In this narrative review, we studied the association of risk factors for chronic kidney disease (CKD) and CKD prevalence at an ecological level and describe potential reasons for international differences in estimated CKD prevalence across European countries. We found substantial variation in risk factors for CKD such as in the prevalence of diabetes mellitus, obesity, raised blood pressure, physical inactivity, current smoking and salt intake per day. In general, the countries with a higher CKD prevalence also had a higher average score on CKD risk factors, and vice versa. There was no association between cardiovascular mortality rates and CKD prevalence. In countries with a high CKD prevalence, the prevention of noncommunicable diseases may be considered important, and, therefore, all five national response systems (e.g. an operational national policy, strategy or action plan to reduce physical inactivity and/or promote physical activity) have been implemented. Furthermore, both the heterogeneity in study methods to assess CKD prevalence as well as the international differences in the implementation of lifestyle measures will contribute to the observed variation in CKD prevalence. A robust public health approach to reduce risk factors in order to prevent CKD and reduce CKD progression risk is needed and will have co-benefits for other noncommunicable diseases.
Individuals with chronic kidney disease (CKD) are at increased risk of all-cause and cardiovascular premature death, and may progress to end-stage renal disease (ESRD) [1–3]. The number of patients starting renal replacement therapy (RRT) due to ESRD varies from < 30 to > 200 people per million population across European countries, and is much higher in most high-income countries compared with countries where access to health care is limited . RRT has a significant impact on both individual patient quality of life [5, 6] and health system budgets [7, 8]. Therefore, where possible, public health initiatives are needed to prevent or delay the development of CKD (primary prevention) and its complications (secondary prevention), including ESRD.
The prevalence of risk factors for CKD, such as lifestyle factors, the difference in general population age distributions as well as differing cardiovascular mortality rates may partly explain international variation in CKD prevalence [9, 10]. Moreover, health policies addressing CKD prevention and treatment also vary widely across the region . A better understanding of this complexity would guide the medical community and policy makers in targeting primary and secondary prevention efforts.
In this narrative review, we will elaborate on the factors potentially underlying observed international differences in CKD prevalence in the elderly within Europe. We will do this by providing an overview of the prevalence of several CKD risk factors and cardiovascular mortality rates by country, and thereafter explore this in relation to the CKD prevalence, focusing on countries with the highest and lowest CKD prevalence. In addition, we will discuss the heterogeneity of general population studies in their assessment of CKD, which likely further contributes to the observed variation.
CKD PREVALENCE IN EUROPE
Figure 1 shows the age- and sex-adjusted prevalence of CKD stages 3–5 in the population aged 65–74 years in European countries and regions based on a study by Brück et al. on behalf of the European CKD Burden Consortium . Figure 1 is adapted to include the CKD prevalence as reported by a recent published study based on the population-representative health survey for England . The CKD prevalence varied from 4.1% [95% confidence interval (95% CI) 2.9–5.3] in the Swiss Bus Santé study to 25.5% (95% CI 22.3–28.8) in the Northeast German SHIP study .
The comparability of CKD prevalence across countries and regions was increased by using the same definition of CKD, i.e. CKD stages 3–5 was defined as the presence of estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m2 as calculated by the CKD epidemiology collaboration (CKD-EPI) equation, from a single serum creatinine measurement. Furthermore, the influence of different age and sex distributions was reduced through age and sex standardization to the European Union 27 population (with exception of the CKD prevalence reported for England). Moreover, as IDMS calibration standardization reduces the bias and improves the interlaboratory comparability , Figure 1 distinguishes between studies with isotope dilution mass spectrometry (IDMS) and non-IDMS results.
