The main objective of the current study was to identify metabolic markers that are associated with traditional cardiovascular disease risk factors and MetS in Arab population. Our study showed sexual dimorphism in plasma level of leptin, PAI-1, hsCRP, resistin and adiponectin with women showing higher levels of these metabolic markers in Arabs. hsCRP showed significant positive correlation with all risk factors except for HDL where it was inversely associated. Leptin, PAI-1 and adipsin showed positive correlation with BMI. Visfatin level was not associated BMI level unlike other adipokines. Adiponectin showed highest association with CVD risk factors as it negatively associated with age, BMI, sex, SBP, HOMA-IR and LDL. It was also negatively associated with increased odds of MetS and increasing MetS components. Lepetin and PAI on the other hand were associated with higher odds of MetS and with increasing number of MetS components. Taken together, this data highlights some of the potential role these metabolic markers play in development of MetS and CVD as well as their diagnostic value in this population.
HsCRP, PAI-1 and Leptin Association with Metabolic Risk Factors and MetS
MetS is defined by different health organizations as a cluster of metabolic risk factors that constitute an increased risk for developing T2D and CVD . Central obesity, dislypidemia hypertension and hyperglycemia are key components in the definition of MetS [21, 22]. MetS is also characterized by a chronic low grade inflammation condition [20, 23]. Current definitions of MetS lack a component that measures inflammation status. HsCRP has been suggested as a sensitive marker that is predictive of MetS as well as development of cardiovascular problems [20, 22–24]. Our data showed that hsCRP correlated with all the measured risk factors and was most predictive of the development of MetS in this population. HsCRP strong association with increased number of MetS components highlights its beneficial use as prognostic marker for metabolic disorders in this population as well.
Using a wide range of adipokines our current study showed that women possessed higher circulating levels of leptin, PAI-1 and adiponectin. Even though all of these markers showed a significant difference, it was highest for leptin at almost three folds followed by adiponectin at 1.6 fold difference. This difference in leptin level for example is consistent with levels reported in other studies such as Zuo et al., reported that men had 1.45 ng/ml compared to 8.32 ng/ml in women . This difference can be explained by the difference in sex hormones especially in women as a result of menopause which reduces estrogen production a factor that has been shown to lead to obesity in women . Consistent with data reported in literature for other populations, leptin showed positive association with BMI and HOMA-IR.
PAI-1 and adiponectin were remarkably associated with seven and eight of the studied risk factors respectively. PAI-1 is a serine protease inhibitor that plays an important role in fibrinolysis due to its inhibition of plasminogen activation . Elevated blood levels of PAI-1 have been linked to high reoccurrence events of myocardial infarction ([28–30]), Atherosclerosis  and MetS. The positive association of PAI-1 with all of the risk factors, except for HDL, is in agreement with literature . In accordance to our data, PAI-1 has been shown to positively correlate with atherosclerosis risk factors such as obesity, hyperinsulinemia and hypertriglyceridemia [32, 33]. In general, PAI-1 positively associates with MetS and its components in our study. Other studies have shown that PAI-1 level are reduced in response to interventional therapies that leads to reduction in MetS components such as insulin resistance or weight reduction . Thus, the predictive ability of PAI-1 for CVD diminishes after adjusting for MetS components. This suggests that MetS is a prerequisite to high PAI-1 blood circulating level . It can also explain the lack of association with as many markers in the multivariate analysis in our population.
Adiponectin, a cardio-protective adipokine
Adiponectin on the other hand associated with most CVD risk factors at the univariate and multivariate level. It associated with six of the risk factors showing the strength of this marker in predicting CVD risk. We also showed that decreasing concentration of adiponectin associated with an increasing number of CVD risk factors particularly in women highlighting the gender difference. The importance of adiponectin in CVD is well documented as its plasma level has been shown to be attenuated by cardiac pathologies such as coronary artery disease [35, 36], hypertension  and myocardial infarction . Further, low plasma levels of adiponectin are associated with obesity, T2D and MetS which are main CVD risk factors . Adiponectin is believed to be a cardio-protective cytokine due to its insulin sensitizing, anti-inflammatory and antiatherogenic properties [11, 40]. In Human, a reported single nucleotide polymorphism is associated with a reduction in adiponectin plasma levels leaving individuals with these mutations predisposed to insulin resistance . Iindividuals carrying this mutation were highly susceptible to hypertension and coronary artery disease supporting the protective role this protein plays .
