Geographic and thematic imbalance in household food waste research (2011–2025): a CiteSpace-based scoping review

Abstract

Introduction:

Household food waste poses a major barrier to achieving UN SDG Target 12.3. Despite rapid research growth, most existing bibliometric reviews focus on counting publications and mapping trends, leaving a critical question unanswered: whose problems, whose contexts, and whose solutions are being prioritized?

Methods:

This study conducted a CiteSpace-based scoping review of 616 articles (2011–2025) from Web of Science, CNKI, SciELO, and AJOL, with cross-validation using Scopus.

Results:

Three core findings emerged. (1) Global knowledge production was geographically and thematically uneven. The Global South, with the notable exception of China, produced few publications and occupied peripheral network positions. Researchers from lower-income countries rarely led domestic research before 2025. High-income country research focused on behavioral drivers, whereas lower-income research prioritized food security and infrastructure. Direct keyword connections between Northern and Southern clusters totaled only two edges. (2) The field was split between behavioral reduction and end-of-pipe management. Behavioral and engineering clusters shared less than one percent of cross-connected edges and mixed cliques. (3) The Theory of Planned Behavior, long the dominant framework, showed a declining relative share. Its publication share peaked at 30.4% in 2021 and declined to 20.9% by 2025. Scholars were supplementing it with emotion-related and context-specific variables.

Discussion:

This study identified geographic inequity, thematic misalignment, and disciplinary fragmentation as structural constraints. The field has evolved from engineering and short-term behavioral approaches toward macro-contextual interventions and long-term lifestyle-oriented strategies. This study calls for inclusive collaboration, context-sensitive theorization, and investment in foundational data from low-resource settings to support evidence-based progress toward SDG 12.3.

1 Introduction

Food waste has become one of the most urgent challenges facing the global food system (Alkindi et al., 2026). According to reports from the Food and Agriculture Organization (FAO) and the United Nations Environment Programme, approximately 1/3 of all food produced for human consumption is lost or wasted each year, amounting to 1.3 billion tons (Chepeliev and Gatto, 2024; Teku et al., 2025). This results in nearly one trillion US dollars in economic losses and contributes 8 to 10% of global greenhouse gas emissions, exacerbating resource depletion, biodiversity loss, and food insecurity (Sjerven, 2019). Within the global sustainable development agenda, the United Nations Sustainable Development Goal (SDG) Target 12.3 explicitly calls for halving per capita global food waste at the retail and consumer levels by 2030 and reducing food losses along production and supply chains (Ardra and Barua, 2022). The European Union, the United States, China, India, and the African Union have subsequently introduced national targets and action plans, elevating food waste from an environmental issue to a global core agenda with economic, social, health, and ethical dimensions (Hu et al., 2024).

The household, as the core setting of food consumption, is a primary source of global food waste. Reducing household food waste directly contributes to SDG 2 (Zero Hunger), SDG 12 (Responsible Consumption), SDG 13 (Climate Action), and SDG 17 (Partnerships) (De Visser-Amundson, 2022; Oláh et al., 2022; Theodoridis and Zacharatos, 2022). According to the United Nations Environment Programme’s Food Waste Index Report 2024, approximately 1.05 billion tons of food are wasted annually at the household, food service, and retail levels, with households contributing about 60% of this total (Mungkung et al., 2026). More broadly, sustainable waste management has emerged as a cross-cutting research domain covering resource recovery, waste-derived materials, mechanical recycling systems (Agarwal, 2018; Sathish et al., 2025), as well as business model innovations for the circular economy (Fraccascia et al., 2021). Household food waste still stands out, as it is closely tied to daily routines, consumer psychology and retail environments (González-Santana et al., 2022). Household food waste behavior is influenced by multiple factors, including consumption habits, cultural norms, economic conditions, family structure, and purchasing and storage practices, exhibiting significant individual variation and contextual specificity (Farr-Wharton et al., 2014; Liao et al., 2018; Lin and Guan, 2021; Loxton et al., 2020; Wang et al., 2015). These characteristics make household food waste an indispensable component of food waste management systems, and the effectiveness of its management directly affects the achievement of SDG Target 12.3 (Manzoor et al., 2024). In recent years, it has become a focal point of multidisciplinary research on sustainable waste-management practices, resource utilization, and environmental management frameworks.

As literature has expanded rapidly, the academia has produced a number of valuable reviews. Thematic reviews have systematically identified behavioral drivers, intervention strategies, and cross-cultural patterns (Chia et al., 2024b; Morkunas et al., 2024; Nguyen and Tran, 2025; Salume et al., 2026). Descriptive bibliometric studies have mapped publication trends, keyword evolution, and country collaboration networks (Salume et al., 2026; Yang and Kazumitsu, 2024). Despite these contributions, existing reviews leave a critical question unanswered: whose problems, whose contexts, and whose solutions are being prioritized? Descriptive bibliometric studies report what has been published and where, but they do not examine whose problems and contexts are being prioritized. Thematic reviews identify behavioral drivers across countries, but they rarely ask whether the global research agenda is structurally aligned with the diverse realities of food waste across different developmental contexts—where the primary drivers of waste may shift from individual attitudes to infrastructural constraints, such as unreliable electricity or limited cold chain access. In other words, the field has focused on measuring and explaining household food waste, but it has not critically examined the geographic and thematic priorities embedded in its own knowledge production. A second, related limitation concerns methodology. Most bibliometric reviews rely on a single database, such as Web of Science or Scopus alone (Rodrigues et al., 2025; Salume et al., 2026). While each database has its own coverage strengths and biases, relying on a single database risk omitting relevant literature. Conversely, simply merging multiple databases introduces practical challenges, including inconsistent data formats and duplicate records (Donthu et al., 2021). As a result, there is no consensus on the optimal database strategy for bibliometric reviews.

This study offers a complementary perspective by re-reading existing bibliometric patterns through the lens of knowledge production inequality. Its contribution is threefold. First, it moves beyond counting publications to examine who studies whom: using author–study context classifications, it reveals whether research on the Global South is led by Global South scholars or remains dependent on Global North framing. Second, it quantifies thematic priorities between high-income and low-income country research, showing whether behavioral and psychological frameworks dominate in all contexts or give way to infrastructure and food security concerns in lower-income settings. Third, it provides a methodologically transparent approach to database validation: Web of Science serves as the primary source for CiteSpace analysis. Scopus is used for confirmatory testing of geographic publication patterns without merging (Farooq, 2023; Pasko et al., 2021; Rodrigues et al., 2025), as harmonizing heterogeneous metadata structures between the two databases for comparative network analysis would introduce methodological inconsistencies (Donthu et al., 2021). CNKI, SciELO and AJOL records, by contrast, were integrated into the main analytical dataset after manual standardization of metadata formats. Accordingly, this study aims to: first, clarify the temporal and geographical patterns of knowledge production in household food waste research; second, identify the evolutionary trajectory of research themes and paradigm shifts; and third, to examine whether global knowledge production is aligned with the diverse realities of food waste across different development contexts. By highlighting patterns consistent with geographic concentration and thematic priorities, this review seeks to stimulate reflection on geographic and thematic equity and to provide a context-sensitive foundation for advancing progress toward SDG Target 12.3 worldwide.

2 Methodology

2.1 Data sources and integration

This study integrated literature from four databases to ensure comprehensive coverage of global household food waste research. Web of Science Core Collection served as the primary international database, selected for its standardized bibliographic metadata across all indexed records, including consistent author keywords and citation references essential for reliable CiteSpace analysis (Donthu et al., 2021). To address the under-indexing of Global South journals in Web of Science, three regional databases were also included: CNKI for Chinese-language literature, SciELO for Latin American literature, and AJOL for African literature. This multi-database integration strategy directly responds to concerns about database coverage bias in bibliometric studies.

Merging multi-source data presents practical challenges, including inconsistent metadata formats and duplicate records (Donthu et al., 2021). To address these issues, author affiliations were manually standardized following a predefined protocol, keyword formats were unified according to CiteSpace requirements, and duplicate records were identified using database detection tools and manually verified. Non-English records from CNKI and SciELO were translated and cross-checked for accuracy, then formatted to align with Web of Science metadata standards prior to merging.

Author country affiliation was determined based on the corresponding author’s address, as recorded in each database. This follows standard bibliometric practice for assigning research output to countries and avoids double counting in internationally collaborative publications (Gauffriau, 2021; Harsh et al., 2021).

