Prevalence and Cardiovascular Risk Profile of CKD
Prevalence and Cardiovascular Risk Profile of CKD
Early European studies have shown a variable CKD prevalence, with significant differences in men and women in most cases. Observed variability may be explained at least in part by heterogeneity of study design, such as differences in population sampling, including age and extent of geographic area, equation used to estimate eGFR and examined stages. The peculiarities of the main CKD surveys are summarized in Table 5. CARHES is the first study estimating CKD prevalence in an European country at the national level and based on a large sample size (the number of screened subjects/resident population was greater when compared with the other four national surveys outside Europe) and adequate measures (CKD stages were defined by ACR and CKD-EPI eGFR).
Unadjusted prevalence of CKD averaged 7% in men and women, with predominance of early stages (Table 2). Similar figures were observed after age and gender adjustment; indeed, standardized rates were 6.3% for all stages of CKD, 3.8% for CKD Stage G1–2 and 2.5% for Stage G3–5. Accordingly, we can estimate 2.2 million adult persons with CKD in Italy, 60% with Stages I and II. Noteworthy, among those with CKD, only 1 of 10 was aware of the pathology, and the perception of disease was similarly scarce also in those with low eGFR (18%), thus confirming that low awareness of CKD is a global problem.
CARHES findings differ from the results of the other four main surveys in nationally representative samples of the general population (Table 5). In the National Health and Nutrition Examination Survey (NHANES) 1999–2004, CKD prevalence was 13.1% overall, and the dimension of overt phases (Stages G3–5) was remarkably higher versus Italy. In Canada, comparable estimates with those in the USA were reported; however, at variance with NHANES, there were higher rates for Stages G1–2 (9.4%) than for Stages G3–5 (3.1%) were found. The survey in China yielded comparable results with US and Canadian surveys in terms of overall CKD prevalence; however, prevalence of Stage G1–2 was higher than that reported in NHANES 1999–2004 and the overt stages of disease were poorly represented. In Australia, CKD prevalence has been recently recalculated according to the CKD-EPI equation in the original cohort representative of the adult population examined in 1999–2000; also in this country, the overall prevalence is high; however, at variance with the other national surveys, the prevalence rates of early and advanced stages are remarkably similar. An additional peculiarity of CKD in Italy is the lower prevalence of albuminuria with decreasing eGFR. Indeed, a reverse association has been shown in the USA, Canada and China, whereas a similar prevalence by CKD stages was found in Australia. The Gubbio study specifically addressed the relationship between albuminuria and GFR in a sample of Italian adult subjects of the general population of a town in central Italy. The authors did not find any correlation, with high albuminuria and low eGFR that provided complementary information in defining kidney dysfunction and predicting CV risk. In this regard, it is interesting that a recent large population-based study found a strong association between hyperhomocysteinaemia and higher prevalence of albuminuria that was independent of eGFR level. Therefore, the association between renal function and albuminuria may not be as simple as expected and deserves further ad hoc studies.
Conversely, in agreement with other surveys, we found that males were characterized by higher prevalence of early, albuminuric CKD stages versus females, but the difference was not maintained for more advanced disease. It is possible to hypothesize that also in Italy men are more prone to develop proteinuric CKD due to the higher rates of hypertension, diabetes, smoking and overweight and, due in part to these differences, being exposed to faster progression to ESRD and premature death with respect to women.
The different findings reported by national surveys of the general population, including albuminuria distribution by eGFR, indicate that epidemiology of non-dialysis CKD is mainly influenced by features inherent to the population examined; this hypothesis is supported by the different rates of ESRD reported by national registries of dialysis and transplantation. Of note, the comparison of CKD prevalence between our study and the surveys in USA and China may be affected by the different methods of eGFR estimation to define CKD; indeed, NHANES 1999–2004 used the MDRD equation and the survey in China used a modified MDRD equation. In this regard, in NHANES 1999–2006, prevalence of CKD was slightly lower (11.5%) when based on the CKD-EPI versus MDRD equation (13.1%). This difference has been consistently found in other surveys that reported similar overestimation of prevalence rate when GFR was calculated by means of MDRD versus CKD-EPI. Nonetheless, the high prevalence of CKD in USA (14.0%) has been confirmed by the most recent report of NHANES (2005–10) publically available (http://www.usrds.org/2013/pdf/v1_ch1_13.pdf), where the CKD-EPI equation was used.
