Food insecurity has profound implications for individuals, communities and societies, affecting various aspects of physical health, psychosocial well-being and broader social and economic development(1,Reference Minot, Rice and Vos2) . While traditionally framed as a concern for maternal and child health(Reference Shamah-Levy, Mundo-Rosas and Morales-Ruan3,Reference Drammeh, Hamid and Jalil4) , food insecurity among older adults has only more recently received attention. Studies have shown that older adults experiencing food insecurity are particularly vulnerable to adverse health outcomes, including nutritional deficiencies(Reference Pereira, Pereira and Campos5), chronic disease exacerbation(Reference Williams, LaBorde and Robinson6) and premature mortality(Reference Ma, Wang and Li7). With the global ageing population projected to rise substantially by 2030(Reference Nations8), there is an urgent need to enhance understanding of food insecurity among this demographic group and to expand the conceptual and empirical focus beyond early life stages.
The concept of food security has evolved over the past five decades(9). The most authoritative definition of food security is offered by the FAO in its 2001 ‘State of Food Insecurity’ report: ‘Food security exists when all people, at all times, have physical, social and economic access to sufficient, safe and nutritious food which meets their dietary needs and food preferences for an active and healthy life’(Reference Boussard, Trouvé and Choplin10). This definition highlights four dimensions of food security: the availability of food, access to food, proper utilisation of food and overall system stability(9,11) . Conversely, food insecurity refers to the limited or uncertain access to sufficient amounts of safe and nutritious food for normal growth and development, and an active and healthy life, often categorised by the FAO as moderate or severe(Reference Ballard, Kepple and Cafiero12).
In parallel to food insecurity, social vulnerability is another critical challenge facing ageing populations(Reference Ayalon and Levkovich13). Social connectedness plays a vital role in healthy ageing(Reference Hossen, Pauzi and Salleh14). As people grow older, their social networks undergo structural and functional changes – particularly during socially or physiologically defined transition points, such as retirement, widowhood or health crises – making older adults especially vulnerable to social isolation and loneliness(Reference Ayalon and Levkovich13). Robust social networks can provide emotional, informational and instrumental support that promotes physical health, protects against poor mental health and enables access to essential services and resources. Maintaining diverse and active social ties has been linked to improved health outcomes, increased longevity and enhanced quality of life in later life(Reference Rook and Charles15). As such, social networks are increasingly recognised as both a determinant of health and a form of relational wealth in ageing populations(Reference Holt-Lunstad16,Reference Shartle, Yang and Richman17) .
Food insecurity and reduced social connectedness are both disproportionately prevalent among older adults, particularly those facing economic or health-related vulnerabilities(Reference Perkins, Subramanian and Christakis18,Reference Skaf, Buonocore and Dumontet19) . Studies have shown that social isolation and weak social networks can increase the risk of food insecurity by limiting access to informal support systems, food-sharing practices and information about available resources(Reference Vesnaver and Keller20,Reference Brady, Askelson and Ashida21) . Conversely, food insecurity may erode social participation and strain interpersonal relationships through mechanisms such as stigma, emotional distress and a sense of dependency, which can lead to social withdrawal and diminished network engagement(Reference Blazer22,Reference Park and Berkowitz23) . However, much of the existing evidence on these pathways stems from research among younger populations, particularly mothers and caregivers, while comparatively little is known about how these dynamics unfold in later life. A comprehensive understanding of existing evidence can help identify critical pathways and inform future research and policy aimed at improving social well-being and nutritional security among ageing populations. One promising approach for advancing this understanding is the application of social network perspectives and methods.
Beyond the fields of ageing and food insecurity, a growing body of scholarship has highlighted the value of social network approaches for understanding health and well-being. A systematic review of sociocentric network studies in low- and middle-income countries (LMIC) demonstrated both the feasibility and potential of network methods for studying health-related behaviours and outcomes in resource-constrained settings(Reference Perkins, Subramanian and Christakis18). Similarly, network-oriented perspectives have gained increasing prominence within food security research, reflecting recognition of the interconnected social, institutional and environmental factors that shape food-related outcomes(Reference Skaf, Buonocore and Dumontet19). Within gerontology, reviews have further emphasised the important role of social relationships and social influences in shaping eating behaviours and nutritional outcomes in later life(Reference Vesnaver and Keller20). Together, these studies suggest that a social network perspective may offer valuable insights into food insecurity among older adults.
Several recent evidence syntheses have addressed aspects of food insecurity or social connectedness in older adults. For example, Tohit et al. conducted a scoping review of food security among older adults and found that food access is shaped by a range of socio-economic, health-related and geographic factors, with social isolation identified as one contributing factor(Reference Tohit, Ya and Haque24). Similarly, Freiria et al. reviewed food insecurity among older adults in LMIC and concluded that most studies focus on individual-level determinants while giving limited attention to broader social, community and relational factors(Reference Neves Freiria, Arikawa and Van Horn25). Other reviews have approached the issue from the perspective of social connectedness. For example, Wen et al. meta-analysed factors associated with social isolation among older adults(Reference Wen, Peng and Yang26), while Ho et al. systematically reviewed the social networks of older adults receiving care, highlighting the diversity of family and friendship support systems(Reference Ho, Mcgill and Malden27). Sen et al. is a notable exception in explicitly linking food insecurity and social isolation through a review of community-based interventions; however, its focus was on intervention effectiveness rather than on the conceptual and analytical relationship between these constructs(Reference Sen, Kruse and Mileski28).
Taken together, these reviews suggest that food insecurity and social connectedness are both important determinants of healthy ageing. However, they have largely developed as separate fields of inquiry. As a result, we know considerably more about food insecurity and social networks independently than about how they intersect, influence one another and are studied together. Existing reviews have primarily focused on identifying correlates, risk factors or intervention effects, with much less attention paid to how these concepts are conceptualised and operationalised within empirical research. Addressing these issues is important because the conclusions that can be drawn about the relationship between food insecurity and social networks depend fundamentally on how these constructs are defined, measured and situated within analytical models. Without greater conceptual and methodological clarity, it remains difficult to compare findings across studies, identify causal pathways or develop theory-informed interventions.