FACTORS POTENTIALLY UNDERLYING THE INTERNATIONAL DIFFERENCES IN CKD PREVALENCE
The prevalence of risk factors for CKD, the cardiovascular mortality in the general population as well as public health policies may affect CKD prevalence. Figure 2 shows the prevalence of risk factors, i.e. diabetes mellitus , obesity , raised blood pressure , physical inactivity , current smoking  as well as salt intake . Table 1 provides an overview of the estimated CKD stages 3–5 prevalence, together with an average score on the above-mentioned risk factors for CKD in that particular country (see Table 1 for more details). In Table 1, we did not include countries with multiple estimations of CKD prevalence (France, Germany, Italy and the Netherlands, see Figure 1). Moreover, England was not reported in Table 1 as an exact estimation of CKD stages 3–5 prevalence and a 95% CI in age group 65–74 years was unavailable.3 presents the number of public health policies targeting noncommunicable diseases . We used the most recent data available from various sources such as the World Health Organization (WHO) and World Bank (see figure legends for more details). Whenever possible, the estimates presented were age-standardized. The categories presented in each figure are based on quartiles. However, countries with a very similar prevalence of a risk factor were assigned to the same category.
|Countries||CKD prevalence (95% CI)a||Average score on risk factors for CKDb|
|Countries||CKD prevalence (95% CI)a||Average score on risk factors for CKDb|
The average score on risk factors for CKD was based on the prevalence of diabetes mellitus, obesity, raised blood pressure, physical inactivity, current smoking and salt intake (see Figure 2). The 13 countries with a known prevalence of CKD as shown in Figure 1 were ranked based on their prevalence of each risk factor (13 points for the country with the highest prevalence of that risk factor). Thereafter, the scores were averaged by country and categorized in quartiles (red, orange, green and light green, with red indicating the highest average risk score and light green the lowest average risk score). In Table 1, we did not include countries with multiple estimations of CKD prevalence (France, Germany, Italy and the Netherlands, see Figure 1). Moreover, England was not reported in Table 1 as an exact estimation of CKD stages 3–5 prevalence and a 95% CI in age group 65–74 years was unavailable.
Throughout this article, we will focus on those countries that have an estimation of CKD stages 3–5 prevalence in the age group 65–74 years as presented in Figure 1.
The prevalence of diabetes mellitus varied from 2.5% in the Republic of Moldova to 14.8% in Turkey (Figure 2a). Portugal had a relatively high prevalence of diabetes mellitus (9.6%), along with a high prevalence of CKD (Figure 1).
Body mass index
The percentage of obesity (body mass index ≥30 kg/m2) in the general population varied from 14.9% in Republic of Moldova to 29.5% in Turkey (Figure 2b). The UK ranked second with a prevalence of 28.1%. The low prevalence of diabetes mellitus (Figure 2a) in that country is, therefore, remarkable. Switzerland, on the other hand, had a relatively low prevalence of obesity (19.4%), together with a relatively low prevalence of CKD (Figure 1).
Raised blood pressure
The percentage of age-standardized raised blood pressure varied from 15.2% in the UK to 31.7% in Estonia (Figure 2c). Although with some exceptions, the percentage of age-standardized raised blood pressure was lower in countries in the Western part and higher in countries in the Eastern part of Europe.
The percentage of age-standardized physical inactivity varied from 9.5% in Russia to 38.7% in Serbia (Figure 2d). After Serbia, the UK had the highest percentage of age-standardized physical inactivity (37.3%), followed by Ireland (35.1%) and Portugal (34.9%).
The percentage of age-standardized current smokers of any tobacco product varied substantially from 17.5% in Iceland to 43.3% in Greece (Figure 2e). Behind Iceland and Denmark, the UK had the lowest percentage of age-standardized current smokers (20.3%).
The mean daily salt intake in grams in the adult population was among the highest in Portugal (≥12 g/day) and lowest in Germany (6–7 g/day) (Figure 2f).
Average score on risk factors for CKD
Table 1 provides an overview of estimated CKD stages 3–5 prevalence in the population aged 65–74 years of age, together with the average score on the risk factors for CKD (i.e. diabetes mellitus, obesity, raised blood pressure, physical inactivity, current smoking and salt intake) in that particular country. In general, countries with a higher CKD prevalence had a higher average score on the risk factors for CKD and vice versa. For example, Portugal had the highest estimate of CKD prevalence, and also the highest average score on CKD risk factors. On the other hand, Switzerland had the lowest estimate of CKD prevalence and the lowest average score on CKD risk factors.
