Association between prediabetes (defined by HbA1_C, fasting plasma glucose, and impaired glucose tolerance) and the development of chronic kidney disease: a 9-year prospective cohort study

The prevalence of chronic kidney disease (CKD), which is characterized by structural kidney damage and chronic reduction of kidney function, has increased globally with an increase in the aging population [1]. CKD constitutes a major health problem because it increases all-cause mortality and the development of cardiovascular disease (CVD) [2]. Therefore, it is important to identify modifiable risk factors that can act as early predictors of incident CKD.

Diabetes is the leading cause of CKD in most countries, and up to one-third of adults with newly diagnosed diabetes already have CKD, suggesting that CKD may occur in the early stages of diabetes [3, 4]. Prediabetes, which is regarded as a CVD risk factor, is an intermediate state of hyperglycemia, wherein blood glucose is higher than normal but lower than that in diabetes [5, 6]. Considering that prediabetes can be a precursor of diabetes, we hypothesized that prediabetes is associated with incident CKD, through the effect of hyperglycemia. However, very few studies have investigated such an association. In addition, prediabetes can be defined according to impaired fasting glucose (IFG), impaired glucose tolerance (IGT), or HbA1_C level, and a previous study suggested that each of these components might have a different association with CKD [7].

The aim of the present study was to clarify the clinical impact of prediabetes, defined by IFG, IGT, and HbA1_C level, on the development of incident CKD in a Korean adult population using data from the Korea Genome and Epidemiology Study (KoGES).

The mean age of subjects was 52 years, with 52.6% of subjects being male. Mean follow-up duration was 8.7 years. Among the 10,038 patients, 7728 patients were included in this study, after excluding 216 patients with CKD, 1046 with diabetes, 982 patients who were never followed up, and 66 patients with insufficient data.

The anthropometric and biochemical characteristics of the subjects are summarized in Table 1 according to the development of incident CKD. Traditional risk factors for CKD such as older age, obesity, hypertension, and metabolic syndrome were more prevalent in the CKD group. Prediabetes defined according to IGT or HbA1_C level was more frequently noted in the CKD group, but the incidence of IFG was not different between the CKD and non-CKD groups. Interestingly, male sex and regular exercise were associated with a lower incidence of CKD. The incidence of CVD according to IGT is shown in Table 2. The incidence of diabetes and CKD increased in subjects with IGT. The incidence of other CVDs, including stroke and coronary artery disease, did not differ according to IGT. Using a Cox proportional hazards model, we also investigated the clinical impact of IGT on the development of CKD during the follow-up period. Participants with IGT had a higher risk for CKD development (hazard ratio [HR] 1.14, 95% confidence interval [CI] 1.182–1.310, P = 0.043), as shown in Table 3. Moreover, this significant association persisted after adjusting for other CKD risk factors, including age, sex, obesity, hypertension, and metabolic syndrome. Traditional CKD risk factors such as age, obesity, hypertension, and metabolic syndrome also significantly increased the development of CKD after adjusting for confounding factors. We also analyzed the effect of other prediabetes definitions on CKD development by using a Cox model. Consistent with the IGT findings, prediabetes defined by HbA1_C level was also significantly associated with CKD development after adjusting for possible confounding factors (HR 1.39, 95% CI 1.21–1.60, P < 0.001) as shown in Table 4. IFG was not associated with future CKD development.

In this large prospective community-based cohort study of Korean adults, we found that prediabetes defined by IGT or HbA1_C was significantly associated with the development of CKD, independent of traditional CKD risk factors. However, the incidence and relative risk for CKD was not associated with IFG. To the best of our knowledge, this is the first and largest population-based prospective study to evaluate the association between prediabetes, as assessed by IGT, HbA1_C, and IFG, and the development of CKD.

In addition, older age (> 65 years), BMI, hypertension, and metabolic syndrome were revealed as risk factors for CKD in our community-based cohort. These results were in line with previous studies. Older age [9], obesity [10], hypertension [11], and metabolic syndrome [12] are well-recognized risk factors for developing CKD. Other risk factors, including dyslipidemia and hypertension are also associated with the development of CKD in patients with diabetes [13, 14]. Male gender was associated with an increased risk of CKD in our study. Previous studies showed the prevalence of CKD is higher in women than in men including US renal data system (USRDS) annual data report [15‐17]. However, there were also opposing data regarding this issue [18‐20]. Korea data also showed female gender was associated with less reduced renal function (Odds ratio 0.874, 95% CI 0.766–.0.997) and trend toward lower incidence of CKD (Odds ratio 0.953, 95% CI 0.848–1.071) [21]. Thus, we think that there might be a geographic variability in the effect of gender on the prevalence of CKD.

