A Systematic Review of Research on Social Networks of Older Adults

A Systematic Review of Research on Social Networks of Older Adults Abstract Background and Objectives There has been a substantial interest in life course/life span changes in older adults’ social networks and in the relationship between social networks and health and wellbeing. The study embarked on a systematic review to examine the existing knowledgebase on social network in the field of gerontology. Our focus was on studies in which both ego (respondents) and his or her alters (network members) are queried about their social ties. Research Design and Methods We searched for studies published in English before September, 2017, relied on quantitative methods to obtain data from both ego (60 years of age and older) and alters and provided a quantitative account of the social network properties. We searched the following data sets: APA Psychnet, Pubmed, Sociological abstracts, and Ageline. This was followed by a snowball search of relevant articles using Google Scholar. Titles and abstracts were reviewed and selected articles were extracted independently by two reviewers. Results A total of 5,519 records were retrieved. Of these, 3,994 records remained after the removal of duplicates. Ten records reporting on five original samples were kept for the systematic review. One study described a social network of community dwelling older adults and the remaining studies described social networks of institutional older adults. Discussion and Implications The present study points to a lacuna in current understanding of social networks in the field of gerontology. It provides a useful review and possible tools for the design of future studies to address current shortcomings in the field. Systematic review, Sociocentric, Egocentric, Social network The term social network conveys the notion that individuals are embedded within a larger context of relational ties (Borgatti, Mehra, Brass, & Labianca, 2009). In the past few decades, there has been a tremendous amount of research on the role of social networks in the life of older adults (Antonucci & Akiyama, 1987; Sohn et al., 2017; Steinbach, 1992). Research has tended to classify older adults’ networks based on the type, quality, and/or quantity of the relationships (Litwin, 1995; Meeuwesen, Hortulanus, & Machielse, 2001; Nguyen, 2017; Park et al., 2015). For instance, research conducted in Israel has found that networks that consisted of diverse and friends-focused ties fared better in terms of health indicators, whereas community-clan networks were associated with less favorable outcomes (Litwin & Shiovitz-Ezra, 2006). Research conducted in other countries largely supported the relationship between network type and health outcomes, even though the proposed typology was somewhat different (Fiori, Antonucci, & Cortina, 2006; Litwin & Shiovitz-Ezra, 2011). For instance, in the U.S.-based sample (Litwin & Shiovitz-Ezra, 2011), the authors identified five types of social networks, whereas in the Israeli sample (Litwin & Shiovitz-Ezra, 2006), the authors identified six types of networks. Common to both studies was the identification of networks, which were based on friends, family, diverse members, and restricted networks. However, a different U.S.-based sample suggested two restricted networks, rather than one (Fiori, Antonucci, & Cortina, 2006). This line of research has shown that the type of social network one has impacts his or her longevity, mortality, quality of life, and health behaviors. Hence, social networks are thought to play an important role in the life of older adults. A prominent theory in the field is the convoy model of social relations which suggests that both life course and life span influences impact one’s social network (Borgatti et al., 2009). The context and the developmental life stage influence the formation of social relations, which can be characterized across several dimensions, including structure, function, and quality. Social relations are viewed as being multifaceted, constructed of both objective (e.g., number of ties), and subjective (quality of ties) characteristics (Antonucci, Ajrouch, & Birditt, 2014; Antonucci & Akiyama, 1987; Antonucci, Fiori, Birditt, & Jackey, 2010). This model has attracted attention by researchers who have shown that overall, there is a tendency for older adults’ social networks to shrink in old age (Cornwell, Laumann, & Schumm, 2008) and to consist of fewer peripheral network members (English & Carstensen, 2014). The nature of the relationship also changes over time. Research has shown that as older adults’ physical ability declines, members in their social network tend to provide them with higher levels of instrumental and personal support (Ducharme, Lévesque, Lachance, Kergoat, & Coulombe, 2011). In recognition of the importance of older adults’ social networks, several large-scale epidemiological studies have collected data on the topic (e.g., the National Social Life Health and Aging Project; NSHAPE http://www.norc.org/Research/Projects/Pages/national-social-life-health-and-aging-project.aspx; the Survey of Health, Ageing and Retirement in Europe, SHARE http://www.share-project.org/). Although informative, these studies have been conducted from the point of view of the ego, a focal person who provides information about his or her network. There has been very limited research to focus on the entire social network and to incorporate the point of view of various network members related to the ego, also known as alters. This is important because we know that social networks are relational in nature and involve more than a single individual (Seale, 2004). We also know that the individual’s perspective on his or her network does not fully correspond with alters’ perspectives on the network (Marsden, 2002). Hence, certain properties of the network can only be inferred by interviewing all network members. Moreover, studies that have shown that depression or loneliness are “contagious” within the social network (Cacioppo, Fowler, & Christakis, 2009; Rosenquist, Fowler, & Christakis, 2011) could only be conducted if both ego and alters are interviewed. Finally, interventions that target the structure of social ties or the contagion of certain behaviors or beliefs also are likely to benefit from data on the entire social network (Valente, 2012). The present study embarked on a systematic review of the literature to examine the existing knowledgebase on social network in the field of gerontology. We specifically focused only on networks obtained from the point of view of both egos and alters rather than networks that are limited to the ego perspective, given the unique potential for additional information inherent in the former type of networks. In order to facilitate research in the field, we provide detailed information about current methods and findings as well as recommendations for future research. Methods Obtaining the Data We searched the following databases: APA Psychnet, Pubmed, Sociological abstracts, and Ageline. The search was conducted in September, 2017. The following key words were used to guide the search: (“older adult*” OR elder* OR aged OR “nursing home*” OR senior* OR senescent OR “assisted living” OR “long term care” OR “nursing unit” OR “skilled nursing facility*” OR geriatric* OR “residential aged care” OR “adult day care*” OR “continuing care retirement community* OR “lifelong center*”) AND (“social network”). The latter search term was consistent with the term used in a recent review of social network research in a different population (Perkins, Subramanian, & Christakis, 2015). The bibliographies of all relevant review articles were searched. A Google Scholar search using the function “cited by” and “related articles” was used with all articles included in the present review in order to trace additional relevant articles. Selection of Studies All titles and available abstracts were reviewed for relevance by two independent researchers (LA, IL). Disagreements were resolved through a consensus. The following inclusion criteria were employed: (a) articles published on or before September 11, 2017 (when data search was concluded); (b) written in English; (c) relied on quantitative methods to obtain data from both ego and alters; (d) provided a description of data collection methods; (e) provided numeric information on network properties as a whole or at the ego level, and (f) all egos were 60 years old or over. Exclusion criteria: studies in which (a) only two or less alters were queried, as this provides a very limited picture of the whole network; (b) staff or other observers provided network data, rather than self-report of ego and alters; and (c) studies that were not available for a full review, such as conference proceedings. Different studies conducted on the same sample were described in detail but counted only once. Data Extraction Data extraction was conducted independently by the two reviewers. Disagreements between reviewers were discussed and a consensus agreement was established. Extraction details are available in Tables 1 and 2. Table 1. Study Characteristics Author Country Setting (community, adult day center, CCRC, etc.) Scope/definition of the network Number of participants Sample characteristics (age/gender) Abbott et al., 2012 USA One neighborhood in a residential long term care-assisted living facility All assisted living residents and staff 25/65 staff agreed to be photographed, 10/15 residents participated 86 y [82–92 y], 5 male, MMSE = 25.8[16–30] Abbott et al., 2015 USA An assisted living and dementia care unit in a nursing home Assisted living residents and staff, dementia care unit 10/15 residents of an assisted living-Same participants as in Abbott et al., (2012), 10/12 residents of a dementia unit Assisted living-Same participants as in Abbott et al., (2012); Dementia unit- 87 y [82–96 y], 80% female, MMSE = 17.2(14–25) Abbott & Pachucki, 2017 USA A dementia special care unit Dementia special care unit over three consecutive years 10 in wave 1, 10 in wave 2, 17 in wave 3, but only 3 people present across all three waves 90 y, 67–101 y, 80–90% women, dementia, MMSE = 16.9–19.8 Casey et al., 2016 Australia A nursing home: 3 care units, including a dementia unit Nursing home residents 36/94-only 29 reported on relationships 63–94 y, 61.1% female, 67% dementia Hardiman, 2017 USA A residential facility All residents (also asked about relationship with staff, nurses, caregivers in residential facility) 24/76 residents; 16/25 staff, nurses and caregivers provided pictures 100% female, 85 y[72–102 y], 16.6% mild dementia, MMSE = 20–24 Márquez-Serrano et al., 2012 Mexico A neighborhood in Mexico >60 y neighborhood residents who participated in the educational intervention 10/18 65–85 y, 70% female Schafer, 2011 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2013 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2015 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–-96 y], 72% female Schafer, 2016 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 72% female Author Country Setting (community, adult day center, CCRC, etc.) Scope/definition of the network Number of participants Sample characteristics (age/gender) Abbott et al., 2012 USA One neighborhood in a residential long term care-assisted living facility All assisted living residents and staff 25/65 staff agreed to be photographed, 10/15 residents participated 86 y [82–92 y], 5 male, MMSE = 25.8[16–30] Abbott et al., 2015 USA An assisted living and dementia care unit in a nursing home Assisted living residents and staff, dementia care unit 10/15 residents of an assisted living-Same participants as in Abbott et al., (2012), 10/12 residents of a dementia unit Assisted living-Same participants as in Abbott et al., (2012); Dementia unit- 87 y [82–96 y], 80% female, MMSE = 17.2(14–25) Abbott & Pachucki, 2017 USA A dementia special care unit Dementia special care unit over three consecutive years 10 in wave 1, 10 in wave 2, 17 in wave 3, but only 3 people present across all three waves 90 y, 67–101 y, 80–90% women, dementia, MMSE = 16.9–19.8 Casey et al., 2016 Australia A nursing home: 3 care units, including a dementia unit Nursing home residents 36/94-only 29 reported on relationships 63–94 y, 61.1% female, 67% dementia Hardiman, 2017 USA A residential facility All residents (also asked about relationship with staff, nurses, caregivers in residential facility) 24/76 residents; 16/25 staff, nurses and caregivers provided pictures 100% female, 85 y[72–102 y], 16.6% mild dementia, MMSE = 20–24 Márquez-Serrano et al., 2012 Mexico A neighborhood in Mexico >60 y neighborhood residents who participated in the educational intervention 10/18 65–85 y, 70% female Schafer, 2011 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2013 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2015 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–-96 y], 72% female Schafer, 2016 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 72% female Note. MMSE = Mini Mental Status Exam; y = years. View Large Table 1. Study Characteristics Author Country Setting (community, adult day center, CCRC, etc.) Scope/definition of the network Number of participants Sample characteristics (age/gender) Abbott et al., 2012 USA One neighborhood in a residential long term care-assisted living facility All assisted living residents and staff 25/65 staff agreed to be photographed, 10/15 residents participated 86 y [82–92 y], 5 male, MMSE = 25.8[16–30] Abbott et al., 2015 USA An assisted living and dementia care unit in a nursing home Assisted living residents and staff, dementia care unit 10/15 residents of an assisted living-Same participants as in Abbott et al., (2012), 10/12 residents of a dementia unit Assisted living-Same participants as in Abbott et al., (2012); Dementia unit- 87 y [82–96 y], 80% female, MMSE = 17.2(14–25) Abbott & Pachucki, 2017 USA A dementia special care unit Dementia special care unit over three consecutive years 10 in wave 1, 10 in wave 2, 17 in wave 3, but only 3 people present across all three waves 90 y, 67–101 y, 80–90% women, dementia, MMSE = 16.9–19.8 Casey et al., 2016 Australia A nursing home: 3 care units, including a dementia unit Nursing home residents 36/94-only 29 reported on relationships 63–94 y, 61.1% female, 67% dementia Hardiman, 2017 USA A residential facility All residents (also asked about relationship with staff, nurses, caregivers in residential facility) 24/76 residents; 16/25 staff, nurses and caregivers provided pictures 100% female, 85 y[72–102 y], 16.6% mild dementia, MMSE = 20–24 Márquez-Serrano et al., 2012 Mexico A neighborhood in Mexico >60 y neighborhood residents who participated in the educational intervention 10/18 65–85 y, 70% female Schafer, 2011 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2013 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2015 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–-96 y], 72% female Schafer, 2016 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 72% female Author Country Setting (community, adult day center, CCRC, etc.) Scope/definition of the network Number of participants Sample characteristics (age/gender) Abbott et al., 2012 USA One neighborhood in a residential long term care-assisted living facility All assisted living residents and staff 25/65 staff agreed to be photographed, 10/15 residents participated 86 y [82–92 y], 5 male, MMSE = 25.8[16–30] Abbott et al., 2015 USA An assisted living and