Impact of extracellular-to-intracellular fluid volume ratio on albuminuria in patients with type 2 diabetes: A cross-sectional and retrospective cohort study


 Background Body fluid volume imbalance is common in patients with kidney failure and is associated with all-cause mortality. We hypothesized that fluid volume imbalance occurs in diabetic patients without kidney failure and is associated with change in albuminuria. This study aimed to investigate the association between fluid volume imbalance and albuminuria in patients with type 2 diabetic mellitus without kidney failure. Methods We included 432 and 368 participants in the cross-sectional and retrospective studies. Body fluid imbalance was determined by measuring the extracellular water (ECW) to intracellular water (ICW) ratio (ECW/ICW) using bioelectrical impedance analysis. Change in urinary albumin to creatinine ratio (ACR) was defined as the ratio of urinary ACR at follow-up to that at baseline. ECW/ICW ratio was compared to level of albuminuria. Results In this cross-sectional study, ECW/ICW ratio increased with level of albuminuria (0.634 ± 0.002 in normo-, 0.645 ± 0.003 in micro- and 0.656 ± 0.004 in macroalbuminuria, all p <0.001 by Tukey's HSD test). In addition, there was an association between ECW/ICW ratio and logarithms urinary ACR after adjusting for covariates (β = 0.217, p <0.001). Moreover, ECW/ICW ratio was associated with change in the urinary ACR after adjusting for covariates (β = 0.171, p = 0.005) in this retrospective study. Conclusions ECW/ICW ratio is independently associated with level of albuminuria in patients with type 2 diabetes mellitus without kidney failure. This reinforces the importance of monitoring fluid balance in patients with type 2 diabetes mellitus.

Albuminuria, which indicates microvascular endothelial injury, is an independent risk factor for progression to ESRD, particularly in patients with DKD (3,8). Thus, early detection of DKD, especially worsening of albuminuria, is important in management of patients with diabetes mellitus.
In healthy adults, fluid distribution between the intracellular water (ICW) and extracellular water (ECW) compartments is tightly regulated (9). This regulation is however impaired in patients with ESRD and hemodialysis, resulting in fluid overload (10,11). One of the mechanism of fluid overload is thought to be inflammation, which causes hypoalbuminemia and increased vascular permeability (12). The result is extravascular fluid shift which leads to ECF volume overload (13). We hypothesized that subtle fluid imbalance occurs in patients with diabetes mellitus without kidney failure, and that there is an association between fluid imbalance and albuminuria. No studies have reported on this association. In this cross-sectional and retrospective cohort studies, we therefore investigated the association between ECW/ICW ratio and albuminuria in patients with type 2 diabetes mellitus without kidney failure, and the effect of ECW/ICW ratio on the level of albuminuria, using bioelectrical impedance analysis (BIA) which is used to evaluate body composition and fluid compartments (14,15].

Study population
The present study was a sub-analysis of the KAMOGAWA-DM cohort study, the details of which have been described elsewhere (16). Briefly, the KAMOGAWA-DM cohort Study is an ongoing cohort study of subjects at Kyoto Prefectural University of Medicine (Kyoto, Japan) and Kameoka Municipal Hospital (Kameoka, Japan). The purpose of this cohort study is to clarify the natural history of people with diabetes. In the cross-sectional study, we investigated the relationship between ECW/ICW ratio and the prevalence of diabetic nephropathy, and in the retrospective study, we investigated the association between ECW/ICW ratio and change in ACR.
We extracted data of patients with type 2 diabetes mellitus who had body composition analysis, and measured urinary albumin to creatinine ratio (ACR) during the years from 2014 to 2017 from the KAMOGAWA-DM cohort study. The exclusion criteria of the crosssectional study were as follows: missing data of covariates (serum creatinine level and duration of diabetes) and patients with estimated glomerular filtration rate (eGFR) < 30 ml/min/1.73 m 2 (17); and the exclusion criteria of the retrospective study were no follow-up data (including treatment interruption, transfer to another hospital, death and less than three times measurement of urinary ACR).

Variables
Body mass index (BMI), intracellular water (ICW), extracellular water (ECW), total body water (TBW), body fat mass, and skeletal muscle mass were measured with fasting state by BIA (18). ECW/ICW ratio and skeletal muscle index (SMI, kg/m 2 ) was calculated from the obtained data (19,20). Medication data were also collected; medication for diabetes, including sodium-glucose cotransporter 2 (SGLT-2) inhibitors; and medication for hypertension, including renin angiotensin aldosterone (RAS) inhibitor and diuretics. The smoking status was categorized into three groups: never-, ex-, and current smoker.
Urinary albumin and creatinine concentrations were measured using early morning spot urine samples. In this study, a mean value for urinary ACR, which was determined from three urine collections, were used for analyses. According to the Joint Committee on Diabetic Nephropathy, we divided the subjects into three groups; normo-micro-and macro-albuminuria (17). Follow-up examinations were performed one year later, we also collected urine samples for calculation of urinary ACR three times a year. Change in urinary ACR was calculated as follows: dividing the follow up urinary ACR by the baseline urinary ACR (23).

