Raised Glycated Hemoglobin (HbA1c) Level as a Risk Factor for Myocardial Infarction in Diabetic Patients: A Hospital-Based, Cross-Sectional Study in Peshawar

Background Diabetes is a rapidly rising chronic illness in developing countries. The main objective of this research is to compare the frequency of myocardial infarction (MI) in controlled and uncontrolled diabetics in Pakistan, especially in the underprivileged district of Peshawar, and to determine raised blood glucose as a risk factor for MI. Methodology This cross-sectional study involving 237 diabetic patients aged 30-80 years was conducted in three major tertiary care hospitals in Peshawar, Pakistan. The inclusion criteria were diabetic patients with glycated hemoglobin (HbA1C) levels of less than 7% considered to be “controlled diabetics” and above 7% considered to be “uncontrolled diabetics.” Data were collected using structured questionnaires, past medical records, and patient history and were analyzed using SPSS software (IBM Corp., Armonk, NY, USA). The study was concluded in March 2022. Results The highest number of MIs occurred in diabetics with HbA1c levels of 8-9% (47.9% of all MIs). There was a significant association between increasing HbA1c levels and the incidence of MI (p = 0.002). The adjusted prevalence odds ratio for MI in uncontrolled diabetics was 6.105 (95% confidence interval = 2.42-15.43), that is, six times increased incidence of MI in patients with HbA1c of more than 7%. Furthermore, with a 1% increase in HbA1c, there was a 10% increase in the proportion of MIs. Conclusions From this study, it is clear that HbA1c levels of 8-9% were most significantly associated with the risk of MI in uncontrolled diabetics, and with rising levels of HbA1c, the risk of MI increased significantly. Thus, this study highlights the importance of HbA1c control in diabetic patients to prevent a heart attack.


Introduction
The change in lifestyle and food consumption patterns throughout the world has led to a change in trends of disease mortalities which has shifted from infectious diseases, such as tuberculosis, pneumonia, influenza, and malaria, to chronic diseases, such as ischemic heart disease (IHD), stroke, cancers, diabetes mellitus (DM), and hypertension. In particular, DM is crucial because it leads to multiple complications in patients that lead to their untimely death if left unchecked. According to the World Health Organization, by 2020, DM would be the cause of 70% of mortality from all chronic diseases that cause death in developing countries [1]. Even more concerning is the projected 2.5-fold increase in the population of individuals with DM in the developing world from 1995 to 2025 [1].
Cardiovascular disease (CVD) is the major cause of mortality in individuals with DM, especially in uncontrolled long-standing type II DM, and DM remains the number one underlying risk factor for CVD. DM also leads to other conditions that contribute to heart diseases, such as hypertension and hypercholesterolemia. According to the American Heart Association [2], at least 68% of individuals aged 65 or older with diabetes die from some form of heart disease. Adults with diabetes are two to four times more likely to die from heart disease than adults without diabetes. The American Heart Association considers diabetes to be one of the seven major controllable risk factors for CVD. Risk factors other than DM that lead to MI must also be investigated and accounted for as they are a part of the variables and control factors in such research.
Elevated blood glucose level is the most damaging element of uncontrolled DM that may lead to IHD, the most common type of CVD. Elevated blood glucose levels cause vascular damage, especially to the susceptible coronary arteries, by accelerating atherosclerosis, endothelial cell dysfunction, glycosylation of extracellular matrix proteins, and vascular denervation [3]. Persistent vascular damage leads to inflammation, stiffening, stenosis, thrombosis, or even rupture of the coronary arteries which leads to myocardial infarction (MI). Moreover, high blood glucose levels in patients at the time of acute MI have been shown to be associated with a poor prognosis of MI (regardless of whether patients have DM or not) [4]. This also shows the negative impact of high blood glucose on the heart. A crucial indicator of blood glucose level is glycated hemoglobin (HbA1c) levels. It has been reported that every 1% increase in HbA1c levels independently predicts a 19% increase in the odds of experiencing a MI after accounting for other risk factors, including DM [5].
The prevalence of CVD in DM is a worldwide dilemma and one that occurs in both genders of all ethnicities, as suggested by several studies [6]. The INTERHEART case-control study, conducted in 52 countries throughout Europe, Asia, the Middle East, Australia, Africa, and North and South America, found nine modifiable contributors to acute MI that lead to CVD. These modifiable risk factors include hypertension, diabetes, smoking, abdominal obesity, psychological index, lack of exercise, lack of fruits and vegetables, apolipoprotein B/apolipoprotein A1 (ApoB/ApoA1) ratio, and the use of alcohol. In India, a high prevalence of metabolic cardiovascular risk factors has been reported among clinic-based patients with diabetes [7]. It is of global interest to conduct further research into this untreatable disease and how it is affecting the lives of millions.

