Changes in plasma glutamate in hemodialysis-related-headache patients: A pilot study

Abstract

Objective

This study aims to determine the profile of glutamate in hemodialysis-related headache (HDH) patients compared to those without HDH.

Background

HDH is a frequent complication in patients undergoing hemodialysis, although little is known about its pathophysiology.

Methods

Pilot observational and prospective study. Blood samples of patients consecutively recruited with end-stage renal disease undergoing hemodialysis were obtained at onset, at the first 30 min, and at 4 h of the hemodialysis session. Glutamate and ALT, AST plasma levels were analyzed at these three stages. Patients were classified into HDH and non-HDH groups. Hospital charts were revised for clinical history and other laboratory results.

Results

Thirty patients (60% males, mean age 70.2 ± 10.5 years) attending the Hemodialysis Unit were included. Nine (30%) experienced HDH. The whole sample showed a significant reduction of glutamate and an increase of ALT and AST along hemodialysis. Specifically, the magnitude of the reduction of plasma glutamate levels concentration in HDH group was stable during the 30 min and presented a significant posterior reduction, opposite to the evolution in non-HDH patients.

Conclusions

HDH was associated with a higher variation in plasma glutamate levels with a particular hemodialysis clearance profile, useful to detect patients with a higher susceptibility of developing HDH and receive preventive measures. In conclusion, HDH-group had a later drop in glutamate that could be related to HDH development. Other modalities with more prolonged solute clearance could be considered. We hope the findings of this study will help promote more research to interpret the role of glutamate variations in patients with hemodialysis related headaches.

Resumen

Objetivo

Este estudio tiene como objetivo determinar el perfil del glutamato en pacientes con cefalea relacionada con la hemodiálisis (CRH) en comparación con aquellos sin CRH.

Antecedentes

La CRH es una complicación frecuente en los pacientes sometidos a hemodiálisis, aunque su fisiopatología aún se conoce poco.

Métodos

Estudio piloto observacional y prospectivo. Se obtuvieron muestras sanguíneas de pacientes con enfermedad renal terminal en tratamiento con hemodiálisis, reclutados de manera consecutiva, al inicio, a los treinta minutos y a las cuatro horas de la sesión de hemodiálisis. Los niveles plasmáticos de glutamato, alanina aminotransferasa (ALT) y aspartato aminotransferasa (AST) fueron analizados en las tres etapas. Los pacientes se clasificaron en dos grupos: con CRH y sin CRH. Se revisaron las historias clínicas hospitalarias para obtener antecedentes clínicos y otros resultados de laboratorio.

Resultados

Se incluyeron treinta pacientes (60% varones, edad media 70,2 ± 10,5 años) atendidos en la Unidad de Hemodiálisis. Nueve (30%) presentaron CRH. En el conjunto total se observó una reducción significativa del glutamato y un incremento de ALT y AST durante la hemodiálisis. Específicamente, en el grupo con CRH la reducción de los niveles plasmáticos de glutamato se mantuvo estable durante los primeros 30 minutos y presentó una disminución significativa posterior, en contraste con la evolución observada en los pacientes sin CRH.

Conclusiones

La CRH se asoció con una mayor variación de los niveles plasmáticos de glutamato y un perfil específico de depuración durante la hemodiálisis, lo que podría ayudar a identificar a pacientes con mayor susceptibilidad y aplicar medidas preventivas. El grupo con CRH presentó una disminución tardía del glutamato, posiblemente relacionada con el desarrollo de la cefalea, lo que podría justificar considerar modalidades de diálisis con depuración más prolongada. Estos hallazgos podrían promover nuevas investigaciones sobre el papel del glutamato en la cefalea asociada a la hemodiálisis.

Introduction

Hemodialysis-related headache (HDH) is an usual complication in patients undergoing hemodialysis. When defined according to the International Classification of Headache Disorders (ICHD) criteria,¹ most series find a prevalence of HDH in around 30% of the dialyzed patients, although reported frequencies ranged from 6% to 48%.^2–7

Despite the important impact of HDH in quality of life, very few studies have tried to elucidate its underlying pathophysiology, since it was first described by Bana et al. five decades ago.² Previous research has identified factors that could contribute to this disorder, such as prominent shifts in blood urea nitrogen levels and arterial blood pressure,^3,5 high sodium and low magnesium levels,⁴ type of dialysate⁶ ultrafiltration rate⁸ or high calcitonin gene related peptide plasma levels.⁹

Glutamate is the principal and most widespread excitatory neurotransmitter in the brain. It is known to be neurotoxic at elevated concentrations and to play an important role both in brain insults such as stroke or traumatic brain injury¹⁰ and in pain-related conditions such as migraine and central sensitization,^11,12 although its participation in other types of headache is not well known.