The search was conducted on 12 June 2026 using the following Boolean search string across all databases: TS = (“household food waste” OR “home food waste” OR “domestic food waste” OR “food waste at home” OR “household food wastage” OR “consumer food waste” OR “domestic food loss” OR “avoidable food waste” OR “kitchen waste behavior”). The time span was limited to 1 January 2011 to 31 December 2025, covering the rapid development phase of household food waste research following the adoption of SDG Target 12.3.

The initial search yielded 953 records from Web of Science, 59 from CNKI, 21 from SciELO, and 0 from AJOL. After removing duplicates and applying consistent screening criteria across all databases, a final sample of 616 empirical journal articles was included for bibliometric analysis. Of these, 579 were from Web of Science, 31 from CNKI, and 6 from SciELO. The screening process followed the PRISMA protocol (Figure 1). Three explicit exclusion criteria were applied uniformly: (1) studies focusing exclusively on non-household food waste settings, including catering service, retail, wholesale, agricultural production, and industrial processing; (2) purely technical engineering studies that only investigated technical parameters such as anaerobic digestion efficiency, biogas production, fermentation performance, and sludge treatment, without involving any behavioral, policy, social, or socioeconomic dimensions; (3) conceptual reviews, theoretical syntheses, and perspective papers without original empirical data collection and empirical analysis.

Figure 1

The exclusion of theoretical reviews may have omitted some influential conceptual contributions relevant to thematic framing. Nevertheless, this review deliberately restricts its scope to empirical research output to maintain consistent analytical boundaries and comparability across temporal and geographic patterns.

No language restriction was imposed during the initial database search. However, for non-English publications to be included, they required complete standardized bibliometric metadata (title, abstract, keywords, author affiliation, publication year) compatible with CiteSpace processing. After manual standardization, CNKI and SciELO records that met this requirement were retained; those that did not were excluded.

Two researchers completed title and abstract screening independently. Inter-rater reliability was assessed using Cohen’s Kappa, with an overall agreement rate of 92% and a Kappa value of 0.85, indicating almost perfect consistency beyond chance (Kalaycioglu et al., 2023). All screening disagreements were resolved through discussion to reach a consensus. Duplicate papers were identified using the built-in detection tools of each database and then verified manually. Papers passing the first screening entered full-text evaluation, following the same independent double-check and consensus adjudication procedure.

2.2 Cross-validation with Scopus

To assess whether the core findings derived from the integrated dataset generalize across a different bibliometric platform, a confirmatory analysis was performed using the Scopus database. Scopus was selected for cross-validation because it is the most widely used alternative database with comparable coverage but a different indexing structure (Dong and Wu, 2026). Using the identical search strategy and the same inclusion and exclusion criteria, a validation dataset comprising 775 articles was retrieved from Scopus. The Scopus dataset was analyzed separately for geographic publication patterns. Thematic priorities, burst detection, and clustering structure were not cross-validated due to differences in indexing and metadata structure between the databases. The thematic priorities, burst detection, and clustering structure reported in Sections 3.4–3.5 were based on the full integrated dataset, which included records from Web of Science, CNKI, AJOL and SciELO. However, these findings may not generalize to other databases such as Scopus due to differences in indexing and metadata structure. These findings may not generalize to other databases due to differences in indexing and metadata structure. This validation design complies with mainstream practices for multi-database bibliometric research (Farooq, 2023; Pasko et al., 2021; Rodrigues et al., 2025).

2.3 Change-point detection for phase division

To identify statistically significant shifts in publication volume, a change-point detection analysis was conducted using the Binary Segmentation method (Killick et al., 2012). Binary Segmentation was selected because it allows researchers to specify the maximum number of breakpoints and is computationally efficient for this type of longitudinal bibliometric data. Alternative methods such as the Pruned Exact Linear Time (PELT) method are more sensitive to minor fluctuations and tend to identify numerous breakpoints that may obscure broader evolutionary patterns. The analysis therefore adopted a three-phase division (Q = 2) to capture major shifts in research development, which is consistent with the broader narrative of the field’s evolution. The results of this analysis are reported in Section 3.1.

2.4 Bibliometric analysis using CiteSpace

CiteSpace 6.4.R1 was used to construct knowledge graphs for bibliometric analysis (Lai et al., 2024). Keyword normalization was performed by merging plural and singular forms (e.g., ‘behavior’ and ‘behaviors’), consolidating common synonyms (e.g., ‘theory of planned behavior’ and ‘TPB’), and unifying British and American English variants (e.g., ‘behavior’ and ‘behavior’) based on manual verification of keyword lists.

To ensure network validity and interpretability, the core parameters were configured as follows: a 1-year time slicing, and a modified g-index (k = 25) for selection criteria. The modified g-index (k = 25) was selected because lower k values produced overly sparse networks, while higher k values generated excessive connectivity that obscured cluster boundaries (Lang et al., 2023). Pruning was kept at default settings (pathfinder network scaling) to preserve complete evolutionary paths; sensitivity tests with alternative pruning strategies showed no substantive changes in major clusters. Cluster analysis used the log-likelihood ratio (LLR) algorithm for its ability to extract feature words with high cohesion and low coupling (Zhang et al., 2026). To assess cluster stability, the network was recalculated using different time slicing options (1-year and 2-year windows), and the core cluster structure remained consistent. To address semantic limitations of machine clustering, the results were manually validated by tracing back the titles and abstracts of key node literature in each core cluster, excluding invalid or semantically overlapping clusters.

The analysis focuses on keyword co-occurrence and burst detection rather than document co-citation analysis or author coupling. This choice prioritizes mapping research fronts and thematic shifts over intellectual lineages and knowledge bases, which is consistent with the review’s aim of tracing thematic evolution. Based on this, the study systematically analyzed annual publication trends, geographical distribution of research output, co-occurrence networks, thematic clusters, temporal evolution of research themes, and burst detection.

3 Results

3.1 Sample and publication trends

The final sample comprised 616 empirical journal articles. To assess the sample’s interdisciplinary coverage, the included literature was analyzed by Web of Science subject category (Figure 2). The sample was disciplinarily diverse, concentrated in “Environmental Sciences & Ecology” (73), “Food Science & Technology” (67), “Business & Economics” (55), and “Green & Sustainable Science & Technology” (42), with notable contributions also from “Behavioral Sciences” (34), “Psychology, Multidisciplinary” (25), and “Nutrition & Dietetics” (24). This disciplinary diversity provides a solid foundation for analyzing the multidimensional drivers of household food waste.

Figure 2

From 2011 to 2025, the total number of publications increased from 5 to 91, representing a more than 18-fold growth over the 15-year period (Figure 3). A change-point detection analysis was conducted using the Binary Segmentation method (Killick et al., 2012), limited to a maximum of two breakpoints. The algorithm identified change points at positions 8 and 10, corresponding to the years 2018 and 2020. These breakpoints support the division of the 15-year period into three distinct phases.

Figure 3

3.1.1 Phase 1 (2011–2017)

Annual publications remained low, with a mean of 11.4 papers per year. All studies were authored by researchers from high-income countries (HIC) and primarily focused on HIC contexts. A small number of studies by HIC authors targeting upper-middle-income countries (UMC) appeared from 2013 onward. No studies involved authors or study contexts from lower-middle-income (LMC) or low-income (LIC) countries during this period.

3.1.2 Phase 2 (2018–2020)

Annual publications increased sharply, with a mean of 45.7 papers per year. This period saw a continued increase in publications by UMC authors, who contributed to both UMC-focused and HIC-focused studies. However, LMC and LIC countries remained absent in both authorship and study contexts throughout this phase.

3.1.3 Phase 3 (2021–2025)

Annual publications reached a mean of 79.8 papers per year, peaking at 91 in 2025. From 2021 onward, HIC authors began publishing studies focused on LMC/LIC contexts, reflecting growing attention to low-resource settings. The first publications by authors from LMC/LIC contexts emerged in 2025, including three LMC/LIC → HIC studies and one LMC/LIC → UMC study. Notably, LMC/LIC → LMC/LIC publications remained absent across the entire study period.

