1. Introduction
Diabetic retinopathy (DR) has been identified as a common microvascular complication of diabetes, with a global prevalence of approximately 22.27% among people with diabetes [1] . According to an 8-year prospective cohort study, fluctuating blood glucose contents will induce DR onset [2] . A recent DCCT and UKPDS study found that strictly controlling HbA1c can slow the progression of retinopathy [3] . Conversely, GA reflects a shorter glycemic control period and greater glucose fluctuations than HbA1c [4] . GA level was significantly correlated with DR incidence in type 2 diabetes (T2D) cases [5] [6] [7] . According to mounting evidence, the GA/HbA1c ratio can now be used to evaluate and diagnose disease in blood glucose testing [8] . The GA/HbA1c ratio has a shorter half-life and represents short-term glycemic control when compared to GA [9] . Furthermore, the GA/HbA1c ratio has been linked to fluctuating blood glucose levels and pancreatic β-cell activity [8] . According to the findings of a previous study, the GA/HbA1c ratio can be used to predict blood glucose (BG) control, which is closely related to MAGE in T2D cases [8] [10] . Therefore, GA/HbA1c is most likely a significant predictor of DR incidence. Although DR is considered potentially preventable and treatable, there is currently a lack of awareness of DR complications among domestic people. However, almost no research has been conducted on the relationship between the GA/HbA1c ratio and the severity of DR in T2DM patients. The current study examined the relationship between the GA/HbA1c ratio and DR in T2DM patients, intending to provide clinicians with correlative therapeutic strategies for DR management.
2. Methods
2.1. Subject
From January 2018 to January 2019, 1571 T2DM cases were treated at the endocrinology ward of Harbin Medical University’s Second Affiliated Hospital. This study excluded patients with type 1 diabetes (T1DM), diabetes of other types, primary renal disease, advanced liver disease, thyroid disease, severe anemia, acute inflammatory diseases, pregnancy, and tumor.
T2DM was defined as diabetes mellitus (DM) symptoms, fasting BG (FBG) ≥ 7 mmol/L, random BG ≥ 11.1 mmol/L, and 2-h BG ≥ 11.1 mmol/L after the oral glucose tolerance test (in line with the 1999 criteria for T2DM released by World Health Organization).
DR was graded according to the International Classification of Diabetic Retinopathy in all cases by a medical technologist with extensive experience using the retinal camera: 1) Non-DR (Group A); 2) mild non-proliferative DR (Group B); 3) moderate non-proliferative DR (NPDR) (Group C); 4) severe NPDR (Group D); 5) proliferative DR (PDR) (Group E).
2.2. Data Collection
Medical records were used to obtain information such as age, gender, DM course, height, and weight. Diastolic blood pressure (DBP), systolic blood pressure (SBP), glucose laboratory tests, serum C-peptide, GA, HbA1c, UACR, and serum creatinine were all measured after an overnight fast. The HbA1c level was determined using high-performance liquid chromatography (HPLC), and GA was measured using an enzymatic assay. Body mass index (BMI) was calculated in this study by dividing body weight (kg) by height squared (m2).
2.3. Statistical Analysis
For statistical analysis, SPSS25.0 software was used. The Kruskal-Wallis test was used to compute statistical differences between subgroups (the statistic H). Normally distributed variables are represented by means ± standard deviation (the statistic F), while abnormally distributed variables are represented by the median (interquartile range). Using the non-parametric Mann-Whitney test, two data groups with non-normal distribution or uneven variance were compared (the statistic U). The chi-square test was used to compare the two groups. The effects of variables on the likelihood of DR progression were investigated using multivariate logistic regression, where p < 0.05 represented statistical significance.
3. Results
In total, 1571 T2DM patients were enrolled. There are 619 DR groups and 952 non-DR groups. Table 1 shows T2DM cases which developed DR had a longer disease course, older age, weight, BMI, SBP, UACR, TC, TG, apoB, GA, HbA1c, and GA: HbA1c. Patients with T2DM in the Non-DR group had higher c-peptide 0 min and c-peptide 120 min (p < 0.05).
![]()
Table 1. T2DM features of DR and non-DR groups.
Values are represented by means ± standard deviation or median [interquartile range] or as case numbers (percentages). *indicates p < 0.05. Abbreviations: BMI, body mass index; GA, glycated albumin; HbA1C, glycated hemoglobin A 1C; UACR, urine albumin-creatinine ratio; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
Based on subsequent classification according to DR severity, DM duration, BMI, GA, HbA1C, GA/HbA1C, C-peptide (fasting), C-peptide 30 min, C-peptide 60 min, C-peptide 120 min, and UACR, the four groups were comparable. Table 2 shows clinical features classified according to DR severity.
According to the findings, the relationship of GA with HbA1c in four groups was investigated. As shown in Figure 1, the HbA1c in Group D increased when compared with Group B (p = 0.015). Group D had higher GA than Group B, and Group E had higher GA than Groups B and C (p = 0.000). Group D had higher GA: HbA1c than Group B (p = 0.001); Group E had higher GA: HbA1c than Groups B and C (p = 0.000); Group D had higher GA: HbA1c than Group C (p = 0.011); and Group E had higher GA: HbA1c than Group D (p = 0.001).
![]()
Table 2. Clinical features of participants based on DR progression.
Values are represented by means ± standard deviation or median [interquartile range] or as case numbers (percentages). * indicates p < 0.05.