dementia care unit in a nursing home Assisted living residents and staff, dementia care unit 10/15 residents of an assisted living-Same participants as in Abbott et al., (2012), 10/12 residents of a dementia unit Assisted living-Same participants as in Abbott et al., (2012); Dementia unit- 87 y [82–96 y], 80% female, MMSE = 17.2(14–25) Abbott & Pachucki, 2017 USA A dementia special care unit Dementia special care unit over three consecutive years 10 in wave 1, 10 in wave 2, 17 in wave 3, but only 3 people present across all three waves 90 y, 67–101 y, 80–90% women, dementia, MMSE = 16.9–19.8 Casey et al., 2016 Australia A nursing home: 3 care units, including a dementia unit Nursing home residents 36/94-only 29 reported on relationships 63–94 y, 61.1% female, 67% dementia Hardiman, 2017 USA A residential facility All residents (also asked about relationship with staff, nurses, caregivers in residential facility) 24/76 residents; 16/25 staff, nurses and caregivers provided pictures 100% female, 85 y[72–102 y], 16.6% mild dementia, MMSE = 20–24 Márquez-Serrano et al., 2012 Mexico A neighborhood in Mexico >60 y neighborhood residents who participated in the educational intervention 10/18 65–85 y, 70% female Schafer, 2011 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2013 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2015 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–-96 y], 72% female Schafer, 2016 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 72% female Note. MMSE = Mini Mental Status Exam; y = years. View Large Table 2. Main Findings and Study Design Author Study design (cross-sectional/ longitudinal, etc.) Network types measured and analyzed (friendship/ familiarity etc.) Exact questions used to derive the network Method used to obtain the network (name generator, census report) Statistical method Basic network properties reported Map provided Main findings Abbott et al., 2012 Cross-sectional Social interactions in the past week The AL residents were asked the same 3 questions about people in their own and other neighborhoods within the same AL facility and about staff who worked at the AL facility: “5 people you spent most time with in the past week, 5 people who provided help to you in the past week, 5 people you provided help to in the past week.” Photos were used only when asked about residents in own community. Sociocentric with pictures of all AL residents and staff Descriptive Degree centrality, in-degree, out-degree Yes Descriptive analysis of AL- indegree, outdegree overall degree centrality ranges between 1 and 23 Abbott et al., 2015 Cross-sectional Social interactions in the past week/recently Sociocentric as in Abbott et al.,(2012). In the dementia unit, “recently spent time” with rather than “over the past week” was used for reference (Egocentric-not relevant for this review) and sociocentric with photos Descriptive Sociogram, network size, degree centrality, reciprocity Yes Sociocentric vs. egocentric approach nominations. Mean sociocentric network size 7[0–14]. Residents with a higher MMSE score nominated more individuals. Reciprocity in AL 55% [20%–75%] Abbott & Pachucki, 2017 Cross-sectional—three waves of data but only 3 residents went through all waves of data collection Social interactions recently Photos were provided. 5 questions asked: who they spent time with, listened to problems, helped with something, who helped them, who listened to them. Sociocentric with photos Associations of quality of life and cognition with network characteristics Personal network size, in-degree, out-degree, betweeness- centrality, closeness- centrality, network density Yes Average personal network size 2–4.5. Half the ties were reciprocated and there was a positive association between integration- betweeness centrality and quality of life, but inconsistent associations with cognitive functioning. Friendship ties were more frequent among people of adjacent cognitive status categories Casey et al., 2016 Cross-sectional Social support, friendship Showing photos and asking to identify friends; followed-up by questions about true friend vs. casual friend; (integrated with qualitative and observational data) Sociocentric with photos Correlations between ties and assessment scores Out-degree, in-degree, reciprocity, density, path length Yes When all potential relationship were included in the unit, the median size was 0 [0–1]. Low density, low reciprocity (22.2%), high levels of isolation; cognitive impairment was negatively correlated with network size and reciprocity, friendship correlated with reciprocity Hardiman, 2017 Cross-sectional Spend time, provide help, receive help 1. Pick up to or point out five people with whom you spent the most time in the past two weeks. 2. Pick up to or point out five people who provided help to you in the past week. 3. Pick up to or point out five people you provided help to in the past week. *Sociocentric questions are to be asked about residents of the NH and with regards to staff, nurses, or caregivers. A list of names and photos of those who agreed to have their picture taken- both staff, nurses, caregivers and residents Correlational Total residents and staff nominated No Staff had a higher in-degree centrality. A correlation between number of connections with others, quality of life and MMSE score Márquez-Serrano et al., 2012 Pre-post: but networks measured only once Relationship types Indicate the individuals with whom they have relationships: family, friends, neighbors or work colleagues Egocentric Network characteristics in relation to flu infection Structure, density, degree Yes Low density: .0167. Out-degree: 1.57. No knowledge transfer within the network Schafer, 2016 Cross-sectional Spend time, confidant Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. List the people they talked to about important matters. A map of the continuing care retirement community Regression analysis in-degree, out- degree, Bonacich centrality Yes Time spent (mean 20.09) had more ties than confidant relations (mean 2.24). Less asymmetry in confidant relations (.2) than in time spent (.33). Health as a predictor of Bonacich centrality, in-degree and out-degree. Significant results for some but not all types of ties. Schafer, 2013 Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Regression analysis Constraint, integration No Health predicted structural position. Residents with the best health had positional advantage in the network. Schafer, 2015 Cross-sectional Confidant “From time to time, people will often talk with others about things that are important to them. This could include sharing good news, or bad news, sharing about concerns they might have, or otherwise just talking about things they find very important or significant. Who are the people here at [RC] that you can talk to about things that are important to you?” A map of the continuing care retirement community Exponential random graph Close ties sent and received, density Yes An overall density score of .02. Close relations are influenced by physical proximity. Physical proximity intensified health based a-symmetry. (Schafer. 2016) Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Exponential random graph Total ties, density, geodesic distances, reciprocated ties, centrality measures No Healthier individuals received more social tie nominations. Only modest support for health-based homophily. Author Study design (cross-sectional/ longitudinal, etc.) Network types measured and analyzed (friendship/ familiarity etc.) Exact questions used to derive the network Method used to obtain the network (name generator, census report) Statistical method Basic network properties reported Map provided Main findings Abbott et al., 2012 Cross-sectional Social interactions in the past week The AL residents were asked the same 3 questions about people in their own and other neighborhoods within the same AL facility and about staff who worked at the AL facility: “5 people you spent most time with in the past week, 5 people who provided help to you in the past week, 5 people you provided help to in the past week.” Photos were used only when asked about residents in own community. Sociocentric with pictures of all AL residents and staff Descriptive Degree centrality, in-degree, out-degree Yes Descriptive analysis of AL- indegree, outdegree overall degree centrality ranges between 1 and 23 Abbott et al., 2015 Cross-sectional Social interactions in the past week/recently Sociocentric as in Abbott et al.,(2012). In the dementia unit, “recently spent time” with rather than “over the past week” was used for reference (Egocentric-not relevant for this review) and sociocentric with photos Descriptive Sociogram, network size, degree centrality, reciprocity Yes Sociocentric vs. egocentric approach nominations. Mean sociocentric network size 7[0–14]. Residents with a higher MMSE score nominated more individuals. Reciprocity in AL 55% [20%–75%] Abbott & Pachucki, 2017 Cross-sectional—three waves of data but only 3 residents went through all waves of data collection Social interactions recently Photos were provided. 5 questions asked: who they spent time with, listened to problems, helped with something, who helped them, who listened to them. Sociocentric with photos Associations of quality of life and cognition with network characteristics Personal network size, in-degree, out-degree, betweeness- centrality, closeness- centrality, network density Yes Average personal network size 2–4.5. Half the ties were reciprocated and there was a positive association between integration- betweeness centrality and quality of life, but inconsistent associations with cognitive functioning. Friendship ties were more frequent among people of adjacent cognitive status categories Casey et al., 2016 Cross-sectional Social support, friendship Showing photos and asking to identify friends; followed-up by questions about true friend vs. casual friend; (integrated with qualitative and observational data) Sociocentric with photos Correlations between ties and assessment scores Out-degree, in-degree, reciprocity, density, path length Yes When all potential relationship were included in the unit, the median size was 0 [0–1]. Low density, low reciprocity (22.2%), high levels of isolation; cognitive impairment was negatively correlated with network size and reciprocity, friendship correlated with reciprocity Hardiman, 2017 Cross-sectional Spend time, provide help, receive help 1. Pick up to or point out five people with whom you spent the most time in the past two weeks. 2. Pick up to or point out five people who provided help to you in the past week. 3. Pick up to or point out five people you provided help to in the past week. *Sociocentric questions are to be asked about residents of the NH and with regards to staff, nurses, or caregivers. A list of names and photos of those who agreed to have their picture taken- both staff, nurses, caregivers and residents Correlational Total residents and staff nominated No Staff had a higher in-degree centrality. A correlation between number of connections with others, quality of life and MMSE score Márquez-Serrano et al., 2012 Pre-post: but networks measured only once Relationship types Indicate the individuals with whom they have relationships: family, friends, neighbors or work colleagues Egocentric Network characteristics in relation to flu infection Structure, density, degree Yes Low density: .0167. Out-degree: 1.57. No knowledge transfer within the network Schafer, 2016 Cross-sectional Spend time, confidant Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. List the people they talked to about important matters. A map of the continuing care retirement community Regression analysis in-degree, out- degree, Bonacich centrality Yes Time spent (mean 20.09) had more ties than confidant relations (mean 2.24). Less asymmetry in confidant relations (.2) than in time spent (.33). Health as a predictor of Bonacich centrality, in-degree and out-degree. Significant results for some but not all types of ties. Schafer, 2013 Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Regression analysis Constraint, integration No Health predicted structural position. Residents with the best health had positional advantage in the network. Schafer, 2015 Cross-sectional Confidant “From time to time, people will often talk with others about things that are important to them. This could include sharing good news, or bad news, sharing about concerns they might have, or otherwise just talking about things they find very important or significant. Who are the people here at [RC] that you can talk to about things that are important to you?” A map of the continuing care retirement community Exponential random graph Close ties sent and received, density Yes An overall density score of .02. Close relations are influenced by physical proximity. Physical proximity intensified health based a-symmetry. (Schafer. 2016) Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Exponential random graph Total ties, density, geodesic distances, reciprocated ties, centrality measures No Healthier individuals received more social tie nominations. Only modest support for health-based homophily. Note. AL = assisted living; betweeness centrality = number of shortest paths from all nodes (alters) that path through the ego; Bonacich centrality = takes into account the number of connections within the ego network; degree centrality = the number of ties one has; density = actual ties/all possible ties; in-degree = number of incoming ties; MMSE = Mini Mental Status Exam; NH = nursing home; out-degree = number of outgoing ties; reciprocity = the likelihood of a tie to be mutually linked; RC = retirement community. View Large Table 2. Main Findings and Study Design Author Study design (cross-sectional/ longitudinal, etc.) Network types measured and analyzed (friendship/ familiarity etc.) Exact questions used to derive the network Method used to obtain the network (name generator, census report) Statistical method Basic network properties reported Map provided Main findings Abbott et al., 2012 Cross-sectional Social interactions in the past week The AL residents were asked the same 3 questions about people in their own and other neighborhoods within the same AL facility and about staff who worked at the AL facility: “5 people you spent most time with in the past week, 5 people who provided help to you in the past week, 5 people you provided help to in the past week.” Photos were used only when asked about residents in own community. Sociocentric with pictures of all AL residents and staff Descriptive Degree centrality, in-degree, out-degree Yes Descriptive analysis of AL- indegree, outdegree overall degree centrality ranges between 1 and 23 Abbott et al., 2015 Cross-sectional Social interactions in the past week/recently Sociocentric as in Abbott et al.,(2012). In the dementia unit, “recently spent time” with rather than “over the past week” was used for reference (Egocentric-not relevant for this review) and sociocentric with photos Descriptive Sociogram, network size, degree centrality, reciprocity Yes Sociocentric vs. egocentric approach nominations. Mean sociocentric network size 7[0–14]. Residents with a higher MMSE score nominated more individuals. Reciprocity in AL 55% [20%–75%] Abbott & Pachucki, 2017 Cross-sectional—three waves of data but only 3 residents went through all waves of data collection Social interactions recently Photos were provided. 