Developing the NOMI framework for food and nutrition security in a climate-exposed delta: qualitative evidence from the Indian Sundarbans

Abstract

Introduction

Conventional food and nutrition security frameworks often priorities availability and access but insufficiently capture lived experience, socioecological interactions, equity, adaptive capacity, and knowledge plurality in climate-exposed settings. This study develops the NOMI framework to analyse food and nutrition security in the Indian Sundarbans, a delta shaped by salinity intrusion, recurrent cyclones, livelihood precarity, and institutional fragmentation.

Methods

An exploratory qualitative study was conducted across six administrative blocks: Sagar, Namkhana, Kakdwip, Patharpratima, Gosaba, and Sandeshkhali II. Data comprised household interviews (n = 40), focus group discussions (n = 7), expert consultations (n = 5), and ethnographic field notes. Bengali interviews and discussions were audio-recorded with consent, transcribed verbatim, and analysed in ATLAS.ti using reflexive thematic analysis and abductive synthesis.

Results

Ten thematic clusters showed that food and nutrition security is negotiated through rice-centred diets, seasonal monotony, limited animal-source foods, ecological degradation, institutional frictions, embodied stress, adaptive coping, gendered allocation, territorial marginalisation, and changing local foodways. Cross-theme synthesis produced four NOMI pillars: Nutrition as a Socioecological System, Observational Resilience, Multi-Lens Equity, and Inclusive Knowledge. These patterns were shaped by ecological stressors such as salinity intrusion, cyclones, and erosion, alongside fragile livelihoods and fuel and water constraints affecting utilisation. Households sustained food access through storage, substitution including wild foods, safety nets, livelihood diversification, and migration. However, vulnerability remained unevenly distributed. Women consistently ate last and least, while landless and island-based households faced compounded disadvantage. Traditional foodways supported resilience but were eroding and remained under-recognised.

Conclusions

NOMI provides a middle-range, mechanism-focused framework for climate-vulnerable coastal contexts. By foregrounding ecological feasibility, adaptive practice, structural inequity, and knowledge legitimacy, it can strengthen interpretation of performance and equity gaps in platforms such as ICDS, school meals, and the Public Distribution System, while supporting nutrition-sensitive One Health programming.

Highlights

• •Food and nutrition security in the Sundarbans is shaped by coupled ecological, institutional and social constraints.
• •The NOMI framework organises findings into ecological feasibility, observational resilience, equity and knowledge legitimacy.
• •Rice-centred diets and public grain transfers stabilise calories but do not ensure dietary diversity.
• •Women, older adults, landless households and isolated island communities absorb unequal nutritional burdens.
• •Co-produced nutrition programming should combine climate adaptation, social protection and local food knowledge.

1. Introduction

Global food security continues to embody a paradox of progress and vulnerability [1], [2], [3]. Although global food systems can meet aggregate calorie demand, hunger, micronutrient deficiencies, and diet-related non-communicable diseases persist alongside rising obesity, producing the widely recognised double burden of malnutrition [4], [5], [6], [7], [8], [9], [10], [11]. These outcomes reflect not only inequities in access and affordability but also structural and ecological constraints that shape what households can grow, purchase, prepare, and sustain across seasons and shocks [3], [12], [13], [14]. Rapidly changing food environments with increased availability of ultra-processed foods further compound these dynamics [8], [15], [16], [17], [18].

Since the 1996 World Food Summit, food security has been defined as a condition in which “all people, at all times, have physical, social and economic access to sufficient, safe and nutritious food that meets their dietary needs and food preferences for an active and healthy life” [19]. This definition consolidated four classical pillars: availability, access, utilisation, and stability that have underpinned measurement and policy [19]. Letter expanded by the High-Level Panel of Experts (HLPE) to include agency and sustainability, emphasising participation, power, and ecological grounding in food system decision-making [20]. Food-systems scholarship has further highlighted resilience, inclusivity, and environmental sustainability as conditions for long-term security under climate change and compounding shocks [21], [22], [23], reflecting growing recognition that nutritional outcomes are shaped across the life course by interacting biological, social, and environmental pathways [9], [24], [25].

Despite these conceptual advances, operationalisation has lagged. In practice, food security, nutrition security, and food and nutrition security are often treated as interchangeable, enabling programmes to claim nutritional sensitivity while remaining centred on calorie sufficiency and anthropometric endpoints [26], [27], [28]. Dominant indicators (cereal availability, caloric intake, stunting, and wasting) often underrepresent dietary quality, seasonality, culturally shaped food practices, and environmental feedbacks that influence utilisation and feasibility [29], [30], [31], [32], [33], while obscuring within-household coexistence of undernutrition and overweight [34], [35]. In this paper, we use food and nutrition security to denote the sustained ability of households to access, utilise, and benefit from safe, diverse, and culturally acceptable diets in ways that are equitable and resilient to shocks.

Experience-based measurement has long shown that food insecurity is also lived as anxiety about food supply, perceived inadequacy of food quality and quantity, and progressive reductions in intake; ethnographically grounded qualitative item development combined with quantitative validation improves classification and policy relevance [36], [37], [38]. Evidence from diverse contexts demonstrates that experiential food insecurity is strongly associated with undernutrition and household-level double burden outcomes beyond what income or calorie measures capture [39]. However, food-related fields still face reporting challenges when qualitative components are appended without sufficient transparency or alignment with qualitative standards [40], [41], [42]. Together, these points indicate the need for approaches that capture mechanisms, lived constraints, and implementation realities, rather than relying on indicators alone [24].

A One Health perspective reinforces this need in climate-exposed landscapes. One Health is now formalized as an integrated, unifying approach that seeks to balance and optimize the health of people, animals, and ecosystems, explicitly recognizing their interdependence and the need for coordinated, multi-sector action on shared threats, including safe and nutritious food, clean water, and climate change [43], [44]. Nutritional outcomes co-evolve with environmental change (e.g., salinity intrusion, flooding, erosion, ecosystem degradation), livelihood interfaces with animals and aquatic systems, and health risks linked to water quality, food safety, and infectious disease ecology [45], [46], [47]. In deltas where fisheries and aquaculture are central to livelihoods and diets, aquatic foods can be a major source of micronutrients while also linking nutrition to ecosystem health and exposure risks across human–animal–environment interfaces [48]. These interactions are mediated by infrastructure and governance capacity, shaping not only what foods are available and affordable but also whether they can be safely accessed, stored, prepared, and converted into health under recurrent disruption. For public health, this implies that programme effectiveness depends on risk environments, service reach, household coping capacity, and equity conditions that are often weakly captured by indicator-only assessments.

This gap is evident in India. Large-scale safety nets such as the Public Distribution System have reduced calorie deprivation and stabilised consumption during crises, yet their cereal-centric design limits dietary diversity gains and broader food and nutrition security outcomes [49], [50], [51], [52], echoing global evidence that nutrition-insensitive programmes can unintentionally reinforce poor-quality diets [15]. Similarly, maternal and child programmes such as Integrated Child Development Services (ICDS) and school feeding have expanded coverage and routine monitoring, but often approach food insecurity primarily through biomedical endpoints, with limited attention to labour demands, care practices, seasonality, and local food environments that determine whether recommended diets are feasible in everyday life [53], [54]. This disconnect is reinforced by disciplinary fragmentation in the evidence base, with agriculture prioritising yields, economics emphasising prices and markets, public health focusing on deficiencies and morbidity, environmental science tracking ecosystem change, and social sciences examining labour, inequality, and culture often analysed in isolation [24], [55], [56], [57]. The result is robust monitoring of prevalence and trends, but weaker explanation of how food insecurity is negotiated at household level and why programme benefits vary across contexts.

The Indian Sundarbans delta concentrates these dynamics in an extreme but policy-relevant setting. Ecologically rich yet highly vulnerable, it is shaped by recurrent cyclones, saline intrusion, riverbank erosion, and precarious livelihoods [58], reflecting well-documented climate nutrition pathways in which environmental shocks disrupt food access, care practices, environmental health, and disease risk [9], [25], [47]. High burdens of food insecurity and malnutrition intersect with poverty, weak infrastructure, and constrained healthcare access [59], [60], consistent with evidence that climate-exposed populations face compounded nutritional risk through interacting social and ecological vulnerabilities [8], [24].

Here, hunger is not only the absence of calories but the inability to withstand ecological shocks, manage inequitable intra-household food distribution, and sustain culturally embedded coping practices under instability [61], [62], [63]. This aligns with qualitative and mixed-methods evidence showing that food insecurity is experienced as uncertainty, social negotiation, and risk management rather than simple intake shortfalls [39], [64]. Food is negotiated through obligations, norms, risk management, and adaptation; nourishment is social as much as biological and frequently co-produced with environmental conditions and institutional functioning [9].

Existing frameworks clarify pillars and outcomes but often under-specify the mechanisms that link ecological disturbance, learning and adaptation, stratified vulnerability, and competing knowledge systems to programme-relevant decision points in fragile, climate-exposed contexts. This critique echoes broader assessments that nutrition and food security frameworks remain indicator-heavy and weakly explanatory, particularly with respect to food environments, social power, and lived constraints [24], [28], [34], [35] To address this gap, we draw on qualitative fieldwork across six Sundarbans blocks, integrating household interviews, focus group discussions, expert consultations, and ethnographic field notes - approaches increasingly recognised as essential for understanding nutrition dynamics in complex systems [41], [42], [64].

Iterative coding and abductive analysis synthesised four cross-cutting relational dynamics, which crystallised into the NOMI framework: Nutrition as a Socioecological System (NSS), Observational Resilience (OR), Multi-Lens Equity (MLE), and Inclusive Knowledge (IK). Such middle-range, mechanism-oriented frameworks are increasingly advocated in nutrition and public health research to bridge micro-level experience and macro-level policy design [15], [24]. We position NOMI as a middle-range framework in the sense of [65], linking grounded patterns in the Sundarbans to transferable analytical concepts and practical questions of diagnosis, programme design, and evaluation, consistent with calls for systems-oriented and context-sensitive approaches to addressing the double burden of malnutrition [8], [9], [10].