![]()
(a) ![]()
(b) ![]()
(c)
Figure 1. Correlation of diabetic retinopathy with GA, HbA1c, and GA: HbA1c. (a) Relation of GA with DR of four Groups. (b) Relation of HbA1c with DR of four groups. (c) Relation of GA: HbA1c with DR in four groups. DN, diabetic retinopathy; HbA1c, hemoglobin A1c; GA, glycated albumin.
3.1. Association between Serum GA/HbA1c Ratio and Clinical Features
Age was not found to be a significant predictor of DR. The duration of DM, BMI, c-peptide (fasting), and c-peptide 120 min all had a clear relationship with the GAHbA1c ratio (Table 3).
3.2. Multivariate Regression on Factors Affecting DR Incidence in T2DM Patients
Finally, multiple logistic regressions that included the C-peptide, UACR, and GA/HbA1c ratio as independent variables revealed that the c-peptide (fasting), UACR, and GA/HbA1c ratio independently predicted diabetic retinopathy (p = 0.004, p = 0.000, p = 0.000, respectively). Fasting C-peptide is a protective factor, i.e., fasting C-peptide has a significant and negative impact on DR development. UACR and GA/HbA1C were risk factors; they had a significant and positive effect on DR. The rest were not DR-related influencing factors (Table 4).
4. Discussion
Umayahara Y et al. found that the GA/HbA1c ratio independently predicted the occurrence of DR [11] . GA changes as blood glucose levels rise or fall, reflecting BG fluctuations in a short period of time. The GA/HbA1c ratio increased due to the recent loss of glycemic control [10] . This study found that GA and HbA1c levels are related to DR in T2DM patients. A previous study found that the
![]()
Table 3. Analysis between GA/HbA1c ratio and related indexes.
![]()
Table 4. Logistic regression analysis of the factors affecting the progression of fundus lesions.
OR: odds ratio; CI: confidence interval; Model: adjusted for DM duration and BMI.
changes in GA and HbA1c among T1DM patients increased when compared to T2DM patients [12] . Therefore, if a patient with type 1 diabetes only tested GA once a year, the results may not accurately reflect overall glycemic control. Although the GA and HbA1c were only used once in this study, they were chosen in accordance with previous research. It could be related to the fact that changes in GA and HbA1c among T2DM cases are lower than in non-T2DM cases. The values roughly reflect the overall glycemic control over the course of a year. In conclusion, it can be stated that GA/HbA1c is a factor that influences DR.
Dyslipidemia is common in T2DM patients. The relationship between apolipoproteins and DR was stronger than the relationship between traditional serum lipid levels and DR [13] . This study found that the TC, TG, and Apo-B levels in the DR groups increased when compared to the NPDR group (p < 0.05). These significant differences support the notion that dyslipidemia has a significant impact on DR. Diabetic retinopathy and nephropathy are characterized by diabetic microangiopathy caused by long-term hyperglycemia. This study’s logistic regression analysis reveals that Urinary Albumin Creatinine Ratio (UACR) is a factor that influences DR, and diabetic patients with abnormal UACR should be screened as soon as possible to prevent DR. According to the findings of Selvin et al., glycated albumin and fructosamine have prognostic utility in predicting DR in the community [14] . Prospective studies should be conducted to determine whether GA/HbA1c can be used to predict DR. Time in range (TIR) represented the time proportion in the 3.9 - 10.0 mmol/L glucose range. According to recent research, TIR is strongly linked to the occurrence of microvascular complications [15] . The hazard ratio (HR) for DR occurrence increased by 64% for every 10% decrease in TIR points, while the HR for microalbuminuria outcome increased by 40%. The HR for DR occurrence increased by 64% for every 10% decrease in TIR points, while the HR for microalbuminuria outcome increased by 40% [16] . Recent research also discovered that when TIR increased by 10%, the risk of abnormal carotid intima-media thickness (CIMT) decreased by 6.4% [17] . The inevitable trend of blood glucose monitoring is continuous blood glucose monitoring. Future research should concentrate on the relationship between GA/HbA1c and BG fluctuation and DR using glycemic variability metrics based on continuous glucose monitoring (CGM). This study classified the DR status into five groups based on the International Classification of Diabetic Retinopathy. As a result, this experiment can provide a more comprehensive understanding of the relationship between blood glucose monitoring indicators and the development of DR.
Due to the limited number of subjects with proliferative diabetic retinopathy (PDR) in this study (n = 83), it was not possible to further categorize PDR stages. Additionally, it should be noted that cross-sectional research does not establish causal relationships.
To confirm the association between glucose fluctuations, diabetic retinopathy (DR), and the GA/HbA1c ratio, it is recommended to conduct future large-scale multicenter studies utilizing continuous glucose monitoring systems. Such studies would provide more comprehensive and reliable data.
Acknowledgements
Thank the Department of Endocrinology of the Second Affiliated Hospital of Harbin Medical University and the Department of Critical Care Medicine of the Seventh Affiliated Hospital of Sun Yat-sen University for their assistance in this study. This work was supported by the Shenzhen Fundamental Research Program (JCY20190809121801664).
Data Availability
The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding authors.
Ethics Approval and Consent to Participate
The study was approved by the ethics review board of The Seventh Affiliated Hospital, Sun Yat-sen University (KY-2022-038-01) in accordance with the Declaration of Helsinki.