Ecuador household survey links farm diversity with lower market food dependence

Three food-acquisition pathways were measured separately

A study published in Frontiers in Nutrition on 8 September 2026 examined production, consumption from own production and food purchase among 133 smallholder farming households in Los Ríos, Ecuador. Households were recruited through community field activities in April 2025 using non-probability methods. The authors describe household-level associations rather than representative estimates for the province or its cantons.

The same ten food groups were used to classify each pathway. A household could produce, self-consume and purchase foods within the same group. Median production diversity was three groups, while median diversity reportedly consumed from own production was two. The survey did not use a standardised dietary recall period, so the self-consumption measure is a food-group proxy rather than a validated measure of total dietary diversity.

Exact overlap distinguishes production from household retention

The new Farm-to-Diet Coupling Index counts groups reported as both produced and self-consumed by the same household, divided by its number of produced groups. Its median was 0.50. It measures correspondence between food-group inventories, without establishing quantities consumed or the economic value of retained production. Two households produced no primary food groups and therefore had no defined index.

Eighteen households reported at least one self-consumed group without a matching production record. Those responses were retained for auditing but excluded from the exact-overlap numerator. The authors consider the index a locally constructed operational measure requiring independent validation. They note that a single matching or non-matching group can substantially affect it when a household produces few groups.

Vegetables, fish and red meat showed the largest positive differences between purchase and production prevalence. Vegetables were purchased by 89 households (66.9%) and produced by 26 (19.5%). Fruits and the combined cereals, grains and legumes group were more often produced than purchased. The authors stress that market acquisition can complement production by providing foods unavailable locally or seasonally.

The main market-dependence association varied across analyses

The Market Food Dependency Index counted purchased groups as a share of groups acquired through self-consumption or purchase; its median was 0.67. High market dependence was operationally defined as at least 0.75. In the main adjusted model, 49 of 124 households met that threshold. Each additional produced group was associated with a lower prevalence of high dependence (adjusted prevalence ratio 0.708; 95% confidence interval 0.603–0.832).

Each 0.10-unit increase in exact overlap had an adjusted prevalence ratio of 0.933 (95% confidence interval 0.874–0.996). The authors describe this association as modest and less precise in broader adjustment, canton-clustered and penalised analyses. With an alternative high-dependence threshold of 0.50, neither overlap nor production diversity was statistically distinguishable from the null. Continuous-outcome analyses supplied complementary inverse associations.

Overlap did not consistently indicate wider food-system outcomes

Exact overlap was not independently associated with food insecurity, market-dependent vulnerability or low food-system buffer resilience in the core analyses. The authors caution against treating production diversity as equivalent to dietary diversity or food security, or interpreting the findings as support for replacing markets with own production. The cross-sectional design cannot establish temporal order or causality.

The authors identify self-reporting, a modest nonrepresentative sample, classification uncertainty and incomplete measurement of income, prices and market access as limitations. One month of observations could not capture seasonal changes. They call for independent validation, probability-based and longitudinal studies, repeated seasonal measurements and standardised dietary assessment. ECOTEC University funded the study and covered the publication charge; the authors declared no commercial or financial conflicts of interest.