Abstract
The aim of this study was to determine the prevalence of key cardiovascular risk factors in the Middle East region. We conducted a systematic review of the literature through searches in the MEDLINE/PubMed and PARLINE databases between January 1980 and April 2005. Cohort studies published from 1980, in English, which included at least 1000 participants that reported the prevalence of at least one of the following; diabetes mellitus, obesity (body mass index ≥ 30 kg/m2), hypertension, hyperlipidemia, and smoking in the Middle East region. Data were abstracted using standardized data abstraction forms. Studies were combined using random-effect models. In total, 51 studies (267 537 participants) were included. On the basis of a random-effect model, the overall prevalence of obesity was 24.5% [95% confidence interval (CI): 21.8-27.5; I 2: 99.3%; 24 studies], diabetes mellitus was 10.5% (95% CI: 8.6-12.7%; I 2: 99.4%; 24 studies), hypertension was 21.7% (95% CI: 18.7-24.9; I 2: 99.5%; 24 studies), smoking was 15.6% (95% CI: 12.3-19.6%; I 2: 99.7%; 21 studies). Smoking was more common in men than women, whereas obesity and hypertension were more common in women. The overall prevalence was not calculated because of marked variations in the definition of dyslipidemia among studies. There is a high prevalence of diabetes mellitus, obesity, hypertension, and smoking in the Middle East. The prevalence of obesity and hypertension was higher in women, whereas prevalence of smoking was higher in men. These data suggest that cardiovascular disease will be a major health problem in the Middle East.
Keywords
Introduction
In 2005, 60% of deaths worldwide were because of chronic diseases [1], with cardiovascular diseases (CVD) being the most common cause [2]. In terms of global burden of disease, the World Health Organization (WHO) has identified obesity, hypertension, hypercholesterolemia, and smoking among the top 10 risk factors for premature death and disability [3]. Over the last century, in most populations, there has been a dramatic shift in the causes of death and disability with a decrease in deaths from nutritional deficiencies and infectious diseases, and an increase in deaths from noncommunicable diseases such as CVD. This ‘epidemiologic transition’ [4] initially began in high-income countries, but has now extended to many middle and low-income countries including those in the Middle East region. These transitions are in part driven by decrease in childhood mortality from malnutrition and infectious diseases, increase in tobacco consumption, and urbanization leading to decrease in physical activity, and changes in diet leading to obesity and increase in other risk factors [5, 6].
The INTERHEART study [7] has showed that nine modifiable risk factors are responsible for more than 90% of the population-attributable risk for CHD. The most important risk factors are obesity, diabetes mellitus, hypercholesterolemia, smoking, and hypertension. As these risk factors are usually evident many years before developing cardiovascular disease, knowledge of their prevalence in populations undergoing ‘epidemiologic transition’ allows us to predict the likely burden of cardiovascular disease in the approaching decades [8].
We conducted a systematic review of the currently available data from the Middle East, to summarize the available data on the prevalence of obesity, diabetes mellitus, hypercholesterolemia, smoking, and hypertension and determine whether the prevalence of these risk factors varied by sex.
Methods
Data sources
A comprehensive search in the MEDLINE/PubMed was conducted for articles published from January 1980 to April 2005 in the Middle East region, that reported the prevalence of any of the following: diabetes mellitus, hyperlipidemia, smoking, hypertension, and obesity. This region based on PARLINE database [9] included the following countries: Saudi Arabia, Kuwait, Oman, Qatar, Lebanon, Iran, Iraq, Israel, Gaza and the West Bank, Lebanon, Syria, Egypt, Yemen, United Arab Emirates, Turkey, and Bahrain. The MeSH search term used were the name of above countries with each of above risk factors as well as the following terms: ‘prevalence’, ‘epidemiology’, ‘Eastern Mediterranean’, and ‘Middle East’. The search was carried out on PUBMed and MEDLINE. Bibliographies of retrieved articles were scanned for additional articles.
Study selection
The eligibility criteria applied for all studies were: (i) prospective and retrospective cohort studies published since 1980 (studies published before 1980 were excluded to facilitate the generation of contemporary estimates that used standardized definitions), (ii) studies that included at least 1000 people, (iii) reported the prevalence of at least one of the following risk factors: hypertension, diabetes mellitus, hypercholesterolemia, and obesity and, (iv) studies where a full manuscript was published in English. For reporting of individual risk factors, the following criteria was required: (i) body mass index (BMI), calculated as weight (kilogram)/height (square meter), was used to define obesity (BMI ≥ 30 kg/m2) and, (ii) diabetes mellitus was defined based on either fasting glucose or glucose tolerance cut-point (studies that used only self-report of history of diabetes without any plasma glucose measurements were excluded). Studies that included prevalence of risk factors for only men or women were also not included.
