Full Text
Type 2 diabetes mellitus (T2DM) is a chronic metabolic disorder characterized by progressive impairment of insulin secretion in the context of insulin resistance. Its clinical importance extends beyond persistent hyperglycemia to the development of cardiovascular, renal, neurological, and retinal complications. Because metabolic abnormalities may precede overt symptoms by several years, laboratory assessment provides an important opportunity to identify dysglycemia before clinical deterioration becomes apparent. Timely recognition, followed by appropriate preventive or therapeutic intervention, may reduce subsequent disease burden and support individualized care.
Established laboratory measures, including fasting plasma glucose, glycated hemoglobin (HbA1c), and plasma glucose measured during an oral glucose tolerance test, remain central to the diagnosis of diabetes. These tests capture different aspects of glucose regulation and do not invariably identify the same individuals. Fasting glucose primarily reflects glucose homeostasis in the fasting state, whereas the oral glucose tolerance test can reveal abnormalities in the response to a standardized glucose challenge. HbA1c estimates longer-term glycemic exposure and offers practical advantages because fasting is unnecessary. Nevertheless, altered erythrocyte turnover, certain hemoglobin variants, and other clinical conditions can affect its interpretation. In asymptomatic individuals, abnormal diagnostic results generally require confirmation unless hyperglycemia is unequivocal.
For monitoring established T2DM, HbA1c is widely used to evaluate glycemic control, but it does not adequately characterize short-term fluctuations or hypoglycemic episodes. Fructosamine and glycated albumin reflect glycemic exposure over a shorter interval and may offer complementary information when HbA1c is unreliable or when recent changes in control are clinically relevant. Their interpretation, however, can be affected by changes in protein concentration or turnover. Additional laboratory measurements, such as C-peptide, lipid parameters, urinary albumin-to-creatinine ratio, and estimated glomerular filtration rate, address distinct dimensions of disease assessment, including endogenous insulin secretion, cardiometabolic risk, and kidney involvement. These measures should not be regarded as interchangeable with diagnostic glycemic tests.
Interest has also increased in biomarkers associated with inflammation, adipose tissue dysfunction, and metabolic stress. Although some candidate markers are associated with insulin resistance or future diabetes, association alone does not establish diagnostic utility. Their potential contribution depends on reproducibility, clinically meaningful incremental information beyond established measures, and feasibility across diverse patient populations. Interpretation must also account for age, adiposity, medication exposure, and comorbid conditions that may influence biomarker concentrations independently of glycemic status.
Against this background, the present study aims to evaluate the contribution of selected laboratory biomarkers to the early identification of T2DM and the monitoring of glycemic status over time. By examining their relationships with established glycemic measures and relevant clinical characteristics, the study seeks to clarify whether complementary biomarker assessment provides useful information beyond conventional testing. This approach distinguishes markers of dysglycemia from those reflecting associated metabolic dysfunction or complications, while recognizing the need for clinically interpretable and accessible laboratory strategies.
2. Materials and Methods
A prospective, single-center observational study was designed to evaluate laboratory biomarkers for the early detection and longitudinal monitoring of type 2 diabetes mellitus. Recruitment was planned over 12 months from primary care and endocrinology clinics, with each participant followed for 12 months. Eligible adults were aged 30–70 years, had no previous diabetes diagnosis, and had at least one established risk factor, including overweight or obesity, hypertension, dyslipidemia, or a first-degree family history of diabetes. Exclusion criteria included pregnancy, suspected autoimmune or secondary diabetes, recent acute illness, systemic glucocorticoid treatment, active malignancy, advanced liver disease, and estimated glomerular filtration rate below 30 mL/min/1.73 m². Conditions affecting erythrocyte survival or hemoglobin measurement were recorded and considered in sensitivity analyses. Recruitment would begin only after institutional ethics approval and written informed consent.
The target sample comprised 360 consecutively enrolled participants. Assuming approximately 25% had previously unrecognized diabetes and 10% had incomplete diagnostic data, this sample was expected to provide approximately 80 evaluable diabetes cases. Sample-size planning focused on estimating diagnostic discrimination with reasonable precision rather than detecting small differences between individual biomarkers. Enrollment numbers, exclusions, incomplete assessments, and reasons for withdrawal were to be documented prospectively.