5 questions asked: who they spent time with, listened to problems, helped with something, who helped them, who listened to them. Sociocentric with photos Associations of quality of life and cognition with network characteristics Personal network size, in-degree, out-degree, betweeness- centrality, closeness- centrality, network density Yes Average personal network size 2–4.5. Half the ties were reciprocated and there was a positive association between integration- betweeness centrality and quality of life, but inconsistent associations with cognitive functioning. Friendship ties were more frequent among people of adjacent cognitive status categories Casey et al., 2016 Cross-sectional Social support, friendship Showing photos and asking to identify friends; followed-up by questions about true friend vs. casual friend; (integrated with qualitative and observational data) Sociocentric with photos Correlations between ties and assessment scores Out-degree, in-degree, reciprocity, density, path length Yes When all potential relationship were included in the unit, the median size was 0 [0–1]. Low density, low reciprocity (22.2%), high levels of isolation; cognitive impairment was negatively correlated with network size and reciprocity, friendship correlated with reciprocity Hardiman, 2017 Cross-sectional Spend time, provide help, receive help 1. Pick up to or point out five people with whom you spent the most time in the past two weeks. 2. Pick up to or point out five people who provided help to you in the past week. 3. Pick up to or point out five people you provided help to in the past week. *Sociocentric questions are to be asked about residents of the NH and with regards to staff, nurses, or caregivers. A list of names and photos of those who agreed to have their picture taken- both staff, nurses, caregivers and residents Correlational Total residents and staff nominated No Staff had a higher in-degree centrality. A correlation between number of connections with others, quality of life and MMSE score Márquez-Serrano et al., 2012 Pre-post: but networks measured only once Relationship types Indicate the individuals with whom they have relationships: family, friends, neighbors or work colleagues Egocentric Network characteristics in relation to flu infection Structure, density, degree Yes Low density: .0167. Out-degree: 1.57. No knowledge transfer within the network Schafer, 2016 Cross-sectional Spend time, confidant Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. List the people they talked to about important matters. A map of the continuing care retirement community Regression analysis in-degree, out- degree, Bonacich centrality Yes Time spent (mean 20.09) had more ties than confidant relations (mean 2.24). Less asymmetry in confidant relations (.2) than in time spent (.33). Health as a predictor of Bonacich centrality, in-degree and out-degree. Significant results for some but not all types of ties. Schafer, 2013 Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Regression analysis Constraint, integration No Health predicted structural position. Residents with the best health had positional advantage in the network. Schafer, 2015 Cross-sectional Confidant “From time to time, people will often talk with others about things that are important to them. This could include sharing good news, or bad news, sharing about concerns they might have, or otherwise just talking about things they find very important or significant. Who are the people here at [RC] that you can talk to about things that are important to you?” A map of the continuing care retirement community Exponential random graph Close ties sent and received, density Yes An overall density score of .02. Close relations are influenced by physical proximity. Physical proximity intensified health based a-symmetry. (Schafer. 2016) Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Exponential random graph Total ties, density, geodesic distances, reciprocated ties, centrality measures No Healthier individuals received more social tie nominations. Only modest support for health-based homophily. Author Study design (cross-sectional/ longitudinal, etc.) Network types measured and analyzed (friendship/ familiarity etc.) Exact questions used to derive the network Method used to obtain the network (name generator, census report) Statistical method Basic network properties reported Map provided Main findings Abbott et al., 2012 Cross-sectional Social interactions in the past week The AL residents were asked the same 3 questions about people in their own and other neighborhoods within the same AL facility and about staff who worked at the AL facility: “5 people you spent most time with in the past week, 5 people who provided help to you in the past week, 5 people you provided help to in the past week.” Photos were used only when asked about residents in own community. Sociocentric with pictures of all AL residents and staff Descriptive Degree centrality, in-degree, out-degree Yes Descriptive analysis of AL- indegree, outdegree overall degree centrality ranges between 1 and 23 Abbott et al., 2015 Cross-sectional Social interactions in the past week/recently Sociocentric as in Abbott et al.,(2012). In the dementia unit, “recently spent time” with rather than “over the past week” was used for reference (Egocentric-not relevant for this review) and sociocentric with photos Descriptive Sociogram, network size, degree centrality, reciprocity Yes Sociocentric vs. egocentric approach nominations. Mean sociocentric network size 7[0–14]. Residents with a higher MMSE score nominated more individuals. Reciprocity in AL 55% [20%–75%] Abbott & Pachucki, 2017 Cross-sectional—three waves of data but only 3 residents went through all waves of data collection Social interactions recently Photos were provided. 5 questions asked: who they spent time with, listened to problems, helped with something, who helped them, who listened to them. Sociocentric with photos Associations of quality of life and cognition with network characteristics Personal network size, in-degree, out-degree, betweeness- centrality, closeness- centrality, network density Yes Average personal network size 2–4.5. Half the ties were reciprocated and there was a positive association between integration- betweeness centrality and quality of life, but inconsistent associations with cognitive functioning. Friendship ties were more frequent among people of adjacent cognitive status categories Casey et al., 2016 Cross-sectional Social support, friendship Showing photos and asking to identify friends; followed-up by questions about true friend vs. casual friend; (integrated with qualitative and observational data) Sociocentric with photos Correlations between ties and assessment scores Out-degree, in-degree, reciprocity, density, path length Yes When all potential relationship were included in the unit, the median size was 0 [0–1]. Low density, low reciprocity (22.2%), high levels of isolation; cognitive impairment was negatively correlated with network size and reciprocity, friendship correlated with reciprocity Hardiman, 2017 Cross-sectional Spend time, provide help, receive help 1. Pick up to or point out five people with whom you spent the most time in the past two weeks. 2. Pick up to or point out five people who provided help to you in the past week. 3. Pick up to or point out five people you provided help to in the past week. *Sociocentric questions are to be asked about residents of the NH and with regards to staff, nurses, or caregivers. A list of names and photos of those who agreed to have their picture taken- both staff, nurses, caregivers and residents Correlational Total residents and staff nominated No Staff had a higher in-degree centrality. A correlation between number of connections with others, quality of life and MMSE score Márquez-Serrano et al., 2012 Pre-post: but networks measured only once Relationship types Indicate the individuals with whom they have relationships: family, friends, neighbors or work colleagues Egocentric Network characteristics in relation to flu infection Structure, density, degree Yes Low density: .0167. Out-degree: 1.57. No knowledge transfer within the network Schafer, 2016 Cross-sectional Spend time, confidant Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. List the people they talked to about important matters. A map of the continuing care retirement community Regression analysis in-degree, out- degree, Bonacich centrality Yes Time spent (mean 20.09) had more ties than confidant relations (mean 2.24). Less asymmetry in confidant relations (.2) than in time spent (.33). Health as a predictor of Bonacich centrality, in-degree and out-degree. Significant results for some but not all types of ties. Schafer, 2013 Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Regression analysis Constraint, integration No Health predicted structural position. Residents with the best health had positional advantage in the network. Schafer, 2015 Cross-sectional Confidant “From time to time, people will often talk with others about things that are important to them. This could include sharing good news, or bad news, sharing about concerns they might have, or otherwise just talking about things they find very important or significant. Who are the people here at [RC] that you can talk to about things that are important to you?” A map of the continuing care retirement community Exponential random graph Close ties sent and received, density Yes An overall density score of .02. Close relations are influenced by physical proximity. Physical proximity intensified health based a-symmetry. (Schafer. 2016) Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Exponential random graph Total ties, density, geodesic distances, reciprocated ties, centrality measures No Healthier individuals received more social tie nominations. Only modest support for health-based homophily. Note. AL = assisted living; betweeness centrality = number of shortest paths from all nodes (alters) that path through the ego; Bonacich centrality = takes into account the number of connections within the ego network; degree centrality = the number of ties one has; density = actual ties/all possible ties; in-degree = number of incoming ties; MMSE = Mini Mental Status Exam; NH = nursing home; out-degree = number of outgoing ties; reciprocity = the likelihood of a tie to be mutually linked; RC = retirement community. View Large Results Figure 1 demonstrates the study flow chart according to PRISMA guidelines (Moher, Liberati, Tetzlaff, Altman, & PRISMA Group, 2009). The PRISMA diagram maps the flow of information through the various review stages. The diagram outlines the number of articles retrieved, number of articles kept for full review, number of articles excluded, reasons for exclusion and number of articles kept for analysis. A total of 5,519 records were retrieved into an endnote library from the four data sets searched. Of these, 3,994 records remained after the removal of duplicates and 60 records were maintained for a thorough review because they appeared to be potentially relevant based on their title and abstract. Four additional records were obtained through Google Scholar search of already identified articles. Please see Figure 1 for details concerning study flow and reasons for exclusion. Figure 1. View largeDownload slide Study flow. Figure 1. View largeDownload slide Study flow. Overall, 10 records reporting on five original samples were kept for the systematic review. One study addressed community dwelling older adults in Mexico (Márquez-Serrano, González-Juárez, Castillo-Castillo, González-González, & Idrovo, 2012). This study used an egocentric method which relied on a name generator to identify the network. The remaining studies addressed older adults in long-term care institutes in the United States and Australia. Three of the studies presented data on social networks among older adults with dementia or mild cognitive impairment (Abbott, Bettger, Hampton, & Kohler, 2012; Abbott, Bettger, Hampton, & Kohler, 2015; Abbott & Pachucki, 2017; Casey, Low, Jeon, & Brodaty, 2016). Under these circumstances, a common approach appeared to be the use of photographs to construct a complete social network. One study employed a pre-post design, but network measures were obtained only once (Márquez-Serrano et al., 2012). Another study had three waves of data collection, but because of the variability in the network over time, the analysis focused on each network separately rather than on changes in network dynamics (Abbott & Pachucki, 2017). All other studies employed a cross-sectional design. Studies provided descriptive data on the ego network and correlational data to examine the associations of various types of social relations indicators with quality of life, cognitive functioning, and other health measures (Abbott & Pachucki, 2017; Casey et al., 2016; Hardiman, 2017). Two records presented results based on an exponential random graph to examine relationship quality or network position and structure as outcomes (Schafer, 2015, 2016), and two other records used regression analyses to examine health as a predictor of social network properties (Schafer, 2011, 2013). Studies have alluded to a bi-directional relationship between health and well-being and various network properties calculated at the ego level. See Tables 1 and 2 for details. Discussion The present study is the first systematic review of whole social networks among older adults. The findings suggest that in contrast to the plethora of research that has examined social networks of older adults from the ego perspective (Nyqvist, Forsman, Giuntoli, & Cattan, 2013; Smith, Banting, Eime, O’Sullivan, & van Uffelen, 2017), a few studies have examined both ego and alters in the field of gerontology. This finding is disappointing given the growing interest in social networks for the understanding of health and wellbeing in the general population (Lazer et al., 2009). The findings suggest that the study of whole social networks among older adults is feasible, even in the case of older adults with mild-to-moderate dementia (Abbott et al., 2015; Hardiman, 2017). Moreover, these studies allude to a unique opportunity available to those interested in the study of whole social networks of older adults in institutions. This is because the boundaries of the network are already pre-defined. Hence, this could provide an incentive to conducting a sociocentric study in which all network members are queried. Studies conducted in the United States and Australia relied exclusively on this property to examine an entire unit or institute in order to produce whole social network data (Abbott et al., 2015; Casey et al., 2016; Hardiman, 2017; Schafer, 2011). In contrast, a study conducted in Mexico (Márquez-Serrano et al., 2012) relied on the close-knit nature of community dwellers who participated in an educational intervention to develop a model which capitalized on overlap in ego networks (Márquez-Serrano et al., 2012). A community-dwelling living arrangement, which allows for the use of ego networks to construct full social networks due to overlap in ego networks, is less common in urban places, which characterize the global North. A potentially interesting and useful tool for collecting whole network data can be the use of snowball sampling several degrees away from a focal person (Antonucci & Israel, 1986; Bear, 1990). Such an approach can potentially help researchers to understand the social network not only from the perspective of the older adult, but