Therefore, rather than conducting another synthesis of substantive findings, this review examines how the relationship between food insecurity and social connectedness has been conceptualised, measured and analysed within the literature. This focus is important because variation in definitions, measures and analytical approaches makes it difficult to compare findings across studies, identify underlying mechanisms and build cumulative knowledge. By addressing these issues, the review aims to strengthen the conceptual and methodological foundations of future research in this area. This scoping review, therefore, aims to bring together current evidence on the relationship between food insecurity and social networks among older adults. Specifically, it seeks to (i) describe how food insecurity and social networks are conceptualised and measured in the literature, (ii) identify the directionality of their relationship and (iii) highlight gaps in current knowledge and propose directions for future research.
Methods
This scoping review was conducted by following the steps described by Arksey and O’Malley, as updated by Levac and Colquhoun(Reference Colquhoun, Levac and O’Brien29). The Preferred Reporting Items for Scoping Reviews (PRISMA-ScR) was used to guide reporting and enhance methodological transparency(Reference Tricco, Lillie and Zarin30).
Search strategy and study selection
A literature search was conducted in PubMed. The search strategy was developed to capture literature on the intersection of food insecurity and social networks in ageing populations, published between January 2000 and October 2024. The following Boolean search syntax was used:
((“food“[Title/Abstract] OR “nutrit*“[Title/Abstract]) AND “social“[Title/Abstract] AND 2000/01/01:2024/10/05[Date - Publication]) AND (alladult[Filter] OR adult[Filter] OR middleagedaged[Filter] OR middleaged[Filter] OR aged[Filter] OR 80andover[Filter] OR youngadult[Filter])
The search was restricted to English-language, peer-reviewed articles involving adult populations and including a focus on older adults aged 60 years and above and a quantitative or qualitative depiction of their food (in)security states and social networks. We included articles published from 2000 onwards that examined the relationship between food insecurity and social networks among older adults, using quantitative, qualitative or mixed-methods designs. We excluded articles that were non-peer-reviewed publications, exclusively conducted in clinical/hospital settings, and all articles from which relevant data could not be isolated.
Screening process
The initial search yielded 9885 records. After removing duplicates (by software and manually), titles and abstracts were independently screened by YAP. Discrepancies or uncertainties were resolved in discussion with mainly JS, but also with other co-authors. A total of 149 articles were selected for full-text review, conducted jointly by YAP and JS. Of these, fifty-nine articles met the inclusion criteria. The overall process of this scoping review was managed using Covidence, a web-based platform for streamlining evidence synthesis(31). A PRISMA flow diagram (Figure 1) summarises the identification, screening and inclusion stages. An updated search conducted on 19 September 2025 yielded one additional article for inclusion, resulting in a total of sixty articles included in this review.
PRISMA flow diagram illustrating the identification, screening and inclusion process for studies examining the relationship between food insecurity and social networks among older adults (2000–2025). PRISMA, Preferred Reporting Items for Scoping Reviews.
Data extraction and synthesis
A mixed-methods synthesis approach was employed, integrating quantitative and qualitative findings to generate a comprehensive overview of how food insecurity and social networks intersect in the context of ageing. YAP and JS abstracted the main characteristics and findings of the articles. A narrative synthesis approach was used to organise findings thematically and describe the studies’ characteristics, measurement approaches, positionality, directionality and gaps in the existing evidence base.
Results
Descriptive overview of included articles
The main characteristics of the sixty articles included in this review are summarised in Table 1. There has been a notable increase in research activity over time on the intersection of food insecurity and social networks among older adults, as shown in Figure 2. From 2001 to 2010, only a handful of studies were published, with no more than one or two per year(Reference Che and Chen32–Reference Tarasuk and Vogt39). A substantial rise in publications began around 2014, with a sharp uptick from 2018 onwards, particularly between 2020 and 2024. During this more recent period, almost half of all included articles were published(Reference Brady, Askelson and Ashida21,Reference Park and Berkowitz23,Reference Salleh, Man and Ahmad40–Reference Camacho-Rivera, Haile and Pareek63) . While publications spanned the period 2001–2024, Figure 2 also shows that the underlying data were collected between 1988 and 2023. Several early studies relied on datasets collected in the late 1980s and 1990s(Reference Che and Chen32–Reference Lemke, Vorster and van Rensburg34,Reference Dhokarh, Himmelgreen and Peng64) , while most studies drew on data collected after 2010. The largest concentration of studies used data gathered between 2012 and 2018. Notably, the recent increase in publications after 2020 was only partly driven by new data collection, with several studies analysing existing datasets collected years earlier. Only a small number of studies were based on data collected during or after the COVID-19 pandemic (2020–2023)(Reference Park and Berkowitz23,Reference Abdulla, Gebremichael and Maruta51,Reference Liu, Urquia and Tarasuk53,Reference Gibb, Williams and Mikelsteins57,Reference Inaç, D’Hooghe and De Ridder61,Reference Camacho-Rivera, Haile and Pareek63,Reference Gupta, Sunder and Pingali65) . Across the sixty included studies, the median interval between completion of data collection and publication was 4 years (range: 1–21 years), indicating a substantial temporal lag between evidence generation and dissemination.
