In this population-based study (2008–2011) about 13% of urban Tehranian residents were found to have CHD, of whom about 9% and 4% had clinical CHD and silent MI respectively. Generally, in comparison with women, men had higher prevalence of total and clinical CHD. In addition, the prevalences of total CHD and silent MI were about 18% and 3% among male diabetic participants and were about 19% and 6% among their female counterparts, respectively. Despite, the higher prevalence of clinical CHD among diabetic participants, our findings did not show a higher prevalence of silent MI among diabetic participants compared to other glycemic categories in either gender. Finally, the NDM group and those suffering from both IFG and IGT showed similar high prevalences for total CHD (about 16%), which were higher than that of the NGT group.
In our previous study, using self-reported history of CHD and ECG-defined ischemia for defining CHD, a 15.1% prevalence of CHD was reported for Tehranian adults in 1999–2001[4]. The differences between the prior study and the current study might be attributable to the following factors. Firstly, in a previous of ours, the prevalence of silent ischemia was reported to be > 10% in both sexes, significantly higher than those reported in the current study. It should be noted that in our previous study, we included both probable and possible Minnesota coding of CHD, while in the current study, only definite Minnesota coding of MI was included as diagnostic criteria[26, 27]. Secondly, for history of CHD, it was considered positive, only when its hospital records were provided and then confirmed by the outcome committee. Hence despite a strong association between Rose angina and incident CHD events which reported among tehranian population[28], we did not include it in the clinical CHD definition. Hence, in the current study, considering the stringent criteria for definition of clinical CHD and silent MI led to underestimations for prevalence of CHD.
It is important to note that due to different diagnostic criteria for CHD definition and different study age-group categorization, comparing our results with other population-based studies is somewhat difficult. Abbasi et al. reported that among an Iranian population, aged over 20 years, the national prevalence of self-reported CHD was 5.3% (5.6% among urban residents) in 2011[29]; their values for prevalence of self-reported CHD were significantly lower than our reports. Moreover, in comparison with our results for silent MI, in a cross-sectional population-based study in southern Iran, Nabipour et al. reported that 1.4% of their participants (1.9% of men and 1% of women), aged 25–64 years were diagnosed to have ECGs with evidence of previous MI (Minnesota codes 1.1 and 1.2)[30]. Comparing our results with the two studies previously mentioned suggests that Tehranian residents rank high in the prevalence of CHD among Iranian populations.
Compared to developed countries, based on a national study, the American Heart Association reported the total CHD prevalence to be 6.7% among US adults, aged ≥ 20 years (7.4% for men and 6.2% for women)[31]. Furthermore, data from the Quebec Integrated Chronic Disease Surveillance System (QICDSS) indicated (ⅰ) that in 2012/2013, the crude prevalence of CHD was 9.4% among an adult Canadian population and (ⅱ) that from 2000/2001 to 2012/2013 the age-standardized prevalence had increased by 14%, despite having a slight decreasing trend since 2009/2010[32]. For the United Kingdom (UK), data from the Quality and Outcomes Framework (QOF) indicated that the prevalence of CHD remained constant at about 3% in England and 4% in Scotland, Wales, and Northern Ireland between 2004/2005 and 2014/2015[33]. Among Asian countries, results of the China National Diabetes and Metabolic Disorders Study showed that the prevalences of self-reported CHD were 0.72% and 0.48% among male and female adult Chinese, respectively[34]. Also the prevalence of CHD from national studies varies from 2–4% in India[35]. Furthermore, in Saudi Arabia, as a Middle Eastern country, the age-adjusted prevalence of CHD was reported to be 5.9% among men and 4.4% among women, aged 30–70 years[36]. Generally, it seems that the estimated prevalence of CHD among Tehranian populations are alarmingly higher than corresponding figures in US[31], Canada[32], UK[33], China[34], India[35] and Saudi Arabia[36], an issue previously addressed in 2015 by Zhu et al.[37] As we reported previously, modifiable risk factors including diabetes, hypertension, smoking and dyslipidemia totally had population attributable fraction of 36.6% and 50.2% of CHD among male and female Tehranian populations, respectively[5]. Other reasons that might justify the high prevalence of CHD among Tehranian populations are related to impact of air pollution[38–40] and stress[41] both of which are common in Tehran.