Data abstraction
For each study, one author (B.M.) abstracted the following data: (i) proportion of patients with hypertension, diabetes mellitus, obesity, hypercholesterolemia, and/or smoking, (ii) mean or median age, (iii) sex, and (iv) proportion of participants living in rural or urban areas.
Statistical analysis
Statistical analysis was performed using the Statistical Package for Social Sciences (SPSS 13.0, SPSS Production Facility, Chicago, Illinois, USA) and Comprehensive Metaanalysis software packages (Comprehensive Meta-analysis, version 2, Biostat, Englewood, New Jersey, USA). Anticipating heterogeneity, the summary estimate for prevalence of obesity, diabetes, hypertension, and smoking were calculated using a random-effect model, using weighted data based on the sample size. P values less than 0.05 were considered significant. Tests of heterogeneity were carried out, although we expected significant heterogeneity between studies because studies were conducted in different countries at different time periods by using different sampling methods.
Results
The systematic literature search of articles identified 51 studies (with a total of 267 537 participants) that met our eligibility criteria (Appendix A). The characteristics of the study populations, including mean age, response rate, and the sampling methods, are summarized in Appendix A. Diabetes mellitus was defined using the WHO Criteria 1985 [fasting plasma glucose (FPG) levels of ≥ 7.8 mmol/l or 2-h 75-g Oral Glucose Tolerance Test of ≥ 11.1 mmol/l] in 15 studies [10], the WHO Criteria 1999 (FPG ≥ 7mmol/l or random plasma glucose of ≥ 11.1 mmol/l) in five studies [11], and FPG ≥ 6.7 mmol/l or 2-h 75-g Oral Glucose Tolerance Test of ≥ 10 mmol/l in three studies. Although the majority used systolic blood pressure (≥ 140 mmHg) or diastolic blood pressure ≥ 90 mmHg to define hypertension, three studies used systolic blood pressure of ≥ 160 or diastolic blood pressure of ≥ 95 mmHg [12, 13].
Prevalence of obesity
The overall prevalence of obesity (BMI ≥ 30 kg/m2), reported in 25 studies (135 048 participants), was 24.5% [95% confidence interval (CI): 21.8-27.5; I 2: 99.3%]. In the 21 studies (125 519 participants) that reported sex, the prevalence of obesity was greater in women than men [odds ratio (OR): 2.2; 95% CI: 1.9-2.6; I 2: 96.4%], with an estimated prevalence of 30.6% (95% CI: 27.1-34.3; I 2: 98.9%) in women and 16.6% (95% CI: 14.0-19.6; I 2: 98.7%) in men (Appendix C).
Prevalence of diabetes mellitus
The overall prevalence of diabetes mellitus, reported in 26 studies (152 279 participants), was 10.5% (95% CI: 8.6-12.7%; I 2: 99.4%). In the 18 studies that reported sex (84 865 participants), diabetes mellitus was comparable in women and men (OR: 1.0; 95% CI: 0.9-1.2; I 2: 86.2%) with an estimated prevalence of 9.3% (95% CI: 7.4-11.7; I 2: 97.4%) in women and 9.6% (95% CI: 8.0-11.6; I 2: 98.1%) in men. In studies that reported both obesity and diabetes mellitus (16 studies, 95 036 participants) we found a significant correlation between the two risk factors (r 2 = 0.6; P = 0.01).
Prevalence of hypertension
The overall prevalence of hypertension, reported in 24 studies (135 873 participants), was 21.7% (95% CI: 18.7-24.9; I 2: 99.5%). In the 18 studies that reported sex (85 132 participants), hypertension was greater in women than men (OR: 1.2; 95% CI: 1.04-1.4; I 2: 91.8%), with an estimated prevalence of 23.0% (95% CI: 19.0-27.7; I 2: 99.0%) in women and 20.1% (95% CI: 16.6-24.0; I 2: 98.6%) in men.
Prevalence of smoking
The overall prevalence of smoking, reported in 21 studies (148 023 participants), was 15.6% (95% CI: 12.3-19.6%; I 2: 99.7%). Smoking was more common in men than women (OR: 13.7; 95% CI: 9.8-19.2; I 2: 97.4%) in the 16 studies (102 471 participants) that reported sex distribution. The prevalence of smoking was 2.9% (95% CI: 1.8-4.6; I 2: 99.1%) in women and 28.8% (95% CI: 23.3-35.0; I 2: 98.3%) in men.
