Association of Abnormal Renal Profiles with Subretinal Fluid in Diabetic Macular Edema

Purpose To investigate the risk factors for subretinal fluid (SRF) in diabetic macular edema (DME). Methods We reviewed the records of 66 patients with DME. Systemic parameters, including hypertension, glycosylated hemoglobin, serum fasting glucose, total cholesterol, triglycerides, and diabetic kidney disease, were evaluated. Renal parameters for diabetic kidney disease included serum albumin, serum creatinine, albuminuria, and estimated glomerular filtration rate. Ocular factors included visual acuity and diabetic retinopathy, and the following parameters are evaluated through optical coherence tomography examination: disorganization of the retinal inner layers, loss of ellipsoid zone, central subretinal fluid thickness, central macular thickness, and presence of SRF. Results Higher albuminuria (odds ratio, 3.431; 95% confidence interval, 1.039–11.334; P=0.043) was associated with the presence of SRF in patients with DME. Lower serum albumin levels (beta = −14.028, se = 6.646, P=0.044) were associated with increased SRF thickness. Conclusions Poor kidney function was associated with the presence of SRF in DME. Screening for SRF in DME in patients with higher albuminuria and lower serum albumin levels should be routinely performed.


Introduction
Diabetic macular edema (DME) is a major cause of visual impairment in patients with diabetes [1] and can be classifed into cystoid, difuse, serous, or mixed types. DME occurs because of the disruption of the bloodretinal barrier. Diabetic kidney disease and DME may share a similar microvascular pathophysiology [2]. Serum albumin, creatinine, albuminuria, and estimated glomerular fltration rate (eGFR) have been used as markers of diabetic kidney disease. Previous studies have shown that albuminuria and macroalbuminuria are closely associated with DME [2][3][4]. However, eGFR does not seem to have any efect on the severity or pattern of DME [5,6].
Subretinal fuid (SRF) is found in 15%-30% of DME cases [2]. Persistent SRF may be detrimental to the retinal pigment epithelium and photoreceptors [7]. Few studies have reported the association between SRF in DME and diabetic kidney disease. Koo et al. [8] found a signifcantly higher frequency of SRF in DME in patients with albuminuria. In this study, we investigated the association between diabetic kidney disease and the presence and severity of SRF in patients with diabetes.

Materials and Methods
Tis study adhered to the tenets of the Declaration of Helsinki. Ethical approval for this study was waived by the Local Ethics Committee of Eye Hospital, China Academy of Chinese Medical Sciences in view of the retrospective nature of the study. glycosylated hemoglobin (HbA 1c ), triglycerides, total cholesterol, serum albumin, serum creatinine, urinary albumin creatinine ratio, eGFR, and diabetic retinopathy (DR) severity, were recorded. Disorganization of the retinal inner layers, loss of ellipsoid zone, central SRF thickness, central macular thickness (CMT), and presence of SRF were documented using optical coherence tomography. Disorganization of the retinal inner layers was defned as the inability to distinguish the boundaries between the ganglion cell-inner plexiform layer complex, inner nuclear layer, and outer plexiform layer [9]. DR severity was categorized as mild nonproliferative retinopathy (mild NPDR), moderate NPDR, severe NPDR, and proliferative diabetic retinopathy. Overnight frst-void urine samples were collected. Albuminuria was defned as a urinary albumin-to-creatinine ratio >30 mg/g. eGFR was calculated by using the chronic kidney disease epidemiology collaboration equation and was categorized into the following fve groups: <30 mL/min/1.73 m 2 (stage 1), 30-44 mL/min/1.73 m 2 (stage 2), 45-59 mL/min/1.73 m 2 (stage 3), 60-89 mL/min/1.73 m 2 (stage 4), or >90 mL/ min/1.73 m 2 (stage 5). SRF, a categorical variable, was divided into two categories: the presence and absence of SRF. Hypertension was analyzed as a categorical variable, with patients divided into two categories according to the presence or absence of hypertension. Data from the worst eye were used for analyses.

Statistical Analyses.
Statistical analyses were performed using SPSS Statistics for Windows, version 26.0 (IBM, Armonk, NY). Summary statistics included the mean-± standard deviation, where appropriate. Binary and multiple logistic regression models were used to assess the risk factors for SRF. Univariate and multivariable linear regressions were used to evaluate the correlation between central SRF thickness and clinical parameters. Univariate linear regression was used to evaluate the correlation between CMT and clinical parameters. P values <0.05 were considered statistically signifcant.