also from the perspective of his or her alters even in settings, where boundaries are not predefined and there is no expectation for a natural overlap of ego networks. Although our review identified several studies that relied on such a method, these studies were limited to one or two additional alters or examined a focal person who did not meet our age criteria (Carpentier & Ducharme, 2007; Koehly, Ashida, Schafer, & Ludden, 2015). Despite the complexity of this type of design, it is particularly valuable for use in populations that do not have a-priori set boundaries, such as urban, community dwelling older adults. Based on the studies reviewed, one can infer that the concept of friendship can be used to describe older adults’ networks (Abbott et al., 2012; Casey et al., 2016). In general, networks in institutions are characterized by low density (number of actual ties divided by the number of all possible ties) and reciprocity (e.g., if ego knows alter, alter also knows ego) and high levels of isolation (no incoming/outgoing ties) (Casey et al., 2016; Schafer, 2011). This appears to be the case also in the community (Márquez-Serrano et al., 2012). Although some of the studies found a correlation between health, quality of life, cognitive functioning, and network characteristics (Hardiman, 2017; Schafer, 2013, 2015), the relationship appears to be bidirectional and given the cross-sectional design of the studies, it is impossible to determine its exact direction. Unfortunately, the small and varied nature of the studies reviewed does not allow determining the size or direction of these potential effects. Implications To sum, despite a plethora of research on social networks in older adults, there is only a handful of studies on whole social networks in this population. Hence, many studies disregard the fact that social interactions are bidirectional, with all stakeholders contributing to the shared meaning of the encounter (Seale, 2004). Moreover, the limited research comparing ego networks to sociocentric networks has shown that the two are not always comparable and that valuable information can be obtained by querying both ego and alters, rather than relying only on the ego (Chung, Hossain, & Davis, 2005). Examining the structure and the function of the entire network is important for several reasons. First, this can provide additional insights and enrich existing theories in the field of aging which have been exclusively based on the ego-perspective. Second, such an approach can provide information about the network as a whole, which is not limited to the ego perspective. Third, there is a growing interest in social networks as an intervention tool (Valente, 2012). These interventions address the structure of the network. For instance, by identifying those individuals who are more prone to isolation, one can initiate interventions for re-integration into the network. Alternatively, interventions can instigate a change in the network through contagion. For instance, interventions that promote the adaption of specific health behaviors within the network (Valente & Davis, 1999). The present study points to a lacuna in current understanding of social networks in the field of gerontology. It also provides useful tools for the design of future studies to address current shortcomings in the field. However, as is always the case with systematic reviews, there is a chance that our search had failed to include papers that should have been included. Also, given the very small number of articles and their heterogeneity, we cannot make concrete predictions about the nature of older adults’ full social networks nor about their correlates. Finally, our inclusion/exclusion criteria might have resulted in the exclusion of important studies in the field. For instance, we excluded qualitative studies because of our interest in numeric network properties (e.g., density, degree centrality) and in self-report of social ties. It is important to note, though, that this is in line with the methodology of systematic reviews, which requires strict and coherent inclusion/exclusion criteria, as long as they follow a clear rationale. Funding The study was funded by the Israel Science Foundation 537/16. Conflict of Interest None reported. References Abbott , K. M. , Bettger , J. P. , Hampton , K. , & Kohler , H. P . ( 2012 ). Exploring the use of social network analysis to measure social integration among older adults in assisted living . Family and Community Health , 35 , 322 – 333 . doi: 10.1097/FCH.0b013e318266669f Google Scholar CrossRef Search ADS PubMed Abbott , K. M. , Bettger , J. P. , Hampton , K. N. , & Kohler , H. P . ( 2015 ). The feasibility of measuring social networks among older adults in assisted living and dementia special care units . Dementia (London, England) , 14 , 199 – 219 . doi: 10.1177/1471301213494524 Google Scholar PubMed Abbott , K. M. , & Pachucki , M. C . ( 2017 ). Associations between social network characteristics, cognitive function, and quality of life among residents in a dementia special care unit: A pilot study . Dementia (London, England) , 16 , 1004 – 1019 . doi: 10.1177/1471301216630907 Google Scholar PubMed Antonucci , T. C. , Ajrouch , K. J. , & Birditt , K. S . ( 2014 ). The convoy model: Explaining social relations from a multidisciplinary perspective . The Gerontologist , 54 , 82 – 92 . doi: 10.1093/geront/gnt118 Google Scholar CrossRef Search ADS PubMed Antonucci , T. C. , & Akiyama , H . ( 1987 ). Social networks in adult life and a preliminary examination of the convoy model . Journal of Gerontology , 42 , 519 – 527 . doi: 10.1093/geronj/42.5.519 Google Scholar CrossRef Search ADS PubMed Antonucci , T. C. , Fiori , K. L. , Birditt , K. , & Jackey , L. M. H . ( 2010 ). Convoys of Social Relations: Integrating Life-Span and Life-Course Perspectives . In The Handbook of Life-Span Development . New York, NY : John Wiley & Sons, Inc . Google Scholar CrossRef Search ADS Antonucci , T. C. , & Israel , B. A . ( 1986 ). Veridicality of social support: A comparison of principal and network members’ responses . Journal of Consulting and Clinical Psychology , 54 , 432 – 437 . doi: 10.1037/0022-006X.54.4.432 Google Scholar CrossRef Search ADS PubMed Bear , M . ( 1990 ). Social network characteristics and the duration of primary relationships after entry into long-term care . Journal of Gerontology , 45 , S156 – S162 . doi: 10.1093/geronj/45.4.S156 Google Scholar CrossRef Search ADS PubMed Borgatti , S. P. , Mehra , A. , Brass , D. J. , & Labianca , G . ( 2009 ). Network analysis in the social sciences . Science (New York, N.Y.) , 323 , 892 – 895 . doi: 10.1126/science.1165821 Google Scholar CrossRef Search ADS PubMed Cacioppo , J. T. , Fowler , J. H. , & Christakis , N. A . ( 2009 ). Alone in the crowd: the structure and spread of loneliness in a large social network . Journal of Personality and Social Psychology , 97 , 977 – 991 . doi: 10.1037/a0016076 Google Scholar CrossRef Search ADS PubMed Carpentier , N. , & Ducharme , F . ( 2007 ). Social network data validity: The example of the social network of caregivers of older persons with Alzheimer-type dementia . Canadian Journal on Aging = La revue Canadienne Du Vieillissement , 26 ( Suppl. 1 ), 103 – 115 . doi: 10.3138/cja.26.suppl_1.103 Google Scholar CrossRef Search ADS Casey , A. N. , Low , L. F. , Jeon , Y. H. , & Brodaty , H . ( 2016 ). Residents perceptions of friendship and positive social networks within a nursing home . The Gerontologist , 56 , 855 – 867 . doi: 10.1093/geront/gnv146 Google Scholar CrossRef Search ADS PubMed Chung , K. K. , Hossain , L. , & Davis , J . ( 2005 ). Exploring sociocentric and egocentric approaches for social network analysis . Paper presented at the Proceedings of the 2nd international conference on knowledge management in Asia Pacific , Victoria University of Wellington, New Zealand . Cornwell , B. , Laumann , E. O. , & Schumm , L. P . ( 2008 ). The social connectedness of older adults: A national profile* . American Sociological Review , 73 , 185 – 203 . doi: 10.1177/000312240807300201 Google Scholar CrossRef Search ADS PubMed Ducharme , F. , Lévesque , L. , Lachance , L. , Kergoat , M. J. , & Coulombe , R . ( 2011 ). Challenges associated with transition to caregiver role following diagnostic disclosure of Alzheimer disease: A descriptive study . International Journal of Nursing Studies , 48 , 1109 – 1119 . doi: 10.1016/j.ijnurstu.2011.02.011 Google Scholar CrossRef Search ADS PubMed English , T. , & Carstensen , L. L . ( 2014 ). Selective narrowing of social networks across adulthood is associated with improved emotional experience in daily life . International Journal of Behavioral Development , 38 , 195 – 202 . doi: 10.1177/0165025413515404 Google Scholar CrossRef Search ADS PubMed Fiori , K. L. , Antonucci , T. C. , & Cortina , K. S . ( 2006 ). Social network typologies and mental health among older adults . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 61 , P25 – P32 . doi: 10.1093/geronb/61.1.P25 Google Scholar CrossRef Search ADS Hardiman , K. M . ( 2017 ). Social networks, depression, and quality of life among women religious in a residential facility . Doctoral dissertation. Marywood University . Koehly , L. M. , Ashida , S. , Schafer , E. J. , & Ludden , A . ( 2015 ). Caregiving networks-using a network approach to identify missed opportunities . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 70 , 143 – 154 . doi: 10.1093/geronb/gbu111 Google Scholar CrossRef Search ADS Lazer , D. , Pentland , A. S. , Adamic , L. , Aral , S. , Barabasi , A. L. , Brewer , D. , … Gutmann , M . ( 2009 ). Life in the network: the coming age of computational social science . Science (New York, NY) , 323 , 721 – 723 . doi: 10.1126/science.1167742 Google Scholar CrossRef Search ADS Litwin , H . ( 1995 ). The social networks of elderly immigrants: An analytic typology . Journal of Aging Studies , 9 , 155 – 174 . doi: 10.1016/0890-4065(95)90009-8 Google Scholar CrossRef Search ADS Litwin , H. , & Shiovitz-Ezra , S . ( 2006 ). Network type and mortality risk in later life . The Gerontologist , 46 , 735 – 743 . doi: 10.1093/geront/46.6.735 Google Scholar CrossRef Search ADS PubMed Litwin , H. , & Shiovitz-Ezra , S . ( 2011 ). Social network type and subjective well-being in a national sample of older Americans . The Gerontologist , 51 , 379 – 388 . doi: 10.1093/geront/gnq094 Google Scholar CrossRef Search ADS PubMed Márquez-Serrano , M. , González-Juárez , X. , Castillo-Castillo , L. E. , González-González , L. , & Idrovo , A. J . ( 2012 ). Social network analysis to evaluate nursing interventions to improve self-care . Public Health Nursing , 29 , 361 – 369 . doi: 10.1111/j.1525-1446.2012.01014.x Google Scholar CrossRef Search ADS PubMed Marsden , P. V . ( 2002 ). Egocentric and sociocentric measures of network centrality . Social Networks , 24 , 407 – 422 . doi: 10.1016/S0378-8733(02)00016-3 Google Scholar CrossRef Search ADS Meeuwesen , L. , Hortulanus , R. , & Machielse , A . ( 2001 ). Social contacts and social isolation: A typology . Netherlands’ Journal of Social Sciences , 37 , 132 – 154 . Moher , D. , Liberati , A. , Tetzlaff , J. , & Altman , D. G .; PRISMA Group . ( 2009 ). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement . Annals of Internal Medicine , 151 , 264 – 9, W64 . doi: 10.7326/0003-4819-151-4-200908180-00135 Google Scholar CrossRef Search ADS PubMed Nguyen , A. W . ( 2017 ). Variations in social network type membership among older African Americans, Caribbean Blacks, and Non-Hispanic Whites . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 72 , 716 – 726 . doi: 10.1093/geronb/gbx016 Google Scholar CrossRef Search ADS Nyqvist , F. , Forsman , A. K. , Giuntoli , G. , & Cattan , M . ( 2013 ). Social capital as a resource for mental well-being in older people: A systematic review . Aging and Mental Health , 17 , 394 – 410 . doi: 10.1080/13607863.2012.742490 Google Scholar CrossRef Search ADS PubMed Park , N. S. , Jang , Y. , Lee , B. S. , Ko , J. E. , Haley , W. E. , & Chiriboga , D. A . ( 2015 ). An empirical typology of social networks and its association with physical and mental health: A study with older Korean immigrants . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 70 , 67 – 76 . doi: 10.1093/geronb/gbt065 Google Scholar CrossRef Search ADS Perkins , J. M. , Subramanian , S. V. , & Christakis , N. A . ( 2015 ). Social networks and health: A systematic review of sociocentric network studies in low- and middle-income countries . Social Science and Medicine (1982) , 125 , 60 – 78 . doi: 10.1016/j.socscimed.2014.08.019 Google Scholar CrossRef Search ADS Rosenquist , J. N. , Fowler , J. H. , & Christakis , N. A . ( 2011 ). Social network determinants of depression . Molecular psychiatry , 16 , 273 – 281 . doi: 10.1038/mp.2010.13 Google Scholar CrossRef Search ADS PubMed Schafer , M. H . ( 2011 ). Health and network centrality in a continuing care retirement community . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 66 , 795 – 803 . doi: 10.1093/geronb/gbr112 Google Scholar CrossRef Search ADS Schafer , M. H . ( 2013 ). Structural advantages of good health in old age: Investigating the health-begets-position hypothesis with a full social network . Research on Aging , 35 , 348 – 370 . doi: 10.1177/0164027512441612 Google Scholar CrossRef Search ADS Schafer , M. H . ( 2015 ). On the locality of asymmetric close relations: Spatial proximity and health differences in a senior community . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 70 , 100 – 110 . doi: 10.1093/geronb/gbu043 Google Scholar CrossRef Search ADS Schafer , M. H . ( 2016 ). Health as status? Network relations and social structure in an American retirement community . Ageing and Society , 36 , 79 – 105 . doi: 10.1017/S0144686X14000993 Google Scholar CrossRef Search ADS Seale , C . ( 2004 ). Researching society and culture . London, UK : Sage . Smith , G. L. , Banting , L. , Eime , R. , O’Sullivan , G. , & van Uffelen , J. G . ( 2017 ). The association between social support and physical activity in older adults: a systematic review . International Journal of Behavioral Nutrition and Physical Activity , 14 , 56 . doi: 10.1186/s12966-017-0509-8 Google Scholar CrossRef Search ADS PubMed Sohn , S. Y. , Joo , W. T. , Kim , W. J. , Kim , S. J. , Youm , Y. , Kim , H. C. , … Lee , E . ( 2017 ). Social network types among older Korean adults: Associations with subjective health . Social Science and Medicine (1982) , 173 , 88 – 95 . doi: 10.1016/j.socscimed.2016.11.042 Google Scholar CrossRef Search ADS Steinbach , U . ( 1992 ). Social networks, institutionalization, and mortality among elderly people in the United States . Journal of Gerontology , 47 , S183 – S190 . Google Scholar CrossRef Search ADS PubMed Valente , T. W . ( 2012 ). Network interventions . Science (New York, N.Y.) , 337 , 49 – 53 . doi: 10.1126/science.1217330 Google Scholar CrossRef Search ADS PubMed Valente , T. W. , & Davis , R. L . ( 1999 ). Accelerating the diffusion of innovations using opinion leaders . The Annals of the American Academy of Political and Social Science , 566 , 55 – 67 . doi: 10.1177/000271629956600105 Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Gerontologist Oxford University Press

A Systematic Review of Research on Social Networks of Older Adults

Loading next page...
 
/lp/ou_press/a-systematic-review-of-research-on-social-networks-of-older-adults-xi7sbPqDHe
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0016-9013
eISSN
1758-5341
D.O.I.
