The renal tubular damage marker urinary N-acetyl-β-d-glucosaminidase may be more closely associated with early detection of atherosclerosis than the glomerular damage marker albuminuria in patients with type 2 diabetes

Cardiovascular disease (CVD) is a major cause of morbidity and mortality in patients with type 2 diabetes mellitus (T2D), who have a twofold to fourfold increase in CVD incidence compared with patients without diabetes [1]. Therefore, the development of methods to estimate the risk of CVD in diabetic patients with more accuracy is highly desirable and will aid in preventing CVD events. Several surrogate measures for CVD that estimate subclinical atherosclerosis have been developed. Acceptable surrogates include pulse wave velocity (PWV), which is the gold standard for determining arterial stiffness and plays an important role in the occurrence of atherosclerosis [2, 3], carotid artery intima-media thickness (IMT) and the presence of carotid plaques, which, despite different pathogenic characteristics, correlate with coronary artery disease and stroke [4], and albuminuria, which is highly predictive not only of the development of diabetic kidney disease (DKD) but also of subsequent CVD and death [5]. Furthermore, several previous studies report positive independent associations between albuminuria and indices of macrovascular complications such as PWV and coronary artery calcification in patients with T2D or essential hypertension [6‐8]. However, the relationship between IMT and albuminuria is still controversial [2, 9]. This controversy might be explained by differences in characteristics of study populations such as age, duration of diabetes, and ethnic diversity as well as the use of different methods to measure carotid atherosclerotic lesions [10].

Recently, several renal tubular damage markers have gained considerable attention because of their clinical relevance as sensitive and specific biomarkers for predicting the development and progression of early-stage DKD [11, 12]. The urinary enzyme N-acetyl-β-d-glucosaminidase (NAG) exists in the lysosomes of proximal tubule epithelial cells. Increased NAG excretion in urine is caused exclusively by proximal tubular cell injury. Furthermore, increases in urinary NAG already occur in normal to mildly increased albuminuric patients with T2D [11]. There is accumulating evidence that urinary NAG is correlated or associated not only with nephropathy but also with vascular complications of T2D, including retinopathy [13], neuropathy [14], and macrovascular disease [15]. However, little research has focused on the relationship between urinary NAG excretion and carotid IMT or plaques in patients with T2D.

The aim of this study was to investigate the association between urinary NAG and carotid IMT and plaques in patients with T2D. By comparing the predictive value of NAG with that of albuminuria, we also aimed to determine whether urinary NAG, a marker for renal tubulopathy, is a better predictor of carotid atherosclerosis than albuminuria, a marker for renal glomerulopathy.

Renal tubular damage markers, including urinary NAG, have gained considerable attention because of their clinical relevance as sensitive and specific biomarkers for predicting the development and progression of early-stage DKD [11, 12]. Furthermore, there is accumulating evidence that urinary NAG is correlated or associated with vascular complications of T2D, including not only nephropathy but also retinopathy [13], neuropathy [14], and macrovascular disease [15]. However, there is no information on the relationship between urinary NAG excretion and carotid atherosclerosis in patients with T2D. To our knowledge, only one study in patients with untreated hypertension has shown that urinary NAG excretion is positively associated with carotid IMT [20]. The aim of our study was to examine whether urinary NAG, a marker for renal tubulopathy, is more closely associated with carotid IMT and plaques in patients with T2D than albuminuria, a marker for renal glomerulopathy. In this study, we demonstrated several important findings. First, regarding carotid IMT, mean, maximum, and mean of maximum carotid IMT were significantly higher in the high NAG group than the low NAG group. In correlation analyses, urinary NAG showed a significant positive correlation with carotid IMT, whereas urinary ACR did not correlate significantly with carotid IMT. After adjustment for clinical and laboratory risk variables, including age, sex, waist circumference, current smoking status, hypertension, HDL cholesterol, eGFR, and urinary ACR, urinary NAG independently predicted measures of carotid IMT. Second, in terms of carotid plaques, both urinary NAG and ACR were significantly higher in participants with carotid plaques than in those without carotid plaques. In addition, having an above-median level of urinary NAG was associated with approximately twice the risk for the presence of carotid plaques compared with urinary NAG levels at or below the median. After adjustment for confounding variables, both urinary NAG and urinary ACR were significantly associated with the presence of carotid plaques.