3.2 Cross-validation using Scopus

A cross-database comparison was conducted based on the Scopus database. Adopting identical search strategies and screening criteria, a total of 775 articles were retrieved to form the Scopus validation dataset. The two datasets were not integrated. Given the differences in indexing coverage and metadata structure between the two databases, this auxiliary robustness check was only restricted to geographic publication patterns. Table 1 lists the top ten most productive countries in two databases.

The Spearman correlation coefficient of country rankings between Web of Science and Scopus was 0.82, indicating a moderate level of consistency. China ranked first in both databases, while Italy occupied the second position in both. England, Australia, Germany and Greece all featured in the top ten rankings across the two datasets. This consistency suggests that the geographic concentration pattern derived from Web of Science is not substantially affected by database selection bias.

3.3 Geographical distribution of research output

China ranked first with 102 papers and a betweenness centrality of 0.21, followed by the United States (61 papers, centrality 0.17), Italy (55 papers, centrality 0.27), England (45 papers, centrality 0.12), and Australia (44 papers, centrality 0.05). Germany ranked sixth with 35 papers (centrality 0.14), followed by Greece (28 papers, centrality 0.07), the Netherlands (26 papers, centrality 0.05), Canada (25 papers, centrality 0.01), and Japan (23 papers, centrality 0.08).

Betweenness centrality measures the proportion of shortest paths between any two nodes that pass through a given node. For a node , it is calculated as , where is the total number of shortest paths from node to node , and is the number of those paths that pass through . Higher values indicate greater bridging importance in the network (Freeman, 2000).

Countries with high betweenness centrality, including Italy (0.27), China (0.21), South Africa (0.20), the United States (0.17), and Spain (0.16), acted as structural bridges in the collaboration network. A betweenness centrality of zero, observed for countries such as Finland, indicated that these countries did not serve as bridging nodes and were more peripheral in the collaboration network.

Among the top 25 most productive countries, multiple economies belong to the Global South. China ranked first with 102 publications (centrality 0.21); South Africa produced 18 articles (centrality 0.20); Brazil and Iran each contributed 17 papers (centrality 0.10 for both); India published 15 studies (centrality 0.08); Malaysia had 11 publications (centrality 0.01); Indonesia yielded 8 papers (centrality 0.02). Although China occupied the leading position in total output, the remaining Global South nations generally recorded far lower publication counts relative to high-income Global North economies.

Country affiliation reflected the corresponding author’s address, which may not capture the full contribution of internationally collaborative teams. A country’s rank did not necessarily indicate that research focused on its domestic food waste context; Section 3.1 provided a separate analysis of research focus by country context. Betweenness centrality measured structural position in the collaboration network, not research quality or impact.

High-income countries dominated in both publication volume and network centrality, while most Global South countries had lower output and more peripheral network positions. This pattern was consistent with the cross-database validation reported in Section 3.2.

3.4 Keyword co-occurrence and thematic priorities

The keyword co-occurrence network was constructed using CiteSpace. The network contained 500 nodes and 1,114 links, with a network density of 0.0089. Table 2 presents the top 25 high-frequency keywords along with their frequency and betweenness centrality. Food waste ranked first with 302 occurrences, followed by behavior (189), consumer (137), attitudes (124), and household food waste (114).

Centrality analysis. Betweenness centrality reflects the capability of a keyword to act as an intermediary linking different research clusters. Among the top 25 keywords, China (0.20), anaerobic digestion (0.15), household (0.12), consumer (0.10), and reduction (0.10) exhibited the highest betweenness centrality, indicating their critical bridging roles in connecting diverse research themes. By contrast, terms such as management (0) and sustainability (0) had centrality values of zero, meaning they were relatively confined within individual research domains without cross-cluster linkage.

Thematic priorities across different income contexts. Table 3 compares the ranking of the top 25 keywords between HIC and UMC focused studies and LMC and LIC focused studies. Several clear thematic differences can be identified.

First, several HIC and UMC core keywords were not ranked within the top 25 in LMC and LIC research, including generation, management, anaerobic digestion, sustainability, theory of planned behavior, and quantification. The absence from the LMC and LIC top 25 does not equate to a complete lack of relevant research, as author-supplied keywords are influenced by journal norms and academic conventions. Nevertheless, such systematic omission indicated evident differences in explicit research priorities between the two contexts.

Second, overlapping keywords such as consumer behavior (ranked 12th in LMC&LIC, down from 3rd), attitudes (14th, down from 4th), and behavior (10th, down from 5th) appeared lower in LMC and LIC rankings, suggesting a relatively weaker emphasis on individual behavioral and psychological perspectives in lower-income country settings.

Third, several macro and development-related keywords rose markedly in LMC and LIC rankings. Specifically, China ranked 6th (up from 9th), food loss ranked 8th (up from 19th), food safety ranked 9th (up from 25th), and determinants ranked 9th (up from 10th). In contrast, behavior ranked 10th (down from 5th). These higher positions highlighted that research in lower-income settings placed greater emphasis on food security and macro-systemic sustainability issues.

Interpretation of these ranking differences requires appropriate methodological considerations. The smaller literature base for LMC and LIC studies may cause valuable niche topics to be excluded from the top 25 keyword list due to lower occurrence frequency. Additionally, author-assigned keywords cannot capture the full scope of each study. A missing keyword does not imply a lack of relevant research. This comparison only reflected different thematic priorities instead of absolute research gaps. Hence, Table 3 illustrates differing research priorities between the two contexts, rather than proving the complete absence of certain research topics. Further quantitative validation of these differences is provided in Section 3.5.3 via clique and k-core analysis.

3.5 Keyword clustering analysis

3.5.1 Thematic clusters and timeline analysis

Figure 4 presents the keyword clustering map of household food waste research, identifying 10 distinct clusters with high modularity (Q = 0.783) and weighted mean silhouette scores (S = 0.9079), confirming well-structured and internally consistent thematic groups. The harmonic mean of Q and S reached 0.8409, further validating the high reliability of the clustering outcomes.

Figure 4

The clusters were spatially and thematically segregated into two broad research domains. One domain was dominated by engineering and technology-focused themes, including Cluster #2 biogas, Cluster #4 bio-oil, Cluster #9 biomass, which centered on waste treatment, biomass utilization and resource recovery technologies such as anaerobic digestion. The other domain comprised social science, consumption and household-oriented themes, covering Cluster #0 sustainability, Cluster #1 fruit and vegetable purchases, Cluster #3 food security, Cluster #5 covid-19, Cluster #6 europe, Cluster #7 lifestyle and Cluster #8 household food waste. This branch mainly focused on consumer purchasing behavior, household-level food waste generation, public health shocks, regional research disparities, lifestyle factors and sustainable food system governance. The clear spatial separation between these two domains, together with the low inter-group connectivity implied by the sparse network density of 0.0089, indicated that the technical resource recovery stream and the social-behavioral household food waste research stream evolved largely independently, with limited integration between waste end-treatment management and source-side food waste reduction perspectives.

Following the three-phase periodization established in Section 3.1 (2011–2017, 2018–2020, 2021–2025), Figure 5 presents a keyword timeline view spanning 2011–2025, with high modularity (Q = 0.783) and weighted mean silhouette score (S = 0.9079), indicating robust clustering. The harmonic mean of Q and S reached 0.8409, further verifying the reliability of the keyword clustering results.

Figure 5

Research from 2011 to 2017 laid the foundation for food waste research. In the early part of this timeframe, studies mainly revolved around anaerobic digestion, biogas, municipal solid waste, biomass, focusing on waste resource conversion and energy recovery. Other emerging topics included food supply chain and food security. Later within this period, research extended to social and policy fields, with attitude, policy, barrier, China turning into popular research hotspots, and sustainability was gradually incorporated into relevant research.

From 2018 to 2020, academic research shifted to household food waste governance and empirical analysis. Scholars paid close attention to driver, quantification, intervention, self-reporting, exploring the generation mechanism of household food waste and designing corresponding waste reduction schemes. Meanwhile, supply chain research related to restaurants also gained increasing academic interest.

The years spanning 2021 to 2025 coincided with the COVID-19 pandemic, during which research gradually moved toward behavioral and household characteristic analysis. Drawing on core keywords from the dataset including domestic waste, healthy, extended theory, household income, scholars analyzed residents’ food waste behaviors and the differentiated influences of household attributes on food waste. Cross-regional research represented by Europe and Italy also enriched the global research system of food loss and waste.