Alanine aminotransferase (ALT) and aspartate aminotransferase (AST), formerly called serum glutamic pyruvic transaminase and serum glutamic oxaloacetic transaminase, respectively, are blood resident enzymes involved in reversible glutamate scavenging in peripheral blood.

The efficacy of plasma glutamate reduction with extracorporeal methods such as hemodialysis and peritoneal dialysis has been previously proved in humans and animals.¹³ In addition, higher plasma glutamate concentrations, as well as lower levels of ALT and AST were observed in patients on hemodialysis compared with healthy controls.^14,15

To the best of our knowledge, no studies have explored glutamate levels in dialysis-related headache so far. This study is aimed to determine the pattern and rate of clearance of glutamate in HDH patients during hemodialysis. Understanding the underlying mechanisms of HDH could be useful to establish change in hemodialysis modality to improve hemodialysis-patients quality of life.

Methods

This observational and prospective study comprised thirty consecutively recruited patients with end-stage renal disease undergoing hemodialysis in the tertiary care La Princesa University Hospital in Madrid (Spain). The recruitment period lasted 18 months. We included end-stage renal disease adult patients who underwent hemodialysis for longer than a month. Exclusion criteria were: (i) clinical instability; (ii) cognitive impairment; (iii) refusal or inability to obtain the informed consent. The study followed the Declaration of Helsinki and Good Clinical Practice and the protocol was approved by the Drugs Research Ethics Committee of La Princesa University Hospital. All participants signed the informed consent. Biological samples were stored as part of the registered collection from the Instituto de Salud Carlos III C.0003640.

Hemodialysis procedure

All patients were on standard hemodialysis, performed three times weekly for 4 h. One type of high-flux dialyzer was used: Toray NS21. No reuse technique was performed. Blood flow rate ranged between 350 and 400 mL/min.

Baseline clinical features

Data regarding demographic features were obtained from digital medical charts. Other clinical features collected were vascular risk factors, renal disease etiology, treatments, arterial blood pressure and other current diseases that could increase AST/ALT levels, such as alcohol consumption.

Headache questionnaire

Patients were interviewed by a neurologist about their headaches through a questionnaire (Fig. 1) based on the ICHD-3 criteria (Table 1), specially designed for this study. Firstly, they were asked about the presence of headache before and after the initiation of hemodialysis (HD). Depending on the presence of previous headache, patients were classified into different groups: (i) presence of headache that fulfilled the diagnostic criteria of hemodialysis-related headache, in our study HDH patients, (ii) presence of headache before starting hemodialysis for the first time (with the same or different phenotype) (non-HDH patients), (iii) and patients with frequent headache before the hemodialysis that experienced an improvement of frequency (<12 per year) following the start of hemodialysis. Other items included were time of headache onset during hemodialysis when present, frequency of the attacks and pain characteristics. Considered pain features were: intensity (mild, moderate or severe based on the verbal analogue scale from 0 to 10: 1–2 mild, 3–7 moderate, 8–10 severe); quality type (oppressive, throbbing, stinging, burning or others); duration (less than 4 h, 4–72 h, more than 72 h); location (holocraneal, hemicraneal, retro-ocular, occipital, mixed); associated symptoms (nausea, vomiting, photophobia, phonophobia, red eye, tearing, ptosis, nasal congestion); worsening with exercise; and influence of pain on the daily activities. The type of medication taken for pain relief and their efficacy on pain relief was also collected.

Figure 1.

Questionnaire based on ICHD-3 beta criteria, designed to assess pre- and post-hemodialysis headaches. Questionnaire used and designed specifically for this study, based on the ICHD-3 beta version criteria, to provide information on possible headaches presented by patients before, during and after hemodialysis.

Table 1.

Adapted from the ICHD-3 diagnostic criteria for dialysis headache (10.2).

