In the map and horizontal bar chart, the data are presented in five classes created using the quantile classification method. Each class contains 20% of countries, which is easy to interpret. The quintile intervals are labeled sequentially from Quintile 1, also called the bottom quintile which includes the lowest fifth (0 to 20%) of data points to Quintile 5 (or top quintile), which includes the top fifth (80% to 100%) of data points.
Anemia refers to low levels of hemoglobin in the blood, which seriously affects health.
Anemia in women: women aged 15−49 years with a hemoglobin concentration lower than 120 g/L for non-pregnant women and lactating women, and lower than 110 g/L for pregnant women, adjusted for altitude and smoking.
Anemia in children: children aged 6−59 months with a hemoglobin concentration less than 110 g/L, adjusted for altitude.
DATA SOURCES AND METHODS
This topic and visualization used data from the 2021 edition of the WHO global anemia estimates (see the references section).
The methodology for the 2021 edition of the WHO global anemia estimates is detailed in the technical document "WHO methods and data sources for mean hemoglobin and anemia estimates in women of reproductive age and pre-school age children 2000-2019", available online as a pdf file.
In brief, data on the prevalence of anemia and/or mean hemoglobin in women of reproductive age, collected between 1995 and 2019 were obtained from 408 population-representative data sources from 124 countries worldwide. A Bayesian hierarchical mixture model was used to estimate hemoglobin distributions and systematically address missing data, non-linear time trends, and representativeness of data sources. Full details on data sources are available on the WHO Global Health Observatory: Anemia in women and children page.
Full details on statistical methods may be found here: Finucane MM, Paciorek CJ, Stevens GA EM. Semiparametric Bayesian density estimation with disparate data sources: a meta-analysis of global childhood undernutrition. J Am Stat Assoc. 2015;110(511):889–901.