Search results1 – 3 of 3
The purpose of this paper is to assess food security as a risk factor in the development of poor mental health among younger populations in the USA over an eight-year…
The purpose of this paper is to assess food security as a risk factor in the development of poor mental health among younger populations in the USA over an eight-year period using a nationally representative cross-sectional sample.
Using data from individuals who participated in the National Health and Nutrition Examination Survey between 2005 and 2012, respondents were classified as either having “poor mental health” or “good mental health.” Multivariate logistic regression models based on this dichotomy are employed to estimate the odds ratios in the association of household food security and mental health using three cut-off points that correspond to these models.
Respondents from very low food security had higher odds (OR=2.06, p<0.0001; OR=1.98, p<0.0001; OR=1.94, p=0.01) of suffering from poor mental health compared with participants from fully food secure households. These findings indicate the robustness of the results across all three separate regression models.
Causality cannot be determined from the cross-sectional design. Although potential endogeneity could invalidate the conclusions, these findings inform public policy that food security is a contributory factor in the development of poor mental health at an early age. It suggests that interventions to alleviate food insecurity could improve mental health among younger populations in the USA.
Several cut-off points are developed to distinguish between “poor” and “good” mental health to assess the robustness of the findings. This approach has the potential to minimize the misclassification of mental health outcomes. Very low food security is a strong predictor of poor mental health regardless of the cut-off point used.
Confounding is of central importance in epidemiologic studies. Its definition has been under wide debate over the past decades. The classical definition is straightforward…
Confounding is of central importance in epidemiologic studies. Its definition has been under wide debate over the past decades. The classical definition is straightforward and easy-to-implement. Nevertheless, it is data-driven and has drawbacks. The more recent counterfactual definition captures the essential roles a confounder plays in causal inference. It would be beneficial for researchers to grasp substantive knowledge in causal structure and broadly adopt the latter definition. There are various methods of handling confounding issues. The choice of one option over another depends on various factors, including the nature of the study, sample size and rarity of events.