Code	10.2 dialysis headache
Description	A headache with no specific characteristics, occurring during and caused by hemodialysis. It resolves spontaneously within 72 h after completion of the hemodialysis session.
Diagnostic criteria	A. At least three episodes of acute headache fulfilling criterion C.B. The patient is receiving hemodialysis.C. Evidence of causation demonstrated by at least two of the following: 1. Each headache occurs during a hemodialysis session. 2. One or both of the following: (a) Each headache worsens during the hemodialysis session. (b) Each headache resolves within 72 h after the end of the hemodialysis session. 3. The headaches cease completely after successful renal transplantation and termination of hemodialysis. D. Not better accounted for by another ICHD-3 diagnosis.

Source: International Classification of Headache Disorders, 3rd edition (ICHD-3), Section 10.2, Headache attributed to dialysis. Available at: https://ichd-3.org.

Laboratory parameters

Blood samples for the determination of glutamate were obtained in the first hemodialysis session of the week at three different stages: (i) immediately after the initiation of the hemodialysis session (baseline, 0 h); (ii) during the first 30 ± 15 min of the hemodialysis onset (0.5 h); (iii) right after the disconnection from the hemodialysis machine (4 h). Plasma levels of AST and ALT were also determined.

For the determination of glutamate, blood samples were recollected in EDTA tubes and were centrifuged immediately after collection and the plasma was stored at −80 °C until use. A previously validated method was used for plasma glutamate determinations.¹³ Briefly, a high-performance liquid chromatography (HPLC) instrument coupled to a triple quadrupole mass spectrometer (MS/MS) (Agilent 1200 Series, 6410B MS) was used. Chromatographic separations were performed at 25 °C on an ACE C18-PFP column (4.6 mm × 150 mm, 3-μm, SYMTA, Madrid, Spain). The mobile phase was a combination of a phase A (0.2% formic acid in water) and B (pure acetonitrile) which was run in a 12-min gradient followed by a 4-min equilibration step. Samples were prepared by precipitating proteins with 1% acetonitrile, evaporating the supernatant and reconstituting the pellet in the mobile phase. 5 μL of sample or calibrator was injected. The quantitative determination of AST and ALT was performed using biochemistry tubes and following standard procedure of the hospital clinical labs on Roche automated clinical chemistry analyzers (Cobas 8000 mod 701 Roche). Data on other blood test results regarding hemoglobin, calcium, phosphate, potassium, sodium, urea or creatinine serum levels in these patients were also collected.

Statistical analysis

All statistical analyses were conducted using SPSS for iOs version 20.0 (IBM Corp., Armonk, NY, USA). Chi-square and Fisher's exact tests were carried out for clinical qualitative variables. Since ALT, AST and Glutamate values were not normally distributed, non-parametric procedures were used. For the evaluation of the differences in the behavior of ALT, AST and glutamate between HDH and non-HDH patients, we performed three main analyses in the sample. Firstly, we compared blood levels of ALT, AST and glutamate at the dissimilar stages using the Mann–Whitney test and confirmed results with a two factor ANOVA. Secondly, we compared the variation in their blood levels between the dissimilar stages with the Friedman test, followed by the Wilcoxon post-hoc test with Bonferroni correction when necessary. Thirdly, we performed correlations between the parameters at the different research stages using the Spearman correlation coefficient. For other biochemical values, Student's T-test (parametric data) and Mann–Whitney test (non-parametric data) were performed regarding its Gaussian distribution. A p < 0.05 was considered statistically significant for the 95% confident intervals.

Outliers and missing data

An outlier patient with ALT and AST values far above the others was found. The cause of the rise was unknown. Some blood samples were hemolyzed, which led to missing ALT values at 30 min and 4 h in two patients, and missing AST values at 30 min and 4 h in two and three patients, respectively. There are also missing data for glutamate in three patients, one of them with HDH. To ensure that the outlier and missing data were not interfering the results, two new sets of patients were additionally analyzed: set 2, dismissing the ALT-AST outlier patient, and set 3, dismissing the outlier patient and imputing the missing ALT (4 values) and AST (5 values) data by a multiple regression analysis using predicting variables of each dependent variable.

Results were stable in the three different sets of patients previously described, whether only raw data were analyzed (set 1), the outlier was subtracted (set 2) or missing values were imputed by a multiple regression analysis (set 3).

