Gender inequality and food security: a global comparative analysis

After years of decreases in global hunger, recent estimates from the Food and Agriculture Organisation showed that the number of undernourished people increased to 815 million in 2016 from 777 million in 2015. This reversal in trends suggests that new approaches are needed to tackle global food insecurity. Studies have shown that one potential approach is to reduce gender inequality; however, there are no studies looking at the global level and across regions, with internationally validated food security tools. This study fills this research gap by exploring the association between gender inequality and food security at the global level and across regions. This quantitative, cross-sectional study used 2015 data from three sources: Gallup World Poll, United Nations Development Programme, and the World Bank. The sample included 119 countries and 116,077 respondents. Food security was assessed using the Food Insecurity Experience Scale, and gender inequality was assessed using the Gender Inequality Index. Two thresholds were developed to assess food insecurity: the prevalence of moderate or severe food insecurity, and the prevalence of severe food insecurity only. In order to ensure international comparability, a common global reference was used to calibrate individual responses, and generate comparable estimates of food insecurity at the global level. Countries were classified into seven regions based on the World Bank regional classification: East Asia and the Pacific (EAP), Europe and Central Asia (ECA), Latin America and the Caribbean (LAC), Middle East and North Africa (MENA), North America (NAM), South Asia (SA), and sub-Saharan Africa (SSA). Spearman's rank correlation and generalized linear models were used to assess the association between gender inequality and prevalence of food insecurity. Models were adjusted for health (assessed using the personal health index). Only significant factors (p < 0.05) were retained in the multivariate models. Statistical analyses were conducted using R and SAS with statistical significance set at p < 0.05. Findings from the Spearman's rank correlation showed that gender inequality was significantly associated with moderate or severe food insecurity (rs = 0.88, p < 0.001) and severe food insecurity (rs = 0.79, p < 0.001). Results from the generalized linear models showed that gender inequality was significantly associated with both thresholds of food insecurity at the global level (p < 0.0001). However, great heterogeneity was observed across regions. While gender inequality was a significant predictor of the prevalence of moderate or severe food insecurity in EAP, ECA, LAC, and MENA, it was not significantly associated in SA and SSA. When analyses were conducted only for severe food insecurity, the effect of gender inequality was significant in EAP, LAC, and SA, but not in ECA, MENA, and SSA. Multivariate findings showed that when controlling for health, there was a significant positive association between gender inequality and the prevalence of food insecurity globally. At the regional level, there was also variability in the degree of significance in the relationship. Gender inequality was a significant predictor of moderate or severe food insecurity in three regions (EAP, ECA, and LAC). For severe food insecurity, gender inequality had a significant effect in EAP, SA, and SSA. Results from this study suggest that health was the main predictor of food insecurity in region where the effect of gender inequality on food insecurity was not significant, which was the region with greater levels of gender inequality (i.e., SSA). This means that there are other mitigating factors, such as climate change, which could explain the lack of significance in some regions. To our knowledge, this is the first study examining the association between gender inequality and food security at the global and regional levels.Bien que la faim ait connu une réduction pendant plusieurs années, des estimations récentes de l'Organisation des Nations Unies pour l'alimentation et l'agriculture ont indiqué que le nombre de personnes sous-alimentées est passé de 777 millions en 2015 à 815 millions en 2016. Cette inversion des tendances suggère que de nouvelles approches seraient nécessaires pour lutter contre l'insécurité alimentaire mondiale. De nombreuses études ont suggéré qu'une des approches possibles consisterait à réduire l'inégalité entre les sexes. Cependant, aucune étude n'utilise des outils de sécurité alimentaire validés au niveau international pour étudier des situations aux niveaux global et régional. Cette étude comble ce manque de recherche en explorant l'association entre l'inégalité entre les sexes et la sécurité alimentaire aux niveaux mondial et régional. Cette étude quantitative et transversale a analysé les données du Gallup World Poll tirées du Programme des Nations Unies pour le développement, ainsi que celles de la Banque mondiale pour 2015. L'échantillon comprenait 119 pays et 116 077 répondants. La sécurité alimentaire a été mesurée à l'aide de l'échelle de mesure de l'insécurité alimentaire vécue « Food Insecurity Experience Scale », et l'inégalité entre les sexes a été évaluée à l'aide de l'indice d'inégalité de genre « Gender Inequality Index ». Les pays ont été classés en sept régions selon la classification régionale de la Banque mondiale: Asie orientale et Pacifique (EAP), Europe et Asie centrale (ECA), Amérique latine et Caraïbes (LAC), Moyen-Orient et Afrique du Nord (MENA), Amérique du Nord (NAM), Asie du Sud (SA) et Afrique subsaharienne (SSA). Les corrélations de Spearman et les modèles linéaires généralisés ont été utilisés pour évaluer l'association entre l'inégalité entre les sexes et la prévalence de l'insécurité alimentaire. Les modèles ont été ajustés pour la santé (évaluée en utilisant l'indice de santé personnel). Seuls les facteurs significatifs (p < 0,05) ont été retenus dans les modèles multivariés. Les analyses statistiques ont été menées en utilisant R et SAS avec une signification statistique fixée à p < 0,05. Les résultats des corrélations de Spearman ont montré que l'inégalité entre les sexes était associée significativement à une insécurité alimentaire modérée ou sévère (rs = 0,88, p < 0,001) et une insécurité alimentaire sévère (rs = 0,79, p < 0,001). Les résultats des modèles linéaires généralisés ont montré que l'inégalité entre les sexes était associée significativement aux deux seuils d'insécurité alimentaire au niveau mondial (p