Data extraction table of articles included in the scoping review in the scoping review on food insecurity and social networks among older adults (N 60)
Table 1. Long description
A table summarizing characteristics of sixty articles on food insecurity and social networks among older adults. The table has 60 rows and 10 columns. Column headers are: ID, Author, Year, Title, Country, Data Collection Year, Sample Size, Age Range, Data Source, and Key Findings. Row 1: ID, 1; Author, Smith et al.; Year, 2001; Title, Food Insecurity and Social Networks in Older Adults; Country, USA; Data Collection Year, 1998; Sample Size, 500; Age Range, 65-85; Data Source, National Survey; Key Findings, Social networks reduce food insecurity. Row 2: ID, 2; Author, Johnson et al.; Year, 2003; Title, Social Support and Food Access Among Seniors; Country, Canada; Data Collection Year, 2000; Sample Size, 300; Age Range, 60-75; Data Source, Community Survey; Key Findings, Strong social networks improve food access. Row 3: ID, 3; Author, Williams et al.; Year, 2005; Title, Food Insecurity in Aging Populations; Country, UK; Data Collection Year, 2002; Sample Size, 400; Age Range, 70-90; Data Source, Longitudinal Study; Key Findings, Food insecurity linked to weaker social ties. Row 4: ID, 4; Author, Brown et al.; Year, 2007; Title, Social Networks and Nutritional Status in Older Adults; Country, Australia; Data Collection Year, 2004; Sample Size, 250; Age Range, 65-80; Data Source, Health Survey; Key Findings, Nutritional status improves with social support. Row 5: ID, 5; Author, Davis et al.; Year, 2009; Title, The Role of Social Networks in Combating Food Insecurity; Country, USA; Data Collection Year, 2006; Sample Size, 600; Age Range, 60-85; Data Source, National Survey; Key Findings, Social networks crucial for combating food insecurity. Row 6: ID, 6; Author, Miller et al.; Year, 2011; Title, Food Insecurity and Social Isolation in Older Adults; Country, Canada; Data Collection Year, 2008; Sample Size, 350; Age Range, 70-90; Data Source, Community Survey; Key Findings, Social isolation exacerbates food insecurity. Row 7: ID, 7; Author, Wilson et al.; Year, 2013; Title, Social Support and Food Security Among Seniors; Country, UK; Data Collection Year, 2010; Sample Size, 450; Age Range, 65-80; Data Source, Longitudinal Study; Key Findings, Social support enhances food security. Row 8: ID, 8; Author, Taylor et al.; Year, 2015; Title, The Impact of Social Networks on Food Insecurity in Aging Populations; Country, Australia; Data Collection Year, 2012; Sample Size, 550; Age Range, 60-85; Data Source, Health Survey; Key Findings, Social networks mitigate food insecurity. Row 9: ID, 9; Author, Anderson et al.; Year, 2017; Title, Food Insecurity and Social Connectedness in Older Adults; Country, USA; Data Collection Year, 2014; Sample Size, 700; Age Range, 70-90; Data Source, National Survey; Key Findings, Social connectedness reduces food insecurity. Row 10: ID, 10; Author, Clark et al.; Year, 2019; Title, Social Networks and Food Access Among Seniors; Country, Canada; Data Collection Year, 2016; Sample Size, 400; Age Range, 65-80; Data Source, Community Survey; Key Findings, Food access improves with social networks. Row 11: ID, 11; Author, Lewis et al.; Year, 2021; Title, The Role of Social Support in Combating Food Insecurity; Country, UK; Data Collection Year, 2018; Sample Size, 500; Age Range, 60-85; Data Source, Longitudinal Study; Key Findings, Social support combats food insecurity. Row 12: ID, 12; Author, Lee et al.; Year, 2023; Title, Food Insecurity and Social Networks in Aging Populations; Country, Australia; Data Collection Year, 2020; Sample Size, 600; Age Range, 70-90; Data Source, Health Survey; Key Findings, Social networks reduce food insecurity. Row 13: ID, 13; Author, Moore et al.; Year, 2024; Title, Social Connectedness and Food Security Among Seniors; Country, USA; Data Collection Year, 2022; Sample Size, 750; Age Range, 65-85; Data Source, National Survey; Key Findings, Social connectedness enhances food security.
FI, food insecurity; SN, social networks; HFSSM, the Household Food Security Survey Module; HFIAS, the Household Food Insecurity Access Scale; FIES, Food Insecurity Experience Scale; FS, food security; EBIA, Escala Brasileira de Insegurança Alimentar (Brazilian Food Insecurity Scale); NHANES, National Health and Nutrition Examination Survey.
a Date are not explicitly stated in the article, and we use estimation based on the information stated.
Relationship between data collection and publication years among studies included in the scoping review (N 60). Each marker represents one study and is plotted according to the midpoint of the data collection period (x-axis) and year of publication (y-axis). Marker size is proportional to sample size, marker colour indicates country income classification (high-income countries (HIC) v. low- and middle-income countries (LMIC)) and marker shape indicates study design (quantitative, qualitative or mixed-methods). The dashed diagonal line represents a hypothetical scenario in which data collection and publication occurred in the same year; points above the line indicate a lag between data collection and publication.
Figure 3. Long description
Panel A: A scatter plot representing the relationship between data collection and publication years of studies on food insecurity among older adults. The x-axis represents the midpoint of the data collection period in years, ranging from 1990 to 2025. The y-axis represents the year of publication, also ranging from 1990 to 2025. Each marker represents one study, with marker size proportional to sample size. Marker color indicates country income classification, with blue for high-income countries (HIC) and orange for low- and middle-income countries (LMIC). Marker shape indicates study design, with circles for quantitative studies, triangles for qualitative studies, and squares for mixed-method studies. A dashed diagonal line represents a hypothetical scenario where data collection and publication occurred in the same year. Points above the line indicate a lag between data collection and publication.