Prevalence of hypercholesterolemia
The overall prevalence for this risk factor was not calculated because of lack of a consistent definition for dyslipidemia in this study sample. However, among studies carried out in the 1990s or later, the rate of hypercholesterolemia in men ranged from 7.0% in Saudi Arabia [14] to 42.2% in Kuwait [15], whereas for women it ranged from 6.0% in Turkey [16] to 33.3% in Kuwait [15]. The prevalence of elevated low-density lipoprotein (LDL) in the Middle East varied between 6.6-36.3% for men and 9.0-38.7% in women (Appendix E).
Discussion
Our findings reveal a high prevalence of obesity, diabetes mellitus, hypertension, and smoking in the Middle Eastern region, although there was considerable variation in reported prevalence of risk factors among studies.
The overall prevalence of obesity in the Middle East, based on the studies in this review, is comparable with some of the rates reported from several high-income and western countries, such as the United States (28.0% of men and 34.0% of women) [17] and the United Kingdom (23.0% in men and 25.0% in women) [18]. Among the different countries in the Middle East region there was marked heterogeneity between studies. Areas including Jordan (49.7%), Oman (30.8%), Qatar (40.8%), and Gaza and West Bank (41.5%) had extremely high prevalence of obesity. Overall, obesity was also found to be more prevalent in urban areas compared with rural regions (Appendix B). This is most likely because of rapid economic and nutritional transitions in this region. Our analysis showed that the overall prevalence of obesity in the Middle East region was considerably higher among women compared with the men. Although the exact cause of such sex variations is not entirely clear, it has been reported that women are less active compared with men in certain areas [19, 20]. Physical and cultural barriers to physical activity have been reported among women in Egypt and Saudi Arabia. These include climatic conditions of extreme heat in the summer, limited exercise facilities devoted solely for women, lack of physical education or an emphasis on its importance in schools, and absence of women's participation in organized sports for most part. Physical inactivity and sedentary lifestyle are major risk factors for the development of obesity and CVD [7].
The prevalence of diabetes mellitus also varied among studies, with highest rates in Israel, Jordan, Qatar, and Saudi Arabia (Appendix D). Surprisingly, we did not observe a difference in prevalence of diabetes mellitus between men and women given differences observed for prevalence of obesity. However, in studies that reported the prevalence of both diabetes mellitus and obesity, we found a significant correlation between the prevalence of these two risk factors. According to the World Health Report 2002 [3], approximately 58% of diabetes mellitus globally occurs in individuals with a BMI above 21 kg/m2 [21]. For countries that included studies separated by approximately 10 years or more (Turkey, Oman, and Saudi Arabia), our findings suggest an increasing prevalence in diabetes mellitus, similar to trends observed for obesity. However, the limited number of studies included and differing study methodologies preclude definitive conclusions.
The prevalence of hypertension in the Middle East, which we report seems similar to rates observed in North America (28%), but lower than Europe (44%) [22]. Overall, hypertension was more prevalent in women than men. This may, in part, be explained by the higher prevalence of obesity in women, as several studies have reported an association between obesity and hypertension [23, 24]. The number of studies that have compared rates of hypertension in urban and rural areas in the Middle East is limited and we are unable to determine whether differences exist.
The definitions used for elevated cholesterol, LDL, and triglyceride levels, varied considerably among studies. These variations, as well as the availability of limited data from this region, make it difficult to conduct any cross-country comparisons with confidence. However, among studies published after 1990 (Appendix F), the rates of hypercholesterolemia varied from 7.0 to 42.2% in men and 6.0 to 33.3% in women. The prevalence of elevated LDL in the Middle East ranged between 6.6-36.3% for men and 9.0-38.7% in women. Notable observations include the very high prevalence of hypercholesterolemia (> 5.2 mmol/l) in Kuwait and Oman [15, 25], and the very low levels of HDL (< 35mg/dl) in Turkey, Oman, and Saudi Arabia (Appendix F).
The prevalence of smoking is high in the Middle East in men, but it is relatively uncommon in women, in large part because it is still not culturally acceptable for women to smoke in this region (Appendix G). The rates of smoking in Turkey were among the highest in the region, with one study reporting a prevalence of 64.8% between men and 20.1% women [26]. Smoking is known to play a major role in cardiovascular mortality. According to Ezzati et al. [27], an estimated 11% of total global cardiovascular deaths in 2000 were attributed to smoking, half of which occurred in developing countries. The proportion of deaths attributed to smoking in developing nations is proposed to increase, as smoking rates continue to increase in these countries, whereas gradually declining in developed world [28].
