Abstract

Background and objectives: For decades the global population has been experiencing diabetic epidemic. The risks related to obesity, diabetes mellitus (DM) and coronary artery diseases (CAD) are well known. This study aimed to assess the prevalence of coronary artery disease (CAD) and its related risks in Bangladeshi diabetics.

Materials and methods: The study was conducted at Bangladesh Institute of Research and Rehabilitation in Diabetes, Endocrine and Metabolic Disorders (BIRDEM), a largest referral center for diabetes in Bangladesh. Socio-demographic and clinical history including biochemical investigation report were collected from the BIRDEM registry. The eligible criteria of study participants were: age 30 – 60 year, having DM, non-smoker, free from retinopathy, nephropathy and neuropathy. The prevalence of CAD, systolic hypertension (SHTN) and diastolic hypertension (DHTN) in the registered diabetic patients were estimated. Additionally, the study addressed the risk and predictors of CAD among those with DM.

Investigations included – anthropometry, blood pressure, blood glucose, serum lipids and electrocardiogram (ECG). CAD was diagnosed on: (a) history of angina plus positive ECG - either on rest or on stress, post-myocardial infarction (MI) with Q-wave MI or non-Q-MI or echocardiographic evidences. Lipids namely triglycerides (TG), total cholesterol (T-Chol), high density lipoproteins (HDL) and low-density lipoproteins (LDL) were estimated by Hitachi-704 auto-analyzer using enzymatic method.

Results: A total of 693 (M /W =295/398) participants volunteered. The prevalence of CAD, SHTN, DHTN and mean arterial hypertension (MAH) were 18.6%, 23.2%, 13.6% and 17.7%, respectively. Their mean (±SD) values of age, body mass index (BMI - kg/m²), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR) and mean arterial pressure (MAP) were 47 (8.6) years, 24.6 (3.5), 0.98(0.05), 0.56(0.06) and 101(11.3) mmHg, respectively. The mean (±SD) of FBG (mmol/L), T-Chol, TG and HDL (mg/dl) were 10.2 ± 4.0, 206 ± 44, 218 ± 86 and 47.5 ± 9.3 respectively. The women had significantly higher BMI (p<0.001), WHtR (p<0.001), SBP (<0.001), MAP (p<0.001), T-Chol (p<0.001) and TG (p=0.043) than men. The risk variables were categorized into quartiles and Chi-sq trend determined whether the increasing prevalence of CAD were significant. Higher quartile of age was found consistently significant (p<0.001). Of the obesity indices, only higher quartile of WHtR was significant (p< 0.05). For BP measures, higher MAP quartiles showed the trend significant (p<0.001). Likewise, for lipids, higher quartiles of TG (p<0.001) and lower quartile of HDL (p<0.001) were significant.

Finally, logistic regression estimated the risk related to CAD. The highest age-quintile (>55y: 95% CI: 1.09 - 43.7) and highest TG-quintile (281mg/dl: 95% CI: 1.45-59.7) were proved to be significant predictor of CAD and HDL highest quintile (>54mg/dl) was proved to be significant protecting factor for CAD (95% CI: 0.005-0.583).

Conclusion: The study observed the importance of MAP, TG, HDL, T-Chol/HDLR (T-Chol -to HDL ratio) and TG/HDLR (triglycerides-to HDL ratio) as risks for CAD among diabetics. Further study with investigations of echocardiogram, ETT, coronary angiogram and coronary calcium scoring would be helpful in confirming these findings related to CAD risks.

IMC J Med Sci. 2023; 17(1): 002.  DOI: https://doi.org/10.55010/imcjms.17.002

*Correspondence:M Abu Sayeed, Department of Community Medicine, Ibrahim Medical College, 1/A, Ibrahim Sarani, Segunbagicha, Dhaka 1000, Bangladesh. Email: [email protected]

Introduction

According to World Health Organization (WHO) cardiovascular diseases (CVDs) are the leading cause of death, taking approximately 18 million lives annually globally [1]. There have been many newer published reports highlighting high arterial pressure and dyslipidemia as important risk for coronary artery disease (CAD) [2,3]. Four out of five CVD deaths are due to heart attacks and strokes [1]. One third of these deaths occur prematurely below (<70y). Several risk factors are shared between Type2 diabetes (T2D) and CAD, including obesity, insulin resistance and dyslipidemia [4,5]. CAD can precede type 2 diabetes (T2D), which is a major risk factor for CVD [6]. It may be recalled that 1in 10 adults of the world are now living with diabetes [7]. For Bangladeshi diabetic population the findings of coronary risks were reported mainly on age, sex, geographical site, occupation, obesity, hypertension, and glycemic control [8]. This study revisited the published report comparing the risk of CAD related to obesity, elevated mean arterial hypertension and dyslipidemia. It also showed the effect of high total cholesterol (T-Chol), high triglyceride (TG) and low high-density lipoprotein (HDL) on CAD separately. Additionally, it demonstrated the effect of T-Chol-to-HDL ration (T-Chol/ HDLR) and TG-to-HDL ratio (TG/HDLR) on CAD.

