Adaptive selection at G6PD and disparities in diabetes complications

Abstract

Diabetes complications occur at higher rates in individuals of African ancestry. Glucose-6-phosphate dehydrogenase deficiency (G6PDdef), common in some African populations, confers malaria resistance, and reduces hemoglobin A1c (HbA1c) levels by shortening erythrocyte lifespan. In a combined-ancestry genome-wide association study of diabetic retinopathy, we identified nine loci including a G6PDdef causal variant, rs1050828 -T (Val98Met), which was also associated with increased risk of other diabetes complications. The effect of rs1050828 -T on retinopathy was fully mediated by glucose levels. In the years preceding diabetes diagnosis and insulin prescription, glucose levels were significantly higher and HbA1c significantly lower in those with versus without G6PDdef. In the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, participants with G6PDdef had significantly higher hazards of incident retinopathy and neuropathy. At the same HbA1c levels, G6PDdef participants in both ACCORD and the Million Veteran Program had significantly increased risk of retinopathy. We estimate that 12% and 9% of diabetic retinopathy and neuropathy cases, respectively, in participants of African ancestry are due to this exposure. Across continentally defined ancestral populations, the differences in frequency of rs1050828 -T and other G6PDdef alleles contribute to disparities in diabetes complications. Diabetes management guided by glucose or potentially genotype-adjusted HbA1c levels could lead to more timely diagnoses and appropriate intensification of therapy, decreasing the risk of diabetes complications in patients with G6PDdef alleles.

Figure 1.

Manhattan plots depicting the diabetic retinopathy meta-analysis. A–C, Manhattan plot of the diabetic retinopathy meta-analysis for (A) combined-, (B) NH-AFR-, and (C) NH-EUR-Ancestry. The y-axis shows the -log10[p], and the x-axis shows the chromosomal positions. The horizontal red line represents the threshold of p = 5 × 10⁻⁸ for genome-wide significance. All p values are computed for associations between genotyped or imputed SNPs and diabetic retinopathy as dependent variables in multivariable adjusted logistic regression models.

Figure 2.

Box plots summarizing differences in HbA1c, mean plasma glucose, and predicted HbA1c. A–E, Box plot of (A) mean HbA1c by diabetic retinopathy and G6PDdef allele status, (B) mean plasma glucose by diabetic retinopathy and G6PDdef allele status. Scatter plot of (C) mean HbA1c and mean plasma glucose in men with diabetes, (D) men of NH-AFR ancestry with diabetes, and (E) men with the G6PDdef risk allele with an overlay of the difference in HbA1c and predicted HbA1c from (A). The y-axis shows either mean HbA1c (%) or plasma glucose (mg/dL), the x-axis shows either G6PDdef allele status or mean plasma glucose (mg/dL), the colored lines of (C – E) represent difference in HbA1c and predicted HbA1c (blue = top tertile, orange = middle tertile, grey = bottom tertile).

Figure 3.

Line graph summarizing the fitted log-odds of diabetic retinopathy and corresponding 95% confidence band by mean HbA1c post diabetes diagnosis, stratified by G6PDdef risk allele status in men of NH-AFR ancestry with diabetes. The y-axis shows mean fitted log-odds of diabetic retinopathy, the x-axis shows HbA1c (%), the colored lines represent G6PDdef risk allele status, between the dotted lines represents 99% of the data.

Figure 4.

PheWAS plots summarizing the association of the G6PDdef risk allele with PheCodes in men with and without diabetes with NH-AFR ancestry. A–B, Plot of PheWAS results for individuals with diabetes (A) and plot of PheWAS results for individuals without diabetes (B). The y-axis shows the -log10[p], and the x-axis shows the PheCode Groups. The horizontal red line represents the threshold of p = 1.55 × 10⁻⁵ for genome-wide significance.