A pilot study of thiamine and folic acid in hemodialysis patients with cognitive impairment

Background: This study aimed to explore the effectiveness of thiamine and folic acid supplementation on the improvement of the cognitive function in patients with maintenance hemodialysis. Methods: In the present study, we randomly assigned patients undergoing hemodialysis who had the Montreal Cognitive Assessment (MoCA) score lower than 26 to treatment group (n=25, thiamine 90mg/day combined with folic acid 30mg/day) or control group (n=25, nonintervention). All subjects were followed up for 96 weeks. The primary outcome was the improvement of the MoCA score. The secondary outcomes included homocysteine level, survival and safety. Results: Patients in treatment group had an increase of the MoCA score from 21.95±3.81 at baseline to 25.68±1.96 at week 96 (P<0.001, primary outcome), as compared with the MoCA score from 20.69±3.40 to 19.62±3.58 in control group. Thiamine combined with folic acid treatment also resulted in lower level of serum homocysteine in treatment group compare with control group at week 96 (P<0.05, secondary outcome). 3 patients and 9 patients died during follow-up period in treatment and control group respectively (P=0.048). The proportion of adverse events in treatment group was signicantly lower than that in control group. Conclusion: Hemodialysis patients with cognitive impairment treated with thiamine and folic acid had a signicant improvement in MoCA score.


Background
End Stage Renal Disease (ESRD) is a public health issue of global concern. United States Renal Data System (USRDS) reported that the overall prevalence of ESRD in 2018 is 1,943 per million people in the United States [1] . The number of dialysis patients in Asia is expected to rise to 2.162 million by 2030 [2] .
Cognitive Impairment (CI) is one of the common neurological complications in patients with ESRD [3] . Our previous single-center cohort observational study showed that the incidence of CI in patients with maintenance hemodialysis (MHD) was 51.6% and 3 times higher than that of non-Kidney disease patients [4] . MoCA score which has been con rmed in the previous literature [5,6] was used to determine cognitive impairment in MHD patients in this study. Furthermore, the 3-year survival rate of MHD patients with CI was signi cantly lower than that of patients with normal cognitive function [4] .
An association between cognitive dysfunction and oxidative stress has been discussed [3] . Animal studies demonstrated that the cognitive function of ESRD mice which was assessed by the radial arm water maze test decreased compared with the sham-operated control mice. Furthermore, treatment with tempol (an antioxidant) in ESRD mice improved cognitive function compared to treatment with vehicle. Histological assessment suggested indicated increased oxidation in the hippocampal neurons and increased numbers of neurons undergoing apoptosis or necrosis, compared to control mice and those treated with tempol [7] .
Recently, studies have found that thiamine, as a cofactor of transferenolase, plays an important role in reducing the production of reactive oxygen species (ROS) in the nervous system and reducing oxidative stress [8,9] . Folic acid has a direct antioxidant effect, which interacts with enzyme endothelial nitric oxide synthase (eNOS) and affects the bioavailability of nitric oxide cofactor [10] . Moreover, recent animal study have also found that oxidative stress index including malondialdehyde (MAD), protein carbonyl, 8hydroxydeoxyguanosine (8-OHdG) and nitric oxide (NO) were closely related to thiamine and folic acid de ciency, leading to oxidative stress and cognitive impairment [11] .Clinical studies showed that both thiamine and folic acid supplementation can reduce the level of oxidative stress index including homocysteine in blood [10,12] , thereby reducing oxidative stress and bringing bene ts to the treatment of cardiovascular complications in MHD patients [12] , particularly in those intensive treatment or with a drop in homocysteine levels of more than 20 percent [12] . However, recent systematic review and Meta-Analysis indicated that thiamine or folic acid alone can not improves cognitive function healthy older people [13,14] . Therefore, we designed this pilot study to explore whether the intensive combination of these two vitamin B can improve cognitive function in MHD patients with CI. The dose of thiamine 90mg/d is refer to the treatment dose of Wernicke's encephalopathy [15] and refeeding syndrome [16] . The dose of folic acid 30mg/d is according to the treatment of cardiovascular complications in MHD patients [12] .
The present study conducted a randomized controlled pilot study to explore whether patients on MHD with CI can improve their MoCA score and survival by thiamine combined with folic acid.