The aim of this paper is twofold: first, to foreground lived experiences of food, uncertainty, and care in the Sundarbans; and second, to demonstrate how NOMI can structure interpretation of these experiences for public health and One Health-relevant intervention planning and assessment, responding directly to evidence that climate change, nutrition, and health must be addressed through integrated, cross-sectoral frameworks [15], [43], [44], [47].

2. Methodology

2.1. Study design and epistemological orientation

This study adopted an exploratory qualitative design informed by ethnographic principles to examine food security and nutrition dynamics in climate-vulnerable communities of the Sundarbans delta, India [66], [67], [68]. The research was situated within a constructivist epistemological framework, recognising that experiences of food security are socially constructed and embedded in specific socioecological contexts [69]. An iterative research strategy was employed, whereby insights emerging from early fieldwork informed the refinement of subsequent data collection tools and analytical focus [70]. The analysis followed an abductive logic, enabling movement between inductively generated themes and existing theoretical constructs on socioecological systems, resilience, equity, and local knowledge systems [71], [72], [73]. Ethnographic principles were operationalised through field observations and detailed field notes to contextualise interview and group accounts. The analytic objective was to characterise the feasibility space within which dietary practices in the Sundarbans are constructed, sustained, or compromised under recurrent climate disturbance.

2.2. Study area

Fieldwork was conducted across six administrative blocks in North and South 24 Parganas districts: Sagar, Namkhana, Kakdwip, Patharpratima, Gosaba, and Sandeshkhali II. Fig. 1 locates the six study blocks and visually distinguishes the island and mainland settings that shape market connectivity, exposure to cyclones and salinity, and access to services. Sites were purposively selected to capture diversity in island and mainland contexts, exposure to cyclones and salinity, infrastructure and service access, livelihood portfolios, and socioeconomic vulnerability. Participant characteristics and site coverage are summarised in Table 1, demonstrating representation across blocks, livelihood types, and household roles relevant to food-system exposure.

Fig. 1.

Study Area Map: Six Administrative Blocks of the Indian Sundarbans

Table 1.

Demographic profile of participants across six administrative blocks of the Indian Sundarbans

Sl. No.	Study Area / Institute (with brief description)	Methods	Respondents	Number of Interviews (Age; Gender)	Rationale
1	Institutions working on food security, salinity, aquaculture, livestock, and deltaic ecology.	IDIs	Academic and technical experts	5 (48-62 yrs; 4M, 1F)	To incorporate scientific and institutional understanding of food systems, salinity, and resilience.
2	Namkhana: A coastal mainland block exposed to cyclones, salinity intrusion, and seasonal migration; livelihoods include fishing, cultivation, petty trade.	IDIs	Fishers, cultivators, livestock rearers, labourers, traders, drivers, forest collectors, migrants	6 (45-65 yrs; 3M, 2F)	Known for high climate exposure and mixed land–sea livelihoods, capturing gendered coping roles.
3	Patharpratima: Deltaic islands affected by embankment breaches, tidal flooding, and saline ingress; dependent on agriculture, fishing, and wage labour.	IDIs	Inland fishers, small farmers, labourers, shop owners, migrants, widow-led households	5 (20-70 yrs; 3F, 5M)	Strong dependence on fragile ecosystems; high exposure to riverine hazards.
4	Sagar and Ghoramara: Highly eroding islands facing tidal surges and farmland loss; cultivation and fishing are major occupations.	IDIs	Farmers, fishers, livestock keepers, petty traders, migrant families, household women	13 (30-70 yrs; 7F, 11M)	Among the most climate-affected sites, illustrating recurrent shocks to food systems.
5	Canning (Pakhiralay, Gosaba): Riverine areas with high SC/ST populations; livelihoods include fishing, agriculture, and forest resource extraction; highly flood-prone.	IDIs	Fishing families, cultivators, forest product collectors, migrant families, elderly women	5 (25-60 yrs; 3M, 2F)	Selected for heavy dependence on forest-wetland ecosystems and social vulnerability.
6	Kakdwip: Semi-urban transitional zone with both farm and non-farm livelihoods; exposed to saline flooding and cyclones.	IDIs	Fishers, cultivators, daily wage earners, small business operators, transport workers, household women	5 (30-65 yrs; 2F, 3M)	To capture livelihood transitions and environmental pressures affecting food access.
7	Sandeshkhali II: Inland delta block with high SC/ST concentration; dominated by agricultural labour, cultivation, and craft-based work; flood-prone.	IDIs	Small farmers, inland fishers, wage workers, craft households, household women	6 (35-70 yrs; 1F, 5M)	Important for understanding caste-linked marginality under environmental stress.
8	All study sites	FGDs	Local workers, fishers, teachers, health staff, officials, ration dealers, NGOs	1 FGD per site	To triangulate community, institutional, and sectoral perspectives on food and nutrition security.
Total - 5 expert IDIs, 40 household IDIs, and 7 FGDs

2.3. Participants, sampling, and data collection

Data collection was conducted between August 2024 and November 2024 and comprised forty in-depth household interviews, seven focus group discussions (FGDs), and five expert consultations. Expert consultations with domain specialists in coastal livelihoods/ food systems, and public-sector nutrition programming. These consultations were used to contextualize sectoral mechanisms (e.g., livelihood transitions, food access constraints) and to triangulate the plausibility of cross-cutting patterns emerging from community data. Sampling was purposive and maximum-variation oriented to capture diversity across gender roles, livelihood categories, and social position, and to include both chronic and episodic experiences of food insecurity. Participants were recruited through community intermediaries and local networks, with eligibility defined as adult residents with primary responsibility for household food decisions and/or livelihood activities. Sampling continued until no substantively new themes emerged across blocks and participant groups.

Data collection was guided by three explicit questions: (1) What ecological, social, and institutional constraints shape household food and nutrition security, diet quality, and nutritional well-being in the Sundarbans? (2) How do households cope and adapt to seasonal variability and climate-related disruption? (3) What changes do residents prioritise for a more secure, equitable, and dignified food future? Semi-structured interview and FGD guides were iteratively refined and included probes on (i) daily meals, food sourcing, and seasonal diet change; (ii) livelihood-food linkages and market access; (iii) programme access and perceived functioning (Public distribution system (PDS), Integrated Child Development Services (ICDS), school meals) including disruption during monsoon/cyclones; (iv) water and food safety practices and illness experiences affecting utilisation; and (v) intra-household allocation, gendered labour, and care constraints. Household interviews and FGDs were conducted in Bengali by the authors, audio-recorded with oral consent, transcribed verbatim, and supplemented with field notes on household food environments, preparation spaces, and local markets. Expert consultations captured local service delivery and institutional perspectives relevant to food, health, and livelihoods.

Interview and FGD guides were drafted by the research team in English, translated into Bengali, reviewed by community intermediaries for cultural appropriateness, and piloted with three to five households in one of the six study blocks before main fieldwork commenced; pilot data were excluded from the analytic corpus. Oral consent was audio-recorded at the start of each interview and FGD as documentation; FGD consent was sought verbally after introductions, with participants reminded of their right to withdraw at any time, and a field-note entry was logged for each session given low literacy. Approximately 50 households approached across the six blocks, 40 enrolled; non-participation was attributed primarily to time constraints during the agricultural season and to unavailability during visit windows. No systematic differences in livelihood type or gender composition were apparent between participants and non-participants.

Household interviews lasted between 60 and 90 minutes (median approximately 75 minutes); FGDs lasted between 90 and 120 minutes; sessions were conducted in participants' homes, courtyards, or community spaces with no third-party present unless invited by the participant.

2.4. Data analysis and framework synthesis

Data were analysed using ATLAS.ti Version 25 following principles of reflexive thematic analysis [74], [75], [76]. Transcripts were systematically analysed to identify salient experiences and constraints. Each transcript was assigned a unique identifier indicating the data type (household interview [HH], focus group discussion [FGD], or expert consultation [EXP]), the study block (denoted by the first three letters of the block name), and the interview sequence number (e.g., HH-SAG-01 for the first household interview conducted in Sagar). This system ensured systematic organization and traceability across the six study blocks. The process began with open coding, which identified recurring ideas across transcripts such as rice dependence, dietary monotony, gendered sacrifice, and salinity stress. Given the multidimensional nature of lived experiences in climate-exposed settings, certain excerpts were intentionally multiply coded to retain layered meanings rather than force singular categorisation. This approach aligns with established qualitative practice recognising that a single data segment can legitimately inform multiple thematic interpretations [74], [77]. Codes were then grouped through axial coding into broader categories (sub-themes), and these categories were consolidated into higher-order thematic clusters. Cross-theme synthesis linked these clusters to four integrative pillars constituting the NOMI framework: Nutrition as a Socioecological System (NSS), Observational Resilience (OR), Multi-Lens Equity (MLE), and Inclusive Knowledge (IK). NOMI was not prespecified as a coding framework. It was developed after reflexive thematic analysis, as an abductive synthesis of cross-cutting patterns across themes; pillar definitions were iteratively refined by re-checking against the coded dataset and relevant theory. Abductive engagement was maintained to ensure findings remained grounded while strengthening conceptual transferability [71], [72], [73]. This analytic progression is summarised in Fig. 2, which shows how open codes were clustered into thematic groupings and then synthesised into the four NOMI pillars.

Fig. 2.

Analytic traceability from data sources to codes, themes, and NOMI pillars.

2.5. Trustworthiness and reflexivity

Trustworthiness was strengthened through triangulation across households, FGDs, expert accounts, and field notes; iterative team discussions during coding and interpretation; and an audit trail maintained within ATLAS.ti. Interpretive validation was supported by revisiting emerging summaries with selected participants and/or community intermediaries. Reflexive notes and debriefs were used to consider how the researcher positioning, interview setting, local power relations and hierarches may have shaped disclosure, particularly regarding intra-household allocation and programme experiences.