Adipsin, visfatin and resistin association with metabolic markers
Our data sheds light on the association between visfatin and adipsin with CVD risk factors in Arab population. No independent association between BMI and visfatin was observed. Association between obesity and visfatin has been contested with some studies showing positive correlation with obesity  while others failed to show any correlation . One of the factors suggested for this discrepancy was ethnicity . The fact that we show no association between visfatin and BMI emphasizes the importance of ethnic variations and establishing the association between the different biomarkers in different ethnicities.
Adipsin on the other hand is a serine protease that has been shown to be upregulated in obesity. It is involved in triglyceride metabolism through the cleavage of complement factor C3 to C3a that stimulates TG production in adipose tissue . Maresh et al. showed that adipsin was upregulated in a type 1 diabetes model possibly to compensate for the increased fat utilization in insulin-deficient animals . Our data shows a trend of increased adipsin production in subjects with higher BMI and WC that is consistent with the known function of adipsin.
Resistin is another adipokines that was found to be increased in a diet induced obesity mouse model, as well as genetically modified diabetic and obese mouse models . Resistin was also found to be a mediator of insulin resistance in rodents. Unlike rodents, resistin was predominantly expressed in monocytes and macrophages . Even though, resistin was associated with obesity and insulin resistance in rodents, human data is conflicting in this regard. Some studies show association with obesity and insulin resistance while others do not show any association or negative association [47–49]. Our data shows negative association between resistin and HOMA-IR, highlighting the importance of ethnic studies to better understand the function of these metabolic markers.
Ethnicity and metabolic markers
Difference in the level of various metabolic markers between different ethnic groups is well documented and is suggested as a main reason for the discrepancies reported in different studies [14–17, 50–52]. For example, Caucasians have a higher level of adiponectin than other ethnicities such as Asians or African Americans [13, 53]. In a multiethnic study, Morimoto et al. reported significant ethnic differences in the level of leptin, adiponectin, IL-6 and CRP between Caucasians, Japanese Americans, Latinos, African Americans and Native Hawaiians . TNF-α on the other hand was the only biomarker that did not show any significant difference . This ethnic difference in the plasma level of metabolic markers is also reflected on the association between these markers and various risk factors [16, 45]. For example, Sulistyoningrum et al. showed that decrease in adiponectin level was associated with greater increase in insulin resistance as measured by HOMA-IR in Aboriginals, Chinese, and South Asians compared to Europeans .
Available data regarding the level of metabolic markers in the Arab population are scarce. Therefore, this study represents an important milestone in understanding the role of these metabolic markers in Arabs. On the other hand, studies linking various metabolic markers to the genetic background and comparing Arabs to other ethnicities are also scarce. However; Genetic variation in adiponectin and its association with metabolic disorders is one of the well studied markers in Arabs [54–57]. For example, Mtiraoui et al. reported association between adiponectin single nucleotide polymorphisms and T2D in Tunisian Arabs . In another study, Zadjali et al. showed that rs266729 in the adiponectin gene was associated with traits defining obesity in Arab population . As a result, such studies looking at the association between metabolic markers genetic variants and metabolic disorders in Arabs compared to other ethnicities will be crucial to better understand the ethnic variation in this ethnicity.
Study strengths and limitations
The strength of the current study is the fact that it comprehensively investigates the correlation between the CVD risk factors and adipokines in Arab population. It also highlights the importance of hsCRP in predicting MetS and potential use as a prognostic marker for increased risk of T2D and CVD. Nonetheless, we have a few limitations in this study that can be overcome in future studies. The first one is the lack of comparison with other populations especially Caucasians to establish the level of adipokines in Arabs to a well studied population. Establishing such a cohort will be a difficult task to do in Kuwait where percentage of Caucasians is limited. The second limitation regarding the study design, the cross-sectional design used here makes it impossible to determine any causal relationship between circulation levels of adipokines and CVD. It would also be important to study the effect of nutrients in future studies as they have been shown to modulate adipokines as recently shown by Juanola-Falgarona et al. .