Recently, attention to renal tubular damage markers as potential risk factor indicators for CVD is growing. There is accumulating evidence that urinary NAG is associated with various vascular diseases in different populations. In the GISSI-Prevenzione trial targeting subjects with chronic heart failure [21], urinary NAG level was consistently and strongly related to poor outcomes. In studies of type 1 or type 2 diabetes, retinopathy and peripheral neuropathy correlated with urinary NAG [13, 14, 22]. Weitgasser et al. [15] showed the association of urinary NAG with macrovascular disease in elderly T2D patients. That study investigated patients with T2D during a median follow-up of 7 years and showed that urinary NAG is comparable to albuminuria as an indicator of the preexistence and development of severe macrovascular disease, including myocardial infarction and peripheral vascular disease. Regarding subjects without diabetes, Ouchi et al. [23] reported that elevated urinary NAG is related to elevated arterial stiffness assessed by brachial-ankle PWV. In a study including a low-risk general population, urinary NAG and ACR were independently associated with first-ever myocardial infarction, first-ever ischemic stroke, and all-cause mortality. However, NAG did not add to the risk predicted by traditional cardiovascular risk factors, such as eGFR and ACR [24].

With respect to the relationship between albuminuria as a glomerular damage marker and subclinical atherosclerosis as assessed by IMT, although the association of albuminuria with CVD has been well-established in a number of populations, including diabetics, hypertensive patients, and general subjects [25, 26], the relationship between carotid IMT and albuminuria is still controversial, as some studies show conflicting results. Among studies of patients with T2D, Ishimura et al. [2] and Ito et al. [27] reported that the significant association between these two parameters was lost after adjustment for traditional risk factors such as blood pressure and waist-hip ratio. There may be several explanations for this discrepancy. First, diabetes itself has such a powerful influence on carotid IMT that any effect of albuminuria could be masked [9]. Second, the study populations differed in terms of age, duration of diabetes, eGFR, ethnicity, and the use of methods of measuring carotid atherosclerotic lesions [27].

Only a single study, to our knowledge, has investigated the association between subclinical atherosclerosis of IMT and urinary NAG [20]. It showed that urinary NAG and carotid IMT or PWV were positively correlated and that carotid IMT and PWV had significant positive effects on urinary NAG in regression analyses of patients with untreated hypertension. In the present study, we observed that measures of carotid IMT were significantly associated with urinary NAG rather than with urinary ACR in simple correlation and regression analyses. Because of the cross-sectional design of our study, possible mechanisms for the relationship between urinary NAG and carotid IMT remain unclear. There is evidence that chronic tubulointerstitial hypoxia and increased oxidative stress may play a pathogenic role in early stages of DKD [28], which in turn is related to the development of CVD. Furthermore, increased urinary NAG is a sensitive marker of tubular injury related to inflammation and oxidative stress [24]. Therefore, the elevation of urinary NAG, a renal interstitial marker, might reflect the severity of interstitial changes that result from atherosclerosis and subsequent ischemic injury [23]. In addition, the association between urinary NAG and carotid IMT might be mediated by several common risk factors. Consistent with previous reports [29], we observed that both urinary NAG and measures of carotid IMT were positively correlated with age and negatively correlated with eGFR, which are well-known risk factors for both DKD and systemic atherosclerosis. Nevertheless, we found several differences in the correlations of urinary NAG versus measures of carotid IMT with metabolic variables, including glucose-homeostasis and blood pressure. We previously suggested that increased urinary NAG is related to glycemic parameters reflecting glucose fluctuation (e.g., postprandial glucose rather than fasting glucose, GA rather than HbA_1C) and decreased insulin secretory capacity as assessed by HOMA-β or postprandial C-peptide-to-glucose ratio (PCGR) in patients with T2D [12]. Regarding plausible explanations for associations between glucose fluctuation and increases in urinary NAG excretion, we suggested the possibility of a physiologic increase in urinary NAG for metabolizing urinary glucose and a nephrotoxic effect of glycated end products on renal proximal tubules. NAG is an enzyme involved in carbohydrate metabolism; thus, when the proximal tubules are exposed to high urinary glucose, NAG might be more highly secreted in the urine, depending on urinary glucose concentrations [30]. In addition, the peptides derived from advanced glycation end products could have a nephrotoxic effect on the proximal tubule, thus contributing to the occurrence of proximal tubule injury [31]. We previously suggested [32] that serum GA is more closely associated with postprandial glucose than fasting glucose and might be a useful index for monitoring fluctuating and poorly controlled glycemic excursions. The clinical relevance of GA could be attributed to observations that increased GA rather than HbA_1C is significantly correlated with insulin secretory beta-cell function as assessed by PCGR and HOMA-β and that GA increases as the duration of diabetes lengthens [33]. Despite the lack of association between HbA_1C and IMT, consistent with previous studies [34], increasing waist circumference was associated only with measures of carotid IMT, which could be explained by the association of insulin resistance with increased IMT [35, 36]. However, neither urinary NAG nor measures of carotid IMT were significantly associated with BMI [37]. Urinary NAG was not related to high blood pressure, as was noted in a previous study [15], whereas measures of carotid IMT were negatively correlated with diastolic blood pressure.