2.6. Ethical considerations

Ethical approval was obtained from the institutional ethics committee. This study received ethical approval from the Institutional Human Ethics Committee (IHEC), Belda College (Certificate Ref. BC/IHEC/ECC/10/24, dated 10 Aug 2024; IHEC meeting 08 Aug 2024). Given low literacy and possibility that written consent forms can feel intimidating or raise suspicion oral informed consent was used after explaining the study purpose, voluntary participation, the right to withdraw, and confidentiality protections. Transcripts were de-identified by removing personal identifiers and specific locational markers; access to audio files and transcripts was restricted to the research team, and only anonymised excerpts are reported.

3. Results

Grounded fieldwork across the Indian Sundarbans demonstrates that food and nutrition security is not experienced as a discrete or stable condition. It is produced through tightly coupled ecological, social, economic, institutional, embodied, and cultural processes. Ten inductive thematic clusters were consolidated into four analytically interdependent NOMI pillars (Table 2, Table 3). Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8, Fig. 9, Fig. 10, Fig. 11 visualise selected pathways operating across households and territories, while Fig. 12 synthesises the integrated framework.

Table 2.

Mapping of inductive thematic findings (theme 1 to 10) to NOMI pillars and mechanism signals.

Themes	Mechanism signal (what the evidence shows)	Quotation(s)	NOMI pillar(s) most directly supported
1.Rice-centred dietary dependence & low dietary diversity	Rice functions as a cultural and structural anchor of satiety; dietary diversity becomes episodic (“whatever is available”), with fruit/milk aspirational	"We eat rice three times a day - Panta Bhat… hot rice… Pusto Bhat at night." (HH-SAG-5) "In the morning we eat Panta Bhat… children eat the same." (FGD-GOS-1) "Fruits… only when we go to Kakdwip. Milk is only for those who have cows." (HH-NAM-1)	NSS (core); MLE (linked via distribution mechanisms)
2. Limited animal-source foods & seasonal crises	Animal-source foods become contingent due to ecological change, market extraction/price, and regulatory constraints; climatic shocks compress diets into dry staples	"During the ban, fish availability drops." (HH-NAM-1) "After the flood, most pond fish perished." (FGD-GOS-2) "Sometimes we just boil potato because fish is too costly." (FGD-SAG-1) "During storms… vegetables and fish become impossible to get." (FGD-NAM-1)	NSS (core); OR (core); MLE (linked via protein allocation patterns)
3.Ecological stressors & environmental change	Ecological processes operate as upstream feasibility constraints on production, fisheries, and water safety; repeated cyclones/erosion generate prolonged recovery and diet narrowing	"We only do Aman paddy… there is no alternative cropping." (HH-NAM-1) "Cyclones like Amphan and Yaas have completely devastated this island." (HH-SAG-3) "Around ten percent of the people here have become landless." (FGD-SAG-1) "The salinity causes skin diseases and diarrhoea among people." (FGD-SAG-2) "After the storm… weeks before we can get back to regular meals." (FGD-NAM-1)	NSS (core); OR (core); MLE (linked via differentiated exposure and recovery)
4. Institutional safety nets & contradictions	Programmes buffer hunger but translate imperfectly into diet quality because of delivery/quality frictions and utilisation constraints (fuel, transport, service access)	"People usually have rice or flour at home because the government provides grains." (HH-NAM-1) "The people who don't eat the government rice, they sell it." (HH-PAT-1) "Sometimes at night we have to rent a boat to reach the hospital." (FGD-GOS-2) "Cooking fuel is very scarce here… I cook only once a day." (HH-SAG-10)	NSS (core); MLE (core-linked); OR (linked via buffer reliability during shocks)
5. Physiological & psychological factors	Health status and stress constrain what is tolerable/feasible, sometimes narrowing diets at household level	"After I developed diabetes, I had to stop eating fermented rice." (HH-NAM-5) "We have mental tension… we can't sleep at night." (HH-SAG-5)	NSS (core); OR (linked-chronic stress as compounding disturbance); MLE (linked via gendered caregiving burdens)
6. Coping & adaptive practices	Observed coping repertoire spans anticipation, absorption, adaptation, and boundary points where household action cannot substitute for infrastructure/governance	What we eat is a daily issue… based on whatever I earn today." (HH-NAM-1) "Before the flood… we collect puffed rice and flattened rice." (FGD-GOS-2) "We cook together and share meals while staying at the cyclone shelter." (FGD-GOS-2) "My house is right next to the embankment. It breaks every year." (HH-SAG-5) "People survived by eating Giriya shak [Suaeda maritima]." (FGD-SAG-2)	OR (core); IK (core-linked); MLE (linked via labour/time costs)
7. Gendered malnutrition & household inequities	Scarcity is managed through allocation hierarchies; women/elderly absorb nutritional compromise; caloric sufficiency coexists with fatigue/“hidden hunger”	"By the time food reaches me, only rice and dal are left… fish and eggs go to the children and their father." (HH-SAG-10) "When rice is less, I eat only once so they can eat twice." (FGD-GOS-1) "I usually eat three times a day… but always feel sleepy and tired." (HH-NAM-2)	MLE (core); NSS (linked-quality constraints under cereal sufficiency)
8. Inequalities, marginalisation & adaptive challenges	Vulnerability and coping capacity are stratified by place, assets, livelihood type, and social identity; adaptation redistributes burdens (time poverty, autonomy constraints, migration effects)	"In Patharpratima they are about forty percent behind… Gosaba is more advanced." (EXP-SAG-2) "We don't have land… must buy everything from the market." (HH-SAG-10) "Our sons go outside for work… only old people and women remain." (HH-PAT-1) "I cook only once a day… often don't get breakfast on busy days." (HH-SAG-10)	MLE (core); OR (linked-coping costs); IK (linked-erosion of reciprocity/transfer)
9. Traditional foodways & cultural practices	Cultural repertoires function as practical resilience technologies (seasonality, combinations, rituals, fermentation) but are pressured by changing preferences and loss of communal tools	"We live in the forest, so we worship Bonbibi… to protect ourselves from tigers." (FGD-GOS-1) "Our daughter likes biryani and chowmein… every village now has such shops." (HH-NAM-5) "Earlier we used to pound rice using the dhenki… now the dhenki is gone." (FGD-SAG-2)	IK (core); OR (linked-adaptive repertoire)
10. Aspirations & change	Communities specify enabling conditions for dignified, durable food and nutrition security: skills, local production, women’s livelihoods, and reliable systems (water, markets, storage, transport, services)	"Hunger doesn't happen now… many nights we went to sleep hungry. Now things have improved." (HH-NAM-1) "If we ensure local foods such as banana blossoms, Colocasia, leafy greens we will get good nutrition." (EXP-SAN-1) "We couldn't continue because there was no proper support system." (HH-SAG-9)	Integrative (NSS + OR + MLE + IK)-community-defined conditions for feasible, equitable, resilient change

This table provides an early navigation crosswalk showing how each Results subsection contributes to one or more NOMI pillars. The mapping is not intended as a one-to-one assignment: overlap is expected because participants’ accounts describe interdependent socioecological constraints, coping repertoires, stratified burdens, and knowledge systems.

Note

Quote IDs are reported to preserve traceability across participant type and site. HH = household interview; FGD = focus group discussion; EXP = expert/community informant. Site abbreviations follow your field coding (e.g., SAG, NAM, GOS, PAT, KAK, SAN).

NOMI pillar overlap is expected because mechanisms co-occur in lived accounts (e.g., shock impacts simultaneously constrain feasibility, activate coping repertoires, and redistribute burdens).

NOMI = Nutrition as a Socioecological system, Observational Resilience, Multi-Lens Equity, Inclusive Knowledge; NSS = Nutrition as a Socioecological System; OR = Observational Resilience; MLE = Multi-Lens Equity; IK = Inclusive Knowledge.

Table 3.

Inductive themes consolidated into NOMI pillars Nutrition as a Socioecological System, Observational Resilience, Multi-Lens Equity, and Inclusive Knowledge

