Abstract
Background
Atherosclerosis, a precursor to macrovascular complications in type 2 diabetes, has shown links to blood glucose fluctuations. Carotid intima-media thickness (CIMT) serves as a non-invasive tool for early atherosclerosis assessment.
Purpose
This study explores associations between specific glycaemic variability markers and CIMT in young type 2 diabetes patients.
Methodology
A cross-sectional study was conducted on 52 patients aged ≤50, assessing mean blood glucose (MBG), standard deviation (SD), mean amplitude of glucose excursion (MAGE), largest amplitude of glucose excursion (LAGE) and CIMT. Statistical analyses included descriptive statistics and Spearman’s rank correlation tests.
Results and Discussion
Participants exhibited diverse characteristics and poor glycaemic control (mean glycated hemoglobin (HbA1c): 8.7%). Nephropathy (52%), neuropathy (44.2%) and diabetic retinopathy (67.3%) were prevalent. Continuous glucose monitoring (CGM) revealed elevated glycaemic parameters. Significant correlations were found between MBG, SD, MAGE, LAGE and CIMT. Despite regional variations and elevated risk factors, findings align with prior studies, emphasising the multifactorial nature of atherosclerosis. The study highlights the importance of targeted interventions in diabetes management.
Conclusion
The study reveals suboptimal glycaemic control and a high prevalence of microvascular complications. Contrary to some studies, it establishes a significant correlation between glycaemic variability and CIMT, emphasising the importance of targeted interventions.
Keywords
Introduction
Atherosclerosis is a complex and multifactorial disease that is often a precursor to the onset of macrovascular complications in patients with diabetes. Research indicates that blood glucose fluctuations may contribute to the progression of atherosclerosis in individuals with type 2 diabetes. In this context, carotid intima-media thickness (CIMT) has emerged as a robust and non-invasive tool for the evaluation of early atherosclerotic lesions in type 2 diabetes, aiding in the assessment of disease severity and progression. 1 Previous studies have shown the association between the averaged mean levels of glycemia and macrovascular complications.2, 3 However, intensive glycaemic control does not significantly reduce cardiovascular disease in patients with type 2 diabetes mellitus (T2DM). 4
Early detection of diabetes and assessment of the risk of future cardiovascular diseases in young adults may contribute to the gradual control of the metabolic disorder pandemic in India and worldwide.
Along with raised blood glucose levels and glycated hemoglobin (HbA1c) levels, radiological markers of subclinical atherosclerosis may contribute to early detection of the risk of future cardiovascular and other complications in young diabetic patients.
CIMT is a non-invasive measurement obtained through ultrasonography, which is acknowledged as a reliable indicator of subclinical atherosclerosis. Research has demonstrated a direct correlation between increased carotid IMT and a heightened risk of developing cardiovascular and cerebrovascular diseases. 5
There are multiple factors explaining this expedited progression of this phenomenon, including dyslipidaemia with elevated levels of atherogenic low-density lipoprotein (LDL), hyperglycaemia, augmented oxidative stress and heightened inflammation. 6
Diabetic patients exhibit elevated atherogenic LDL levels, leading to intracellular lipid deposition. Modified LDL, with altered properties compared to native LDL, prompts significant lipid accumulation. 7 These modified LDLs, with altered properties, accumulate in atherosclerotic lesions, interacting with proteoglycans, leading to extended residency. Reduced affinity for the LDL receptor results in non-specific phagocytosis, contributing to intracellular cholesterol accumulation and foam cell development. 8 Furthermore, insulin inhibits the production of apoprotien - B - 48 (apoB48) and secretion of chylomicrons, which contribute to the packaging and absorption of lipids from the intestine. This inhibition is reversed in diabetic patients, thereby contributing significantly to postprandial lipidaemia and liver lipids.9, 10
Unquestionably, atherosclerosis represents a multifactorial anomaly and glycaemic variability constitutes a crucial element of dysglycemia. 11 In contrast to chronic hyperglycaemia, the connection between atherosclerosis and glycaemic variability remains inadequately comprehended, primarily owing to constraints inherent in glucose monitoring methodologies. 12 In light of the advent of continuous glucose monitoring (CGM), glycaemic variability may now be accurately investigated.
Our study endeavours to comprehensively examine the links between specific glycaemic variability markers, such as mean blood glucose (MBG), standard deviation (SD), mean amplitude of glucose excursion (MAGE), largest amplitude of glucose excursion (LAGE) and CIMT in young patients diagnosed with T2DM.