Cardiovascular mortality in the general population
The age-standardized cardiovascular mortality per 100 000 deaths is lower in high-income European countries and higher in low-income European countries (http://www.who.int/healthinfo/mortality_data; data not shown). There was no association between cardiovascular mortality rates and CKD stages 3–5 prevalence across these countries.
Public health policy
The WHO Country Profiles 2014 lists national response systems that have been implemented to prevent and control noncommunicable diseases . For the purpose of this review we used data on the implementation of the following national response systems per country, comprising an operational (multisectorial) national policy, strategy or action plan: (i) that integrates several noncommunicable diseases and shared risk factors; (ii) to reduce the harmful use of alcohol; (iii) to reduce physical inactivity and/or promote physical activity; (iv) to reduce the burden of tobacco use; or (v) to reduce unhealthy diet and/or promote healthy diets. Each country received a score, ranging from 0 to 5, based on the number of incorporated national response systems. A score of 0 indicates that none of these five national response systems has been implemented in that particular country, and a score of 5 that all five national response systems have been implemented. The results are shown in Figure 3. In Portugal, a country with a high prevalence of CKD stages 3–5, the prevention of noncommunicable diseases may be considered important and therefore all five national response systems may have been implemented.
HETEROGENEITY OF STUDIES ON CKD PREVALENCE
Differences in CKD prevalence may be partly explained by heterogeneity in study populations. All studies assessing CKD prevalence presented in Figure 1 were designed to be representative of the respective regional or national general populations. Nevertheless, their sample selections varied considerably. Sample selection methods may have an effect on both the coverage of the population and the response rate. As a consequence, they may influence the representativeness of the sample . Nonresponse analyses among several studies assessing CKD prevalence suggested that participants were equally healthy or healthier in comparison with nonparticipants . Possibly, underrecruitment of patients with more advanced CKD provided a lower estimated CKD prevalence. Furthermore, as CKD prevalence may change over time [19, 20], some of the international differences in CKD prevalence may be explained by differences in the time period the study was performed.
Assessment of CKD prevalence
Differences in CKD prevalence may be due to the heterogeneity of methodology used in the included studies with regard to measurements of creatinine including not only differences in assays, but also in handling and storage conditions [21, 22]. Moreover, regardless of the use of IDMS standardization, we cannot rule out that some interlaboratory variability still exists even within the IDMS studies .
IMPLEMENTATION OF LIFESTYLE MEASURES
Previous studies have shown that by implementing lifestyle measures, the prevalence of risk factors for CKD can be reduced at the general population level. For example, in several countries physical inactivity [24, 25], smoking  and salt intake  were all reduced after implementation of nationwide measures. However, great variation exists among European countries with regard to this implementation and this will likely have had an impact on the observed prevalence of risk factors of CKD.
First, the effect of lifestyle measures may vary across countries. For instance, the effect of raising taxes to reduce smoking differs across countries as smokers in high-income countries seem to be less responsive to price changes . Nevertheless, the percentage of current smokers was relatively low in most high-income countries in comparison with low-income countries. Moreover, besides taxation measures, other effective tobacco use-reduction strategies may vary across countries, such as comprehensive bans on advertising and promotion of tobacco products, strong restrictions on smoking in work places and public spaces, education and counteradvertising campaigns and improved product warning labels . It is possible that strategies addressing combinations of lifestyle measures may have a stronger effect than those addressing each lifestyle factor individually.
Furthermore, the time period in which lifestyle measures were implemented may differ. For example, Finland began working on salt reduction as early as the 1970s, including public awareness campaigns . In 1993, mandatory salt labelling was introduced, and products containing particularly high levels of salt were also required to bear warning labels. This was accompanied by the introduction of a ‘better choice’ logo, supported by the Finnish Heart Association, which identified low-salt options. Daily salt intake in Finland dropped from ∼12 g/day in the late 1970s to as low as 6.8 g/day among women already in 2002 . In many European countries, however, national initiatives to reduce salt intake were not initiated before the new millennium . For instance, in Portugal, the first national initiative to reduce salt was implemented in 2009 by setting a limit on the maximum amount of salt in bread (≤1.4 g salt/100 g). Additionally, in 2012, Portugal levied taxes on salt products . Obviously, in Portugal, the effect of these implemented measures on salt intake may be seen only after 2009, which is much later than in Finland.
