Fasting Single-Spot Urine pH Is Associated with Metabolic Syndrome in the Japanese Population

• In this study, a significant, inverse relationship between urine pH and metabolic syndrome and the number of its components were identified after controlling for confounding factors, including serum uric acid levels. The examination of urine pH could be a practical screening tool for metabolic syndrome.

Metabolic syndrome (MetS) is associated with a variety of physiological and metabolic alterations, and is clinically recognized by numerous constitutive traits, including abdominal obesity, hypertension, dyslipidemia, and hyperglycemia [1]. This cluster of features is strongly associated with type 2 diabetes, cardiovascular disease, and increased cardiovascular and all-cause mortality [1].

Recently, a relationship between low urine pH (pH <5.5) and diabetes [2] and insulin resistance [3] had been reported. Obesity, another feature of MetS, is also associated with a low urine pH [4]. In addition, non-high-density lipoprotein cholesterol (HDL-C) could have a significant association with a lower urinary pH [3]. In a recent cohort study, it was proposed that low urine pH is a causative and predictive factor for the development of MetS [5]. However, the relationship between urine pH and blood pressure has been investigated in only a few studies [5,6], and only a limited number of studies has considered confounding variables such as smoking habits and alcohol intake [7,8]. Epidemiological studies have revealed a close relationship between serum uric acid (UA) levels and the presence of MetS [9,10]. Nevertheless, while patients with high serum UA levels were reported to have a low urine pH [5,6,11], it is unclear whether or not low urine pH relates to MetS in a manner independent of serum UA levels. Therefore, to ascertain whether or not urine pH is related to MetS and its components, a cross-sectional study was conducted in a Japanese population.

This study included 5,721 individuals who underwent routine health check-ups in Iida Municipal Hospital, Iida City, Japan, from January 2013 to December 2015. When data were available for individuals from multiple visits, only data acquired at the first visit were included. Exclusion criteria were: the potential for advanced renal dysfunction such as in those with an estimated glomerular filtration rate (eGFR) <30 mL/min/1.73 m², evidence of albuminuria as indicated by urine dipstick albumin >2+, and advanced age (>80 years). Consequently, of the 5,721 individuals, 5,430 subjects were included in the final analysis (4,691 patients without MetS and 739 patients with MetS). This study was approved by the Institutional Ethics Review Board of Iida Municipal Hospital and performed in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants.

According to the classification described by the National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III) criteria [12], MetS is defined by the presence of ≥3 components of the following 5 abnormalities: (a) waist circumference (WC) ≥90 cm and ≥80 cm in Japanese men and women, respectively; (b) systolic blood pressure (SBP) ≥130 mm Hg and/or diastolic blood pressure (DBP) ≥85 mm Hg, or treatment for previously diagnosed hypertension; (c) triglyceride (TG) levels ≥150 mg/dL or receiving specific treatment for hypertriglycemia; (d) HDL-C levels <40 mg/dL in males and <50 mg/dL in females, or receiving specific treatment for low HDL-C; (e) fasting plasma glucose (FPG) levels ≥100 mg/dL or previously diagnosed diabetes. Individuals with <3 criteria for the diagnosis of MetS served as controls.

WC was measured between the lower rib margin and iliac crest after a normal expiratory breath. SBP and DBP were measured using an automated measuring device (HEM-7080IT, Omron Co., Tokyo, Japan). Two separate measurements were taken after 15 min of resting, with an interval of at least 30 s between the 2, and the mean of these was considered.

Information regarding various comorbidities, including hypertension, dyslipidemia, and diabetes, were collected by in-person interviews. The smoking status and alcohol consumption of study participants were assessed using a standardized questionnaire. The smoking status was considered as current smoking (daily and occasional smoking) and nonsmoking (never and former smoking). Alcohol consumption was classified as current drinkers (at least once per week) and noncurrent drinkers.

Blood samples were obtained after an overnight fast. To obtain a serum sample, the blood sample was centrifuged in a plain vacutainer within 20-30 min of blood clotting. Total serum cholesterol, TG, HDL-C, low-density lipoprotein cholesterol, UA, FPG, high-sensitivity C-reactive protein (hs-CRP), and creatinine levels were measured using standard methods (Hitachi 47 automatic analyzer, Hitachi Ltd., Japan). eGFR was calculated using the formula of the Japanese Society of Nephrology according to the equation: 194× serum creatinine−^1.094 × age−^0.287 mL/min/1.73 m², further multiplied by 0.739 for female subjects [13]. An early morning midstream urine sample was collected for urine chemistry following an overnight fasting period, and was analyzed immediately using an automated urine dipstick analyzer (Aution Max AX-4030, ARKRAY Inc., Japan). Kraut and Madias [14] indicate that in the absence of impaired renal acidification, acidic urine should be defined as having a pH of ≤5.5, neutral urine a pH of 6.0-7.5, and alkaline urine a pH of ≥8.0.

Statistical analyses were performed using SPSS software v21.0 (SPSS Inc. IL, USA). Characteristics of the subjects with and without MetS were compared using the independent t test for continuous variables and the χ² test for dichotomous variables. Skewed distribution variables were analyzed after logarithmic transformation. Correlations between urine pH and variables of MetS components were analyzed using the Pearson correlation analysis, and partial correlation analysis was performed to control confounding factors (age, gender, eGFR, serum UA, hs-CRP, and smoking and drinking status). In these correlation analyses, the subjects taking antihypertensive, antihyperglycemic, and antidyslipidemic agents were excluded to avoid the effects of drugs on the variables associated with MetS components. Analysis of covariance models were used to examine the relationship between urine pH and the number of MetS abnormalities. We used multiple logistic regression analyses to adjust confounding factors to determine the adjusted odds ratios (ORs) for MetS incidence between different urine pH categories (pH <5.5, 5.5-6.0, and >6.0).

