Introduction

Hypertension is commonly associated with type 2 diabetes (T2DM) [1]. Cardiovascular disease is a major cause of morbidity and mortality in T2DM patients [2], indicating that cardiovascular risk management is the most crucial issue for patients with T2DM.

High sodium and low potassium intake is a principal risk factor for cardiovascular diseases [3, 4]. However, estimation of sodium intake in the Japanese population has been reported to show a mean of 5529 mg/day in men and 4650 mg/day in women. Thus, the sodium intake in Japan is much higher than the WHO recommendation for adults of 2000 mg/day [5]. Furthermore, the potassium intake is also lower than the recommended values [6]. Because accurately measuring sodium and potassium intakes from dietary surveys are difficult, 24-h urine collection is highly recommended as a standard method to estimate these values [7,8,9]. Compared to urinary excretions of sodium or potassium alone, the sodium-to-potassium ratios in 24-h-collected urine have been reported to have strong associations with blood pressure [10,11,12,13] and an increased risk of cardiovascular diseases [14]. Although dietary balance of sodium and potassium is important for preventing cardiovascular diseases, adherence is poor in T2DM patients [15].

In addition, patients with T2DM frequently have multiple risk factors for cardiovascular disease [16]. Obesity is a well-known predictor of T2DM and an independent risk factor for cardiovascular diseases [17,18,19]. Obesity also modifies the association between sodium and potassium intake and blood pressure [20]. However, the impact of obesity on the relationship between the sodium-to-potassium ratio and cardiovascular risk in T2DM patients remains unclear.

Asymmetric dimethylarginine (ADMA) is an endogenous modified amino acid that competitively inhibits the synthesis of endothelium-derived nitric oxide (NO, a putative vasodilator) [21]. Several studies have reported elevated ADMA levels in patients with T2DM [22, 23], which are potentially associated with diabetes-related complications [24] and increased cardiovascular risk and mortality [25, 26]. Therefore, reducing ADMA levels are important for patients with T2DM to prevent cardiovascular complications. An animal experiment study and some intervention studies have reported that the dietary sodium and potassium intakes influence the ADMA levels in animals and humans. In hypertensive Dahl salt-sensitive rats, a high salt intake increased ADMA production accompanied by decreased NO production, increased blood pressure, and a significant increase in urinary excretion of ADMA [27]. In other studies, salt loading has led to increased plasma ADMA levels in normotensive salt-sensitive people [28], postmenopausal women [29], patients with essential hypertension [30], and patients with end-stage renal disease [31]. Conversely, salt restriction and dietary potassium supplementation reduced the plasma ADMA levels [28, 30, 31]. However, the effect of obesity on the relationship between the urinary sodium-to-potassium ratio and the serum ADMA level has not been defined in patients with T2DM.

In this cross-sectional study, we investigated the relationship between 24-h urinary sodium-to-potassium ratios and serum ADMA levels in Japanese T2D patients with or without obesity to clarify the impact of obesity on the association between the dietary sodium–potassium balance and cardiovascular risk in T2DM patients.

Subjects and methods

Study design and patients

The Ethics Committees of Mukogawa Women’s University and Ikeda Hospital approved the study protocol for this cross-sectional study. The number of patients with diabetes who were hospitalized for the first time for diabetes education in the hospital during the study period between June 2015 and August 2016 determined the sample size. A total of 344 patients were recruited during this period. All patients provided written informed consent before participation. Of these patients, we included 322 patients aged 30–85 years with a confirmed diagnosis of T2DM. However, 41 patients who required insulin treatment were excluded because previous reports suggested a possible stage-dependent association of ADMA with diabetes complications [32]. We also excluded patients with a current or past history of cancer (n = 18), who were receiving medication with steroidal (n = 2) or hormonal (n = 2) agents, who were shown by a dietary assessment questionnaire to have a habitual energy intake <800 kcal/day (n = 3), or who had missing data (n = 1). In addition, four subjects with a high daily urine volume (>4000 ml/day) and one subject with an outlier urinary sodium-to-potassium ratio (>11) were excluded because the dietary sodium and potassium levels were estimated from urinary excretions. Seven hemolyzed serum samples for measurement of the ADMA levels were also eliminated due to erythrocyte ADMA contamination. After these exclusions, 243 subjects were analyzed in this study.