Focusing on diabetes status, as reported by a national study in 2016, among Iranian diabetic patients, aged ≥ 18 years, the crude prevalences of clinical CHD were 25.1% for men and 23.2% for women[42]; the corresponding values were 27.88% and 25.66% among our diabetic population. It seems that similar to total papulations, among the Tehranian diabetic population, the prevalence of CHD was higher than that reported in nationwide studies. The age-standardized prevalences of clinical CHD were 4.43% and 4.76% among male and female Chinese patients with T2DM, respectively[43]. Moreover, the prevalence of CHD among Thai patients with diabetes was 3.54% in 2013[44]. Also among Swedish diabetic patients, aged 45–74 years, the crude prevalences of CHD were 24.9% for men and 18.0% for women, which were lower than our results, despite the fact that their population was older than ours[45]. In addition, a significant racial difference was reported in the prevalence of CHD between White and African diabetic patients in a hospital-based study[46]. It has been suggested that there is a racial susceptibility for CHD among diabetic patients, which could make Iranian diabetic patients more prone to developing CHD, compared to Asian, African and European ethnicities. Furthermore, although CVD risk factors among Iranian diabetic populations have been controlled to some degree, during recent years, most diabetic participants still have uncontrolled CVD risk factors[47] which could also have led to high prevalence of CHD among our diabetic population.
In the current study as expected the highest prevalence of CHD was shown among diabetic participants; however, we also found that the NDM group showed > 16% prevalence of CHD. We have previously reported that during a 7.6 years follow-up, Tehranian adults with NDM, especially in men, exhibited a CHD risk comparable to non-DM with a prior CHD[48]. Also, regarding prediabetes status, the prevalence of CHD become more prominent among participants with combined IFG and IGT, similar to NDM groups (Fig. 2). Based on angiographic data, among a non-diabetic population, it was reported that participants with combined IFG and IGT had higher prevalence of significant CHD and higher severity of disease; however, there were no significant differences among subjects with NGT, I-IFG, and I-IGT[49].
In the current study, we found that about 25% (4.48/18.18) of total CHD among the diabetic population was attributed to silent MI. Previous studies reported different prevalences of silent MI among diabetic populations, considering the different tools used to assess infarction, ranging from 3.9% (by ECG-criteria) to 37% (by dipyridamole thallium scintigraphy)[16]. We also observed a high prevalence of silent MI reach to 6.8% among combined the IFG and IGT population, although our findings did not support significant difference in the prevalence of silent MI across glycemic status categories, probably due to the limited number of silent MI in each group. Recently among an elderly population it was also shown that subjects with IFG had no increased risk of silent MI than those with NFG. Furthermore, Bhatt et al. found an independent association between Q-waves and the homeostasis model assessment of insulin resistance (HOMA-IR)[50]. To the best of our knowledge, no previous study has examined the prevalence of silent MI among strictly defined glucose tolerance groups, using both FPG and 2 h-PCPG criteria. However two studies showed participants with IGT[51] and IFG[52] had higher prevalence of silent MI than the NGT group. Some authors have speculated that prediabetes and diabetes status cause cardiovascular autonomic neuropathy which could imply some degree of cardiac pain suppression[52]. Therefore, in the light of this mechanism, diabetic and prediabetic populations are more prone to silent CHD and silent MI.
Strengths of this study include using standardized ECG procedures for defining stringent criteria of MI, definite documented CHD ascertained by an outcome committee and finally the determination of CHD prevalence among the whole population, based on glycemic categories, using the glucose challenge test.
Several limitations need to be acknowledged. First, our study shows an optimistic picture of CHD prevalence among our population since inclusion of subjects in an ongoing study can improve the level of attention paid to controlling their health risks (cohort effect). Therefore, the burden of CHD will be much higher in the context of the community. Second, this investigation was conducted among residents of Tehran as a metropolitan city and our results might not be generalizable to rural zones. Third, we have only examined non-fatal CHD, whereas the prevalence of CHD is much higher, by when considering the burden of fatal CHD. Fourth, because of our large population-based sample, it’s not practical and reasonable to apply some modalities including stress test, positron emission tomography and angiography for all participants to detect silent ischemia[53]. Finally, the sensitivity and specificity of ECG criteria for detecting previous MI is limited[27] and patients with acute non-Q-wave MI do not necessarily develop Q waves eventually[15]. Also some of the Q waves disappear after MI[26].