Journal of Affective Disorders

Journal of Affective Disorders 175 (2015) 5365
Contents lists available at ScienceDirect
Journal of Affective Disorders

journal homepage: www.elsevier.com/locate/jad
Review
The association between social relationships and depression:

A systematic review
Ziggi Ivan Santini a,n, Ai Koyanagi a, Stefanos Tyrovolas a, Catherine Mason b,
Josep Maria Haro a
a
Parc Sanitari Sant Joan de Du, Universitat de Barcelona, Fundaci Sant Joan de Du/CIBERSAM, Dr Antoni Pujades, 42, 08830, Sant Boi de Llobregat,
Barcelona, Spain
b
Faculty of Social and Human Sciences, Academic Unit Psychology, Building 44, Higheld Campus, University of Southampton, Southampton SO17 1BJ, UK
art ic l e i nf o
a b s t r a c t
Article history:
Received 30 September 2014
Received in revised form
17 December 2014
Accepted 19 December 2014
Available online 31 December 2014
Background: Depression is one of the most prevalent mental disorders globally and has implications for
various aspects of everyday-life. To date, studies assessing the association between social relationships
and depression have provided conicting results. The aim of this paper was to review the evidence on
associations between social relationships and depression in the general population.
Methods: Studies investigating the association of social support, social networks, or social connectedness
with depression were retrieved and summarized (searches using Pubmed, ScienceDirect, PsycNet were
conducted in May 2014).
Results: Fifty-one studies were included in this review. The strongest and most consistent ndings were
signicant protective effects of perceived emotional support, perceived instrumental support, and large,
diverse social networks. Little evidence was found on whether social connectedness is related to
depression, as was also the case for negative interactions.
Limitations: Due to the strict inclusion criteria relating to study quality and the availability of papers in
the domain of interest, the review did not capture gray literature and qualitative studies.
Conclusion: Future research is warranted to account for potential bias introduced by the use of subjective
measures as compared to objective measures of received support and actual networks. Due to the
heterogeneity between available studies on the measure of social relationships, the inclusion of
comparable measures across studies would allow for more valid comparisons. In addition, welldesigned prospective studies will provide more insight into causality. Future research should address
how social support and networks interact and together affect risks for depression. Social connectedness
and negative interactions appear to be underutilized as measures in population-based studies.
& 2014 Elsevier B.V. All rights reserved.
Keywords:
Social support
Social networks
Social connectedness
Depression
Contents
1.
2.
3.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.
Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.
Data extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.
Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.
Social support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.
Social networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.
Social connectedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.4.
Other salient ndings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.5.
Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.6.
Intrapersonal characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Corresponding author.
E-mail address: [email protected] (Z.I. Santini).
http://dx.doi.org/10.1016/j.jad.2014.12.049
0165-0327/& 2014 Elsevier B.V. All rights reserved.
54
55
55
55
56
56
56
58
60
60
60
60
54
Z.I. Santini et al. / Journal of Affective Disorders 175 (2015) 5365
3.7.
Chronic physical illness and disability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.
Social support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.
Social networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.
Social connectedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.
Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.5.
Intrapersonal characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.6.
Chronic physical illness and disability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.7.
Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Role of funding source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conict of interest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.
1. Introduction
Depression is arguably one of the most prevalent and prominent
mental disorders of our time, touching the lives of people across
nations, ages, and social and cultural groups. The Global Burden of
Disease Study 2010 has identied depressive disorders as the second
leading cause of years lived with disability, designating it a as a major
public-health priority (Ferrari et al., 2013). Besides depression being a
serious and debilitating disorder in and of itself, it has various negative
consequences for physical health (Prince et al., 2007). At worst,
depression can lead to suicide (Ferrari et al., 2013), and 15% of those
who are clinically depressed die by suicide (Verster et al., 2008).
Further, the economic consequences of depression are substantial. The
costs of mood disorders in Europe for the year 2010 has been
estimated to be PPP 113.4 billion (Gustavsson et al., 2011). Almost
half of these costs is the result of productivity loss, implying the vast
negative impact of depression on populations' economy and sustainability. Depression has been found to be associated with a wide range
of factors such as female gender, somatic illness, and cognitive and
functional impairments. Additionally, loss of close social contacts has
been found to be an important predictor of depression, along with
various other variables relating to social relationships (Djernes, 2006).
Social relationships have important implications for both physical
and mental health. The state of one's social relationships can affect
that person's overall health. Specically, according to evidence from a
recent review, good social relationships can prolong survival by 50%
(Holt-Lundstad et al., 2010). A seven-decade follow-up study discovered that social relationships are better predictors of health than a
range of biological and economic factors (Vaillant, 2008). Similarly,
Holt-Lundstad et al. (2010) reported that having poor social relationships is potentially more harmful than excessive drinking and
smoking, obesity, and lack of exercise.
Studies investigating social relationships generally pertain to three
major domains: social support, social networks, and social connectedness (Ashida and Heaney, 2008; Barratt et al., 2006; Noone and
Stephens, 2014; Ottmann et al., 2006; Stone, 2003). Social support has
long been known to exert considerable inuence on mental health
and wellbeing (Thoits, 2011). The literature distinguishes between
perceived and received (or enacted) social support. Perceived support
is the subjective feeling of being supported by one's relationships,
while received support refers to the actual support provided. The
literature on social support further distinguishes between emotional
support (e.g. someone being available to listen or offer sympathy
during times of crisis or hardship, or someone available to give advice)
and instrumental support (e.g. someone available to offer help with
issues that require physical effort or nancial aid). All these different
forms of social support appear to have different implications for
mental health (Nurullah, 2012). Support may also be provided to or
received from different sources, such as spouse, children, relatives,
61
61
61
61
61
62
62
62
62
62
63
63
63
63
friends, and co-workers. It has been demonstrated that the role and
effect of social support on health and psychological well-being varies
depending on the source of support (Li et al., 2014; Walen and
Lachman, 2000). Finally, as an opposite measure of supportive
behaviors, some studies also include negative interactions (strain),
such as tension, criticism, or placing too strong demands on others
(Schuster et al. 1990).
Social relationships can be more clearly distinguished in terms of
their network structure (i.e. social networks) and network function
(i.e. social support) (House, 1987). Thus, while the term social support
relates to the functional content in social relationships, social networks is a concept that relates to the formal structure of social
relationships, such as size, composition, contact frequency, boundedness, etc. (Prince et al., 1997). Studies on social networks have been
characterized by investigating the web of social connections that
surrounds an individual and this has been shown to have important
associations with both physical and mental health (Smith and
Christakis, 2008). Two distinct approaches exist for assessing social
networks: egocentric and sociocentric approaches. Egocentric models
include data from individuals about members in their network,
whereas sociocentric models utilize not only the network information from each individual, but also information from the network
members themselves. As a consequence, sociometric models often
yield more novel insights, but also make much greater demands of
data (Smith and Christakis, 2008). Another insightful way to examine
social networks is to categorize them into network types, such as
network composition (e.g. the level of diversity of family members,
friends, and coworkers in a network) and other measures, such as
social engagement. Network types thus tend to encompass a more
comprehensive assessment of network properties and can be particularly informative (e.g. Litwin and Landau, 2000).
Finally, social connectedness refers to the degree to which a
person experiences belongingness, attachment, relatedness, togetherness, or entrenchment in one's social relationships. Thus, it refers
more to subjective feelings and attitudes towards oneself in relation
to the social context, rather than specic social relationships
(Townsend and McWhirter, 2005; Williams and Galliher, 2006).
The literature also makes use of the term social disconnectedness
to refer to conditions of social isolation, such as living alone, physical
separation from others, widowhood, etc. (Cornwell and Waite, 2009a,
2009b; Zavaleta et al., 2014).
A number of recent reviews have been published that address
the inuence of social relationships on populations' health status
(Holt-Lundstad et al., 2010; Smith and Christakis, 2008; Tajvar
et al., 2013; Uchino, 2006). However, the evidence in the literature
concerning the inuence of social relationships on mental illness,
particularly for depression, is often conicting or sparse (Nurullah,
2012). Moreover, many studies on social relationships employ
non-probability or convenience sampling, which considerably
limits the degree to which one can extrapolate ndings and draw
conclusions about the inuence of global network and support
properties in the general population (Smith and Christakis, 2008).
Thus, this work focuses on population-based studies that investigate the association between social relationships (social support,
social networks, and social connectedness) and depression.
2. Methods
2.1. Search strategy
The electronic databases of PsycInfo/PsycNet, PubMed/MedLine,
and ScienceDirect were searched for studies measuring the association
between social networks, social support, or social connectedness, and
depression. Search words were customized for each database, and
each search involved combining key word searches for a list of social
relationships variables (social support, peer support, emotional
support, social networks, social relationship, social connectedness,
belongingness) and specic terms relating to outcomes of depression
(depression, depressive symptoms). Medical Subject Headings
(MeSH) (i.e. social support, depression, adult) were used whenever possible. MeSH is the National Library of Medicine's controlled
vocabulary for the purpose of indexing journal articles in a hierarchical
structure, which makes it possible to conduct a more comprehensive
search. This terminology is commonly used in review articles to
identify relevant studies. Databases were searched for studies published in English, Spanish, French, Scandinavian (Danish, Swedish,
Norwegian), or Ex-Yugoslavian (Croatian, Bosnian, Serbian) languages.
These languages were selected based on the availability of people who
understand these languages in the research group.
The search was limited to studies published between 2004 and
2014 in order to obtain the most recent scientic articles. Studies were
55
included if: (a) they published empirical quantitative research examining the association between social support, social networks,
and/or social connectedness, and depression or depressive symptoms;
(b) social support, social networks, and/or social connectedness were
the predictor/independent variable or one of the main predictors/
independent variables of the study; and (c) depression or depressive
symptoms was the outcome/dependent variable of the study. Due to
the bias introduced when designating particular target groups as
participants, only studies involving community samples were included
(Tsuang et al., 2011). Thus, studies not employing random sampling
were excluded, and included studies had to have been carried out
according to conventional standards relating to appropriate sampling
procedures to reect the general population (Bonita et al., 2006). Due
to the focus on adults, published studies exclusively on infants,
children, and adolescents were also excluded.
Two authors independently reviewed potential articles to be
included based on the inclusion criteria. The level of agreement was
assessed with kappa statistics (kappa 0.82; SE 0.05), and disagreements were resolved subsequently by consensus. The initial search
yielded 1737 articles with duplicates removed. Inspection of abstracts
and titles found that 1675 articles did not fulll the inclusion criteria.
Sixty-two articles were identied as potentially relevant, but 11 of
those were later excluded as closer examination revealed that they did
not match the inclusion criteria. An overview of the search process is
illustrated in Tables 1 and 2 and Fig. 1.
2.2. Data extraction
The search strategy resulted in a total of 51 papers being
included in the review. All the information from each article that
was relevant to the research question and in line with the
inclusion criteria was extracted and tabulated. Extracted data
comprised publication data, country, language, setting and aims
Table 1
List of sources searched and search terms used for systematic review.
Electronic databases
Pubmed/Medline
PsycNet/PsycInfo
ScienceDirect
Search terms
Social support [MeSH] OR peer support [keyword] OR emotional support [keyword] OR social networks [keyword] OR social relationship [keyword] OR social
connectedness [keyword] OR belongingness [keyword]
AND
Depression [MeSH] OR depressive symptoms [keyword]
AND
Adult [MeSH]
AND
Year: 2004 TO 2014
Table 2
Overview of the search terms used in each search database.
Pubmed
ScienceDirect
PsycNet
Search terms
Social support [MeSH]
Peer support [keyword]
Emotional support [keyword]
Social networks [keyword]
Social relationship [keyword]
Social connectedness [keyword]
Belongingness [keyword]
Depression [MeSH]
Depressive symptoms [keyword]
Adult [MeSH]
Boolean operators
OR
OR
OR
OR
OR
OR
AND
OR
AND
Search terms
Social support [keyword]
Depression [keyword]
Boolean operators
OR
OR
OR
OR
OR
OR
AND
OR
Search terms
Social support [keyword]
Depression [keyword]
Adulthood (18yrs & older)
Year: 2004 TO 2014
Boolean operators
OR
OR
OR
OR
OR
OR
AND
OR
AND
AND
Hits:
1550
Hits:
167
Hits:
58
56
3. Results
1775 citations
initially retrieved
38 duplicates excluded
1737 citations
preliminary
assessed for
inclusion based on
their abstracts and
titles
1675 were excluded because:
62 full-text papers
read
51 included for
review
Not general
populations
(n=1392)
Not adults (n=23)
Not relevant
(n=260)
11 were excluded because they