Clinical assessments and laboratory measurements

Anthropometric data were assessed at the time of hospital admission. The body composition was measured using an InBody 770 body composition analyzer (BioSpace Inc., Cerritos, CA, USA). The body mass index (BMI) was calculated as weight (kg) divided by height (m) squared. The brachial–ankle pulse wave velocity (baPWV) was determined using a volume-plethysmography apparatus (Form PWV/ABI, Omron Colin Co., Tokyo, Japan). The carotid artery intima–media thickness (IMT) was measured using an Aloka SSD-Alpha 10 Diagnostic Ultrasound System (Hitachi Aloka Medical Ltd., Tokyo, Japan), and the maximum and mean values were calculated (max-IMT and mean-IMT, respectively). Blood samples were drawn during a routine assessment in the morning after overnight fasting. Serum biochemical data, high-sensitivity C-reactive protein (hsCRP), and tumor necrosis factor (TNF)-α were evaluated using standard laboratory methods (FALCO Biosystems Ltd., Kyoto, Japan, LSI Medience Corp., Tokyo, Japan). Glycated hemoglobin (HbA1c) was measured in the hospital using high-performance liquid chromatography (ADAMS A1C HA8180, Arkray Inc., Kyoto, Japan). The serum ADMA levels were measured in our laboratory using a competitive enzyme-linked immunosorbent assay, as previously described [33].

Urine collection

Collection of the subjects’ urine started immediately after admission to the hospital and lasted for 24 h. The urinary sodium and potassium concentrations were measured by the electrode method (LSI Medience Corporation, Tokyo, Japan), and the 24-h urinary sodium-to-potassium ratio was calculated using the following formula: sodium-to-potassium ratio (mmol/mmol) = urinary sodium concentration (mmol/l)/urinary potassium concentration (mmol/l).

Assessment of dietary intake

The dietary intake was assessed using a brief self-administered questionnaire (BDHQ) that was developed for the general Japanese adult population [34, 35]. The BDHQ assesses dietary habits during the previous month. The questionnaire comprises four pages with 80 questions about general eating patterns and food preparation, the frequency and amount of intake of five alcoholic beverages, and the frequency of consumption of 50 selected foods and nonalcoholic beverages. Food groups were calculated to energy-adjusted values using the density method (the amount of each nutrient or food/1000 kcal of daily energy intake) as described by Kobayashi et al. [35].

Statistical analysis

We divided the subjects into two groups according to their BMIs (<25 and ≥25 kg/m2). For the clinical characteristics, quantitative variables are expressed as the mean ± standard deviation except for the duration of diabetes. The duration of diabetes is shown as the median (interquartile range) because of its non-normal distribution. Comparisons between the non-obese (BMI <25) and obese (BMI ≥25) groups were performed using the unpaired t-test or the Mann–Whitney U test as appropriate. Comparisons of categorical variables were performed using the χ2-test. Age- and sex-adjusted and multivariable-adjusted correlations between the urinary sodium-to-potassium ratios and serum ADMA levels were analyzed by multivariable linear regression analysis. Using a logistic regression model, odds ratios (ORs) with 95% confidence intervals (CIs) for the cardiovascular disease risk were estimated for each serum ADMA-level quartile group. We set 17.0 m/s as the cutoff value of baPWV, indicating a high risk of cardiovascular disease [36,37,38]. Carotid atherosclerosis was defined as max-IMT >1.1 mm [39]. The baPWV and IMT tests were performed on 130 and 129 patients, respectively. Univariate and multivariate correlations between food groups and the urinary sodium-to-potassium ratios were calculated using Pearson’s correlation analysis and multivariable linear regression analysis, respectively. All statistical analyses were performed using IBM SPSS Statistics 22.0 (IBM Corp., Armonk, NY, USA). Two-tailed P values <0.05 were considered statistically significant.