The incidence of MetS in all subjects, males, and females were 13.6, 18.3, and 7.3%, respectively. The physical and biochemical characteristics of the study subjects are summarized in Table 1. Subjects with MetS had a significantly lower urine pH (5.9 ± 0.7) than those without MetS (6.0 ± 0.7) (p < 0.001). Serum UA levels were significantly higher in subjects with MetS than in those without MetS (6.2 ± 1.4 and 5.3 ± 1.4, respectively) (p < 0.001).

Some of MetS components (SBP, DBP, TG, and FPG levels) were negatively correlated with urine pH, whereas HDL-C showed a positive correlation, even after adjusting for confounding factors, including serum UA (Table 2). However, WC was not a correlation factor with urine pH.

Mean urine pH adjusted for confounding factors is shown in Figure 1. Urine pH decreased from 6.00, 6.01, 5.97, and 5.89, to 5.81, along with an increasing number of metabolic abnormalities (i.e., with 0, 1, 2, 3, and 4 or 5 abnormalities, respectively). Subjects with ≥3 metabolic abnormalities had a significantly lower urine pH than those with ≤2 metabolic abnormalities (p_{for trend} < 0.001).

Unadjusted and adjusted ORs of the presence of MetS according to different urine pH categories are shown in Table 3. The ORs increased with a decrease in urine pH. Compared with subjects with a urine pH >6.0, the adjusted ORs (with 95% confidence intervals [CIs]) for the presence of MetS in subjects with a urine pH of 5.5-6.0 and <5.5 were 1.34 (1.04-1.73) and 1.52 (1.09-2.13), respectively.

In this study, there was a significant inverse relationship between urine pH and MetS and the number of abnormal metabolic components. In addition, urine pH was correlated with variables of MetS components. These results did not change even after adjusting for a range of confounding factors such as age, gender, eGFR, hs-CRP, UA, and smoking and drinking status.

Our findings are consistent with those of previous studies [2,3,4,5] regarding the relationship between urine pH and metabolic parameters. For example, urine pH has been found to be significantly lower in patients with type 2 diabetes than in normal volunteers [2]. In another study, a low urine pH showed an association with an increase in WC, and FPG and TG levels and a decrease in HDL-C levels [6]. Furthermore, participants with MetS had a significantly lower 24-h urine pH than those without MetS [15], and low urine pH could be a predictive factor for the development of MetS [5]. Our results and these previous studies indicate that urine pH correlates with metabolic parameters and that a low urine pH is associated with the incidence of MetS.

Although the precise reasons for the relationship between MetS and a lower urine pH remain unclear at this time, a low urine pH could result from increased acid excretion, impaired urinary buffering, or both. Ammonium is the primary buffer in the urine [16], and its renal production and excretion are regulated by the ambient acid-base environment. Insulin is known to influence both of these processes [17], and mechanisms of acid-base homeostasis may be altered in a state of insulin resistance [18]. Considering our data along with those from previous studies [2,3,4,5], we hypothesize that acidic urine may be a renal manifestation of insulin resistance resulting from MetS.

Although serum UA is not included in the diagnostic criteria for MetS, epidemiological studies have established a close relationship between elevated serum UA levels and the increasing prevalence of MetS [9,10,19]. Serum UA, a weak acid, may also affect urine pH. Indeed, previous studies have reported that subjects with higher serum UA levels have lower urine pH [5,6,11]. In this study, serum UA levels were negatively correlated with urine pH, which is consistent with the results of previous reports. In addition, the relationship between lower urine pH and MetS persisted, even after adjusting for serum UA, suggesting that high serum UA levels alone cannot account for the more acidic urine. Aging [20] and renal dysfunction [21] are also known to be related with a decrease in urine pH. Our results are consistent with the following previous findings: age had an inverse relationship with urine pH whereas eGFR had a positive relationship. In this study, individuals who exhibited potential signs of advanced renal dysfunction and/or advanced age were excluded. However, the relationship between MetS and its components with urine pH was unchanged even after adjusting for age and eGFR.

In this study, fasting single-spot urine samples, and not urine collected over 24 h, were used to determine the urine pH. Although the diurnal variation in urine pH has been confirmed, Capolongo et al. [22 ]demonstrated that fasting urine pH is significantly correlated with 24-h urine pH in a large cohort. Additionally, the measurement of urine pH using the indicator method is known to be less reliable than using electrodes; however, the indicator method might be more appropriate for routine health check-ups because of its convenience and lower cost.

The strength of the study was its relatively large, well-characterized, community-based population. However, it also has some limitations which include its cross-sectional design which did not permit the determination of causality. In addition, nutritional records were not available for the patients, and so we could not include dietary patterns in the analyses. Furthermore, the study population consisted of Japanese patients only; consequently, it is uncertain whether these findings could be generalized to other ethnic groups.

In this study, there was a relationship between MetS and low urine pH, in a manner which is independent of serum UA levels. In addition, a progressive decline in urine pH was noted with an increasing number of MetS features. Hence, the examination of fasting urine pH could be a practical screening tool for MetS, and this practice might be beneficial in recognizing the segment of a Japanese population with lower urine pH, particularly urine pH <5.5. Further studies are needed to elucidate the mechanisms responsible for the lower urine pH in individuals with MetS.

The authors gratefully acknowledge Nobuo Shimosawa for preparing an electronic database of the participants' clinical records. We also acknowledge Naoki Nakadaira, a medical technologist in the Iida Municipal Hospital, for the interpretation of the study data.

There were no conflicts of interest.