OBJECTIVE: This study assessed the diet quality of households by their type of engagement in homestead aquaculture and/or horticulture. Socio-demographic determinants of diet quality were also studied.

METHOD: Diet quality was assessed using a nutrient adequacy ratio (NAR), based on the preceding 7 days' dietary recall at the household level. Adult male equivalent units (AMEs) were used for age- and sex-specific intra-household distribution of household intakes. Mean adequacy ratios (MAR) were computed as an overall measure of diet quality, using NAR.

RESULTS: Better diet quality (mean ± SD) was associated with households engaged in both homestead aquaculture and horticulture (0.43 ± 0.23; p < 0.001) compared to only one type of agriculture (0.38 ± 0.20) or none (0.36 ± 0.20). Tukey's post-hoc test confirmed significant differences in diet quality between both and either engagement (0.05 ± 0.01, p < 0.001), both and no engagement (0.07 ± 0.01, p < 0.001), and either and no engagement households (0.02 ± 0.01, p < 0.001). Beyond farm production of nutrient-rich foods, generalized estimating equations showed that diet quality was influenced by the higher educational level and occupation of adult household members, higher daily per capita food expenditure, sex, family size and region.