3.5.2 Research fronts and burst detection

Following the three-phase periodization established in Section 3.1 (2011–2017, 2018–2020, 2021–2025), Table 4 lists the top 25 keywords with citation bursts alongside their burst strength and active time windows. Across all keywords, “generation” registered the highest burst strength at 5.90 (2018–2019), followed by “methodology” (4.44, 2019–2021) and “city” (3.92, 2023–2025). “Food waste prevention” featured the longest continuous burst spanning 2011 to 2020, while “biogas” maintained sustained research attention from 2011 to 2019; their long durations revealed persistent academic interest in household food waste reduction and waste-to-energy recycling technologies.

3.5.2.1 Early bursts (2011–2017)

This initial stage centered on environmental engineering and waste resource recovery research. Representative burst keywords included “anaerobic digestion” (4.00, 2011–2015), “organic waste” (2.01, 2011–2012), “greenhouse gas emissions” (1.90, 2011–2012), and “bioethanol” (1.87, 2014–2017). Both “biogas” (2.40, 2011–2019) and “food waste prevention” (2.22, 2011–2020) initiated their burst status in this period and lasted into subsequent research phases. Toward the later years of this stage, emerging burst keywords covered “waste prevention” (2.27, 2015–2018), “impact” (2.13, 2015–2017), “chain” (1.97, 2015–2019), “consumer behavior” (2.61, 2016–2020), “loss” (2.18, 2017–2019) and “system” (3.12, 2017–2018), gradually shifting research focus from technical treatment to behavioral and supply chain governance.

3.5.2.2 Middle bursts (2018–2020)

Research further shifted toward food waste generation mechanism analysis and multidisciplinary methodological exploration. The keyword “generation” reached the global peak burst strength of 5.90 from 2018 to 2019. Other prominent burst terms in this phase included “sewage sludge” (2.87, 2018–2020) and “anaerobic co-digestion” (2.16, 2019–2020). “Methodology” (4.44, 2019–2021) started its citation burst at the end of this stage and continued into the most recent research period, reflecting growing demand for standardized empirical research frameworks in food waste studies. Several keywords such as “household waste,” “prevention,” “fruit” and “perception” also triggered burst signals starting around 2020, laying a research foundation for the subsequent post-pandemic stage.

3.5.2.3 Recent bursts (2021–2025)

Research frontiers in this phase focused on household micro-level behavior, urban governance and behavioral theoretical application, with the keyword ‘covid-19’ appearing in publications from 2021 to 2023. The keyword “covid-19” obtained a burst strength of 2.57 from 2021 to 2023. Emerging hot topics included “rural household” (2.08, 2021–2022) and “intention” (3.13, 2023–2025). From 2022 onwards, behavioral theory and urban-scale research dominated emerging trends: “extended theory” (3.76, 2022–2025) and “city” (3.92, 2023–2025) maintained strong citation bursts until 2025, representing the latest research directions of food waste governance.

Whether the Theory of Planned Behavior (TPB) remains the dominant framework in household food waste research has been debated. To examine the framework’s actual attention over time, the annual share of TPB-related articles was calculated.

As shown in Figure 6, TPB-related articles first appeared in 2014 (5.9%) and peaked at 30.4% in 2021, with notable fluctuations (rising to 21.6% in 2016, dropping to 7.0% in 2017) before declining to 20.9% by 2025. The late bursts of these constructs (2023–2025) suggested renewed attention to specific TPB dimensions even as the framework’s overall share declined from its peak. A comparison with Social Practice Theory was attempted, but the framework appeared in too few articles (n < 10) to support a meaningful longitudinal analysis. This absence itself suggested that alternative frameworks remained marginal in household food waste research compared to TPB.

Figure 6

3.5.3 Network metrics (clique and K-core analysis)

The keyword clustering analysis (Section 3.5.1) showed a qualitative separation between behavioral and engineering research themes, and the keyword ranking comparison (Table 3) indicated thematic differences between HIC/UMC and LMC/LIC contexts. However, these observations were descriptive and did not quantify the structural relationships between thematic clusters. To provide quantitative, network-based evidence for thematic fragmentation and agenda separation, clique and k-core analyses were conducted using the igraph package in R.

The keyword co-occurrence network for this analysis contained 268 nodes and 992 edges, with a network density of 0.0277 (Table 5). These nodes were keywords with a minimum frequency of two, selected to ensure a stable network structure while including moderately frequent terms that bridge thematic clusters.

3.5.3.1 Clique analysis

A total of 1,255 cliques of size three (triangles) were identified, and the maximum clique size was 7. This indicated the existence of a strongly bonded group of 7 keywords that frequently co-occurred in the literature. Four behavioral keywords (including “theory of planned behavior,” “attitude,” “consumer behavior,” and “tpb”) and three engineering keywords (including “anaerobic digestion,” “biogas,” and “waste management”) were identified in the network. The number of cross-clique edges between behavioral and engineering keywords was only 2, accounting for 0.20% of the total network edges. Mixed cliques containing both behavioral and engineering keywords numbered 2, representing only 0.16% of all 1,255 triangles. These metrics provided quantitative evidence of thematic fragmentation between behavioral reduction research and end-of-pipe management research.

3.5.3.2 K-core analysis

The k-core analysis revealed a maximum k-core value of 8, containing 13 core keywords (k = 8). The distribution of k-core values showed that the largest layer (k = 8) contained 13 keywords, followed by 12 keywords at k = 6, 9 at k = 5, 12 at k = 4, 1 at k = 3, and 3 at k = 2 (Figure 7). These core keywords included “household,” “food waste,” “consumer behavior,” “households,” “food security,” “COVID-19,” “survey,” “household food waste,” “sustainability,” “consumers,” “waste,” “food waste behavior,” and “consumption.” Notably, engineering-oriented keywords such as “anaerobic digestion,” “biogas,” and “waste management” were absent from this core layer. This observation, combined with the extremely low cross-clique connectivity (only 2 edges, 0.20%) and the minimal number of mixed cliques (2 out of 1,255 triangles, 0.16%) reported above, suggests that behavioral reduction research and end-of-pipe engineering management research may constitute structurally fragmented thematic clusters.

Figure 7

3.5.3.3 North-South agenda separation

Based on the keyword classifications in Table 3, keywords associated with HIC/UMC contexts (e.g., “theory of planned behavior,” “tpb,” “attitude,” “consumer behavior”) and those associated with LMC/LIC contexts (e.g., “food security,” “rural household,” “storage,” “nutrition”) were identified. The direct connections between these two groups in the co-occurrence network totaled only 2 edges (both connecting “food security” with “consumer behavior” and “household”). This remarkably low connectivity indicated that HIC-focused and LIC-focused research agendas formed structurally separate clusters.

4 Discussion

Research output expansion did not equate to balanced global knowledge construction. The landscape of household food waste research was shaped by researchers’ geographic locations, practical demands and adopted theoretical paradigms, which constituted an inherent structural feature of global academic development rather than inappropriate academic malpractice.

4.1 Geographic imbalance: structural constraints on knowledge applicability

Geographic concentration of research output does not stem from methodological limitations. Consistent country rankings from Scopus and supplementary searches of regional databases validate this stable distribution pattern. Latin American and African regions show scarce indexed publications in household food waste research. Additional CNKI literature does not change China’s leading research status. These findings support a unified conclusion. Global indexed studies on household food waste are concentrated in the indexed literature within a small number of high-income countries. Such distribution represents an authentic feature of contemporary global academic production.

Publication disparities generate critical challenges beyond quantitative gaps. Current research structures weaken the cross regional applicability of global academic findings for Global South contexts. Low- and middle-income countries feature diverse economic foundations, livelihood modes and dietary cultures. Sub Saharan African peri urban areas suffer food spoilage mainly from unstable power supply and insufficient cold storage facilities (Aworh, 2021; Sipho and Tilahun, 2020). Rapidly industrializing Southeast Asian cities encounter waste issues driven by evolving consumption patterns and expanding supermarket coverage (Arfanuzzaman and Dahiya, 2019). High income research frameworks rely on mature infrastructural conditions. Their theoretical models, behavioral interpretations and intervention schemes cannot fully adapt to diversified Global South scenarios. Public education on portion control fails to mitigate spoilage caused by unstable power facilities (Masinga and Madzivhandila, 2023; Snyder et al., 2024). Psychological indicators centered on guilt show limited explanatory power in regions that prioritize daily food security (Carolan, 2013; Myers et al., 2020).

