Evaluation of diagnostic procedures of Dysglycemia: A study in a tertiary care hospital in Bangladesh

Introduction

Diabetes mellitus (DM) poses a significant global health burden, affecting 451 million adults (8.8%) worldwide, with projections reaching 693 million (9.9%) by 2045 [1] . The Asia-Pacific region, a hotspot for diabetes, is impacted by rapid economic growth and sedentary lifestyles [2] . Dysglycemia, encompassing diabetes and prediabetes, is characterized by hyperglycemia linked to an increased risk of diabetic retinopathy, cardiovascular disease, and metabolic syndrome [3] . Prediabetes includes impaired fasting glucose (IFG), impaired glucose tolerance (IGT), or elevated HbA1c levels [4] . Up to 70% of individuals with prediabetes progress to overt diabetes over their lifetime. Annual diabetes incidence varies, with rates of 6-9% in isolated IFG, 4-6% in isolated IGT, and 15-19% in combined IFG-IGT cases. HbA1c levels of 5.7-6.4% indicate a 7.5-year diabetes risk [5] . Late diagnosis, poor follow-up, and noncompliance may exacerbate the diabetes burden in this region [6] . Previously, dysglycemia diagnosis relied on FPG and 2-hour post-load plasma glucose (2hPG) via OGTT. Despite its sensitivity, this approach is limited by poor reproducibility and fasting requirements [7] . A meta-analysis of 96 population studies revealed HbA1c's sensitivity and specificity for diabetes detection as 30.5% and 99.7%, respectively, compared to FPG and 2hPG [8] . Statistically significant differences were observed among FPG, 2hPG, and HbA1c as diagnostic tools.

The diagnostic sensitivity was highest for HbA1c (69.6%), followed by 2hPG (64.2%), and FPG (43.2%) [9] . A community-based study in high-risk Thai participants reported dysglycemia in 51% of individuals, with diabetes prevalence determined by 2hPG being twice as high as that by FPG (11.0% vs. 5.4%) [10] . Early diabetes detection is essential for timely management and complication prevention [11, 12] . The American Diabetes Association (ADA) defines diabetes as FPG≥126 mg/dL, 2hPG≥200 mg/dL, or HbA1c≥6.5% [12] . The ADA guidelines recommend repeating the same test for confirmation if only one of two diagnostic tests is above the cut-off, rather than conducting an additional test [12] . Currently, no method reliably ensures early detection of undiagnosed diabetes. While FPG and 2hPG are widely used for identifying at-risk individuals, many diabetic patients do not meet these criteria and remain undiagnosed. Such cases could be detected using OGTT-derived 2hPG [13] . Moreover, a recent study highlighted that HbA1c has a limited capacity to identify diabetes compared to 2hPG and FPG at current cut-points [14] . However, HbA1c offers practical advantages: it is standardized, exhibits low intraindividual variability, requires no fasting or preparation, and remains stable at room temperature post-collection. A crosssectional study of 512 subjects reported normal glucose tolerance in 43.0% of participants, impaired glucose tolerance in 37.3%, and diabetes in 19.7%. The prevalence of diabetes using 2hPG was nearly double that identified by HbA1C (19.7% vs. 11.1%), with poor agreement between prediabetes and diabetes classifications defined by 2hPG and HbA1C [15] . Similar discordance has been observed in other Asian studies, suggesting that many diabetes cases may be missed when relying solely on HbA1C compared to 2hPG from OGTT [8] . This highlights the need for further investigation into the use of HbA1C for identifying prediabetes in the Asia Pacific region to enhance early dysglycemia detection. Further research is essential to determine optimal HbA1C cut-off points for diagnosing diabetes in high-risk Southeast Asian populations [15] . Discordance between diagnostic criteria was most evident in diabetes and prediabetes categories, with minimal differences in the normal group. Among those identified as diabetic by HbA1C, only 59% were confirmed by FPG, while 95% of those classified as normal by HbA1C were similarly categorized by FPG [7] . Overall, HbA1C detected more individuals at risk for diabetes than FPG. A cross-sectional study of 1113 adults found that 259 participants met at least one DM criterion (FPG, HbA1C, or 2hPG). Diabetes prevalence was 2.7% by FPG, 5.0% by HbA1C, and 22.5% by 2hPG, with a total of 23.3% classified as diabetic by any of these methods [15] . A cohort study of 1241 overweight and obese individuals revealed a prediabetes prevalence of 46.8% and newly diagnosed diabetes in 11.9% based on ADA criteria. Using HbA1C alone, 78 subjects were identified as diabetic, but 47.3% of newly diagnosed cases would have been missed without 2hPG testing. Among 581 prediabetic subjects, 1.4% had isolated IFG, 30.5% had IGT, 5.1% had combined IFG and IGT, and 9.8% had isolated elevated HbA1C. Notably, 44.2% of prediabetes cases (HbA1C<5.7%) would have been missed without 2hPG [8] . The study highlights the superior sensitivity of 2hPG in diagnosing diabetes, identifying more cases than FPG or HbA1C [16] . Despite HbA1C's limitations, its role as a single screening tool remains uncertain. Combining multiple tests does not notably enhance diabetes detection, and relying solely on one test is inadequate for a thorough evaluation of dysglycemia [16] .