10.1093/geront/gnx218
Publisher site
See Article on Publisher Site

Abstract

Abstract Background and Objectives There has been a substantial interest in life course/life span changes in older adults’ social networks and in the relationship between social networks and health and wellbeing. The study embarked on a systematic review to examine the existing knowledgebase on social network in the field of gerontology. Our focus was on studies in which both ego (respondents) and his or her alters (network members) are queried about their social ties. Research Design and Methods We searched for studies published in English before September, 2017, relied on quantitative methods to obtain data from both ego (60 years of age and older) and alters and provided a quantitative account of the social network properties. We searched the following data sets: APA Psychnet, Pubmed, Sociological abstracts, and Ageline. This was followed by a snowball search of relevant articles using Google Scholar. Titles and abstracts were reviewed and selected articles were extracted independently by two reviewers. Results A total of 5,519 records were retrieved. Of these, 3,994 records remained after the removal of duplicates. Ten records reporting on five original samples were kept for the systematic review. One study described a social network of community dwelling older adults and the remaining studies described social networks of institutional older adults. Discussion and Implications The present study points to a lacuna in current understanding of social networks in the field of gerontology. It provides a useful review and possible tools for the design of future studies to address current shortcomings in the field. Systematic review, Sociocentric, Egocentric, Social network The term social network conveys the notion that individuals are embedded within a larger context of relational ties (Borgatti, Mehra, Brass, & Labianca, 2009). In the past few decades, there has been a tremendous amount of research on the role of social networks in the life of older adults (Antonucci & Akiyama, 1987; Sohn et al., 2017; Steinbach, 1992). Research has tended to classify older adults’ networks based on the type, quality, and/or quantity of the relationships (Litwin, 1995; Meeuwesen, Hortulanus, & Machielse, 2001; Nguyen, 2017; Park et al., 2015). For instance, research conducted in Israel has found that networks that consisted of diverse and friends-focused ties fared better in terms of health indicators, whereas community-clan networks were associated with less favorable outcomes (Litwin & Shiovitz-Ezra, 2006). Research conducted in other countries largely supported the relationship between network type and health outcomes, even though the proposed typology was somewhat different (Fiori, Antonucci, & Cortina, 2006; Litwin & Shiovitz-Ezra, 2011). For instance, in the U.S.-based sample (Litwin & Shiovitz-Ezra, 2011), the authors identified five types of social networks, whereas in the Israeli sample (Litwin & Shiovitz-Ezra, 2006), the authors identified six types of networks. Common to both studies was the identification of networks, which were based on friends, family, diverse members, and restricted networks. However, a different U.S.-based sample suggested two restricted networks, rather than one (Fiori, Antonucci, & Cortina, 2006). This line of research has shown that the type of social network one has impacts his or her longevity, mortality, quality of life, and health behaviors. Hence, social networks are thought to play an important role in the life of older adults. A prominent theory in the field is the convoy model of social relations which suggests that both life course and life span influences impact one’s social network (Borgatti et al., 2009). The context and the developmental life stage influence the formation of social relations, which can be characterized across several dimensions, including structure, function, and quality. Social relations are viewed as being multifaceted, constructed of both objective (e.g., number of ties), and subjective (quality of ties) characteristics (Antonucci, Ajrouch, & Birditt, 2014; Antonucci & Akiyama, 1987; Antonucci, Fiori, Birditt, & Jackey, 2010). This model has attracted attention by researchers who have shown that overall, there is a tendency for older adults’ social networks to shrink in old age (Cornwell, Laumann, & Schumm, 2008) and to consist of fewer peripheral network members (English & Carstensen, 2014). The nature of the relationship also changes over time. Research has shown that as older adults’ physical ability declines, members in their social network tend to provide them with higher levels of instrumental and personal support (Ducharme, Lévesque, Lachance, Kergoat, & Coulombe, 2011). In recognition of the importance of older adults’ social networks, several large-scale epidemiological studies have collected data on the topic (e.g., the National Social Life Health and Aging Project; NSHAPE http://www.norc.org/Research/Projects/Pages/national-social-life-health-and-aging-project.aspx; the Survey of Health, Ageing and Retirement in Europe, SHARE http://www.share-project.org/). Although informative, these studies have been conducted from the point of view of the ego, a focal person who provides information about his or her network. There has been very limited research to focus on the entire social network and to incorporate the point of view of various network members related to the ego, also known as alters. This is important because we know that social networks are relational in nature and involve more than a single individual (Seale, 2004). We also know that the individual’s perspective on his or her network does not fully correspond with alters’ perspectives on the network (Marsden, 2002). Hence, certain properties of the network can only be inferred by interviewing all network members. Moreover, studies that have shown that depression or loneliness are “contagious” within the social network (Cacioppo, Fowler, & Christakis, 2009; Rosenquist, Fowler, & Christakis, 2011) could only be conducted if both ego and alters are interviewed. Finally, interventions that target the structure of social ties or the contagion of certain behaviors or beliefs also are likely to benefit from data on the entire social network (Valente, 2012). The present study embarked on a systematic review of the literature to examine the existing knowledgebase on social network in the field of gerontology. We specifically focused only on networks obtained from the point of view of both egos and alters rather than networks that are limited to the ego perspective, given the unique potential for additional information inherent in the former type of networks. In order to facilitate research in the field, we provide detailed information about current methods and findings as well as recommendations for future research. Methods Obtaining the Data We searched the following databases: APA Psychnet, Pubmed, Sociological abstracts, and Ageline. The search was conducted in September, 2017. The following key words were used to guide the search: (“older adult*” OR elder* OR aged OR “nursing home*” OR senior* OR senescent OR “assisted living” OR “long term care” OR “nursing unit” OR “skilled nursing facility*” OR geriatric* OR “residential aged care” OR “adult day care*” OR “continuing care retirement community* OR “lifelong center*”) AND (“social network”). The latter search term was consistent with the term used in a recent review of social network research in a different population (Perkins, Subramanian, & Christakis, 2015). The bibliographies of all relevant review articles were searched. A Google Scholar search using the function “cited by” and “related articles” was used with all articles included in the present review in order to trace additional relevant articles. Selection of Studies All titles and available abstracts were reviewed for relevance by two independent researchers (LA, IL). Disagreements were resolved through a consensus. The following inclusion criteria were employed: (a) articles published on or before September 11, 2017 (when data search was concluded); (b) written in English; (c) relied on quantitative methods to obtain data from both ego and alters; (d) provided a description of data collection methods; (e) provided numeric information on network properties as a whole or at the ego level, and (f) all egos were 60 years old or over. Exclusion criteria: studies in which (a) only two or less alters were queried, as this provides a very limited picture of the whole network; (b) staff or other observers provided network data, rather than self-report of ego and alters; and (c) studies that were not available for a full review, such as conference proceedings. Different studies conducted on the same sample were described in detail but counted only once. Data Extraction Data extraction was conducted independently by the two reviewers. Disagreements between reviewers were discussed and a consensus agreement was established. Extraction details are available in Tables 1 and 2. Table 1. Study Characteristics Author Country Setting (community, adult day center, CCRC, etc.) Scope/definition of the network Number of participants Sample characteristics (age/gender) Abbott et al., 2012 USA One neighborhood in a residential long term care-assisted living facility All assisted living residents and staff 25/65 staff agreed to be photographed, 10/15 residents participated 86 y [82–92 y], 5 male, MMSE = 25.8[16–30] Abbott et al., 2015 USA An assisted living and dementia care unit in a nursing home Assisted living residents and staff, dementia care unit 10/15 residents of an assisted living-Same participants as in Abbott et al., (2012), 10/12 residents of a dementia unit Assisted living-Same participants as in Abbott et al., (2012); Dementia unit- 87 y [82–96 y], 80% female, MMSE = 17.2(14–25) Abbott & Pachucki, 2017 USA A dementia special care unit Dementia special care unit over three consecutive years 10 in wave 1, 10 in wave 2, 17 in wave 3, but only 3 people present across all three waves 90 y, 67–101 y, 80–90% women, dementia, MMSE = 16.9–19.8 Casey et al., 2016 Australia A nursing home: 3 care units, including a dementia unit Nursing home residents 36/94-only 29 reported on relationships 63–94 y, 61.1% female, 67% dementia Hardiman, 2017 USA A residential facility All residents (also asked about relationship with staff, nurses, caregivers in residential facility) 24/76 residents; 16/25 staff, nurses and caregivers provided pictures 100% female, 85 y[72–102 y], 16.6% mild dementia, MMSE = 20–24 Márquez-Serrano et al., 2012 Mexico A neighborhood in Mexico >60 y neighborhood residents who participated in the educational intervention 10/18 65–85 y, 70% female Schafer, 2011 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2013 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2015 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–-96 y], 72% female Schafer, 2016 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 72% female Author Country Setting (community, adult day center, CCRC, etc.) Scope/definition of the network Number of participants Sample characteristics (age/gender) Abbott et al., 2012 USA One neighborhood in a residential long term care-assisted living facility All assisted living residents and staff 25/65 staff agreed to be photographed, 10/15 residents participated 86 y [82–92 y], 5 male, MMSE = 25.8[16–30] Abbott et al., 2015 USA An assisted living and dementia care unit in a nursing home Assisted living residents and staff, dementia care unit 10/15 residents of an assisted living-Same participants as in Abbott et al., (2012), 10/12 residents of a dementia unit Assisted living-Same participants as in Abbott et al., (2012); Dementia unit- 87 y [82–96 y], 80% female, MMSE = 17.2(14–25) Abbott & Pachucki, 2017 USA A dementia special care unit Dementia special care unit over three consecutive years 10 in wave 1, 10 in wave 2, 17 in wave 3, but only 3 people present across all three waves 90 y, 67–101 y, 80–90% women, dementia, MMSE = 16.9–19.8 Casey et al., 2016 Australia A nursing home: 3 care units, including a dementia unit Nursing home residents 36/94-only 29 reported on relationships 63–94 y, 61.1% female, 67% dementia Hardiman, 2017 USA A residential facility All residents (also asked about relationship with staff, nurses, caregivers in residential facility) 24/76 residents; 16/25 staff, nurses and caregivers provided pictures 100% female, 85 y[72–102 y], 16.6% mild dementia, MMSE = 20–24 Márquez-Serrano et al., 2012 Mexico A neighborhood in Mexico >60 y neighborhood residents who participated in the educational intervention 10/18 65–85 y, 70% female Schafer, 2011 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2013 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2015 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–-96 y], 72% female Schafer, 2016 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 72% female Note. MMSE = Mini Mental Status Exam; y = years. View Large Table 1. Study Characteristics Author Country Setting (community, adult day center, CCRC, etc.) Scope/definition of the network Number of participants Sample characteristics (age/gender) Abbott et al., 2012 USA One neighborhood in a residential long term care-assisted living facility All assisted living residents and staff 25/65 staff agreed to be photographed, 10/15 residents participated 86 y [82–92 y], 5 male, MMSE = 25.8[16–30] Abbott et al., 2015 USA An assisted living and dementia care unit in a nursing home Assisted living residents and staff, dementia care unit 10/15 residents of an assisted living-Same participants as in Abbott et al., (2012), 10/12 residents of a dementia unit Assisted living-Same participants as in Abbott et al., (2012); Dementia unit- 87 y [82–96 y], 80% female, MMSE = 17.2(14–25) Abbott & Pachucki, 2017 USA A dementia special care unit Dementia special care unit over three consecutive years 10 in wave 1, 10 in wave 2, 17 in wave 3, but only 3 people present across all three waves 90 y, 67–101 y, 80–90% women, dementia, MMSE = 16.9–19.8 Casey et al., 2016 Australia A nursing home: 3 care units, including a dementia unit Nursing home residents 36/94-only 29 reported on relationships 63–94 y, 61.1% female, 67% dementia Hardiman, 2017 USA A residential facility All residents (also asked about relationship with staff, nurses, caregivers in residential facility) 24/76 residents; 16/25 staff, nurses and caregivers provided pictures 100% female, 85 y[72–102 y], 16.6% mild dementia, MMSE = 20–24 Márquez-Serrano et al., 2012 Mexico A neighborhood in Mexico >60 y neighborhood residents who participated in the educational intervention 10/18 65–85 y, 70% female Schafer, 2011 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2013 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2015 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–-96 y], 72% female Schafer, 2016 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 72% female Author Country Setting (community, adult day center, CCRC, etc.) Scope/definition of the network Number of participants Sample characteristics (age/gender) Abbott et al., 2012 USA One neighborhood in a residential long term care-assisted living facility All assisted living residents and staff 25/65 staff agreed to be photographed, 10/15 residents participated 86 y [82–92 y], 5 male, MMSE = 25.8[16–30] Abbott et al., 2015 USA An assisted living and dementia care unit in a nursing home Assisted living residents and staff, dementia care unit 10/15 residents of an assisted living-Same participants