Methods
This is a cross-sectional study that screened 100 diabetic patients presenting to a tertiary health care centre general medicine outpatient department (OPD)/ward and diabetic clinic, of whom 52 were enrolled and 48 were excluded based on various exclusions. Sample size calculation was done using G Power statistics based on the results obtained from a previous study, 1 taking a power of 90% and an alpha error of 5%.
Outpatients and inpatients of the Department of General Medicine, including patients in the general medicine OPD, diabetic clinic and OPD of lifestyle disorders who met the following inclusion and exclusion criteria and were already on 14-day CGM, were included in the study after obtaining written informed consent.
All male and female patients aged ≤50 years with fasting blood sugar (FBS) ≥ 126 mg/dl or HbA1c ≥ 6.5% (diabetic according to American Diabetes Association (ADA) guidelines 2021).
Previously known cases of diabetes aged ≤50 years.
Patients giving negative consent for the study.
Full 14-day CGM data unavailable/ Incomplete CGM data.
Known type 1 Diabetes.
Recent acute complications, including diabetic ketoacidosis and hyperglycaemic hyperosmolar state.
Liver disorders or end-stage kidney disorders and other factors that affect glucose changes.
Recent acute cerebral stroke, acute myocardial infarction, malnutrition and cancers.
Patients with a history of surgical procedures or radiation therapy over the neck.
Patients with structural neck deformity.
Pregnant or lactating females.
Data Collection Process
Patients were enrolled based on inclusion and exclusion criteria. Informed written consent was obtained. Procedures included direct interviews, physical exams, 14-day FreeStyle Libre Pro glucose monitoring and CIMT measurement using high-resolution ultrasound. Data was entered into Microsoft Excel and a master chart was created. Statistical analysis, as defined, is employed for data analysis.
Statistical Analysis
Descriptive statistics for demographic and study variables were calculated, including mean, median, SD, range (continuous variables) and relative frequencies (categorical variables). Skewness, Kurtosis and the Shapiro-Wilk test were used to detect data skewness. Spearman correlation analysis was performed between CIMT and other quantitative variables from CGM. Significance was set at a p value below .05. Statistical analysis utilised the Statistical Package for Social Sciences trial version for Windows (version 25.0).
Patient and Public Involvement
Patient and public involvement were not incorporated into the design and implementation of this research study. The research question, design and conduct were developed without direct input from patients or the public. Furthermore, the choice of outcome measures, recruitment strategies and methods for disseminating study results were determined without involving patients or the public in the decision-making process.
Results
The study included 52 participants, with a mean age of 43.79 years and a SD of 4.57. The population exhibited a negatively skewed distribution, indicating non-normality, with a single mode, making it unimodal. Gender distribution showed 57.7% males and 42.3% females.
The duration of diabetes had a mean of 11.46 years, a median of 10 and a wide range from 1 to 30 years. The data exhibited positive skewness and negative kurtosis, indicating a flat but not normal distribution. Despite these skewed characteristics, the Shapiro-Wilk test confirmed non-normality.
Regarding medical history, 51.9% used oral anti-diabetic drugs, 21.2% received insulin therapy and 26.9% opted for none or alternative medicine. Additionally, 42.3% had a history of hypertension, 13.5% had a prior stroke, 21.2% had a history of coronary artery disease (CAD) and 40.4% reported alcohol intake. Smoking history was present in 23.1% of participants.
The mean HbA1c was 8.70%, with a median of 8.20 and a range of 4.9–14. Skewness indicated a lack of substantial skewness, while kurtosis suggested a normal distribution, although the Shapiro-Wilk test indicated deviation from normality.
Nephropathy was present in 51.9%, neuropathy in 44.2% and retinopathy in 67.3% of participants.
MBG data exhibited a positive skewness, suggesting non-normality, confirmed by the significant Shapiro-Wilk test. The SD of blood glucose showed normal distribution, but with multiple peaks, indicating non-normality. MAGE appeared normally distributed, while LAGE showed positive skewness, deviating from normality.
In summary, the study population displayed diverse characteristics, including skewed distributions in some parameters, emphasising the importance of statistical considerations in data interpretation.
Basic Details of the Participants.
