Update on pathogenesis and diagnosis flow of normoalbuminuric diabetes with renal insufficiency

Over the years, albuminuria has been regarded as a pivotal indicator of diabetic kidney disease (DKD) and invariably precedes estimated glomerular filtration rate (eGFR) loss. However, this classical view has been challenged by evidence that a considerable proportion of patients with diabetes have renal insufficiency in the absence of albuminuria, referred to normoalbuminuric diabetic kidney disease [1, 2] or nonalbuminuric diabetic nephropathy [1, 3] or diabetic kidney disease without albuminuria [4], or nonalbuminuric renal insufficiency in type 2 diabetes [5], defined as eGFR < 60 mL/min per 1.73 m² and in the absence of albuminuria (urinary albumin excretion rate (UAER) < 30 mg/24 h or urinary albumin creatinine ratio (UACR) < 30 mg/g), has been identified as albuminuria-independent phenotype of DKD [6, 7]. The Italian Diabetes Society and Italian Society of Nephrology also well described the nonalbuminuric renal impairment phenotype in a joint position statement on the natural history of diabetes mellitus in 2020 [8].

The combination of renal insufficiency and normoalbuminuria in diabetes was first highlighted by Lane et al. in 1992 [9], and has been confirmed for decades. The prevalence of the normoalbuminuric diabetes with renal insufficiency (NADRI) increased and has become a prevailing phenotype of DKD at present [10]. Even though these patients carry a lower risk for chronic kidney disease (CKD) progression, end-stage renal disease, or cardiovascular disease or all-cause mortality compared to those with albuminuric diabetes with renal insufficiency (ADRI) [7, 11‐19], it significantly contributes to the burden of advanced CKD or end-stage renal disease morbidity [6, 13, 20]. Moreover, mortality rates generally increased in these patients [10].

Currently, the natural course and potential pathogenesis of NADRI are still elusive. In this article, we review the latest studies addressing the epidemiology, clinical characteristics, and pathology of NADRI. Based on the studies of clinical features and renal histopathologic changes, we attempt to propose an underlying pathogenesis model and a flow chart for diagnosing NADRI.

Two early studies reported that renal insufficiency may occur in both type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM) with normoalbuminuria [9, 21], and these observations have been confirmed in subsequent studies.

In T2DM, the prevalence of NADRI ranges from 14 to 57% in cross-sectional studies, with differences among surveys depending on the geographic area and the calculation formula of eGFR [5, 13, 14, 21‐36]. A higher incidence of NADRI (more than 60%) was observed in recent years [37‐41]. Similar high prevalence was noted in multicenter multinational interventional researches, in which values were affected by different inclusion criteria, nevertheless [42‐45]. Different from T2DM, the prevalence variation of NADRI in T1DM is significant, which ranges from 2 to 58% [9, 15, 38, 46‐48]. A recent study indicated that the proportion of NADRI was remarkably high (58.6%) [49].

These data are in line with the increasing prevalence of declined eGFR and the decreasing incidence of albuminuria observed in diabetes [50, 51]. The increasing prevalence of NADRI may be partly due to the wide application of renin–angiotensin–aldosterone system (RAAS) inhibitors, advances in the treatment of antihyperglycemic, antihypertensive and hypolipidemic, and smoking cessation [19, 24, 52‐54].

The reported clinical features of NADRI include a higher proportion of females [7, 15, 16, 36, 37, 47], more likely to be older [15, 16, 47] and nonsmokers [7, 15, 16, 36, 37, 47]; a shorter diabetes duration [36, 55, 56]; lower levels of systolic blood pressure [7, 13, 15, 16, 36, 37, 47, 55, 56], diastolic blood pressure [13, 15, 16, 47, 55], hemoglobin A1c (HbA1c) [7, 15, 36, 37, 47], triglycerides[15, 36, 37, 47], total cholesterol [13, 15, 47, 55], and low-density lipoprotein cholesterol [13, 15, 47]; higher levels of high-density lipoprotein cholesterol [15, 18, 36, 47]; and lower prevalence of diabetic retinopathy [7, 15, 18, 37, 55, 56], neuropathy [55], and cardiovascular disease [7, 16, 37], as compared to patients with ADRI (Table 1).

Compared to patients with micro-ADRI and/or macro-ADRI, the clinical features of NADRI include a higher proportion of females [5, 18, 57], more likely to be older [5, 20] and nonsmokers [18, 23, 58], a shorter diabetes duration [57]; lower levels of systolic blood pressure [20], diastolic blood pressure [20], HbA1c [20, 23, 57], and triglycerides [20]; higher levels of high-density lipoprotein cholesterol[20]; and lower prevalence of hypertension [57], diabetic retinopathy [23, 57], and cardiovascular disease [23] (Table 2).

