European Journal of Cardiovascular Medicine

Type 2 Diabetes Mellitus (T2DM) is a chronic metabolic disorder of global significance, with its prevalence and associated healthcare burden rising exponentially [1]. The primary therapeutic goal in managing T2DM is to prevent or delay the onset of long-term complications, which are broadly categorized as microvascular (retinopathy, nephropathy, neuropathy) and macrovascular events [2]. For decades, the measurement of glycated hemoglobin (HbA1c) has been the cornerstone of glycemic monitoring, reflecting average blood glucose over the preceding 2–3 months. Landmark trials have unequivocally demonstrated that lowering HbA1c reduces the risk of microvascular complications [3, 4].

However, HbA1c has inherent limitations; it provides a static measure of mean glycemia and fails to capture the dynamic intraday and interday fluctuations in blood glucose, collectively known as glycemic variability (GV) [5]. Two patients with identical HbA1c levels can have vastly different daily glucose profiles, one with stable glucose levels and another with dramatic swings between hyperglycemia and hypoglycemia. There is a growing consensus that these glycemic excursions may exert a unique pathogenic effect, independent of chronic sustained hyperglycemia [6]. The proposed mechanisms linking high GV to tissue damage include the overproduction of reactive oxygen species, induction of endothelial dysfunction, and activation of pro-inflammatory pathways, which are particularly detrimental to the microvasculature of the retina, glomeruli, and peripheral nerves [7, 8].

With the advent and increasing accessibility of continuous glucose monitoring (CGM) technology, it is now possible to quantify GV with high precision using various metrics such as the standard deviation (SD) of glucose, coefficient of variation (CV), and the Mean Amplitude of Glycemic Excursions (MAGE) [9]. Several cross-sectional studies have reported associations between high GV and the presence of microvascular complications [10, 11]. However, the evidence from longitudinal studies, which is critical for establishing a predictive relationship, remains sparse. Most prospective studies have been limited by short follow-up durations, small sample sizes, or a focus on a single complication rather than a composite endpoint [12].

This research gap is clinically important. If high GV is confirmed as an independent, long-term risk factor for complications, it could shift the paradigm of diabetes management from a sole focus on HbA1c to a more comprehensive approach that also aims to stabilize glucose fluctuations. Therefore, this study was designed to fill this gap. Our primary aim was to longitudinally assess GV using CGM over a five-year period and to determine its independent association with the incidence and progression of a composite of microvascular complications—diabetic retinopathy, nephropathy, and neuropathy—in a well-characterized cohort of T2DM patients.

Study Design and Population: This was a prospective, single-center, longitudinal cohort study and a total of 352 patients with T2DM were enrolled.

Inclusion and Exclusion Criteria: Eligible participants were adults aged 30–75 years with a diagnosis of T2DM for at least two years (according to American Diabetes Association criteria), treated with stable doses of oral anti-diabetic agents or insulin for at least three months prior to enrollment.

Exclusion criteria included Type 1 diabetes, a history of gestational diabetes, severe renal impairment (estimated Glomerular Filtration Rate [eGFR] < 30 mL/min/1.73 m²), severe hepatic disease, active malignancy, pregnancy or lactation, and any condition that would preclude the accurate use of a CGM device (e.g., skin conditions, allergy to adhesive).

Data Collection and Procedures: At the baseline visit, detailed demographic data, medical history, and medication use were recorded. A physical examination including blood pressure and body mass index (BMI) was performed. Fasting blood samples were collected for measurement of HbA1c, lipid profile, and serum creatinine.

Glycemic Variability Assessment: All participants underwent a 14-day period of blinded professional CGM (FreeStyle Libre Pro, Abbott Diabetes Care) at baseline and annually for five years. Participants were instructed to maintain their usual diet, physical activity, and medication regimens. CGM data were considered valid if the sensor was worn for at least 10 days with >70% of available data. The primary GV metric was MAGE, calculated using dedicated software. The 5-year average MAGE was calculated for each patient and used to stratify the cohort into tertiles: Tertile 1 (Low GV), Tertile 2 (Moderate GV), and Tertile 3 (High GV).

Assessment of Microvascular Complications: Standardized assessments for microvascular complications were performed at baseline and at the final 5-year follow-up visit.

