Recent Trends in the Prevalence of Chronic Kidney Disease

Not the Same Old Song

Raymond K. Hsu; Neil R. Powe


Curr Opin Nephrol Hypertens. 2017;26(3):187-196. 

In This Article

Update in Chronic Kidney Disease Prevalence in the United States

In a study published in 2016, Murphy et al.[10] examined temporal trends in CKD in the United States using data from NHANES,[4] a health examination survey that uses a probability sampling design to select participants representative of the US civilian population and involves a combination of in-person interviews, physical examinations, and laboratory data. The investigators analyzed data from NHANES III (conducted in 1988–1994) through NHANES 2011–2012 and limited the study to participants aged 20 years or older with available serum creatinine measurements. CKD was defined in this study as estimated glomerular filtration rate (eGFR) less than 60 ml/min/1.73 m2, determined using serum creatinine measurements and the CKD Epidemiology Collaboration (CKD-EPI) creatinine equation.[11] Consistent with older analyses of US CKD prevalence using NHANES,[5,6,12] the investigators excluded those with eGFR less than 15 ml/min/1.73 m2 due to the relatively small sample size and inability to differentiate from ESRD (i.e., whether those individuals were receiving maintenance dialysis). They performed a secondary analysis using an expanded definition of CKD to also include those with eGFR at least 60 ml/min/1.73 m2 and a one-time urine albumin-to-creatinine ratio (ACR) at least 30 mg/g. Diabetes status in this study was defined as self-reported physician diagnosis, use of glucose-lowering medication (oral hypoglycemic medication or insulin), or a laboratory-measured hemoglobin A1c (HbA1c) level of at least 6.5%.

The crude prevalence of CKD (eGFR < 60 ml/min/1.73 m2) increased from 4.8% in 1998–1994 to 6.9% in 2003–2004, but largely stabilized thereafter to 6.9% in 2011–20122 (Table 1).[5–7,12–19] This stability in trend was seen despite a significant increase in prevalence of diabetes (7.4% in 1988–1994 to 11.5% in 2011–2012) and an overall increase in age (mean 44.8 years in 1988–1994 to 47.3 years in 2011–2012). In subgroup analysis, prevalence of CKD was consistently higher in older age groups. For example, among persons aged 65–79 years, CKD prevalence increased from 19.4% in 1988–1994 to 25.1% in 2003–2004, then stabilized/decreased to 21.7% in 2011–2012. In analysis adjusted for age, sex, race/ethnicity, and diabetes status, there was also no significant increase in CKD prevalence after the early 2000s (P for interaction = 0.26). Figure 1 summarizes the temporal trend in crude prevalence of CKD (eGFR < 60 ml/min/1.73 m2) in the United States (along with international comparisons as discussed below).[10,20–22] In their secondary analysis using an expanded definition of CKD to include those with microalbuminuria, the investigators also found little change in crude CKD prevalence after the early 2000s (prevalence of 14.0% in 2003–2004 to 14.2% in 2011–2012, Table 1).

Figure 1.

Temporal trends in the prevalence of chronic kidney disease (estimated glomerular filtration rate <60 ml/min/1.73 m2) using nationally representative surveys. Recent publications using nationally representative surveys from several developed countries demonstrate stable prevalence of chronic kidney disease (estimated glomerular filtration rate <60 ml/min/1.73m2) since the early 2000s. Data from [10,20–22].

Notably, among all subgroups examined, only non-Hispanic blacks had a progressive increase in CKD prevalence throughout the study period, including from the early 2000s through 2011–2012. Crude CKD prevalence (eGFR < 60 ml/min/1.73 m2) among non-Hispanic blacks increased from 3.7% in 1998–1994 to 4.9% in 2003–2004 to 6.2% by 2011–2012.[10] This steady increase in CKD prevalence among non-Hispanic blacks persisted in both adjusted analysis and in secondary analysis using the expanded definition of CKD.

These latest results from Murphy et al.[10] suggest a reversal in the temporal trend in overall CKD prevalence in several previous NHANES analyses, starting with an analysis by the same research group that showed a significant increase in CKD prevalence (eGFR < 60 ml/min/1.73 m2) from 2.0% in 1976–1980 to 2.5% in 1988–1994, a mean change in prevalence by +1.7% per year.[5] Subsequent analyses[6,12] comparing overall CKD prevalence from the NHANES III (1988–1994) period with the early 2000s (1999–2004 period) have shown variable results in the temporal trend, depending on the definition of CKD employed and the method of calibration for creatinine, which was necessary to account for laboratory drift across those earlier NHANES cycles.[23] In one of these studies, Coresh et al.[6] reported a 3.5% annual increase in CKD prevalence (eGFR < 60 ml/min/1.73 m2) from 1988–1994 to 1999–2004 and a 2.6% annual increase in prevalence of eGFR less than 60 ml/min/1.73 m2 or albuminuria in the same study period. Ensuing studies[7,15] using serum cystatin C as an alternative filtration marker also showed discrepant temporal trends in CKD prevalence from the same era, with the latest study by Grams et al.[7] (employing the most updated eGFR cystatin C-estimating equation[17] and technique for cystatin C calibration) reporting a significant ~5.0% annual increase in CKD prevalence (eGFR < 60 ml/min/1.73 m2) from 1988–1994 to 1999–2004. Table 1 also summarizes these older published studies that used NHANES data.

Reasons for the stabilization in CKD prevalence in the United States in the most recent decade are not absolutely clear. Although Murphy et al.[10] did not attempt to directly model any potential mediating factors for this current trend (due to the cross-sectional nature of the NHANES survey design), it has been well reported from other NHANES studies that blood pressure (BP) control among those with hypertension[24,25] and glycemic control among those with diabetes[26] have improved over time. The use of renin–angiotensin system (RAS) inhibitors have also become more pervasive over time,[27,28] rendering a potential renoprotective effect. This latest update is strengthened by its large sample size generalizable to the US population, the use of NHANES recommended methods for serum creatinine calibrations, and the robustness of results in analyses adjusted for demographics and diabetes along with analyses using an expanded definition of CKD including albuminuria. In summary, this latest and most scientific rigorous update[10] in CKD prevalence in the United States paints a favorable picture of overall stabilization of the epidemic of CKD. Despite an aging population and growing prevalence of diabetes and obesity, we should be cautiously optimistic that our collective efforts at preventing and treating CKD and its associated risk factors have likely helped to abate the CKD 'epidemic' overall in the United States within the last 10–15 years.