Glucose 6-phosphate dehydrogenase deficiency and Southeast Asian-specific mutations lower HbA1c levels in a Thai population: implications for diabetes diagnosis.

Abstract

Aims/hypothesis: Glucose 6-phosphate dehydrogenase (G6PD) deficiency, the most common inherited enzymopathy, can affect HbA_1c levels and the diagnosis of type 2 diabetes. This cross-sectional study aimed to investigate the association between G6PD deficiency, its common mutations (G6PD Viangchan^G871A, G6PD Mahidol^G487A) and HbA_1c levels in a Thai cohort.

Methods: Blood samples from 1007 healthy hospital staff were collected during annual health checkups. Individuals with diabetes, diabetes medication use and conditions affecting erythrocyte turnover were excluded. HbA_1c levels were measured by enzymatic assay and HPLC, while fasting plasma glucose (FPG) and haematological variables were obtained from checkup records. Fructosamine levels and G6PD activity (classified as deficiency, intermediate, normal) were measured by spectrophotometric assay. Genotyping was performed using TaqMan SNP, PCR-Restriction Fragment Length Polymorphism (RFLP) and Sanger sequencing. Prediabetes and diabetes were diagnosed based on two abnormal results from FPG and HbA_1c at the same time, following the modified ADA criteria. Optimal HbA_1c cutoffs were determined using receiver operating characteristic curve analysis with bootstrapping in RStudio, optimising Youden's index and the harmonic mean of sensitivity and specificity.

Results: HbA_1c levels were significantly lower in individuals with G6PD deficiency (25.68 mmol/mol [4.50%] by enzymatic assay; 27.33 mmol/mol [4.65%] by HPLC; n=28; p<0.001) compared with those with normal G6PD levels (34.05 mmol/mol [5.27%] by enzymatic assay; 36.61 mmol/mol [5.50%] by HPLC; n=492). Similarly, individuals with G6PD Viangchan^G871A (29.46 mmol/mol [4.85%] by enzymatic assay; 31.15 mmol/mol [5.00%] by HPLC; n=52; p<0.001) and G6PD Mahidol^G487A (28.63 mmol/mol [4.77%] by enzymatic assay; 31.15 mmol/mol [5.00%] by HPLC; n=15; p<0.001) had significantly lower HbA_1c levels. HbA_1c levels positively correlated with G6PD activity (r=0.208, p<0.001 by enzymatic assay; r=0.211, p<0.001 by HPLC). The optimal HbA_1c cutoffs for predicting prediabetes in participants with G6PD mutation were 33 to <42 mmol/mol (5.2% to <6.0%) by enzymatic assay (sensitivity 70%; specificity 88.64%; accuracy 86.74%) and 34 to <43 mmol/mol (5.3% to <6.1%) measured by HPLC (sensitivity 72.73%; specificity 86.21%; accuracy 84.70%). For diabetes, the optimal cutoffs were ≥42 mmol/mol (≥6.0%) by enzymatic assay (sensitivity 100%; specificity 97.92%; accuracy 97.96%) and ≥43 mmol/mol (≥6.1%) by HPLC (sensitivity 100%; specificity 96.88%; accuracy 96.94%). Using the mutation-specific HbA_1c cutoffs resulted in the proportion of individuals being diagnosed with diabetes remaining the same but the proportion diagnosed with prediabetes rose from 8.2% (ADA criteria) to 18.4% using enzymatic assay and from 9.2% to 21.4% using HPLC.

Conclusions/interpretation: HbA_1c levels were positively correlated with G6PD activity, with individuals carrying G6PD Viangchan^G871A and G6PD Mahidol^G487A exhibiting significantly lower HbA_1c levels. Our findings highlight the need to consider G6PD deficiency and G6PD mutations when using HbA_1c to diagnose type 2 diabetes in Southeast Asian populations.