There is a well-established association between urinary ACR and carotid plaques in patients with T2D [10, 38]. In this study, we found that the presence of carotid plaques was associated with age, long duration of diabetes, history of hypertension, coronary artery disease, or ischemic stroke, reduction in eGFR, and increases in both urinary ACR and urinary NAG. An above-median level of urinary NAG (OR 1.96, 95% CI 1.12–3.41, p = 0.02) was significantly associated with the presence of carotid plaques. However, urinary ACR levels (OR 2.33, 95% CI 0.97–5.63, p = 0.06) were not significantly related to the presence of carotid plaques in an unadjusted model. After adjustment for age, sex, and history of hypertension, both urinary NAG and urinary ACR were significantly associated with the presence of carotid plaques. After further adjustment to use lipid-lowering drugs and eGFR CKD-EPI, the associations between urinary NAG or urinary ACR and carotid plaques remained significant. These results show that urinary NAG is comparable to urinary ACR in predicting the presence of carotid plaques. To our knowledge, this is the first time the relationship between urinary NAG and carotid plaques has been investigated.

The current study has several strengths. First, our study provides clinical evidence supporting associations between urinary NAG and various vascular complications of diabetes. Second, we investigated whether urinary NAG, a renal tubulopathic marker, is more closely associated with carotid atherosclerosis than the glomerulopathic marker albuminuria, which might reinforce the comparability between markers. Additionally, we conducted the study with a relatively large number of subjects in a single center trial, which strengthens the statistical reliability of the results. However, the current study also has some limitations. First, because of the inherent drawbacks of cross-sectional studies, we cannot infer any causal or temporal relationships between urinary NAG and carotid atherosclerosis. Second, the values of Spearman’s correlation coefficients between urinary NAG and measures of carotid IMT were statistically significant but of only slight to fair magnitude. The process of increasing IMT is a complex phenomenon that is not only determined by pure atherosclerotic risk factors [39]. Thus, multifactorial determinants of IMT may reduce the statistical significance of urinary NAG. Finally, we did not compare the sensitivity of detecting carotid atherosclerosis of urinary NAG using other renal tubular damage markers. For example, urinary liver-type fatty acid binding protein is also a sensitive marker of renal injury [40].

We evaluated the association of urinary NAG with subclinical atherosclerosis as assessed by carotid IMT and revealed a novel finding that increased carotid IMT and the presence of carotid plaques are related to elevated urinary NAG, a marker of renal tubular damage, in patients with T2D. Furthermore, we found that urinary NAG might be a more sensitive urinary biomarker than urinary albumin for early detection of atherosclerosis. Further longitudinal prospective studies are needed to investigate the causal relationship between renal tubular damage and systemic atherosclerosis and its mechanisms.