Knowledge on the early, albuminuric, stages of disease is critical. Pathological albuminuria acts as an independent predictor of de novo development of renal function impairment, ESRD and mortality in the general population, and its remission heralds a better cardio-renal prognosis. CARHES shows a lower prevalence of CKD Stages I–II in Italy when compared with the other main surveys in the rest of the world. Specifically, prevalence was 4.2%, and 3.8% after standardization to the resident population, that corresponds to a concomitance of albuminuria and eGFR ≥60 in ~1.3 million of adult persons in Italy. In this regard, our data suggest that urine testing would be especially indicated in elderly, smokers, diabetics, hypertensives and those with a history of CV disease, being that these factors are significantly associated with the risk of pathological albuminuria (Table 4). Noteworthy, we found albuminuria to be more prevalent in overt CKD than among individuals with hypertension or diabetes, which are conditions where albuminuria testing is traditionally common. The knowledge that coexistence of albuminuria and low eGFR confers a substantial increase in the risk for ESRD and all-cause mortality should therefore stimulate physicians to routinely perform urine testing also in subjects with impaired eGFR.
The reason why the prevalence of CKD is lower in Italy is puzzling, and this holds particularly true when taking into account age and CV risk profile. In the whole population, the mean age was ~10 years higher with respect to surveys in North Europe (Table 5), USA, Canada, China and Australia. Older age is expected to be a factor that predisposes to CKD development because aging is itself associated with reduced GFR and because older age associates with increased prevalence of obesity, hypertension and diabetes that are all well-known determinants of CKD. Noteworthy, in the whole sample, >50% of subjects had hypertension, one of the four was obese, 20% were smokers and over 10% subjects were affected by diabetes. This picture is not dissimilar when compared with data from the other countries of the Western world. In particular, in the USA, which is the country with the highest prevalence of CKD, hypertension is less frequent (33%) but obesity is more prevalent (35%), while similarities are observed for diabetes, smoking and lipid profile as well. Interestingly, as observed for CKD, also the prevalence of persons with a positive history of CV, disease is significantly lower in Italy (Table 1) than in the USA. Therefore, it is possible that features inherent to the population living in Italy may be protective not only for CV events, but also for CKD development, even in the presence of unfavourable risk profile. Whether this is a 'renal' aspect of the genetic low background risk, and/or dependent on the still high adherence to the Mediterranean diet, deserves further studies.
CARHES has three limitations that are shared with the other main surveys on CKD prevalence. First, it is limited by the single measurement of serum creatinine and ACR, while correct identification of CKD requires confirmation of abnormalities in eGFR and/or albuminuria over at least a 3-month period. Secondly, the 53% response rate of the first ones on the list may introduce a bias because these subjects may be healthier or sicker than the rest of population. Thirdly, the dimension of CV disease may not be accurately quantified being mostly based on questionnaires. Finally, as a further potential confounder, we observed a different rate of response to survey, higher in the north versus the rest of Italy, which is compatible with the higher educational and economic level of this region.
In conclusion, in Italy, when compared with other countries, CKD prevalence is relatively low, being ~7.0% in men and women, with predominance of the early stages (59%). The prevalence of CKD appears to be unexpectedly lower when considering the older age and the unfavourable CV risk profile of the whole population. Low background risk (genetic factors) and/or dietary habits (Mediterranean diet) may play a protective role.
The consequences of CKD, in terms of life years lost and ESRD incidence, vary significantly worldwide and even within Europe. Comparison of CARHES data with those obtained in the other national surveys outside Europe suggests that CKD may be considered as a 'geographic pathology' also in terms of disease prevalence. Country-level studies on epidemiology of CKD are therefore needed to attain proper estimates of the burden of this high-risk condition.