did not meet the inclusion
criteria or did not sufficiently
address the review topic
Fig. 1. Selection of studies for systematic review.
of the study, study design, sample characteristics, data collection

methods, key measures, theoretical framework, and main ndings.
Extracted data also comprised quality assessment of each study
following the guidelines of the Effective Public Health Practice
Project (Armijo-Olivo et al., 2012; Thomas et al., 2004). An overview and description of included studies including quality assessment can be found in Tables 3 and 4. The quality assessment
comprised of six components: (1) selection bias; (2) study design;
(3) confounders; (4) blinding; (5) data-collection method; and
(6) withdrawals and drop-outs. Each component was rated as
weak, moderate, or strong, and a nal rating was made of each
study. A study was rated as low quality if it had received two or
more weak ratings; moderate quality if it had received one weak
rating; and high quality if it had not received any weak ratings.
Any discrepancies in terms of rating were resolved between the
two reviewers. The complete details of the quality assessment
procedure can be found on the Effective Public Health Practice
Project website: http://www.ephpp.ca/tools.html.
2.3. Synthesis
Due to the broad nature of the search and the different studies
included in the review, a narrative synthesis was conducted to synthesize the information (see Table 1 and Fig. 1). The synthesis was
guided by the methods described by Popay et al. (2006). This guide
provides some basic steps to the process of conducting a narrative
synthesis which is more systematic and transparent, and also minimizes bias in both the assessment of studies and decision made by
reviewers. The guide includes topics on developing a theoretical
model and a preliminary synthesis for a narrative review, exploring
the relationships in the data, and assessing the robustness of the
synthesis product. As a component to this approach, different studies
in this review were grouped into overarching conceptually or thematically related categories.
The review included 28 cross-sectional and 23 prospective

studies. According to quality criteria set forth by the Effective
Public Health Practice Project (Armijo-Olivo et al., 2012; Thomas
et al., 2004), nine studies were rated as low quality (9/51 17.6%),
30 studies were rated as moderate quality (30/51 58.8%), and 12
studies were rated as high quality (12/51 23.6%). Of those rated
as low, seven were cross-sectional, and two were prospective
studies. Of those rated as moderate, 21 were cross-sectional, and
nine were prospective studies. Finally, the 12 studies rated as
high were all prospective studies.
3.1. Social support
Perceived emotional support was the most frequently utilized
social support variable, being used in 35 of the studies. This
variable was signicantly associated with depressive symptoms
in 32 of the 35 ve studies (32/3591.4%). In all these studies,
higher levels of perceived emotional support were protective
against depression, and lower levels were associated with the
presence, onset or development of depression [15.6% (5/32) of the
studies were cross-sectional with low quality (Chen et al., 2005;
Harvey et al., 2010; Ostberg and Lennartsson, 2007; Virtanen et al.,
2008; Zhang and Li, 2011); 40.6% (12/32) were cross-sectional with
moderate quality (Alexandrino-Silva et al., 2011; Chazelle et al.,
2011; Choi and Ha, 2011; Fiori et al., 2006; Fiori and Denckla, 2012;
Glaesmer et al., 2011; Grav et al., 2012; Leung et al., 2007; Li and
Liang, 2007; Mair et al., 2010; Millan-Calenti et al., 2013; Sicotte
et al., 2008); 6.3% (2/32) were prospective with low quality
(Heponiemi et al., 2006; Tiikkainen and Heikkinen, 2005); 21.9%
(7/32) were prospective with moderate quality (Bierman and
Statland, 2010; Jokela et al., 2007; Koizumi et al., 2005; Pettit
et al., 2011; Plaisier et al., 2007; Rugulies et al., 2006; Stoetzer et
al., 2009); and nally 15.6% (5/32) were prospective with high
quality (Fauth et al., 2012; Huang et al., 2011; Taylor and Lynch,
2004; Teo et al., 2013; Yang, 2006)]. The remaining three studies
(3/35 8.6%) did not nd perceived emotional support to be a
signicant correlate of depression [33.3% (1/3) cross-sectional moderate (Litwin, 2011); 66.7% (2/3) prospective moderate (Smith and
Bielecky, 2012; Tsai et al., 2005)].
The ndings from the ve studies which measured both perceived
emotional support and perceived negative interactions were mixed.
One study found that neither were signicantly associated with
depression [20% (1/5) cross-sectional moderate (Litwin, 2011)]. Two
studies found that emotional support was negatively associated with
depression, and that negative interactions were positively associated
with depression with similar magnitudes in opposite directions [20%
(1/5) cross-sectional moderate (Li and Liang, 2007); 20% (1/5) crosssectional moderate (Fiori et al., 2006)]. One study found that only
negative interactions were signicantly and positively associated with
depression [20% (1/5) cross-sectional low (Ford et al., 2011)], while
another study found that both emotional support and negative
interactions were associated with depression in opposite directions
(i.e. emotional support being protective, and negative interactions
being a risk factor), but with the effect of negative interactions being
modestly stronger [20% (1/5) prospective high (Teo et al., 2013)].
Findings for received emotional support were similar to ndings regarding perceived emotional support. Low levels of received
emotional support were signicantly associated with depression
in eight out of 12 studies (8/12 66.7%) [37.5% (3/8) cross-sectional
moderate (Fiori and Denckla, 2012; Leggett et al., 2012; Mair et al.,
2010); 25% (2/8) prospective moderate (Rugulies et al., 2006;
Stoetzer et al., 2009), and 37.5% (3/8) prospective high (Chao,
2011; Garcia-Pena et al., 2013; Sonnenberg et al., 2013)]. One study
out of the twelve (1/12 8.3%) reported that receiving emotional
57
Table 3
Overview of the 28 cross-sectional studies included in the review.
Reference
Location
of study
Number of
participants
and age
group
N 6767
All adults
(18 years)
N 5033
6085 years
old
Grav et al.
Norway
N 40659
(2012)
2089 years
old
Harvey et al.
Norway
N 40401
(2010)
2089 years
old
Leggett et al. Vietnam N 600
(2012)
55 years old
and above
Li and Liang
China
N 2943
(2007)
6094 years
old
Sweden
N 5053
Ostberg and
Lennartsson
1875 years
(2007)
old
Zhang and Li China
N 1428
(2011)
55 years old
and above
Choi and Ha
USA
N 2924
(2011)
5785 years
old
Litwin (2010) Europe
N 9054
60 years old
and above
Litwin (2012) USA
N 1349
65 years old
and above
Mair et al.
USA
N 3105
All adults
(2010)
(18 years)
Millan-Calenti Spain
N 579
et al. (2013)
65 years old
and above
Tsai et al.
Taiwan
N 1200
(2005)
65 years old
and above
Fiori et al.
USA
N 1669
(2006)
60 years old
and above
Litwin (2011) USA
N 1350
6585 years
old
Sicotte et al.
Cuba
N 1905
60 years old
(2008)
and above
Wilby (2011) USA
N 91
65 years old
and above
Tiedt (2010)
Japan
N 3807
65 years old
and above
Chan et al.
Singapore N 4489
(2011)
60 years old
and above
Chen et al.
China
N 1600
60 years old
(2005)
and above
N 367
Alexandrino- Brazil
Silva et al.
60 years old
(2011)
and above
Ford et al.
UK
N 9377
45 years old
(2011)
USA
Fiori and
Denckla
(2012)
Glaesmer
Germany
et al. (2011)
Leung et al.
(2007)
China
N 507
65 years old
and above
Depression outcome measure
Methodological
quality
CrossSupport (perceived; received; provided;

sectional instrumental, emotional)
The Center for Epidemiologic StudiesDepression scale (CES-D) short form
Moderate
CrossSupport (perceived emotional)

sectional
The Depression Module of the Patient

Health Questionnaire (PHQ-9)
Moderate

sectional
The Hospital Anxiety and Depression

scale for depression (HADS-D)
Moderate

sectional
The Hospital Anxiety and Depression

scale for depression (HADS-D)
Low
CrossSupport (received emotional)

sectional
The Center for Epidemiologic StudiesDepression scale (CES-D)
Moderate
CrossSupport (perceived instrumental and emotional)

sectional and negative interactions
Moderate

sectional
Self-reporting of the occurrence of

depression
Low

sectional
Low

sectional
Moderate
Network (contact frequency, size, structure) and

Crosssectional social engagement
The EURO-D depression Scale
Moderate
CrossNetwork (contact frequency; size) and network

sectional type
Moderate
CrossCombined network (ties) and support (perceived

sectional emotional and reciprocal emotional and
instrumental)
CrossCombined network (contact frequency) and
sectional support (perceived emotional and instrumental)
Moderate
The Geriatric Depression Scale (GDS)

short form
Moderate
CrossCombined network (size) and support (perceived

sectional emotional)

short form
Moderate
CrossSupport (perceived emotional) and negative

sectional interactions.
Network types
CrossCombined network (social ties) and support
sectional (perceived emotional). Negative interactions and
network types
CrossCombined network (social ties) and support
sectional (perceived and received; network type emotional
and instrumental)
CrossCombined network (social ties; contact frequency)
sectional and support (received emotional)
The Center for Epidemiologic StudiesDepression scale (CES-D) short form
Moderate
Moderate
Moderate
Low
Moderate
Moderate
Design
and
study
length
Social relationships measure
CrossCombined network (size; contact frequency) and

sectional support (perceived and received emotional and
instrumental)
CrossCombined network (size; contact frequency) and
sectional support (perceived emotional)
CrossCombined network (contact frequency) and
sectional support (perceived emotional)
sectional
CrossCombined network (size; social ties; contact
sectional frequency) and support (perceived and received
emotional and instrumental). Negative interactions
sectional
The Geriatric Mental State (GMS) and the Low