Several studies have shown that prediabetes and metabolic syndrome are associated with CKD development, after adjustment for established CKD risk factors [22‐26]. In one meta-analysis that included 9 cohort studies (8 IFG-defined prediabetes and 1 HbA_1C-defined prediabetes study) showed that the overall relative risk of CKD was 1.12 (95% CI. 1.02–1.21), after adjustments for established risk factors [7]. However, most of the studies defined prediabetes according to IFG, since IGT determination requires 2-h post-glucose measurement. Our study is different from earlier studies that reported an association between prediabetes and CKD because we used three different definitions of prediabetes. Interestingly, there was no difference in baseline fasting glucose level or the incidence of IFG according to the presence of overt CKD in our subjects. In contrast, baseline IGT- or HbA1_C-defined prediabetes was more prevalent in the CKD group, and the hazard ratio was also robustly higher after adjusting for traditional CKD risk factors. It is difficult to know exactly whether this is an incidental finding or a meaningful difference. In previous studies, only one definition was studied, or both IGT and IFG were included in the definition of prediabetes [26]. Although, both IGT and IFG are included in prediabetes, the clinical impact or underlying mechanism might be different. We think IGT or HbA1_C level may better reflect long-term exposure to non-diabetic hyperglycemia than IFG. Previous study reported that IFG reflects defects in insulin secretion and IGT reflects insulin resistance [27]. Milicevic et al. [28] reported that IGT and HbA1_C were more strongly associated with cardiovascular outcomes than IFG was. Therefore, clinical significance of each indicator is expected to be studied in the future.

Previous researchers suggested some potential mechanisms by which prediabetes might affect the initiation and progression of CKD. Individuals with prediabetes can be as insulin-resistant as individuals with diabetes. Thus, dysfunction in the metabolism of lipids and carbohydrates and loss of protein stores, which are often observed in patients with insulin-resistance, may be associated with kidney damage in prediabetic individuals [29, 30]. Cross-sectional studies have revealed that borderline hyperglycemia or prehypertension is associated with glomerular hyperfiltration, which is a marker of early renal damage [31, 32]. Thus, in subjects with prediabetes or prehypertension, especially with comorbid glomerular hyperfiltration, earlier treatment of hyperglycemia or high blood pressure may be necessary to prevent the development of kidney damage. In addition, insulin resistance may lead to CKD by alteration of endothelial function and inflammation, sodium retention, and activation of the sympathetic nervous system [33, 34].

The present study has several advantages over previous studies. As mentioned above, we examined all the three commonly used prediabetes definition parameters (IFG, IGT, and HbA1_C) to identify their relationship with CKD. Second, this study prospectively followed a relatively large number of subjects for an average of 9 years. Development of incidental CKD is preceded by glomerular hyperfiltration, which is an important link between prediabetes and CKD development. Thus, short or intermediate follow-up periods might not capture the association between prediabetes and CKD. Moreover, although most of the published studies have relied on GFR estimated using the Modification of Diet in Renal Disease Study (MDRD) equation [7], we defined CKD using the CKD-EPI equation, which performs better than the MDRD equation, especially at higher GFR, with less bias and greater accuracy [35]. Because most individuals with prediabetes have higher GFR, the CKD-EPI equation might better reflect kidney function.

This study has several limitations. First, GFR was not measured directly, but was estimated using a serum creatinine-based equation, which might have overestimated or underestimated the actual GFR. Second, we did not have information on urine analysis including microalbuminuria, a known risk factor for the development of kidney disease in individuals with diabetes. Considering CKD is defined by the presence of albuminuria as well as by eGFR, overall CKD may be underestimated in this study. Third, because all participants were required to attend follow-up examination, those who died or developed serious comorbidities were excluded; thus, we may have introduced survival bias. Fourth, this study was restricted to only Koreans. Therefore, these results might not be applicable to other ethnic groups.

The results of this study show that subjects with prediabetes defined by IGT or HbA1_C level have an increased risk of incident CKD; however, IFG was not associated with CKD. Thus, our findings indicate that IGT or HbA1C are predictors of CKD, and might enable clinicians to better prevent the development of CKD in individuals with prediabetes, potentially resulting in better outcomes.