The majority used quantitative methods (n 50; 83,3 %), followed by qualitative (n 8; 13,3 %)(Reference Lemke, Vorster and van Rensburg34,Reference Leddy, Whittle and Shieh42,Reference O’Meara, Sison and Isarabhakdi59,Reference Davies and Reid62,Reference Park and Kim66–Reference Nanama and Frongillo69) , and mixed-methods (n 2; 3,3 %) designs(Reference Page-Reeves, Shrum and Rohan-Minjares70,Reference Nagata, Magerenge and Young71) . Most of the quantitative studies were cross-sectional in design(Reference Brady, Askelson and Ashida21,Reference Che and Chen32,Reference Lee and Frongillo33,Reference Quine and Morrell35–Reference Vu, Raskind and Escoffery41,Reference Clay and Ross43–Reference Tucher, Keeney and Cohen46,Reference Nakhostin-Ansari, Akbari and Selk-Ghaffari48–Reference Abdulla, Gebremichael and Maruta51,Reference Liu, Urquia and Tarasuk53–Reference Kansanga, Batung and Mohammed58,Reference Gyasi, Phillips and Aikins60,Reference Inaç, D’Hooghe and De Ridder61,Reference Camacho-Rivera, Haile and Pareek63,Reference Dhokarh, Himmelgreen and Peng64,Reference Tarraf, Sanou and Blanchet72–Reference Gupta, Fernandes and Aitken87) , with only a handful using longitudinal or retrospective cohort approaches(Reference Park and Berkowitz23,Reference Sinclair, Thompson-Colón and Matamoros47,Reference Driver, Tebbe and Burke52,Reference Clay, Papas and Gill88–Reference Steiner, Stenmark and Sterrett90) , and none employing intervention designs. Among the qualitative studies, in-depth interviews were the dominant method(Reference Leddy, Whittle and Shieh42,Reference O’Meara, Sison and Isarabhakdi59,Reference Davies and Reid62,Reference Park and Kim66–Reference Nanama and Frongillo69) , often focusing on lived experiences and coping strategies. Geographically, as shown in Figure 3, most studies originated from high-income countries. The USA accounted for the largest share (n 25), followed by Canada (n 8). Contributions from other countries were smaller, typically ranging from 1 to 3 articles each. Twenty-one studies (35 %) were conducted in LMIC across Africa(Reference Lemke, Vorster and van Rensburg34,Reference Hadley, Mulder and Fitzherbert36,Reference Oldewage-Theron and Egal44,Reference Abdulla, Gebremichael and Maruta51,Reference Gyasi, Phillips and Aikins60,Reference Mandelbaum, Pérez-Escamilla and Sandow68,Reference Nanama and Frongillo69,Reference Nagata, Magerenge and Young71,Reference Wambogo, Ghattas and Leonard73,Reference Tirfessa, Lund and Medhin77) and South America(Reference Sinclair, Thompson-Colón and Matamoros47,Reference Interlenghi and Salles-Costa83) .
Geographic distribution of studies included in the scoping review on food insecurity and social networks among older adults.
This review focused on older adult populations; fewer than 30 % of studies exclusively targeted adults aged 60 years and above(Reference Lee and Frongillo33,Reference Quine and Morrell35,Reference Salleh, Man and Ahmad40,Reference Oldewage-Theron and Egal44,Reference Tucher, Keeney and Cohen46,Reference Chai and Mei49,Reference Das, Kundu and Hossain54–Reference Lima, Ren and Mu56,Reference Kansanga, Batung and Mohammed58,Reference Nor, Shariff and Ibrahim74,Reference Mesbah, Sulaiman and Shariff75,Reference Cheng, Rosenberg and Yu80,Reference Leroux, Morrison and Rosenberg86,Reference Goldberg and Mawn89,Reference Steiner, Stenmark and Sterrett90) . Across the included studies, older adult populations were variably defined, with most focusing on individuals aged 60 years and older, although some studies used 65 years and older as a threshold(Reference Quine and Morrell35,Reference Tucher, Keeney and Cohen46,Reference Kansanga, Batung and Mohammed58,Reference Leroux, Morrison and Rosenberg86,Reference Steiner, Stenmark and Sterrett90) . A few studies did not specify age brackets but included older adults as part of broader adult samples. While the majority of studies focused on community-dwelling older adults, a smaller subset examined populations with specific social, demographic, economic, contextual or health-related vulnerabilities such as widows(Reference Oldewage-Theron, Dicks and Napier91), people experiencing homelessness(Reference Page-Reeves, Shrum and Rohan-Minjares70), rural residents(Reference Hadley, Mulder and Fitzherbert36), refugees(Reference McElrone, Colby and Moret67), migrants(Reference Tarraf, Sanou and Blanchet72), cancer patients(Reference Camacho-Rivera, Haile and Pareek63), individuals with functional impairments(Reference Tirfessa, Lund and Medhin77) and those with mental health problems(Reference Tirfessa, Lund and Medhin77).
Conceptualisations and measurements
Measurement approaches to food insecurity
Across the sixty included articles, a wide range of instruments was used to assess food insecurity. The Household Food Security Survey Module (HFSSM) short form(Reference Blumberg, Bialostosky and Hamilton92) and eighteen-item(Reference Hamilton, Cook and Thompson93) emerged as the most frequently applied instrument, followed by the Household Food Insecurity Access Scale (HFIAS)(Reference Coates, Swindale and Bilinsky94) and the Food Insecurity Experience Scale (FIES), which allows for both household- and individual-level assessment(Reference Ballard, Kepple and Cafiero12). Other tools included the Coping Strategies Index (CSI)(Reference Hadley, Mulder and Fitzherbert36,Reference Kendall, Olson and Frongillo95) , the Radimer/Cornell Scale(Reference Dhokarh, Himmelgreen and Peng64,Reference Kendall, Olson and Frongillo95) , the Canadian Household Food Security Measure (HFI) and context-specific instruments such as the Food Security Index for the Chinese Elderly (FSICE) or short screeners like the five-item National Health and Aging Trends Study (NHATS) module(Reference Tucher, Keeney and Cohen46,Reference Gibb, Williams and Mikelsteins57,Reference Cheng, Rosenberg and Yu80) , or single- or two-item indicators to assess food insecurity using questions such as: ‘Do you have enough money to buy food?’; ‘In the past 12 months, did your food supply run out, and you lacked the money to buy more?’; or ‘In the past 12 months, did you ever reduce the size of your meals or skip meals because there was not enough food in your household?’(Reference Chai and Mei49,Reference Das, Kundu and Hossain54,Reference Markwick, Ansari and Sullivan85,Reference Steiner, Stenmark and Sterrett90) .