Automated Geriatric Examination for
Computer Assisted Taxonomy (AGECAT)
The Composite International Diagnostic
Moderate
Interview version 1.1 (CIDI 1.1)
The abbreviated revised Clinical
Interview Schedule (CIS-R)
Low
The Symptom Checklist-90 revised (SCL90-R)
Moderate
58
Table 3 (continued )
Reference
Location
of study
Chazelle et al. Ireland

(2011)
Virtanen et al. Finland
(2008)
Waldenstrom Sweden
et al. (2008)
Russell and
Taylor
(2009)
USA
Number of
participants
and age
group
Design
and
study
length
N 9978
All adults
(18 years)
N 3374
3064 years
old
N 672
2064 years
old
N 947
All adults
(18 years)
CrossCombined network (composition) and support

sectional (perceived emotional)
Methodological
quality
CrossSupport (perceived instrumental)

sectional
The short form of the Composite

Moderate
International Diagnostic Interview (CIDISF)
Low
The WHO Composite International
Diagnostic Interview (M-CIDI) for
depressive disorder
DSM-IV depressive syndromes assessed
Moderate
via SCAN interviews

sectional

sectional
Moderate
Methodological quality was rated in terms of six components: (1) selection bias; (2) study design; (3) confounders; (4) blinding; (5) data-collection method; (6) withdrawals
and drop-outs. Each component was rated as weak, moderate or strong, and a nal rating was made of each study. A study was rated as 'low quality' if it had received two or
more weak ratings; 'moderate quality' if it had received one weak rating, and 'high quality' if it had not received any weak ratings. Any discrepancies in terms of rating were
resolved between the two reviewers.
support was a predictor of depression [prospective high (Bisschop

et al., 2004)], while the remaining studies (3/12 25%) did not
reach statistical signicance [33.3% (1/3) cross-sectional low (Ford
et al., 2011); 33.3% (1/3) prospective moderate (Smith and
Bielecky, 2012), 33.3% (1/3) prospective high (Taylor and Lynch,
2004)].
Low levels of perceived instrumental support also emerged as
an important predictor of depression in eleven out of twelve
studies (11/12 91.6%) [27.3% (3/11) cross-sectional low (Harvey et
al., 2010; Ostberg and Lennartsson, 2007; Virtanen et al., 2008);
45.4% (5/11) cross-sectional moderate (Fiori and Denckla, 2012;
Grav et al., 2012; Leung et al., 2007; Li and Liang, 2007; Sicotte
et al., 2008); 9.1% (1/11) prospective low (Tiikkainen and
Heikkinen, 2005); 9.1% (1/11) prospective moderate (Koizumi
et al., 2005); 9.1% (1/11) prospective high (Huang et al., 2011)],
while only one study (1/12 8.4%) examining perceived instrumental support did not nd it to be a signicant correlate [crosssectional moderate (Millan-Calenti et al., 2013)]. Three studies
included both perceived and received support, and all of these
found that perceived support was a more important predictor of
depression than received support, which was found to have either
a less strong or non-signicant association [33.3% (1/3) crosssectional moderate (Fiori and Denckla, 2012), 66.7% (2/3) prospective high (Taylor and Lynch, 2004; Yang, 2006)].
Findings were more mixed for received instrumental support.
Only two out of ten studies (2/10 20%) reported protective effects
of instrumental support receipt [50% (1/2) cross-sectional moderate (Waldenstrom et al., 2008); 50% (1/2) prospective high
(Muramatsu et al., 2010)]. Three studies (3/10 30%) found receipt
of instrumental support to be a signicant risk factor for depression [33.3% (1/3) cross-sectional low (Ford et al., 2011); 33.3% (1/3)
cross-sectional moderate (Tiedt, 2010); 33.3% (1/3) prospective
high (Bisschop et al., 2004)], while four studies (4/10 40%) did
not reach signicance [50% (2/4) cross-sectional moderate (Fiori
and Denckla, 2012; Tsai et al., 2005); 50% (2/4) prospective high
(Taylor and Lynch, 2004; Yang, 2006)]. The remaining study
[prospective high (Garcia-Pena et al., 2013)] reported that low
levels of both received emotional and instrumental support predicted deterioration of depressive symptoms for people with
depression at baseline, but neither was a signicant predictor of
depressive symptoms for people without depression at baseline.
Eight studies utilized variables on both emotional and instrumental
support. Five of those (5/8 62.5%) found emotional support to be
more strongly related to depression than instrumental support [60%
(3/5) cross-sectional moderate (Fiori et al., 2006; Leung et al., 2007;
Millan-Calenti et al., 2013), 40% (2/5) prospective high (Chao, 2011;

Yang, 2006)], while the remaining three studies (3/837.5%) concluded the opposite [66.6% (2/3) cross-sectional low (Ford et al., 2011,
Ostberg and Lennartsson, 2007); 33.3% (1/3) prospective moderate
(Koizumi et al., 2005)].
In terms of the source of social support, ve studies out of seven
studies (5/771.4%) concluded that social support from friends was
equally important in terms of predicting depression as family support
[20% (1/5) cross-sectional low (Zhang and Li, 2011); 60% (3/5) crosssectional moderate (Choi and Ha, 2011; Leggett et al., 2012; Russell
and Taylor, 2009), 20% (1/5) prospective high (Muramatsu et al.,
2010)], while two studies (2/7 28.6%) reported that only family
support exerted a signicant inuence on depression [50% (1/2)
prospective moderate (Pettit et al., 2011), 50% (1/2) prospective high
(Teo et al., 2013)].
In terms of studies focusing on the role of support in work-settings,
four studies provided evidence that emotional and instrumental
support in the workplace was protective against depression [25%
(1/4) cross-sectional moderate (Waldenstrom et al., 2008); 75% (3/4)
prospective moderate (Plaisier et al., 2007; Rugulies et al., 2006;
Stoetzer et al., 2009)]. Stoetzer et al. (2009) [prospective moderate]
found that work-related social support had a protective effect, however, when stratied by gender, this relationship was only found to be
protective for men. Rugulies et al. (2006) [prospective moderate]
reported that low levels of support from workplace supervisors
signicantly predicted depression, but only for women. Low coworker
support was not found to be signicant for either gender. One study
did not nd evidence that work-related social support predicted
depression (Smith and Bielecky, 2012) [prospective moderate]. Two
of the studies reported that negative working conditions were strong
predictors of depression, and that social support was not effective
enough to buffer against the effect of this association [100% (2/2)
prospective moderate (Plaisier et al., 2007; Smith and Bielecky, 2012)].
3.2. Social networks
Nine studies (9/1369.2%) reported that a larger social network
was an important protective factor against depression [11.1% (1/9)
cross-sectional low (Ford et al., 2011); 33.3% (3/9) cross-sectional
moderate (Chan et al., 2011; Sicotte et al., 2008; Tsai et al., 2005);
55.5% (5/9) prospective high (Chao, 2011; Garcia-Pena et al., 2013;
Kuchibhatla et al., 2012; Rosenquist et al., 2011; Sonnenberg et al.,
2013)] while the remaining four studies (4/1330.8%) did not nd a
signicant association between social network size and depression
[25% (1/4) cross-sectional low (Wilby, 2011); 25% (1/4) cross-sectional
59
Table 4
Overview of the 23 prospective studies included in the review.
Reference
Location of Number of
study
participants and
age group
Garcia-Pena
et al.
(2013)
Huang et al.
(2011)
Mexico
Koizumi
et al.
(2005)
Pettit et al.
(2011)
Japan
Teo et al.
(2013)
USA
Kuchibhatla
et al.
(2012)
Rosenquist
et al.
(2011)
Tiikkainen
and
Heikkinen
(2005)
Chaoet al.
(2011)
USA
Sonnenberg
et al.
(2013)
Heponiemi
et al.
(2006)
Jokela et al.
(2007)
Bierman and
Statland
(2010)
Bisschop
et al.
(2004)
Fauth et al.
(2012)
Taylor and
Lynch
(2004)
Yang (2006)
Muramatsu
et al.
(2010)
Plaisier et al.
(2007)
Rugulies
et al.
(2006)
Smith and
Bielecky
(2012)
Stoetzer
et al.
(2009)
Cacioppo
et al.
(2010)
Taiwan
USA
USA
Finland
N 7449
60 years old
and above
N 1017
65 years old
and above
N 1178
70 years old
and above
N 816
2130 years old
N 4642
2575 years
old
N 4162
65 years old
and above
N 12067
30 years old
and above
N 133
80 years old
N 4049
60 years and
above
The
N 2823
Netherlands 5585 years
and above
Finland
N 3596
1530 years old
Taiwan
N 341
All adults
(18 years)
USA
N 1167
65 years and
older
The
N 2288
Netherlands 5585 years
old
Sweden
N 779
70 years old
and above
USA
N 3876
65 years and
older
USA
N 1149
65 years old
and above
USA
N 6535
70 years old
and above
The
N 7076
Netherlands 1864 years old
Finland
Denmark
Canada
Sweden
USA
N 4470
All adults
(18 years)
N 3753
1574 years old
N 4040
2064 years
old
N 229
5068 years
old
Design and
study length
Methodological
quality
Prospective;
3 time points
over 3 years
Prospective;
4 time points
over 10 years
Prospective;
2 time points
over 1 year
Prospective;
4 time points
over 14 years
Prospective;
2 time points
over 10 years
Prospective;
4 time points
over 10 years
Prospective;
3 time points
over 18 years
Prospective;
2 time points
over 5 years
Combined network (size) and support (received

emotional and instrumental)
The Geriatric Depression Scale

(GDS)
High
Support (perceived instrumental and emotional)
The Center for Epidemiologic

Studies-Depression scale (CES-D)
High
Support (perceived instrumental and emotional)
The Geriatric Depression Scale

(GDS)
Moderate
Support (perceived emotional)