Diabetes and CVD pathophysiology
The pathophysiology of coronary heart disease due to hyperglycemia includes the processes of atherosclerosis, oxidative stress, inflammatory changes, abnormal protein folding, increased coagulability, and disruption of the endothelial NO synthase system of the coronary endothelium which leads to endothelial dysfunction, impaired blood flow, and reduced tissue perfusion of the myocardium. Oxidative stress is the major factor that leads to MI. Another factor worth mentioning here is that fluctuations in blood glucose levels have more oxidative stress on the microvasculature than sustained hyperglycemia [8]; hence, measuring blood glucose values other than HbA1c is also necessary to evaluate the degree of diabetes and its related complications [9].
These factors are studied further in detail to examine the critical effects of uncontrolled diabetes on the micro and macrovasculature of the body.
Atherosclerosis is the major outcome of the destructive effects of hyperglycemia in diabetic patients [10]. High blood glucose causes several metabolic and structural disturbances in the blood vessels. The process by which glucose brings about these potentially destructive changes is called glycation. Glucose binds with protein residues (e.g., lysine, arginine, etc.) via this process to form Amadori complexes, of which HbA1c is a well-known example. These complexes then change to advanced glycation end-products (AGE) by oxidative, non-oxidative, and spontaneous pathways. These products react with circulating hemoglobin to form hemoglobin-AGE, and their levels are elevated in diabetic patients. AGE products bind to their receptors (RAGE) present on smooth muscles, endothelium, and cells of the immune system and bring about oxidative stress (via protein kinase C, overproduction of superoxide anion and other reactive oxygen species (ROS), arterial stiffness, atherosclerotic changes, induction of cytokines, etc.) Thus, they play a major role in degenerative diseases such as diabetes, Alzheimer's disease, and chronic kidney diseases. Apart from this, insulin resistance itself induces pro-inflammatory cytokines such as C-reactive protein and interleukin 6 which contribute to atherosclerosis. All of these destructive changes and atherosclerotic processes that are more often seen in diabetic patients cause plaque and thrombus formation which manifests as MI [11].

Rationale
Pakistan is the sixth most populous country in the world, and DM has been a growing public health issue in Pakistan [12], with an 11.77% prevalence of diabetes [12], that is, more than 25 million people according to the current population of Pakistan. The vast majority of the population belongs to a poor socioeconomic background and low educational status. Because DM is vastly concerned with a healthy lifestyle and awareness about diabetic control measures, this research has been conducted to assess the extent of diabetes management and outcomes among diabetic patients in the study area. We also hope to create general awareness about the risk of MI in diabetic patients and guide their perceptions of it, so that necessary lifestyle changes and medical help may be provided in time.

Objectives
The primary objectives of this study were to determine an association between high HbA1c levels and fasting blood sugar levels in diabetic patients with the prevalence of MI and to compare the frequency of MI in controlled and uncontrolled diabetics. The secondary objective of the study was to find concomitant risk factors along with DM which contribute to MI. We hypothesize that there is a positive relationship between uncontrolled diabetes and the risk of MI.

Study design and study population
A hospital-based, analytical, cross-sectional study was conducted in three tertiary care hospitals, namely, Khyber Teaching Hospital, Hayatabad Medical Complex, and Peshawar Institute of Cardiology, in the city of Peshawar, KPK. These hospitals were chosen using a purposive strategy as they mainly cater to the 4 million-plus population of Peshawar and are situated in different geographical regions of the district. Moreover, such a study has not been conducted in these hospitals yet. Data were mainly collected from the Cardiology, Endocrinology, and Medical units of these hospitals. Our study population included clinically diagnosed type II diabetic patients aged 30-80 years, both males and females with or without MI. The study period was from April 2021 to March 2022 during which data were collected within two months of time, as shown by the Gantt chart inserted below ( Table 1).

Sampling
The sample size of 236 was achieved using the Cochran formula where the estimated prevalence of MI in diabetics was taken to be 0.19 (19%) [5]. The z-value was taken as 1.96 and the margin of error as 0.05. A purposive sampling (judgment sampling) strategy was devised for this study. It is a type of non-probability sampling that requires researchers to have prior knowledge about the purpose of their studies so that they can properly choose and approach eligible participants for surveys. Researchers use purposive sampling when they want to access a particular subset of people as all participants of a survey are selected because they fit a particular profile. The main objective of a purposive sample is to produce a sample that can be logically assumed to be representative of the population.
The researchers visited the selected hospitals to collect data from only confirmed diabetic patients regardless of their history of MI. For patients who had an MI, only those were included who had their MI after their diabetes was diagnosed, and both good and bad glycemic control patients were included. Patients with both HbA1c more than 7% and less than 7% were purposefully selected and included in the study (maximum variation). Patients who fulfilled the inclusion criteria were selected to complete the sample size. Inclusion criteria included male and female known diabetic patients between 30 and 80 years of age who consented to participate in the study. Diabetic patients who were not willing to cooperate were excluded from our study. Moreover, patients who developed diabetes after MI were also excluded.