Methodology

This cross-sectional study was conducted at the Department of Biochemistry & Molecular Biology, Bangabandhu Sheikh Mujib Medical University, Dhaka, Bangladesh, from March 2019 to February 2020. A total of 1165 individuals attending the outpatient department of Endocrinology, BSMMU, for dysglycemia screening (prediabetes and diabetes) were included. Non-probability sampling was used for sample selection. The study received approval from the ethical committee of the hospital. This study included patients aged 20-75 years, of both genders, suspected to have dysglycemia based on a history of polydipsia, polyuria, polyphagia, family history of type 2 diabetes in first-degree relatives, history of gestational diabetes, obesity (BMI≥30 kg/m²), history of PCOS in females, and sudden weight loss or gain. Exclusion criteria comprised of patients with diagnosed diabetes mellitus, elevated SGPT (hepatic impairment), high serum creatinine (renal impairment), known thyroid, growth hormone, or adrenal disorders, anemia, hemoglobinopathies, BMI<18.5 kg/m², pregnancy, malignancy, acute or chronic infection, or a history of gastric bypass surgery. Data were analyzed using SPSS version 23.0.

Results

In this study, diabetes and prediabetes were defined according to the ADA criteria. The prevalence of diabetes was found to be 21.2% based on 2hPG, 15.9% based on FPG, and 23.3% based on HbA1C. Regarding prediabetes, the prevalence was 28.4% based on FPG, 24.5% based on 2hPG, and 37.8% based on HbA1C. Proportion tests revealed no significant difference between 2hPG and HbA1C in detecting diabetes mellitus. However, FPG significantly underestimated the frequency of diabetes mellitus compared to both 2hPG and HbA1C. Concordance among the three diagnostic tools for detecting diabetes mellitus occurred in only 12.4% of cases. The highest number of diabetic patients was detected using the combination of 2hPG and HbA1C (25.6%). However, when FPG was added to this combination, the detection rate dropped significantly to 12.4%. The frequency of diabetes detected using all three diagnostic tools was significantly different from that detected using any two diagnostic tools. Moreover, all three diagnostic tools identified more diabetic patients with increasing age, with a sharp contrast in detection between individuals below and above 40 years of age. Age was found to be a significant independent variable influencing the diagnosis of diabetes mellitus across all diagnostic tools (FPG, 2hPG, and HbA1C). The Kappa test indicated a moderate level of agreement between FPG and 2hPG (Kappa value 0.64) and between HbA1C and 2hPG (Kappa value 0.65) for diagnosing diabetes mellitus. Both values suggest a good agreement, though not classified as very good or excellent. Additionally, FPG showed less sensitivity compared to HbA1C. The specificity and accuracy of both FPG and HbA1C were approximately similar.