Moderate
Support (perceived emotional) and negative

interactions

High
International Diagnostic Interview
(CIDI-SF)
High
Prospective;
5 time points
over 14 years
Prospective;
5 time points
over 14 years
Prospective;
2 time points
over 5 years
Prospective;
2 time points
over 4 years
Prospective,
2 time points
over 2 years
Prospective;
3 time points
over 6 years
Prospective;
3 time points
over 12 years
Prospective;
4 time points
over 12 years
Prospective;
2 time points
over 6 years
Prospective;
5 time points
over 10 years
Prospective;
3 time points
over 4 years
Prospective;
2 time points
over 5 years
Prospective;
3 time points
over 5 years
Prospective;
2 times points
over 3 years
Prospective;
5 time points
over 5 years
Network (contact frequency; size)
Network (social ties)

High
Social connectedness (perceived togetherness)

Low
Combined network (social ties; composition;

contact frequency) and support (received
emotional and instrumental)
Combined network (size) and support (received
emotional)

High

High
Beck's Depression Inventory (BDImodied)
Low
Beck's Depression Inventory (BDImodied)
Moderate
Four items from the Hopkins

Symptoms Checklist
Moderate
Combined network (size; contact frequency) and

support (received instrumental and emotional)

High

High
Support (perceived emotional; received emotional The Center for Epidemiologic

and instrumental)
High
Combined network (size; contact frequency) and

support (perceived emotional; received
instrumental)
Support (perceived and received instrumental)

High

short form
The CIDI interview
High
Support (perceived emotional and instrumental)
Support (perceived emotional and instrumental)
Moderate
Moderate
The ve-item Mental Health
Inventory (MHI-5) of the ShortForm Health Survey
Moderate
International Diagnostic Interview
(CIDI-SF)
The Major Depression Inventory
Moderate
(MDI)
Moderate
Methodological quality was rated in terms of six components: (1) selection bias; (2) study design; (3) confounders; (4) blinding; (5) data-collection method; (6) withdrawals
and drop-outs. Each component was rated as weak, moderate or strong, and a nal rating was made of each study. A study was rated as 'low quality' if it had received two or
more weak ratings; 'moderate quality' if it had received one weak rating, and 'high quality' if it had not received any weak ratings. Any discrepancies in terms of rating were
resolved between the two reviewers.
60
moderate (Millan-Calenti et al., 2013); 50% (2/4) prospective high

(Bisschop et al., 2004; Yang, 2006)]. The ndings on the role of
frequency of social contact on depression were less consistent. Two
studies out of four (2/450%) reported that lower frequency of social
contact predicted depression [50% (1/2) cross-sectional moderate
(Chan et al., 2011); 50% (1/2) prospective high (Chao, 2011)], while
two other studies (2/450%) did not nd any signicant associations
between frequency of contact and depression [50% (1/2) crosssectional moderate (Millan-Calenti et al., 2013), 50% (1/2) prospective
high (Teo et al., 2013)].
In terms of studies utilizing more comprehensive measures of
network properties, four studies analyzed the effect of the type of
social network on depression, and these studies consistently found
that diverse social networks were protective against depression as
opposed to more restricted network types, i.e., networks including
family, relatives, and several networks of friends had favorable
effects on depression outcomes [100% (4/4) cross-sectional moderate (Fiori et al., 2006; Litwin, 2011, 2012; Sicotte et al., 2008)].
Chao (2011) [prospective high] concluded that a network consisting of 2550% family and 5075% friends was the most benecial
in terms of protecting against depression. Rosenquist et al. (2011)
[prospective high] analyzed the distribution of depression across
social ties over time and discovered that depression levels in one
person were positively correlated with depression levels in friends
and neighbors, and that these correlations remained signicant up
to three degrees of separation. In the events of nancial strain,
Sicotte et al. (2008) [cross-sectional moderate] reported that social
networks served as a buffer against depression. Living with others
and having a great diversity of social ties were important deterrents of depression under conditions of insufcient income.
In terms of living arrangements, two studies reported that
multigenerational co-residence was a protective factor against
depression, i.e. people who live with their relatives or children
benet from the social networks of their living arrangements [50%
(1/2) cross-sectional moderate (Sicotte et al., 2008); 50% (1/2)
prospective high (Chao, 2011)]. While maintaining relationships to
one's family and relatives is important, Fiori et al. (2006) [crosssectional moderate] concluded that having only a family network
with few or no friends was more detrimental and posed a greater
risk of depression than having a network with friends but no
family.
3.3. Social connectedness
In relation to social connectedness, Tiikkainen and Heikkinen
(2005) [prospective low] assessed the impact of perceived togetherness on the association between loneliness and depression
over time and concluded that people who feel mutual proximity
and security in their social environment suffer signicantly less
often from depression.
In terms of social isolation, three studies out of twelve (3/1225%)
did not nd signicant associations between living alone or without a
partner and depression [33.3% (1/3) cross-sectional low (Wilby, 2011);
66.6% (2/3) prospective high (Garcia-Pena et al., 2013; Teo et al., 2013)].
However, nine studies (9/1275%) reported that living alone or
without a partner were signicant predictors of depression [11.1%
(1/9) cross-sectional low (Chen et al., 2005); 44.4% (4/9) crosssectional moderate (Chan et al., 2011; Fiori et al., 2006; Russell and
Taylor, 2009; Sicotte et al., 2008); 11.1% (1/9) prospective low
(Tiikkainen and Heikkinen, 2005); 11.1% (1/9) prospective moderate
(Cacioppo et al., 2010); 22.2% (2/9) prospective high (Bisschop et al.,
2004; Sonnenberg et al., 2013)]. Four studies were consistent in
reporting that living alone or without a partner was a greater risk
factor of depression for men [75% (3/4) cross-sectional moderate
(Chan et al., 2011; Fiori et al., 2006; Sicotte et al., 2008); 25% (1/4)
prospective high (Sonnenberg et al., 2013)], while three studies
suggested that women were more at risk from the loss of close friends
[100% (3/3) cross-sectional moderate (Alexandrino-Silva et al., 2011;
Choi and Ha, 2011; Fiori et al., 2006)]. Zhang and Li (2011) [crosssectional moderate] reported that widows either had or perceived
themselves as having less emotional support from the family than the
married elderly, and this lower level of support predicted higher levels
of depression. Russell and Taylor (2009) [cross-sectional moderate]
similarly concluded that people who live alone seem to benet less
from social support than people who live with a partner.
3.4. Other salient ndings
As a result of sorting all included papers into overarching
thematically or conceptually related categories, several distinct
topics emerged on how various factors play a role in the association between social relationships and depression. These factors
were gender (13 studies), intrapersonal characteristics (5 studies),
and chronic physical illness and disability (8 studies).
3.5. Gender
In terms of associations by sex, eight studies out of 13 (8/13
61.5%) reported that lack of social support was a signicant predictor
of depression for women, while this was not the case for men [12.5%
(1/8) cross-sectional low (Virtanen et al., 2008); 50% (4/8) crosssectional moderate (Fiori and Denckla, 2012; Mair et al., 2010; Sicotte
et al., 2008; Tiedt, 2010); 25% (2/8) prospective moderate (Koizumi
et al., 2005; Rugulies et al., 2006); 12.5% (1/8) prospective high (Huang
et al., 2011)]. Four studies (4/1330.8%) discovered that this same
pattern was signicant for men, but not for women [50% (2/4) crosssectional moderate (Alexandrino-Silva et al., 2011; Choi and Ha, 2011);
50% (2/4) prospective moderate (Plaisier et al., 2007; Stoetzer et al.,
2009)]. One study (1/137.7%) reported that lack of emotional
support was a signicant predictor only in women, while lack of
instrumental support was a signicant predictor only in men [crosssectional moderate (Grav et al., 2012)].
Only two studies included measures on support provision. Fiori
and Denckla (2012) [cross-sectional moderate] reported that
provision of emotional support was protective against depression
for both men and women, while women especially seemed to
benet from having a network in need of emotional support.
Sicotte et al. (2008) [cross-sectional moderate] concluded that
women seem to benet from reciprocity of social support, i.e.
balanced exchanges of social support play a protective role against
depression only for women.
3.6. Intrapersonal characteristics
Yang (2006) [prospective high] reported that a favorable effect of
perceived support in the association between disabilities and depression could be explained by psychological mechanisms, in the sense
that perceived emotional support bolstered an individual's sense of
control. Two studies showed results that implied the opposite effect,
but with instrumental support. In these studies, receipt of instrumental support exacerbated symptoms of depression, suggesting that
such support may undermine feelings of control [50% (1/2) crosssectional moderate (Tiedt, 2010); 50% (1/2) prospective high (Bisschop
et al., 2004)]. Heponiemi et al. (2006) [prospective low] investigated the inuence of perceived emotional support on progression
to depression, while taking into account childhood anger and later
hostility. They reported an overall protective effect of support - an
effect that was associated with decreases in depressive symptoms over
time. Although hostility itself was a risk factor for depression, its
presence often caused interpersonal conicts, and this in turn,
effectively reduced an individual's opportunities for receiving support.
Sonnenberg et al. (2013) [prospective high] investigated the inuence
of individuals' need for social afliation the need to actively seek out
support from social connections in the relation between received
emotional support and late-life depression. Across both genders, a
high need for social afliation seemed to worsen outcomes under low
levels of emotional support.
Finally, Jokela et al. (2007) [prospective moderate] assessed the
inuence of a particular genetic factor on the association between
perceived social support and depression. Over a four year period, low
social support predicted an increase in depressive symptoms only in
individuals carrying both dominant alleles of the Tryptophan hydroxylase 1 gene, implying a greater risk for individuals with this
particular genetic feature under conditions of low social support.
3.7. Chronic physical illness and disability
Eight studies reported that perceived social support played a
signicant protective role in the association between disability or
chronic illness and depression [12.5% (1/8) cross-sectional moderate (Leung et al., 2007), 12.5% (1/8) prospective moderate
(Bierman and Statland, 2010), 75% (6/8) prospective high (Chao,
2011; Fauth et al., 2012; Huang et al., 2011; Muramatsu et al., 2010;
Taylor and Lynch, 2004; Yang, 2006)]. In a similar fashion, Huang
et al. (2011) [prospective high], Yang (2006) [prospective high] and
Bierman and Statland (2010) [prospective moderate] found that
adults with high levels of perceived social support experienced
less distress from disability and limitations in activities of daily
living. Leung et al. (2007) [cross-sectional moderate] also reported
a similar protective effect of perceived emotional support in the
association between chronic illnesses and depression. Fauth et al.
(2012) [prospective high] found that greater levels of perceived
social support predicted fewer depressive symptoms at disability
onset, and a more pronounced decline in depressive symptoms at
later time points, thus implying a protective effect. Both Taylor and
Lynch (2004) [prospective high] and Muramatsu et al. (2010)
[prospective high] reported that changes in trajectories of perceived emotional support over time mediated the relationships
between worsening disability and depressive symptoms, and can
act as a buffer against the detrimental effects of disability on
depression. Finally, Chao (2011) [prospective high] concluded that
provision of instrumental support was protective when it was
received on a short-term basis, which can be indicative of a less
severe illness, while the opposite was the case when it was
received on a long-term (permanent) basis, as can be expected
with more severe illnesses.
4. Discussion
4.1. Social support
There was a general consensus that perceived support is more
important than received support, and there was also overwhelming evidence that the perceived emotional variant consistently
played a protective role against depression across general populations. A considerable amount of good quality studies demonstrated
a causal direction from perceived social support to depression
through prospective analyses. Also, there was some indication that
negative interactions are important risk factors for depression.
However, only few studies utilized measures relating to negative
interactions.
The evidence was mixed for receipt of instrumental support, and
it was difcult to pinpoint exactly how this inuences depression.
The evidence was conicting across studies that were different in
terms of both quality and design. Other reviews note that the effect
of received social support on mental health is generally inconclusive
(Haber et al., 2007), which seems to also be the case for received
61
instrumental support in this review. A number of the included