as in Abbott et al., (2012), 10/12 residents of a dementia unit Assisted living-Same participants as in Abbott et al., (2012); Dementia unit- 87 y [82–96 y], 80% female, MMSE = 17.2(14–25) Abbott & Pachucki, 2017 USA A dementia special care unit Dementia special care unit over three consecutive years 10 in wave 1, 10 in wave 2, 17 in wave 3, but only 3 people present across all three waves 90 y, 67–101 y, 80–90% women, dementia, MMSE = 16.9–19.8 Casey et al., 2016 Australia A nursing home: 3 care units, including a dementia unit Nursing home residents 36/94-only 29 reported on relationships 63–94 y, 61.1% female, 67% dementia Hardiman, 2017 USA A residential facility All residents (also asked about relationship with staff, nurses, caregivers in residential facility) 24/76 residents; 16/25 staff, nurses and caregivers provided pictures 100% female, 85 y[72–102 y], 16.6% mild dementia, MMSE = 20–24 Márquez-Serrano et al., 2012 Mexico A neighborhood in Mexico >60 y neighborhood residents who participated in the educational intervention 10/18 65–85 y, 70% female Schafer, 2011 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2013 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 69% female Schafer, 2015 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–-96 y], 72% female Schafer, 2016 USA A continuing care retirement community All residents of a continuing care retirement community- independent living only 123/158, 91% participation rate 86 y [74–96 y], 72% female Note. MMSE = Mini Mental Status Exam; y = years. View Large Table 2. Main Findings and Study Design Author Study design (cross-sectional/ longitudinal, etc.) Network types measured and analyzed (friendship/ familiarity etc.) Exact questions used to derive the network Method used to obtain the network (name generator, census report) Statistical method Basic network properties reported Map provided Main findings Abbott et al., 2012 Cross-sectional Social interactions in the past week The AL residents were asked the same 3 questions about people in their own and other neighborhoods within the same AL facility and about staff who worked at the AL facility: “5 people you spent most time with in the past week, 5 people who provided help to you in the past week, 5 people you provided help to in the past week.” Photos were used only when asked about residents in own community. Sociocentric with pictures of all AL residents and staff Descriptive Degree centrality, in-degree, out-degree Yes Descriptive analysis of AL- indegree, outdegree overall degree centrality ranges between 1 and 23 Abbott et al., 2015 Cross-sectional Social interactions in the past week/recently Sociocentric as in Abbott et al.,(2012). In the dementia unit, “recently spent time” with rather than “over the past week” was used for reference (Egocentric-not relevant for this review) and sociocentric with photos Descriptive Sociogram, network size, degree centrality, reciprocity Yes Sociocentric vs. egocentric approach nominations. Mean sociocentric network size 7[0–14]. Residents with a higher MMSE score nominated more individuals. Reciprocity in AL 55% [20%–75%] Abbott & Pachucki, 2017 Cross-sectional—three waves of data but only 3 residents went through all waves of data collection Social interactions recently Photos were provided. 5 questions asked: who they spent time with, listened to problems, helped with something, who helped them, who listened to them. Sociocentric with photos Associations of quality of life and cognition with network characteristics Personal network size, in-degree, out-degree, betweeness- centrality, closeness- centrality, network density Yes Average personal network size 2–4.5. Half the ties were reciprocated and there was a positive association between integration- betweeness centrality and quality of life, but inconsistent associations with cognitive functioning. Friendship ties were more frequent among people of adjacent cognitive status categories Casey et al., 2016 Cross-sectional Social support, friendship Showing photos and asking to identify friends; followed-up by questions about true friend vs. casual friend; (integrated with qualitative and observational data) Sociocentric with photos Correlations between ties and assessment scores Out-degree, in-degree, reciprocity, density, path length Yes When all potential relationship were included in the unit, the median size was 0 [0–1]. Low density, low reciprocity (22.2%), high levels of isolation; cognitive impairment was negatively correlated with network size and reciprocity, friendship correlated with reciprocity Hardiman, 2017 Cross-sectional Spend time, provide help, receive help 1. Pick up to or point out five people with whom you spent the most time in the past two weeks. 2. Pick up to or point out five people who provided help to you in the past week. 3. Pick up to or point out five people you provided help to in the past week. *Sociocentric questions are to be asked about residents of the NH and with regards to staff, nurses, or caregivers. A list of names and photos of those who agreed to have their picture taken- both staff, nurses, caregivers and residents Correlational Total residents and staff nominated No Staff had a higher in-degree centrality. A correlation between number of connections with others, quality of life and MMSE score Márquez-Serrano et al., 2012 Pre-post: but networks measured only once Relationship types Indicate the individuals with whom they have relationships: family, friends, neighbors or work colleagues Egocentric Network characteristics in relation to flu infection Structure, density, degree Yes Low density: .0167. Out-degree: 1.57. No knowledge transfer within the network Schafer, 2016 Cross-sectional Spend time, confidant Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. List the people they talked to about important matters. A map of the continuing care retirement community Regression analysis in-degree, out- degree, Bonacich centrality Yes Time spent (mean 20.09) had more ties than confidant relations (mean 2.24). Less asymmetry in confidant relations (.2) than in time spent (.33). Health as a predictor of Bonacich centrality, in-degree and out-degree. Significant results for some but not all types of ties. Schafer, 2013 Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Regression analysis Constraint, integration No Health predicted structural position. Residents with the best health had positional advantage in the network. Schafer, 2015 Cross-sectional Confidant “From time to time, people will often talk with others about things that are important to them. This could include sharing good news, or bad news, sharing about concerns they might have, or otherwise just talking about things they find very important or significant. Who are the people here at [RC] that you can talk to about things that are important to you?” A map of the continuing care retirement community Exponential random graph Close ties sent and received, density Yes An overall density score of .02. Close relations are influenced by physical proximity. Physical proximity intensified health based a-symmetry. (Schafer. 2016) Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Exponential random graph Total ties, density, geodesic distances, reciprocated ties, centrality measures No Healthier individuals received more social tie nominations. Only modest support for health-based homophily. Author Study design (cross-sectional/ longitudinal, etc.) Network types measured and analyzed (friendship/ familiarity etc.) Exact questions used to derive the network Method used to obtain the network (name generator, census report) Statistical method Basic network properties reported Map provided Main findings Abbott et al., 2012 Cross-sectional Social interactions in the past week The AL residents were asked the same 3 questions about people in their own and other neighborhoods within the same AL facility and about staff who worked at the AL facility: “5 people you spent most time with in the past week, 5 people who provided help to you in the past week, 5 people you provided help to in the past week.” Photos were used only when asked about residents in own community. Sociocentric with pictures of all AL residents and staff Descriptive Degree centrality, in-degree, out-degree Yes Descriptive analysis of AL- indegree, outdegree overall degree centrality ranges between 1 and 23 Abbott et al., 2015 Cross-sectional Social interactions in the past week/recently Sociocentric as in Abbott et al.,(2012). In the dementia unit, “recently spent time” with rather than “over the past week” was used for reference (Egocentric-not relevant for this review) and sociocentric with photos Descriptive Sociogram, network size, degree centrality, reciprocity Yes Sociocentric vs. egocentric approach nominations. Mean sociocentric network size 7[0–14]. Residents with a higher MMSE score nominated more individuals. Reciprocity in AL 55% [20%–75%] Abbott & Pachucki, 2017 Cross-sectional—three waves of data but only 3 residents went through all waves of data collection Social interactions recently Photos were provided. 5 questions asked: who they spent time with, listened to problems, helped with something, who helped them, who listened to them. Sociocentric with photos Associations of quality of life and cognition with network characteristics Personal network size, in-degree, out-degree, betweeness- centrality, closeness- centrality, network density Yes Average personal network size 2–4.5. Half the ties were reciprocated and there was a positive association between integration- betweeness centrality and quality of life, but inconsistent associations with cognitive functioning. Friendship ties were more frequent among people of adjacent cognitive status categories Casey et al., 2016 Cross-sectional Social support, friendship Showing photos and asking to identify friends; followed-up by questions about true friend vs. casual friend; (integrated with qualitative and observational data) Sociocentric with photos Correlations between ties and assessment scores Out-degree, in-degree, reciprocity, density, path length Yes When all potential relationship were included in the unit, the median size was 0 [0–1]. Low density, low reciprocity (22.2%), high levels of isolation; cognitive impairment was negatively correlated with network size and reciprocity, friendship correlated with reciprocity Hardiman, 2017 Cross-sectional Spend time, provide help, receive help 1. Pick up to or point out five people with whom you spent the most time in the past two weeks. 2. Pick up to or point out five people who provided help to you in the past week. 3. Pick up to or point out five people you provided help to in the past week. *Sociocentric questions are to be asked about residents of the NH and with regards to staff, nurses, or caregivers. A list of names and photos of those who agreed to have their picture taken- both staff, nurses, caregivers and residents Correlational Total residents and staff nominated No Staff had a higher in-degree centrality. A correlation between number of connections with others, quality of life and MMSE score Márquez-Serrano et al., 2012 Pre-post: but networks measured only once Relationship types Indicate the individuals with whom they have relationships: family, friends, neighbors or work colleagues Egocentric Network characteristics in relation to flu infection Structure, density, degree Yes Low density: .0167. Out-degree: 1.57. No knowledge transfer within the network Schafer, 2016 Cross-sectional Spend time, confidant Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. List the people they talked to about important matters. A map of the continuing care retirement community Regression analysis in-degree, out- degree, Bonacich centrality Yes Time spent (mean 20.09) had more ties than confidant relations (mean 2.24). Less asymmetry in confidant relations (.2) than in time spent (.33). Health as a predictor of Bonacich centrality, in-degree and out-degree. Significant results for some but not all types of ties. Schafer, 2013 Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Regression analysis Constraint, integration No Health predicted structural position. Residents with the best health had positional advantage in the network. Schafer, 2015 Cross-sectional Confidant “From time to time, people will often talk with others about things that are important to them. This could include sharing good news, or bad news, sharing about concerns they might have, or otherwise just talking about things they find very important or significant. Who are the people here at [RC] that you can talk to about things that are important to you?” A map of the continuing care retirement community Exponential random graph Close ties sent and received, density Yes An overall density score of .02. Close relations are influenced by physical proximity. Physical proximity intensified health based a-symmetry. (Schafer. 2016) Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Exponential random graph Total ties, density, geodesic distances, reciprocated ties, centrality measures No Healthier individuals received more social tie nominations. Only modest support for health-based homophily. Note. AL = assisted living; betweeness centrality = number of shortest paths from all nodes (alters) that path through the ego; Bonacich centrality = takes into account the number of connections within the ego network; degree centrality = the number of ties one has; density = actual ties/all possible ties; in-degree = number of incoming ties; MMSE = Mini Mental Status Exam; NH = nursing home; out-degree = number of outgoing ties; reciprocity = the likelihood of a tie to be mutually linked; RC = retirement community. View Large Table 2. Main Findings and Study Design Author Study design (cross-sectional/ longitudinal, etc.) Network types measured and analyzed (friendship/ familiarity etc.) Exact questions used to derive the network Method used to obtain the network (name generator, census report) Statistical method Basic network properties reported Map provided Main findings Abbott et al., 2012 Cross-sectional Social interactions in the past week The AL residents were asked the same 3 questions about people in their own and other neighborhoods within the same AL facility and about staff who worked at the AL facility: “5 people you spent most time with in the past week, 5 people who provided help to you in the past week, 5 people you provided help to in the past week.” Photos were used only when asked about residents in own community. Sociocentric with pictures of all AL residents and staff Descriptive Degree centrality, in-degree, out-degree Yes Descriptive analysis of AL- indegree, outdegree overall degree centrality ranges between 1 and 23 Abbott et al., 2015 Cross-sectional Social interactions in the past week/recently Sociocentric as in Abbott et al.,(2012). In the dementia unit, “recently spent time” with rather than “over the past week” was used for reference (Egocentric-not relevant for this review) and sociocentric with photos Descriptive Sociogram, network size, degree centrality, reciprocity Yes Sociocentric vs. egocentric approach nominations. Mean sociocentric network size 7[0–14]. Residents with a higher MMSE score nominated more individuals. Reciprocity in AL 55% [20%–75%] Abbott & Pachucki, 2017 Cross-sectional—three waves of data but only 3 residents went through all waves of data collection Social interactions recently Photos were provided. 