Materials and methods

Study design: Subjects and methods have been detailed in the previous published report [8]. Briefly the selection procedure is elaborated in the Figure-1. The duration of diabetes (mean ±S.D.) was 13.6 ± 3.6 (range 2–18) months. Informed consent was taken and they were interviewed for the clinical history related to initial investigations and diagnosis, smoking habits, family history of diabetes, HTN and atherosclerotic cardiovascular events and their drug history (if any). They were also interviewed for past illness about HTN and CAD followed by general and systemic examination.

Figure-1: Selection of study participants from the BIRDEM diabetes registry. CAD-coronary artery disease; SHTN-systolic hypertension; DHTN-diastolic hypertension; MAH-mean arterial hypertension (MAP >110mmHg)

Then, based on clinical findings, relevant investigations were undertaken in BIRDEM for confirmation ofthe diagnosis. The subjects with secondary HTN, cerebrovascular stroke, foot ulcer, nephropathy and retinopathy were excluded from the study. Those who were using corticosteroid and contraceptive pills were also excluded.

Anthropometric assessment included body mass index (BMI), waist-to-hip ration (WHR) and waist-to-height ration (WHtR). Mean arterial pressure (MAP) was estimated as (MAP = dbp + 1/3(sbp – dbp) [9]. For this study elevated (>110mmHg) MAP was considered. Systolic (SHTN) and diastolic (DHTN) hypertension were taken as SBP ≥140 and DBP ≥90 mmHg, respectively. Hypertensive subjects previously diagnosed were also included. Their BP was taken 2 days after cessation of anti-hypertensive drugs. World Health Organization (WHO) diagnostic criteria were used to diagnose diabetes mellitus.

The measurements of plasma glucose were done by glucose-oxydase peroxydase method using Technicon M-II autoanalyzer. All subjects underwent ECG-tracing except those with recent ECG reports. The diagnosis of CAD was based on: (a) history of angina plus positive ECG either on rest or on stress, post-myocardial infarction (MI) with Q-wave MI or non-Q-MI in echocardigraphic evidences. Lipids (TG, Chol, HDL, LDL) were estimated by Hitachi-704 auto-analyzer using enzymatic method. LDL-cholesterol (LDL) was measured using formula: LDL-C = 0.9 TC- (0.9 TG/5)-28 [18].

Statistical analysis: The prevalence rates (qualitative variables) were given in percentages. The quantitative variables were presented in means with standard deviation (SD). The comparisons between groups were estimated by unpaired t-test. The associations between anthropometrics and lipid fractions were estimated by Pearson’s correlations co-efficient. The prevalence trends (increasing / decreasing) were estimated by Chi-sq. Binary logistic regression analysis showed the effects of independent variables (obesity, blood pressure, lipids) on the dependent variables CAD. The significance levels of all statistical tests were taken at 0.05.

Results

A total of 693 (M=295, F=398) registered diabetic patients of age 30 – 60 year volunteered the study (Figure1). The prevalence of CAD was 18.6% (men vs. women = 16.6 vs. 20.2%; p = 0.139). The prevalence of systolic hypertension (SHTN) was 23.2% (men vs. women = 19.3 vs. 26.1%, p<0.05) and prevalence of diastolic hypertension (DHTN) was 13.6% (men vs. women = 11.2 vs. 15.3%, p =0.07). The prevalence of mean arterial hypertension (MAH) was 17.7% (men vs. women = 14.9 vs. 19.8, p = 0.058).

The biophysical characteristics of all participants were shown in Table-1a. The comparisons of these characteristics between men and women were shown in Table-1b. The comparison between age-matched 295 men and 398 women showed that the women had significantly higher BMI (p<0.001), WHtR (p<0.001), SBP (p<0.001), MAP (p = 0.002), T-Chol (p<0.001) and TG (p<0.05) than their male counterpart. Thus, most of the obesity and blood pressure related variables were higher in women than men, except WHR, which was significantly higher (p<0.001) in men.