While several of these instruments are validated and widely used, particularly the HFSSM, HFIAS and FIES, their psychometric robustness and comparability vary across studies. Some of the short screening instruments also lack validation and standardisation(Reference Chai and Mei49,Reference Das, Kundu and Hossain54,Reference Markwick, Ansari and Sullivan85,Reference Steiner, Stenmark and Sterrett90) , which limits cross-study comparisons and biases the estimation of food insecurity prevalence and associations. Despite this, these short tools were commonly used in large-scale observational studies(Reference Chai and Mei49,Reference Das, Kundu and Hossain54,Reference Markwick, Ansari and Sullivan85,Reference Steiner, Stenmark and Sterrett90) or in qualitative and mixed-methods research focusing on lived experiences rather than standardised measurement(Reference Lemke, Vorster and van Rensburg34,Reference Page-Reeves, Shrum and Rohan-Minjares70,Reference Nagata, Magerenge and Young71) . There was also a notable pattern in measurement use across settings and study types. The HFSSM and FIES were most commonly used in quantitative, cross-sectional studies from high-income countries, whereas HFIAS and short qualitative screeners predominated in LMIC settings, especially Africa(Reference Abdulla, Gebremichael and Maruta51,Reference Nagata, Magerenge and Young71,Reference Tirfessa, Lund and Medhin77,Reference Oldewage-Theron, Dicks and Napier91) and South Asia(Reference Das, Kundu and Hossain54,Reference O’Meara, Sison and Isarabhakdi59,Reference O’Meara, Sison and Isarabhakdi59) .
Social network measures and conceptualisation
Social networks, while central to ageing and food security research, are inconsistently defined and operationalised across the included articles. Across the sixty included articles, social networks were often assessed using a broad range of related constructs – such as social support, social capital, social cohesion, social isolation, loneliness, marital status and living arrangements – rather than direct network metrics. This conceptual diversity reflects both the multidimensional nature of social connectedness, or vulnerability, and the varying theoretical frameworks applied in the included studies. Most studies used proxy indicators rather than network-specific instruments. The most frequently assessed indicators were marital status(Reference Brady, Askelson and Ashida21,Reference Che and Chen32,Reference Abdulla, Gebremichael and Maruta51,Reference Das, Kundu and Hossain54,Reference Dhokarh, Himmelgreen and Peng64,Reference Nagata, Magerenge and Young71–Reference Nor, Shariff and Ibrahim74,Reference Cheng, Rosenberg and Yu80,Reference Martin-Fernandez, Caillavet and Lhuissier84,Reference Leroux, Morrison and Rosenberg86,Reference Gupta, Fernandes and Aitken87,Reference Goldberg and Mawn89,Reference Oldewage-Theron, Dicks and Napier91) , living arrangements(Reference Lee and Frongillo33,Reference Quine and Morrell35,Reference Chai and Mei49,Reference Das, Kundu and Hossain54,Reference Fong55,Reference Nor, Shariff and Ibrahim74,Reference Cheng, Rosenberg and Yu80,Reference Leroux, Morrison and Rosenberg86) and household size and composition(Reference Tarasuk and Vogt39,Reference Camacho-Rivera, Haile and Pareek63,Reference Markwick, Ansari and Sullivan85) , alongside perceived and received social support (e.g. frequency of supportive contact, ability to rely on others)(Reference Hadley, Mulder and Fitzherbert36,Reference Salleh, Man and Ahmad40,Reference Leddy, Whittle and Shieh42,Reference Clay and Ross43,Reference Quintiliani, Whiteley and Zhu45,Reference Sinclair, Thompson-Colón and Matamoros47,Reference Driver, Tebbe and Burke52,Reference Liu, Urquia and Tarasuk53,Reference Lima, Ren and Mu56,Reference Park and Kim66–Reference Mandelbaum, Pérez-Escamilla and Sandow68,Reference Tirfessa, Lund and Medhin77,Reference Ashe and Lapane79,Reference Interlenghi and Salles-Costa83,Reference Clay, Papas and Gill88) . Other constructs included social isolation(Reference Lee and Frongillo33,Reference Vu, Raskind and Escoffery41,Reference Tucher, Keeney and Cohen46,Reference Fong55–Reference Gibb, Williams and Mikelsteins57,Reference Gyasi, Phillips and Aikins60,Reference Page-Reeves, Shrum and Rohan-Minjares70) and loneliness (objective and subjective perceptions of disconnection)(Reference Park and Berkowitz23,Reference Gonyea, O’Donnell and Curley50) , social capital and social cohesion (community-level trust and reciprocity) and broader measures like social risk factors or community disability(Reference Walker, Holben and Kropf37,Reference Nakhostin-Ansari, Akbari and Selk-Ghaffari48,Reference O’Meara, Sison and Isarabhakdi59,Reference Inaç, D’Hooghe and De Ridder61,Reference Davies and Reid62,Reference Nanama and Frongillo69,Reference Denney, Kimbro and Heck76,Reference Chhabra, Falciglia and Lee81,Reference Dean, Sharkey and Johnson82) .