Such structural mismatches hinder the implementation of SDG Target 12.3. This global target requires effective governance practices across all economies (Li et al., 2026). Existing policy evidence overly depends on empirical findings from wealthy regions. Local academic studies led by Global South scholars serve as essential analytical tools. These context specific investigations identify localized waste drivers covering infrastructural, cultural and economic factors (Awino and Apitz, 2024). They also support the formulation of targeted and practical mitigation strategies. The betweenness centrality data reported in Section 3.3 show near-zero values for most Global South nations. These numerical results demonstrate peripheral academic status in publication volume and agenda setting capacity.

This study does not deny the existence of local non-indexed research. Regional reports and non-core journals contain abundant valuable grassroots findings. The layered database design of this research fully considers inherent indexing limitations. Mainstream indexed academic discussions show clear inclination toward high income contexts. Global food waste research and corresponding policy guidance largely serve the needs of developed economies. Research skew correction is not merely an issue of academic equity. It provides essential localized evidence to promote inclusive SDG 12.3 progress across all regions worldwide.

4.2 Thematic evolution: from individual psychology to systemic constraints

Household food waste research has changed its focus over time. An examination of keyword distribution reveals a clear intellectual trajectory. Early research prioritized waste to energy technologies such as anaerobic digestion, biogas production, and fermentation. Behavioral psychology then emerged as a mainstream direction, with studies drawing heavily on the Theory of Planned Behavior and its core constructs of attitudes, intentions, guilt, and social norms. Recent trends show a different pattern. Extended theory, motivation, perception, nutrition, household heterogeneity, and intervention design have gained growing attention. These keywords signal a departure from the narrow boundaries of rational choice models (Kannengiesser and Gero, 2019).

This shift matters because it reflects a fundamental change in how the field understands the drivers of food waste. The early focus on individual psychology assumed that waste is primarily a matter of personal choice. If people had better attitudes and stronger intentions, the logic went, they would waste less (Al Mamun et al., 2024; Schanes et al., 2018). But this assumption has not held up well against empirical evidence. The intention behavior gap is real and persistent (Garza, 2018). People who sincerely want to waste less often fail to do so. This failure is not a personal shortcoming (Chia et al., 2024a). It is a sign that the model is incomplete.

The recent turn toward systemic, infrastructural, and cultural explanations acknowledges that individual behavior is always embedded in larger contexts. A person cannot reduce food waste if their electricity supply is unreliable, if their refrigerator breaks down frequently, or if their local market only sells food in large packages. These are not psychological problems. They are structural problems. The field is slowly recognizing that effective waste reduction requires changing not just minds but also the material conditions in which people live (Parvatiyar and Sheth, 2023).

The evolution of research themes also has implications for how knowledge is produced and shared. Researchers working in contexts with reliable infrastructure can afford to focus on the finer grained question of individual behavior. Researchers facing infrastructure shortages are logically drawn to themes of food security, storage, and postharvest loss. Neither approach is superior. Each responds to what are most salient in its own context. The problem arises when one tradition assumes its findings apply universally. A behavioral intervention that works in a high-income setting may fail entirely in a low-income setting, not because the intervention is bad, but because the underlying conditions are different (Rizvi et al., 2024).

No single research tradition holds absolute superiority. The future of the field lies not in replacing one orthodoxy with another but in building an integrated science. Such integration must address how individual behavior and systemic constraints interact across diverse economic and infrastructural backgrounds. This requires contributions from both psychological and engineering perspectives, both quantitative metrics and qualitative insights, both Northern and Southern researchers. The evolution of keywords is not merely a bibliometric trend. It maps where the field needs to go next.

4.3 Theoretical transition: from TPB’S dominance to theoretical diversification

The Theory of Planned Behavior (TPB) has gradually lost its absolute dominance in household food waste research after over a decade of widespread application. Its publication proportion peaked at 30.4% in 2021 and declined steadily to 20.9% in 2025. Core TPB constructs including behavioral intention and planned behavior have maintained steady publication output from 2023 to 2025. This trend indicates ongoing theoretical enrichment through emotional, nutritional and contextual variables instead of comprehensive theoretical abandonment. TPB is being decentered within the theoretical system rather than completely replaced.

The declining dominance of TPB has been attributed to inherent limitations of rational choice models, particularly the prevalent intention behavior gap. TPB theorizes that attitudes and subjective norms shape individual intentions and further translate into practical waste reduction actions (Barnett, 2024; Yeow and Loo, 2022). Empirical evidence does not consistently support this assumption, as positive waste reduction intentions rarely convert into consistent behavioral changes (Chakraborty and Mattila, 2025; Olthof, 2012). The gap expands significantly in underdeveloped regions where unstable power supply and inadequate cold chains dominate food waste generation.

This discovery complemented behavioral science debates on rational model boundaries. Multiple alternative theoretical frameworks offered expanded analytical perspectives. However, the present study did not test the causal direction of this theoretical transition; it may also reflect publication preferences, journal scopes, or research training traditions rather than a deliberate shift away from TPB.

A comparison with Social Practice Theory was conducted to contextualize the declining share of TPB. This framework appeared in too few articles (n < 10) to support a meaningful longitudinal analysis. This absence itself was meaningful. It suggested that alternative behavioral frameworks remained marginal in household food waste research, even as TPB’s dominance had eroded. This posed a puzzle: if TPB was declining, why were alternative frameworks not rising to replace it? The answer may lie in the different analytical levels at which these frameworks operate.

TPB is a micro-level rational choice model. It explains behavior through individual attitudes, subjective norms, and perceived behavioral control (Malau et al., 2025). It assumes deliberate, reasoned decision-making. Social Practice Theory, by contrast, is a macro-level framework. It situates behavior within collective routines, material arrangements, and shared meanings, emphasizing habitual and socially embedded action rather than conscious intention (Evans, 2012; Shove et al., 2012; Heidenstrøm, 2022; Shove et al., 2017). These two frameworks are not direct substitutes. They address different questions and operate at different scales. This may explain why the decline of TPB has not been accompanied by a corresponding rise in Social Practice Theory. The field had not abandoned micro-level analysis; it had sought to enrich it.

TPB’s declining dominance was associated with a growing recognition that food waste behavior was not always the outcome of rational calculation. Many wasteful acts were habitual, automatic, or driven by emotional responses such as guilt, worry, and the desire to provide for one’s family (Watson and Meah, 2012). Portion sizes, package formats, and refrigerator organization were often shaped by social norms and material infrastructure rather than conscious intention (Evans, 2012; Waitt and Phillips, 2016). A micro-level framework such as TPB, which assumed rational deliberation, may not fully capture these non-rational and context-dependent drivers. The emerging diversification observed in this study, including keywords such as “anticipated guilt,” “motivation,” “perception,” and “extended theory,” can be interpreted as scholars supplementing TPB with emotional and contextual variables. Rather than being replaced by a new orthodoxy, TPB was being expanded through the incorporation of these supplementary constructs. This pattern appeared consistent with theoretical enrichment rather than paradigm replacement.

The theoretical diversification observed in this study points to the field’s gradual transition toward pluralistic analytical paradigms. For future research, this suggests that theoretical frameworks should be selected according to the specific research question and context, rather than defaulting to a single model. Multi-level approaches that integrate micro-level psychological variables with macro-level structural conditions may offer more comprehensive explanations of household food waste behavior. This does not imply that TPB should be abandoned, but rather that it should be situated within a broader theoretical toolkit. Effective interventions may require multi-dimensional optimization involving habitual behavior reshaping, emotional antecedent regulation and contextual choice architecture redesign, as suggested by recent intervention research (Zhang et al., 2025). These dimensions align with the emerging keyword patterns observed in this study, including motivation, perception, and extended theory.

4.4 Disciplinary fragmentation: a structural divide with practical consequences

Clique analysis and k-core calculation results quantitatively confirm long-standing disciplinary segregation in food waste research, which has only been described qualitatively in previous studies. Cross-domain keyword connections and mixed research cliques between behavioral reduction studies and terminal waste treatment research account for merely 0.20 and 0.16% of the total network structure, respectively. Core anaerobic digestion keywords are excluded from the highest k-core layer. Such segregation constitutes a fundamental structural feature of this research field, manifested in citation tendencies, journal scope positioning and keyword co-occurrence patterns (Harvey et al., 2022).