Table-1a: Biophysical characteristics of the total participants

Table-1b: Comparison of biophysical characteristics between men and women

The Pearson’s correlation test, controlling age and sex, was used to determine the associations between obesity related variables (BMI, WHR, WHtR) and lipid fractions (T-Chol, TG, HDL, LDL) [Table-2]. Significant correlations of lipid-fractions were neither found with general (BMI) nor with central obesity (WHR, WHtR). The correlations of lipid fractions with BP measures (SBP, DBP and MAP) were shown in Table-3. Of the lipids, TG showed significant positive and HDL significant negative correlations with all BP measures though these correlations with T-Chol and LDL were not significant.

Table-2: Correlations (controlling for age and sex) between obesity and lipids related variables

Table-3: Correlations of lipid related variables (T-Chol, TG, HDL, LDL) with SBP, DBP and MAP (controlling for age and sex).

Table-4 depicted correlations (controlling for age and sex) between blood pressure and metabolic variables (T-Chol, TG, HDL, LDL, FBG). All BP measures (SBP, DBP and MAP) correlated significantly with T-Chol/HDLR and TG/HDLR, but not with FBG.

Table-4: Correlations of BP measures with metabolic variables like ratios of lipid fractions (CHOL/HDLR, TG/HDLR) and FBG.

Whether the prevalence of CAD was related to advancing age, increasing mean arterial pressure (MAP), TG and decreasing with increasing HDL level are shown in Figure-2. The prevalence of CAD according to quartiles of age, MAP, TG and HDL were estimated by chi-sq trend with level of significance (chi-sq, p). The trends were significant for the quartiles of age (33.6, <0.001), (MAP: 75.7, p<0.0001), TG (23.5, <0.001). As expected, HDL had inverse association with CAD prevalence (20.2, <0.001), which indicated that low HDL level had higher risk of developing CAD.

Figure-2:Prevalence (%) of CAD according to quartiles (Q1, Q2, Q3, Q4) of age, MAP, TG and HDL. Age (y): Q1<40, Q2 41- 47, Q3 48-55, Q4 >55; HDL mg /dl: Q1 <41, Q2 41-48, Q3  48-53, Q4 >53; MAP mmHg: Q1 <93, Q2 94-100, Q3 101- 106, Q4 >106.

The trend of CAD prevalence according to the quartiles of TG, HDL, T-Chol/HDLR and TG/HDLR are shown in Figure-3 for comparison. Very high prevalence of CAD was found in the highest quartile of TG and lowest quartile of HDL (for both, p<0.001). Importantly, the increasing quartiles of T-Chol / HDL ratio and TG / HDL ratio showed significant increasing trend of CAD prevalence. The trend of CAD prevalence with increasing obesity (quartiles of BMI, WHR, WHtR) is shown in Figure-4. The trends were not significant for BMI and WHR. The highest quartile of WHtR (Q4 >0.6) was found significant (p = 0.02).

Figure-3: Prevalence (%) of CAD according to quartiles (Q1, Q2, Q3, Q4) of TG, HDL, T-Chol / HDL ratio and TG / HDL ratio. The trends were significant (chi Sq, P) for increasing quartiles of TG (23.5, <0.001) and decreasing HDL (20.2, <0.001), T-Chol/HDL ratio (30.7<0.001) and TG/HDL ratio (30.7, <0.001). Quartile values of TG, mg / dl: Q1 <153, Q2 154 - 201, Q3 202 - 280, Q4 >280; Quartile values of HDL mg / dl: Q1 <41, Q2 41 - 48, Q3 48 - 53, Q4 >53; Quartile values of TG/HDL; Q1 <3.03, Q2 3.04 - 4.23, Q3 4.24 - 6.77, Q4 >6.77; and Quartile values of T-Chol/HDL ratio: Q1 <3.75, Q2 3.76-4.83, Q3 4.84-5.63 and Q4 >5.63. * Values for HDL, cholesterol and TG, and ratios were estimated in mg/dL.

Figure-4: Prevalence (%) of CAD according to quartiles (Q1, Q2, Q3, Q4) of BMI, WHR and WHtR.. Quartile values of BMI: Q1 <22.2, Q2 22.3 - 24.3, Q3 24.4 - 26.5, Q4 >26.5; Quartile values of WHR: Q1 <0.95, Q2 0.96 - 0.98, Q3 0.99 -1.01, Q4 >1.01; Quartile values of WHtR Q1 <0.52, Q2 0.53 - 0.55, Q3 0.56 - 0.6, Q4 >0.6.

Some inconsistent findings emerged when we tried to determine the risks related to CAD among the our diabetic study population. The investigated variables were age, sex, sites (urban/rural), family history of NCDs, obesity, blood pressures and lipids. Of these risk factors, which were more significant remained unclear.