At enrollment, trained staff collected demographic characteristics, medical and medication histories, smoking status, family history, and physical activity information using a standardized questionnaire. Weight, height, waist circumference, and blood pressure were measured according to uniform procedures. Body mass index was calculated as weight in kilograms divided by squared height in meters. Medication changes, intercurrent illnesses, and diabetes-related treatment were documented throughout follow-up. Clinical care was not delayed or determined by research biomarker findings.
Following an overnight fast of at least eight hours, venous blood was collected for fasting plasma glucose, glycated hemoglobin (HbA1c), fructosamine, glycated albumin, insulin, C-peptide, lipid profile, creatinine, albumin, and high-sensitivity C-reactive protein. Participants underwent a 75-g oral glucose tolerance test, with plasma glucose measured two hours after glucose ingestion. Glucose specimens were processed promptly using validated procedures to minimize glycolysis. HbA1c was measured using an internationally standardized assay; other analytes were measured using validated enzymatic or immunochemical methods. Manufacturer-specified quality controls were included in each analytical run. Samples requiring batch analysis were stored at −80°C, avoiding repeated freeze–thaw cycles, and laboratory personnel were blinded to diagnostic classification.
The primary reference classification was based on two-hour oral glucose tolerance results, independently of the biomarkers under evaluation. Values below 140 mg/dL indicated normal glucose tolerance, 140–199 mg/dL indicated impaired glucose tolerance, and values of at least 200 mg/dL indicated diabetes. In asymptomatic participants, diabetes-range results required confirmation by a repeat oral glucose tolerance test within two weeks. Discordant results prompted additional clinical assessment, and unresolved classifications were excluded from the primary accuracy analysis. An endocrinologist reviewed clinical features to assess whether confirmed diabetes was consistent with type 2 diabetes.
Follow-up assessments occurred at three, six, and twelve months. Biomarkers and clinical measurements were repeated at each visit; oral glucose tolerance testing was repeated at twelve months in participants without baseline diabetes. Baseline diagnostic discrimination constituted the primary outcome. Secondary outcomes included incident diabetes and within-person biomarker changes after routine treatment initiation. For monitoring analyses, changes in fructosamine and glycated albumin were compared with contemporaneous and subsequent HbA1c changes.
Receiver operating characteristic curves quantified diagnostic performance, with bootstrap confidence intervals and paired comparisons of areas under the curve. Prespecified multivariable models assessed incremental discrimination beyond age, sex, body mass index, and family history. Mixed-effects models evaluated longitudinal associations while accounting for treatment changes and repeated measurements. Missing-data patterns were examined, with multiple imputation used when assumptions were defensible. Secondary analyses were considered exploratory.
In the simulated study cohort, 360 adults undergoing assessment for dysglycemia were classified by repeat oral glucose tolerance testing as having normal glucose tolerance (n = 120), prediabetes (n = 120), or newly diagnosed type 2 diabetes mellitus (n = 120). Mean age was 51.8 ± 10.6 years, and 188 participants (52.2%) were women. Mean body mass index increased across the three groups, from 25.8 ± 3.4 to 28.1 ± 4.0 and 30.0 ± 4.6 kg/m², respectively. Age and family history of diabetes differed between groups, whereas sex distribution was comparable. No participant with newly diagnosed diabetes had received glucose-lowering medication before baseline sampling.
Baseline glycemic biomarkers showed progressively higher concentrations across glucose tolerance categories. Mean fasting plasma glucose was 5.1 ± 0.4 mmol/L in participants with normal glucose tolerance, 5.8 ± 0.6 mmol/L in those with prediabetes, and 7.6 ± 1.8 mmol/L in those with diabetes. Corresponding mean glycated hemoglobin (HbA1c) values were 5.3 ± 0.3%, 5.9 ± 0.4%, and 7.1 ± 1.0%. Mean glycated albumin increased from 12.8 ± 1.6% to 15.2 ± 2.0% and 19.0 ± 3.7%. Between-group differences for these biomarkers remained evident after adjustment for age, sex, and body mass index, although substantial overlap persisted between adjacent categories.
Fasting insulin and the homeostasis model assessment of insulin resistance were also higher with worsening glucose tolerance. Median insulin resistance estimates were 1.7 (interquartile range, 1.2–2.4), 2.9 (2.0–4.1), and 4.5 (3.0–6.7) across the three groups. High-sensitivity C-reactive protein concentrations were modestly higher in participants with diabetes, but the association weakened after adjustment for adiposity. Triglyceride concentrations increased and high-density lipoprotein cholesterol decreased across categories. These metabolic and inflammatory markers exhibited wider distributions than the glycemic biomarkers and did not clearly separate individual participants with prediabetes from those with diabetes.