Limitations
A number of limitations deserve mention. First, the studies included in this study varied in design, population included (e.g. eligibility by age), definition of risk factor, and year of publication. Expectedly, we observed considerable heterogeneity between studies with I 2 in excess of 90%. Therefore, it is arguable whether summary estimate should be presented because of the considerable heterogeneity. However, our objective was not to provide very accurate summary estimates, but rather, present a general approximation of the prevalence of these risk factors to facilitate the message. We chose a random-effect model for all summary estimates because of the anticipated heterogeneity. Second, our search was restricted to studies published in English. Therefore, it may be possible that we missed some studies published solely in local journals in Arabic or Persian, although such studies are expected to have smaller sample sizes. Third, the lack of standardization of definitions of dyslipidemia limits our ability to provide summary estimates for this important risk factor.
Conclusion
We report a high prevalence of key cardiovascular risk factors in the Middle East. Given the ongoing and rapid urbanization of a relatively young population in this region, a large rise in cardiovascular disease over the next few decades is anticipated. Targeted educational programs on nutrition and healthier lifestyle for higher risk individuals such as children, and women who play an important role in family nutrition, may help to reduce the prevalence of these risk factors in the region.
Footnotes
Acknowledgements
The authors acknowledge Rahim Rezaie and Dr CH Goldsmith for their editorial contribution and constructive feedback. There are no disclosures and funding.
Appendix
Prevalence of smoking in the Middle East
|
|
|||||
|---|---|---|---|---|---|
| Prevalence (%) | |||||
| Country study | Year(s) of survey | Number (n) | Male | Female | Total |
|
|
|||||
| Bahrain | |||||
| Hamadeh et al. [29] Age: ≥ 19 (years) | 1981-1983 | 9282 | 33.1 | 9.2 | 21.5 |
| Iran | |||||
| Sarrafzadegan and Amini Nik [33] Age: 19-70 (years) | 1993-1994 | 8639 | NA | NA | U: 20.0 R: 21.0 |
| Sarrafzadegan et al. [34] Age: 19-70 (years) | 1999 | 2200 | 26.0 | 1.4 | 11.0 |
| Sarrafzadegan et al. [12] Age: ≥ 19 (years) | 1999 | 8624 | 29.0 | 1.6 | 12.5 |
| Azizi et al. [37] Age: > 15 (years) | 1999-2001 | 12030 | 22.3 | 2.1 | 10.6 |
| Ahmadi et al. [38] Age: 16-90 (years) | 2001 | 1335 | 26 | 3.6 | 16.7 |
| Sarrafzadegan et al. [35] Age: > 19 (years) | 2004 | 2626 | 21.2 | 6.7 | 12.9 |
| Jordan | |||||
| Ajlouni et al. [13] Age: ≥ 25 (years) | 1994-1996 | 2836 | 31.7 | 6.2 | 15.6 |
| Kuwait | |||||
| Olusi et al. [46] Age: ≥ 15 (years) | 2001 | 7609 | 31.7 | 1.4 | 17.4 |
| Memon et al. [45] Age: ≥ 18 (years) | 1996 | 3859 | 34.4 | 1.9 | 17.0 |
| Oman | |||||
| Hasab et al. [50] Age: ≥ 18 | 1994 | 4732 | 23.5 | 1.6 | 11.8 |
| Al Riymi and Afifi [25] Age: ≥ 20 (years) | 2004 | 7011 | 13.4 | 0.5 | 7.9 |
| Qatar | |||||
| Bener et al. [53] Age: 25-65 (years) | 2003 | 1208 | 21.7 | 11.6 | 15.8 |
| Saudi | |||||
| Al-Nuaim [58] Age: 30-64 (years) | 1990-1993 | 2059 | 21 | 1.0 | 11.1 |
| Al-Nuaim et al. [14] Age: > 15 (years) | 1990-1993 | 4539 | 20.0 | 0.9 | 12.0 |
| Jarallah et al. [59] Age: > 15 (years) | 1990-1993 | 8310 | 21.1 | 0.9 | 12.0 |
| Al-Nozha et al. [65] Age: 30-70 (years) | 1995-2000 | 16 917 | NA | NA | 12.8 |
| Turkey | |||||
| Mahley et al. [16] Age: 20-100 (years) | 1990-1993 | 9000 | 55.3 | 19.3 | 43.6 |
| Satman et al. [68] Age: ≥ 20 (years) | 1997-1998 | 24788 | 50.9 | 10.9 | 28.9 |
| Tezcan et al. [26] Age: 25-64 (years) | 1999 | 1672 | 64.8 | 20.1 | 37.9 |
|
|
|||||
NA, not available; R, rural; U, urban.
