The relationship between brain-derived neurotrophic factor levels, oxidative and nitrosative stress and depressive symptoms: a study on peritoneal dialysis

Abstract

Background: Depression is one of the most commonly encountered psychiatric problems in peritoneal dialysis (PD) patients. Our aim was to investigate the associations between oxidative and nitrosative stress (O&NS) and brain-derived neurotrophic factor (BDNF) in PD patients with elevated depressive symptoms (EDS). Methods: Eighty-three patients with PD and 84 healthy controls were enrolled in this study. In PD patients, two subgroups were formed: 28 with and 55 without EDS. EDS were defined as a Beck Depression Inventory (BDI) score ≥17 in patients. Serum malondialdehyde (MDA) erythrocyte, glutathione (GSH) levels measured spectrophotometrically. Serum superoxide dismutase (SOD) activity, nitric oxide (NO) and BDNF levels were determined by ELISA. Results: While MDA and NO levels were higher, levels of SOD, GSH and BDNF were lower in PD patients compared to controls (p < 0.001). The patients with EDS had higher levels of MDA and lower levels of BDNF as compared to those without EDS (p < 0.005). In linear regression analysis, the BDNF levels were dependently associated with SOD levels in PD patients (B: 0.274, p: 0.043). In addition, while a negative correlation existed between BDI scores with BDNF levels (r = –0.312, p = 0.004), a positive correlation was present between BDI scores and MDA levels (r = 0.320, p = 0.005) in PD patients. Conclusion: Our results suggest the presence of high O&NS and low antioxidant capacity accompanied with decreased levels of BDNF in PD patients, especially those with EDS were deeper. These may represent the risk factors for cellular injury and might reveal part of the mechanism causing the depressive state in PD patients.

• Brain-derived neurotrophic factor
• chronic kidney disease
• nitrosative stress
• oxidative stress
• peritoneal dialysis

Introduction

Peritoneal dialysis (PD) is commonly used in patients with end-stage renal disease (ESRD).1 Due to the protracted course, the patients with chronic kidney disease (CKD) are vulnerable to other disorders both physically and mentally.

Depression is one of the most commonly encountered psychiatric problems in dialysis patients.2–6 The prevalence of depressive disorder in dialysis patients is reported between 33% and 50%.7 The incurability and the fatal outcome of CKD are common sources of stress. Among PD patients, such psychological stress contributes seriously to the induction of different psychological disorders, especially depression.8^,9

Brain-derived neurotrophic factor (BDNF) has been described as a protein that improves the survival of peripheral neurons. BDNF is currently known as a factor that is expressed profusely in the nervous system, with the maximum levels of mRNA and protein established in the hippocampus and frontal cortex. Moreover, BDNF regulates neuronal survival, migration, cell differentiation, growth of axons and dendrites and synapse formation.10

Oxidative stress has been documented as consequences of an imbalance between free radical production and insufficient endogenous antioxidant defense mechanisms in uremic patients. In addition to the reactive oxygen species (ROS), the reactive nitrogen species (RNS) are also known as free radicals and are generated from aerobic metabolism in biological systems.11

The associations between depression, oxidative and nitrosative stress (O&NS) and BDNF levels in CKD have not been reported extensively. Although some researchers have paid attention to depression, oxidative balance and BDNF levels individually or in pairs in CKD, we have not encountered a study that includes all of these factors.11–13 The purpose of this study is to investigate the associations between depressive symptoms, O&NS and BDNF in PD patients.

Materials and methods

The patient and control groups

Eighty-three patients [28 with and 55 without elevated depressive symptom (EDS)] and 84 healthy volunteers agreed to participate in this study. This cross-sectional study, approved by the ethics committee of the Medical Faculty of Kocaeli University, was conducted between June 2013 and November 2013. Written informed consent was obtained from each participant. Patients receiving PD for at least three months in our hospital and healthy volunteers of >18 years of age were enrolled in this study. Etiologies of ESRD in the PD patients were hypertensive nephropathy (n = 38), polycystic kidney disease (n = 7), diabetic nephropathy (n = 22), chronic glomerulonephritis (n = 3) and unknown (n = 13). The PD patients were divided into two groups using the Beck Depression Inventory (BDI) scoring system to determine those with and those without EDSs. The healthy control group was comprised of volunteers with no history of physical illness or depression. The severity of depressive symptoms in PD patients was evaluated by means of BDI. The BDI is a 21-item self-report rating inventory measuring the characteristic symptoms of depression. The 21 items are answered on a four-point Likert scale in which 0 represents the absence of a problem and 3 represents an extreme problem, with a possible total score ranging from 0 to 63. Validation of the BDI in the Turkish population was done by Hisli.14 A BDI score of ≥17 was the standard cut-off point to define depression in Turkish individuals, thus this value was used to classify patients with EDS.