Study design and participant
This study was designed as a randomized, controlled, single-center study. Eligible patients were adults (18-80 years of age) with end-stage kidney disease who had been undergoing hemodialysis for at least three months and who had cognitive impairment (the MoCA score lower than 26 and <25 if the patients get education less than 12 years) [5] .The detail inclusion and exclusion criteria were showed in the Supplemental Figure 1. MoCA [6] was evaluated by professional neurologists to assess cognitive ability. Every participant completed the MoCA test before or during the rst hour of haemodialysis at baseline, week 48 and week 96.

Randomization
Randomization sequence was generated by an independent data manager using SAS (version 9.3) and stored within sealed opaque envelopes. The study physician opened an envelope to obtain the treatment assignment every time a patient was consented to enter the trial. Assignments were balanced in a 1:1 ratio between the two groups. No masking was applied.

Data collection
Hemodialysis patients in Renji Hospital, School of Medicine, Shanghai Jiao Tong University were screened. Eligible patients assigned to treatment group received thiamine 90mg/day combined with folic acid 30mg/day for 96 weeks, and those assigned to control group had nonintervention. Concomitant treatment with stable doses of anticoagulant medicine, antihypertension medicine, hypoglycemic medicine and statins. The initiation of new improved cognitive medicine was prohibited. The subjects who signed the informed consent were grouped by the investigator according to the treatment allocation.
During the 96-weeks intervention period, urine volume and the following information of hemodialysis treatment was recorded: average weekly ultra ltration (L/session), intradialytic hypotension (The de nition of intradialytic hypotension is systolic pressure <90 mmHg or diastolic pressure <60 mmHg, accompanied by dizziness and other symptoms of hypotension during dialysis treatment within one week prior to follow-up), body weight of pre and post haemodialysis, blood pressure (BP) of pre and post haemodialysis.

E cacy Outcomes
The primary outcome was the mean change from baseline at week 96 in the MoCA scores. The secondary outcomes included the mean change of serum homocysteine from baseline to week 96, survival and safety during follow-up.
Estimated the serum thiamine, folic acid and homocysteine Vitamin detector LINBIAO LK3000VI (Tianjin lanbiao electronic technology development co., LTD. China) and the original matching thiamine content determination kit were used to detect serum thiamine levels of two groups. Beckman Coulter's UniCel DxI 800 Access Immunoassay System (Beckman Coulter, inc., United States) and the original folic acid assay kit were used to detect serum folic acid levels of two groups by immunoluminescence assay. The Hitachi 7600-020 automatic biochemical analyzer (Hitachi, Japan) and the corresponding homocysteine assay kit (Ningbo meicang biotechnology co., LTD. China) were used to detect serum homocysteine levels.

Statistical analyses
The primary analysis only included patients with completed primary outcomes. The comparisons among MoCA scores and other secondary outcomes over time within subjects was conducted by using repeated measure ANOVA, and pairwise t-test with Bonferroni adjustment was performed for multiple comparisons. On the other hand, for comparisons between treatment and control group, simple one-way ANOVA or Wilcoxon rank sum test was used. The chisquare method was used for statistical analysis of the categorical data. For missing data in secondary e cacy outcomes, they were assumed to be missing at random and multiple imputation was conducted. Kaplan-Meier survival curve was used for survival analysis.
Variables of normal distribution were presented by means with SD while skewed ones were reported as the median and the inter-quartile ranges. Counting data were expressed as constituent ratios or percentages.
A p < 0.05 was considered statistically signi cant difference. Statistical analyses were conducted with SPSS (Version 20.0, SPSS Inc., Chicago, IL, USA).