Procedures
The data collection tool was a structured questionnaire and an on-site survey involving live, face-to-face interviews by trained interviewers. It was also complemented with secondary data from patients' medical records that were available for every patient included in the study. Data were recorded by the investigators themselves. The questionnaire was validated using face and content validity by CVD/diabetes and public health specialists. Second, it was reviewed after a pilot study on 10% of the calculated total sample size. Moreover, all cut-off values/ranges for variables in the questionnaire were taken with regard to the National Institute for Health and Care Excellence (NICE) UK guidelines and according to the significant effect of the variable on MI.
All values of HbA1c and other variables were recorded before the occurrence of MI in patients so that a temporal relationship could be established. HbA1c levels were taken from patient files, ranging from three months to one year before the occurrence of MI in the patients, as indicated in the questionnaire (Q14: Before the occurrence of MI, if MI occurred). For patients who did not have an MI, the HbA1c levels were recorded from three months to one year before inclusion in the study. Where multiple HbA1c tests were available during that period, a mean of the level was recorded.
The study variables included HbA1c levels and fasting blood glucose levels as independent variables and MI attacks as dependent variables. Other risk factors studied included hypertension, family history of MI, body mass index (BMI), and lipid profiles. The questionnaire included questions related to the onset of diabetes and the age of occurrence of MI. It also included basic personal information (age, gender, marital status, education level, occupation type, family income, etc.) and the prevalence of behavioral risk factors (smoking, diet, physical activity, etc.). No more than 5% of the data were missing for any variable apart from lipid profile where a significant amount of data was not available in patient records. Because lipid profile was not a part of our primary objectives, it had little significance on the findings of the study and was not included in the logistical regression analysis.

Biases
To reduce the information bias, structured questionnaires were used along with secondary data from the patients' medical files. Interviewers were the researchers themselves and were trained to conduct the surveys using local language skills and technical knowledge of the questions. To reduce the selection bias (sampling bias), purposive sampling was done on a calculated large sample that most closely represented the CVD diabetic population of Peshawar. Moreover, the study area included three major hospitals situated in different locations and covering the vast majority of the district area. Furthermore, both inpatient and outpatient data were recorded. All these steps were done to increase the external validity of the research. Finally, the effect of potential risk factors (confounding bias) was adjusted using analytical techniques such as logistical regression models, and only the significantly associated variables with MI were adjusted for.

Analysis
For analysis, we used the Statistical Package for the Social Sciences (SPSS) software version 23 (IBM Corp., Armonk, NY, USA). To determine the statistical significance of our results we performed descriptive statistics, chi-squared tests, prevalence odds ratio, and regression models, and included 95% confidence interval (CI) values as well. P-values of <0.05 were considered significant. Diabetic patients aged 30-80 years with HbA1C levels less than 7% were categorized as "controlled diabetics" and HbA1C levels of more than 7% were categorized as "uncontrolled diabetics." The confidentiality of respondents was maintained and no names or photographs were used. No forced participation in research without consent was assured. Verbal consent was taken after explaining the questionnaire-based survey and that no sensitive information was recorded. It was confirmed that no offensive questions were asked from the volunteers.

Results
Data of 237 patients were recorded.  The association of patients' HbA1c levels with the prevalence of MI is depicted as a bar chart in Figure 1 and as a line graph in Figure 2. Table 3 shows the crosstab for MI attacks among patients with HbA1c of more than 7% (uncontrolled diabetes) and less than 7% (controlled diabetes); this is further illustrated as a bar chart in Figure 3. To remove the effect of other significant risk factors such as hypertension and family history, a logistical regression model was used, as shown in Table 4. Table 5 shows the prevalence of MI with each percentage increase in HbA1c level. As this study is a cross-sectional prevalence study, we calculated the prevalence odds ratio instead of the odds ratio, which is calculated in case-control and cohort studies. The association of fasting blood sugar levels with the prevalence of MI is shown in Figure 4. Table 6 shows the crosstabs of the additional risk factors with MI.      The figure shows the association between fasting blood sugar above and below 126 mg/dL with MI attacks. A significant association was found between the two (chi-square (1, N = 221) = 9.58, p = 0.002  The association of other risk factors with MI using chi-square statistics. Blood pressure and family history showed a significant association with MI.

