Background

Diabetic nephropathy is a major complication of type 2 diabetes mellitus (T2DM), and is the most common cause of end-stage renal disease. The prevalence of diabetic nephropathy is increasing worldwide [1]. Further, diabetic nephropathy, especially macroalbuminuria, is reported to be the primary risk factor for cardiovascular disease [2], and, eventually, mortality [3]. Thus, the prevention and treatment of diabetic nephropathy, especially macroalbuminuria, are important for the prevention of renal and cardiovascular events and death.

The role of fatty acids (FAs) in inflammation and related chronic diseases is well established. Originally, albumin is an FA transport protein, and hyperfiltration in diabetes causes a high concentration of FAs to enter the tubular space, along with albumin [4]. Circulating FAs are composed of saturated FAs without double bonds, monounsaturated FAs with a single double bond, and polyunsaturated FAs (PUFAs) with multiple double bonds. Several studies have reported that increased influx of these FAs, especially saturated FAs, contributes to the worsening of tubular damage [5, 6]. In addition to saturated FAs, we should focus on the effects of the other types of FAs in organ damage through lipotoxicity. Previous studies reported that n-3 PUFAs, such as eicosapentaenoic acid (EPA) and docosahexaenoic acid, have anti-inflammatory and cardiovascular protective properties [7,8,9]. On the other hand, n-6 PUFA, such as arachidonic acid (AA), produces a series of lipid mediators called eicosanoids (prostaglandins and leukotrienes), which play a central role in enhancing vascular permeability, and neutrophil infiltration and activation during the early stages of the inflammatory response [10]. In addition, a low ratio of serum EPA to AA has been reported to be a risk factor for cardiovascular events [11,12,13]. Conversely, several previous studies demonstrated that moderate intake of dietary n-6 PUFAs lowers the risk of coronary artery disease [14] and circulating n-6 PUFAs were inversely associated with total mortality and coronary artery disease mortality [15].

In recent years, it has become clear that the concentration of each circulating FAs in patients with diabetes differed from those of people without diabetes and that these differences may contribute to the development of organ damage [16,17,18]. However, the association between these circulating FAs, especially detail components of FAs, and diabetic nephropathy has not yet been clarified.

Thus, we conducted the present cross-sectional study with an aim to determine the relation between circulating FAs and diabetic nephropathy.

Methods

Study design and participants

The KAMOGAWA-DM cohort study, is an ongoing prospective cohort study that began in 2014 [19], included the outpatient clinics of the Kyoto Prefectural University of Medicine (Kyoto, Japan) and the Kameoka Municipal Hospital (Kameoka, Japan). For the present cross-sectional study, we included T2DM patients in whom FA measurements were performed. We excluded patients taking EPA medication. This study was approved by the hospital’s Ethics Committee. T2DM was diagnosed based on the criteria by the American Diabetes Association [20].

Data collection

The background factors of patients (i.e., sex, age, disease duration, smoking, and the history of cardiovascular disease (CVD)) was gathered from their medical records. CVD was defined as myocardial infarction and stroke. Blood pressure was measured in an outpatient setting. After an overnight fast, venous blood samples were collected from the patients and fasting plasma glucose, triglycerides, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, creatinine, uric acids, and C-peptide were measured. Hemoglobin A1c (HbA1c) levels were measured by high-performance liquid chromatography, ant the values are presented as the National Glycohemoglobin Standardization Program unit. An index of insulin resistance using serum C-peptide concentration, 20/(fasting C-peptide × fasting plasma glucose) [21], was used. All patients submitted spot morning urine samples, and urinary albumin excretion (UAE) was measured by immunoturbidimetric assay. Mean UAE in the present study was used as the average of the three urinary measurements. Macroalbuminuria was defined as UAE > 300 mg/g creatinine [22].

Diabetic neuropathy was diagnosed based on the criteria by The Diagnostic Neuropathy Study Group [23]. Retinopathy was classified into four categories according to the ophthalmologist’s diagnosis as follows: NDR, no diabetic retinopathy; SDR, simple diabetic retinopathy; PPDR, pre-proliferative diabetic retinopathy; and PDR, proliferative diabetic retinopathy.

Estimation and assessment of the patients’ habitual diet and nutritional intake

A brief-type self-administered diet history questionnaire (BDHQ) was adopted for the assessment of each patient’s habitual diet and nutritional intake [24]. The details of the BDHQ are provided elsewhere [24]. In brief, the BDHQ estimates the dietary intake of 58 food items over the past month from respondents’ memories using a computer algorithm, based on the Standard Tables of Food Composition in Japan [25]. And then, the dietary total energy (kcal/day), carbohydrate (g/day), protein (g/day), total fat (g/day), saturated FAs (g/day), monosaturated FAs (g/day), PUFAs (g/day), n-3 FAs (g/day), n-6 FAs (g/day), total fiber (g/day), cholesterol (g/day), alcohol (g/day), and sodium (g/day) intake were estimated using this calculation program of the BDHQ.