studies suggested that receipt of instrumental support has negative
implications for depression. Previous research has also noted associations between receipt of instrumental support and adverse mental
health outcomes (Deelstra et al., 2003; Kawachi and Berkman, 2001;
Nurullah, 2012).
Across included studies, there was some evidence that workrelated social support was an important determinant of depression,
but the evidence was limited and sometimes conicting. Similarly,
several studies found more important determinants relating to
negative working conditions. Previous research reports that the effect
of work-related social support seems to be sensitive to other factors,
and that several psychosocial work characteristics must be taken into
account to predict physical health (Hoogendoorn et al. 2000; Michie
and Williams, 2003).
4.2. Social networks
The protective effect of larger social networks against depression
has been conrmed in the literature (Smith and Christakis, 2008). In
this review, over half of the studies arriving at this same conclusion
were of the highest quality and had a prospective design. Notably, the
diverse types of social networks were consistently found to be
associated with favorable depression outcomes. It seems that having
close family relationships combined with a number of different peer
networks from different contexts is particularly benecial. Finally,
depression can spread through social networks. The contagious properties of depression have been corroborated by other independent
studies and meta-analyses (Haeffel and Hames, 2014; Joiner and Katz,
1999). Although this may seem disconcerting, Smith and Christakis
(2008) note that positive and benecial emotional states, such as
happiness and optimism, also seem to spread through social networks.
4.3. Social connectedness
Summarizing the limited evidence for social connectedness, it
seems that people in a social environment with relationships in which
they feel mutual proximity and belongingness are less likely to be at
risk of depression. Ashida and Heaney (2008) found that social
connectedness may be more important for the health status of older
adults, possibly because higher levels of feeling socially connected
offer more proximity to social networks, and greater likelihood of
feeling comfortable while relying on networks for support. Thus, social
connectedness may play a protective role against depression by
mediating the positive effects of social relationships on mental health
(Williams and Galliher, 2006). Previous reviews and independent
studies have conrmed the protective effect of social connectedness
against various mental health outcomes (Townsend and McWhirter,
2005; Lee et al., 2001; Lee and Robbins, 1998; Resnick et al., 1997).
However, this review highlights the current situation of social connectedness being underutilized in studies assessing social determinants of depression.
There was good evidence in terms of both prospective designs
and quality to support that social isolation is particularly detrimental
in terms of predicting depression. Multi-generational living arrangements was a general protective factor against depression. This nding
may be considered in the light of research demonstrating that
transitions to institutional settings predict depressive symptoms
(Pot et al., 2005). There may be several reasons for this, but it could
be attributed to feelings of proximity and social inclusion. Living
alone or without a partner were generally found to be strong
predictors of depression, and bereavement in particular played an
important role. The importance of psychological factors of social
isolation and bereavement in the association with depression in
community-dwelling elderly has been emphasized in previous
reviews (Cole and Dendukuri, 2003; Kawachi and Berkman, 2001).
62
Reviews have conrmed that bereavement can be more damaging to

the physical and mental health of men, while women are strongly
affected by the loss of close friends (Smith and Christakis, 2008;
Kawachi and Berkman, 2001). This may be explained by the
consistent nding that men tend to rely on spouses for support,
while women seem to rely more on friends as condants (Glynn
et al., 1999; Reevy, 2007; Schultz and Schwarzer, 2004; van Daalen
et al., 2005).
4.4. Gender
The literature supports signicant associations between social
support and depression particularly and sometimes exclusively for
women (Kawachi and Berkman, 2001), and also that reciprocity of
support seems to be an important protective factor for women
(Belle, 1987).
4.5. Intrapersonal characteristics
A variety of psychological characteristics play important roles in the
relationship between social support and depression. In terms of
feelings of control, previous research conrms that different types of
social support may either promote self-efcacy or have detrimental
effects on self-esteem by reinforcing dependence (Kawachi and
Berkman, 2001). Only three studies in this review included measures
of feelings of control. More research is warranted to explore sense of
control and self-efcacy as a possible mechanism in the link between
social relationships and outcomes on mental disorders. A high need
for social afliation also seems to be important, especially under
conditions of low levels of social support. Similar conclusions have
been drawn in studies utilizing the variables of social neediness in
relation to depression (Rude and Burham, 1995). Another intrapersonal characteristic, hostility, was a risk factor for depression, which
often resulted in interpersonal conicts and reduced opportunities for
receiving support. Thus, psychological factors and levels of social
support seem to have a reciprocal effect on each other in these
instances, which ultimately have important consequences for depression outcomes. A moderating role of psychological characteristics on
physical health outcomes has been noted in a previous review and
may be equally relevant for mental health (Uchino, 2006). Uchino
further provides evidence for the inuence of social support on the
relationship between various biological characteristics and disease. In
this review, we identied one particular genetic composition which
seems to have a role to play in the support-depression relationship.
4.6. Chronic physical illness and disability
Perceived social support seems to offer a sense of security and
reliance on social networks for people that are physically disadvantaged or people that are experiencing disease. This was a robust
nding across a range of different included studies, with three quarters
being prospective high quality studies. Previous research has found
similar favorable associations between perceived social support and
outcomes relating to depression, levels of functional impairment, and
ability to cope with physical disabilities (Evers et al., 1998; Greenglas
et al., 2006; Holahan et al., 1997, 1995, 1996).
The included studies were less informative in terms of the
inuence of received support, particularly instrumental support.
The reason for this may be that long-term received instrumental
support is often referred to under the term informal care, which
was beyond the scope of this review. However, it should be noted
that an abundance of studies demonstrate that long-term instrumental support for various reasons is associated with adverse
mental health outcomes (Blazer, 2003; Kwak et al., 2014; Martire
et al., 2002; Newsom and Schultz, 1998; Roe et al., 2001; Beach
et al., 2005; Christie et al., 2009; Jang et al., 2002; Krause and
Rook, 2003; Wallsten et al., 1999; Wolff and Agree, 2004).
4.7. Limitations
A number of limitations should be born in mind when interpreting
the results. First, reporting bias may exist for perceived social support.
For example, lack of perceived support may appear to be a risk factor
for depression, but this may be due to the tendency of individuals with
depression to form rather cynical and pessimistic perceptions of the
reality of their social surroundings, and this may not necessarily mean
that support is actually lacking. This could explain the discrepancies
observed for the associations between perceived or received support
and depression. In addition, reverse causality may exist between
depression and social relationships, where people with depression
may be less likely to engage in social activities or seek social support.
This is a limitation of particular importance for cross-sectional studies,
as prospective studies are less likely to be affected by this type of bias.
Further, due to the large number of studies available in the domain of
interest, only studies from peer-reviewed journals were included.
Publication bias and the fact that gray literature (i.e. papers published
in databases that are not controlled by academic publishers) were not
included in our review may have limited our ndings. One should also
note that our emphasis on studies assessing the inuence of social
relationships in populations-based samples did not allow for the
inclusion of qualitative studies. Although qualitative studies are less
able to address how one variable inuences another, they can be
useful in terms of offering explanations of psychosocial phenomena
that are often not accessible through quantitative research. Finally,
given that this review included studies from a wide range of cultural
and national contexts, cultural differences may exist and have implications for how ndings should be interpreted. A growing body of
literature has examined how social relationships vary across culture,
such as social support seeking in individualistic as opposed to
collectivistic cultures (Kim et al., 2008). However, the articles included
in this review generally focused on the population in question, and we
did not come across any studies drawing particularly insightful
conclusions regarding culture differences. This is an area that should
be further explored.
5. Conclusion
This review provides some conrmation that perceived social
support and larger, diverse social networks, in particular, play
important protective roles against depression in the general
population, including those with chronic somatic illness or disability. However, some challenges persist and must be addressed.
First, studies are needed to address how people's perceptions of
social support and networks, as compared to objective measures,
affect the association between social relationships and depression.
Second, it was difcult to compare studies due to the utilization of
different support and network variables that are distinct and not
necessarily comparable even when pertaining to social support or
social networks. Thus, the best comparisons were derived from
studies that utilized several social support variables (e.g. perceived
and received support) or social network variables (e.g. size and
contact frequency). Studies on social support and social networks
could benet from more coherence in terms of incorporating
several relevant measures, and research should aim to include a
broad range of social relationships variables, rather than just one
or two single measures. This would pave the way for a more robust
research base for social support and social networks, and ultimately strengthen the comparability across studies. Third, measures on negative interactions (as opposed to emotional support)
were largely underutilized, which made it difcult to make conc-
lusions about how supportive relationships and negative interactions may interact and impact on health. Thus, studies on social
support are more informative when also including measures on
negative interactions in social relationships. Forth, although the
included studies on social support and social networks were vast
and comprehensive, many of the studies did not include variables
on both social networks and social support, and often when they
did, they still did not include all the necessary and relevant
variables. On the contrary, most studies focused either on social
support or social networks, and were not able to evaluate how
these two constructs relate to each other. The most informative
studies were those that did not focus exclusively on social support
or social networks, but those that considered these two in their
analyses as being two distinct aspects of basically the same
construct. As emphasized previously, social networks and social
support have sometimes been referred to as network structure
and network function, respectively. This may be a particularly
useful way to engage in research with these variables, as this
approach seems to take both elements into account. There is a
strong call for future research to determine how social networks
and social support interact and ultimately affect risks for depression. Finally, social connectedness is underutilized as a social
relationships measure in studies assessing depression outcomes
in the general population. Future research has yet to assess the
degree to which this variable constitutes a vital social determinant
of depression in population-based samples.
Role of funding source