Empirical Theme Cluster (from field data)	Conceptual Pillar (abductive synthesis)
Socio-Ecological-Dietary Feedback Loops: Respondent explained that ecological conditions such as soil salinity, cyclones, and mangrove biodiversity directly shape dietary outcomes, including crop failure, seasonal hunger, and reliance on wild foods. These outcomes are further influenced by social structures such as land ownership and gender hierarchies, cultural practices including religious food restrictions and traditional knowledge, and economic dependencies like livelihood decline and market reliance. Together, these interactions show that nutrition outcomes emerge from coupled human-environment dynamics.	Nutrition as a Socioecological System (NSS): Reconceptualizes nutrition as emerging from coupled human-environment systems wherein food security outcomes result from continuous feedback between ecological conditions and the full spectrum of social relations, cultural practices, and economic structures. This pillar recognizes that dietary quality and nutritional status are co-produced through multiple reinforcing cycles linking environmental health, social equity, cultural knowledge systems, and economic opportunities. This conceptualization draws on Social-Ecological Systems framework [82] to analyse coupled human-environment sustainability, the Food Systems Framework to connect social and ecological dimensions of food security, and eco-nutrition theory linking ecosystem health to nutrition [94], [127], [128]. It also aligns with the Nutrition Continuum, which highlights the dynamic interplay between food systems, dietary behaviors, and health outcomes [86]
Observe-Learn-Adapt Routines: Community members demonstrated sophisticated monitoring and adaptive response systems involving continuous observation of environmental cues (weather patterns, crop changes, animal behavior, water conditions), iterative learning through experimentation and knowledge sharing, and strategic behavioral adjustments spanning anticipatory actions (pre-stocking food based on weather signs), absorptive tactics (community resource sharing during floods), adaptive strategies (switching to salt-tolerant crops), and transformative changes (livelihood diversification). These routines represent accumulated community wisdom embedded in everyday practices, seasonal preparedness cycles, and disaster response protocols that enable households to navigate climate variability and maintain food security despite recurring environmental shocks.	Observational Resilience (OR): Conceptualizes resilience as a dynamic process of continuous observational learning and adaptive response rather than static capacity or fixed infrastructure. This pillar emerged from recognizing how communities achieve stability through vigilant monitoring, iterative experimentation, and behavioral flexibility based on accumulated local knowledge. OR extends classical stability concepts by positioning human learning and adaptation as core mechanisms for achieving food security persistence over time, particularly under conditions of high environmental uncertainty and social change. This framework integrates Adaptive Resilience Theory emphasizing food system actors adapting through robustness, recovery, and reorientation, Social Learning Theory focusing on community-based learning and knowledge adaptation, and Panarchy Theory [129] conceptualizing system's ability to maintain structure while adapting to disruptions [84], [130], [131].
Layered Inequities in Food Access: Data revealed intersectional vulnerabilities wherein food security experiences vary systematically based on multiple, intersecting social positions. Gender hierarchies create systematic disadvantages (women eating last, restricted mobility, time poverty constraining market access), while economic stratification determines differential impacts (landless households experiencing greatest vulnerability during crises versus land-owning families maintaining greater buffers). Caste discrimination affects service delivery and aid distribution, geographic isolation creates unequal access to markets and services (island versus mainland disparities), and age intersects with gender to produce particular vulnerabilities (elderly women's nutritional deprivation). These layered inequities operate through social equity to food security feedback loops wherein structural disadvantages create differential food access, leading to unequal health outcomes that reinforce existing social hierarchies and perpetuate systematic exclusion from food security benefits.	Multi-Lens Equity (MLE): Introduces an explicitly justice-oriented analytical framework interrogating "Food security for whom?" across multiple, intersecting dimensions of social difference. This pillar recognizes that food insecurity is not experienced uniformly but varies systematically based on gender, class, caste, geography, age, and other social positions that intersect to create compounding disadvantages. MLE demands simultaneous application of multiple analytical lenses to identify who is excluded from food security benefits and why, challenging interventions to address structural inequities rather than assuming aggregate improvements benefit all groups equally. This pillar draws from 4I framework emphasizing inequality, inclusion, [13], Intersectionality Theory [132] analyzing multiple intersecting social identities creating compounding disadvantages, Capabilities Approach [133] focusing on what people can actually do and be, Environmental Justice Framework addressing differential impacts across social groups [134].
Knowledge Frictions Between Local and External Systems: Fieldwork revealed systematic disconnects between indigenous knowledge systems (traditional ecological knowledge, local food processing techniques, weather prediction methods, salt-tolerant crop varieties) and external interventions (government programs, NGO initiatives, scientific recommendations). Participants described knowledge practice feedback loops wherein effective local innovations were undervalued by formal systems, leading to promotion of unsuitable external solutions, poor outcomes, and further erosion of traditional knowledge. Examples included aid agencies introducing inappropriate crop varieties that failed in saline conditions, nutrition programs promoting packaged foods over nutritionally superior local preparations, and development projects ignoring community expertise regarding flood patterns and farming cycles. These frictions create epistemic injustices wherein local knowledge holders are marginalized in decisions affecting their food security, while potentially effective traditional practices are abandoned in favour of less suitable external approaches.	Inclusive Knowledge (IK): Prioritizes co-creation of food security solutions through systematic integration of indigenous, local experiential, and scientific knowledge systems. This pillar emerged from recognizing that effective interventions require bridging different epistemological traditions rather than privileging external expertise over local knowledge. IK challenges epistemic hierarchies that marginalize community knowledge holders, instead positioning them as equal partners in identifying problems, designing solutions, and evaluating outcomes. This framework integrates Participatory Development Theory [135] emphasizing community participation in development processes, Decolonial Epistemology challenging knowledge hierarchies and epistemic injustices, Knowledge Co-production Framework enabling integration of multiple knowledge systems [136], and Traditional Ecological Knowledge (TEK) Theory supporting indigenous knowledge systems integration [137], [138], [139], [140].

Fig. 3.

Ecological stressors and environmental change shaping dietary feasibility in the Sundarbans

Fig. 4.

Human-induced pressures on ecosystems and downstream food–nutrition impacts

Fig. 5.

Livelihood and economic pathways shaping food access and diet diversity.

Fig. 6.

Institutional and infrastructure constraints mediating conversion of entitlements into realized diets.

Fig. 7.

Psychological and physiological pressures shaping food choices, utilization, and household adjustment.

Fig. 8.

Observed resilience strategies: preparedness, food-system adaptation, livelihood diversification, and social solidarity.

Fig. 9.

Multi-Lens Equity: structural, territorial, and intrahousehold pathways shaping unequal food access.

Fig. 10.

Traditional food practices, cultural nutrition norms, and their erosion under stress.

Fig. 11.

Community voices: co-created priorities for nutrition resilience and systems improvement.

Fig. 12.

NOMI conceptual framework diagram, illustrating the integration of four pillars -Nutrition as a Socioecological System (NSS), Observational Resilience (OR), Multi-Lens Equity (MLE), and Inclusive Knowledge (IK).

3.1. Nutrition as a Socioecological System (NSS)

Nutrition as a Socioecological System (NSS) recognises that dietary outcomes in the Sundarbans emerge from interacting ecological, economic, institutional and embodied conditions rather than from food availability or individual choice alone. Across sites, participants described a bounded dietary feasibility space in which staple security was prioritised while dietary diversity and nutrient density remained constrained by salinity intrusion, climatic volatility, livelihood uncertainty, water, fuel, transport, service access and health burdens.

3.1.1. Rice-centred dietary dependence and low dietary diversity (NSS core; OR & MLE secondary)

Rice was the dominant food, a cultural marker of satiety and the practical basis of household food security. Participants described panta bhat [overnight soaked or fermented rice], hot rice at lunch and pusto bhat [cooked rice mixed with water] at night as normal daily patterns. This stabilised hunger but compressed diversity, especially at breakfast and during lean periods.

“We eat rice three times a day. Panta bhat in the morning, hot rice at lunch, and pusto bhat at night. We rarely eat roti [wheat flatbread], only if we are unwell. (HH-SAG-5)”

“In the morning, we eat panta bhat with mashed potatoes; if there is nothing else, just green chillies and salt are enough. Children eat the same. (FGD-GOS-1)”

Vegetables, pulses, fish, fruit and milk were desired but irregular. A Namkhana household reported that fruit was bought mainly when travelling to larger markets and that milk was consumed regularly only by households owning cows. Rice therefore functioned as both nutritional foundation and constraint: it secured calories while normalising low diversity.

3.1.2. Limited animal-source foods and seasonal crises (NSS core; OR & MLE Secondary)

Fish and meat were culturally valued but increasingly contingent. Participants linked declining fish access to salinity, pond losses after flooding, seasonal bans, market extraction, deep-sea shifts and rising prices. Meat was typically reserved for festivals, children’s requests or special occasions. Crisis periods pushed households toward dry staples such as puffed rice and flattened rice.

“After the flood, most pond fish perished. (FGD-GOS-2)”

“During the ban, fish availability drops. (HH-NAM-1)”

“During storms, we survive on flattened rice or puffed rice. Vegetables and fish become impossible to get. (FGD-NAM-1)”

Animal-source foods therefore appeared as episodic and shock-sensitive resources rather than stable components of daily diets. This had particular implications for micronutrient adequacy and for intrahousehold allocation, because protein-rich foods were often reserved for children or male earners when scarce.

3.1.3. Ecological stressors and environmental change (NSS core; OR & MLE Secondary)

Ecological fragility actively structured what households could produce, obtain and safely consume. Fig. 3 synthesises how slow-onset stressors such as soil and water salinity, altered river hydrology and erosion interacted with rapid-onset shocks including cyclones, storm surges and extreme rainfall. Fig. 4 extends this pathway by showing how shrimp aquaculture, deforestation, pollution and destructive fishing degrade soils, fish stocks, mangrove buffers and local biodiversity. Fig. 5 maps the livelihood consequences: reduced fishing viability, agrarian livelihood collapse, market dependence, income instability and debt.

“Even if we try farming, the land becomes saline after one flood; then it is of no use for five years. If saline water enters, it is over. (HH-SAG-5)”

“Cyclones like Amphan and Yaas have completely devastated this island. People are still trying to recover. (HH-SAG-3)”

“Around 10 percent of the people here have become landless. Their grandfathers used to be landlords; now almost everyone has lost their homes and land to the river. (FGD-SAG-1)”

Participants also described shrimp farming as a double-edged livelihood. It could generate income for some households, but saline water spread beyond pond boundaries and reduced surrounding cultivation. Destructive fishing and juvenile fish capture were similarly framed as undermining future aquatic food availability.

“If I do shrimp farming on one bigha [0.33 acres], the nearby ten bighas [3.3 acres] become uncultivable. (FGD-GOS-2)”

“They catch everything, big or small fish, with no consideration. They even catch the juvenile fish before they can grow and reproduce. (HH-NAM-1)”

Health effects were also narrated through ecological pathways. Saline water was associated with skin disease and diarrhoea, while storm periods disrupted normal meals for weeks. Ecological change therefore operated simultaneously as a production, livelihood, water, disease and nutrition pathway.

3.1.4. Institutional safety nets and contradictions (NSS core; OR & MLE Secondary)

Safety-net programmes buffered extreme hunger but did not guarantee dietary adequacy. The PDS was described as a lifeline because subsidised grain ensured that most households had some rice or flour. Yet grain quality, supply regularity, biometric authentication, electricity, network connectivity, transport and informal resale shaped whether entitlements became nutritionally adequate meals. Fig. 6 maps these service and infrastructure bottlenecks.

“Even if people do not have much else, they usually have rice or flour at home because the government provides grains. (HH-NAM-1)”

“The people who do not eat the government rice sell it. I bought that. (HH-PAT-1)”

“During the monsoon, there is a lot of network trouble; the machine only works outside. (FGD-SAG-1)”

ICDS and school meals were valued for children and for the sociality of eating together, but participants described menus as repetitive and sometimes low in protein. During monsoon and cyclone disruption, the platforms expected to buffer insecurity became harder to access. Cash transfers such as Lakshmi Bhandar, old-age pensions and widow pensions helped with groceries, health and education expenses, but respondents repeatedly located the binding constraint beyond cash: water, fuel, roads, transport, health services and storage determined whether money could be converted into food.