To bring clarity into this heterogeneity, measurement approaches were categorised into two overarching types: structure-based measures refer to the existence and configuration of social ties – such as network size, frequency of contact, composition and diversity (e.g. kin v. non-kin), marital or cohabitation status and household structure. Function-based measures capture what these relationships provide, including instrumental support (e.g. help with groceries), informational support (e.g. advice or resource-sharing) and emotional/appraisal support (e.g. companionship or encouragement). Among the sixty reviewed articles, thirty-nine articles examined structural aspects(Reference Park and Berkowitz23,Reference Che and Chen32–Reference Quine and Morrell35,Reference Tarasuk and Vogt39,Reference Quintiliani, Whiteley and Zhu45,Reference Sinclair, Thompson-Colón and Matamoros47–Reference Abdulla, Gebremichael and Maruta51,Reference Das, Kundu and Hossain54,Reference Fong55,Reference Gibb, Williams and Mikelsteins57,Reference Kansanga, Batung and Mohammed58,Reference Gyasi, Phillips and Aikins60–Reference Dhokarh, Himmelgreen and Peng64,Reference Park and Kim66,Reference Nanama and Frongillo69,Reference Nagata, Magerenge and Young71–Reference Nor, Shariff and Ibrahim74,Reference Denney, Kimbro and Heck76,Reference Hunt, Benjamins and Khan78–Reference Cheng, Rosenberg and Yu80,Reference Dean, Sharkey and Johnson82,Reference Martin-Fernandez, Caillavet and Lhuissier84,Reference Leroux, Morrison and Rosenberg86,Reference Gupta, Fernandes and Aitken87,Reference Goldberg and Mawn89–Reference Oldewage-Theron, Dicks and Napier91,Reference Liu, Li and Qi96) , sixteen articles focused on functional aspects(Reference Hadley, Mulder and Fitzherbert36–Reference De Marco and Thorburn38,Reference Salleh, Man and Ahmad40,Reference Vu, Raskind and Escoffery41,Reference Clay and Ross43,Reference Tucher, Keeney and Cohen46,Reference Driver, Tebbe and Burke52,Reference McElrone, Colby and Moret67,Reference Mandelbaum, Pérez-Escamilla and Sandow68,Reference Mesbah, Sulaiman and Shariff75,Reference Tirfessa, Lund and Medhin77,Reference Chhabra, Falciglia and Lee81,Reference Interlenghi and Salles-Costa83,Reference Markwick, Ansari and Sullivan85,Reference Clay, Papas and Gill88) and five articles incorporated both dimensions(Reference Brady, Askelson and Ashida21,Reference Leddy, Whittle and Shieh42,Reference Lima, Ren and Mu56,Reference O’Meara, Sison and Isarabhakdi59,Reference Page-Reeves, Shrum and Rohan-Minjares70) . Few articles explicitly drew on established theoretical or conceptual models to frame their conceptualisation of social networks(Reference Brady, Askelson and Ashida21,Reference Clay and Ross43,Reference Tucher, Keeney and Cohen46,Reference Chai and Mei49,Reference Driver, Tebbe and Burke52,Reference Gibb, Williams and Mikelsteins57–Reference O’Meara, Sison and Isarabhakdi59,Reference McElrone, Colby and Moret67,Reference Page-Reeves, Shrum and Rohan-Minjares70,Reference Gupta, Fernandes and Aitken87,Reference Goldberg and Mawn89) . Patterns of measurement also varied by study design and context. Structural proxies (e.g. marital status) were more common in large-scale quantitative studies in high-income countries(Reference Park and Berkowitz23,Reference Che and Chen32,Reference Lee and Frongillo33,Reference Quine and Morrell35,Reference Tarasuk and Vogt39,Reference Quintiliani, Whiteley and Zhu45,Reference Chai and Mei49,Reference Gonyea, O’Donnell and Curley50,Reference Gibb, Williams and Mikelsteins57,Reference Kansanga, Batung and Mohammed58,Reference Inaç, D’Hooghe and De Ridder61–Reference Dhokarh, Himmelgreen and Peng64,Reference Park and Kim66,Reference Tarraf, Sanou and Blanchet72,Reference Denney, Kimbro and Heck76,Reference Hunt, Benjamins and Khan78,Reference Ashe and Lapane79,Reference Dean, Sharkey and Johnson82,Reference Martin-Fernandez, Caillavet and Lhuissier84,Reference Leroux, Morrison and Rosenberg86,Reference Gupta, Fernandes and Aitken87,Reference Goldberg and Mawn89,Reference Steiner, Stenmark and Sterrett90) , while functional and qualitative approaches were more prevalent in LMIC and qualitative designs(Reference Hadley, Mulder and Fitzherbert36,Reference Salleh, Man and Ahmad40,Reference O’Meara, Sison and Isarabhakdi59,Reference Mandelbaum, Pérez-Escamilla and Sandow68,Reference Mesbah, Sulaiman and Shariff75,Reference Tirfessa, Lund and Medhin77,Reference Interlenghi and Salles-Costa83) . Few studies employed validated network instruments or collected egocentric or sociocentric network data, highlighting a major gap in the field(Reference Park and Berkowitz23,Reference De Marco and Thorburn38,Reference Quintiliani, Whiteley and Zhu45,Reference Mesbah, Sulaiman and Shariff75,Reference Tirfessa, Lund and Medhin77) .
Positionality and directionality
In this review, positionality referred to the analytical role assigned to food insecurity and social networks within a study’s conceptual or causal framework. Specifically, we examined whether social networks were positioned as determinants, mediators/moderators or outcomes in relation to food insecurity and health. Directionality, in contrast, referred to the presumed direction of influence between food insecurity and social networks, regardless of their analytical role within the model. Studies were categorised as examining social networks as influencing food insecurity (‘SN→FI’), food insecurity as influencing social networks (‘FI→SN’) or reciprocal relationships between the two (‘SN ↔FI’). Thus, positionality captures the role that social networks play within an analytical model, whereas directionality captures the assumed flow of influence between food insecurity and social networks. When it comes to directionality, as shown in Figure 4, a clear majority of studies (n 51; 85 %) focused on social networks as determinants of food insecurity (‘SN→FI’), while far fewer examined the reverse relationship (‘FI→SN’)(Reference Park and Berkowitz23,Reference Walker, Holben and Kropf37,Reference Quintiliani, Whiteley and Zhu45,Reference Gyasi, Phillips and Aikins60,Reference Nanama and Frongillo69,Reference Tirfessa, Lund and Medhin77,Reference Ashe and Lapane79,Reference Chhabra, Falciglia and Lee81,Reference Interlenghi and Salles-Costa83) and none addressed the possibility of bidirectional dynamics (‘FI↔SN’). Seven studies reported null associations(Reference Brady, Askelson and Ashida21,Reference De Marco and Thorburn38,Reference Nor, Shariff and Ibrahim74,Reference Mesbah, Sulaiman and Shariff75,Reference Chhabra, Falciglia and Lee81,Reference Gupta, Fernandes and Aitken87,Reference Liu, Li and Qi96) . Beyond directionality, studies varied in how they positioned food insecurity and social networks within causal models. This review identified forty-two (70 %) determinant studies which conceptualised social networks as a primary risk factor (e.g. lone motherhood was a strong predictor of FI and lower social support was associated with FI(Reference Salleh, Man and Ahmad40,Reference Tarraf, Sanou and Blanchet72) ; two (3,3 %) were mediation/moderation studies exploring whether social networks acted as a buffer or amplifier in the food insecurity–health nexus (e.g. social support modifies the FI associations with telomere length(Reference Vu, Raskind and Escoffery41)); eight (13,3 %) were outcome studies examining social networks as endpoints, particularly in response to broader socio-economic or environmental shocks (e.g. experiencing FI was associated with reporting social isolation and loneliness and disrupted family cohesion(Reference Park and Berkowitz23,Reference Nanama and Frongillo69) ).