Discussion

This study aimed to evaluate the prevalence and diagnostic accuracy of diabetes mellitus using three common diagnostic tools: FPG, 2-hour post load plasma glucose (2hPG), and HbA1C, based on the American Diabetes Association (ADA) criteria. The results highlighted differences in the detection rates of diabetes across these tools and their varying diagnostic reliability. The study found that the prevalence of diabetes was highest when measured by HbA1C (23.3%), followed by 2hPG (21.2%), and lowest with FPG (15.9%). These results are in agreement with several studies that suggest HbA1C may detect a higher proportion of diabetes cases compared to FPG [17] , which could be due to HbA1C's ability to reflect long-term glycemic control, unlike FPG, which measures a single point in time [18] . In contrast, the use of FPG as a screening tool significantly underestimated the prevalence of diabetes compared to both 2hPG and HbA1C, as demonstrated by the lower detection rates with FPG. This finding is consistent with other research indicating that FPG has lower sensitivity in diagnosing diabetes, especially when compared to HbA1C and 2hPG [19] . Regarding prediabetes, a higher prevalence was observed using HbA1C (37.8%), followed by FPG (28.4%) and 2hPG (24.5%). The disparity in prediabetes detection is aligned with previous studies suggesting that HbA1C may capture more individuals at risk for developing diabetes due to its broader criteria for impaired glucose regulation [20] . Furthermore, the study revealed that combining 2hPG and HbA1C detected the most diabetic cases (25.6%), underscoring the utility of using multiple diagnostic methods for a more accurate diagnosis, as supported by prior research advocating a combined approach for better diagnostic sensitivity [21] . However, the addition of FPG to this combination led to a notable reduction in the detection rate (12.4%), highlighting FPG's relatively lower diagnostic capacity in comparison to the other methods. The low concordance rate of 12.4% among all three diagnostic tests suggests that no single test is fully reliable in detecting diabetes mellitus, supporting the recommendation for using a combination of tests for a more comprehensive assessment of glucose metabolism disorders [22] . The finding that the diagnostic tools were more effective in detecting diabetes in older individuals, particularly those above 40 years of age, emphasizes the importance of considering age as a critical factor in the diagnosis of diabetes [23] . Age-related differences in insulin sensitivity and glucose tolerance likely explain this trend, as older adults are at a higher risk for developing diabetes. The good agreement between FPG and 2hPG (Kappa value 0.64) and between HbA1C and 2hPG (Kappa value 0.65) suggests that these diagnostic tests provide similar results but are not perfectly aligned. These findings are consistent with studies that report good agreement between different diagnostic tools for diabetes diagnosis [24] . The relatively lower sensitivity of FPG further supports its limitations in accurately identifying diabetes, especially when compared to HbA1C, which has better diagnostic sensitivity and is widely used for monitoring long-term glycemic control [25] . However, the specificity and accuracy of FPG and HbA1C were similar, which indicates that while both tests are reliable, HbA1C may be a more comprehensive tool for diagnosing and monitoring diabetes. Overall, this study highlights the complexities of diagnosing diabetes and the varying effectiveness of different diagnostic tools. While HbA1C appears to be the most reliable tool for detecting diabetes, combining multiple tests may provide the best diagnostic accuracy, especially in populations at higher risk. The findings also underscore the importance of considering factors such as age and the limitations of FPG in clinical decision-making regarding diabetes diagnosis and management.

Conclusion & Recommendation

Based on these findings, the study highlights the variability in diagnostic procedures for dysglycemia, with HbA1C and 2hPG providing higher detection rates compared to FPG. While FPG significantly underestimates diabetes prevalence, 2hPG and HbA1C show better concordance in identifying diabetic patients. Age is a significant factor influencing diagnosis across all diagnostic tools. The good agreement among FPG, 2hPG, and HbA1C suggests the need for careful consideration of diagnostic criteria in identifying dysglycemia, especially in different age groups.