Discussion
The main focus of our research was to highlight the prevalence of heart attacks in uncontrolled diabetes patients. There have been many studies on microvascular changes caused by poor glycemic control and the association of HbA1c with the severity of CAD but few on the occurrence of MI, especially in our study area.
A cross-sectional study was carried out in Dhaka, Bangladesh on 100 patients who were divided into two groups: HbA1c >6.5% (Group A) and HbA1c <6.5% (Group B) [13]. Group A had six times more prevalence of acute coronary syndrome (MI) than Group B. The study showed a statistically significant correlation between HbA1c and the severity of CAD (p < 0.001). Positive family history of IHD was also significantly associated with CAD (p = 0.001). All these findings conform to our findings. This study also showed a significant association between CAD incidents with dyslipidemia (p = 0.001). However, in our study, the association of MI with increased cholesterol levels was insignificant due to insufficient data concerning lipid profiles.
A study carried out in Peshawar Lady Reading Hospital in 2020 studied the degree of severity of CAD with HbA1c levels categorizing patients into two groups, namely, poor glycemic control (HbA1c >7.5%) and good glycemic control (HbA1c <7.5%) [14]. The study found a significant association between poor glycemic control and the severity of CAD. Fisher's exact test for DM and hypertension with glycemic control showed a significant association (p = 0.005). We observed similar results in our study with blood pressure and MI (p = 0.009). In their study, the two-tailed Mann-Whitney U test was significant (p < 0.001) for age with glycemic control but insignificant for BMI. As poor glycemic control is directly related to the severe outcomes of CAD, these findings are consistent with our study findings. We observed a similar statistical association between patient age with MI (p < 0.001) and patient's BMI with MI (p = 0.299).
A cohort study carried out in Norway involving over 4,000 patients reported that the risk of acute MI is higher in obese men but lower in overweight women [15]. In our study, 63.4% of obese patients were females; hence, the phenomenon of the obesity paradox was seen and an insignificant association was noted between BMI and MI. Many other studies have also observed the same paradoxical effect [16]. The reason for this effect is beyond the scope of our study but has been explained by others to be the presence of natural reserves to tackle the effects of pathological stress and at the molecular level about its protective effect.
Other studies that disagree with this false perception report accountable factors such as the limitations of adjusting all confounding variables, the difference in the nature of comorbidities among study groups (e.g., age), and that BMI itself is not the best measure of morbid adiposities.
According to a cohort study by Selvin et al, a 1% increase in HbA1c levels increased the risk of having a cardiovascular event by 18% [17]. Our study found that with a 1% increase in HbA1c (6% and above), there was a 10.4% (95% CI = 2.3-18.5) increase in the prevalence of MI.
In our study, the association of fasting blood glucose with the prevalence of MI was statistically significant (p = 0.017). With a fasting blood glucose level of 126 mg/dL, there is a fourfold increase in MI compared to those who had a fasting blood glucose level of <126 mg/dL. Increased fasting blood glucose is a good indicator of the risk of developing cardiovascular disease and MI. Many studies have demonstrated the effect of increasing fasting blood glucose on the outcome of CAD and MI [18].
A cohort study was conducted on 1,687 patients to examine the effect of HbA1c on cardiovascular outcomes in diabetic patients with and without the presence of vascular disease [19]. The study found no statistically significant association of HbA1c with cardiovascular outcomes in patients with vascular disease but a significant association between HbA1c and cardiovascular outcomes in patients without vascular disease. In our study, only 9.7% of patients had vascular disease; hence, conclusively, our analysis showed a strong association between HbA1c and MI.

Limitations
This study had some limitations. First, the sampling technique that could be a probability/randomized method of patient selection instead of the non-probability technique used in this study. This was due to limited access to patients and reduced patient load in hospitals, pertaining to strict rules/lockdowns during the coronavirus disease 2019 pandemic. Moreover, because complete patient medical records were missing for lipid profiles, they could not be measured accurately. Third, due to the study setting being hospitals, patients with controlled diabetes were fewer in number as mostly complicated and serious patients are admitted with high blood glucose levels. Therefore, uncontrolled diabetics were more prevalent in hospitals.
We recommend a case-control study to follow this cross-sectional prevalence study where case-control groups could be formed and matching can be done. This would further endorse the findings of this study.

Conclusions
This study was conducted in three major hospitals in Peshawar and represented a healthy proportion of the CVD diabetic population. We found a clear strong association between increased blood glucose and the prevalence of MI. A large proportion of diabetics who had poor glycemic control showed the largest proportion of MIs. Furthermore, as the HbA1c and fasting blood sugar levels increased, so did the proportion of patients who had an MI. Every 1% increase in HbA1c increased the proportion of patients having an MI by over 10%. In addition to blood glucose, we observed blood pressure and family history to significantly increase the adverse outcomes.
This study highlights the importance of glycemic control in diabetic patients and how essential it is to maintain HbA1c levels under 7% to potentially prevent a heart attack.   Additional Information