In addition, compared to patients with normoalbuminuric preserved renal function (eGFR > 60 mL/min/1.73 m²), individuals with NADRI are older [16, 28, 59], more frequently females [16, 28, 59] and nonsmokers [16, 28], and have a higher prevalence of hypertension [28, 60] and hyperlipidemia (including triglycerides, total cholesterol, or low-density lipoprotein cholesterol) [28, 59] and metabolic syndrome [59] and cardiovascular disease [24, 28].

A retrospective study found albuminuria or/and low eGFR was absent in nearly 20% [20 of 106] of individuals throughout life showed histopathologic changes characteristic of DKD. Moreover, structural changes were greatly variable, including almost all histopathologic classification of DKD[61].

Currently, the pathogenesis of NADRI remains to be fully elucidated. The proposed pathogenic mechanisms that underlie NADRI are multifactorial, complex and may involve age-associated renal senescence, hypertension, dyslipidemia, obesity, renal hypertensive and interstitial fibrosis, vascular disease, arteriosclerosis, cholesterol microemboli, lipid toxicity, inflammation, and masking of albuminuria by RAAS inhibitors [4, 5, 22, 28, 58, 70‐72]. Based on the clinical research of identified clinical and pathological characteristics, we attempt to propose a hypothetical model for the development of NADRI (Fig. 1).

Kidney Disease Outcome Quality Initiative (KDOQI) guidelines recommend that the diagnosis of DKD should be based on monitoring UACR or UAER, fundus changes, and evaluation of eGFR. Based on KDOQI guidelines recommendations and clinical studies of NADRI, we attempt to propose the diagnostic criteria of NADRI: (1) to meet the latest diabetes diagnostic standards of World Health Organization or American Diabetes Association (ADA); (2) to test and prove within 6 months, UAER < 30 mg/24 h or UAER < 20 µg/min or UACR < 30 mg/g, at least twice in three measurements, to avoid confounding by transient increases in conditions, such as exercise, fever, hematuria, urinary tract infection, and congestive heart failure[140]; (3) to detect within 6 months, eGFR < 60 mL/min/ 1.73 m², at least twice in three measurements, and exclude eGFR reduction caused by AKI and other reasons; and (4) to exclude other secondary nondiabetic kidney diseases (Fig. 2).

eGFR estimation formula based on serum creatinine is generally recommended for diagnosis of DKD, including Cockcroft-Gault formula, modification of diet in renal disease formula, and CKD epidemiology collaboration (CKD-EPI) formula [141, 142]. However, serum creatinine level is susceptible to be influenced by factors, like muscle mass, dietary protein intake, hyperlipidemia, hemolysis, etc.; the sensitivity in eGFR is insufficient. Therefore, in recent years, the Kidney Disease: Improving Global Outcomes (KDIGO) and ADA proposed to use the eGFR calculation formula based on serum creatinine and cystatin C, such as CKD-EPI cystatin C formula and CKD-EPI creatinine-cystatin C formula [143]. Consequently, according to the latest KDIGO guidelines, we recommend adding the detection of serum cystatin C, and the eGFR may be calculated according to the cystatin C and creatinine-cystatin C formula.

Epidemiological data demonstrate that NADRI is a prevalent clinical phenotype of DKD. Moreover, NADRI exhibits distinct clinical characteristics and a wide heterogeneity of histology features. Based on the studies of clinical features and renal histopathologic changes of NADRI, we attempt to propose an underlying pathogenesis model and a flow chart for diagnosing NADRI, which increases the practicality and reliability of the study. The underlying pathogenesis associated with NADRI involves multiple factors and comprises the usual suspects, including obesity, hyperlipidemia, tubulointerstitial lesion and vascular disease, age, gene polymorphism, female gender, immune inflammation and oxidative stress, drugs, and AKI, while appears to be at least partially independent of hyperglycemia and hypertension and does not exclude the possible role of hyperuricemia. Early diagnosis and a deeper understanding of pathogenesis may provide theoretical guidance for the therapy of NADRI and prevent or delay the CKD progression, end-stage renal disease, and diabetes complications as far as possible.

Although many of factors are discussed in this review, the list is not exhaustive. Multilevel, multicenter, large-sample clinical studies are needed to further clarify the pathogenesis of the disease, discover specific biomarkers, establish clear and precise laboratory diagnostic standards for the disease, and provide an important theoretical basis for preventing the occurrence, development, early intervention, and treatment of the disease.