We used binary logistic regression taking the risk factors as independent and CAD as dependent variable. Of the independent variables (sex, area, age, BMI, WHR, WHtR) only higher age quartile (Q3 and Q4) was proved to be a significant risk for CAD [Table-5]. In Table-6, the independent variables were sex, age and lipid fractions. The highest quartile of age and TG, and the lowest quartile of HDL were found significant risk for CAD.

Table-5: Binary logistic regression taking coronary artery disease (no =0, yes =1) as a dependent variable; and sex, area, age, BMI, WHR, WHtR as independent variables. The categorical variables are depicted below

Table-6: Binary logistic regression taking coronary artery disease (no =0, yes =1) as a dependent variable and sex, age, cholesterol, TG, HDL as independent variables. The categorical variables are depicted below.  

Discussions

The study was conducted on purposively selected registered diabetic patients of a referral center, BIRDEM. The prevalence of CAD was 18.6% [Figure-1], which was more or less consistent with the findings of the systemic review report published earlier [9]. In the review, 21.2% had coronary heart disease (42 articles, N = 3,833,200). Higher prevalence of cardiovascular disease in patients with type 2 DM was 37.4% (95% CI: 31.4-43.8) in Iran [10]. Another study from Bangladesh reported the prevalence of CAD as 17.2% [11]. The study found that it had no significant difference between gender and CAD. The present study observed higher prevalence of CAD in women than men (20.2% vs. 16.6 %) though not significant.

Interestingly, the age matched women had significantly higher BMI (p<0.001), WHtR (p<0.001), SBP (p<0.001), MAP (p = 0.002), T-Chol (p<0.001) and TG (p<0.05) than men. Only, WHR was significantly higher in men than women (p<0.001).

The measures of obesity (BMI, WHtR), blood pressure (SBP, DBP, MAP) and TG were higher in women than men and consistent with the higher prevalence of CAD in women, though the difference was not significant. Most of the cited studies reported higher CAD prevalence in men than women [5,7,10,11]. This contradictory finding could have been explained if in the study there were equal numbers of female participants from rural population. The BIRDEM diabetes registry revealed that more than 30% of women were occupationally urban housewives and they lack physical activity resulting obesity with dyslipidemia.

The associations between variables of obesity and lipids (Table-2), and lipids and blood pressures (Table-3) revealed that none of lipid fractions correlated with obesity significantly. Of the lipid fractions, TG and HDL (and not T-Chol and LDL) showed very significant association with SBP, DBP and MAP. This indicated that T-Chol and LDL were not related to blood pressure. These findings could not be compared with any published data that studied lipid fractions separately in relation to SBP, DBP and MAP. Anika et al showed significant correlation between total cholesterol and systolic blood pressure, also between triglyceride and diastolic blood pressure [12]. Other studies found total cholesterol was positively associated with IHD mortality in both middle and old age [13,14].

Interestingly, this study revealed that T-Chol/HDLR and TG/HDLR correlated with all types of BP measures (Table-4 and Figure-3). This observation is very much consistent with other Bangladeshi report [15]. Of the obesity indices only WHtR proved to have significant risk at Q4 (>0.6). This study proved WHtR to be a better obesity index for CAD. The highest quartile of mean arterial pressure (MAP >106mmHg) was found to be a significant risk for CAD. The importance of MAP was also emphasized by Gao et al [2]. The study showed the importance of T-Chol/HDLR and TG/HDLR for predicting CAD in diabetic population. This observation is very much consistent with the findings of other studies. [15-17].

The study had some limitations. Firstly, glycemic control could not be monitored for the follow-up period after registration. Secondly, the number of women was not proportionate to the geographical sites (urban/rural). Thirdly, physical activity of the study participants could not be graded. Lastly, the diagnosis of CAD was based on only on ECG findings.

Conclusions

The study revealed the prevalence of coronary artery disease (CAD), systolic hypertension and diastolic hypertension in the registered Bangladeshi diabetic patients. It identified the risk factors for developing CAD. Additionally, the study addressed the possible predictors of CAD among those with DM. The study observed the importance of MAP, TG, HDL, T-Chol/HDLR and TG/HDLR as predictors of CAD. Further study along with the investigation of echocardiography, ETT, coronary angiogram and coronary calcium scoring would be helpful in confirming these findings related to CAD risks.

Acknowledgements – We are grateful to those participants who actively volunteered the study. We are also indebted to the doctors and other official staff of BIRDEM-OPD. The BIRDEM registry office helped in obtaining the records of the newly registered patients with the “REFERECE No” and laboratory technician with biochemical reports. We would like to convey our gratitude to the departed souls of Dr. Fazlul Hoque and Dr. Khandoker Abul Ahsan. We commemorate both of them for their whole hearted cooperation to get the study complete.