Disciplinary fragmentation generates tangible practical obstacles for food waste governance. Behavioral intervention strategies targeting personal habits show limited efficacy in regions suffering from infrastructural defects. Public education on portion control cannot reduce food spoilage caused by unstable power supply (Awino and Apitz, 2024). Conversely, terminal engineering solutions such as anaerobic digester construction and biogas efficiency optimization ignore the potential of source reduction (Ammar et al., 2026). End of pipe treatment alleviates landfill pressure but fails to eliminate waste generation fundamentally. SDG Target 12.3 requires substantial reduction in per capita food waste, which cannot be achieved solely through terminal governance. Source reduction serves as the core solution that integrates behavioral transformation and infrastructural support (Urugo et al., 2024).

Research fragmentation also reshapes disciplinary funding and evaluation mechanisms. Funding agencies divide research support by disciplinary boundaries. Social science grants exclusively support behavioral research while engineering funds focus on terminal waste treatment technologies. Journal scopes and citation networks further strengthen disciplinary homogeneity (Harvey et al., 2022). Such segregation originates from institutional academic incentive designs rather than artificial exclusion. Cross disciplinary researchers thus face higher publication and evaluation barriers, which further reinforce self-perpetuating disciplinary silos (Basore et al., 2025).

Prior studies elaborated disciplinary segregation primarily through qualitative discussion. The present study quantified fragmentation degrees to support longitudinal dynamic tracking. Future research could explore temporal changes in disciplinary gaps and evaluate the effectiveness of cross-disciplinary funding programs and journal special issues in breaking academic barriers. Although the present study confirmed the existence of disciplinary fragmentation, root causes remained open to further exploration. Insufficient granular household waste data hindered accurate attribution of waste drivers among behavioral factors, infrastructural conditions and systemic mechanisms. Policy interventions could play a role in enabling such data infrastructure, as shown in European cases where waste management policies facilitated industrial symbiosis and resource recovery (Costa et al., 2010). Funding rules, journal positioning and academic career incentives also provided plausible explanatory dimensions. The present study did not test the causal direction of this fragmentation; it may also reflect funding structures, journal scopes, or career incentives rather than deliberate exclusion. These competing explanations warrant further empirical investigation. The present study established essential quantitative baselines for subsequent causal exploration and academic debate.

4.5 Limitations

First, this analysis relies on author provided keywords. Indexed keyword information cannot fully capture all implicit research content within full texts. This is an unavoidable inherent limitation of keyword based bibliometrics. Second, the exclusion of non-English literature is forced by inconsistent metadata standards across non-English journals, inevitably introducing mild linguistic indexing bias despite supplementary regional database inspection. Third, this research adopts the corresponding author’s affiliation to classify national attributes, failing to capture detailed contribution distribution of multi author international teams beyond correspondence authors. Fourth, all claims about ‘global’ knowledge production in this study are conditional on the coverage of Web of Science, CNKI, SciELO, and AJOL (with zero records retrieved from AJOL). Scopus verification is limited to national publication ranking only. Thematic structure and keyword burst cannot be cross validated due to inconsistent metadata formats between databases. Non-English and non-indexed research from the Global South may alter the observed patterns. All above limitations are objectively recognized to avoid overgeneralization of the core research conclusions.

4.6 Practical implications

This study’s core novelty lies in quantifying widely noted North-South disparities with measurable metrics, instead of repeating known descriptive observations.

The findings deliver practical value for five groups. Global South researchers may use these data to argue for better research visibility in grant and publication applications (Panait et al., 2025). Journal editors could optimize keyword classification to avoid implicit preference for Western behavioral theories. Funding agencies are suggested to launch targeted grants for developing-country food waste research. Policymakers need context-specific SDG 12.3 indicators to avoid high-income biased policies. Bibliometric scholars can adopt our layered database and network-based scheme to assess thematic fragmentation and indexing bias in future work.

This paper does not label uneven knowledge distribution as unjust but provides solid quantitative evidence of structural asymmetry in global research. As food waste research remains dominated by high-income contexts, gradual institutional adjustments are required to advance inclusive SDG 12.3 progress.

5 Conclusion

This bibliometric scoping review analyzed 616 empirical papers (2011–2025) from Web of Science, CNKI, SciELO, and AJOL, with cross-validation using Scopus. Three core findings were summarized below.

First, household food waste research clusters geographically in high- and upper-middle-income nations. Global South scholars lack independent agenda-setting capacity, and thematic priorities diverge distinctly across income contexts; merely two network links connect high-income and low-income research keywords.

Second, the Theory of Planned Behavior showed a declining relative share. After peaking at 30.4% in 2021, its publication share fell to 20.9% in 2025. During the same period, studies increasingly integrated emotion, nutrition and contextual factors alongside classic TPB constructs. Relevant research has evolved sequentially from food waste engineering, through short-term behavioral intention analysis, to diversified macro-contextual interventions, and further toward long-term value-driven and lifestyle-oriented approaches.

Third, behavioral and engineering research streams remained largely disconnected, with cross-domain edges and mixed cliques accounting for only 0.20 and 0.16% of total network structures, respectively.

Advancing SDG 12.3 requires gradual, evidence-based improvement rather than systemic overhaul. This study delivered quantifiable metrics and a multi-layered validation framework that highlights global research asymmetry, helping the field move toward greater inclusiveness.

References

• 1

  AgarwalA. (2018). A Pioneering Tactic for Sustainable Waste Management: Design of a Cost-Effective Recycling Machine. Les Ulis: EDP Sciences, 3003.

  • Google Scholar

• 2

  Al MamunA.MaY.Nurul Hassan RezaM.AhmadJ.Wan Mohd Hirwani WanH.LiliZ. (2024). Predicting attitude and intention to reduce food waste using the environmental values-beliefs-norms model and the theory of planned behavior. Food Qual. Prefer.120:105247. doi: 10.1016/j.foodqual.2024.105247

  • CrossRef
  • Google Scholar

• 3

  AlkindiM.HaddadA.SaiduM. B.AbushammalaH. (2026). Exploring habitual and emotional determinants of food waste behavior among multicultural consumers in Dubai, United Arab Emirates. Front. Sustain. Food Syst.10:1736201. doi: 10.3389/fsufs.2026.1736201

  • CrossRef
  • Google Scholar

• 4

  AmmarE. E.AbdelmaksoudA. A.FaragallahM. H.SherifS.AlwakeelS. A.ElshahatA.et al. (2026). Anaerobic digestion of organic waste for bioenergy production: optimization, performance, and sustainability—a review. Int. J. Energy Res. doi: 10.1155/er/3449272

  • CrossRef
  • Google Scholar

• 5

  ArdraS.BaruaM. K. (2022). Halving food waste generation by 2030: the challenges and strategies of monitoring UN sustainable development goal target 12.3. J. Clean. Prod.380:135042. doi: 10.1016/j.jclepro.2022.135042

  • CrossRef
  • Google Scholar

• 6

  ArfanuzzamanM.DahiyaB. (2019). Sustainable urbanization in Southeast Asia and beyond: challenges of population growth, land use change, and environmental health. Growth Chang.50, 725–744. doi: 10.1111/grow.12297

  • CrossRef
  • Google Scholar

• 7

  AwinoF. B.ApitzS. E. (2024). Solid waste management in the context of the waste hierarchy and circular economy frameworks: an international critical review. Integr. Environ. Assess. Manag.20, 9–35. doi: 10.1002/ieam.4774,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8

  AworhO. C. (2021). Food safety issues in fresh produce supply chain with particular reference to sub-Saharan Africa. Food Control123:107737. doi: 10.1016/j.foodcont.2020.107737

  • CrossRef
  • Google Scholar

• 9

  BarnettE. (2024). Leftovers: a History of Food Waste and Preservation. London, UK: Bloomsbury Publishing.

  • Google Scholar

• 10

  BasoreC.MohammedS.TirrellB.MarhefkaJ.HamiltonK.ZhangT.et al. (2025). Post-award grant collaboration: facilitators and hindrances to cross-disciplinary authorship. Front. Educ.9:1400595. doi: 10.3389/feduc.2024.1400595

  • CrossRef
  • Google Scholar

• 11

  CarolanM. S. (2013). Reclaiming Food Security. Abingdon, Oxon: Routledge.

