The distribution of haemoglobin C and its prevalence in newborns in Africa

Abstract

Haemoglobin C (HbC) is one of the commonest structural haemoglobin variants in human populations. Although HbC causes mild clinical complications, its diagnosis and genetic counselling are important to prevent inheritance with other haemoglobinopathies. Little is known about its contemporary distribution and the number of newborns affected. We assembled a global database of population surveys. We then used a Bayesian geostatistical model to create maps of HbC frequency across Africa and paired our predictions with high-resolution demographics to calculate heterozygous (AC) and homozygous (CC) newborn estimates and their associated uncertainty. Data were too sparse outside Africa for this methodology to be applied. The highest frequencies were found in West Africa but HbC was commonly found in other parts of the continent. The expected annual numbers of AC and CC newborns in Africa were 672,117 (interquartile range (IQR): 642,116-705,163) and 28,703 (IQR: 26,027-31,958), respectively. These numbers are about two times previous estimates.

Figure 1. Global distribution of surveys on HbC.

Green dots and orange triangles indicate surveys which found HbC to be present and absent from the population sample respectively. Created with ESRI ArcGIS 10.1.

Figure 2. Summary map of HbC predicted allele frequency in Africa.

Raster map (5 km × 5 km) of HbC allele frequency (posterior mean) generated by a Bayesian model-based geostatistical framework. Created with ESRI ArcGIS 10.1.

Figure 3. Uncertainty map in HbC predicted allele frequency in Africa.

Interquartile range (50% probability) of the per-pixel predicted allele frequency. Created with ESRI ArcGIS 10.1.

Figure 4. Schematic overview of the approach.

Blue diamonds describe input data. Orange boxes denote models and experimental procedures. Green rods indicate output data. Created with Microsoft Office Visio 2007.