DISCUSSION
This study is the first to report CKD prevalence using GFR estimated by using CKD-EPI equations together with ACR for albuminuria in the Korean population.
The prevalence of CKD was 7.9% in the KNHANES samples of 2011 and 2012 in Korean adults aged ≥ 19, and the CKD risk prognosis was reported in Table 2.
The CKD risk prognosis from moderately increased to very high reported in this study was lower than that in the USA [15]. GFR and albuminuria are independent and complementary predictors of important clinical outcomes, including CKD progression, end-stage renal disease, acute kidney injury, cardiovascular mortality, and all-cause mortality [16-19].
The frequencies of GFR < 60 mL/min/1.73 m2 showed sharp increases starting from around age 50 for men and 60 for women, whereas those of ACR ≥ 30 mg/g increased relatively gradually, starting from earlier ages, as shown in Table 4 and 5.
The rate of decline of measured GFR was reported to double in kidney transplantation donors aged > 45 years [20].
The prevalence of estimated GFR < 60 mL/min/1.73 m2 using CKD-EPI equations and the MDRD Study equation was 2.2% and 2.0% in this study, but 2.6% and 3.2%, respectively in the previous study [7].
The statistical method used in the current study was complex sample analysis incorporating sampling weight, in contrast to the previous study which did not consider weight in the analysis.
The frequencies of GFR stage G1 and G2 by CKD-EPI equations were 67.7% and 29.8%, whereas those by the MDRD Study equation were 59.4% and 38.7%, respectively, in this study.
These numbers are comparable to the previous study, considering the different analytical methods; the frequencies of GFR stage G1 and G2 by CKD-EPI equations were 64.5% and 28.9%, whereas those by the MDRD Study equation were 47.6% and 49.2%, respectively [7].
These results confirm the tendency to underestimate GFR by the MDRD Study equation when GFR is 60 mL/min/1.73 m2 or greater.
The previous report of lower prevalence of GFR stage G3 using CKD-EPI equations than with the MDRD Study equation was not observed in this study [5].
Among the population of the previous study, the frequencies of G3 by CKD-EPI equations and the MDRD Study equation (33.2% and 37.2%, respectively) were far greater than those among the Korean population (2.1% and 1.9%, respectively).
This difference in the study population might explain the current result.
The prevalence of ACR ≥ 30 mg/g in the entire study population, and ACR ≥ 30 mg/g in individuals with GFR ≥ 60 mL/min/1.73 m2 was 6.5% and 5.7%, respectively, in this study. The prevalence of ACR ≥ 30 mg/g in the entire study population and ACR ≥ 30 mg/g in individuals with GFR ≥ 60 mL/min/1.73 m2 aged ≥ 35 years was 10.2% and 8.7%, respectively, in the previous Korean population [11].
Besides the different age criteria, the populations of the previous and current study are quite different; the previous study was designed by the researcher group considering age, gender, and city factors not us-ing the KNHANES sample data.
Although the absolute values are different, the important point in common is that a large number of people have albuminuria before GFR decreases to < 60 mL/min/1.73 m2 (85.3% and 87.7% of the albuminuria cases did not have decreased GFR in the previous and current study, respectively).
As shown in Table 3, test strips did not detect albuminuria in 81.5% of cases.
At present, laboratory tests for albuminuria are not routinely performed during regular physical examinations in Korea. This might result in the failure of early detection of kidney disorders.