  • Google Scholar

• 12

  ChakrabortyD.MattilaA. S. (2025). The effect of self-conscious emotions of guilt and regret on consumers' intentions to reduce food waste. J. Sustain. Tour.33, 1871–1889. doi: 10.1080/09669582.2024.2406532

  • CrossRef
  • Google Scholar

• 13

  ChepelievM.GattoA. (2024). Global food loss and waste estimates show increasing nutritional and environmental pressures. Nat. Food5, 136–147. doi: 10.1038/s43016-023-00915-6,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14

  ChiaD.Choo YapC.Kheireddine ArouaM.Ti GewL. (2024a). Theoretical approaches to food waste behaviours: an overview. Appl. Environ. Res.46. doi: 10.35762/AER.2024017

  • CrossRef
  • Google Scholar

• 15

  ChiaD.YapC. C.WuS. L.BerezinaE.ArouaM. K.GewL. T. (2024b). A systematic review of country-specific drivers and barriers to household food waste reduction and prevention. Waste Manag. Res.42, 459–475. doi: 10.1177/0734242X231187559,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 16

  CostaI.MassardG.AgarwalA. (2010). Waste management policies for industrial symbiosis development: case studies in European countries. J. Clean. Prod.18, 815–822. doi: 10.1016/j.jclepro.2009.12.019

  • CrossRef
  • Google Scholar

• 17

  De Visser-AmundsonA. (2022). A multi-stakeholder partnership to fight food waste in the hospitality industry: a contribution to the United Nations sustainable development goals 12 and 17. J. Sustain. Tour.30, 2448–2475. doi: 10.1080/09669582.2020.1849232

  • CrossRef
  • Google Scholar

• 18

  DongY.WuZ. (2026). “Malignancies: a multi-database bibliometric analysis and cross-validation study,” in Challenges and opportunities in tumor metabolomics, eds. ZhengyanC.ChunmiaoC.HongqiT.LeiY.ChengK. (London, UK: Taylor & Francis).

  • Google Scholar

• 19

  DonthuN.KumarS.MukherjeeD.PandeyN.LimW. M. (2021). How to conduct a bibliometric analysis: an overview and guidelines. J. Bus. Res.133, 285–296. doi: 10.1016/j.jbusres.2021.04.070

  • CrossRef
  • Google Scholar

• 20

  EvansD. (2012). Beyond the throwaway society: ordinary domestic practice and a sociological approach to household food waste. Sociology46, 41–56. doi: 10.1177/0038038511416150

  • CrossRef
  • Google Scholar

• 21

  FantomN. J.SerajuddinU. (2016). The World Bank's Classification of Countries by Income. St. Louis: Federal Reserve Bank of St. Louis.

  • Google Scholar

• 22

  FarooqR. (2023). Knowledge management and performance: a bibliometric analysis based on Scopus and WOS data (1988–2021). J. Knowl. Manag.27, 1948–1991. doi: 10.1108/JKM-06-2022-0443

  • CrossRef
  • Google Scholar

• 23

  Farr-WhartonG.FothM.ChoiJ. H.-J. (2014). Identifying factors that promote consumer behaviours causing expired domestic food waste. J. Consum. Behav.13, 393–402. doi: 10.1002/cb.1488

  • CrossRef
  • Google Scholar

• 24

  FraccasciaL.GiannoccaroI.AgarwalA.HansenE. G. (2021). Business models for the circular economy: empirical advances and future directions. Bus. Strateg. Environ.30, 2741–2744. doi: 10.1002/bse.2896

  • CrossRef
  • Google Scholar

• 25

  FreemanL. C. (2000). Centrality in social networks. A conceptual clarification. Polít. Soc.33, 131–148.

  • Google Scholar

• 26

  GarzaK. B. (2018). How do we close the intention–behavior gap?J. Am. Pharm. Assoc.58, 240–241. doi: 10.1016/j.japh.2018.04.008,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27

  GauffriauM. (2021). Counting methods introduced into the bibliometric research literature 1970–2018: a review. Quant. Sci. Stud.2, 932–975. doi: 10.1162/qss_a_00141

  • CrossRef
  • Google Scholar

• 28

  González-SantanaR. A.BlesaJ.FrígolaA.EsteveM. J. (2022). Dimensions of household food waste focused on family and consumers. Crit. Rev. Food Sci. Nutr.62, 2342–2354. doi: 10.1080/10408398.2020.1853033,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29

  HarshM.BalR.WeryhaA.WhatleyJ.OnuC. C.NegroL. M. (2021). Mapping computer science research in Africa: using academic networking sites for assessing research activity. Scientometrics126, 305–334. doi: 10.1007/s11192-020-03727-8

  • CrossRef
  • Google Scholar

• 30

  HarveyJ.Nica-AvramG.SmithM.HibbertS.MuthuriJ. (2022). Mapping the landscape of consumer food waste. Appetite168:105702. doi: 10.1016/j.appet.2021.105702,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31

  HeidenstrømN. (2022). The utility of social practice theory in risk research. J. Risk Res.25, 236–251. doi: 10.1080/13669877.2021.1936608

  • CrossRef
  • Google Scholar

• 32

  HuW.XieD.LeY.FuN.ZhangJ.YinS.et al. (2024). Evolution of food trade networks from a comparative perspective: an examination of China, the United States, Russia, the European Union, and African countries. Foods13:2897. doi: 10.3390/foods13182897,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33

  KalayciogluI.RiouxB.BriardJ. N.NehmeA.ToumaL.DansereauB.et al. (2023). Inter-rater reliability of risk of bias tools for non-randomized studies. Syst. Rev.12, 227–228. doi: 10.1186/s13643-023-02389-w,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34

  KannengiesserU.GeroJ. S. (2019). Design thinking, fast and slow: a framework for Kahneman’s dual-system theory in design. Des. Sci.5:e10. doi: 10.1017/dsj.2019.9

  • CrossRef
  • Google Scholar

• 35

  KillickR.FearnheadP.EckleyI. A. (2012). Optimal detection of change points with a linear computational cost. J. Am. Stat. Assoc.107, 1590–1598. doi: 10.1080/01621459.2012.737745

  • CrossRef
  • Google Scholar

• 36

  LaiY.JiangJ.ZhangH.GongK. (2024). Bibliometric analysis of curcumin based on CiteSpace: landscapes, hotspots, and Frontiers. Drug Des. Devel. Ther.18, 5743–5758. doi: 10.2147/DDDT.S494758,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 37

  LangX.GuoJ.LiY.YangF.FengX. (2023). A bibliometric analysis of diagnosis related groups from 2013 to 2022. Risk Manag. Healthc. Policy.16, 1215–1228. doi: 10.2147/RMHP.S417672,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 38

  LiT.QiaoG.CongN.KohK. (2026). Understanding “anti-food waste” efforts in China: a policy text analysis based on the period 2009–2021. Environ. Dev. Sustain. doi: 10.1007/s10668-026-07652-3

  • CrossRef
  • Google Scholar

• 39

  LiaoC.ZhaoD.ZhangS.ChenL. (2018). Determinants and the moderating effect of perceived policy effectiveness on residents’ separation intention for rural household solid waste. Int. J. Environ. Res. Public Health15:726. doi: 10.3390/ijerph15040726,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 40

  LinB.GuanC. (2021). Determinants of household food waste reduction intention in China: the role of perceived government control. J. Environ. Manag.299:113577. doi: 10.1016/j.jenvman.2021.113577,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 41

  LoxtonM.TruskettR.ScarfB.SindoneL.BaldryG.ZhaoY. (2020). Consumer behaviour during crises: preliminary research on how coronavirus has manifested consumer panic buying, herd mentality, changing discretionary spending and the role of the media in influencing behaviour. J. Risk Financ. Manag.13:166. doi: 10.3390/jrfm13080166

  • CrossRef
  • Google Scholar

• 42

  MalauL. R. E.KhaliqiM.YulniT.RambeK. R.NovandaR. R.WibowoR. P. (2025). Why we waste: a meta-analysis of food waste behavior using the theory of planned behavior (TPB). Sustain. Futures10:101421. doi: 10.1016/j.sftr.2025.101421

  • CrossRef
  • Google Scholar

• 43

  ManzoorS.FayazU.DarA. H.DashK. K.ShamsR.BashirI.et al. (2024). Sustainable development goals through reducing food loss and food waste: a comprehensive review. Future Foods9:100362. doi: 10.1016/j.fufo.2024.100362