For identifying oral glucose tolerance test–defined diabetes against the combined nondiabetic groups, HbA1c yielded an area under the receiver operating characteristic curve of 0.90 (95% confidence interval, 0.86–0.93). Corresponding estimates were 0.87 (0.83–0.91) for fasting plasma glucose and 0.86 (0.82–0.90) for glycated albumin. At the conventional HbA1c threshold of 6.5%, sensitivity was 73.3% and specificity was 95.0%. Fasting plasma glucose at 7.0 mmol/L had a sensitivity of 66.7% and specificity of 96.7%. Thus, neither fasting glucose nor HbA1c independently detected all participants whose diabetes was established by repeated post-load glucose measurements.
An exploratory model combining fasting plasma glucose and HbA1c achieved an apparent area under the curve of 0.93 (0.90–0.96), with a bootstrap optimism-corrected estimate of 0.92. Adding glycated albumin increased the apparent estimate to 0.94, but the incremental improvement was small and its confidence interval included no improvement. Adding fasting insulin or high-sensitivity C-reactive protein produced no clear additional discrimination. Classification of prediabetes versus normal glucose tolerance was less accurate: areas under the curve were 0.79 for HbA1c, 0.76 for fasting glucose, and 0.75 for glycated albumin. These exploratory estimates were not evaluated in an external validation cohort.
Among participants with newly diagnosed diabetes, 108 (90.0%) completed both the 6-week and 12-week assessments after initiation of lifestyle measures and clinician-selected treatment. Baseline HbA1c and fasting glucose were similar between completers and noncompleters, although the small number lost to follow-up limited assessment of attrition bias. In completers, mean HbA1c decreased from 7.1 ± 1.0% at baseline to 6.8 ± 0.9% at 6 weeks and 6.5 ± 0.8% at 12 weeks. Mean fasting glucose declined from 7.5 ± 1.7 to 6.5 ± 1.2 and 6.2 ± 1.0 mmol/L, respectively. Glycated albumin declined from 18.9 ± 3.6% to 16.6 ± 2.8% and 15.8 ± 2.4%.
The reduction in glycated albumin at 6 weeks correlated with the reduction in HbA1c at 12 weeks (r = 0.61; 95% confidence interval, 0.48–0.72). The corresponding correlation for early fasting glucose reduction was 0.44 (0.27–0.58). The glycated albumin association persisted after adjustment for baseline HbA1c and treatment category. However, individual responses remained heterogeneous, and early biomarker changes did not consistently predict attainment of HbA1c below 7.0%. Changes in inflammatory and lipid markers were smaller and showed weak associations with HbA1c improvement. Because treatment was not randomized, these longitudinal findings describe biomarker trajectories rather than comparative treatment effects.
Laboratory biomarkers play complementary roles in the early diagnosis and longitudinal monitoring of type 2 diabetes mellitus, but their value depends on the clinical question, analytical reliability, and patient characteristics. The central implication of evaluating these markers together is that no single measurement captures the full spectrum of dysglycemia, insulin resistance, β-cell dysfunction, and emerging complications. Glycemic biomarkers remain the foundation of diagnosis, whereas lipid, inflammatory, and renal measurements provide additional information about metabolic risk and disease burden. These roles should remain distinct: association with diabetes does not necessarily establish usefulness for diagnosing it.
Fasting plasma glucose and glycated hemoglobin (HbA1c) offer practical but different perspectives on glucose regulation. Fasting glucose primarily reflects glucose homeostasis at the time of sampling and is influenced by hepatic glucose production, fasting duration, acute illness, and recent behavioral changes. HbA1c integrates glycemic exposure over approximately two to three months, with greater weighting toward recent weeks, and does not require fasting. Their combined interpretation can therefore reduce reliance on an isolated physiological state. Nevertheless, discordance between these measurements should prompt clinical evaluation rather than automatic classification of one result as erroneous.
Such discordance is particularly relevant during early disease, when postprandial hyperglycemia may precede a consistent elevation in fasting glucose. An oral glucose tolerance test can identify impaired glucose tolerance or diabetes in individuals whose fasting glucose and HbA1c remain below diagnostic thresholds. Its greater procedural burden and within-person variability, however, limit routine application. A selective approach is reasonable when clinical suspicion remains high despite inconclusive initial testing. In the absence of unequivocal hyperglycemia, abnormal diagnostic findings require confirmation according to established criteria. This distinction is important because risk stratification, screening, and definitive diagnosis are related but not interchangeable activities.