Results
Baseline characteristics of patients A total of 50 patients with MHD met the inclusion criteria and included in the present study. According to the study protocol, subjects were randomly divided 1:1 into treatment group (n=25) and control group (n=25) (Supplemental Figure 1).
There was no signi cant difference in demographic data, medical history and hemodialysis data between treatment group and control group except ultra ltration volume (P=0.039;  Table 2). In addition, MoCA scores increased signi cantly from baseline to 96 weeks in treatment group, but not control group ( Table 2). The proportion of patients with MoCA scores greater than or equal to 26 was signi cantly higher in treatment group than that in control group at week 96 (72.7% vs 6.2%, P<0.001; Supplemental Table 2).
Seven items of MoCA score were compared between the treatment group at baseline and the 96 weeks of follow-up. We found that the item of abstraction and delayed recall were improved by thiamine and folic acid treatment signi cantly (Supplemental Table 3).

Secondary analysis
At 96 weeks of follow-up, the level of serum homocysteine in treatment group and control group were 32.58±13.42 μmol/L and 45.69±18.54 μmol/L, P=0.016. We also found that the levels of serum thiamine and folic acid in treatment group were signi cantly higher than those in control group (Table 2).
In addition, the level of folic acid, but not serum thiamine or homocysteine had signi cantly changed over time in treatment group ( Table 2).
There was no signi cant difference in hemodialysis data and laboratory test between treatment group and control group at 48 weeks or 96 weeks of follow-up (Supplemental Table 4 and Supplemental Table  5). Notes: The normal range for blood thiamine concentration is approximately 70-180nmol/L [17] and the normal range for serum levels of folate acid is approximately 2.7-17ng/ml [18] in the general population.
A total of 12 patients died during a follow-up of 96 weeks. The mortality rate in treatment group and control group was 12.0% and 36.0%, P=0.047. The Kaplan-Meier survival curve analysis showed that the survival rate of treatment group was lower than that of control group (P=0.048; Figure 1). The causes of death were cardiovascular and cerebrovascular diseases (9 cases, 75.0%), malignant tumors (2 cases, 16.7%), and infection (1 case, 8.3%) ( Table 3).