5 questions asked: who they spent time with, listened to problems, helped with something, who helped them, who listened to them. Sociocentric with photos Associations of quality of life and cognition with network characteristics Personal network size, in-degree, out-degree, betweeness- centrality, closeness- centrality, network density Yes Average personal network size 2–4.5. Half the ties were reciprocated and there was a positive association between integration- betweeness centrality and quality of life, but inconsistent associations with cognitive functioning. Friendship ties were more frequent among people of adjacent cognitive status categories Casey et al., 2016 Cross-sectional Social support, friendship Showing photos and asking to identify friends; followed-up by questions about true friend vs. casual friend; (integrated with qualitative and observational data) Sociocentric with photos Correlations between ties and assessment scores Out-degree, in-degree, reciprocity, density, path length Yes When all potential relationship were included in the unit, the median size was 0 [0–1]. Low density, low reciprocity (22.2%), high levels of isolation; cognitive impairment was negatively correlated with network size and reciprocity, friendship correlated with reciprocity Hardiman, 2017 Cross-sectional Spend time, provide help, receive help 1. Pick up to or point out five people with whom you spent the most time in the past two weeks. 2. Pick up to or point out five people who provided help to you in the past week. 3. Pick up to or point out five people you provided help to in the past week. *Sociocentric questions are to be asked about residents of the NH and with regards to staff, nurses, or caregivers. A list of names and photos of those who agreed to have their picture taken- both staff, nurses, caregivers and residents Correlational Total residents and staff nominated No Staff had a higher in-degree centrality. A correlation between number of connections with others, quality of life and MMSE score Márquez-Serrano et al., 2012 Pre-post: but networks measured only once Relationship types Indicate the individuals with whom they have relationships: family, friends, neighbors or work colleagues Egocentric Network characteristics in relation to flu infection Structure, density, degree Yes Low density: .0167. Out-degree: 1.57. No knowledge transfer within the network Schafer, 2016 Cross-sectional Spend time, confidant Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. List the people they talked to about important matters. A map of the continuing care retirement community Regression analysis in-degree, out- degree, Bonacich centrality Yes Time spent (mean 20.09) had more ties than confidant relations (mean 2.24). Less asymmetry in confidant relations (.2) than in time spent (.33). Health as a predictor of Bonacich centrality, in-degree and out-degree. Significant results for some but not all types of ties. Schafer, 2013 Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Regression analysis Constraint, integration No Health predicted structural position. Residents with the best health had positional advantage in the network. Schafer, 2015 Cross-sectional Confidant “From time to time, people will often talk with others about things that are important to them. This could include sharing good news, or bad news, sharing about concerns they might have, or otherwise just talking about things they find very important or significant. Who are the people here at [RC] that you can talk to about things that are important to you?” A map of the continuing care retirement community Exponential random graph Close ties sent and received, density Yes An overall density score of .02. Close relations are influenced by physical proximity. Physical proximity intensified health based a-symmetry. (Schafer. 2016) Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Exponential random graph Total ties, density, geodesic distances, reciprocated ties, centrality measures No Healthier individuals received more social tie nominations. Only modest support for health-based homophily. Author Study design (cross-sectional/ longitudinal, etc.) Network types measured and analyzed (friendship/ familiarity etc.) Exact questions used to derive the network Method used to obtain the network (name generator, census report) Statistical method Basic network properties reported Map provided Main findings Abbott et al., 2012 Cross-sectional Social interactions in the past week The AL residents were asked the same 3 questions about people in their own and other neighborhoods within the same AL facility and about staff who worked at the AL facility: “5 people you spent most time with in the past week, 5 people who provided help to you in the past week, 5 people you provided help to in the past week.” Photos were used only when asked about residents in own community. Sociocentric with pictures of all AL residents and staff Descriptive Degree centrality, in-degree, out-degree Yes Descriptive analysis of AL- indegree, outdegree overall degree centrality ranges between 1 and 23 Abbott et al., 2015 Cross-sectional Social interactions in the past week/recently Sociocentric as in Abbott et al.,(2012). In the dementia unit, “recently spent time” with rather than “over the past week” was used for reference (Egocentric-not relevant for this review) and sociocentric with photos Descriptive Sociogram, network size, degree centrality, reciprocity Yes Sociocentric vs. egocentric approach nominations. Mean sociocentric network size 7[0–14]. Residents with a higher MMSE score nominated more individuals. Reciprocity in AL 55% [20%–75%] Abbott & Pachucki, 2017 Cross-sectional—three waves of data but only 3 residents went through all waves of data collection Social interactions recently Photos were provided. 5 questions asked: who they spent time with, listened to problems, helped with something, who helped them, who listened to them. Sociocentric with photos Associations of quality of life and cognition with network characteristics Personal network size, in-degree, out-degree, betweeness- centrality, closeness- centrality, network density Yes Average personal network size 2–4.5. Half the ties were reciprocated and there was a positive association between integration- betweeness centrality and quality of life, but inconsistent associations with cognitive functioning. Friendship ties were more frequent among people of adjacent cognitive status categories Casey et al., 2016 Cross-sectional Social support, friendship Showing photos and asking to identify friends; followed-up by questions about true friend vs. casual friend; (integrated with qualitative and observational data) Sociocentric with photos Correlations between ties and assessment scores Out-degree, in-degree, reciprocity, density, path length Yes When all potential relationship were included in the unit, the median size was 0 [0–1]. Low density, low reciprocity (22.2%), high levels of isolation; cognitive impairment was negatively correlated with network size and reciprocity, friendship correlated with reciprocity Hardiman, 2017 Cross-sectional Spend time, provide help, receive help 1. Pick up to or point out five people with whom you spent the most time in the past two weeks. 2. Pick up to or point out five people who provided help to you in the past week. 3. Pick up to or point out five people you provided help to in the past week. *Sociocentric questions are to be asked about residents of the NH and with regards to staff, nurses, or caregivers. A list of names and photos of those who agreed to have their picture taken- both staff, nurses, caregivers and residents Correlational Total residents and staff nominated No Staff had a higher in-degree centrality. A correlation between number of connections with others, quality of life and MMSE score Márquez-Serrano et al., 2012 Pre-post: but networks measured only once Relationship types Indicate the individuals with whom they have relationships: family, friends, neighbors or work colleagues Egocentric Network characteristics in relation to flu infection Structure, density, degree Yes Low density: .0167. Out-degree: 1.57. No knowledge transfer within the network Schafer, 2016 Cross-sectional Spend time, confidant Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. List the people they talked to about important matters. A map of the continuing care retirement community Regression analysis in-degree, out- degree, Bonacich centrality Yes Time spent (mean 20.09) had more ties than confidant relations (mean 2.24). Less asymmetry in confidant relations (.2) than in time spent (.33). Health as a predictor of Bonacich centrality, in-degree and out-degree. Significant results for some but not all types of ties. Schafer, 2013 Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Regression analysis Constraint, integration No Health predicted structural position. Residents with the best health had positional advantage in the network. Schafer, 2015 Cross-sectional Confidant “From time to time, people will often talk with others about things that are important to them. This could include sharing good news, or bad news, sharing about concerns they might have, or otherwise just talking about things they find very important or significant. Who are the people here at [RC] that you can talk to about things that are important to you?” A map of the continuing care retirement community Exponential random graph Close ties sent and received, density Yes An overall density score of .02. Close relations are influenced by physical proximity. Physical proximity intensified health based a-symmetry. (Schafer. 2016) Cross-sectional Spend time Given a map of the floor. Asked if spent time interacting or socializing with the occupant of the apartment: “spend time interacting or socializing with [NAME] in a given week, beyond just passing by or saying hello.” If the answer was an affirmative, participants were then asked to approximate how much time they spent socializing “in a typical, or average, week.” ≥30 min of interaction was coded as spend time with. A map of the continuing care retirement community Exponential random graph Total ties, density, geodesic distances, reciprocated ties, centrality measures No Healthier individuals received more social tie nominations. Only modest support for health-based homophily. Note. AL = assisted living; betweeness centrality = number of shortest paths from all nodes (alters) that path through the ego; Bonacich centrality = takes into account the number of connections within the ego network; degree centrality = the number of ties one has; density = actual ties/all possible ties; in-degree = number of incoming ties; MMSE = Mini Mental Status Exam; NH = nursing home; out-degree = number of outgoing ties; reciprocity = the likelihood of a tie to be mutually linked; RC = retirement community. View Large Results Figure 1 demonstrates the study flow chart according to PRISMA guidelines (Moher, Liberati, Tetzlaff, Altman, & PRISMA Group, 2009). The PRISMA diagram maps the flow of information through the various review stages. The diagram outlines the number of articles retrieved, number of articles kept for full review, number of articles excluded, reasons for exclusion and number of articles kept for analysis. A total of 5,519 records were retrieved into an endnote library from the four data sets searched. Of these, 3,994 records remained after the removal of duplicates and 60 records were maintained for a thorough review because they appeared to be potentially relevant based on their title and abstract. Four additional records were obtained through Google Scholar search of already identified articles. Please see Figure 1 for details concerning study flow and reasons for exclusion. Figure 1. View largeDownload slide Study flow. Figure 1. View largeDownload slide Study flow. Overall, 10 records reporting on five original samples were kept for the systematic review. One study addressed community dwelling older adults in Mexico (Márquez-Serrano, González-Juárez, Castillo-Castillo, González-González, & Idrovo, 2012). This study used an egocentric method which relied on a name generator to identify the network. The remaining studies addressed older adults in long-term care institutes in the United States and Australia. Three of the studies presented data on social networks among older adults with dementia or mild cognitive impairment (Abbott, Bettger, Hampton, & Kohler, 2012; Abbott, Bettger, Hampton, & Kohler, 2015; Abbott & Pachucki, 2017; Casey, Low, Jeon, & Brodaty, 2016). Under these circumstances, a common approach appeared to be the use of photographs to construct a complete social network. One study employed a pre-post design, but network measures were obtained only once (Márquez-Serrano et al., 2012). Another study had three waves of data collection, but because of the variability in the network over time, the analysis focused on each network separately rather than on changes in network dynamics (Abbott & Pachucki, 2017). All other studies employed a cross-sectional design. Studies provided descriptive data on the ego network and correlational data to examine the associations of various types of social relations indicators with quality of life, cognitive functioning, and other health measures (Abbott & Pachucki, 2017; Casey et al., 2016; Hardiman, 2017). Two records presented results based on an exponential random graph to examine relationship quality or network position and structure as outcomes (Schafer, 2015, 2016), and two other records used regression analyses to examine health as a predictor of social network properties (Schafer, 2011, 2013). Studies have alluded to a bi-directional relationship between health and well-being and various network properties calculated at the ego level. See Tables 1 and 2 for details. Discussion The present study is the first systematic review of whole social networks among older adults. The findings suggest that in contrast to the plethora of research that has examined social networks of older adults from the ego perspective (Nyqvist, Forsman, Giuntoli, & Cattan, 2013; Smith, Banting, Eime, O’Sullivan, & van Uffelen, 2017), a few studies have examined both ego and alters in the field of gerontology. This finding is disappointing given the growing interest in social networks for the understanding of health and wellbeing in the general population (Lazer et al., 2009). The findings suggest that the study of whole social networks among older adults is feasible, even in the case of older adults with mild-to-moderate dementia (Abbott et al., 2015; Hardiman, 2017). Moreover, these studies allude to a unique opportunity available to those interested in the study of whole social networks of older adults in institutions. This is because the boundaries of the network are already pre-defined. Hence, this could provide an incentive to conducting a sociocentric study in which all network members are queried. Studies conducted in the United States and Australia relied exclusively on this property to examine an entire unit or institute in order to produce whole social network data (Abbott et al., 2015; Casey et al., 2016; Hardiman, 2017; Schafer, 2011). In contrast, a study conducted in Mexico (Márquez-Serrano et al., 2012) relied on the close-knit nature of community dwellers who participated in an educational intervention to develop a model which capitalized on overlap in ego networks (Márquez-Serrano et al., 2012). A community-dwelling living arrangement, which allows for the use of ego networks to construct full social networks due to overlap in ego networks, is less common in urban places, which characterize the global North. A potentially interesting and useful tool for collecting whole