HbA1c also requires interpretation in the context of erythrocyte biology. Anemia, altered red-cell survival, recent transfusion, and some hemoglobin variants can produce values that do not accurately reflect circulating glucose. Advanced kidney disease may further complicate interpretation through anemia, treatment effects, and changes in erythrocyte turnover. Consequently, agreement between HbA1c and plasma glucose should be examined before treatment decisions are based on HbA1c alone. Where reliability is uncertain, glucose-based assessment is generally more appropriate for diagnosis. These considerations also support standardized assays and consistent preanalytical procedures when comparing measurements across participants or visits.
For monitoring, HbA1c remains useful because it summarizes sustained glycemic exposure and is linked to the risk of microvascular complications. However, it cannot distinguish stable glucose concentrations from alternating hyperglycemia and hypoglycemia that produce a similar average. A reduction in HbA1c should therefore not automatically be interpreted as improvement in every dimension of glycemic control. Blood glucose profiles and, where indicated and accessible, continuous glucose monitoring provide complementary information about variability and hypoglycemia. Monitoring targets should also account for age, comorbidities, treatment burden, and vulnerability to adverse effects rather than applying identical expectations to every patient.
Fructosamine and glycated albumin may offer shorter-term perspectives, generally reflecting glycemia over the preceding two to three weeks. They can be useful when rapid treatment changes are being assessed or when HbA1c is difficult to interpret. Their utility is nevertheless constrained by conditions affecting protein turnover, including substantial proteinuria, liver disease, and altered albumin metabolism. Differences in assay availability and standardization also limit direct substitution for HbA1c. These markers are best viewed as selective adjuncts, not universal replacements or independently established diagnostic tests for type 2 diabetes.
Insulin-related biomarkers provide a different layer of biological information. Fasting insulin and indices such as the homeostasis model assessment of insulin resistance can characterize metabolic dysfunction in research settings, potentially before overt hyperglycemia develops. Their clinical interpretation is limited by assay variability, population-dependent distributions, and the absence of universally accepted diagnostic thresholds. C-peptide can inform endogenous insulin secretion in selected cases, particularly when diabetes classification or insulin deficiency is uncertain, but it is not a routine screening marker for type 2 diabetes. Moreover, renal clearance and prevailing glucose concentrations influence C-peptide interpretation. These measurements should therefore support specific questions rather than expand testing indiscriminately.
Lipid abnormalities and inflammatory biomarkers may identify a broader adverse metabolic phenotype. Elevated triglycerides and reduced high-density lipoprotein cholesterol commonly accompany insulin resistance, while high-sensitivity C-reactive protein may reflect low-grade inflammation associated with adiposity and metabolic dysfunction. These relationships are nonspecific and susceptible to confounding by obesity, smoking, infection, medications, and cardiovascular disease. Renal markers require a similarly differentiated interpretation. Urinary albumin-to-creatinine ratio and estimated glomerular filtration rate are important for complication surveillance and treatment planning, but neither establishes the presence of diabetes. Persistent albuminuria should be distinguished from transient elevations before clinical conclusions are drawn.
Several methodological issues constrain interpretation of biomarker studies. Associations measured at one time point cannot establish whether a marker predicts incident diabetes or merely reflects existing dysglycemia. Medication exposure, disease duration, adiposity, and comorbidities may influence both biomarker concentrations and clinical outcomes. Selection from a single clinical setting can limit generalizability, while within-person variation may weaken conclusions based on single measurements. Demonstrating added diagnostic value also requires more than statistically significant group differences: discrimination, calibration, external validation, and clinical utility should be evaluated against readily available standard tests.
Overall, an integrated but selective biomarker strategy is more defensible than an expanded panel applied uniformly. Standardized glycemic testing should anchor early diagnosis, with additional measurements chosen to resolve uncertainty, assess short-term change, or detect complications. Future prospective studies should determine whether candidate biomarkers improve prediction and patient outcomes beyond conventional testing, while considering cost and equitable access. The practical goal is not simply earlier biochemical labeling, but timely identification of clinically meaningful dysglycemia and monitoring that supports safe, individualized care.
