Factors Influencing Nutritional Adequacy among Rural Households in Nigeria: How Does Dietary Diversity Stand among Influencers?

ABSTRACT

This study examined the influence of food consumption diversity on adequate intakes of food calories, proteins and micronutrients among households in rural Nigeria within the framework of panel data econometrics using a nationally representative data. We found that substantial proportion of households suffered deficiency of calories, proteins and certain micronutrients; with higher percentage of sufferer households occurring in the post-planting season. The different measures of dietary diversity (constructed and used for analysis) consistently indicate significant and positive influence of dietary diversity on the likelihood of adequate consumption of food nutrients. While higher level of income, education and non-farm enterprise engagement may strongly stimulate adequate nutrient intakes, increases in the number of adolescents would substantially diminish it. Although our findings call for renewed attention on diet diverseness, we stress the complementary/synergistic roles of education and rural income improvement, especially through non-farm enterprise diversification in tackling multiple nutritional deficiencies in rural Nigeria.

KEYWORDS:

• Dietary diversity
• food security
• Nigeria
• nonfarm income
• nutrition
• rural households

Notes

¹ There are situations where certain individuals in households regarded as nutrient deficient would meet their food nutrient requirements while some persons living in households categorized as nutritionally adequate may still have their nutrient needs unsatisfied. Since households belonging to the two categories are not selected in any systematic manner, the two possible errors of misclassification are presumed to be randomly distributed, with their effects assumed to cancel out such that the data can be used to estimate the proportion of people who are nutritionally deficient in the population (Smith and Subandoro 2007).

² The use of a pooled probit model was motivated by the following. First, the Hausman tests of model comparison between estimated fixed-effects and random-effects models established the suitability of random-effects models (over the fixed-effects models) for all the nutrients examined. Second, the likelihood ratio tests associated with the random effects of the probit models estimated for most of the nutrients examined suggest that the intraclass error correlation in the latent variable (of the random-effects probit model) is not significantly different from zero, meaning that the random effects are substantially lower in comparison with the idiosyncratic error. Hence, a pooled model (albeit corrected for autocorrelations) can be employed for analysis with results unlikely to differ significantly from those of the random-effect probit model. Third, the estimated pooled probit models yielded more meaningful results than the pooled logit models.

³ In terms of the expected influence of the explanatory variables on outcomes, the signs of coefficients of zonal dummies, age and sex classifications, household size composition, and nonfarm enterprise diversification could be negative or positive while the coefficients associated with food consumption diversity, education, and income classification dummies are expected to be positive.

⁴ Calcium and iron were excluded because they have lower rates of deficiency compared to other minerals (such as phosphorous and potassium). Prevalence of iron deficiency is the lowest compared to the deficiencies of other nutrients, and that of calcium deficiency may be understated because calcium from drinking water are not captured. Apart from the fact that vitamins B1, B3 (niacin), and B6 (which were excluded) perform some basic functions similar to vitamin B12 (which was included), they also had lower rates of deficiency across seasons than vitamin B12. Some key variables (predictors) with similar influence across all nutrients are X₁₇, X₁₈, and X₁₉.