“People here can still arrange food somehow, but they remain insecure when it comes to health services, riverbank erosion and floods. (FGD-GOS-2)”

3.1.5. Physiological and psychological factors (NSS core; MLE secondary)

What households ate was shaped not only by availability and affordability, but also by health, fatigue, and bodily limits. Fig. 7 shows how chronic illness, heat stress, pregnancy, fatigue, psychological distress and caregiving demands filtered what households could cook, tolerate and prioritise. Diabetes, hypertension, allergies and stress altered meals at household level because foods were usually cooked collectively.

“After I developed diabetes, I had to stop eating fermented rice. (HH-NAM-5)”

“We have mental tension. We cannot sleep at night. (HH-SAG-5)”

NSS therefore shows that diets in the Sundarbans are assembled within a bounded feasibility field. Rice and entitlements stabilise hunger, but ecological degradation, livelihood compression, service gaps, water and fuel constraints and embodied stress narrow diet quality.

3.2. Observational resilience (OR)

Observational resilience (OR) explains how households act within the NSS feasibility space over time. In the Sundarbans, resilience was rarely enacted through formal infrastructure alone. It emerged through continuous observation, experiential learning and iterative adjustment embedded in everyday life. People monitored tides, rainfall, soil salinity, wind direction, fish behaviour, crop performance, fuel availability, prices and bodily fatigue to guide food, livelihood and care decisions.

3.2.1. Coping and adaptive practices (OR core; NSS, MLE & IK secondary)

Household strategies for managing climate-related food and livelihood pressures were grouped under anticipatory, absorptive, adaptive and transformative capacities. As shown in Fig. 8 anticipatory practices included storing grain, puffed rice, flattened rice, fuel and fodder before floods or cyclones. Absorptive strategies included shared cooking, temporary dietary simplification and drawing on kin or neighbour support. Adaptive practices included livelihood switching, migration, pond restocking, mixed aquaculture, kitchen gardening and use of salt-tolerant foods. Transformative capacities, such as collective advocacy for infrastructure or policy change, were present but rare and institutionally dependent.

“Before the flood happens, we collect puffed rice and flattened rice. Announcements help us prepare. (FGD-GOS-2)”

“We place the gas stove on a cot and cook there. (FGD-NAM-1)”

“We cook together and share meals while staying at the cyclone shelter. (FGD-GOS-2)”

Food-based resilience included foraging and survival foods. Bhater fena [starchy rice water] with tamarind, Colocasia, yam, water lily stems and salt-tolerant greens were described as crisis foods. Giriya shak [Suaeda maritima], a salt-tolerant wild green, was remembered as especially important after saline flooding.

“When there was nothing else, we ate Colocasia, yam and water lily stems. (FGD-GOS-2)”

“People survived by eating this Giriya shak [Suaeda maritima]. (FGD-SAG-2)”

Households diversified livelihoods through wage labour, migration, livestock, small shops, aquaculture and women’s informal income activities. Yet diversification generated trade-offs. Shrimp farming could provide cash while salinising adjacent land; migration sustained purchases through remittances while weakening local labour and intergenerational food knowledge. Fuel-efficient cooking was also a resilience strategy, with women using residual heat, one-pot meals and reduced cooking frequency when fuel was scarce.

“Cooking fuel is very scarce here, so I cook only once a day. (HH-SAG-10)”

“My house is right next to the river embankment. It breaks every year, so we cannot build anything permanent. (HH-SAG-5)”

OR therefore reveals resilience as an adaptive repertoire, but not an equal or cost-free one. The same practice that sustains meals in the short term may increase unpaid work, reinforce gendered sacrifice or postpone structural recovery.

3.3. Multi-lens equity (MLE)

Multi-lens equity (MLE) asks who benefits from food security measures, who absorbs sacrifice and whose adaptive options remain constrained. Fig. 9 synthesises intrahousehold, territorial, asset, livelihood and institutional inequalities that filter both exposure and response. These inequities determine who absorbs nutritional compromise, who benefits from protection and whose adaptation remains limited.

3.3.1. Gendered malnutrition and household inequities (MLE core; NSS & OR secondary)

Women and older adults frequently absorbed scarcity through serving hierarchies. Women reported eating after men and children, receiving less fish, eggs or protein-rich foods and sometimes skipping meals to protect children during shortages. These patterns were routine and intensified during crisis periods.

“By the time food reaches me, only rice and a little dal [lentil soup] are left. Fish and eggs go to the children and their father. (HH-SAG-10)”

“When rice is less, I eat only once so they can eat twice. (FGD-GOS-1)”

“I usually eat three times a day, but always feel sleepy and tired. (HH-NAM-2)”

Participants distinguished being full from being nourished. This distinction is critical because household-level food security can mask hidden hunger within households. Scarcity was socially absorbed through gendered and age-based allocation norms rather than evenly distributed.

3.3.2. Inequalities, marginalisation and adaptive challenges (MLE core; OR & IK secondary)

Food insecurity also varied by territory, livelihood, land access, caste and migration status. Participants contrasted better connected areas with isolated islands and riverine blocks, where market access, health services and programme delivery were more fragile. Landless households were fully exposed to market prices, while erosion converted formerly landholding families into market-dependent households.

“In Patharpratima they are about forty percent behind compared to here. In contrast, if we go to Gosaba, they are more advanced than we are. (EXP-SAG-2)”

“We do not have land for farming and must buy everything from the market; we are completely dependent on shops. (HH-SAG-10)”

Migration was both adaptation and social loss. Remittances sustained food purchases, but women and older adults remained responsible for farms, care and food routines. Refrigeration and marketisation also weakened older reciprocity practices in which surplus fish or food was shared with neighbours.

“Our sons go outside for work; only old people and women are here. We manage as we can. (HH-PAT-1)”

“The practice of sharing food is no longer seen these days. Earlier, if someone caught a lot of fish, they would share some with neighbours. Now the extra catch is stored in refrigerators. (HH-NAM-5)”

Women’s time poverty cut across these inequalities. Cooking, collecting fuel, managing children, buying food, caring for livestock, supporting shops and negotiating scarcity were concentrated in women’s daily routines. Marriage also relocated women into new food environments where they adjusted to unfamiliar fish types, household tastes and senior-family preferences.

“I have to do everything: cooking, household work, collecting fuel, going to the market and helping in our shop. On busy days like Tuesday and saturday I often do not get breakfast. (HH-SAG-10)”

MLE therefore reframes food security from aggregate adequacy to distributive justice. It shows that vulnerability is patterned through intrahousehold, intersectional, intergenerational, territorial and institutional filters.

3.4. Inclusive knowledge (IK)

Inclusive Knowledge (IK) examines how knowledge systems shape what is considered feasible, appropriate and legitimate in food practice. In the Sundarbans, food decisions were guided by ecological observation, cultural meaning, bodily experience, religious rules, women’s cooking repertoires and elder knowledge as much as by formal nutrition information.

3.4.1. Traditional foodways and cultural practices (IK core; NSS & OR secondary)

Local food knowledge was marked by both continuity and erosion, as illustrated in Fig. 10. Continuity appeared in seasonal dietary shifts, culinary biodiversity, fermented and Indigenous preparations, ritual food rules and ecological-spiritual practices. Erosion appeared through loss of shared tools, declining intergenerational transmission, processed food penetration and loss of shared preparation spaces.

“We live in the forest, so we worship Bonbibi to protect ourselves from tigers. (FGD-GOS-1)”

“This is a salty sea area; we mostly eat sour fish curry made with tamarind and mango. (HH-KAK-2)”

“Earlier, we used to pound rice using the dhenki [traditional wooden foot-operated rice pounder]. But now the dhenki is gone. (FGD-SAG-2)”

Ritual days such as amabasya [new moon], purnima [full moon] and Ambubachi [monsoon ritual observance] shaped animal-source food restriction and plant-based alternatives. These practices sometimes narrowed animal-source intake but also maintained cultural continuity and seasonal ecological awareness. Younger people increasingly preferred purchased foods such as biryani, chowmein, egg rolls and fried snacks, which participants interpreted as a shift in the everyday spaces where food knowledge is learned.

3.4.2. Aspirations and change (IK core; NSS, OR & MLE secondary)

Participants did not frame nutrition challenges as information deficiency alone, but as consequences of the practical conditions that shape everyday food access and use. As summarised in Fig. 11 they called for practical food skills, school-based local food literacy, kitchen gardens, community seed banks, Indigenous food revival, women’s livelihoods, better water and sanitation, storage, transport, markets and reliable services.

“If we ensure our daily food through different local foods, like banana blossoms, banana stems, Colocasia and green leafy vegetables, we will get good nutrition and our health will remain good. (EXP-SAN-1)”

“We could not continue because there was no proper support system. Though we tried growing mushrooms, riverine transport and lack of storage led to spoilage. (HH-SAG-9)”

“Women do not have kormosonthan [employment] here. The government should look into it. (HH-NAM-3)”

IK therefore shows that nutritional resilience depends on epistemic recognition. Where local food knowledge remains valued and linked to services, households retain adaptive options. Where knowledge erodes or is displaced by externally imposed models, diet options narrow.

3.5. Integrated synthesis: the NOMI framework

NOMI brings these findings together as an integrated explanatory framework rather than a linear pathway. As shown in Fig. 12, NSS forms the central socioecological system, showing how nutritional outcomes are shaped through interactions among food practices, health and wellbeing, access, services, livelihoods, environment, policy, and infrastructure. The surrounding lenses explain how this system operates under climate stress. OR captures how households observe, learn, and adapt over time. MLE shows how burdens and benefits are unevenly distributed across gender, age, class, livelihood, territory, and institutions. IK shows how local, experiential, scientific, institutional, and participatory knowledge systems shape what responses are possible and legitimate.