Results

Profiles of the subjects

The clinical characteristics of the study subjects categorized according to non-obesity or obesity are shown in Table 1. Obesity was present in approximately half of the study population (the prevalence of obesity was 50.2%). Within comparisons with the non-obese patients, the obese patients were younger and had a shorter diabetes duration and a higher prevalence of dyslipidemia. Lower HDL-cholesterol and higher triglyceride levels were found in the obese patients than in the non-obese patients. Regarding renal functions, the eGFR levels in the obese patients were lower than those in the non-obese patients, whereas the albumin excretion levels were not different between the obese and non-obese patients. The serum hsCRP levels in the obese patients were significantly higher than those in the non-obese patients. No significant differences in the urinary sodium-to-potassium ratios or serum ADMA levels were found between the two groups (P = 0.194 and P = 0.763, respectively).

Table 1 Clinical characteristics of the study patient (n = 243)
Full size table

Urinary sodium-to-potassium ratios positively correlated with serum ADMA levels in non-obese patients with T2DM

The associations between the urinary sodium-to-potassium ratios and the serum ADMA levels were examined using multivariable linear regression analysis (Table 2). In the overall study population, the urinary sodium-to-potassium ratios tended to be associated with the serum ADMA levels in the age- and sex-adjusted model (β = 0.123, P = 0.053), although their association was attenuated after further adjustments for relevant variates. Positive correlations between the urinary sodium-to-potassium ratios and the serum ADMA levels were only observed in the non-obese patients (β = 0.265, P = 0.003; age- and sex-adjusted model), and not in the obese patients. These correlations were also observed in the fully adjusted model.

Table 2 Multivariate regression analysis of serum ADMA level
Full size table

Furthermore, we performed a stratified analysis to assess the impact of hypertension on the association between the urinary sodium-to-potassium ratio and the serum ADMA levels (Table 2). The results showed a positive correlation between the urinary sodium-to-potassium ratio and serum ADMA levels in non-obese patients with hypertension (age- and sex-adjusted β = 0.357, P = 0.019, and n = 42) but not in the non-obese patients without hypertension (age- and sex-adjusted β = 0.189, P = 0.091, and n = 79). However, these correlations were not observed in the fully adjusted model. No association between the urinary sodium-to-potassium ratio and the serum ADMA level was observed in the obese patients with or without hypertension.

Elevation of the serum ADMA level correlated with the degree of atherosclerotic progression

Next, we estimated whether the serum ADMA level was related to the degree of atherosclerosis in the subjects of this study. As shown in Table 3, the baPWV measurement showed that the ORs for increased baPWV (>17.0 m/s) for subjects in the third and highest quartiles of the serum ADMA levels were significantly higher than that for subjects in the lowest quartile in the unadjusted model (third quartile: OR, 4.66; 95% CI, 1.44−15.10; fourth quartile: OR, 7.23; 95% CI, 2.23−23.44) and in the multivariate-adjustment model (third quartile: OR, 4.11; 95% CI, 0.86−19.57; fourth quartile: OR, 7.89; 95% CI, 1.71−36.42). IMT measurement showed that the OR for increased max-IMT (>1.1 mm) for subjects in the highest quartile of the serum ADMA levels was higher than those in the lowest quartile in both the unadjusted and multivariate-adjusted models (unadjusted model: OR, 3.55; 95% CI, 1.21−9.30; multivariate model: OR, 3.49; 95% CI, 1.02−11.95). These results indicate that elevation of serum ADMA levels may be associated with the degree of atherosclerotic progression.