network data can be the use of snowball sampling several degrees away from a focal person (Antonucci & Israel, 1986; Bear, 1990). Such an approach can potentially help researchers to understand the social network not only from the perspective of the older adult, but also from the perspective of his or her alters even in settings, where boundaries are not predefined and there is no expectation for a natural overlap of ego networks. Although our review identified several studies that relied on such a method, these studies were limited to one or two additional alters or examined a focal person who did not meet our age criteria (Carpentier & Ducharme, 2007; Koehly, Ashida, Schafer, & Ludden, 2015). Despite the complexity of this type of design, it is particularly valuable for use in populations that do not have a-priori set boundaries, such as urban, community dwelling older adults. Based on the studies reviewed, one can infer that the concept of friendship can be used to describe older adults’ networks (Abbott et al., 2012; Casey et al., 2016). In general, networks in institutions are characterized by low density (number of actual ties divided by the number of all possible ties) and reciprocity (e.g., if ego knows alter, alter also knows ego) and high levels of isolation (no incoming/outgoing ties) (Casey et al., 2016; Schafer, 2011). This appears to be the case also in the community (Márquez-Serrano et al., 2012). Although some of the studies found a correlation between health, quality of life, cognitive functioning, and network characteristics (Hardiman, 2017; Schafer, 2013, 2015), the relationship appears to be bidirectional and given the cross-sectional design of the studies, it is impossible to determine its exact direction. Unfortunately, the small and varied nature of the studies reviewed does not allow determining the size or direction of these potential effects. Implications To sum, despite a plethora of research on social networks in older adults, there is only a handful of studies on whole social networks in this population. Hence, many studies disregard the fact that social interactions are bidirectional, with all stakeholders contributing to the shared meaning of the encounter (Seale, 2004). Moreover, the limited research comparing ego networks to sociocentric networks has shown that the two are not always comparable and that valuable information can be obtained by querying both ego and alters, rather than relying only on the ego (Chung, Hossain, & Davis, 2005). Examining the structure and the function of the entire network is important for several reasons. First, this can provide additional insights and enrich existing theories in the field of aging which have been exclusively based on the ego-perspective. Second, such an approach can provide information about the network as a whole, which is not limited to the ego perspective. Third, there is a growing interest in social networks as an intervention tool (Valente, 2012). These interventions address the structure of the network. For instance, by identifying those individuals who are more prone to isolation, one can initiate interventions for re-integration into the network. Alternatively, interventions can instigate a change in the network through contagion. For instance, interventions that promote the adaption of specific health behaviors within the network (Valente & Davis, 1999). The present study points to a lacuna in current understanding of social networks in the field of gerontology. It also provides useful tools for the design of future studies to address current shortcomings in the field. However, as is always the case with systematic reviews, there is a chance that our search had failed to include papers that should have been included. Also, given the very small number of articles and their heterogeneity, we cannot make concrete predictions about the nature of older adults’ full social networks nor about their correlates. Finally, our inclusion/exclusion criteria might have resulted in the exclusion of important studies in the field. For instance, we excluded qualitative studies because of our interest in numeric network properties (e.g., density, degree centrality) and in self-report of social ties. It is important to note, though, that this is in line with the methodology of systematic reviews, which requires strict and coherent inclusion/exclusion criteria, as long as they follow a clear rationale. Funding The study was funded by the Israel Science Foundation 537/16. Conflict of Interest None reported. References Abbott , K. M. , Bettger , J. P. , Hampton , K. , & Kohler , H. P . ( 2012 ). Exploring the use of social network analysis to measure social integration among older adults in assisted living . Family and Community Health , 35 , 322 – 333 . doi: 10.1097/FCH.0b013e318266669f Google Scholar CrossRef Search ADS PubMed Abbott , K. M. , Bettger , J. P. , Hampton , K. N. , & Kohler , H. P . ( 2015 ). The feasibility of measuring social networks among older adults in assisted living and dementia special care units . Dementia (London, England) , 14 , 199 – 219 . doi: 10.1177/1471301213494524 Google Scholar PubMed Abbott , K. M. , & Pachucki , M. C . ( 2017 ). Associations between social network characteristics, cognitive function, and quality of life among residents in a dementia special care unit: A pilot study . Dementia (London, England) , 16 , 1004 – 1019 . doi: 10.1177/1471301216630907 Google Scholar PubMed Antonucci , T. C. , Ajrouch , K. J. , & Birditt , K. S . ( 2014 ). The convoy model: Explaining social relations from a multidisciplinary perspective . The Gerontologist , 54 , 82 – 92 . doi: 10.1093/geront/gnt118 Google Scholar CrossRef Search ADS PubMed Antonucci , T. C. , & Akiyama , H . ( 1987 ). Social networks in adult life and a preliminary examination of the convoy model . Journal of Gerontology , 42 , 519 – 527 . doi: 10.1093/geronj/42.5.519 Google Scholar CrossRef Search ADS PubMed Antonucci , T. C. , Fiori , K. L. , Birditt , K. , & Jackey , L. M. H . ( 2010 ). Convoys of Social Relations: Integrating Life-Span and Life-Course Perspectives . In The Handbook of Life-Span Development . New York, NY : John Wiley & Sons, Inc . Google Scholar CrossRef Search ADS Antonucci , T. C. , & Israel , B. A . ( 1986 ). Veridicality of social support: A comparison of principal and network members’ responses . Journal of Consulting and Clinical Psychology , 54 , 432 – 437 . doi: 10.1037/0022-006X.54.4.432 Google Scholar CrossRef Search ADS PubMed Bear , M . ( 1990 ). Social network characteristics and the duration of primary relationships after entry into long-term care . Journal of Gerontology , 45 , S156 – S162 . doi: 10.1093/geronj/45.4.S156 Google Scholar CrossRef Search ADS PubMed Borgatti , S. P. , Mehra , A. , Brass , D. J. , & Labianca , G . ( 2009 ). Network analysis in the social sciences . Science (New York, N.Y.) , 323 , 892 – 895 . doi: 10.1126/science.1165821 Google Scholar CrossRef Search ADS PubMed Cacioppo , J. T. , Fowler , J. H. , & Christakis , N. A . ( 2009 ). Alone in the crowd: the structure and spread of loneliness in a large social network . Journal of Personality and Social Psychology , 97 , 977 – 991 . doi: 10.1037/a0016076 Google Scholar CrossRef Search ADS PubMed Carpentier , N. , & Ducharme , F . ( 2007 ). Social network data validity: The example of the social network of caregivers of older persons with Alzheimer-type dementia . Canadian Journal on Aging = La revue Canadienne Du Vieillissement , 26 ( Suppl. 1 ), 103 – 115 . doi: 10.3138/cja.26.suppl_1.103 Google Scholar CrossRef Search ADS Casey , A. N. , Low , L. F. , Jeon , Y. H. , & Brodaty , H . ( 2016 ). Residents perceptions of friendship and positive social networks within a nursing home . The Gerontologist , 56 , 855 – 867 . doi: 10.1093/geront/gnv146 Google Scholar CrossRef Search ADS PubMed Chung , K. K. , Hossain , L. , & Davis , J . ( 2005 ). Exploring sociocentric and egocentric approaches for social network analysis . Paper presented at the Proceedings of the 2nd international conference on knowledge management in Asia Pacific , Victoria University of Wellington, New Zealand . Cornwell , B. , Laumann , E. O. , & Schumm , L. P . ( 2008 ). The social connectedness of older adults: A national profile* . American Sociological Review , 73 , 185 – 203 . doi: 10.1177/000312240807300201 Google Scholar CrossRef Search ADS PubMed Ducharme , F. , Lévesque , L. , Lachance , L. , Kergoat , M. J. , & Coulombe , R . ( 2011 ). Challenges associated with transition to caregiver role following diagnostic disclosure of Alzheimer disease: A descriptive study . International Journal of Nursing Studies , 48 , 1109 – 1119 . doi: 10.1016/j.ijnurstu.2011.02.011 Google Scholar CrossRef Search ADS PubMed English , T. , & Carstensen , L. L . ( 2014 ). Selective narrowing of social networks across adulthood is associated with improved emotional experience in daily life . International Journal of Behavioral Development , 38 , 195 – 202 . doi: 10.1177/0165025413515404 Google Scholar CrossRef Search ADS PubMed Fiori , K. L. , Antonucci , T. C. , & Cortina , K. S . ( 2006 ). Social network typologies and mental health among older adults . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 61 , P25 – P32 . doi: 10.1093/geronb/61.1.P25 Google Scholar CrossRef Search ADS Hardiman , K. M . ( 2017 ). Social networks, depression, and quality of life among women religious in a residential facility . Doctoral dissertation. Marywood University . Koehly , L. M. , Ashida , S. , Schafer , E. J. , & Ludden , A . ( 2015 ). Caregiving networks-using a network approach to identify missed opportunities . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 70 , 143 – 154 . doi: 10.1093/geronb/gbu111 Google Scholar CrossRef Search ADS Lazer , D. , Pentland , A. S. , Adamic , L. , Aral , S. , Barabasi , A. L. , Brewer , D. , … Gutmann , M . ( 2009 ). Life in the network: the coming age of computational social science . Science (New York, NY) , 323 , 721 – 723 . doi: 10.1126/science.1167742 Google Scholar CrossRef Search ADS Litwin , H . ( 1995 ). The social networks of elderly immigrants: An analytic typology . Journal of Aging Studies , 9 , 155 – 174 . doi: 10.1016/0890-4065(95)90009-8 Google Scholar CrossRef Search ADS Litwin , H. , & Shiovitz-Ezra , S . ( 2006 ). Network type and mortality risk in later life . The Gerontologist , 46 , 735 – 743 . doi: 10.1093/geront/46.6.735 Google Scholar CrossRef Search ADS PubMed Litwin , H. , & Shiovitz-Ezra , S . ( 2011 ). Social network type and subjective well-being in a national sample of older Americans . The Gerontologist , 51 , 379 – 388 . doi: 10.1093/geront/gnq094 Google Scholar CrossRef Search ADS PubMed Márquez-Serrano , M. , González-Juárez , X. , Castillo-Castillo , L. E. , González-González , L. , & Idrovo , A. J . ( 2012 ). Social network analysis to evaluate nursing interventions to improve self-care . Public Health Nursing , 29 , 361 – 369 . doi: 10.1111/j.1525-1446.2012.01014.x Google Scholar CrossRef Search ADS PubMed Marsden , P. V . ( 2002 ). Egocentric and sociocentric measures of network centrality . Social Networks , 24 , 407 – 422 . doi: 10.1016/S0378-8733(02)00016-3 Google Scholar CrossRef Search ADS Meeuwesen , L. , Hortulanus , R. , & Machielse , A . ( 2001 ). Social contacts and social isolation: A typology . Netherlands’ Journal of Social Sciences , 37 , 132 – 154 . Moher , D. , Liberati , A. , Tetzlaff , J. , & Altman , D. G .; PRISMA Group . ( 2009 ). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement . Annals of Internal Medicine , 151 , 264 – 9, W64 . doi: 10.7326/0003-4819-151-4-200908180-00135 Google Scholar CrossRef Search ADS PubMed Nguyen , A. W . ( 2017 ). Variations in social network type membership among older African Americans, Caribbean Blacks, and Non-Hispanic Whites . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 72 , 716 – 726 . doi: 10.1093/geronb/gbx016 Google Scholar CrossRef Search ADS Nyqvist , F. , Forsman , A. K. , Giuntoli , G. , & Cattan , M . ( 2013 ). Social capital as a resource for mental well-being in older people: A systematic review . Aging and Mental Health , 17 , 394 – 410 . doi: 10.1080/13607863.2012.742490 Google Scholar CrossRef Search ADS PubMed Park , N. S. , Jang , Y. , Lee , B. S. , Ko , J. E. , Haley , W. E. , & Chiriboga , D. A . ( 2015 ). An empirical typology of social networks and its association with physical and mental health: A study with older Korean immigrants . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 70 , 67 – 76 . doi: 10.1093/geronb/gbt065 Google Scholar CrossRef Search ADS Perkins , J. M. , Subramanian , S. V. , & Christakis , N. A . ( 2015 ). Social networks and health: A systematic review of sociocentric network studies in low- and middle-income countries . Social Science and Medicine (1982) , 125 , 60 – 78 . doi: 10.1016/j.socscimed.2014.08.019 Google Scholar CrossRef Search ADS Rosenquist , J. N. , Fowler , J. H. , & Christakis , N. A . ( 2011 ). Social network determinants of depression . Molecular psychiatry , 16 , 273 – 281 . doi: 10.1038/mp.2010.13 Google Scholar CrossRef Search ADS PubMed Schafer , M. H . ( 2011 ). Health and network centrality in a continuing care retirement community . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 66 , 795 – 803 . doi: 10.1093/geronb/gbr112 Google Scholar CrossRef Search ADS Schafer , M. H . ( 2013 ). Structural advantages of good health in old age: Investigating the health-begets-position hypothesis with a full social network . Research on Aging , 35 , 348 – 370 . doi: 10.1177/0164027512441612 Google Scholar CrossRef Search ADS Schafer , M. H . ( 2015 ). On the locality of asymmetric close relations: Spatial proximity and health differences in a senior community . The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences , 70 , 100 – 110 . doi: 10.1093/geronb/gbu043 Google Scholar CrossRef Search ADS Schafer , M. H . ( 2016 ). Health as status? Network relations and social structure in an American retirement community . Ageing and Society , 36 , 79 – 105 . doi: 10.1017/S0144686X14000993 Google Scholar CrossRef Search ADS Seale , C . ( 2004 ). Researching society and culture . London, UK : Sage . Smith , G. L. , Banting , L. , Eime , R. , O’Sullivan , G. , & van Uffelen , J. G . ( 2017 ). The association between social support and physical activity in older adults: a systematic review . International Journal of Behavioral Nutrition and Physical Activity , 14 , 56 . doi: 10.1186/s12966-017-0509-8 Google Scholar CrossRef Search ADS PubMed Sohn , S. Y. , Joo , W. T. , Kim , W. J. , Kim , S. J. , Youm , Y. , Kim , H. C. , … Lee , E . ( 2017 ). Social network types among older Korean adults: Associations with subjective health . Social Science and Medicine (1982) , 173 , 88 – 95 . doi: 10.1016/j.socscimed.2016.11.042 Google Scholar CrossRef Search ADS Steinbach , U . ( 1992 ). Social networks, institutionalization, and mortality among elderly people in the United States . Journal of Gerontology , 47 , S183 – S190 . Google Scholar CrossRef Search ADS PubMed Valente , T. W . ( 2012 ). Network interventions . Science (New York, N.Y.) , 337 , 49 – 53 . doi: 10.1126/science.1217330 Google Scholar CrossRef Search ADS PubMed Valente , T. W. , & Davis , R. L . ( 1999 ). Accelerating the diffusion of innovations using opinion leaders . The Annals of the American Academy of Political and Social Science , 566 , 55 – 67 . doi: 10.1177/000271629956600105 Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

The GerontologistOxford University Press

Published: Jan 29, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off