Non-parametric Spearman correlation tests were applied due to the non-normal distribution of at least one variable in each analysis. The study revealed significant correlations between glycaemic variability markers and CIMT. MBG showed a moderate positive correlation with CIMT (rho = 0.32, p = .021), indicating that for every 1 unit increase in MBG, CIMT increases by 0.00 units. A strong positive correlation was found between SD of blood glucose and CIMT (rho = 0.64, p < .001), with a 0.01 unit increase in CIMT corresponding to a 21.16 unit increase in SD of blood glucose. Similarly, MAGE (mg/dL) and LAGE (mg/dL) exhibited strong positive correlations with CIMT (rho = 0.7, p < .001 and rho = 0.71, p < .001, respectively), emphasising their potential role in influencing CIMT changes. For both, every 1 unit increase in CIMT corresponded to an increase in MAGE by 55.97 units and LAGE by 92.41 units.
Correlation of CIMT with Glycaemic Variability.
Discussion
The study comprised 52 participants, aged ≤50 years (mean age: 43.79 ± 4.57 years), falling within the 32–50 age range. The data on age displayed a negative skew, aligning with the common diagnosis of type 2 diabetes after the fourth decade. Gender distribution favoured males (57.7%, n = 30) over females (42.3%, n = 22). The majority of participants originated from northern India, with a few seeking medical care from other regions. The study population exhibited a relatively brief diabetes duration, with a mean of 11.46 years and a SD of 6.94 years.
Regarding comorbidities, 42.3% of the participants had a documented history of hypertension. Furthermore, 13.5% of patients had a prior history of stroke, reported by seven individuals. Notably, the prevalence of CAD in our study population (21.2%) is closely aligned with the national data published in 2020, which reported a prevalence of 21.4%. 13 Specifically, 11 patients (21.2%) had a documented history of CAD. Alcohol consumption was observed in 40.4% of the participants, which strongly aligns with the prevalent higher alcohol consumption practices within the region. 14 A noteworthy finding in our study was the higher prevalence of tobacco smoking practices among participants, with approximately 23.1% having a history of smoking. This proportion is more than twice the overall smoking prevalence reported in India according to a recent demographic survey conducted in 2019–2020. 15
Furthermore, the study findings demonstrated a mean HbA1c of 8.7% with a SD of 2.37, indicating a notable range from 4.9 to 14. These results highlight the presence of relatively uncontrolled blood glucose levels within the study population. The observed higher proportion of participants not receiving pharmacological treatment or opting for alternative medications may serve as a contributing factor to this phenomenon. Additionally, potential factors such as improper dosing or non-adherence to prescribed medication regimens could further exacerbate the suboptimal glycaemic control observed among our study participants.
In terms of microvascular complications, the distribution of nephropathy, defined by the presence of micro or macroalbuminuria, exhibited a prevalence of approximately 52%. This prevalence exceeds the previously reported levels of 48% in the START-INDIA study conducted in 2017. 16 The higher incidence of nephropathy, despite the relatively shorter duration of diabetes in our study population, raises concerns. This finding suggests potential factors contributing to this trend, such as suboptimal glycaemic control, inadequate patient education, societal stigma and poor acceptance of insulin therapy, as well as challenges with adherence to prescribed diets and medications. These factors are further evidenced by the elevated HbA1c levels and other parameters observed in the CGM data, which will be discussed in subsequent sections.
A similar trend was observed in the prevalence of neuropathy, as it was found to be present in 44.2% of the participants. This percentage surpasses the estimated prevalence of 31% reported in a study conducted in southern India. 17 The higher prevalence of neuropathy in our study population, despite potential variations in demographic and regional factors, raises significant clinical interest. It suggests the presence of underlying factors contributing to the elevated occurrence of neuropathy, which warrants further investigation and evaluation in subsequent analyses.
Diabetic retinopathy exhibited a remarkably high prevalence of 67.3% within the study population. Further subgroup analysis revealed that 28.8% had grade 1 non-proliferative diabetic retinopathy (NPDR), 21.2% had grade 2 NPDR, 9.6% had grade 3 NPDR and 7.7% had proliferative diabetic retinopathy. The substantial prevalence of diabetic retinopathy can be attributed to its well-established status as the most common microvascular complication of T2DM. 18 Notably, diabetic retinopathy can manifest even 7 years before the formal diagnosis of diabetes. 19 These observations, in conjunction with the relatively poor glycaemic control evident in our study, likely contribute to the heightened prevalence of diabetic retinopathy among the participants.
Distribution of Parameters of Continuous Glucose Monitoring
In our study, CGM data revealed key glycaemic parameters: Median MBG 167 mg/dL (IQR: 134–217), median SD 63 mg/dL (IQR: 40–76), median MAGE 176 mg/dL (IQR: 125.75–194) and median LAGE 233 mg/dL (IQR: 178–280).