Results

Clinical and demographic features

We included a total of 30 consecutive patients, 18 males (60%) and 12 females (40%), mean age 70.2 ± 10.5 years, range 47–90, attended at the Hemodialysis Section of the Nephrology Department. Median time since the initiation of the hemodialysis treatment was 18.5 months, range 14–166. Of the 30 patients, nine (30%) presented headache that met the ICHD-3 criteria of HDH. Two patients (7%) had headaches before the initiation of hemodialysis for the first time: one of them, with a migrainous pattern, fulfilled hemodialysis-related headache ICHD-3 criteria; and the other, with tensional type, became headache-free after initiating hemodialysis.

Table 2 summarizes demographic and clinical data. We did not find any differences in age, sex ratio, vascular risk factors, time since initiation of chronic hemodialysis therapy, blood pressure measures at the beginning of the hemodialysis session or renal disease etiology between the HDH and non-HDH groups. We also compared biochemical characteristics and found no significant differences between the two groups in their basal blood levels of hemoglobin, calcium, phosphate, potassium, sodium, urea or creatinine. None of the patients consumed alcohol regularly.

Table 2.

Demographic and clinical features of the hemodialysis headache (HDH) group and the non-HDH control group.

	HDH group(N = 9)	Non-HDH group(N = 21)	p-Value	All patients(N = 30)
Sex			0.418
Male	4 (44%)	14 (67%)		18 (60%)
Female	5 (56%)	7 (33%)		12 (40%)
Age (years)*	68 ± 10	71 ± 11	0.384	70 ± 11
Vascular risk factors
Hypertension	7 (78%)	18 (86%)	0.622	25 (83%)
Diabetes mellitus	1 (11%)	4 (19%)	1.000	5 (17%)
Dyslipidemia	5 (56%)	14 (67%)	1.000	19 (63%)
Underlying renal disease			0.948
Glomerulonephritis	3 (33%)	6 (28%)		9 (30%)
DiabeticNephropathy	1 (11%)	3 (14%)		4 (13%)
ADPKD	1 (11%)	2 (10%)		3 (10%)
ObstructiveUropathy	0 (0%)	3 (14%)		3 (10%)
Rare cause	1 (11%)	1 (5%)		2 (7%)
Unknown	3 (33%)	5 (24%)		8 (27%)
Multifactorial	0 (0%)	1 (5%)		1 (3%)
BP systolic (mmHg)*	135 ± 21	135 ± 22.7	0.974	135 ± 21.9
BP diastolic (mmHg)*	76 ± 13	72 ± 11	0.350	73 ± 11.9
Time since HD initiation (months)^#	19 (11–36)	18 (15–40)	1.000	18.5 (14–37)

Data are presented as N and %, as mean ± SD (*) or as median (Q1–Q3) (^#). Notes: HDH: hemodialysis headache; ADPKD: autosomal dominant polycystic kidney disease; BP: blood pressure at the beginning of the hemodialysis session; HD: hemodialysis; SD: standard deviation.

Hemodialysis-related headache clinical features

Clinical features of HDH are shown in Table 3. Headache quality was oppressive in all cases, although location varied, with 33% having unilateral pain and 22% fulfilled criteria for migraine without aura. None had trigemino-autonomic symptoms. Headache intensity was predominantly moderate. In a third of the HDH group, headache interfered with their daily life. Pain onset in HDH patients started mostly between the second and third hour of the hemodialysis session (67%). During the first 2 h it began in 33% of cases and none began during the last hour. This headache, in more than half of the cases (56%) lasted between 4 and 72 h and in no case did it continue beyond 72 h. All patients took medication at least once for headache relief (mostly acetaminophen). It was effective in 89% (eight patients out of nine).

Table 3.

Hemodialysis headache (HDH) clinical characteristics.

	N (%)
Time of pain onset during hemodialysis
First 2 h	3 (33%)
Between 2 h and 3 h	6 (67%)
From 3 h to 4 h	0 (0%)
HDH pain location
Holocranial	3 (33%)
Hemicranial	3 (33%)
Retro-ocular	0 (0%)
Occipital	2 (22%)
Mixed	1 (11%)
HDH pain intensity
Mild	0 (0%)
Moderate	7 (78%)
Severe	2 (22%)
HDH pain duration
Less than 4 h	4 (44%)
4–72 h	5 (56%)
More than 72 h	0 (0%)
Worsening with exercise or movement
No	7 (78%)
Yes	2 (22%)
Symptoms associated
Light sensitivity	1 (11%)
Phonophobia	2 (22%)
Nausea/vomiting	2 (22%)
Pain aggravation by movement	2 (22%)
Red eye	0 (0%)
Tearing	0 (0%)
Ptosis	0 (0%)
Nasal congestion	0 (0%)

Data are presented as N and %.

