Our data present four important findings derived from a large diverse population of participants with CKD. The first finding is the high prevalence of metabolic syndrome in CKD. Nearly 2 out of 3 participants (65%) in the CRIC Study satisfy the ATP-III diagnostic criteria for metabolic syndrome, with a noteworthy proportion (44.3%) of participants without diabetes fulfilling metabolic syndrome criteria. Such a high proportion of metabolic syndrome in diabetics (87% in the current study) was also noted by Alexander et al. [15
] in the NHANES-III population. Reported prevalence of metabolic syndrome in the USA has been increasing, in step with the growing problem of obesity. In recent studies, the metabolic syndrome prevalence in adults in the USA increases with age and ranges from 22 to 35% [10
]. Even in the absence of diabetes, the prevalence of metabolic syndrome in our CKD population is nearly twice as high as in other recently reported cohorts. The recently reported association between metabolic syndrome and decline in kidney function with more components of the syndrome showing a greater decline underscores the clinical importance of this finding. In the Atherosclerosis Risk in Communities (ARIC) Study, those with all 5 metabolic syndrome components had a nearly 2.5-fold higher odds of developing CKD compared with no components during a 9-year follow-up [18
]. The mechanism by which metabolic syndrome participates in accelerating the loss of kidney function may be through hyperfiltration, supported by studies such as that of Chagnac et al. [19
], which demonstrated a greater degree of hyperfiltration in subjects with severe obesity compared with similar aged controls. In addition, elevated insulin concentrations, a consequence of the insulin resistance thought to underlie the metabolic syndrome, is another potential mediator by increased renal blood flow through its vasodilatory effects on the kidney circulation [20
Secondly, there is substantial variability in the prevalence of the individual components of the metabolic syndrome with elevated BP present more than twice as often as elevated triglyceride levels. Non-CKD populations like NHANES [15
] show a higher prevalence of elevated triglyceride concentration, but similar prevalence in waist circumference, HDL and BP criterion.
Third, metabolic syndrome was associated with prevalent CVD at enrollment into CRIC overall, but it improves model fit as measured by BIC only among those without diabetes when stratified for this. This latter point is particularly noteworthy and suggests that perhaps the designation of metabolic syndrome is less important than understanding the individual risk factors present among CKD patients with diabetes.
Lastly our data show that simply counting up the number of metabolic syndrome components (i.e. 1–5 components being present in any individual) showed the best model characteristics as evaluated using the BIC compared with the presence of metabolic syndrome or the individual components of it. This suggests that dichotomization into a binary diagnosis of metabolic syndrome unnecessarily reduces information for predicting CVD. While the use of the score or its individual components may somewhat improve prediction, this comes at a price of increased model complexity and difficulty of interpretation and use which may not be warranted in many clinical settings.
The metabolic syndrome was described to characterize cardiovascular risk capitalizing on the well-known finding that CV risk factors tend to ‘cluster’ in individuals with upper body obesity [21
]. In recent years, the association of metabolic syndrome with CVD has been called into question by arguments that cast doubt on its value in CVD risk recognition since the components alone appear to be equally predictive [22
]. As depicted in table , the only individual component with a stronger association with prevalent CVD among participants without diabetes as assessed by the adjusted PR was hypertension, which was present in 87.6% of our participants. The presence of hypertension doubles the likelihood of having metabolic syndrome when compared to the general non-hypertensive population and metabolic syndrome in hypertensives generally heightens the risk of CVD [23
]. Our data showed that hypertension appeared to be ‘protective’ from CVD in the diabetics with metabolic syndrome. Although this at first seems counterintuitive, it may be that those with diabetes and increased BP were treated earlier and more aggressively compared to those diabetics without a diagnosis of hypertension with drugs blocking the renin-angiotensin system. Sixty-seven percent of the diabetics without metabolic syndrome reported taking an ACE inhibitor or an angiotensin receptor blocker, while 80% of the diabetics with metabolic syndrome were treated with these drugs in CRIC. Each of these drug classes has an agent with an indication for high cardiovascular risk (‘Expanded indication for Telmisartan FDA’ http://www.theheart.org/article/1014115.do
and ‘Ramipril FDA approval for high CV risk’ http://www.pslgroup.com/dg/1e36a6.htm
both accessed January 11, 2011). In addition, the dialysis experience also somewhat paradoxically suggests that lower BPs are associated with a greater risk of cardiovascular death [24
Our findings regarding metabolic syndrome in CKD add to the literature, given the large size of our CKD population, about half of whom were diabetic and about half of whom were African-American. We acknowledge several limitations including the cross-sectional nature of the analyses and the self-report of prevalent CVD.
In summary, we observed a high prevalence of metabolic syndrome in a population of subjects all of whom have CKD compared to that reported in the general population without diabetes. Hypertension was the component most often present in our CKD population fulfilling the criteria for metabolic syndrome. The presence of metabolic syndrome was associated with a significant PR for CVD at enrollment into the CRIC study, but only among those without diabetes. The longitudinal nature of the CRIC study provides an ideal setting to determine prospectively the predictive value of metabolic syndrome, compared to individual components, both on kidney function decline and on worsening of existing CVD and incident cardiovascular outcomes in an already high-risk population, those with established CKD.