Measurement of circulating free fatty acids

The composition of FAs in frozen serum samples was measured by GC/MS, Agilent 7890B/5977B (Agilent Technologies, Santa Clara, CA, USA). We metylated 25 μl of serum using a FA methylation kit (Nacalai Tesque, Kyoto, Japan), and loaded the final product onto a Varian capillary column (DB-FATWAX UI; Agilent Technologies). The capillary column used for FA separation was CP-Sil 88 for FAME (100 m × an inner diameter of 0.25 mm × membrane thickness of 0.20 μm, Agilent Technologies). We set the temperature in column at 100 °C for 4 min and then increased gradually by 3 °C/min to 240 °C and held there for 7 min. We injected the samples in split mode at a split ratio of 5:1. Each FA methyl ester was detected in the selected ion monitoring mode. All results were normalized to the peak height of the C17:0 internal standard [26].

Statistical analysis

JMP (ver. 13.0) software (SAS Institute, Cary, NC, USA) was used for all analyses of the patients’ data. Probability values of < 0.05 were accepted as significant. Categorical variables were expressed as numbers, and continuous variables were presented as the mean ± SD. To assess the statistical significance of the differences between groups, the chi-square test for categorical variables and the Wilcoxon signed-rank test for continuous variables was used because the continuous variables did not follow a normal distribution. Spearman rank correlation coefficients were used to investigate the association between the logarithm of FAs and the logarithm of UAE. Then, because AA was found to be significantly correlated with the logarithm UAE, the odds ratio (OR) and 95% confidence interval (CI) of AA and other variables in the presence of macroalbuminuria were calculated by performing univariate and multivariate logistic regression analyses. We adjusted for age and sex (Model 2) and further adjusted for body mass index (BMI), disease duration, index of insulin resistance, total cholesterol, systolic blood pressure, smoking status, physical activity, total energy, usage of renin-angiotensin system inhibitors, creatinine and the history of CVD (Model 3). In addition, the area under the curve (AUC) of circulating arachidonic acid concentration for the presence of macroalbuminuria was calculated by the receiver operating characteristic (ROC) curve.

Results

In this study, the BDHQ was administered to 426 patients (234 men and 192 women) with T2DM. Among them, we measured circulating FAs in 197 patients (96 men and 101 women) using the GC/MS system. Moreover, we excluded 7 patients with EPA medication usage (Fig. 1).

Fig. 1
figure 1

The registration of patients

Full size image

The clinical characteristics of the 190 patients with T2DM are summarized in Table 1. Mean age, HbA1c, and UAE were 62.7 ± 12.1 years, 7.5 ± 1.6%, and 202.0 ± 583.0 mg/gCr, respectively. The number (percentage) of patients with macroalbuminuria was 27 (13.7%).

Table 1 Clinical characteristics of the study patients
Full size table

Mean circulating FA concentrations are shown in Table 2. Pentadecanoic acid (C15:0), oleic acid (C18:1(9c)), C18:3(9c12t15t) + C18:3(9c12c15t), α-Linolenic acid (C18:3(9c12c15c)), eicosenoic acid (C20:1), eicosadienoic acid (C20:2), AA (C20:4n6), heneicosylic acid (C21:0), behenic acid (C22:0), and tricosanoic acid (C23:0) were significantly lower in patients with macroalbuminuria than in those without macroalbuminuria.

Table 2 Circulating fatty acids of the study patients
Full size table

In addition, the mean dietary intakes of nutrients, including FAs, are shown in Table 3. Energy intake in patients with macroalbuminuria tended to be higher than that in patients without macroalbuminuria (1891.5 ± 809.9 kcal/day vs. 1649.6 ± 551.3 kcal/day, p = 0.069), whereas total fat intake was not significantly different (57.9 ± 22.0 g/day vs. 53.0 ± 20.0 g/day, p = 0.277). The dietary intake of eicosenoic acid (C20:1), AA (C20:4n6), EPA (C20:5n3), erucic acid (C22:1n9), erucic acid (C22:1n9), docosahexaenoic acid (C22:6n3), and nervonic acid (C24:1) in patients with macroalbuminuria was significantly higher than that in patients without macroalbuminuria.