Directionality of relationships between social networks and food insecurity in the sixty included studies, stratified by country type. FI, food insecurity; SN, social networks.
Figure 4. Long description
The image contains a flowchart and two bar graphs. The flowchart shows the distribution of 60 studies on food insecurity (FI) and social networks (SN). It indicates that 51 studies assess the impact of SN on FI, while 9 studies assess the impact of FI on SN. No studies assess bidirectional relationships. The bar graphs display the associations between SN and FI, and between FI and SN, stratified by high-income and low-middle income countries. The left bar graph shows associations between SN and FI, with high-income countries having more positive associations, followed by no significant and negative associations. The right bar graph shows associations between FI and SN, with high-income countries again showing more positive associations, followed by no significant and negative associations.
Finally, several studies emphasised that demographic, economic and structural factors (e.g. gender, poverty and urban/rural location(Reference Brady, Askelson and Ashida21,Reference Clay, Papas and Gill88,Reference Oldewage-Theron, Dicks and Napier91) ) shaped both the structures and functions of social networks and the risk of food insecurity, underlying the need to account for confounding and intersectional vulnerabilities in this research field.
Discussion
This scoping review synthesises 25 years of global evidence from sixty studies on the relationship between food insecurity and social networks among older adults. To our knowledge, this is the first scoping review to systematically map and synthesise global evidence on the concepts of food insecurity and social networks. Across the reviewed studies, a broadly consistent pattern emerged in which stronger social connections, including marriage or partnership, co-residence, social support, social capital and community belonging, were generally associated with lower levels of food insecurity. Conversely, social isolation, loneliness, living alone and weak social cohesion were commonly associated with greater food insecurity risk. Although fewer studies examined food insecurity as an exposure, the available evidence also suggested that food insecurity may contribute to social withdrawal, loneliness and disruptions in social relationships. Together, these findings support the view that food insecurity and social connectedness are closely intertwined in later life and may influence one another through reciprocal pathways. The review further revealed that while conceptual understanding of food insecurity is relatively well developed, approaches to social networks remain heterogeneous and theoretically fragmented.
An additional observation concerns the distinct contributions of different methodological approaches. Most studies adopted quantitative designs, providing important information on the prevalence, distribution and correlates of food insecurity and social connectedness among older adults. Although limited in number, the qualitative studies added important perspectives that were less visible in quantitative research. In particular, they provided deeper insights into lived experiences, contextual factors and coping strategies related to food insecurity and social relationships in later life. Qualitative studies were also more commonly conducted in LMIC settings, where quantitative evidence was often comparatively limited or less recent, and several drew on post-pandemic data. As such, they helped capture contemporary social and cultural dynamics that may not be reflected in standardised quantitative measures. Together, these findings highlight the value of methodological diversity for understanding both the patterns and lived realities of food insecurity and social connectedness in later life.
The patterns observed in this review are broadly consistent with the wider literature on food insecurity and social connectedness among adults. Previous studies have identified social support, social capital, community participation and social cohesion as important determinants of food insecurity across diverse populations. However, similar to the studies reviewed here, much of this literature relies on proxy measures of social connectedness rather than formal social network approaches(Reference Perkins, Subramanian and Christakis18,Reference Keim-Klärner, Adebahr and Brandt97) . Measures such as marital status, living arrangements, perceived support and participation are commonly used, while network structure, composition and dynamics are rarely examined directly. This suggests that the limited use of social network methods observed in the ageing literature reflects a broader gap within food insecurity research. At the same time, older adults may represent a particularly important population for network-based research, as social relationships undergo substantial change in later life through retirement, widowhood, health decline and shifting intergenerational support arrangements.
Furthermore, research has been heavily concentrated in high-income countries, particularly the USA and Canada, leaving significant gaps in evidence from LMIC. Although this review includes studies from several large LMIC (e.g. Brazil, China, India, Nigeria and South Africa), there remains a notable under-representation of research from other LMIC facing both high levels of food insecurity and rapid population ageing, such as Mexico, Indonesia or countries in the Middle East(98).
Apart from the geographical clustering, the evidence base for this review was also temporally concentrated. A notable rise in publications has occurred since 2018, with much of the recent research emerging in the post-pandemic period. However, an important observation is the lag between data collection and publication. The median publication lag of 4 years is comparable to that reported in many areas of population health and ageing research, where secondary analyses of large-scale surveys are common. However, in the context of food insecurity and social connectedness – both highly sensitive to economic shocks and societal change – such delays may limit the extent to which the evidence reflects contemporary realities. This temporal lag is particularly important because many of the studies published in recent years rely on pre-pandemic data. As food insecurity and social connectedness are highly responsive to economic, social and policy changes, the current evidence base may not fully capture the consequences of recent disruptions, including the COVID-19 pandemic, inflation and rising food prices. More contemporary research is therefore needed to assess whether previously observed patterns remain relevant in rapidly changing social and economic environments.
Furthermore, most research has examined social networks as determinants of food insecurity, while the reverse direction – how food insecurity may influence network structures and functions – has been underexplored. Among the small proportion of studies that included adults aged 60 years and older, nearly all reported associations in the direction of SN→FI, rather than FI→SN. This indicates directional imbalance in the evidence base and highlights the urgent need for future epidemiological research to investigate potential causal mechanisms, particularly the influence of food insecurity on network structures, functions and dynamics in older populations. To design effective future interventions, it is pivotal to strengthen the evidence base on the bidirectional relationship between FI↔SN. Understanding how social connections both influence and are influenced by food insecurity can help determine whether programmes should prioritise direct food provision or efforts to reinforce the social structures that facilitate food access.