Glutamate and ALT and AST blood levels

In all patients undergoing HD, including HDH and non-HDH, we observed a statistically significant decrease in glutamate between baseline and the fourth hour. Conversely, ALT and AST were significantly increased (Table 4).

Table 4.

Evolution of glutamate and ALT, AST blood levels along the different research stages in the whole hemodialysis group (HDH and non-HDH).

	Baseline (0 h)	0.5 h	4 h
Glutamate (μM)	98.9 (72.3–117.7)	83.9 (53.6–144.6)	73.1 (53.7–123.1)^$
ALT (UI/L)	11.0 (8.0–15.0)	11.0 (8.0–15.0)	12.0 (9.0–15.0)**
AST (UI/L)	13.0 (10.0–16.0)	14.0 (10.0–19.0)	15.0 (12.0–20.0)**

Data are presented as median (Q1–Q3). **p = 0.001, ^$p < 0.001 vs baseline (0 h).

When comparing glutamate and AST, ALT levels at each research stage (baseline, 0.5 h and 4 h) between the HDH and non-HDH groups, we did not find significant differences (Table 5), although glutamate levels were higher in HDH patients at the three stages.

Table 5.

Glutamate blood levels (μM) at different research stages in HDH and non-HDH groups.

Research stages	HDH group(N = 9)	Non-HDH group(N = 21)	p-Value
Glutamate (μM)
Baseline (0 h)^B	104.2 (82.8–141.7)	87.5 (67.6–113.2)	0.360
0.5 h^B	106.4 (87.7–141.2)	67.3 (49.4–144.6)	0.132
4 h^B	85.3 (59.3–125.9)	70.5 (51.5–123.1)	0.418
ALT (IU/L)
Baseline (0 h)	12.5 (10.3–15.5)	11.0 (6.0–15.0)	0.163
0.5 h^A	12.5 (10.3–16.0)	10.5 (7.0–13.5)	0.132
4 h	12.0 (11.3–17.0)	12.5 (7.8–15.0)	0.486
AST (IU/L)
Baseline (0 h)	14.0 (10.3–17.0)	12.5 (9.8–14.3)	0.230
0.5 h^A	15.0 (10.3–19.0)	13.5 (9.8–15.3)	0.403
4 h^B	16.0 (11.3–18.8)	15.0 (11.8–20.3)	0.802

Data are presented as median (Q1–Q3). Notes: A: N = 28; B: N = 27.

However, when specifically considering the variation values of the plasma parameters (0.5 h–baseline, 4 h–baseline, 4 h–0.5 h) between the two groups (Fig. 2), we did find that the decrease in glutamate between 0.5 h and 4 h was significantly greater (p = 0.039) in the HDH group than in the non-HDH group [(−14.5 [−27.8–5.3]) vs (−1.3 [−11.1–6.8])]; respectively.

Figure 2.

Glutamate blood level variations during hemodialysis stages in HDH vs non-HDH patients. Violin plots showing the distribution of plasma glutamate concentrations at baseline (0 h), 30 min (0.5 h), and 4 hours (4 h) during the hemodialysis session in patients with and without hemodialysis-related headache (HDH). Each violin displays the kernel density of the data; the white dot represents the median, the thick bar the interquartile range, and the thin line the range of the data.

As shown in Fig. 2, glutamate in HDH patients was stable during the first 30 min, exhibiting an afterwards reduction through the rest of the HD session. This contrasts with the initial decrease and posterior stabilization in the non-HDH group.

Neither ALT nor AST showed any differences in their level of variation in the different study stages among HDH vs non-HDH patients (p > 0.05; data not shown). In HDH patients, AST increased while ALT decreased at 4 h, although this result was not significant.

No significant correlation was found between glutamate levels at any research time and months of duration since the initiation of the hemodialysis treatment (Spearman's rho = 0.042; p = 0.8).

These results were stable in the three different sets of patients previously described, whether only raw data was analyzed (set 1), the outlier was subtracted (set 2) or missing values were imputed by a multiple regression analysis (set 3).