PD patients were on a standard PD program (4 or 5 times daily with 2000 mL of fluid). The etiology of CKD was recorded. The medical history included questions about the presence or absence of the extrarenal co-morbidities such as hypertension, atherosclerotic vascular disease, cardiac failure, cardiac valvular diseases and chronic obstructive pulmonary disease.

Exclusion criteria for PD patients were as follows: severe illness (malignancy, severe heart or respiratory failure), infection in the last two weeks, the presence of any psychiatric illness, cognitive dysfunction, current use of an antidepressant or antipsychotic medication and the use of non-steroidal anti-inflammatory drugs or antioxidant supplements in the last four weeks. We selected healthy control subjects who did not have any medical or psychiatric diagnosis before. The healthy group consisted of volunteers older than 18 years, non-smokers and who were not on supplemental antioxidants or other drugs within the last four weeks.

Analytical procedures

After fasting overnight, between 08:00 and 10:00 in the morning, venous blood samples were drawn from the antecubital veins of all participants and collected into anticoagulant-free tubes and serum was obtained by centrifugation (3000 rpm for 15 min), immediately frozen and stored at −40 °C for the other parameters until subsequent batched for BDNF, malondialdehyde (MDA), nitric oxide (NO) and superoxide dismutase (SOD) analyses. For glutathione (GSH) analyses, the blood samples were collected in standard sterile vacuum tubes with ethylenediamine tetra-acetic acid, and erythrocytes were separated and kept at −40 °C estimation until analysis.

Serum MDA was estimated by the method which was adopted by Philpot.15 Erythrocyte GSH was determined spectrophotometrically using by Ellman & Gan’s method.16 Serum SOD activity, NO and BDNF levels were analyzed with ELISA kits by Alisei Quality System Seac Radin Company analyzer (Cayman Chemical, Ann Arbor, MI; Nitrite/Nitrate Colorimetric Assay – Cayman Chemical, Deerfield, FL and Chemikine™ – Millipore, Temecula, CA, respectively).

Statistical analysis

Data were evaluated using SPSS version 13.0 (SPSS Inc., Chicago, IL). The Kolmogorov–Smirnov test showed that the variables did not fit in normal distribution. Therefore, statistical analyses were performed using an appropriate non-parametric test. The Mann–Whitney U test and Kruskal–Wallis variance analysis with Bonferroni’s correction were used for parametric variables with non-normal distributions. Correlations between blood NO, SOD, MDA, GSH, BDNF concentrations and BDI scores were investigated by Spearman’s correlation test. Since BDNF levels were not distributed normally, logarithmic transformation was performed.

Linear regression analysis was used for BDNF as dependent variable, and the relationships were investigated between other variables. The statistical significance level was set at p < 0.05.

Based on the study of Zoladz et al., group sizes of 18 patients would achieve 95% powers to detect a significance level (α) of 0.05.13

Results

Demographic and clinical characteristics of participants are listed in Table 1. Studied biochemical markers levels of all participants are presented as mean ± SD in Table 2. O&NS parameters (MDA and NO levels) were significantly higher in the PD patients compared with the control group (p < 0.001). The MDA levels were also significantly higher in patients PD with EDS when compared those without EDS (p = 0.008). However, as a nitrosative stress indicator, the NO levels were not significantly different in PD patients with and without EDS (p = 0.347).

Table 1. Demographic characteristics of PD patients and controls.

	PD patients (n = 83)	Controls (n = 84)
Age (years)	51.20 ± 12.30 (22–65)	37.50 ± 7.00 (24–52)
Female/male	44/39	59/25
Total dialysis  duration (months)	50.31 ± 35.79 (3–156)	–

Note: Values are expressed as mean ± SD (minimum–maximum).