Results
A total of 66 patients with DME were enrolled in this study. Te baseline characteristics of the patients are shown in Table 1. None of the patients had undergone dialysis. Of the total patients, 39 (59%) were women. Te mean age was 58.2 ± 11.2 years, the mean duration of diabetes was 10.6 ± 6.9 years, the mean HbA 1c level was 8.0 ± 1.7%, and the mean CMT was 523.2 ± 154 μm. SRF in DME was observed in 30 patients. Table 2 shows the factors associated with the presence of SRF. Binary logistic regression analysis revealed that lower HbA 1c levels (odds ratio [OR], 0.727; 95% confdence interval [CI], 0.529-0.999; P � 0.049) and higher albuminuria (OR, 3.571; 95% CI, 1.112-11.468; P � 0.032) were associated with the presence of SRF among the 66 patients with DME. Multiple regression analysis revealed that only albuminuria was signifcantly associated (OR, 3.431; 95% CI, 1.039-11.334; P � 0.043) with the presence of SRF. Visual acuity, DR severity, hypertension, serum fasting glucose, triglycerides, total cholesterol, eGFR, serum albumin, and serum creatinine showed no signifcant association with the presence of SRF. Table 3 shows the factors associated with SRF thickness according to the univariate linear regression among the 30 patients with SRF. Lower serum albumin levels (beta � −14.028, se � 6.646, P � 0.044) were associated with a greater SRF thickness. Table 4 shows the factors associated with CMT. Univariate linear regression revealed that HbA 1c (beta � −42.17, se � 10.34, P � 0.0001) and serum fasting glucose (beta � −13.24, se � 5.55, P � 0.020) were associated with CMT among the 66 patients with DME. Multivariable linear regression revealed that a lower HbA 1c was significantly associated with a higher CMT (beta � −42.06, se � 13.28, P � 0.002). However, albuminuria, eGFR, serum albumin, and serum creatinine were not associated with CMT.

Discussion
Our study results suggest that diabetic kidney disease plays an important role in the occurrence of SRF in DME. Higher albuminuria showed a better association with the presence of SRF than lower HbA 1c levels, although both were important risk factors for the presence of SRF.
Our data suggested that the presence of SRF had a good correlation with higher albuminuria, while serum albumin was not signifcantly diferent between the patients with or without SRF. In addition, we found that the severity of SRF was negatively correlated with serum albumin levels. Low serum albumin levels were associated with increased SRF thickness. Tsai et al. [7] reported that the presence of SRF was correlated with low serum albumin levels. Tis may be because the correlation of baseline SRF with serum albumin levels was performed to adjust for age and DR severity. Koo et al. [8] also reported a signifcantly higher frequency of SRF in DME in patients with albuminuria. Vascular hyperpermeability is a possible shared pathogenetic mechanism between the kidney (albuminuria) and the eye (SRF of DME). Fluid movement is governed by changes in pressure gradients. Advanced proteinuria with marked protein loss may result in lower intravascular osmotic pressure and higher hydrostatic pressure, leading to fuid retention in the subretinal space. In the early stages, serum albumin defcits may be compensated for by the increased production of albumin molecules in the liver, which may prevent hypoalbuminemia [10]. In the late stage, serum albumin defcits that could not be compensated for may be associated with increased SRF thickness. Additionally, according to our results, no associations could be found between albuminuria and CMT, which is similar to the fndings of previous studies [11][12][13]. We inferred that patients with albuminuria may be prone to more fuid leakage from choroidal vessels and damaged retinal pigment epithelium, which eventually results in SRF in DME. Moreover, previously reported case series on SRF secondary to nephrotic syndrome demonstrated that SRF resolved following systemic furosemide treatment [14][15][16].
Contrary to the general belief [17], we observed that lower HbA 1c levels were more likely to result in SRF and were associated with a higher CMT. Data from a more recent study also indicated that lower HbA 1c levels are more likely to result in SRF in DME [11]. Other supportive evidence showed that better control of HbA 1c did not lead to a greater reduction in CMT in DME [18]. Furthermore, Falavarjani et al. [19] observed reduced CMT after meals compared with that before meals. Tere are several possible explanations for this fnding. First, it may be because of the early worsening efect, which can be caused by the rapid reduction in HbA 1c levels [11]. Second, lowering serum glucose levels can lower  In the present study, no associations were found between eGFR and the severity spectrum of DME or the presence of SRF. Previous studies have also shown that eGFR does not seem to have any efect on the severity or pattern of DME [5,6]. However, Man et al. [2] and Temkar et al. [5] found that lower eGFR was not associated with the presence or severity of DME. Additionally, we did not fnd any significant correlation between DR severity and the presence of SRF. Tis may be because retinopathy and nephropathy may have a similar pathology, and eGFR and DR may fuctuate over time. However, further research is required to clarify this.
Our study has some limitations. It was a retrospective study with a small sample size.

Conclusions
According to our study results, higher albuminuria was associated with the presence of SRF in patients with DME. Lower serum albumin levels were associated with greater SRF thickness. Hence, screening for SRF in DME in all patients with diabetic kidney disease should be emphasized.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.