Discussion
In the present pilot study, we found that the MoCA score of treatment group was signi cantly improved, implicating that supplementation with thiamine and folic acid may improve the cognitive function of patients undergoing MHD with CI. The survival of patients in treatment group was signi cantly improved compared with that in control group. The proportion of adverse events in the control group was signi cantly higher than that in the treatment group, especially in cardiovascular and cerebrovascular events.
The present study pointed out that after 96 weeks of thiamine (90mg/day) and folic acid (30mg/day) supplementation, the MoCA score of the treatment group were signi cantly higher than that of control group, implicating the improvement of cognitive function. Recently, an animal study in mice demonstrated that thiamine de ciency decreased activity of antioxidants and increased activity of malondialdehyde (MAD), protein carbonyl, 8-hydroxydeoxyguanosine (8-OHdG) and nitric oxide (NO) in the cerebral cortex and hippocampus, leading to oxidative stress and cognitive impairment [11] .
Meanwhile, an experimental mouse model simulating the effects of exposure to methotrexate on behavior and cognitive function found acute decrease in serum and CSF levels of folate acid, leading to oxidative stress and displayed cognitive [19] . This may be due to thiamine, as a cofactor of ketolase, can act as an oxygen free radical scavenger and play an important role in reducing the production of reactive oxygen species (ROS) in the nervous system and alleviating oxidative stress [11] . Another reason is the direct antioxidant effect of folic acid, which interacts with endothelial nitric oxide synthase (eNOS) to affect the bioavailability of NO cofactors. Moreover, folic acid is essential for the metabolism of homocysteine into methionine, which can reduce the homocysteine level of MHD patients, thus relieving oxidative stress [10] . However, there is insu cient evidence to support supplementation with thiamine or folic acid alone can improve cognitive functioning in healthy older people or non-chronic kidney diseases (CKD) older adults [13,14] . Therefore, we designed this pilot study to explore whether the combination of these two vitamin B can improve cognitive function in MHD patients with CI.
Furthermore, the normal range for blood thiamine concentration is approximately 70-180nmol/L in the general population [17] . However, in MHD patients, the level of blood thiamine concentration is lower than general population which has been founded in our study. Because thiamine is a water-soluble vitamin, the absorption is affected by uremia toxins. Moreover, it will be cleared by dialysis, especially in patients receiving high-ux dialysis. After 96 weeks of thiamine supplementation, blood thiamine concentration rose, but did not reach the normal range, which may be related to the decreased expression of thiamine transporter 1 and thiamine transporter 2 caused by uremia environment [18] or a higher dose of thiamine may be needed. Further research is needed to con rm this. Moreover, the normal range for serum levels of folate is approximately 2.7-17ng/ml in the general population [18] . However, in MHD patients, the serum levels of folate is at the low end of the normal range which has been founded in our study. After 96 weeks of folate supplementation, serum levels of folate were above the normal range,and this is exactly what our research needs to achieve.
Homocysteine is not only a uremia toxin but a biomarker of oxidative stress in patients with MHD. Many observational studies have suggested that elevated blood homocysteine levels are strongly associated with CI [21][22][23] . Moreover, previous studies have found that there are many mechanisms that cause elevated homocysteine levels in MHD patients including de ciency of vitamins B, especially thiamine and folate acid [24] . In the present study, blood homocysteine levels were signi cantly higher than the normal range in both groups at baseline, suggesting the presence of oxidative stress in patients of MHD with CI [25] . Compared with baseline, homocysteine didn't decrease signi cantly (p=0.063) at week 96. Although, serum homocysteine level at week 96 in treatment group was lower than that in control group, we can't conclude thiamine combined with folic acid treatment improve MoCA scores dependent on decreasing homocysteine. A new study with larger sample size and longer follow-up time should be performed to clarify the role of serum homocysteine.
It is worth mentioning that the survival rate of MHD patients with CI was signi cantly lower than that of patients with normal cognitive function, suggesting that CI is a risk factor for death of MHD patients [4] .
Recent COGNITIVE-HD studies have also con rmed this phenomenon [26] , therefore, it is urgent to nd effective methods or measures to treat or improve the concurrent CI of MHD. In this study, Kaplan-Meier survival curve analysis indicated that the survival rate in treatment group was signi cantly higher than that in control group. The main cause of death in the two groups is cardiovascular and cerebrovascular events (9 cases, accounting for 75%), including 8 cases (88.9%) in the control group and 1 case (33.3%) in the treatment group. We believe that thiamine combined with folic acid treatment doesn't only improve MoCA scores but also decrease the risk of cardiovascular and cerebrovascular disease in patients undergoing hemodialysis.
Several limitations of our study should be considered. First, this study was not designed to use a blind method and placebo control. It may lead to the generation of psychological bias and in uence the MoCA scores. Furthermore, the sample size is small, which will lead to insu cient grasp in statistics and may cause false negative results. Moreover, we only used MoCA score as the basis for judging cognitive functions, which was highly subjective and could not cover all the assessments of cognitive functions.
Future studies need to include more scoring criteria such as neuropsychological battery of 10 tests [27] , and even imaging tests such as functional magnetic resonance imaging [28] , to accurately judge cognitive function. Finally, patients in treatment group were given two medicines, it was di cult to identify which one played a key role.

Conclusion
In conclusion, supplementation of thiamine (90mg/day) combined with folic acid (30mg/day) can Informed Consent Written informed consent was obtained from all individual participants included in the study.
Animal Rights No animal studies were carried out by the authors for this article.

Consent for publication
Not applicable.
Availability of data and materials