Also, one should take into account that the prevalence of risk factors of CKD already varied across countries before the implementation of lifestyle measures. For example, in the UK, the percentage of individuals fulfilling the criteria for sufficient physical activity was relatively low by the end of the previous century (i.e. 32% in men and 21% in women in 1997) , whereas in the Netherlands this percentage was relatively high in the beginning of this century (52% in 2001) . Regardless of the positive effect of the implementation of physical activity measures in the UK, the percentage of people fulfilling the criteria for physical activity was still low in this country in 2012 (43% in men and 32% in women) , whereas in the Netherlands the percentage of people who were physically active had further improved (62% in 2011) .
Noticeably, lowering one risk factor for CKD may also influence other risk factors for CKD or health outcomes. For instance, results of previously performed studies suggest that a decrease in physical inactivity is associated with a decrease in raised blood pressure and cardiovascular disease. Only few studies have described the impact of lifestyle improvements on kidney outcomes. Some evidence exists on the positive effect of salt reduction on both hypertension and proteinuria in people with CKD  and of smoking cessation on kidney function . However, within these studies the outcomes were limited to proteinuria and kidney function instead of hard renal outcomes such as ESRD .
The association of risk factors for CKD with prevalence of CKD among European countries should be interpreted with caution.
As described above, the international comparison of prevalence of CKD may be hampered by heterogeneity of study methods. In addition, it is very likely that the violation of the chronicity criterion by using a single assessment of serum creatinine has led to a systematic overestimation of CKD prevalence in all countries and regions . Furthermore, we were unable to include all European studies that published an estimation of CKD stages 3–5 prevalence, such as the one by Girndt et al., as they did not report the CKD prevalence for the age group 65–74 years .
Along the same lines, the international differences in the prevalence of ‘risk factors’ for CKD (Figure 2) may also have been affected by sample selection. In addition, although several CKD risk factors are presented as age-standardized values, for those risk factors that are non-age-standardized (i.e. diabetes mellitus and salt intake) the differences in prevalence may be partly caused by differences in age distribution across the European countries. Moreover, it is unclear whether the assessment of risk factors for CKD was similar across countries or to what extent they may have suffered from measurement errors. For example, it is likely that a percentage of patients with type 2 diabetes mellitus remain undetected owing to a lack of overt symptoms. Also, lifestyle factors are usually assessed by self-reporting, such as physical inactivity, current smoking and salt intake. Self-reported data are prone to validity problems, as individuals may over- or underreport, might be mistaken or may not remember the amount or frequency of the risk factor. It is also important to recognize that national response systems have been implemented to prevent and control several important noncommunicable diseases and not only one noncommunicable disease, CKD.
Within this narrative review, we have discussed possible ecological associations (i.e. at a population level) between risk factors for CKD and prevalence of CKD. Also, the CKD stages 3–5 prevalence was estimated within the age group 65–74 years, whereas the estimation of the prevalence of lifestyle factors and cardiovascular mortality was assessed in the entire adult general population.
Additionally, one should bear in mind that, although the prevalence of risk factors of CKD and prevalence of CKD is often presented on a national level, it may vary by region . It is possible that in some regions, like the South-west of Germany and the North-west of France, a study on CKD prevalence was started due to the expected high CKD prevalence in that region, which may have resulted in selection bias (i.e. more studies with high CKD prevalence).
Finally, the figures in this article presenting risk factors for CKD contain the most recent data available, although the current CKD prevalence is more likely to be caused by the prevalence of risk factors in previous years.
This narrative overview shows substantial variation among European countries in the prevalence of CKD, in the risk factors for CKD, in cardiovascular mortality rates and in public health policies targeting noncommunicable diseases. In general, the countries with a higher CKD prevalence also have a higher score on risk factors for CKD.
Whether individual countries and regions undertake public health policy and practice interventions based on such data or wait for studies providing more robust evidence of causality is beyond the remit of this article to dictate. However, the general finding that higher prevalence of CKD risk factors may be associated with higher prevalence of CKD suggests a need for action in tackling key risk factors.
CONFLICT OF INTEREST STATEMENT