Together, the four pillars show that food and nutrition security in the Sundarbans is co-produced through ecological conditions, adaptive routines, social stratification, and knowledge legitimacy. NOMI therefore provides a framework for interpreting not only what households eat, but why certain dietary options become feasible, who bears the cost of adaptation, and whose knowledge counts in designing more resilient and equitable food and nutrition interventions.

4. Discussion

This study developed the NOMI framework to explain food and nutrition insecurity in the Indian Sundarbans as a coupled socioecological condition rather than as the failure of any single sector. The findings answer the three questions that framed the study. First, ecological volatility, livelihood precarity, fragmented services, and gendered allocation jointly constrain food and nutrition security, with rice-centred dependence and seasonal protein scarcity emerging as recurrent signatures. Second, households respond through anticipatory storage, absorptive sharing, adaptive livelihood switching, migration, and limited collective action, but these capacities are unevenly distributed by gender, landholding, caste, territory, income, and institutional access. Third, residents prioritise stable entitlements, reliable water and fuel systems, livelihood diversification, and respectful integration of local knowledge as necessary conditions for a secure food future.

These findings crystallise into four interdependent pillars: Nutrition as a Socioecological System (NSS), Observational Resilience (OR), Multi-Lens Equity (MLE), and Inclusive Knowledge (IK). The empirical chain through which interviews, focus-group discussions, expert consultations, and ethnographic observations were consolidated into codes, themes, and the four NOMI pillars is summarised in Fig. 2. NSS defines the feasibility space within which diets are assembled. OR explains how households navigate that feasibility space over time. MLE shows how risks, sacrifices, and benefits are unequally distributed. IK explains how local food knowledge shapes what is considered feasible, safe, legitimate, and nutritionally meaningful. Together, the four pillars shift food and nutrition security analysis from indicator-counting toward mechanism-tracing.

Food and nutrition security in climate-exposed deltas is therefore better conceptualised as a coupled socioecological syndrome. In the Indian Sundarbans, recurrent cyclones, salinity intrusion, riverbank erosion, livelihood disruption, and infrastructural fragility repeatedly reshape what households can grow, purchase, store, cook, and consume. Diets are not merely outcomes of food availability or household income. They are negotiated within a shifting feasibility space where environmental volatility, livelihood risk, service reliability, social differentiation, and knowledge legitimacy operate simultaneously [78], [79], [80], [81]. NOMI contributes to this literature by integrating ecology, resilience, equity, and knowledge into a single diagnostic framework; explaining why identical shocks produce unequal nutritional trajectories; and offering a pathway-based structure for climate-sensitive nutrition programming.

4.1. NOMI and existing frameworks

NOMI builds on established food-systems, socioecological, resilience, equity, livelihood, One Health, and knowledge-system frameworks, but its contribution lies in their simultaneous integration. Existing frameworks illuminate important parts of the problem but do not fully explain the interaction between compound ecological stress, household resilience, equity, and local food knowledge in climate-exposed delta systems.

The HLPE food systems approach maps production, distribution, consumption, food environments, and nutrition outcomes, but it does not provide a dedicated resilience pillar or a sufficiently disaggregated account of power by gender, class, caste, landholding, and territory [79]. Ericksen’s socioecological food-system framework offers a strong ecological architecture for analysing food-system vulnerability, yet it was not designed to explain nutrition outcomes through household allocation, entitlement conversion, or knowledge legitimacy [78]. Ingram’s approach clarifies the relationship between food systems and global environmental change, but remains more conceptual than operational for diagnosing intrahousehold and territorial nutrition inequality [80].

Ostrom’s socioecological systems framework provides a rigorous basis for analysing governance and common-pool resources, but nutrition is not its central outcome domain and compound hazard dynamics are not foregrounded [82]. Resilience frameworks explain absorptive, adaptive, and transformative capacity, but diet quality, food environments, and knowledge legitimacy are often treated indirectly rather than as central analytic domains [83], [84], [85]. The Continuum of Human Nutrition usefully links biological status, dietary intake, household access, livelihoods, and environmental conditions, but it does not fully operationalise intrahousehold and intersectional filtering in settings such as the Sundarbans, where salinity intrusion, livelihood precarity, gendered food allocation, and uneven institutional functioning co-produce nutritional outcomes [86].

The livelihood-system framework of Niehof and Price is especially relevant because it recognises household agency, activity, purpose, quality, environment, and locus, and has been applied in deltaic Bangladesh [87], [88]. NOMI shares this concern with dynamic household systems, but foregrounds specifically nutritional mechanisms: ecological feasibility, observational adaptation, distributive equity, and epistemic legitimacy. One Health and EcoHealth approaches clarify human-animal-environment interdependence, but their application has often been stronger in disease surveillance, zoonotic risk, and antimicrobial resistance than in household diet, food culture, livelihoods, and entitlement conversion [89], [90]. The Food Systems, One Health, and Resilience approach calls for stronger integration across these domains, but does not itself provide pillar-specific indicators, causal pathways, or intervention entry points [91].

NOMI therefore functions as a middle-range framework in the Mertonian sense: more operational than a broad paradigm, but more transferable than a single case description [65]. It assembles insights from socioecological systems theory, resilience theory, entitlement approaches, intersectionality, nutrition equity, traditional ecological knowledge, and co-production scholarship into a single nutrition-focused diagnostic. Table 3 shows how field-derived themes were abductively consolidated with these theoretical traditions. The result is not a replacement for existing frameworks, but an integration layer that explains why interventions in one domain often stall unless ecological feasibility, resilience capacity, equity, and knowledge legitimacy are addressed together.

4.2. NSS: The feasibility space of diets

The NSS pillar reframes the basic diagnostic question from “Is there enough food?” to “What is actually feasible to grow, purchase, prepare, consume, and sustain, for whom, and under what conditions?” In the Sundarbans, the dietary feasibility space is constrained from below by subsistence reliance on rice and public distribution entitlements, and from above by ecological, economic, infrastructural, and institutional ceilings that limit dietary diversification even when households recognise its importance.

The core NSS mechanism is hazard stacking. As shown in Fig. 3, slow-onset stressors such as salinity intrusion, altered river hydrology, soil degradation, and riverbank erosion interact with rapid-onset shocks such as cyclones, storm surges, extreme rainfall, and flooding. These shocks synchronise failures across agriculture, fisheries, drinking water, cooking infrastructure, transport, markets, and household income. When cyclone-induced flooding increases soil salinity, reduces crop diversity, damages ponds, lowers fish availability, disrupts transport, and raises food prices, the result is not a sequence of separate problems. It is a compound nutritional constraint.

This interpretation is supported by evidence from comparable delta settings. Longitudinal analysis from Bangladesh’s climate-exposed coastal regions shows that districts affected by salinity intrusion, cyclones, and seasonal flooding experience higher child stunting and wasting, lower household dietary diversity, and greater zero-fruit intake, even after controlling for income [92]. This finding supports NOMI’s claim that climate exposure narrows diets not only through production loss, but also through fisheries decline, reduced purchasing power, market disruption, and weakened access to substitute foods.

Human-induced pressures further compress the feasibility space. Fig. 4 shows how shrimp aquaculture expansion, land-use change, deforestation, destructive fishing, pollution, and weak resource governance degrade soils, water quality, fish stocks, mangrove buffers, and safe food environments. These ecological pressures do not remain environmental problems alone. They move through livelihood, food safety, disease, income, and food-availability pathways into household nutrition. This aligns with wider evidence linking ecosystem integrity, biodiversity, aquatic resources, and micronutrient sufficiency [93], [94], [95].

Ecological change translates into nutrition outcomes primarily through livelihood pathways. Fig. 5 maps how environmental barriers interact with economic and institutional constraints, including high capital needs for aquaculture, fishing-equipment costs, credit scarcity, middlemen exploitation, weak storage and cold-chain capacity, seasonal bans, permit regimes, and territorial isolation. These constraints produce indebtedness, precarious wage labour, reduced livelihood diversity, and declining purchasing power for diverse foods. Repeated shocks can therefore generate feasibility lock-ins, where recovery consolidates around low-cost staples and episodic protein access, especially for poorer households [96], [97]. Evidence from other South Asian delta settings similarly shows that climate stress undermines food security through income stability, market connectivity, service function, and health risk, not only through production shortfall [62], [98].

The NSS pathway is also visible beyond South Asia. Brouwer’s work in Malawi showed that cooking-fuel constraints affect meal frequency and household food practice, while Leonardo’s study in Mozambique demonstrated that land and labour stratification determine whether households can absorb biofuel or livelihood schemes without compromising food self-sufficiency [99], [100]. These studies reinforce NOMI’s claim that nutrition is shaped by food, energy, labour, land, and infrastructure systems together.

Institutional and infrastructural reliability further condition whether households can convert resources into stable diets. Fig. 6 shows how unsafe or distant water sources, poor transport, limited cold chains, inadequate health facilities, and weak social-protection delivery shape both food access and food utilisation, particularly during and after shocks. This extends Sen’s entitlement approach: deprivation arises not only from food shortage, but from failures in entitlement conversion [101]. The Public Distribution System stabilises cereal access and reduces acute hunger, but its cereal-centric design cannot by itself secure dietary diversity. Its functioning also depends on electricity, mobile connectivity, transport, and administrative capacity, all of which are vulnerable during flooding and cyclones [93], [102], [103].

Food environments are therefore part of NSS. Livelihood shocks and market disruption influence the availability, affordability, convenience, desirability, and safety of foods. Under constraints of cash, fuel, time, storage, and transport, households may shift toward cheaper staples, ready-to-eat foods, or processed products. This aligns with food-environment research showing that diet quality is shaped by retail access, product mix, marketing, affordability, convenience, and social norms, not only by individual knowledge [17], [18], [104], [105], [106].

Finally, NSS includes embodied and psychosocial feasibility. Fig. 7 consolidates the pathways through which chronic illness, heat stress, pregnancy-related vulnerability, fatigue, mental distress, and repeated ecological loss influence appetite, food tolerability, meal timing, caregiving, and dietary decision-making. Evidence linking food insecurity with psychological distress and mental health outcomes supports this interpretation [107], [108], [109]. NOMI therefore treats physiological and psychological constraints not as residual household variation, but as part of the feasibility space of nutrition.