None declared.
Conict of interest
None of the authors have any interests to declare in relation to this submission.
Acknowledgments
The research leading to these results has received funding from the People
Programme (Marie Curie Actions) of the European Union's Seventh Framework
Programme FP7/20072013 under REA Grant agreement no. 316795.
Ai Koyanagi's work was supported by the Miguel Servet contract by CIBERSAM,
Grant agreement no. CP13/00150.
Stefano Tyrovolas' work was funded through a scholarship from the Foundation
for Education and European Culture (IPEP).
References
Alexandrino-Silva, C., Alves, T.F., Tofoli, L.F., Wang, Y.P., Andrade, L.H., 2011.
Psychiatry: life events and social support in late life depression. Clinics
(Sao Paulo) 66 (2), 233238.
Armijo-Olivo, S., Stiles, C.R., Hagen, N.A., Biondo, P.D., Cummings, G.G., 2012.
Assessment of study quality for systematic reviews: a comparison of the
cochrane collaboration risk of bias tool and the effective public health practice
project quality assessment tool: methodological research. J. Eval. Clin. Pract.
18 (1), 1218.
Ashida, S., Heaney, C.A., 2008. Differential associations of social support and social
connectedness with structural features of social networks and the health status
of older adults. J. Aging Health 20 (7), 872893.
Barratt, J., Chambers, L., Graham, T., Keefe, J., Meloche, A., O'Brien-Cousins, S., Scott,
V., 2006. Healthy Aging in Canada: A New Vision, A Vital Investment from
Evidence to Action. Public Health Agency of Canada, Ottawa, ON.
Beach, S.R., Schultz, R., Williamson, G.M., Miller, L.S., Weiner, M.F., Lance, C.E., 2005.
Risk factors for potentially harmful informal caregiver behavior. J. Am. Geriatr.
Soc. 53 (2), 255261.
Belle, D., 1987. Gender differences in the social moderators of stress. In: Barnett, R.C.,
Biener, G.K., Baruch, G.K. (Eds.), Gender and Stress. The Free Press, New York, NY,
pp. 257277.
Bierman, A., Statland, D., 2010. Timing, social support, and the effects of physical
limitations on psychological distress in late life. J. Gerontol. Ser. B: Psychol. Sci.
Soc. Sci. 65 (5), 631639. http://dx.doi.org/10.1093/geronb/gbp128.
Bisschop, M.I., Kriegsman, D.M., Beekman, A.T., Deeg, D.J., 2004. Chronic diseases
and depression: the modifying role of psychosocial resources. Soc. Sci. Med.
59 (4), 721733. http://dx.doi.org/10.1016/j.socscimed.2003.11.038.
Blazer, D.G., 2003. Depression in late life: review and commentary. J. Gerontol., Ser.
A: Biol. Sci. Med. Sci. 58 (3), M249M265.
63
Bonita, R., Beaglehole, R., Kjellstrm, T., 2006. Basic Epidemiology, 2nd ed. World
Health Organization, Geneva, Switzerland.
Cacioppo, J.T., Hawkley, L.C., Thisted, R.A., 2010. Perceived social isolation makes me
sad: 5-year cross-lagged analyses of loneliness and depressive symptomatology
in the Chicago health, aging, and social relations study. Psychol. Aging 25 (2),
453463. http://dx.doi.org/10.1037/a0017216.
Chan, A., Malhotra, C., Malhotra, R., Ostbye, T., 2011. Living arrangements, social
networks and depressive symptoms among older men and women in Singapore. Int. J. Geriatr. Psychiatry 26 (6), 630639. http://dx.doi.org/10.1002/
gps.2574.
Chao, S.F., 2011. Assessing social support and depressive symptoms in older Chinese
adults: a longitudinal perspective. Aging Ment. Health 15 (6), 765774. http:
//dx.doi.org/10.1080/13607863.2011.562182.
Chao, S.F., McCallion, P., Nickle, T., 2011. Factorial validity and consistency of the
Maslach Burnout Inventory among staff working with persons with intellectual
disability and dementia. J. Intellect. Disabil. Res. 55 (5), 529536. http://dx.doi.
org/10.1111/j.1365-2788.2011.01413.x.
Chazelle, E., Lemogne, C., Morgan, K., Kelleher, C.C., Chastang, J.F., Niedhammer, I.,
2011. Explanations of educational differences in major depression and generalised anxiety disorder in the Irish population. J. Affect. Disord. 134 (13),
304314. http://dx.doi.org/10.1016/j.jad.2011.05.049.
Chen, R., Wei, L., Hu, Z., Qin, X., Copeland, J.R., Hemingway, H., 2005. Depression in
older people in rural China. Arch. Intern. Med. 165 (17), 20192025. http://dx.
doi.org/10.1001/archinte.165.17.2019.
Choi, N.G., Ha, J.H., 2011. Relationship between spouse/partner support and
depressive symptoms in older adults: gender difference. Aging Ment. Health
15 (3), 307317. http://dx.doi.org/10.1080/13607863.2010.513042.
Christie, J., Smith, G.R., Williamson, G.M., Lance, C.E., Shovali, T.E., Silva, L.C., 2009.
Quality of informal care is multidimensional. Rehabil. Psychol. 54 (2), 173181.
Cole, M.G., Dendukuri, N., 2003. Risk factors for depression among elderly
community subjects: a systematic review and meta-analysis. Am. J. Psychiatry
160 (6), 11471156.
Cornwell, E.Y., Waite, L.J., 2009a. Measuring social isolation among older adults
using multiple indicators from the NSHAP study. J. Gerontol. Ser. B: Psychol. Sci.
Soc. Sci. 64B (Suppl 1), i38i46.
Cornwell, E.Y., Waite, L.J., 2009b. Social disconnectedness, perceived isolation, and
health among older adults. J. Health Soc. Behav. 50 (1), 3148.
Deelstra, J.T., Peeters, M.C.W., Schaufeli, W.B., Stroebe, W., Zjilstra, F.R.H., 2003.
Receiving instrumental support at work: when help is not welcome. J. Appl.
Psychol. 88, 324331.
Djernes, J.K., 2006. Prevalence and predictors of depression in populations of
elderly: a review. Acta Psychiatr. Scand. 113, 372387.
Evers, A.W.M., Kraaimaat, F.W., Geenen, R., Bijlsma, J.W.J., 1998. Psychosocial
predictors of functional change in recently diagnosed rheumatoid arthritis
patients. Behav., Res. Ther. 36, 179193.
Fauth, E.B., Gerstorf, D., Ram, N., Malmberg, B., 2012. Changes in depressive
symptoms in the context of disablement processes: role of demographic
characteristics, cognitive function, health, and social support. J. Gerontol. Ser.
B: Psychol. Sci. Soc. Sci. 67 (2), 167177. http://dx.doi.org/10.1093/geronb/
gbr078.
Ferrari, A.J., Charlson, F.J., Norman, R.E., Patten, S.B., Freedman, G., Murray, C.J.L.,
Whiteford, H.A., 2013. Burden of depressive disorders by country, sex, age, and
year: ndings from the global burden of disease study 2010. PLoS Med. 10 (11),
112.
Fiori, K.L., Antonucci, T.C., Cortina, K.S., 2006. Social network typologies and mental
health among older adults. J. Gerontol. Ser. B: Psychol. Sci. Soc. Sci. 61 (1),
P2532.
Fiori, K.L., Denckla, C.A., 2012. Social support and mental health in middle-aged
men and women: a multidimensional approach. J. Aging Health 24 (3),
407438. http://dx.doi.org/10.1177/0898264311425087.
Ford, E., Clark, C., Stansfeld, S.A., 2011. The inuence of childhood adversity on
social relations and mental health at mid-life. J. Affect. Disord. 133 (12),
320327. http://dx.doi.org/10.1016/j.jad.2011.03.017.
Garcia-Pena, C., Wagner, F.A., Sanchez-Garcia, S., Espinel-Bermudez, C., JuarezCedillo, T., Perez-Zepeda, M., Gallo, J.J., 2013. Late-life depressive symptoms:
prediction models of change. J. Affect. Disord. 150 (3), 886894. http://dx.doi.
org/10.1016/j.jad.2013.05.007.
Glaesmer, H., Riedel-Heller, S., Braehler, E., Spangenberg, L., Luppa, M., 2011. Ageand gender-specic prevalence and risk factors for depressive symptoms in the
elderly: a population-based study. Int. Psychogeriatr. 23 (8), 12941300. http:
//dx.doi.org/10.1017/s1041610211000780.
Glynn, L.M., Christensen, N., Gerin, W., 1999. Gender, social support, and cardiovascular responses to stress. Psychosom. Med. 61, 234242.
Grav, S., Hellzen, O., Romild, U., Stordal, E., 2012. Association between social support
and depression in the general population: the HUNT study, a cross-sectional
survey. J. Clin. Nurs. 21 (12), 111120. http://dx.doi.org/10.1111/j.13652702.2011.03868.x.
Greenglas, E., Fiksenbaum, L., Eaton, J., 2006. The relationship between coping,
social support, functional disability and depression in the elderly. Anxiety,
Stress Coping: Int. J. 19 (1), 1531.
Gustavsson, A., Svensson, M., Jacobi, F., Allgulander, C., Alonso, J., Beghi, E., Olesen,
J., 2011. Cost of disorders of the brain in Europe 2010. Eur. J. Neuropsychopharmacol. 21, 718779.
Haber, M.G., Cohen, J.L., Lucas, T., Baltes, B.B., 2007. The relationship between selfreported received and perceived social support: a meta-analytic review. Am. J.
Community Psychol. 39 (12), 133144.
64
Haeffel, G.J., Hames, J.L., 2014. Cognitive vulnerability to depression can be