Table 3 Odds ratios for atherosclerosis index according to quartile of serum ADMA level
Full size table

Associations between food intake and the urinary sodium-to-potassium ratio

Next, we examined whether food intake was associated with the urinary sodium-to-potassium ratio in the overall patients (Table 4). In the multivariate analyses, higher intakes of noodles (β = 0.160, P = 0.017), processed meat (β = 0.144, P = 0.034), seasoning, and spices (β = 0.192, P = 0.003) were related to higher urinary sodium-to-potassium ratios. Rice and non-processed meat were not associated with the urinary sodium-to-potassium ratios. Significant inverse associations were observed between the urinary sodium-to-potassium ratios and the consumption of dairy products (β = −0.152, P = 0.028) and green and yellow vegetables (β = −0.143, P = 0.039) in the fully adjusted analyses.

Table 4 Association between food group intake and urinary sodium-to-potassium ratio in type 2 diabetes patients (n = 243)
Full size table

Discussion

In this study, we have demonstrated for the first time that serum levels of the endogenous NO synthase inhibitor ADMA are positively associated with the 24-h urinary sodium-to-potassium ratio only in non-obese patients with T2DM. High serum ADMA levels were associated with atherosclerotic progression in our study patients. These results indicate that sodium−potassium imbalance increases the serum ADMA level and cardiovascular risk in T2DM patients.

The association between the 24-h urinary sodium-to-potassium ratios and the serum ADMA levels was observed only in non-obese patients with T2DM and not in obese patients with T2DM. Obesity could enhance salt sensitivity [40, 41]. Although salt loading was reported to increase serum ADMA levels in several human studies, a reduction of the salt intake did not decrease the serum ADMA levels in normotensive overweight and obese subjects [42]. Borgeraas et al. also showed that plasma ADMA levels were associated with cardiovascular events and death in a low-BMI group but not in a high-BMI group [43]. These results suggest that obesity can blunt the association between salt intake and serum ADMA levels in patients with T2DM.

Previously, a high salt intake has been shown to reduce vascular NO bioavailability by stimulating reactive oxygen species (ROS) generation [44, 45]. Because ROS can strongly inhibit the activity of dimethylarginine dimethylaminohydrase-1 (DDAH-1), which is an ADMA-degrading enzyme, ROS generation may be involved as a possible mechanism for the association between sodium intake and the serum ADMA levels. Previous studies have shown that fat accumulation correlates with systemic oxidative stress in both humans and mice [46], and oxidative markers have been positively correlated with obesity in human clinical studies [47, 48]. In our study, the serum hsCRP levels were significantly higher in the obese patients than in the non-obese patients (Table 1). Thus, upregulated ROS generation by accumulated fat could have interfered with the association between salt intake and the serum ADMA levels in the obese patients in the present study. Contrary to our expectation, no difference was found in the serum ADMA level between non-obese and obese patients (Table 1). Because the obese patients were younger than the non-obese patients (Table 1), the age- and sex-adjusted associations between the BMI and serum ADMA levels were examined using multivariable linear regression analysis. In this analysis, the BMI was positively correlated with the serum ADMA level (β = 0.149, P = 0.034) in T2DM patients, suggesting that the difference between serum ADMA levels of the non-obese and obese patients was masked by other cofounders such as age and sex. Obesity was an exacerbating factor for the serum ADMA levels in the T2DM patients.

Hypertension is apparently frequent in patients with T2DM [49]. T2DM patients with hypertension have a higher risk for cardiovascular diseases than those without hypertension [50,51,52]. Thus, we examined the impact of hypertension on the associations between the urinary sodium-to-potassium ratios and the serum ADMA levels in the non-obese and obese patients. The urinary sodium-to-potassium ratios were related to the serum ADMA levels only in the non-obese and hypertensive patients in the age- and sex-adjusted model, suggesting that hypertension was partially involved in the relationship between the sodium-to-potassium ratios and the serum ADMA levels.