Table 2. Comparison of BDNF, NO, MDA, SOD and GSH levels (means ± SD) between PD patients and controls.

	PD patients (n = 83)	Controls (n = 84)	p Values
MDA (µmol/L)	0.70 ± 0.16	0.57 ± 0.08	<0.001
GSH (mg/dL)	34.47 ± 4.56	47.94 ± 7.00	<0.001
SOD (U/L)	2.59 ± 0.61	10.67 ± 2.80	<0.001
NO (µmol/L)	22.66 ± 4.43	13.91 ± 11.71	<0.001
BDNF (pg/mL)	281.32 ± 179.02	781.50 ± 230.70	<0.001

Notes: BDNF, brain-derived neurotrophic factor; NO, nitric oxide; MDA, malondialdehyde; SOD: superoxide dismutase; and GSH: glutathione.

Antioxidant levels (GSH and SOD) were significantly lower in the PD patients than controls (p < 0.001). However, antioxidant levels (GSH and SOD) were not significantly different between PD patients with and without EDS (p = 0.529, p = 0.526, respectively).

The BDNF levels were significantly lower in PD patients when compared to controls (p < 0.001).The BDNF levels were also significantly lower in patients PD with EDS when compared those without EDS (p < 0.001). Table 2 lists the descriptive statistics and comparisons for the studied markers on PD patients and controls, and Table 3 lists PD patients with and without EDS by BDI scores. While a negative correlation existed between BDI scores with BDNF levels (r = −0.312, p = 0.004), a positive correlation was present between BDI scores and MDA levels (r = 0.320, p = 0.005) in PD patients. On PD patients with EDS analysis, we did not find a statistically significant correlation between O&NS parameters or BDNF levels with BDI scores (Table 4).

Table 3. Comparision of BDNF, NO, MDA, SOD and GSH levels (means ± SD) of PD patients between with and without EDS.

	With EDS patients (n = 28)	Without EDS patients (n = 55)	p Values
MDA (µmol/L)	0.78 ± 0.13	0.67 ± 0.16	0.008
GSH (mg/dL)	33.99 ± 4.29	34.69 ± 4.71	0.529
SOD (U/L)	2.53 ± 0.35	2.62 ± 0.72	0.526
NO (µmol/L)	23.35 ± 6.11	22.33 ± 3.39	0.347
BDNF (pg/mL)	206.81 ± 140.31	323.90 ± 225.82	<0.001

Notes: BDNF, brain-derived neurotrophic factor; NO, nitric oxide; MDA, malondialdehyde; SOD, superoxide dismutase; GSH, glutathione; and EDS: elevated depressive symptoms.

Table 4. Spearman correlation coefficients (r) between BDI scores and analyzed parameters in PD patients and those with EDS.

	r	p
PD patients with EDS (n = 28)
MDA µmol/L	−0.114	0.597
GSH mg/dL	−0.010	0.961
SOD U/L	−0.050	0.803
NO µmol/L	−0.071	0.736
BDNF pg/mL	0.268	0.167
PD patients (n = 83)
MDA µmol/L	0.320	0.005
GSH mg/dL	−0.047	0.684
SOD U/L	0	0.998
NO µmol/L	0.041	0.722
BDNF pg/mL	−0.312	0.004

Notes: BDI, Beck Depression Inventory.

Our study showed that as dependent variable, BDNF was significantly elevated due to raised levels of SOD in PD patients (p: 0.043). Linear regression analysis results on PD patient and those with EDS are listed in Table 5.

Table 5. Association between BDNF and O&NS levels: results of linear regression analyses.

	PD patients with EDS (n = 28)		PD patients (n = 83)
	BDNF
	β-coefficient (95% CI)	p Value	β-coefficient (95% CI)	p Value
MDA	15.53 (−421.09 to 452.15)	0.941	18.65 (−251.28 to 288.58)	0.891
GSH	−1.91 (−13.02 to 9.20)	0.723	−3.98 (−13.99 to 6.03)	0.430
SOD	−63.87 (−229.84 to 102.11)	0.431	78.95 (2.63 to 155.26)	0.043
NO	−0.70 (−12.34 to 10.94)	0.901	−5.03 (−15.67 to 5.61)	0.349

Notes: CI, confidence interval; BDNF, brain-derived neurotrophic factor; and O&NS, oxidative and nitrosative stress.

