of the continued COVID-­19
social isolation. In:​Hajek A, Riedel-­Heller SG, König H-­H (eds). Loneliness and pandemic. J Retail Consum Serv 2023;​72:​103276.
social isolation in old age. London:​Routledge, 2023:​181-90. 165. Kim S, Jang S, Choi W et al. Contactless service encounters among Millennials
138. Holt-­Lunstad J. Social isolation and health. Health Affairs, June 21, 2020. and Generation Z:​the effects of Millennials and Gen Z characteristics on tech­
139. Lim MH, Qualter P, Ding D et al. Advancing loneliness and social isolation as nology self-­efficacy and preference for contactless service. J Res Interact Mark
a global health priority:​taking three priority actions. Public Health Res Pract 2022;​16:​82-100.
2023;​33:​3332320. 166. Ahmed A, Abdulkareem AM. Big data analytics in the entertainment industry:​
140. Rudolph L, Caplan J, Ben-­Moshe K et al. Health in all policies:​a guide for state audience behavior analysis, content recommendation, and revenue maximi­
and local governments. Washington and Oakland:​American Public Health As­ zation. Reviews of Contemporary Business Analytics 2023;​6:​88-102.
sociation and Public Health Institute, 2013. 167. Camilleri MA, Falzon L. Understanding motivations to use online streaming
141. Mahoney JL, Weissberg RP, Greenberg MT et al. Systemic social and emotional services:​integrating the technology acceptance model (TAM) and the uses and
learning:​promoting educational success for all preschool to high school stu­ gratifications theory (UGT). Span J Mark -­ESIC 2021;​25:​217-38.
dents. Am Psychol 2021;​76:​1128-42. 168. Oakman J, Kinsman N, Stuckey R et al. A rapid review of mental and physical
142. Department of Agriculture, Trade and Consumer Protection, State of Wiscon­ health effects of working at home:​how do we optimise health? BMC Public
sin. Farmer Wellness Program. https:​//​datcp.​wi.​gov. Health 2020;​20:​1825.
143. Ali MM, West K, Teich JL et al. Utilization of mental health services in educa­ 169. Holt-­Lunstad J. Fostering social connection in the workplace. Am J Health Pro­
tional setting by adolescents in the United States. J Sch Health 2019;​89:​393-401. mot 2018;​32:​1307-12.
144. Kirby J, Slone S. Social isolation and loneliness:​State policies and interventions 170. Granovetter MS. The strength of weak ties. Am J Sociol 1973;​78:​1360-80.

World Psychiatry 23:3 - October 2024 331

 20515545, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/wps.21224 by Nat Prov Indonesia, Wiley Online Library on [23/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
171. Kim M, Fernandez RM. What makes weak ties strong? Annu Rev Sociol 2023;​ 184. Capraro V, Lentsch A, Acemoglu D et al. The impact of generative artificial
49:​177-93. intelligence on socioeconomic inequalities and policy making. arXiv 2023;​
172. Predmore ZS, Roth E, Breslau J et al. Assessment of patient preferences for tele­ 2401.05377.
health in post-­COVID-­19 pandemic health care. JAMA Netw Open 2021;​4:​ 185. Campbell J. AI’s future:​utopia or dystopia? Experts weigh in on five possible
e2136405. outcomes. https:​//​ainew​stoday.​co.​uk.
173. Rodriguez M, Pratt S, Bellet BW et al. Solitude can be good – if you see it as 186. Jones VK, Hanus M, Yan C et al. Reducing loneliness among aging adults:​the
such:​reappraisal helps lonely people experience solitude more positively. J roles of personal voice assistants and anthropomorphic interactions. Front
Pers 2023;​ doi:​ 10.​1111/​jopy.​12887​. Public Health 2021;​9:​750736.
174. Hoppmann CA, Lay JC, Pauly T et al. Social isolation, loneliness, and solitude 187. Xie T, Pentina I. Attachment theory as a framework to understand relationships
in older adulthood. In:​Coplan RJ, Bowker JC, Nelson LJ (eds). The handbook of with social chatbots:​a case study of Replika. Proceedings of the 55th Hawaii
solitude. Hoboken:​Wiley Blackwell, 2021:​178-89. International Conference on System Sciences, Maui, January 2022.
175. Haney C. Restricting the use of solitary confinement. Annu Rev Criminol 2018;​ 188. Law T, Chita-­Tegmark M, Rabb N et al. Examining attachment to robots:​ben­
1:​285-310. efits, challenges, and alternatives. ACM Trans Hum Robot Interact 2022;​11:​1-
176. Rahbari L. COVID-­19 pandemic and the crisis of care:​wellness discourses, 18.
neoliberal self-­care, and (dis)infodemic. Soc Sci 2023;​12:​137. 189. Baker RS, Hawn A. Algorithmic bias in education. Int J Artificial Intell Educ 2022;​​
177. Baker SA. Alt. Health Influencers:​how wellness culture and web culture have 32:​1-41.
been weaponised to promote conspiracy theories and far-­right extremism dur­ 190. Morewedge CK, Mullainathan S, Naushan HF et al. Human bias in algorithm
ing the COVID-­19 pandemic. Eur J Cult Stud 2022;​25:​3-24. design. Nat Hum Behav 2023;​7:​1822-4.
178. Office of the US Surgeon General. Social media and youth mental health:​the 191. World Health Organization. Social isolation and loneliness among older peo­
US Surgeon General’s Advisory. www.​hhs.​gov. ple:​advocacy brief. Geneva:​World Health Organization, 2021.
179. Twenge JM, Haidt J, Joiner TE et al. Underestimating digital media harm. Nat 192. US Centers for Disease Control and Prevention. Social connectedness. www.​
Hum Behav 2020;​4:​346-8. cdc.​gov.
180. Odgers CL, Schueller SM, Ito M. Screen time, social media use, and adolescent 193. Kovacic M, Mauri C, Nurminen M et al. Key messages and policy recommen­
development. Annu Rev Develop Psychol 2020;​2:​485-502. dations of the one day conference on “Loneliness in the European Union:​poli­
181. Torous J. A path towards progress:​lessons from the hard things about digital cies at work”. Luxembourg:​Publications Office of the European Union, 2023.
mental health. World Psychiatry 2022;​21:​419-20. 194. Lim MH, Smith B. State of the nation report:​social connections in Australia
182. Valkenburg PM, Meier A, Beyens I. Social media use and its impact on ado­ 2023. http:​//​www.​endin​glone​liness.​com.​au.
lescent mental health:​an umbrella review of the evidence. Curr Opin Psychol 195. Fukuoka R, Nagano Y. Law enacted to battle growing problem of loneliness.
2022;​44:​58-68. The Asahi Shimbun, April 2, 2024.
183. Torous J, Myrick K, Aguilera A. The need for a new generation of digital men­
tal health tools to support more accessible, effective and equitable care. World DOI:​10.1002/wps.21224
Psychiatry 2023;​22:​1-2.

332 World Psychiatry 23:3 - October 2024