Funding

Conflict of interest

References

1. Cho NH, Shaw JE, Karuranga S, Huang Y, da Rocha Fernandes JD, Ohlrogge AW, et al. IDF diabetes atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Res Clin Pract. 2018;138:271-281.
2. Hu FB. Globalization of diabetes: the role of diet, lifestyle, and genes. Diabetes Care. 2011;34(6):12491257.
3. American Diabetes Association (ADA). Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes. Diabetes Care. 2018;41(1):S13-S27.
4. Mirasol R, Thai AC, Salahuddin AA, Tan K, Deerochanawong C, Mohamed M, et al. A consensus of key opinion leaders on the management of prediabetes in the Asia-Pacific region. J ASEAN Fed Endocr Soc. 2017;32(1):6-13.
5. Bussychaert M, Medina JL, Bergman M, Shah A, Lonier J. Prediabetes and associated disorders. Endocrine. 2015;48(2):371-393.
6. Huang OS, Tay WT, Ong PG, Sabanayagam C, Cheng CY, Tan GS, et al. Prevalence and determinants of undiagnosed diabetic retinopathy and vision-threatening retinopathy in a multiethnic Asian cohort: The Singapore Epidemiology of Eye Diseases (SEED) study. Br J Ophthalmol. 2015;99(12):1614-1621.
7. Ho-Pham LT, Nguyen UD, Tran TX, Nguyen TV. Discordance in the diagnosis of diabetes: Comparison between HbA1c and fasting plasma glucose. PLoS One. 2017;12(8):e0182192.
8. Meijnikman AS, De Block CEM, Dirinck E, Verrijken A, Mertens I, Corthouts B, et al. Not performing OGTT results in significant underdiagnosis of (pre)diabetes in a high-risk adult Caucasian population. Int J Obes. 2017;41(1):1615-1620.
9. Alqahtani N, Khan WAG, Alhumaidi MH, Ahmed YAAAR. Use of glycated hemoglobin in the diagnosis of diabetes mellitus and pre-diabetes and role of fasting plasma glucose, oral glucose tolerance test. Int J Prev Med. 2013;4(9):1025-1029.
10. Aekplakorn W, Tantayotai V, Numsangkul S, Sripho W, Tatsato N, Burapasiriwat T, et al. Detecting prediabetes and diabetes: Agreement between fasting plasma glucose and oral glucose tolerance test in Thai adults. J Diabetes Res. 2015;2015:1-7.
11. Aschner P, Muñoz OM, Girón D, García OM, Fernández-Ávila D, Casas LÁ, et al. Guía de práctica clínica para la prevención, diagnóstico, tratamiento y seguimiento de la diabetes mellitus tipo 2 en la población mayor de 18 años. Colombia Med. 2016;47(2):109-31.
12. American Diabetes Association (ADA). Classification and Diagnosis of Diabetes. Diabetes Care. 2017;40(1):S11-24.
13. Vlaar EM, Admiraal WM, Busschers WB, Holleman F, Nierkens V, Middelkoop BJ, et al. Screening South Asians for type 2 diabetes and prediabetes: (1) Comparing oral glucose tolerance and hemoglobin A1c test results and (2) Comparing the two sets of metabolic profiles of individuals diagnosed with these two tests. BMC Endocr Disord. 2013;13(1):1-8.
14. Karnchanasorn R, Huang J, Ou HY, Feng W, Chuang LM, Chiu KC, et al. Comparison of the current diagnostic criterion of HbA1c with fasting and 2-hour plasma glucose concentration. J Diabetes Res. 2016;2016:1-11.
15. Thewjitcharoen Y, Elizabeth AJ, Butadej S, Nakasian S, Chotwanvirat P, Another E, et al. Performance of HbA1c versus oral glucose tolerance test (OGTT) as a screening tool to diagnose dysglycemic status in high-risk Thai patients. BMC Endocr Disord. 2019;19(23):1-8.
16. Lopez-Lopez J, Garay J, Wandurraga E, Camacho PA, Higuera-Escalante F, Cohen D, et al. The simultaneous assessment of glycosylated hemoglobin, fasting plasma glucose and oral glucose tolerance test does not improve the detection of type 2 diabetes mellitus in Colombian adults. PLoS One. 2018;13(4):e0194446.
17. American Diabetes Association. 2. Classification and diagnosis of diabetes: Standards of medical care in diabetes-2021. Diabetes Care. 2021;44(Suppl 1):S15S33.
18. Iafusco D, et al. From metabolic syndrome to type 2 diabetes in youth. Children. 2023;10(3):516.
19. Pal P, Palui R, Ray S. Heterogeneity of non-alcoholic fatty liver disease: Implications for clinical practice and research activity. World J Hepatol. 2021;13(11):1584.
20. Leung LY. The effects of low glycemic index (GI) diet on cardiometabolic outcomes among obese Chinese adults. Diss. The Chinese University of Hong Kong (Hong Kong), 2022.
21. Jiang Y, et al. Epidemiology of type 2 diabetic foot problems and predictive factors for amputation in China. Int J Lower Extrem Wounds. 2015;14(1):19-27.
22. Goh JKH, et al. Retinal imaging techniques for diabetic retinopathy screening. J Diabetes Sci Technol. 2016;10(2):282-294.
23. Miettinen H, et al. The impact of age on diagnostic accuracy of diabetes tests. Endocr Rev. 2020;41(3):560570.
24. Anwar S, et al. A review on risk factors, traditional diagnostic techniques, and biomarkers for pneumonia prognostication and management in diabetic patients. Diseases. 2024;12(12):310.
25. Global variation in diabetes diagnosis and prevalence based on fasting glucose and hemoglobin A1c. Nat Med. 2023;29(11):2885-2901.