Discussion

In our study, the PD patients had significantly lower levels of BDNF compared with controls. This suggests that the basal levels of BDNF could be lower in PD patients compared with healthy persons. Prior studies on hemodialysis patients revealed both increased17 and decreased plasma BDNF levels.13

In addition, BDNF levels negatively correlated with BDI scores in PD patients. Within PD patients, the EDS group also had lower levels of BDNF than those without EDS. So we suggest that depressive symptoms could be elevated by reduced levels of BDNF. However, we did not find a correlation between BDI scores and BDNF levels. This may be due to the relatively small number of patients with EDS.

Numerous researchers have proposed a role for BDNF in the pathogenesis of depression. In animals, it has been shown that acute and chronic stress causes decreased expression of hippocampal BDNF protein and mRNA and a difference in modulation of BDNF expression between the amygdala and hippocampus. Furthermore, a significant reduction in the blood and plasma levels of BDNF in depressed patients has been reported.10^,18^,19

Prior studies in which the severity of depression was evaluated by Hamilton Rating Scale for depression (HDRS) only or both HDRS and BDI yielded mixed results (both with and without negative correlation between BDNF levels and HDRS or BDI).19–21

A decrease in BDNF levels20^,21 and an increase in oxidative stress as well as nitrosative stress had been shown in depressed patients, separately.22 The decreased levels of the basal serum BDNF accompanied with enhanced oxidative stress in hemodialysis patients.13

It is known that increased oxidative stress leads to tissue injuries as well as nucleic acid damage in CKD patients.11^,23^,24 Our results are consistent with the previous reports in terms of oxidative balance in CKD. The results of this study showed that MDA and NO levels significantly increase in PD patients compared to controls. These findings suggest that the PD patients are exposed to oxidative stress. The BDI scores correlated with MDA levels in PD patients. However, in our study, the BDNF levels were not affected by the degree of O&NS. As antioxidant defense factors were significantly lower in PD patients than controls as has been shown by other studies.24^,25 We found that the BDNF levels raise in conjunction with the SOD activity in PD patients. SOD acts a major defense system against oxidative damage by enzymatically converting O₂− to H₂O₂. This may result in diminished cellular damage and increased BDNF levels.

There was no significant difference when patients with and without EDS SOD and GSH levels were concerned. On the other hand, as a marker of raised oxidative stress, MDA levels were significantly higher in PD patients with EDS than those without EDS. Although this suggested that with EDS group was more exposed to oxidative stress, the BDNF levels were not affected by the O&NS factors in PD patients with EDS. In addition, the NO levels were significantly higher in PD patients than controls, but there was no significant difference between patients with and without EDS. Prior studies also revealed that increased levels of NO in CKD patients.26–28

Increased O&NS may be the result of raised production of ROS and RNS and/or decreased availability of antioxidant defenses. It may cause damage to cellular components, induce harmful autoimmune responses and ultimately result in failure of normal cellular processes. Prior studies also support a crucial role for O&NS in the pathophysiology of unipolar and bipolar depression.22

Based on our findings, we can speculate that the basal serum levels of BDNF are low in PD patients, especially in those with EDS are deeper. Our results suggest that either a decrease in BDNF or an increase in MDA may lead to depressive symptoms in PD patients.

One of the limitations of our study is relatively small sample sizes of each group. Second, we did not perform other validated screening tools such as Cognitive Depression Index or HDRS. In addition, the control group was not screened for symptoms of depression.

In conclusion, our results suggest that antioxidants and BDNF levels are decreased while O&NS are elevated in PD patients. Furthermore, in PD patients with EDS, the significant decrease in neuroprotective BDNF was accompanied by increased severity of oxidative stress. These results suggest that increased oxidative stress may cause cellular injury that may trigger depressive symptoms.

Further studies are needed to determine the mechanism(s) of O&NS and BDNF in CKD patients with depression.

Acknowledgments

We thank Gür Akansel from Medical School of Kocaeli University for his help with the manuscript.

Declaration of interest

The authors declare that there are no conflicts of interest.