4.3. OR: Observational resilience under recurrent disruption

If NSS defines what is feasible, OR explains how households act within that constrained space. In the Sundarbans, resilience is not only infrastructural or institutional. It is observational, processual, and every day. Farmers, fishers, women, and older adults monitor tides, rainfall, soil salinity, wind direction, fish behaviour, crop conditions, fuel availability, food prices, and bodily tolerance. These observations guide decisions about storage, cooking, sharing, migration, fishing, foraging, and livelihood switching. OR therefore identifies observation itself as a resilience resource.

Household responses ranged from preparation and short-term coping to longer-term adaptation and institutional change. The classification in Fig. 8 organises these practices across anticipatory, absorptive, adaptive, and transformative capacities. Anticipatory capacities include grain storage, fuel pre-positioning, livestock movement, fodder storage, and preparation for cyclone seasons. Absorptive capacities include shared cooking, food sharing, temporary dietary simplification, use of savings, kin support, and reliance on low-cost crisis foods. Adaptive capacities include livelihood switching, seasonal migration, aquaculture adjustment, wage labour, and selective rebuilding of food-producing assets. Transformative capacities, such as collective advocacy for infrastructure, improved services, or policy change, are much rarer and depend heavily on institutional support, social capital, and political opportunity.

This four-capacity structure aligns with resilience scholarship that conceptualises coping not as a single event but as an evolving portfolio shaped by exposure frequency, resources, and institutions (Adger, 2000; Chambers, 1989; Folke, 2006; Walker et al., 2004; Barrett and Constas, 2014; Béné, 2020; Cinner et al., 2018; Tanner et al., 2015). Barrett and Constas’ welfare-based theory is especially useful because it defines resilience not as simple recovery, but as the ability to avoid unacceptable declines in well-being over time (Barrett and Constas, 2014). From this perspective, returning to a pre-shock condition is insufficient if that condition was already nutritionally fragile.

OR also clarifies the limits of resilience. Matin et al.’s concept of equitable resilience argues that resilience is more likely to be durable and just when it accounts for differential access to power, knowledge, and resources (Matin et al., 2018). In the Sundarbans, two households facing the same cyclone may have sharply different recovery trajectories depending on land tenure, women’s agency, remittance access, social networks, institutional reach, and territorial connectivity. Recent regional work on cyclone-linked resilience among smallholders in coastal West Bengal similarly shows how ecological shocks can move households into low-recovery loops unless institutional and livelihood supports are present (Panja et al., 2026).

OR is therefore inseparable from MLE. Coping strategies may sustain household food access in the short term, but they can also transfer labour, sacrifice, and dietary compromise onto those with the least power. Evidence from deltaic Bangladesh shows ordered coping sequences after major floods, moving from savings to kin loans, informal credit, and asset liquidation (Ali, 2005). Evidence from Uganda similarly shows that resilience capacities are shaped by wealth and asset position, meaning that adaptive options are not equally available across households (Wichern, 2019; Ansah, 2021). In the Sundarbans, this inequality appears in the rarity of transformative action and in the tendency of poorer households to remain trapped in absorptive cycles.

4.4. MLE: Why identical shocks produce unequal nutrition outcomes

MLE explains why the nutritional consequences of climate stress are patterned rather than random. In the Sundarbans, structural inequalities operate across intrahousehold, intersectional, intergenerational, territorial, asset-based, and institutional scales. These inequalities filter exposure, coping capacity, entitlement conversion, and diet quality.

Equity pathways explain why coping costs were distributed unevenly within and across households. As synthesised in Fig. 9 intrahousehold power relations, gendered labour burdens, land and asset inequality, territorial disparities, and intergenerational obligations distribute coping costs unevenly. Women, older adults, and sometimes in-married daughters-in-law often absorb scarcity by eating last, eating less, reducing protein intake, or skipping meals. This makes them hidden nutritional shock absorbers within household food insecurity. Household-level indicators can miss this pattern because they often record whether the household has food, not who within the household sacrifices dietary quality.

MLE extends entitlement theory by showing that entitlement conversion is socially filtered [101]. A household may receive subsidised grain, but the nutritional value of that entitlement depends on who controls income, who decides food purchases, who cooks, who eats first, whether transport is available, whether fuel and water are secure, and whether protein-rich foods are affordable. Formal access does not automatically become dietary adequacy.

The intersectional dimension is equally important. Gender does not operate alone. It intersects with caste, class, age, widowhood, landlessness, migration status, household position, and territorial isolation. A lower-caste widow in a landless household in an isolated riverine block faces a different vulnerability profile from a woman in a better-connected household with remittance income and land access. NOMI therefore treats vulnerability as compound rather than additive, aligning with intersectionality theory and nutrition-equity literature that identifies structural, territorial, and intrahousehold inequalities as core drivers of nutrition outcomes [13], [110].

Land and asset relations further shape nutritional resilience. Evidence from rural China shows that perceived land-tenure security can influence migration decisions and agricultural input behaviour [111]. This is relevant to the Sundarbans because land insecurity, erosion, tenancy, and salinity-induced degradation influence whether households can remain agriculturally viable, invest in adaptation, or rely on migration. Comparative evidence from coastal Philippine villages and deltaic Bangladesh also suggests that territorial, gendered, and asset-based inequalities are common features of climate-exposed, market-integrated rural settings [87], [112].

MLE also clarifies why nutrition programmes that target women may fail if they increase responsibility without increasing agency. Women in this study carried substantial responsibility for food procurement, cooking, rationing, and care, but often had limited income autonomy and constrained decision-making authority. This configuration of high responsibility and low power creates a programme risk: women become delivery vehicles for nutrition interventions without gaining control over the resources required to change diets. Evidence that women’s income control and financial decision-making are strongly associated with dietary diversity reinforces this point [113], [114], [115]. At a wider food-system level, weak government support and response remain systemic bottlenecks for nutrition-sensitive food systems [116].

4.5. IK: Knowledge as a condition of nutritional resilience

IK shows that nutritional resilience is not only material or institutional. It is also epistemic. Households sustain diets through knowledge of seasons, soils, fish, wild greens, fermentation, preservation, cooking techniques, bodily tolerance, ritual rules, and culturally acceptable substitutions. This knowledge is not static folklore. It is an adaptive system built through observation, experimentation, memory, and social transmission.

Local food knowledge in the Sundarbans showed both continuity and erosion. As shown in Fig. 10continuity is visible in seasonal dietary shifts, culinary biodiversity, fermentation and preservation practices, ecological-spiritual food rules, and indigenous preparation techniques. Erosion is visible in the loss of traditional tools such as the dhenki, declining intergenerational transmission, migration, ecosystem change, rising reliance on purchased foods, processed food penetration, and weakening of shared preparation spaces. These changes are nutritionally significant because they alter the foods households know how to identify, process, preserve, value, and safely consume.

Berkes’ concept of traditional ecological knowledge is directly applicable because it defines TEK as an adaptive management system refined through human-environment interaction [117], [118]. In the Sundarbans, this knowledge includes seasonal fish availability by species and microhabitat, salt-tolerant edible plants such as Giriya shak, crisis foods, fermentation and preservation practices, and dietary substitution protocols during lean periods or after cyclones. Kuhnlein and Receveur’s multi-community evidence shows that transitions from traditional to market-based food systems often reduce dietary diversity and micronutrient adequacy even when caloric intake is maintained [119], [120]. Vijayan et al. similarly warn that interventions can unintentionally accelerate indigenous food-system erosion if they ignore migration, market integration, and climate-driven ecosystem shifts [121].

Community aspirations and intervention priorities extended beyond information-based nutrition education to the practical conditions that support dietary change. As summarised in Fig. 11, participants emphasised nutrition education grounded in local foods, practical food skills, kitchen gardens, community seed banks, indigenous food revival, stronger local production, women’s leadership, livelihood options, and better-functioning services. This figure is important because it shows that communities did not frame nutrition only as a problem of information deficiency. They understood nutrition improvement as requiring knowledge, services, livelihoods, water, fuel, and institutional support together.

IK therefore requires co-production rather than simple knowledge transfer. Community members, especially women, elders, fishers, farmers, and local cooks, should be treated as knowledge partners who help define feasible foods, acceptable recipes, seasonal constraints, intervention risks, and locally meaningful indicators of success. Norström et al.’s principles of knowledge co-production inclusive, process-oriented, solution-oriented, and transformative provide the methodological foundation for such an approach [122]. Ostrom’s work on polycentric governance reinforces this point: durable resource management depends on local knowledge, participation, and institutional fit [82].

4.6. Cross-pillar integration

The analytical strength of NOMI lies in the links between pillars. As shown in Fig. 12 and Table 2, NSS defines the material feasibility space; OR explains how households navigate that space over time; MLE reveals who bears the costs of coping and who benefits from adaptation; and IK determines which adaptive options are known, legitimate, culturally acceptable, and transmissible. Table 3 further consolidates this abductive integration by linking field-derived themes with socioecological, resilience, equity, entitlement, and knowledge-system literatures.

These pillars interact through feedback loops. Ecological degradation under NSS can erode IK by eliminating the ecosystems that local knowledge describes. OR can deepen MLE when coping depends on women’s unpaid labour, dietary sacrifice, or intensified care burdens. IK erosion can weaken OR because future generations lose the observational intelligence needed for anticipatory action. MLE can constrain IK because marginalised knowledge holders may be excluded from programme design even when they hold crucial adaptive knowledge.

This explains why single-pillar interventions often fail. A crop-diversification programme may not improve diet quality if women lack control over production decisions, saline water limits cultivation, market access is poor, or the promoted crop is not culturally acceptable. A nutrition education campaign may fail if households lack fuel, water, cash, storage, or seasonal access to the recommended foods. A social-protection intervention may stabilise calories but not dietary diversity if it remains cereal-centric and disconnected from local food environments. NOMI makes these interdependencies visible and therefore actionable.

