Abstract
Haemoglobin C (HbC) is a prevalent structural haemoglobin variant. While it causes relatively mild clinical symptoms in homozygous (CC) individuals, its co-inheritance with sickle cell mutations leads to serious complications. This study utilizes a Bayesian geostatistical model to provide the first high-resolution maps of HbC frequency across Africa. By integrating these maps with demographic data, we estimate the annual birth burden of HbC. Our findings reveal that approximately 672,117 heterozygous (AC) and 28,703 homozygous (CC) newborns are born annually in Africa—figures nearly double previous estimates.
Introduction: The Clinical and Geographic Gap
Haemoglobin C (HbC) is a common genetic variant of normal haemoglobin. While individuals with the homozygous "CC" genotype typically experience only mild clinical symptoms, the variant poses a significant public health challenge when co-inherited with the sickle cell mutation (HbS). This combination results in HbSC disease, a condition characterized by chronic haemolytic anaemia and painful crises.
Despite its clinical importance, the contemporary geographic distribution of HbC has remained poorly understood. Previous estimates of the "birth burden"—the number of affected newborns—relied on sparse data and conservative assumptions, such as assuming zero prevalence in regions lacking survey data. This research addresses this gap by creating a continuous, high-resolution map of HbC allele frequencies across Africa and calculating more accurate estimates of the annual birth burden.
Methodology: Bayesian Geostatistical Modeling
To overcome the limitations of previous studies, the research team employed a Bayesian model-based geostatistical (MBG) framework. Unlike simple spatial averaging, this approach predicts allele frequencies across a continuous 5x5 km grid while accounting for spatial correlation and inherent uncertainty in the survey data.
The workflow involved four primary stages:
- Data Assembly: Compiling a global database of HbC allele frequencies from 174 published and unpublished sources, covering over 7.5 million individuals tested between 1950 and 2010.
- Geostatistical Modeling: Using the African subset of data to generate a posterior predictive distribution of HbC frequencies.
- Demographic Integration: Pairing frequency maps with high-resolution (1x1 km) population density data from the Global Rural-Urban Mapping Project (GRUMP) and national crude birth rates from the UN Population Division.
- Uncertainty Quantification: Generating interquartile range (IQR) maps to identify regions where predictions require further validation.
Key Insights: Selection and Migration
The study confirms the "Malaria Hypothesis," showing that the distribution of HbC strongly aligns with historical malaria endemicity. HbC provides significant protection against Plasmodium falciparum malaria—near-full protection for CC individuals and intermediate protection for AC individuals. This evolutionary advantage explains the high prevalence of the trait in West Africa.
The mutation appears to have a unique origin in West Africa, with the highest frequencies (up to 16%) concentrated in eastern Burkina Faso. While historical selection by malaria shaped this distribution, modern human migration is now the primary driver spreading HbC into new regions, including the Americas and Europe.
Results: A Significant Underestimation
The model's results suggest that previous global estimates (such as those by Modell and Darlison in 2008) were highly conservative. By capturing the high heterogeneity of the trait through continuous mapping, this study provides a more accurate picture of the burden:
- Newborn Estimates: In 2010, an estimated 672,117 babies were born with the AC genotype and 28,703 with the CC genotype in Africa.
- Regional Concentration: The burden is highly localized; Burkina Faso, Ghana, and Nigeria alone account for 56% of all AC newborns and 76% of all CC newborns on the continent.
- Model Accuracy: Validation statistics showed a mean error of only 0.012, indicating high accuracy and low bias in the predictive model.
Limitations and Future Directions
While the study provides a breakthrough for Africa, data sparsity prevented the application of the MBG methodology to other regions like the Caribbean or the Middle East. Additionally, the model assumes Hardy-Weinberg equilibrium, which may be slightly affected in areas with high mortality from co-inherited disorders.
The ultimate goal for researchers is the development of multi-allelic models that simultaneously map HbS, HbC, and thalassaemias. Such models would allow for the direct estimation of compound disorders like HbSC disease, which represent a significantly larger public health burden than HbCC alone.
Summary updated