Discussion
Early European studies have shown a variable CKD prevalence, with significant differences in men and women in most cases. Observed variability may be explained at least in part by heterogeneity of study design, such as differences in population sampling, including age and extent of geographic area, equation used to estimate eGFR and examined stages. The peculiarities of the main CKD surveys are summarized in Table 5. CARHES is the first study estimating CKD prevalence in an European country at the national level and based on a large sample size (the number of screened subjects/resident population was greater when compared with the other four national surveys outside Europe) and adequate measures (CKD stages were defined by ACR and CKD-EPI eGFR).
Unadjusted prevalence of CKD averaged 7% in men and women, with predominance of early stages (Table 2). Similar figures were observed after age and gender adjustment; indeed, standardized rates were 6.3% for all stages of CKD, 3.8% for CKD Stage G1–2 and 2.5% for Stage G3–5. Accordingly, we can estimate 2.2 million adult persons with CKD in Italy, 60% with Stages I and II. Noteworthy, among those with CKD, only 1 of 10 was aware of the pathology, and the perception of disease was similarly scarce also in those with low eGFR (18%), thus confirming that low awareness of CKD is a global problem.
CARHES findings differ from the results of the other four main surveys in nationally representative samples of the general population (Table 5). In the National Health and Nutrition Examination Survey (NHANES) 1999–2004, CKD prevalence was 13.1% overall, and the dimension of overt phases (Stages G3–5) was remarkably higher versus Italy. In Canada, comparable estimates with those in the USA were reported; however, at variance with NHANES, there were higher rates for Stages G1–2 (9.4%) than for Stages G3–5 (3.1%) were found. The survey in China yielded comparable results with US and Canadian surveys in terms of overall CKD prevalence; however, prevalence of Stage G1–2 was higher than that reported in NHANES 1999–2004 and the overt stages of disease were poorly represented. In Australia, CKD prevalence has been recently recalculated according to the CKD-EPI equation in the original cohort representative of the adult population examined in 1999–2000; also in this country, the overall prevalence is high; however, at variance with the other national surveys, the prevalence rates of early and advanced stages are remarkably similar. An additional peculiarity of CKD in Italy is the lower prevalence of albuminuria with decreasing eGFR. Indeed, a reverse association has been shown in the USA, Canada and China, whereas a similar prevalence by CKD stages was found in Australia. The Gubbio study specifically addressed the relationship between albuminuria and GFR in a sample of Italian adult subjects of the general population of a town in central Italy. The authors did not find any correlation, with high albuminuria and low eGFR that provided complementary information in defining kidney dysfunction and predicting CV risk. In this regard, it is interesting that a recent large population-based study found a strong association between hyperhomocysteinaemia and higher prevalence of albuminuria that was independent of eGFR level. Therefore, the association between renal function and albuminuria may not be as simple as expected and deserves further ad hoc studies.
Conversely, in agreement with other surveys, we found that males were characterized by higher prevalence of early, albuminuric CKD stages versus females, but the difference was not maintained for more advanced disease. It is possible to hypothesize that also in Italy men are more prone to develop proteinuric CKD due to the higher rates of hypertension, diabetes, smoking and overweight and, due in part to these differences, being exposed to faster progression to ESRD and premature death with respect to women.
The different findings reported by national surveys of the general population, including albuminuria distribution by eGFR, indicate that epidemiology of non-dialysis CKD is mainly influenced by features inherent to the population examined; this hypothesis is supported by the different rates of ESRD reported by national registries of dialysis and transplantation. Of note, the comparison of CKD prevalence between our study and the surveys in USA and China may be affected by the different methods of eGFR estimation to define CKD; indeed, NHANES 1999–2004 used the MDRD equation and the survey in China used a modified MDRD equation. In this regard, in NHANES 1999–2006, prevalence of CKD was slightly lower (11.5%) when based on the CKD-EPI versus MDRD equation (13.1%). This difference has been consistently found in other surveys that reported similar overestimation of prevalence rate when GFR was calculated by means of MDRD versus CKD-EPI. Nonetheless, the high prevalence of CKD in USA (14.0%) has been confirmed by the most recent report of NHANES (2005–10) publically available (http://www.usrds.org/2013/pdf/v1_ch1_13.pdf), where the CKD-EPI equation was used.