contagious. Clin. Psychol. Sci. 2 (1), 7585.
Harvey, S.B., Hotopf, M., Overland, S., Mykletun, A., 2010. Physical activity and
common mental disorders. Br. J. Psychiatry 197 (5), 357364. http://dx.doi.org/
10.1192/bjp.bp.109.075176.
Heponiemi, T., Elovainio, M., Kivimaki, M., Pulkki, L., Puttonen, S., KeltikangasJarvinen, L., 2006. The longitudinal effects of social support and hostility on
depressive tendencies. Soc. Sci. Med. 63 (5), 13741382. http://dx.doi.org/
10.1016/j.socscimed.2006.03.036.
Holahan, C.J., Moos, R.H., Bonin, L.A., 1997. Social support, coping and psychological
adjustment: a resources model. In: Avison, W.R., Gotlib, I.H. (Eds.), Stress and
Mental Health: Contemporary Issues and Prospects for the Future. Plenum,
New York, NY, pp. 213238.
Holahan, C.J., Moos, R.H., Holahan, C.K., Brennan, P.L., 1995. Social support, coping
and depresive symptoms in a late-middle-aged sample. Health Psychol. 14,
152163.
Holahan, C.J., Moos, R.H., Schaefer, J., 1996. Coping, resilience, and growth:
Conceptualizing adaptive functioning. In: Zeidner, M., Endler, N. (Eds.), Handbook of Coping: Research, Theory, and Application. Wiley, New York, NY,
pp. 2443.
Holt-Lundstad, J., Smith, T.B., Layton, J.B., 2010. Social relationships and mortality
risk: a meta-analytic review. PloS Med. 7 (7), 119.
Hoogendoorn, W.E., van Poppel, N.M., Bongers, P., Koes, B.W., Bouter, L.M., 2000.
Systematic review of psychosocial factors at work and private life as risk factors
for back pain. Spine (Phila Pa 1976) 25 (16), 21142125.
House, J.S., 1987. Social support and social structure. Sociol. Forum 2 (1), 135146.
Huang, J.-F., Wong, R.-H., Chen, C.-C., Mao, I.F., Huang, C.-C., Chang, W.-H., Wang, L.,
2011. Trajectory of depression symptoms and related factors in later life a
population based study. J. Affect. Disord. 133 (3), 499508. http://dx.doi.org/
10.1016/j.jad.2011.04.048.
Jang, Y., Haley, W.E., Small, B.J., Mortimer, J.A., 2002. The role of mastery and social
resources in the associations between disability and depression in later life.
Gerontologist 42 (6), 807813.
Joiner, T.E., Katz, J., 1999. Contagion of depressive symptoms and mood: metaanalytic review and explanations from cognitive, behavioral, and interpersonal
viewpoints. Clin. Psychol.: Sci. Pract. 6 (2), 149164.
Jokela, M., Raikkonen, K., Lehtimaki, T., Rontu, R., Keltikangas-Jarvinen, L., 2007.
Tryptophan hydroxylase 1 gene (TPH1) moderates the inuence of social
support on depressive symptoms in adults. J. Affect. Disord. 100 (13), 191197.
http://dx.doi.org/10.1016/j.jad.2006.10.016.
Kawachi, I., Berkman, L.F., 2001. Social ties and mental health. J. Urban Health
78 (3), 458467.
Kim, H.S., Sherman, D.K., Taylor, S.E., 2008. Culture and social support. Am. Psychol.
63 (6), 518526.
Koizumi, Y., Awata, S., Kuriyama, S., Ohmori, K., Hozawa, A., Seki, T., Tsuji, I., 2005.
Association between social support and depression status in the elderly: results of a
1-year community-based prospective cohort study in Japan. Psychiatry Clin.
Neurosci. 59 (5), 563569. http://dx.doi.org/10.1111/j.1440-1819.2005.01415.x.
Krause, N., Rook, K., 2003. Negative interaction in late life: issues in the stability and
generalizability of conict across relationships. J. Gerontol., Ser. B: Psychol. Sci.
Soc. Sci. 58 (2), P88P99.
Kuchibhatla, M.N., Fillenbaum, G.G., Hybels, C.F., Blazer, D.G., 2012. Trajectory classes of
depressive symptoms in a community sample of older adults. Acta Psychiatr. Scand.
125 (6), 492501. http://dx.doi.org/10.1111/j.1600-0447.2011.01801.x.
Kwak, M., Ingersoll-Dayton, B., Bursgard, S., 2014. Receipt of care and depressive
symptoms in later life: the importance of self-perceptions of aging. J. Gerontol.,
Ser. B: Psychol. Sci. Soc. Sci. 69 (2), 325335.
Lee, R.M., Draper, M., Lee, S., 2001. Social connectedness, dysfunctional interpersonal behaviors, and psychological distress: testing a mediator model. J. Couns.
Psychol. 48 (3), 310318.
Lee, R.M., Robbins, S.B., 1998. The relationship between social connectedness and
anxiety, self-esteem, and social identity. J. Couns. Psychol. 45 (3), 338345.
Leggett, A., Zarit, S.H., Nguyen, N.H., Hoang, C.N., Nguyen, H.T., 2012. The inuence
of social factors and health on depressive symptoms and worry: a study of
older Vietnamese adults. Aging Ment. Health 16 (6), 780786. http://dx.doi.org/
10.1080/13607863.2012.667780.
Leung, K.-K., Chen, C.-Y., Lue, B.-H., Hsu, S.-T. (2007). Social support and family
functioning on psychological symptoms in elderly Chinese. Archives of Gerontology and Geriatrics,44(2), 203-213. http://dx.doi.org/10.1016/j.archger.
2006.05.001.
Li, H., Ji, Y., Chen, T., 2014. The roles of different sources of social support on
emotional well-being among Chinese elderly. PloS ONE 9 (3), e90051. http://dx.
doi.org/10.1371/journal.pone.0090051.
Li, L.W., Liang, J., 2007. Social exchanges and subjective well-being among older
Chinese: does age make a difference? Psychol. Aging 22 (2), 386391. http://dx.
doi.org/10.1037/0882-7974.22.2.386.
Litwin, H., 2010. Social networks and well-being: a comparison of older people in
Mediterranean and non-Mediterranean countries. J. Gerontol. Ser. B: Psychol.
Sci. Soc. Sci. 65 (5), 599608. http://dx.doi.org/10.1093/geronb/gbp104.
Litwin, H., 2011. The association between social network relationships and
depressive symptoms among older Americans: what matters most? Int.
Psychogeriatr. 23 (6), 930940. http://dx.doi.org/10.1017/s1041610211000251.
Litwin, H., 2012. Physical activity, social network type, and depressive symptoms in
late life: an analysis of data from the National Social Life, Health and Aging
Project. Aging Ment. Health 16 (5), 608616. http://dx.doi.org/10.1080/
13607863.2011.644264.
Litwin, H., Landau, R., 2000. Social network type and social support among the old
old. J. Aging Stud. 14, 213228.
Mair, C., Diez Roux, A.V., Morenoff, J.D., 2010. Neighborhood stressors and social
support as predictors of depressive symptoms in the Chicago Community Adult
Health Study. Health Place 16 (5), 811819. http://dx.doi.org/10.1016/j.
healthplace.2010.04.006.
Martire, L.M., Stephens, M.A., Druley, J.A., Wojno, W.C., 2002. Negative reactions to
received spousal care: predictors and consequences of miscarried support.
Health Psychol. 21 (2), 167176.
Michie, S., Williams, S., 2003. Reducing work related psychological ill health and
sickness absence: a systematic literature review. Occup. Environ. Med. 60, 39.
Millan-Calenti, J.C., Sanchez, A., Lorenzo-Lopez, L., Cao, R., Maseda, A., 2013.
Inuence of social support on older adults with cognitive impairment,
depressive symptoms, or both coexisting. Int. J. Aging Hum. Dev. 76 (3),
199214.
Muramatsu, N., Yin, H., Hedeker, D., 2010. Functional declines, social support, and
mental health in the elderly: does living in a state supportive of home and
community-based services make a difference? Soc. Sci. Med. 70 (7), 10501058.
http://dx.doi.org/10.1016/j.socscimed.2009.12.005.
Newsom, J., Schultz, R., 1998. Caregiving from the recipient's perspective: negative
reactions to being helped. Health Psychol. 17 (2), 172181.
Noone, J., Stephens, C., 2014. Social Integration, Health and Quality of Life:
Summary report for the New Zealand Longitudinal Study of Ageing. Massey
University, Palmerston North, NZ.
Nurullah, A.S., 2012. Received and provided support: a review of current evidence
and future directions. Am. J. Health Stud. 27 (3), 173188.
Ostberg, V., Lennartsson, C., 2007. Getting by with a little help: the importance of
various types of social support for health problems. Scand. J. Public Health
35 (2), 197204. http://dx.doi.org/10.1080/14034940600813032.
Ottmann, G., Dickson, J., Wright, P., 2006. Social Connectedness and Health: A
Literature Review. Cornell University ILR School, Ithaca, NY.
Pettit, J.W., Roberts, R.E., Lewinsohn, P.M., Seeley, J.R., Yaroslavsky, I., 2011.
Developmental relations between perceived social support and depressive
symptoms through emerging adulthood: blood is thicker than water. J. Family
Psychol. 25 (1), 127136. http://dx.doi.org/10.1037/a0022320.
Plaisier, I., de Bruijn, J.G., de Graaf, R., ten Have, M., Beekman, A.T., Penninx, B.W.,
2007. The contribution of working conditions and social support to the onset of
depressive and anxiety disorders among male and female employees. Soc. Sci.
Med. 64 (2), 401410. http://dx.doi.org/10.1016/j.socscimed.2006.09.008.
Popay, J., Roberts, H., Sowden, A., Petticrew, M., Arai, L., Rodgers, M., Duffy, S., 2006.
Guidance on the Conduct of Narrative Synthesis in Systematic Reviews.
University of Lancaster, UK.
Pot, A.M., Deeg, D.J., Twisk, J.M., Beekman, A.T., Zarit, S.H., 2005. The longitudinal
relationship between the use of long-term care and depressive symptoms in
older adults. Gerontologist 45 (3), 359369.
Prince, M.J., Harwood, R.H., Blizard, R.A., Thomas, A., Mann, A.H., 1997. Social
support decits, loneliness and life events as risk factors for depression in old
age. The Gospel Oak Project VI. Psychol. Med. 27 (2), 323332.
Prince, M., Patel, V., Saxena, S., Maj, M., Maselko, J., Phillips, M.R., Rahman, A., 2007.
No health without mental health. The Lancet 370 (9590), 859877.
Reevy, G.M., 2007. Sex-related differences in the social support-stress relationship.
In: Monat, A., Lazarus, R.S., Reevy, G.M. (Eds.), The Praeger Handbook on Stress
and Coping, vol. 2nd. Praeger, Westport, CT, pp. 349361.
Resnick, M.D., Bearman, P.S., Blum, R.W., Bauman, K.E., Harris, K.M., Jones, J.,
Udry, J.R., 1997. Protecting adolescents from harm: ndings from the National
Longitudinal Study on Adolescent Health. JAMA 278 (10), 823832.
Roe, B., Whattam, M., Young, H., Dimond, M., 2001. Elders' perceptions of formal
and informal care: aspects of getting and receiving help for their activities of
daily living. J. Clin. Nurs. 10 (3), 398405.
Rosenquist, J.N., Fowler, J.H., Christakis, N.A., 2011. Social network determinants of
depression. Mol. Psychiatry 16 (3), 273281. http://dx.doi.org/10.1038/
mp.2010.13.
Rude, S.S., Burham, B.L., 1995. Connectedness and neediness: factors of the DEQ and
SAS dependency scales. Cogn. Ther. Res. 19, 323340.
Rugulies, R., Bultmann, U., Aust, B., Burr, H., 2006. Psychosocial work environment
and incidence of severe depressive symptoms: prospective ndings from a 5year follow-up of the Danish work environment cohort study. Am. J. Epidemiol.
163 (10), 877887. http://dx.doi.org/10.1093/aje/kwj119.
Russell, D., Taylor, J., 2009. Living alone and depressive symptoms: the inuence of
gender, physical disability, and social support among Hispanic and nonHispanic older adults. J. Gerontol. Ser. B: Psychol. Sci. Soc. Sci. 64 (1), 95104.
http://dx.doi.org/10.1093/geronb/gbn002.
Schultz, U., Schwarzer, R., 2004. Long-term effects of spousal support on coping
with cancer after surgery. J. Soc. Clin. Psychol. 23 (5), 716732.
Schuster, T.L., Kessler, R.C., Aseltine, R.H., 1990. Supportive interactions, negative
interactions, and depressed mood. Am. J. Community Psychol. 18, 423438.
Sicotte, M., Alvarado, B.E., Leon, E.M., Zunzunegui, M.V., 2008. Social networks and
depressive symptoms among elderly women and men in Havana, Cuba. Aging
Ment. Health 12 (2), 193201. http://dx.doi.org/10.1080/13607860701616358.
Smith, K.P., Christakis, N.A., 2008. Social networks and health. Annu. Rev. Sociol. 34,
405429.
Smith, P.M., Bielecky, A., 2012. The impact of changes in job strain and its
components on the risk of depression. Am. J. Public Health 102 (2), 352358.
http://dx.doi.org/10.2105/ajph.2011.300376.
Sonnenberg, C.M., Deeg, D.J., van Tilburg, T.G., Vink, D., Stek, M.L., Beekman, A.T.,
2013. Gender differences in the relation between depression and social support
in later life. Int. Psychogeriatr. 25 (1), 6170. http://dx.doi.org/10.1017/