These findings indicate that the participants exhibited poor glucose control, as the values are relatively high. Additionally, when comparing our study to a similar investigation conducted by Taya et al. in 2021, which focused on the Japanese population, 20 of our participants displayed higher blood glucose levels and increased glycaemic variability.
Given these results, it is evident that there is a pressing need for various interventions. These may include implementing a patient education system, ensuring regular follow-ups, establishing screening health programmes and facilitating easy access to medications, particularly in the challenging-to-reach areas of the Uttarakhand district. This is particularly important as the majority of individuals in these regions remain unaware of their diabetes diagnosis.
Correlation of CIMT with Parameters of Glycaemic Variability
In a previous study conducted by Chen et al. in 2010 on the Chinese population, 1 no significant correlation was observed between MBG or SD of blood glucose with CIMT. However, in our study, we identified a significant correlation between MBG (p value .021) and SD of blood glucose (p value < .001) with CIMT. It is important to note that the previous study utilised a three-day CGM period, whereas our study employed a 14-day CGM period. This difference in monitoring duration may contribute to the variation in correlation outcomes. Furthermore, ethnic variations should also be considered, as the previous study focused on the Chinese population, while our study primarily involved individuals from the north Indian population.
Similarly, in a study conducted by Taya et al. in 2021 on the Japanese population, 20 no significant correlation was observed between MBG and SD of blood glucose with CIMT. Conversely, several other studies12, 21, 22 have reported a significant association between these variables. Therefore, the correlation between MBG, SD of blood glucose and CIMT remains controversial.
Our study findings indicate a significant correlation between MBG and SD of blood glucose with CIMT, contrasting the outcomes of previous studies conducted on Chinese and Japanese populations. The disparities in monitoring duration, ethnic composition and conflicting results from various investigations contribute to the ongoing debate surrounding the correlation between these variables.
Consistent with our study, Chen et al. (2010) 1 also discovered a statistically significant correlation between MAGE and LAGE with CIMT. These findings parallel the outcomes of our investigation. However, Taya et al. (2021) 20 did not find a significant correlation between MAGE and LAGE with CIMT in their study. On the contrary, several other studies12, 21, 22 have reported a significant association between these glycaemic variability parameters and CIMT. These divergent findings underscore the impact of ethnic variations on the relationship between CIMT and glycaemic variability.
The consistent results observed in our study, Chen et al. and other studies supporting the significant correlation of MAGE and LAGE with CIMT, along with the conflicting findings from Taya et al. further reinforce the concept of ethnic variations in the CIMT response to glycaemic variability. However, additional research is necessary to obtain further insights and develop a comprehensive understanding of this complex relationship.
The study’s limitations include its single-centre focus on a specific geographic region (Uttarakhand and Uttar Pradesh), potentially limiting generalisability to the broader population. The predominance of patients with poor glycaemic control raises concerns about the generalisability of the correlation between CIMT and glycaemic variability to those with controlled blood sugars. Additionally, the cross-sectional design precludes establishing a causal relationship between glycaemic variability and CIMT. Future research, particularly longitudinal observational studies with extended follow-ups, is warranted to further explore this complex relationship and draw more conclusive insights.
Conclusion
The study identified suboptimal glycaemic control (mean HbA1c: 8.7%) among participants, with a notable proportion not receiving pharmacological treatment. Microvascular complications were prevalent, with nephropathy in 52% and neuropathy in 44.2%. Diabetic retinopathy exhibited a high prevalence of 67.3%, emphasising the need for early screening. Contrary to some studies, our investigation revealed a significant correlation between MBG, SD of blood glucose and CIMT. These findings underscore the importance of targeted interventions for improved glycaemic control, patient education and enhanced healthcare access in managing T2DM.
Authors’ Contribution
SMB and MP conceptualised the study; SMB and ASC collected the data; SMB, MP, OP and ASC analysed the data; SMB, MP, RK, OP, ASC and MD wrote and revised the manuscript; MP, RK and MD supervised the study.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The authors received no financial support for the research, authorship and/or publication of this article.
ICMJE Statement
The study was conducted and reported in accordance with ICMJE guidelines.
Patient Consent
A well-informed written consent was taken from all participants, explaining to them the needs and consequences of the study in their preferred language and ensuring confidentiality and free will.
Statement of Ethics
The study was conducted with ethical clearance from institutional ethics committee (DHR Reg. No.: ECNEW/Inst/2020/1046 CDSCO Reg. No.: ECR/736/Inst/UK/2015/RR-21) received via letter no. AIIMS/IEC/22/608 dated 22/04/2022.