Discussion

As far as we are concerned, this is the first study exploring glutamate plasma profiles in patients with and without HDH. Despite being the most frequently encountered neurological symptom in patients undergoing hemodialysis, its semiological and physiopathological features are still poorly defined.

In this study, we found that HDH is associated with a higher variation in plasma glutamate levels with a particular clearance pattern during hemodialysis and a tendency of higher predialysis glutamate, AST and ALT concentrations.

Clinical HDH features in our sample were similar to previous studies, as we found a 30% prevalence and heterogeneous manifestations.^5–7 All patients reported an oppressive pain, most with moderate intensity and none described trigemino-autonomic symptoms, while 22% fulfilled criteria for migraine without aura, one of them having a known history of migraine. HDH appeared more frequently during the second and the third hour of hemodialysis as described in previous research.⁷ Duration of the pain was most commonly 4–72 h, while in former studies it predominantly lasted less than 4 h.^3,5–7 Differences in semiological aspects present HDH as a miscellaneous syndrome that could be embracing diverse underlying physiopathological mechanisms.

Regarding the glutamate clearance profile, we found different results from those of a previous publication by Rogachev et al.¹⁴ They reported that glutamate concentrations decreased significantly during the 4 h of the hemodialysis session, most prominently in the first hour, with a subtle increase observed in the fourth hour.

In our study, changes in glutamate concentrations, and also in AST and ALT, were more prominent after the first hour in HDH patients, showing an inverse pattern to that of the study. Interestingly, the clearance pattern of plasma glutamate levels in our non-HDH patients was similar to the results of the study mentioned above. Although not statistically significant, glutamate levels tended to be higher in the HDH group at all research stages.

Systemic increase in glutamate levels has been associated with migraine, both ictally and interictally,¹² as well as with other neurological disorders, such as ischemia, traumatic brain injury or epilepsy. In these pathological conditions, where brain glutamate is abnormally elevated, there is a brain-to-blood glutamate efflux following a gradient of concentration. This efflux could explain the increase of glutamate levels in peripheral plasma in these disorders.¹³

We hypothesize that the dysregulation in glutamate homeostasis in HDH patients, revealed by a tendency of higher predialysis glutamate concentration together with its stronger and earlier variation, could be related to the generation of the headache attack.

A first plausible explanation might be that hemodialysis treatment in HDH patients is less effective in scavenging plasma glutamate during the first part of hemodialysis, possibly related to the effects of the higher brain-blood glutamate gradient and the resulting glutamate efflux to blood.¹³ We found no other differences in main clinical characteristics or comorbid conditions that could better explain the disparities in glutamate, AST or ALT clearance pattern or baseline plasma concentrations in the HDH group.

Secondly, fluctuating glutamate levels could lead to transient changes in nociceptive signaling involved in episodic headache attacks. Glutamate and its receptors have been implicated in the activation of the nociceptive neurons along the trigeminovascular pathways and the initiation of headache, especially in migraine, but also in cluster-headache.^12,16

Anteriorly, shifts in laboratory or hemodynamic parameters have already been associated with HDH. A study by Gozubatik-Celik et al. found that variation in serum blood urea nitrogen levels and blood pressure was higher in HDH patients than in controls.³ Although the magnitude of variation was not compared, lower magnesium, higher sodium⁴ and higher calcitonin gene related peptide⁹ predialysis plasma levels have been also observed in patients with HDH. Nevertheless, some of these findings could not be reproduced in other similar studies.⁸

Headache seems to be more frequent in pediatric and adolescent patients with end-stage renal disease undergoing dialysis than patients with end-stage renal disease without hemodialysis treatment.⁸ This supports that changes in plasma products induced by hemodialysis are more relevant in triggering headache than the sole accumulation of toxic metabolites. Also, attacks appear during the changes produced by hemodialysis and not between sessions. The fact that HDH is more frequent in patients with standard hemodialysis treatment compared to other modalities with more prolonged solute clearance, such as peritoneal dialysis⁶ or hemodialysis using lower ultrafiltration rate,⁸ backs up the hypothesis that the greater the solutes shifts are, the higher the probability to induce headache.