s1041610212001202.
Stoetzer, U., Ahlberg, G., Johansson, G., Bergman, P., Hallsten, L., Forsell, Y.,
Lundberg, I., 2009. Problematic interpersonal relationships at work and
depression: a Swedish prospective cohort study. J. Occup. Health Psychol.
51 (2), 144151.
Stone, W., 2003. Ageing, Social Capital and Social Support. Australian Institute of
Family Studies, Melbourne, AU.
Tajvar, M., Fletcher, A., Grundy, E., Arab, M., 2013. Social support and health of older
people in Middle Eastern countries: a systematic review. Australas. J. Ageing 32
(2), 7178. http://dx.doi.org/10.1111/j.1741-6612.2012.00639.x.
Taylor, M.G., Lynch, S.M., 2004. Trajectories of impairment, social support, and
depressive symptoms in later life. J. Gerontol. Ser. B: Psychol. Sci. Soc. Sci. 59 (4),
S238246.
Teo, A.R., Choi, H., Valenstein, M., 2013. Social relationships and depression: tenyear follow-up from a nationally representative study. PLoS One 8 (4), e62396.
http://dx.doi.org/10.1371/journal.pone.0062396.
Thoits, P.A., 2011. Mechanisms linking social ties and support to physical and
mental health. J. Health Soc. Behav. 52, 145161.
Thomas, B.H., Ciliska, D., Dobbins, R.N., Micucci, S., 2004. A process for systematically reviewing the literature: providing the research evidence for public
health nursing interventions. Worldviews Evid.-Based Nurs. 1 (3), 176184.
Tiedt, A.D., 2010. The gender gap in depressive symptoms among Japanese elders:
evaluating social support and health as mediating factors. J. Cross-Cultural
Gerontol. 25 (3), 239256. http://dx.doi.org/10.1007/s10823-010-9122-x.
Tiikkainen, P., Heikkinen, R.L., 2005. Associations between loneliness, depressive
symptoms and perceived togetherness in older people. Aging Ment. Health
9 (6), 526534. http://dx.doi.org/10.1080/13607860500193138.
Townsend, K.C., McWhirter, B.T., 2005. Connectedness: a review of the literature
with implications for counseling, assessment, and research. J. Couns. Dev.
83 (2), 191201.
Tsai, Y.F., Yeh, S.H., Tsai, H.H., 2005. Prevalence and risk factors for depressive
symptoms among community-dwelling elders in Taiwan. Int. J. Geriatr.
Psychiatry 20 (11), 10971102. http://dx.doi.org/10.1002/gps.1413.
Tsuang, M.T., Tohen, M., Jones, P.B., 2011. Textbook of Psychiatric Epidemiology
3rd ed. Wiley-Blackwell, West Sussex, UK.
Uchino, B.N., 2006. Social support and health: a review of physiological processes
potentially underlying links to disease outcomes. J. Behav. Med. 29 (4),
377387.
65
Vaillant, G.E., 2008. Aging Well: Surprising Guideposts to a Happier Life from the
Landmark Study of Adult Development: Hachette Digital, Inc.
van Daalen, G., Sanders, K., Willemsen, T.M., 2005. Sources of social support as
predictors of health, psychological well-being and life satisfaction among Dutch
male and female dual-earners. Women Health 41 (2), 4362.
Sleep and Quality of Life in Clinical Medicine. In: Verster, J., Pandi-Perumal, S.R.,
Streiner, D.L. (Eds.), Humana Press, Totowa, NJ.
Virtanen, M., Koskinen, S., Kivimaki, M., Honkonen, T., Vahtera, J., Ahola, K.,
Lonnqvist, J., 2008. Contribution of non-work and work-related risk factors to
the association between income and mental disorders in a working population:
the Health 2000 Study. Occup. Environ. Med. 65 (3), 171178. http://dx.doi.org/
10.1136/oem.2007.033159.
Waldenstrom, K., Ahlberg, G., Bergman, P., Forsell, Y., Stoetzer, U., Waldenstrom, M.,
Lundberg, I., 2008. Externally assessed psychosocial work characteristics and
diagnoses of anxiety and depression. Occup. Environ. Med. 65 (2), 9096. http:
//dx.doi.org/10.1136/oem.2006.031252.
Walen, H.R., Lachman, M.E., 2000. Social support and strain from partner, family,
and friends: costs and benets for men and women in adulthood. J. Soc. Pers.
Relatsh. 17 (1), 530.
Wallsten, S.M., Tweed, D.L., Blazer, D.G., George, L.K., 1999. Disability and depressive symptoms in the elderly: the effects of instrumental support and its
subjective appraisal. Int. J. Aging Hum. Dev. 48 (2), 145159.
Wilby, F., 2011. Depression and social networks in community dwelling elders: a
descriptive study. J. Gerontol. Soc. Work 54 (3), 246259. http://dx.doi.org/
10.1080/01634372.2010.540074.
Williams, K.L., Galliher, R.V., 2006. Predicting depression and self-esteem from
social connectedness, support, and competence. J. Soc. Clin. Psychol. 25 (8),
855874.
Wolff, J.L., Agree, E.M., 2004. Depression among recipients of informal care: the
effects of reciprocity, respect, and adequacy of support. J. Gerontol., Ser. B:
Psychol. Sci. Soc. Sci. 59 (3), S173S180.
Yang, Y., 2006. How does functional disability affect depressive symptoms in late
life? The role of perceived social support and psychological resources. J. Health
Soc. Behav. 47 (4), 355372.
Zavaleta, D., Samuel, K., Mills, C., 2014. Social Isolation: A Conceptual and
Measurement Proposal. OPHI Working Papers, 67.
Zhang, B., Li, J., 2011. Gender and marital status differences in depressive symptoms
among elderly adults: the roles of family support and friend support. Aging
Ment. Health 15 (7), 844854. http://dx.doi.org/10.1080/13607863.2011.569481.

Journal of Affective Disorders

Uploaded by

Copyright:

Available Formats

Journal of Affective Disorders

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Journal of Affective Disorders

Uploaded by

Copyright:

Available Formats

Journal of Affective Disorders 175 (2015) 5365

Contents lists available at ScienceDirect

Journal of Affective Disorders

The association between social relationships and depression:

Z.I. Santini et al. / Journal of Affective Disorders 175 (2015) 5365

Z.I. Santini et al. / Journal of Affective Disorders 175 (2015) 5365

Z.I. Santini et al. / Journal of Affective Disorders 175 (2015) 5365

1675 were excluded because:

11 were excluded because they

Fig. 1. Selection of studies for systematic review.

of the study, study design, sample characteristics, data collection

The review included 28 cross-sectional and 23 prospective

Z.I. Santini et al. / Journal of Affective Disorders 175 (2015) 5365

Depression outcome measure

CrossSupport (perceived; received; provided;

The Center for Epidemiologic StudiesDepression scale (CES-D) short form

CrossSupport (perceived emotional)

The Depression Module of the Patient

CrossSupport (perceived emotional)

The Hospital Anxiety and Depression

CrossSupport (perceived emotional)

The Hospital Anxiety and Depression

CrossSupport (received emotional)

The Center for Epidemiologic StudiesDepression scale (CES-D)

CrossSupport (perceived instrumental and emotional)

The Center for Epidemiologic StudiesDepression scale (CES-D)

CrossSupport (perceived instrumental and emotional)

Self-reporting of the occurrence of

CrossSupport (perceived emotional)

The Center for Epidemiologic StudiesDepression scale (CES-D)

CrossSupport (perceived emotional)

The Center for Epidemiologic StudiesDepression scale (CES-D)

Network (contact frequency, size, structure) and

The EURO-D depression Scale

CrossNetwork (contact frequency; size) and network

The Center for Epidemiologic StudiesDepression scale (CES-D)

CrossCombined network (ties) and support (perceived

The Center for Epidemiologic StudiesDepression scale (CES-D)

The Geriatric Depression Scale (GDS)

CrossCombined network (size) and support (perceived

The Geriatric Depression Scale (GDS)

CrossSupport (perceived emotional) and negative

The Center for Epidemiologic StudiesDepression scale (CES-D) short form

The Center for Epidemiologic StudiesDepression scale (CES-D)

The Geriatric Depression Scale (GDS)

The Center for Epidemiologic StudiesDepression scale (CES-D)

The Center for Epidemiologic StudiesDepression scale (CES-D)

The Center for Epidemiologic StudiesDepression scale (CES-D)

Social relationships measure

CrossCombined network (size; contact frequency) and

The Geriatric Mental State (GMS) and the Low

The Symptom Checklist-90 revised (SCL90-R)

Z.I. Santini et al. / Journal of Affective Disorders 175 (2015) 5365

Chazelle et al. Ireland

Social relationships measure

Depression outcome measure

CrossCombined network (composition) and support

CrossSupport (perceived instrumental)