• Diabetic Retinopathy: Assessed by seven-field digital fundus photography of both eyes, graded by a trained ophthalmologist blinded to the patient’s clinical data, according to the Early Treatment Diabetic Retinopathy Study (ETDRS) scale.
• Diabetic Nephropathy: Assessed by urinary albumin-to-creatinine ratio (UACR) from a first-morning spot urine sample and eGFR calculated using the CKD-EPI formula.
• Diabetic Peripheral Neuropathy: Assessed using the 10-g Semmes-Weinstein monofilament test at 10 sites on the feet and the Michigan Neuropathy Screening Instrument (MNSI) questionnaire.

Study Outcome: The primary outcome was a composite of incident or progressing microvascular disease, defined as the first occurrence of any of the following during the 5-year follow-up:

1. Retinopathy progression: Worsening of two or more steps on the ETDRS scale, development of proliferative diabetic retinopathy, or the need for pan-retinal photocoagulation or anti-VEGF therapy.
2. Nephropathy progression: Progression from normoalbuminuria to micro- or macroalbuminuria (UACR >30 mg/g), from microalbuminuria to macroalbuminuria (UACR >300 mg/g), or a sustained decline in eGFR of ≥40% from baseline.
3. Incident neuropathy: A new diagnosis of peripheral neuropathy confirmed by a positive monofilament test (inability to feel the filament at ≥4 of 10 sites) in a patient with a negative baseline test.

Statistical Analysis: Data were analyzed using R software (Version 4.1.2). Continuous variables were presented as mean ± standard deviation (SD) and compared across MAGE tertiles using ANOVA. Categorical variables were presented as numbers and percentages (n, %) and compared using the Chi-square test. The cumulative incidence of the composite outcome was estimated using the Kaplan-Meier method, and differences between MAGE tertiles were assessed with the log-rank test. A series of Cox proportional hazards regression models were constructed to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between MAGE tertiles and the primary outcome. Model 1 was unadjusted. Model 2 was adjusted for age, sex, and BMI. Model 3 was further adjusted for key clinical confounders: 5-year mean HbA1c, diabetes duration, mean systolic blood pressure, LDL cholesterol, and smoking status. A p-value <0.05 was considered statistically significant.

Baseline Characteristics: A total of 352 patients were enrolled and completed the baseline assessment. Over the 5-year follow-up, 28 (8.0%) patients were lost to follow-up, leaving 324 patients for the final analysis. The cohort was stratified into tertiles based on their 5-year average MAGE: Low GV (MAGE <55 mg/dL, n=108), Moderate GV (MAGE 55–75 mg/dL, n=110), and High GV (MAGE >75 mg/dL, n=106).

Baseline demographic and clinical characteristics of the study population across the MAGE tertiles are shown in Table 1. The groups were well-matched for age, sex, BMI, and diabetes duration. While the mean HbA1c showed a trend towards being higher in the High GV group, the difference was not statistically significant at baseline. As expected, all CGM-derived GV metrics (SD, CV, and MAGE) were significantly different across the tertiles.

Table 1. Baseline Characteristics of the Study Population by Tertiles of Mean Glycemic Variability (MAGE)

GV: Glycemic Variability; MAGE: Mean Amplitude of Glycemic Excursions; BMI: Body Mass Index; BP: Blood Pressure; LDL: Low-Density Lipoprotein; SD: Standard Deviation.

Incidence of Microvascular Complications: Over the 5-year follow-up period, 102 (31.5%) of the 324 participants experienced the primary composite outcome. The incidence of microvascular events showed a clear dose-response relationship with GV (Table 2). The event rate was lowest in the Low GV tertile (15.7%), intermediate in the Moderate GV tertile (28.2%), and highest in the High GV tertile (51.9%) (p for trend <0.001). This pattern was observed for all individual components of the composite outcome, with the strongest association seen for nephropathy progression.

Table 2. Incidence of Microvascular Complications over 5 Years by MAGE Tertiles

Association between Glycemic Variability and Clinical Outcomes: The Kaplan-Meier survival curves showed a significant separation between the three MAGE tertiles for event-free survival from the composite microvascular outcome (log-rank p<0.001). Patients in the High GV tertile had a significantly lower probability of remaining free of complications over the study period.

