nova página do texto(beta)
Inglês (pdf)
Artigo em XML
Referências do artigo
Tradução automática
Enviar este artigo por email
Citado por SciELO 
Similares em
SciELO 
Permalink
Dear editor: Type 2 diabetes mellitus (T2DM) is a major health problem in Mexico, where it is estimated that by 2030 its prevalence will be between 12 and 18%.1 The diagnosis of T2DM has been based on glucose criteria, either by fasting plasma glucose (FPG) or oral glucose tolerance test (OGTT). An international committee of experts added glycated haemoglobin (A1C) as the newest criteria for the diagnosis of T2DM with levels between 5.7 to 6.4% being diagnostic for prediabetes and levels >6.5% being diagnostic for T2DM.2 Nevertheless, this definition is being questioned by other research groups.3 The World Health Organization reconsidered the A1C criteria and invited each country to evaluate its use and define its appropriate diagnostic cutoff.4
Therefore, we decided to evaluate the diagnostic accuracy of A1C, comparing it to OGTT and FGP in individuals of Monterrey, Mexico. We invited individuals >18 years old, with no medical history of T2DM, with at least one risk factor associated with T2DM. Those risk factors were determined using the Mexican Health Ministry screening questionnaire. Further evaluation with OGTT, A1C and FPG was performed.
A total of 155 participants were included in the study. From them, 52 had abnormal OGTT results: 30 (19.35%) with impaired glucose tolerance (IGT, 140-199 mg/dL) and 22 (14.19%) were in diabetes range (>200 mg/dl).
The A1C median (ranges) for participants with normal OGTT was 5.2% (4.6-6.2%), IGT was 5.65% (4.5-7.1%) and T2DM 7.6% (5.6-13.2%). Participants with A1C >6.5% in the healthy, prediabetes and T2DM groups were 0, 10 and 62% respectively. In our population, 38% of T2DM cases were not diagnosed by the proposed A1C diagnostic cutoff (>6.5%). When we evaluated primary diagnostic tests using ADA cut-point as reference, we observed a lack of sensitivity (60%) and rising of specificity (98%), which is achieved if we remove prediabetic participants from the cohort and include healthy participants and T2DM, especially in specificity (specificity: 100% and sensitivity: 62%). The kappa coefficient for 6.5% cut-off was 0.62 (p<0.001). Therefore, we explored A1C cutoff using ROC method and we observed that 5.95% had a sensitivity of 95% and specificity of 92% (AUC 0.97 [CI95%; 0. 93 to 1.0; p<0.0001])(table I).
Table I Diagnostic test between HbA1c ≥6.5% and ≥5.95%. Monterrey, México
| HbA1c ≥6.5% (CI95%) | HbA1c ≥5.95% (CI95%) | |||
| Sensitivity (%) | 62 (39 to 81) | 95.24 (74 to 99.7) | ||
| Specificity (%) | 98 (92 to 99) | 96.15 (90 to 98.7) | ||
| PPV (%) | 86.6 (58 to 97) | 83.4 (62 to 94.5) | ||
| NPV (%) | 92.7 (86 to 96) | 99 (93 to 99.95) | ||
| LHR (-) (%) | 0.39 (0.23 to 0.67) | 0.05 (0.01 to 0.34) | ||
| LHR (+) (%) | 32.19 (7.8 to 132.24) | 25 (9 to 65) | ||
Our study demonstrates that the incidence of T2DM in people with at least one risk factor was high and by using the A1C diagnostic criteria of > 6.5% we could miss almost 40% of OGTT confirmed diabetic patients. A1C >5.95% could be the diagnostic threshold value for T2DM in this specific population of northern Mexico. Further studies are needed.