4.7. NOMI and one health

NOMI also strengthens the nutrition dimension of One Health. One Health recognises that human, animal, and ecosystem health are interdependent [43], [44], [90]. In the Sundarbans, this interdependence is visible in three pathways. First, environment-to-human pathways include salinity intrusion, unsafe water, flooding, heat stress, and storm damage to cooking spaces. Second, environment-to-animal-to-human pathways include fisheries decline, aquaculture stress, livestock vulnerability, and reduced access to animal-source foods. Third, human-to-environment feedbacks include destructive fishing, intensive shrimp farming, land-use change, pollution, and mangrove degradation.

These One Health pathways are already embedded in the NSS architecture shown in Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, and their social distribution is clarified through OR, MLE, and IK in Fig. 8, Fig. 9, Fig. 10, Fig. 11, Fig. 12. NOMI does not replace One Health. Rather, it operationalises the food-environment-equity-knowledge corner of One Health. Classic One Health and EcoHealth approaches have often been stronger in disease surveillance and zoonotic risk than in household diet, food culture, livelihoods, and entitlement conversion [89]. Milazzo et al.’s review of One Health programmes found that many programmes involved human and animal sectors while failing to adequately incorporate the environment sector, which is precisely the gap NOMI’s NSS pillar addresses [123]. The FOR approach argues for integrating food systems, One Health, and resilience; NOMI advances this proposition by providing operational pillars, causal pathways, and intervention entry points [91].

4.8. Methodological contribution

Methodologically, NOMI is grounded in a transparent chain of evidence from household interviews, focus-group discussions, expert consultations, and ethnographic observations to codes, themes, and framework pillars, as summarised in Fig. 2. This traceability strengthens the credibility of the framework and shifts the Discussion from broad domain listing toward pathway diagnosis. It also responds to critiques that food security measures are often strong in comparability but weaker in causal interpretation and intervention design [29], [124], [125].

The abductive analytic strategy was central. By moving iteratively between empirical themes and theoretical constructs, the study retained contextual depth while producing transferable mechanisms [72], [73]. Multiple coding of excerpts was not a weakness but an analytic necessity because many field observations were simultaneously ecological, social, institutional, and epistemic. This approach is consistent with qualitative coding and reflexive thematic analysis traditions that allow layered meaning rather than forcing complex evidence into single categories [74], [77]

The use of multiple data sources strengthened the analysis. Household interviews captured lived experience and decision-making. Focus-group discussions surfaced shared norms and community negotiation. Expert consultations provided institutional and sectoral interpretation. Field notes documented cooking spaces, water access, fuel use, market conditions, landscape features, and material food environments. Together, these sources allowed NOMI to connect narrative evidence with visible socioecological conditions.

4.9. Policy and programme implications

NOMI supports an education-plus-systems approach to food and nutrition insecurity in climate-vulnerable settings. Fig. 13 presents NOMI as a five-step workflow in which NSS specifies the socioecological conditions that make diets feasible or infeasible, while OR, MLE, and IK operate as interacting filters that determine whether education or programme inputs can translate into sustained improvement. Used this way, NOMI does not replace standard food security metrics. It functions as an interpretation layer that clarifies why outcomes change or stagnate, under what conditions, for whom, and with what supports [126].

Fig. 13.

NOMI application workflow for education plus systems programming.

Programmes specify the decision and minimum viable change, diagnose mechanism level drivers using NOMI, design education paired with enabling supports, build a pillar-based theory of change with a small set of mechanism proximal measures, and implement through adaptive learning with documented feedback loops. NSS represents the system core defining socioecological feasibility, while OR, MLE, and IK shape translation to nutrition outcomes through stability under disturbance, equity filters, and knowledge legitimacy. Solid lines indicate the primary pathway, light lines indicate feedback loops, and dashed lines indicate interactions.

For NSS, climate adaptation should be treated as nutrition policy. Embankment resilience, salinity-tolerant crops, safe drinking water, cooking fuel access, transport, market connectivity, cold storage, fisheries governance, and local food production are nutrition-relevant investments. The PDS should continue to stabilise cereal access, but it must be complemented by nutrition-sensitive social protection and local food-environment strengthening.

For OR, resilience programming should support the practices households already use while reducing their unequal costs. Predictable social protection before cyclone seasons, timely compensation, reliable pension payments, functioning PDS during floods, fuel and water preparedness, and locally managed food reserves can strengthen anticipatory and absorptive capacity. Programmes should document and amplify observational knowledge held by women, elders, fishers, and farmers rather than assuming that external information automatically supersedes local intelligence.

For MLE, equity must be treated as a structural condition for nutritional improvement. Programmes should explicitly assess who eats last, who controls cash, who carries additional labour, who can access markets, and which territories remain underserved. Landless households, isolated island communities, widows, older adults, tribal groups, and women with limited decision-making power require more than generic household-level targeting. Interventions that assign women responsibility without expanding their control over resources risk reproducing inequity.

For IK, nutrition education should be co-produced with local knowledge holders. Community cooking demonstrations, school-based local food literacy, participatory recipe documentation, kitchen gardens, community seed banks, seasonal food calendars, and Indigenous food revival can preserve knowledge while improving dietary agency. These priorities align with the community aspirations shown in Fig. 11 and can be operationalised through the workflow shown in Fig. 13.

Across all four pillars, NOMI shows that effective nutrition programming in climate-exposed deltas must coordinate nutrition, livelihoods, infrastructure, social protection, environmental governance, and knowledge systems. Its practical value lies in helping policymakers and practitioners identify which cross-domain constraint is binding in a specific context, which group bears the cost, which local knowledge can support change, and which intervention can realistically expand the feasibility space for diverse, culturally acceptable, and nutritionally adequate diets.

4.10. Limitations and future research

Several limitations shape the claims of this study. First, as an exploratory qualitative study, the findings support analytic rather than statistical generalisation. The framework identifies mechanisms and relationships, but does not estimate prevalence or effect size. Second, data collection covered a limited time period. The most acute conditions of peak monsoon and cyclone impact were captured partly through retrospective accounts rather than continuous observation, which may underrepresent the full duration and intensity of dietary contraction during severe shocks.

Third, the expert consultation sample was small and partly shaped by existing networks. Greater inclusion of state-level policy actors, private market actors, fisheries regulators, disaster-management officials, and social-protection administrators would strengthen the institutional analysis of NSS and MLE. Fourth, evidence on highly marginalised Indigenous groups, including Lodha households, requires more direct participatory engagement to avoid relying on outsider accounts of survival practices. Fifth, NOMI requires further operationalisation through pillar-specific indicators for routine monitoring, comparative research, and programme evaluation.

These limitations delineate the next phase of the research agenda. The present study constitutes the framework-generating phase of NOMI. The next phase will examine the quantitative reach of the framework using district-level and household-level datasets such as NFHS-5 and NSSO HCES 2022–23, mapping qualitative mechanisms to measurable indicators and classifying evidence as convergence, complementarity, or divergence. A subsequent applied phase will translate the validated framework into a human-in-the-loop application protocol that allows new districts or programmes to conduct an auditable NOMI diagnosis without repeating the full qualitative cycle. In this sequence, the present study builds the mechanism architecture, the future quantitative phase tests its empirical reach, and the applied phase converts the framework into a practical diagnostic tool.

5. Conclusion

This study developed the NOMI framework as a mechanism-oriented, middle-range approach for understanding food and nutrition security in climate-exposed delta systems. Grounded in qualitative evidence from the Indian Sundarbans, NOMI shows that nutrition insecurity is not produced by isolated deficits in food availability or access alone, but through interacting ecological, social, institutional, livelihood, and knowledge processes. By integrating Nutrition as a Socioecological System, Observational Resilience, Multi-Lens Equity, and Inclusive Knowledge, the framework explains how dietary feasibility, adaptive capacity, unequal vulnerability, and knowledge recognition jointly shape nutrition outcomes under chronic environmental stress.

The findings demonstrate that food and nutrition security in climate-exposed deltas requires a shift from sectoral and indicator-based approaches toward cross-domain analysis. Climate adaptation strengthens nutrition only when it stabilises livelihoods and food environments. Social protection builds resilience when it is predictable, inclusive, and responsive to local disruption. Nutrition education becomes meaningful when aligned with ecological feasibility, household negotiation, and local food knowledge. Equity must therefore be treated not as a targeting afterthought, but as a structural condition determining whether interventions translate into improved dietary and health outcomes.

NOMI also strengthens the operational relevance of One Health by tracing the links between environment, livelihood, food systems, diet, and health, while identifying resilience mechanisms and equity filters along these pathways. Although this qualitative study supports analytic rather than statistical generalisation, the framework offers a transferable template for examining nutrition in other climate-vulnerable coastal and delta settings. Future research should operationalise the NOMI pillars through indicators, test NOMI-informed theories of change, and examine how institutions can support coordinated action across health, food systems, livelihoods, environment, and governance.

In conclusion, NOMI reframes food and nutrition insecurity in climate-exposed deltas as a dynamic and stratified system shaped by interacting ecological, social, institutional, and knowledge processes. As a middle-range explanatory framework, it bridges broad paradigms and grounded empirical analysis, providing a practical foundation for designing more resilient, equitable, and context-sensitive nutrition interventions in vulnerable coastal regions.

CRediT authorship contribution statement

Namita Das: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Devastotra Poddar: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Ethical considerations

Ethical approval was obtained from the Institutional Human Ethics Committee (IHEC), Belda College (Certificate Ref. BC/IHEC/ECC/10/24, dated 10 Aug 2024; IHEC meeting 08 Aug 2024). Oral informed consent was secured from all participants, in recognition of local cultural practices where written consent may be intimidating or misunderstood. Participants were informed of the study’s purpose, voluntary nature, and confidentiality protections. Identifiers have been anonymized.

Funding

This work forms part of the doctoral research of Ms. Namita Das at the Natural Science Research Centre, Belda College (Vidyasagar University, Reg. No. 1362/Ph.D./Sc., 22.05.2024), supported by the Swami Vivekananda Merit-cum-Means Scholarship (SVMCM) and the Samposhyam Foundation.

Declaration of competing interest

None.