References

1.     World Health Organization. (Accessed date 20 Dec 2021). Available from: https://www.who.int/ health-topics/cardiovascular-diseases.

2.     Gao Y, Wang Q, Li J, Zhang J, Li R, Sun L, et al. Impact of mean arterial pressure fluctuation on mortality in critically ill patients. Crit Care Med. 2018; 46(12): e1167-e1174.

3.     Pires A, Sena C, Seiça C. Dyslipidemia and cardiovascular changes in children. Curr Opin Cardiol. 2016; 31(1): 95-100.

4.     Laakso M, Kuusisto J. Insulin resistance and hyperglycaemia in cardiovascular disease development. Nat Rev Endocrinol. 2014; 10(5): 293-302.

5.     Silva LF, Vangipurapu J, Laakso M. The "Common Soil Hypothesis" revisited- risk factors for type 2 diabetes and cardiovascular disease. Metabolites. 2021; 11(10): 691.

6.     International Diabetes Federation. (accessed on 20 Dec 2021). Available from: https://www.idf.org/

7.     Pursnani S, Merchant M. South Asian ethnicity as a risk factor for coronary heart disease. Atherosclerosis. 2020; 315: 126-130.

8.     Sayeed MA, Banu A, Malek MA, Khan AKA. Blood pressure and coronary heart disease in NIDDM subjects at diagnosis: prevalence and risks in a Bangladeshi population. Diab Res Clin Pract. 1998; 39: 147-155.

9.     Einarson TR, Acs A, Ludwig C, Panton UH. Prevalence of cardiovascular disease in type 2 diabetes: a systematic literature review of scientific evidence from across the world in 2007–2017. Cardiovasc Diabetol. 2018; 17(1): 1-9.

10.  Kazeminia M, Salari N, Mohammadi M. Prevalence of cardiovascular disease in patients with type 2 diabetes mellitus in Iran: a systematic review and meta-analysis. J Diabetes Res. 2020; 2020. 

11.  Hanif AAM, Hasan M, Khan MSA, Hossain MM, Shamim AA, Hossaine M, et al. Ten-years cardiovascular risk among Bangladeshi population using non-laboratory-based risk chart of the World Health Organization: findings from a nationally representative survey. PLoS One. 2021; 16(5): e0251967.

12.  Anika UL, Pintaningrum Y, Syamsun A. Correlation between serum lipid profile and blood pressure in NTB general hospital. J Hypertens. 2015; 33: e32.

13.  Wong ND, Lopez VA, Roberts CS, Solomon HA, Burke GL, Kuller L, et al. Combined association of lipids and blood pressure in relation to incident cardiovascular disease in the elderly: the cardiovascular health study. Am J Hypertens; 2010; 23(2): 161–167.

14.  Prospective Studies Collaboration, Lewington S, Whitlock G, Clarke R, Sherliker P, Emberson J, et al. Blood cholesterol and vascular mortality by age, sex, and blood pressure: a meta-analysis of individual data from 61 prospective studies with 55,000 vascular deaths. Lancet. 2007; 370(9602): 1829-1839.

15.  Amin MR, Rahman MA, Alam N, Hasan MN, Hasan GS. Relationship between triglyceride HDL-cholesterol ratio and severity of coronary artery disease in patient with acute coronary syndrome. Bangladesh Med J. 2014; 43(3): 157–161.

16.  Calling S, Johansson SE, Wolff M, Sundquist J, Sundquist K. Total cholesterol/HDL-C ratio versus non-HDL-C as predictors for ischemic heart disease: a 17-year follow-up study of women in southern Sweden. BMC Cardiovasc Disord. 2021; 21(1): 1-9.

17.  Holman RR, Coleman RL, Shine BS, Stevens RJ. Non-HDL cholesterol is less informative than the total-to-HDL cholesterol ratio in predicting cardiovascular risk in type 2 diabetes. Diabetes Care. 2005; 28(7): 1796–1797.

18.  Anandaraja S, Narang R, Godeswar R, Laksmy R, Talwar KK. Low density lipoprotein cholesterol estimation by a new formula in Indian population. Int J Cardiol. 2005; 102(1): 117–120.

Cite this article as:

Banu A, Hoque F, Ahsan KA, Sayeed MA. Do obesity, hypertension and dyslipidemia pose significant risks for coronary artery disease among Bangladeshi diabetics? IMC J Med Sci. 2023; 17(1): 002. DOI: https://doi.org/10.55010/imcjms.17.002