In this analysis we demonstrate that higher serum cystatin C and microalbuminuria are independently associated with incident CKD stage 3 in a large representative multi-ethnic cohort. The risks of incident CKD stage 3 associated with increased cystatin C and microalbuminuria were independent of each other and baseline eGFR, and there was no statistically significant interaction between the two variables. Furthermore, there was no overt difference in these relationships among groups based on diabetes or race/ethnicity.
Cystatin C is likely associated with incident CKD stage 3 because it is a sensitive measure of GFR. Interestingly, we noted that cystatin C was a risk factor for incident CKD stage 3 and added information beyond that provided by baseline kidney function. This result is consistent with a sub-analysis from the Cardiovascular Health Study among participants with eGFR > 60 ml/min/1.73m2
, in which cystatin C ≥ 1 mg/L was associated with incident CKD stage 3 independent of baseline creatinine.21
In our study cystatin C was able to distinguish a gradient of risk across quartiles among those with eGFR > 60 ml/min/1.73m2
including those without microalbuminuria. The exact reason why cystatin C adds to eGFR as a risk factor for incident CKD stage 3 is not clear and cannot be evaluated by this study. The two possibilities include that cystatin C is a better estimate of measured GFR in those with eGFR >60 ml/min/1.73m2
,23, 36, 37
the range in which the MDRD Study and CKD EPI equations are known to be less accurate.38, 39
Alternatively, cystatin C reflects other factors independent of measured GFR that are associated with kidney disease progression.40, 41
Several studies have assessed microalbuminuria as risk factor for ESRD; 42–44
however, there are fewer studies that have evaluated microalbuminuria as a risk factor for incident CKD stage 3 in patients without diabetes. There are several reasons why microalbuminuria may be associated with incident CKD stage 3. In patients with diabetes it is well recognized that microalbuminuria may reflect incipient glomerular damage.45
In patients without diabetes, microalbuminuria may also reflect more widespread vascular and endothelial dysfunction,46
which are risk factors for progression of CKD.47–51
Our results are consistent with prior data in patients with Type 152
or Type 2 diabetes53
as well as in those without diabetes.54, 55
In the HOPE (Heart Outcomes Prevention Evaluation) Study, albuminuria below the level of microalbuminuria was associated with development of macroalbuminuria in patients with and without diabetes.56
Similarly, in the PREVEND (Prevention of Renal and Vascular End-stage Disease) Study in individuals with eGFR > 60ml/min/1.73m2
, higher albumin excretion even in the normal range was associated with development of reduced kidney function (ie, <60ml/min/1.73m2
) in the general population.20
The latter is consistent with CVD and all-cause mortality endpoint studies, which have shown a graded increase in risk with incremental increase in ACR, even below the microalbuminuria range.57, 58
In our study, microalbuminuria was associated with incident CKD stage 3, as was ACR on a linear and log linear scale.
There are several potential implications of these results. Cystatin C and microalbuminuria should be evaluated as screening tools to identify individuals at highest risk for development of CKD stage 3. Also, consideration should be given to evaluating interventions that are currently recommended for patients with established CKD, for example, tighter control of blood pressure, use of angiotensin converting enzyme inhibitors or angiotensin receptor blockers, in those without CKD but with high levels of cystatin C and albuminuria.
The strengths of our study include the use of data from a community-based cohort that is large, racially and ethnically diverse, with detailed ascertainment of risk factors, covariates and outcomes. To avoid misclassification of participants due to imprecision of the MDRD Study equation in the range of near normal, we defined incident CKD stage 3 as the presence of both an eGFR < 60 ml/min/1.73m2 at the 3rd or 4th visit and an annual decline in eGFR > 1 ml/min/1.73m2. Furthermore, our results were consistent despite several sensitivity analyses.
There are also several limitations that need to be considered. We examined risk factors for incident CKD stage 3 during only 4.7 years of follow-up; however, we studied a large cohort and noted a 10–12% cumulative incidence of CKD stage 3, so there was adequate power to evaluate the characteristics of interest. The statistical power to detect interactions, particularly between microalbuminuria and diabetes, but also between microalbuminuria and cystatin C, may have been limited. There was only one sample of urine for ACR estimation and ACR is known to vary considerably on a day-to-day basis. Random fluctuations in ACR would have biased our results towards the null and thus we may have underestimated the importance of albuminuria. Our estimate of kidney function is limited, as we do not have actual measurement of GFR. Those with creatinine values in follow up were slightly healthier than those without follow up creatinine values but this most likely would have biased our results to the null. Furthermore, we acknowledge that cystatin C may be influenced somewhat by factors other than GFR such as age, gender, body fat, smoking, and inflammation; despite adjustment for these variables, residual confounding may have remained.40, 41
In summary, serum cystatin C and microalbuminuria were independent risk factors for development of incident CKD stage 3 in this multiethnic cohort. Future studies should evaluate whether these subclinical markers may be useful for screening strategies and whether intervention based on these screening strategies can improve patient outcomes.