A latent growth curve modelling approach to seasonal and spatial dynamics of food security heterogeneities in rural Lake Naivasha Basin, Kenya.

The increasing complexity of food insecurity, malnutrition, and chronic poverty faced by Sub-Saharan Africa warrants urgent categorisation and tracking of household food security along both temporal and spatial dimensions. This will help to effectively target, monitor and evaluate population-level programs and specific interventions aimed at addressing food insecurity. Traditional longitudinal analysis does not address the dynamics of inter- and intrahousehold heterogeneities within the seasonal and spatial context of household-level food security. This study is the first to overcome such limitations by adopting a multi-group piecewise latent growth curve model in the analysis of the food security situation in a statistically representative sample of 601 households involved in subsistence and cut-flower commercial agriculture, around Lake Naivasha. We considered food security as a latent concept, which manifests as food security outcomes in our primary longitudinal dataset from March 2018 to January 2019. Our analysis highlights the temporal and spatial dynamics of food security and advances new evidence on inter- and intrahousehold heterogeneities in food security across different seasons for the subsistence and commercial farming clusters. These heterogeneities were demonstrated primarily during the hunger season from March to June, and persisted in both the clusters and across months, albeit with different intensities. Moreover, our results indicate the importance of commercial agriculture in achieving food security in the hunger season. Our study suggests the need of a multidisciplinary approach to food security and the introduction of well-coordinated interventions for the development of subsistence and commercial agriculture considering the seasonal and cluster-level specificities.