Ethical considerations
This study was approved by the ethics committee of Kyoto Prefectural University of Medicine (Approval number RBMR-E-466-5), and undertaken in accordance with the Declaration of Helsinki. Written informed consent was obtained from all study participants.
To protect the confidentiality of participants, personal identifiable information was removed and medical data stored in a database which was password protected.

Statistical analysis
Statistical analyses were performed using JMP ver. 13.2 software (SAS, Cary, NC). A pvalue < 0.05 was considered significant. For normally distributed continuous variables, data was summarized using mean and standard deviation. Continuous variables with a skewed distribution were summarized using median and inter-quartile range. Categorical variables were described using proportions. Differences between the groups were analyzed as follows: the baseline clinical characteristics of the groups were compared using Pearson's chi-squared test or Fisher's Exact test as appropriate. For normally distributed continuous variables, we compared mean difference between groups using the one-way Analysis of Variance (ANOVA) and Tukey's honestly significant difference (HSD) test.
Because ACR had a skewed distribution, logarithmic transformation was undertaken before correlation and multiple logistic regression analyses. Variables found to be statistically different in bivariate analysis were controlled for in multiple regression analysis. We investigated the relationships between ECW/ICW ratio and logarithmic of ACR or other factors using Pearson's correlation coefficient. Multiple regression analysis for logarithmic of urinary ACR was undertaken.
Furthermore, we also investigated the effect of ECW/ICW ratio on change of urinary ACR by multiple regression analysis. We considered several potential confounders as covariants: age, sex, BMI, HbA1c, creatine, triglycerides, duration of diabetes, smoking status, exercise, and usage of RAS inhibitor and SGLT-2 inhibitor, diuretics, ECW/ICW and logarithmic of urinary ACR at baseline examination.

Results
The inclusion of participants is summarized in Fig. 1. Out of the 481 (261 men and 220 women) participants eligible for the study, 49 (30 men and 19 women) were excluded due to missing data on serum creatinine and duration of diabetes (Fig. 1).
In the retrospective study, out of the 432 people (231 men and 201 women) eligible for the study, 64 (32 men and 32 women) were excluded, resulting in a study population of 368 people (199 men and 169 women) (Fig. 1). Table 4 summarizes the characteristics of study subjects of retrospective study. Table 5 shows the results of the multiple regression analysis with change in the urinary ACR.

Discussion
We investigated the association of fluid volume imbalance and albuminuria in patients with type 2 diabetes without ESRD based on our hypothesis that fluid imbalance occurred in diabetic patients without ESRD, and was associated with changes in albuminuria. Our study's findings support the hypothesis.
Previous studies have demonstrated the association of fluid overload and increased risk of eGFR decline and all-cause or cardiovascular mortality in patients with ESRD and patients on dialysis (10,11,19,24,25,26,27,28). Our study's mean ECW/TBW ratio of 0.390 ± 0.01 was lower than that reported in recent studies of patients with CKD stage 4 or 5 (0.39783-0.512) (26,29,30). This finding suggests that fluid imbalance is less likely to occur in patients with early nephropathy than in patients with ESRD and hemodialysis.
Water shift from ICW to ECW led to change the ECW/ICW ratio (9). Cell volume is regulated by apoptosis, which is a morphological hallmark of programmed cell death (31). The loss in cell volume during apoptosis may play a role in change in balance between ICW and ECW content. In addition, uremic status may also cause cell shrinkage. Previous studies have reported that erythrocytes may undergo suicidal death or eryptosis associated with cell shrinkage, which can be stimulated by uremic toxins (32).
Albuminuria is known to reflect endothelial dysfunction and subclinical inflammation caused by oxidative stress and inflammatory cytokines (8,12,33). Kidney endothelial dysfunction plays an important role in the development of albuminuria by reducing vascular relaxation and inflammatory cell infiltration (34). Under physiological conditions, tubule-glomerular feedback (TGF) signaling maintains stable glomerular filtration rate (GFR) by modulating pre-glomerular arteriole tone. Early in nephropathy, chronic hyperglycemic conditions impair SGLT-2 mediated reabsorption of sodium and glucose in the proximal tubule. Thus, despite increased GFR, the macula densa is exposed to low sodium concentrations. This impairment of TGF signaling likely leads to inadequate arteriole tone and increased renal perfusion. As a result, impairment of TGF cause increased body fluid and fluid imbalance (35,36,37). Previous study show that both human and animals with volume overload have significantly higher pro-inflammatory cytokines such as IL-6 or TNF-α (38), which can be the result of kidney endothelial dysfunction and impairment of TGF. Further, inflammation-induced hypoalbuminemia and increased vascular permeability enhance extravascular fluid shift, thereby resulting in ECF volume overload (13). In fact, the ECW/TBW ratio, which is substantially the same as the ECW/ICW ratio, has been used as a marker of ECW excess (8,39,40).
Increase in ECW/ICW ratio caused by fluid overload effects vascular and endothelial level by oxidative stress, chronic activation of the renin-angiotensin system, sympathetic activation and an increase of inflammation, which leads to atherosclerosis (41). The excess volume status would increase renal efferent pressure and cause glomerular hypertension and eventual decline in eGFR (25   Inclusion and exclusion flow chart of participants.