Laboratory biomarkers remain central to the early identification and longitudinal monitoring of type 2 diabetes mellitus, but their clinical value depends on appropriate selection, interpretation, and integration with individual risk profiles. Fasting plasma glucose and glycated hemoglobin provide complementary information on glycemic status, while an oral glucose tolerance test may help clarify selected cases in which routine measurements are inconclusive or discordant. Abnormal results in asymptomatic individuals generally require confirmation before a diagnosis is established.
For ongoing monitoring, glycated hemoglobin remains a practical indicator of longer-term glycemic exposure, although it does not capture glucose variability or hypoglycemia. Its interpretation should account for conditions that alter erythrocyte turnover, hemoglobin variants, and other sources of measurement bias. Fructosamine or glycated albumin may offer additional information when glycated hemoglobin is unreliable or when shorter-term changes require assessment; however, these markers also have limitations, particularly in disorders affecting protein turnover.
Biomarkers reflecting insulin resistance, inflammation, and metabolic stress may improve understanding of disease heterogeneity and potentially refine risk stratification. Nevertheless, their association with diabetes does not by itself establish diagnostic utility or justify routine implementation. Demonstrating incremental value beyond established clinical factors and standard glycemic tests, together with acceptable reproducibility, accessibility, and cost, remains essential. Measures of kidney function and urinary albumin excretion serve a distinct but complementary role in detecting complications and guiding management rather than diagnosing diabetes itself.
The most defensible clinical approach is therefore a targeted, multimarker strategy rather than indiscriminate expansion of testing. Biomarker results should inform, not replace, clinical assessment and shared treatment decisions. Further prospective, multicenter studies with diverse populations, standardized assays, and clinically meaningful outcomes are needed to determine whether emerging panels improve diagnostic timeliness, therapeutic adjustment, and patient outcomes. Until such benefits are established, validated glycemic biomarkers should remain the foundation of evidence-based diabetes detection and follow-up.
1. American Diabetes Association Professional Practice Committee. (2024). 2. Diagnosis and classification of diabetes: Standards of Care in Diabetes—2024. Diabetes Care, 47(Supplement 1), S20–S42. https://doi.org/10.2337/dc24-S002
2. American Diabetes Association Professional Practice Committee. (2024). 6. Glycemic goals and hypoglycemia: Standards of Care in Diabetes—2024. Diabetes Care, 47(Supplement 1), S111–S125. https://doi.org/10.2337/dc24-S006
3. International Expert Committee. (2009). International Expert Committee report on the role of the A1C assay in the diagnosis of diabetes. Diabetes Care, 32(7), 1327–1334. https://doi.org/10.2337/dc09-9033
4. Sacks, D. B., Arnold, M., Bakris, G. L., Bruns, D. E., Horvath, A. R., Kirkman, M. S., Lernmark, A., Metzger, B. E., & Nathan, D. M. (2011). Guidelines and recommendations for laboratory analysis in the diagnosis and management of diabetes mellitus. Diabetes Care, 34(6), e61–e99. https://doi.org/10.2337/dc11-9998
5. Nathan, D. M., Kuenen, J., Borg, R., Zheng, H., Schoenfeld, D., & Heine, R. J. (2008). Translating the A1C assay into estimated average glucose values. Diabetes Care, 31(8), 1473–1478. https://doi.org/10.2337/dc08-0545
6. Selvin, E., Rawlings, A. M., Grams, M., Klein, R., Steffes, M., & Coresh, J. (2014). Fructosamine and glycated albumin for risk stratification and prediction of incident diabetes and microvascular complications: A prospective cohort analysis of the Atherosclerosis Risk in Communities (ARIC) study. The Lancet Diabetes & Endocrinology, 2(4), 279–288. https://doi.org/10.1016/S2213-8587(13)70199-2
7. Tabák, A. G., Herder, C., Rathmann, W., Brunner, E. J., & Kivimäki, M. (2012). Prediabetes: A high-risk state for diabetes development. The Lancet, 379(9833), 2279–2290. https://doi.org/10.1016/S0140-6736(12)60283-9
8. UK Prospective Diabetes Study (UKPDS) Group. (1998). Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). The Lancet, 352(9131), 837–853. https://doi.org/10.1016/S0140-6736(98)07019-6
9. Radin, M. S. (2014). Pitfalls in hemoglobin A1c measurement: When results may be misleading. Journal of General Internal Medicine, 29(2), 388–394. https://doi.org/10.1007/s11606-013-2595-x