The results of the Cox proportional hazards regression analysis are presented in Table 3. In the unadjusted model, compared to the Low GV tertile, the HR for the composite outcome was 1.95 (95% CI 1.10–3.46) for the Moderate GV tertile and 4.21 (95% CI 2.46–7.21) for the High GV tertile. This association remained strong and highly significant after adjusting for demographic factors (Model 2) and key clinical confounders, including the 5-year mean HbA1c (Model 3). In the fully adjusted model, the HR for the High GV tertile was 2.92 (95% CI 1.68–5.08, p<0.001).

Table 3. Hazard Ratios for the Composite Microvascular Outcome According to MAGE Tertiles

*Model 3 is adjusted for age, sex, BMI, 5-year mean HbA1c, diabetes duration, mean systolic blood pressure, LDL cholesterol, and smoking status.

In this 5-year prospective study of patients with T2DM, we demonstrated that long-term glycemic variability, as measured by MAGE from CGM, is a powerful and independent predictor of the incidence and progression of microvascular complications. Patients in the highest tertile of GV had nearly a threefold increased risk of developing retinopathy, nephropathy, or neuropathy compared to those in the lowest tertile, even after accounting for mean glycemic control (HbA1c) and other traditional risk factors.

Our findings significantly strengthen the evidence base supporting the role of GV as a key contributor to diabetic complications. While previous cross-sectional studies have linked GV to the prevalence of these conditions [11, 13], our longitudinal design provides more robust evidence of a predictive relationship. The "glucose variability hypothesis" posits that intermittent spikes of hyperglycemia, rather than just sustained high glucose, are particularly damaging. These excursions trigger acute bouts of oxidative stress that can overwhelm cellular antioxidant defenses, leading to endothelial cell apoptosis, basement membrane thickening, and pericyte loss in the microvasculature [7, 14]. Our results, showing a graded risk across GV tertiles, strongly support this dose-dependent pathogenic mechanism.

A crucial aspect of our study is the demonstration that GV's impact is independent of HbA1c. The mean HbA1c levels were only marginally different across the GV tertiles, yet the complication rates varied dramatically. This confirms that HbA1c, while indispensable, does not tell the whole story of glycemic health. Our fully adjusted model, which included mean HbA1c as a covariate, confirmed that GV provides unique prognostic information. This implies that a patient with a "good" HbA1c of 7.0% but high variability may be at a greater risk of complications than a patient with the same HbA1c but stable glucose levels. This has profound clinical implications, suggesting that the quality of glycemic control (stability) is as important as the quantity (average glucose).

Our findings are broadly consistent with the few other longitudinal investigations in this area. For instance, a study by Xu et al. found that baseline GV predicted the progression of diabetic nephropathy over 4 years [15]. Our study extends these findings by using a composite endpoint, employing longitudinally assessed GV over five years, and using robust CGM technology, providing a more comprehensive picture of the risk posed by unstable glucose levels.

The clinical implications are clear. As CGM technology becomes more widespread, clinicians will have direct access to GV metrics. Our data suggest that these metrics should be incorporated into routine risk stratification for T2DM patients. For patients with high GV, even with a seemingly acceptable HbA1c, clinicians should consider therapeutic adjustments. Certain classes of anti-diabetic medications, such as GLP-1 receptor agonists and SGLT2 inhibitors, have been shown to reduce GV in addition to lowering HbA1c and may be preferable choices in such patients [16].

This study has several strengths, including its prospective longitudinal design, a relatively long follow-up period, the use of gold-standard CGM for GV assessment, and comprehensive, standardized adjudication of microvascular outcomes. Nevertheless, some limitations should be noted. As an observational study, we cannot establish causality, although the strong, independent, and dose-dependent association is highly suggestive. The study was conducted at a single center, which may limit the generalizability of the findings to other populations. Finally, although we adjusted for major confounders, residual confounding cannot be completely excluded.

compelling evidence that high glycemic variability is a significant and independent risk factor for the development and progression of microvascular complications in individuals with T2DM. The prognostic value of GV is additive to that of traditional markers like HbA1c. These findings advocate for a paradigm shift in diabetes management, moving beyond a singular focus on HbA1c to embrace a more holistic view of glycemic control that includes the assessment and targeting of glucose fluctuations to better protect patients from the devastating long-term consequences of diabetes.