Importantly, the serum ADMA levels were significantly correlated with the degree of atherosclerotic progression, which was assessed by measurement of baPWV and IMT. The patients in the highest quartile of the serum ADMA levels (≥0.53 μM) had a significantly increased prevalence of a high risk for cardiovascular diseases (defined by a baPWV of >17.0 m/s and a max-IMT >1.1 mm) compared to those in the lowest quartile (<0.40 μM), indicating that an increase in the serum ADMA level of only 30% was significantly associated with the degree of atherosclerotic progression in this study. Other studies also showed that a slight increase in the ADMA level induced cardiovascular diseases in patients with T2DM. The plasma ADMA levels were significantly higher in the patients with T2DM who developed cardiovascular complications than in those without cardiovascular complications (0.56 μM vs 0.45 μM) [26]. In a prospective cohort, the highest tertile group of the plasma ADMA levels (mean: 0.56 μM) was associated with a higher risk of long-term cardiovascular events than the lowest tertile group (mean: 0.37 μM) [53]. Taken together, these results suggest that even a modest gain in the circulating ADMA levels can indicate an increased risk for the development of cardiovascular complications in patients with T2DM. Therefore, reducing the serum ADMA levels by adjusting the sodium and potassium intakes may be important for preventing the development of cardiovascular complications in patients with T2DM.

Our results suggest that management of the dietary sodium–potassium balance may be an effective strategy to decrease the serum ADMA levels in patients with T2DM. Thus, we analyzed the major dietary source contributing to the urinary sodium-to-potassium ratio in patients with T2DM. Assessment of dietary intake by the BDHQ suggested that the combination of reducing the intake of salt-rich foods such as noodles, processed meats, seasoning, and spices, and recommending the intake of vegetables and dairy products may be important for adjusting the urinary sodium-to-potassium ratio in patients with T2DM. Notably, this combination is very similar to the Dietary Approaches to Stop Hypertension diet [54]. Dairy products are known to be high in calcium, magnesium, vitamin D, and specific fatty acids, and previous studies have shown inverse associations between some fatty acid biomarkers of dairy consumption and the risk of insulin resistance and diabetes [55]. Evidence suggests that calcium can downregulate the renin–angiotensin system and facilitates a more favorable sodium–potassium balance, which may explain its potential mechanism for reducing blood pressure [56, 57]. Although high consumption of green and yellow vegetables and dairy products was related to a lower urinary sodium-to-potassium ratio and high vegetable consumption has been reported to be associated with lower plasma ADMA levels in a healthy middle-aged population [58], direct associations between vegetable and dairy product intake and serum ADMA levels were not observed in our cross-sectional study. An intervention study is needed to reveal the effects of vegetable and dairy product intakes on reducing serum ADMA levels in patients with T2DM.

This study had several limitations. First, caution must be exercised in the interpretation of the observed associations due to the cross-sectional design. These results should be verified by a future prospective study with interventions that include salt restriction and potassium supplementation. Second, the study was also limited by the sample size. In particular, the number of cases that underwent the baPWV and IMT tests was approximately half of the total. Third, the urinary sodium-to-potassium ratios were calculated with a single 24-h urine collection. Therefore, we could not take into account day-to-day variations in the urinary sodium-to-potassium ratios, and thus the urinary sodium-to-potassium ratios might have been partly reflections of the hospital diet. However, reasonable associations were found between the urinary sodium-to-potassium ratios and dietary habits assessed by the semiquantitative questionnaire. Fourth, the serum ADMA levels were measured a novel ELISA method. However, the ADMA values in our study were similar to those in other studies [26, 53].

In conclusion, to the best of our knowledge, this study is the first to show a positive correlation between the serum levels of ADMA, which is a risk factor for the progression of atherosclerosis, and the urinary sodium-to-potassium ratios in non-obese patients, but not in obese patients, with T2DM. In addition, a combination of restricting the intake of salt-rich foods with the consumption of green and yellow vegetables and dairy products is important for control of the sodium-to-potassium ratio which may reduce the serum ADMA levels and cardiovascular risk in Japanese patients with T2DM.