Table 3 Dietary fatty acids intake of the study patients
Full size table

Moreover, we investigated the association between log circulating FAs and log UAE. Logarithm AA showed a significant correlation (r = − 0.221, p = 0.002) (Table 4). Therefore, we performed logistic regression analyses to investigate the association between logarithm AA and the presence of macroalbuminuria. In univariate logistic regression analysis, logarithm AA was negatively associated with the presence of macroalbuminuria (Model 1, OR of Δ1 incremental: 0.03, 95% CI: 0.01–0.21, p <  0.001). Multivariate logistic regression analysis revealed that logarithm AA was negatively associated with the presence of macroalbuminuria, even after adjusting for covariates (OR of Δ1 incremental: 0.32, 95% CI: 0.10–0.99, p = 0.042) (Table 5).

Table 4 Correlation coefficient between logarithm urinary albumin excretion and serum fatty acids
Full size table
Table 5 Logistic regression analyses for macroalbuminuria
Full size table

Additionally, in ROC analyses, AUC of circulating arachidonic acid concentration was 0.714 (95% CI 0.601–0.827) and the cut-off level of serum AAs with diabetic patients for predicting the presence of macroalbuminuria was 327.5 μg/ml (Fig. 2).

Fig. 2
figure 2

Area under the receiver operating characteristic (ROC) curve (AUC) [95% confidence interval (CI)] of several factors for the presence of macroalbuminuria. AUC of circulating arachidonic acid concentration was 0.714 (95% CI 0.601–0.827), and the cut-off level of serum AAs with diabetic patients for predicting the presence of macroalbuminuria was 327.5 μg/ml

Full size image

Discussion

In the present study, in a cohort of 190 Japanese individuals with T2DM, we investigated the association between circulating FAs and diabetic nephropathy, and we revealed that circulating AA is negatively associated with UAE and the presence of macroalbuminuria.

Possible explanations for the association between circulating AA and diabetic nephropathy are described below. AA is a component of biological membranes and is involved in the generation of prostaglandins, which are the most representative physiologically active lipids, through cyclooxygenase (COX). Prostacyclin (PGI2), a type of prostaglandin, has vasodilatory effects [27], antiplatelet effects [28], angiogenesis [29], and protective effects on vascular endothelium [30]. PGI2 is produced from prostaglandin H2 (PGH2) by prostacyclin synthase [31]. In addition, prostaglandin E1 has a protective effect on diabetic kidney disease by decreasing UAE [32], whereas prostaglandin E2 decrease myocardium contractility and cause cardiorenal syndrome [33]. AA is associated with PGH2, through separation from phospholipids in the cell membrane and receiving COX. PGI2 is well known for its regulation of renal hemodynamics, tubular transport, and renin release and plays an important role by coupling with its receptors and the downstream signals in various types of renal diseases including chronic kidney disease [34]. Therefore, circulating AA levels in patients with macroalbuminuria were significantly lower than those with normoalbuminuria and microalbuminuria, suggesting that the body promoted the production of PGI2 to prevent progression of nephropathy and the other CVD.

On the contrary, AA has been reported to potentiate hypoxia-induced vascular endothelial growth factor (VEGF) expression through the Notch-1, Wnt-1, and HIF-1alpha pathways [35]. In several previous studies, VEGF was proven to have an ameliorating effect on diabetic nephropathy [36], normalization of hyperpermeability in the glomeruli of diabetic nephropathy [37], and in restoring endothelial glycocalyx in diabetic nephropathy [38]. It is suggested that AA is consumed for PGI2 production and the concentration of AA in the body is lowered, which may reduce the effects of VEGF from being fully activated. Collectively, patients with advanced diabetic nephropathy might consume more AA to raise levels of PGI2, which leads to lower circulating AA levels and decreased AA reduce the effect of VEGF and induces further progression of nephropathy. This is potentially the reason why circulating AA levels in patients with diabetic nephropathy are significantly lower than those of patients without this condition, although dietary AA intake in patients with diabetic nephropathy is higher than that in patients without diabetic nephropathy.

There are three major limitations in this study that could be addressed in future research. First, this study had a cross-sectional design; thus, further studies are required to establish the causal relationship between circulating AA levels and diabetic nephropathy. Second, the sample size in this study was relatively small. Thus, further large-scale studies are needed. Third, we did not check circulating prostaglandin level.

Conclusions

In conclusion, this study provides the first demonstration that circulating AA levels are negatively associated with UAE and that these levels are significantly lower in patients with macroalbuminuria than in those without macroalbuminuria. Future prospective studies on the relationship between circulating AA levels and the incidence of diabetic nephropathy and the differences in mechanism between dietary AA and circulating AA are required.