1. Saeedi P, Petersohn I, Salpea P, Malanda B, Karuranga S, Unwin N, et al. Global and regional diabetes prevalence estimates for 2019 and projections for 2030 and 2045: Results from the International Diabetes Federation Diabetes Atlas, 9th edition. Diabetes Res Clin Pract. 2019;157:107843. DOI: 10.1016/j.diabres.2019.107843.
2. Fowler MJ. Microvascular and Macrovascular Complications of Diabetes. Clin Diabetes. 2008;26(2):77-82. DOI: 10.2337/diaclin.26.2.77.
3. The Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med. 1993;329(14):977-86. DOI: 10.1056/NEJM199309303291401.
4. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet. 1998;352(9131):837-53. DOI: 10.1016/S0140-6736(98)07019-6.
5. Beck RW, Connor CG, Mullen DM, Wesley DM, Bergenstal RM. The Fallacy of Average: How Using HbA1c Alone to Assess Glycemic Control Can Be Misleading. Diabetes Care. 2017;40(8):994-9. DOI: 10.2337/dc17-0636.
6. Ceriello A, Monnier L, Owens D. Glycaemic variability in diabetes: clinical and therapeutic implications. Lancet Diabetes Endocrinol. 2019;7(3):221-30. DOI: 10.1016/S2213-8587(18)30136-0.
7. Monnier L, Mas E, Ginet C, Michel F, Villon L, Cristol JP, et al. Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia in patients with type 2 diabetes. JAMA. 2006;295(14):1681-7. DOI: 10.1001/jama.295.14.1681.
8. Quagliaro L, Piconi L, Assaloni R, Martinelli L, Motz E, Ceriello A. Intermittent high glucose enhances apoptosis related to oxidative stress in human umbilical vein endothelial cells: the role of protein kinase C and NAD(P)H-oxidase activation. Diabetes. 2003;52(11):2795-804. DOI: 10.2337/diabetes.52.11.2795.
9. Rodbard D. Continuous glucose monitoring and metrics of glycemic control: an overview. Diabetes Technol Ther. 2017;19(S2):S3-S12. DOI: 10.1089/dia.2017.0027.
10. Suh S, Kim JH. Glycemic Variability: How Do We Measure It and Why Is It Important? Diabetes Metab J. 2015;39(4):273-82. DOI: 10.4093/dmj.2015.39.4.273.
11. Wang C, Lv L, Yang Y, Chen S, Li C, Li L, et al. The role of glycemic variability in the development of cardiovascular autonomic neuropathy in patients with type 2 diabetes. Cardiovasc Diabetol. 2019;18(1):145. DOI: 10.1186/s12933-019-0949-1.
12. Smith-Palmer J, Brändle M, Trevisan R, Orsini-Federici M, Li H, Valentine W. Assessment of the association between glycemic variability and diabetes-related complications in type 1 and type 2 diabetes. Diabetes Res Clin Pract. 2014;105(3):296-304. DOI: 10.1016/j.diabres.2014.06.012.
13. Sartore G, Chilelli NC, Burlina S, Lapolla A. The role of high glucose and glycemic variability in diabetic neuropathy. Curr Diabetes Rev. 2018;14(3):215-22. DOI: 10.2174/1573399813666170213152656.
14. Ceriello A, Esposito K, Piconi L, Ihnat MA, Thorpe JE, Testa R, et al. Oxidative stress and glycemic variability in diabetes: therapeutic strategies. Diabetes Care. 2007;30(10):2604-5. DOI: 10.2337/dc07-1399.
15. Xu Y, Wu H, He X, Zhao Z, Cui J, Yang J, et al. Glycemic variability is an independent risk factor for the progression of diabetic nephropathy in patients with type 2 diabetes. Medicine (Baltimore). 2016;95(30):e4245. DOI: 10.1097/MD.0000000000004245.
16. Cheng Y, Yuan Y, Wang C, Li H, Zhou J. The effects of SGLT2 inhibitors on glycemic variability in patients with type 2 diabetes: a systematic review and meta-analysis. Endocrine. 2021;72(1):36-45. DOI: 10.1007/s12020-020-02559-0.