Although global ID burden has decreased consistently over the decades, we still identified ID inequality in each geographic region and over all, moreover, the inequality has been expanding since 1990. We also observed heavier ID burden in countries with lower HDI, higher ID burden in women than that in men, and smaller between-sex gap of ID burden in countries with higher HDI. Last, the ID inequality seems more severe among men than that in women across the years. To our knowledge, this is among the first studies that examined the overlooked ID inequality issue globally, in different geographic regions, and in each sex.
Although major global efforts have been mounted to address the ID burden in women, sex-related disparity still exists. Despite the narrowed gap of age-standardized DALY rates between women and men, the ID burden has been consistently high in women than that in men. Efforts are still needed to close the gap. Previous studies have found that SES contributes to the lowered ID burden in women, for example, Bentley et al. found that high SES is a protective factor for anemia in Indian women, and Bharati et al. reported that the lower hemoglobin levels in Indian women than men was partly attributed to sex discrimination. This coincides with our observation where the ID burden for women lowered with higher HDI. Therefore, improving SES in women, e.g., by providing universal education and equal salary, will help reduce ID burden.
We observed rather persistent ID burden among men in countries of low- and medium-HDI groups. Interestingly, we also observed higher ID burden among boys than girls younger than 10 years old. This finding is consistent with a previous multi-country survey conducted in Tanzania, Mozambique, Ghana, Malawi, and Indonesia, where boys who were 12–14 years old were more likely to be anemic than girls of the same age . Gupta et al. also observed high ID prevalence in adolescent boys .
Previous studies have called for close collaboration between government and schools for both adolescent girls and boys with ID. For example, India started weekly iron-folic acid (IFA) supplementation in selected schools of urban Puducherry in 2012, where the prevalence of anemia was found to be 62.7% . Further, we noticed significantly fewer ID alleviating programs aiming at boys than girls according to the Global database on the Implementation of Nutrition Action (GINA) that monitors nutrition policies and programmes in 202 countries, and the literatures included in a meta-analysis which focused on interventions to improve adolescent nutrition . We can easily identify programmes such as the Girls’ Iron-Folate Tablet Supplementation (GIFTS) program in Ghana, which is aiming at adolescent girls, but it was difficult to identify similarly highlighted programmes for adolescent boys. To close the gap of ID in adolescent boys, we strongly recommend countries that are suffering from high ID burden pay attention to adolescent boys in making their ID alleviation plans.
East Asia & Pacific and South Asia regions have made big strides in controlling ID, while Sub-Saharan Africa countries still need to work out a strategy to lower their ID burden. China is the among the countries in East Asia region that made the most improvement; the age-standardized DALY rate decreased 75% (95% UI: 72%– 78%) from 1990 to 2017. The rapid advancement in the SES over these years may partly explain this improvement; the HDI of China rose from 0.501 to 0.752 during these years. Additionally, governmental actions such as intervention projects may also contributes to the improvement. Wei et al.  estimated that a 23. 2% decrease in ID prevalence may be attributed to Iron Fortified Soy Sauce Project in China from 2004 to 2013. Other actions, such as Micronutrient Package Project for 6 ~ 24 months infants and Micronutrients Fortified Flour Project may also contribute to alleviation of the ID burden (http://www.chinanutri.cn/FFO/) . Although many iron and folic supplementation programs have been implemented or planned from 1990 to 2027 in South Asia and Sub-Saharan Africa, the ID burden in these regions are very different. (https://extranet.who.int/nutrition/gina/en/map) . In Viet Nam, following a pilot project which distributed weekly iron-folic acid in 2006, together with the de-worming for all women of reproductive age, the prevalence of iron deficiency anemia fell from 38 to 4% at 54 months. Uganda has also implemented public health packages to control ID, but the prevalence of pregnancy anemia remained high and even increased from 41.2% in 2001 to 64.4% in 2006 ; a potential reason is the low prevalence of iron/folic acid supplementation use among pregnant women, suggesting the enforcement of such policies may play an important role.
As in countries with high HDI, the link between SES and ID seemed much weaker. Al Zenki et al.  found that ID prevalence in Kuwait varied by age and gender, but not SES. The majority of very high-HDI level countries seemed to have low ID burdens. On the other hand, even with little changes in GDP, reduction in ID burden can be achieved. For example, Nepal and Bhutan obtained much better progress than most other countries that had similar changes in GDP from 1990 to 2017.
Individual-level SES is closely related to one’s health. According to Phelan et al., individuals and population groups with higher SES are more capable to take advantage of new knowledge, for example, advanced medical treatment . It is quite possible that SES-health gradient, i.e., the disparities in health across SES groups, shift in favor of high SES individuals following the development of new knowledge with empirical evidences , which could also happen to ID. There is an educational gradient in ID , and enhancing ID-preventing knowledge in people may improve the ID burden . A lot of iron supplementary programs have been adopted in areas with high ID prevalence, but it remained unclear why the disparities in the ID burden is increasing. According to Chang and Lauderdale, when we create intervention that is expensive, complicated, time-consuming to carry out, or difficult to imply broadly, we are likely to create health disparities . We may need to question if current ID interventions are easily to be adopted or need additional resources. For example, adding iron to soy bean sauce or regular food may be more effective than recommending supplements to ID patients.
The current study has a few advantages. First, we analyzed ID burden using the longitudinal GBD data that allowed us to examine the trends in ID burden over decades, globally and in different regions. Second, we used HDI to represent the overall SES of the countries, which is a combination of national income, health and education. Third, we innovatively used Gini coefficient and concentration index to quantify the extent of health inequality.
There are also a few limitations. First, our analyses depended on data from GBD 2017 , the validity of the statistical assumptions and data sources used by GBD 2017 may affect our findings. Second, although we provided a global perspective of ID burden and examined its distributions by SES, sex, and geographical regions to discuss ID inequality, we did not include other factors for ID that may also be influential. Therefore, our conclusions may not be generalizable to certain populations. Third, since we used the age-standardized DALY rate as a tool to make the ID burden comparable among different countries, and this metric is not affected by age, we were not able to explore the impact of age in great details.