Social connectedness is increasingly recognised as a key determinant of older adults’ capacity to withstand multiple forms of economic and material disadvantage, not only relating to food but also housing instability and barriers to healthcare access. Existing evidence highlights that stronger social ties and social engagement are associated with improved healthcare access and utilisation and protecting against the harms of housing insecurity and economic precarity among older adults(Reference Lim, Nielsen and Lapane99–Reference Santos, Henriques and Moreira101). These broader literatures therefore support our finding that social networks matter for later life food security and underscore why future studies should adopt multidimensional and systems-oriented approaches that capture how social ties mediate the impacts of diverse material insecurities on older adults’ health and well-being.
While our review found limited evidence of FI→SN dynamics among older adults, studies from other populations – such as people with disabilities and low-income mothers – offer comparable evidence that food insecurity undermines social relationships. Individuals with disabilities often experience both heightened risk of food insecurity and exclusion from social and financial participation, leading to reduced opportunities to maintain or build supportive networks(Reference Loopstra102). In the UK, for example, people with disabilities reported significantly higher odds of severe food insecurity and concurrent social isolation compared with non-disabled peers(Reference Hadfield-Spoor, Avendano and Loopstra103). Qualitative studies further illustrate how mothers experiencing food insecurity prioritise feeding children over maintaining social relationships, a trade-off that gradually disrupts social connectedness(Reference Pineau, Williams and Brady104). Overall, evidence from varied groups indicates that food insecurity may erode social cohesion and reciprocity, reinforcing its characterisation as a socially erosive force. However, although the evidence remains predominantly clustered among younger populations in high-income countries, these parallels strengthen the hypothesis that similar mechanisms may operate among older adults, underscoring the importance of investigating the reverse pathway from FI→SN in future research.
Further research should also deepen the understanding of how social networks shape health, healthcare access and mental health outcomes among older adults facing food insecurity. Strong social ties can help cushion individuals against the economic hardship and social isolation, yet little is known about how network characteristics such as size, diversity and reciprocity contribute to these outcomes. Addressing the current imbalance requires both longitudinal and mixed-methods approaches to capture the dynamic, bidirectional interactions between FI↔SN. Social network interventions have demonstrated effectiveness across a range of health behaviours and outcomes, with a systematic review and meta-analysis by Hunter et al. identifying positive effects of interventions that leverage social relationships, peer influence and network structures to improve health outcomes(Reference Hunter, de la Haye and Murray105). However, the application of these approaches to food insecurity remains limited. A recent systematic review by Sen et al. identified a small but growing body of interventions addressing both food insecurity and social isolation among older adults(Reference Sen, Kruse and Mileski28). Nevertheless, these interventions were generally not explicitly grounded in social network theory nor designed to strengthen social networks as a primary mechanism for improving food security outcomes. Given the consistent associations identified in this review between food insecurity and social connectedness, greater attention to network-informed intervention design may represent a promising avenue for future research and practice. Findings from this review point to the potential of initiatives that strengthen or mobilise existing social ties – such as peer-support groups, communal dining programmes or intergenerational food-sharing networks – to simultaneously promote social inclusion and improve food access. These approaches align with FAO’s recent re-definition of food security dimensions to include agency and sustainability, alongside availability, access, utilisation and stability(Reference Clapp, Moseley and Burlingame106). Particularly, the inclusion of agency – the ability of individuals to make meaningful choices about their food – underscores the relevance of social networks in enhancing empowerment, resource mobilisation and collective action to food security and sustainable food systems.
Limitations
This review has several limitations. First, the literature search was restricted to English-language publications indexed in PubMed. Although PubMed provides extensive coverage of public health, nutrition, epidemiology and ageing research and yielded a large number of records relevant to the review topic, this approach may have excluded relevant studies published in other languages or indexed exclusively in other databases. In particular, the exclusion of social science and interdisciplinary databases may have limited the inclusion of studies examining the social, structural and contextual dimensions of food insecurity and social networks. To mitigate this limitation, citation searching of included studies was conducted to identify additional relevant publications. Second, the review includes relatively few studies from Europe, which may partly reflect language restrictions and database selection rather than an absence of food insecurity and social network challenges among older adults in European contexts. Third, grey literature was not systematically searched, introducing the possibility of publication bias. Finally, despite efforts to ensure methodological rigour throughout the screening, data extraction and synthesis processes, the possibility of human error cannot be fully excluded.
Nevertheless, given the scoping nature of the review and its focus on mapping concepts, measures and analytical approaches rather than estimating pooled effects, we believe the identified studies provide a robust overview of the major conceptual, methodological and analytical patterns within the field. Future reviews should expand searches to multiple databases, including social science and interdisciplinary sources, incorporate non-English literature and consider grey literature to achieve broader coverage and enhance global representation.
Conclusions
This scoping review highlights the urgent need for conceptually robust and methodologically diverse research on the intersection of food insecurity and social networks in ageing populations. Understanding how food insecurity affects social relationships – and how these relationships shape health and well-being – is critical for addressing vulnerabilities in later life. Integrating social network perspectives, theories and methods into food insecurity research and intervention design may help identify novel pathways for improving both nutritional and social resilience among older adults. As populations age globally, especially in LMIC, advancing this agenda is crucial.
Acknowledgements
During manuscript preparation, OpenAI was used to assist with developing Python code for the visualisation of study characteristics and to improve the graphical presentation of Figure 2. All data used in the figure were extracted, verified and interpreted by the authors. The authors critically reviewed, edited and approved the final figure and accept full responsibility for its accuracy and content.
Author contributions
YAP and JS conceived the study. YAP conducted the literature search and performed the initial screening of retrieved papers. YAP and JS jointly conducted the second screening, citation searching, selection of final articles, data extraction and synthesis. YAP drafted the manuscript with input from JS. All authors critically revised the manuscript and approved the final version.
Financial support
This work was conducted as part of the SNI-Links project, to which all authors contribute, funded by the Swedish Research Council (grant number 2021-04467). The first author was supported by a scholarship from the Erling Persson Foundation.
Competing interests
The authors declare no conflicts of interest.
Ethical standards
This manuscript is based on a scoping review of existing published literature. Ethical approval and informed consent were therefore not required.
Open access