Knowledge on the early, albuminuric, stages of disease is critical. Pathological albuminuria acts as an independent predictor of de novo development of renal function impairment, ESRD and mortality in the general population, and its remission heralds a better cardio-renal prognosis. CARHES shows a lower prevalence of CKD Stages I–II in Italy when compared with the other main surveys in the rest of the world. Specifically, prevalence was 4.2%, and 3.8% after standardization to the resident population, that corresponds to a concomitance of albuminuria and eGFR ≥60 in ~1.3 million of adult persons in Italy. In this regard, our data suggest that urine testing would be especially indicated in elderly, smokers, diabetics, hypertensives and those with a history of CV disease, being that these factors are significantly associated with the risk of pathological albuminuria (Table 4). Noteworthy, we found albuminuria to be more prevalent in overt CKD than among individuals with hypertension or diabetes, which are conditions where albuminuria testing is traditionally common. The knowledge that coexistence of albuminuria and low eGFR confers a substantial increase in the risk for ESRD and all-cause mortality should therefore stimulate physicians to routinely perform urine testing also in subjects with impaired eGFR.
The reason why the prevalence of CKD is lower in Italy is puzzling, and this holds particularly true when taking into account age and CV risk profile. In the whole population, the mean age was ~10 years higher with respect to surveys in North Europe (Table 5), USA, Canada, China and Australia. Older age is expected to be a factor that predisposes to CKD development because aging is itself associated with reduced GFR and because older age associates with increased prevalence of obesity, hypertension and diabetes that are all well-known determinants of CKD. Noteworthy, in the whole sample, >50% of subjects had hypertension, one of the four was obese, 20% were smokers and over 10% subjects were affected by diabetes. This picture is not dissimilar when compared with data from the other countries of the Western world. In particular, in the USA, which is the country with the highest prevalence of CKD, hypertension is less frequent (33%) but obesity is more prevalent (35%), while similarities are observed for diabetes, smoking and lipid profile as well. Interestingly, as observed for CKD, also the prevalence of persons with a positive history of CV, disease is significantly lower in Italy (Table 1) than in the USA. Therefore, it is possible that features inherent to the population living in Italy may be protective not only for CV events, but also for CKD development, even in the presence of unfavourable risk profile. Whether this is a 'renal' aspect of the genetic low background risk, and/or dependent on the still high adherence to the Mediterranean diet, deserves further studies.
CARHES has three limitations that are shared with the other main surveys on CKD prevalence. First, it is limited by the single measurement of serum creatinine and ACR, while correct identification of CKD requires confirmation of abnormalities in eGFR and/or albuminuria over at least a 3-month period. Secondly, the 53% response rate of the first ones on the list may introduce a bias because these subjects may be healthier or sicker than the rest of population. Thirdly, the dimension of CV disease may not be accurately quantified being mostly based on questionnaires. Finally, as a further potential confounder, we observed a different rate of response to survey, higher in the north versus the rest of Italy, which is compatible with the higher educational and economic level of this region.
In conclusion, in Italy, when compared with other countries, CKD prevalence is relatively low, being ~7.0% in men and women, with predominance of the early stages (59%). The prevalence of CKD appears to be unexpectedly lower when considering the older age and the unfavourable CV risk profile of the whole population. Low background risk (genetic factors) and/or dietary habits (Mediterranean diet) may play a protective role.
The consequences of CKD, in terms of life years lost and ESRD incidence, vary significantly worldwide and even within Europe. Comparison of CARHES data with those obtained in the other national surveys outside Europe suggests that CKD may be considered as a 'geographic pathology' also in terms of disease prevalence. Country-level studies on epidemiology of CKD are therefore needed to attain proper estimates of the burden of this high-risk condition.
Source...