  • CrossRef
  • Google Scholar

• 44

  MasingaF.MadzivhandilaT. (2023). Loadshedding impact on food spoilage: an analysis of household experiences in South Africa. Afr. J. Gov. Dev.12, 182–197. doi: 10.36369/2616-9045/2023/v12i2a11

  • CrossRef
  • Google Scholar

• 45

  MorkunasM.WangY.WeiJ.GalatiA. (2024). Systematic literature review on the nexus of food waste, food loss and cultural background. Int. Mark. Rev.41, 683–716. doi: 10.1108/IMR-12-2023-0366

  • CrossRef
  • Google Scholar

• 46

  MungkungR.SaejewK.JakrawatanaN.AreerobT.KhantotongS.PrueksakornK. (2026). Measuring food waste national baseline and reduction strategies towards sustainable development goal. J. Mater. Cycles Waste Manag.28, 556–569. doi: 10.1007/s10163-025-02441-y

  • CrossRef
  • Google Scholar

• 47

  MyersC. A.BeylR. A.MartinC. K.BroylesS. T.KatzmarzykP. T. (2020). Psychological mechanisms associated with food security status and BMI in adults: a mixed methods study. Public Health Nutr.23, 2501–2511. doi: 10.1017/S1368980020000889,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 48

  NguyenK. H.TranM. D. (2025). Exploring the landscape of pro-environmental behaviours and sustainable consumer behaviours: unveiling priority research areas through bibliometric insights. SAGE Open15:21582440251388081. doi: 10.1177/21582440251388081

  • CrossRef
  • Google Scholar

• 49

  OláhJ.KaszaG.Szabó-BódiB.SzakosD.PoppJ.LaknerZ. (2022). Household food waste research: the current state of the art and a guided tour for further development. Front. Environ. Sci.10:916601. doi: 10.3389/fenvs.2022.916601

  • CrossRef
  • Google Scholar

• 50

  OlthofT. (2012). Anticipated feelings of guilt and shame as predictors of early adolescents' antisocial and prosocial interpersonal behaviour. Eur. J. Dev. Psychol.9, 371–388. doi: 10.1080/17405629.2012.680300

  • CrossRef
  • Google Scholar

• 51

  PanaitM.RaimiL.GigauriI. (2025). Responsible Behavior Design and Sustainable Future. Switzerland: Palgrave Macmillan.

  • Google Scholar

• 52

  ParvatiyarA.ShethJ. N. (2023). Confronting the deep problem of consumption: why individual responsibility for mindful consumption matters. J. Consum. Aff.57, 785–820. doi: 10.1111/joca.12534

  • CrossRef
  • Google Scholar

• 53

  PaskoO.ChenF.OriekhovaA.BrychkoA.ShalyhinaI. (2021). Mapping the literature on sustainability reporting: a bibliometric analysis grounded in Scopus and web of science Core collection. Eur. J. Sustain. Dev.10:303. doi: 10.14207/ejsd.2021.v10n1p303

  • CrossRef
  • Google Scholar

• 54

  RizviA.KearnsM.DignamM.CoatesA.SharpM. K.MagwoodO.et al. (2024). Effects of guaranteed basic income interventions on poverty-related outcomes in high-income countries: a systematic review and meta-analysis. Campbell Syst. Rev.20:e1414-n/a. doi: 10.1002/cl2.1414,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 55

  RodriguesD.LopesA.BatistaF. (2025). Detecting incoherent citation data among three bibliometric platforms: OpenAlex, Scopus and web of science. J. Inf. Sci. doi: 10.1177/01655515251330579

  • CrossRef
  • Google Scholar

• 56

  SalumeP. K.BarbosaM. W.PintoM. D. R. (2026). Food waste and consumer behavior: a bibliometric and review study and future research directions. Foods15:380. doi: 10.3390/foods15020380,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 57

  SathishT.SelvarajuS.AhalyaN.KumarA.AgarwalA.PrakashC.et al. (2025). Exploring waste-derived catalysts for sustainable biodiesel production: a path towards renewable energy. Carbon Lett.35, 161–172. doi: 10.1007/s42823-024-00773-7

  • CrossRef
  • Google Scholar

• 58

  SchanesK.DobernigK.GözetB. (2018). Food waste matters—a systematic review of household food waste practices and their policy implications. J. Clean. Prod.182, 978–991. doi: 10.1016/j.jclepro.2018.02.030

  • CrossRef
  • Google Scholar

• 59

  ShoveE.PantzarM.WatsonM. (2012). The Dynamics of Social Practice: Everyday Life and How it Changes. Los Angeles: SAGE.

  • Google Scholar

• 60

  ShoveE.SchatzkiT.HuiA. (2017). Matters of Practice. United Kingdom: Routledge.

  • Google Scholar

• 61

  SiphoS.TilahunS. W. (2020). Potential causes of postharvest losses, low-cost cooling technology for fresh produce farmers in sub-Sahara Africa. Afr. J. Agric. Res.16, 553–566. doi: 10.5897/AJAR2020.14714

  • CrossRef
  • Google Scholar

• 62

  SjervenJ. (2019) UN Climate Report Addresses Impact of Diet, Food Waste. Kansas City, MO, USA: Sosland Publishing Company.

  • Google Scholar

• 63

  SnyderA. B.MartinN.WiedmannM. (2024). Microbial food spoilage: impact, causative agents and control strategies. Nat. Rev. Microbiol.22, 528–542. doi: 10.1038/s41579-024-01037-x,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64

  TekuD.MengeshaY. A.KeseteN. (2025). Review of area exclosure practices for sustainable land management and erosion control in Ethiopia: objectives and PRISMA review. Front. Conserv. Sci.6:1568725. doi: 10.3389/fcosc.2025.1568725

  • CrossRef
  • Google Scholar

• 65

  TheodoridisP. K.ZacharatosT. V. (2022). Food waste during Covid-19 lockdown period and consumer behaviour—the case of Greece. Socio Econ. Plan. Sci.83:101338. doi: 10.1016/j.seps.2022.101338

  • CrossRef
  • Google Scholar

• 66

  UrugoM. M.TekaT. A.GemedeH. F.MershaS.TessemaA.WoldemariamH. W.et al. (2024). A comprehensive review of current approaches on food waste reduction strategies. Compr. Rev. Food Sci. Food Saf.23:e70011. doi: 10.1111/1541-4337.70011,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 67

  WaittG.PhillipsC. (2016). Food waste and domestic refrigeration: a visceral and material approach. Soc. Cult. Geogr.17, 359–379. doi: 10.1080/14649365.2015.1075580

  • CrossRef
  • Google Scholar

• 68

  WangL.ChengS.LiuG.LiuX.BaiJ.ZhangD.et al. (2015). Study on theories and methods of Chinese food waste. J. Nat. Resour.30, 715–724. doi: 10.11849/zrzyxb.2015.05.001

  • CrossRef
  • Google Scholar

• 69

  WatsonM.MeahA. (2012). Food, waste and safety: negotiating conflicting social anxieties into the practices of domestic provisioning. Sociol. Rev.60, 102–120. doi: 10.1111/1467-954X.12040

  • CrossRef
  • Google Scholar

• 70

  YangL.KazumitsuM. (2024). Exploring the household food-wasting behaviour: a bibliometric review. Waste Manag. Res. doi: 10.1177/0734242x241259630

  • CrossRef
  • Google Scholar

• 71

  YeowP. H. P.LooW. H. (2022). Antecedents of green computer purchase behavior among Malaysian consumers from the perspective of rational choice and moral norm factors. Sustainable Prod. Consumption32, 550–561. doi: 10.1016/j.spc.2022.05.015

  • CrossRef
  • Google Scholar

• 72

  ZhangW.WangY.JinL. (2025). Digital reading literacy, social reading motivation and social reading behavior of middle school students—an analysis based on the mediating effect model. Front. Hum. Dyn.7:1623865. doi: 10.3389/fhumd.2025.1623865

  • CrossRef
  • Google Scholar

• 73

  ZhangJ.YiM.ZhangG.TangG.LiuJ. (2026). A review of roadside green space ecological benefits using citespace: research hotspots and future trends. Urban Ecosyst.29:96. doi: 10.1007/s11252-026-01955-z

  • CrossRef
  • Google Scholar