It could be also considered that the rapid reduction of plasma glutamate concentration increases the brain-blood osmotic gradient, as described with the reverse urea effect in the hemodialysis disequilibrium syndrome,^3,17 favoring a net water movement into the brain and producing cerebral edema, intracranial hypertension and headache. However, there is some evidence against a major role for other osmolytes in the contribution to the brain-to-blood osmotic gradient. This is the case for glutamine, glutamate, taurine, and myoinositol, which do not significantly rise in the brain after rapid hemodialysis.¹⁷

All our patients received hemodialysis with the same periodicity to avoid that prolonged interval between dialysis sessions may cause a more prominent difference between levels of osmolytes, glutamate or other molecules that could be involved in HDH.² We also found no significant differences between groups in pre-dialysis urea, sodium or other biochemical blood levels, nor in pre-hemodialysis blood pressure measurements, as reported in previous studies. We reasonably excluded the presence of dialysis disequilibrium syndrome in our cohort, as all patients were on chronic hemodialysis programs (three to five sessions per week), and this condition typically occurs in acute patients during their first dialysis sessions. Therefore, the magnitude of urea fluctuations in our sample was not sufficient to induce such a syndrome. Serum sodium and magnesium levels remained stable throughout the study. Sodium levels show minimal intrapatient variability, as the physiological osmostat maintains homeostasis regardless of intake, and the dialysate contains magnesium, preventing hypomagnesemia.

Likewise, blood pressure values – before, during, and after dialysis – were within normal limits and did not show abnormal fluctuations related to headache onset. Caffeine consumption was also recorded, and only one patient reported drinking decaffeinated coffee, which reasonably rules out caffeine withdrawal as a potential cause of headache.

Taken together, these findings suggest that the headaches observed in our study were not attributable to metabolic or hemodynamic imbalances, nor to caffeine withdrawal, but rather to mechanisms directly related to the hemodialysis process itself.

Regarding ALT and AST, it has been suggested that, because both play a crucial role in converting glutamate to the inactive metabolite 2-ketoglutarate, their enhanced activity during hemodialysis may explain, in part, the observed decrease in plasma glutamate levels.^13,14 In our study no significant differences in AST or ALT concentrations were found. A previous research found reduced AST activity in migraine patients compared to controls. They also reported that AST activity was higher during migraine attacks than in interictal periods, but there is lack of evidence of the role of AST and ALT and their relationship with glutamate in HDH patients.

We observed that the evolution of ALT and AST was similar during the first part of the hemodialysis session in both groups as the correlation of their progressive increase was statistically significant at all research stages, except at 4 h in HDH patients, where AST kept on increasing while ALT decreased. Although the difference in their variation was not statistically significant in our study, this could suggest that AST activity has a stronger role than ALT in reducing glutamate levels in HDH patients.

Study limitations

The main limitation of this study is the small size of the sample. Potential contributors to headache such as the presence of a mood disorder or caffeine withdrawal were not recorded either. The study is underpowered to associate glutamate levels and HDH semiological profiles. However, it represents a gateway to future research in this direction.

Conclusions

In conclusion, HDH-group had a later drop in glutamate that could be related to HDH development. Other modalities with more prolonged solute clearance could be considered to improve hemodialysis-patients quality of life. We hope the findings of this study will help promote more research to interpret the role of glutamate variations in patients with hemodyalisis related headache.

Funding

Project supported by a competitive grant. This research has been funded by Instituto de Salud Carlos III (RICORS-RD21/0006/0009) and co-financed with FEDER Funds and/or from the European funds of the Recovery, Transformation and Resilience Plan and by NextGenerationEU. Additionally, this work was supported by SAF 2013-44108-T from MINECO and by FIS No. CA12/00122 to ARN and FPU12/02220 to AW.

Conflict of interest

Gago-Veiga AB has received honoraria as a consultant and speaker for: AbbVie-Allergan, Chiesi, Exeltis, Novartis, Eli Lilly and Teva. Abad-Santos has been a consultant or investigator in clinical trials sponsored by the following pharmaceutical companies: Abbott, Alter, Chemo, Cinfa, FAES, Farmalíder, Ferrer, GlaxoSmithKline, Galenicum, Gilead, Italfarmaco Janssen-Cilag, Kern, Normon, Novartis, Servier, Silverpharma, Teva, and Zambon.

Acknowledgments

We gratefully acknowledge Jesús Garrido, statistician at the Health Research Institute of La Princesa University Hospital, for his outstanding help with the statistical analysis. The authors want to thank to Fundación Teofilo Hernando for its continued support.