Regarding the adverse effect of income inequality

Regarding the adverse effect of income inequality on nutritional status, Subramanian et al. (2007) suggest in their study of Indian society that income inequality can be a marker of both resource maldistribution and inefficient public policy. It is likely that unequal areas are the places where the privileged over-consume while the underprivileged face food insecurity. Equally likely is that, due to the low social purchase SB 203580 hydrochloride as well as other negative externalities associated with a highly skewed income distribution, public policy in a less egalitarian society is prone to manipulation by vested interests, resulting in poor provision of the amenities that are vital for combating malnutrition. In our research, we also found a paradoxical protective effect of facility deprivation on under-nutrition. We initially suspected this to be an artefact of confounding, but it remains unresolved even after fitting multivariate models. While puzzling, this is not an isolated observation (Thomas & Strauss, 1992; Wolff & Maliki, 2008). Perhaps this is attributable to the endogenous, non-random spatial distribution of government programs as a result of the historical priority on placing health facilities and interventions in less healthy areas (Pitt, Rosenzweig, & Gibbons, 1995). Unfortunately, this puzzle cannot simply be addressed using the cross-sectional data we have at hand; it may therefore be pursued further in future research. Other limitations of this study must now be acknowledged. The cross-sectional data that we have do not permit us to incorporate the temporal dimension into our analysis. As a consequence, this study only provides a snapshot capturing the determinants and geographical variations of the double burden of malnutrition in Indonesia in the year 2007. It is known that the burden of obesity gradually shifts to the poor as a nation progresses economically (Brown & Konner, 1987; Popkin, 1998). Whether such a shift has begun to occur in Indonesia is indeed an interesting subject to study, but carrying out the relevant research obviously necessitates the availability of newer data. Another limitation is that the statistical models fitted in this study did not explicitly account for spatial-contextual autocorrelation which may, to some extent, affect the precision as well as the smoothness of the estimated risks. The importance of undertaking such an endeavour cannot be underestimated, but it clearly deserves its own avenue in the vast literature of spatial epidemiology. Despite these limitations, this study does, however, contribute to the literature in several ways. This study is among the few to consider the double burden of malnutrition in Indonesia from the perspective of the general population. As noted earlier, all existing studies have focused rather specifically on Indonesian women (Winkvist et al., 2000) or households (Doak et al., 2005; Oddo et al., 2012; Roemling & Qaim, 2013; Vaezghasemi et al., 2014). This study also adds to the literature by showing that the influence of contextual macro-economic conditions (income inequality and level of economic development) is not negligible with regard to the nutritional well-being of individuals (Block et al., 2004). In addition, this study provides the literature with a principled characterisation of the spatial distribution of nutritional vulnerability within the 17,000-island Indonesian archipelago which, we believe, is indispensable for the purpose of policy targeting. Of course, in the absence of good data, this study would not have been able to offer the present analysis. If any policy implications for dealing with the double burden of malnutrition are to be suggested from the findings of this study, then they should include the following points. Raising the overall level of the socio-economic status of the population through education, employment, and income-enhancing opportunities can help to improve purchasing power, which, in turn, enables individuals to afford enough food to fulfil their needs. That alone, however, is not sufficient; we have already seen that the risk of over-nutrition also increases with every improvement in socio-economic conditions. Therefore, there is a need for a wider public educational campaign that promotes behavioural changes especially in, but not limited to, the spheres of physical activity, dietary pattern and gender equality (Roemling & Qaim, 2012). Furthermore, the need for better nutritional education (Webb & Block, 2004) in academic curricula cannot be overstated as it has become apparent that, at least in our models, more schooling is not always correlated with better nutritional status. Better nutritional education, of course, will not only facilitate behaviour change but also help to shape a healthier body image in society. Simultaneously, as it has been projected that by the year 2030 more people in developing countries will live in cities than in rural areas (Cohen, 2006), the obesogenic urban environment must also be addressed. A recent assessment of Indonesia׳s built environment indicates an environment ‘that is fairly unfriendly to pedestrian physical activity with limited access to healthy foods’ (Shrimpton & Rokx, 2013: 3). This hints that improvement in nutritional health can also be achieved through the provision of a healthier urban planning initiative.