Separate and combined effects of cold dialysis and intradialytic exercise on the thermoregulatory responses of hemodialysis patients


 Background: The separate and combined effects of intradialytic exercise training (IET) and cold dialysis (CD) on patient thermoregulation remain unknown. This study assessed the thermoregulatory responses of hemodialysis patients under four different hemodialysis protocols: a) one typical dialysis (TD) protocol (dialysate temperature at 37°C), b) one cold dialysis (CD) protocol (dialysate temperature at 35°C), c) one typical dialysis protocol which included a single exercise bout (TD+E), d) one cold dialysis protocol which included a single exercise bout (CD+E). Methods: Ten haemodialysis patients (57.2±14.9 years) participated in the study. Core and skin temperatures were measured using an ingestible telemetric pill and by four wireless iButtons attached on the skin, respectively. Body heat storage (S) calculated using the thermometric method proposed by Burton. Results: The TD and TD+E protocols were associated with increased S leading to moderate effect size increases in core body temperature (as high as 0.4°C). The low temperature of the dialysate during the CD and the CD+E protocols prevented the rise in S and core temperature (p>0.05), even during the period that IET took place. Conclusions: TD and IET are accompanied by a moderate level of hyperthermia, which can be offset by CD. We recommended that CD or with IET can prevent the excessive rise of S. Trial registration: Clinical Trial Registry number: NCT03905551 (clinicaltrials.gov), DOR: 05/04/2019, https://clinicaltrials.gov/ct2/show/NCT03905551?term=NCT03905551&draw=2&rank=1


Background
Hemodialysis (HD) represents a signi cant challenge for the thermoregulatory system of HD patients (1,2). Indeed, body temperature rises during HD due to (i) heat transfer into the body via the heated dialysate, (ii) endogenous heat production through normal metabolic processes, and (iii) attenuated heat loss at the skin surface (3). The latter has been hypothesized to occur because cutaneous vessels are constricted during typical dialysis (TD; 37 °C dialysate temperature) due to loss of blood volume towards the extracorporeal circuit (4). This leads to attenuated heat dissipation from the skin surface despite the fact that metabolic heat production remains relatively stable (5). Consequently, heat is accumulated (this is typically referred to as "increased body heat storage") and, soon, core temperature rises during a typical session of hemodialysis (6). This increased heat storage (S) offsets the vasoconstrictive response to hypovolemia (7) and it is one of the responsible contributing factor which leading to the intradialytic hypotension causing patient discomfort and increased mortality (8,9).
Lowering the dialysate temperature (35-36 °C; cold dialysis) has been proposed as a simple and useful method to reduce heat storage during hemodialysis and, therefore, decrease the frequency of intradialytic hypotension episodes (5,10,11). Indeed, cold dialysis (CD) attenuates the risk for patient hyperthermia compared to typical dialysis (dialysate temperature at 37 °C; TD) and leads to cutaneous vasoconstriction (12). These observations have led to a growing interest in the thermal and circulatory adaptations occurring during CD (13). Yet, the precise changes in body heat balance during either CD or TD remain poorly documented and understood.
The above-mentioned bene ts of CD in thermoregulatory and cardiovascular stability are highlighted even further due to the well accepted adoption of intra-dialytic exercise training programs (14). It has been well-established that intra-dialytic exercise leads to bene ts in strength and endurance (14)(15)(16) as well as improved clearance (Kt/v), (17) hemodynamic stability, and patient quality of life (14). The bene cial effects of intradialytic-exercise are due to increased muscle blood ow and reduced intercompartmental resistance by peripheral vasodilation (18,19). Overcoming this resistance seems to be the single most effective method for improved toxin removal during hemodialysis (20,21). Nevertheless, this vasodilation is augmented further during TD by the exercise-induced hyperthermia (18) and may, therefore, increase the frequency of intradialytic hypotension episodes. As indicated above, we hypothesized that CD may be able to minimize the need for peripheral vasodilation, leading to improved patient thermoregulatory and cardiovascular stability. However, despite a strong rationale for the implementation of intradialytic exercise training programs and the aforementioned bene ts of CD, the separate and combined effects of these protocols on patient thermoregulation have not been investigated to date.

Aims
The aim of the current study was to assess the thermoregulatory responses of hemodialysis patients under four different hemodialysis (240 min) protocols: a) one typical dialysis (TD) protocol (dialysate temperature at 37 °C), b) one cold dialysis (CD) protocol (dialysate temperature at 35 °C), c) one typical dialysis protocol which included a single exercise bout (TD + E), d) one cold dialysis protocol which included a single exercise bout (CD + E).

Study Design
Ten haemodialysis patients were recruited from a single haemodialysis unit at the General Hospital of Trikala, Thessaly, Greece. All study measurements were performed at a hospital climate control room using the metabolic ward of the General Hospital of Trikala, Greece. The mean age was 57.2 ± 14.9 years (Table 1). Patients enrolled by a research assistant assigned into the study while the order that the patients assigned to the rst scenario was random using a computer random number generator. Each patient was monitored during a) one protocol of typical dialysis with dialysate temperature at 37 °C (TD), b) one protocol of cold dialysis with dialysate temperature at 35 °C (CD), c) one protocol of typical dialysis which included a single exercise bout (TD + EP) and d) one protocol of cold dialysis which included a single exercise bout (CD + EP). The patient participants underwent hemodialysis therapy Page 4/16 (Fresenius 4008B, Oberursel, Germany) three times per week with low ux, hollow ber dialyzers and bicarbonate buffer, with the hemodialysis protocol lasting 4 hours. The dialysis protocols were performed using dialysis ow at 550 ml/min and mean average of conductivity dialysance at 137-140 mEq/L. The net ultra ltration weight was the same in all sessions. All patients were clinically stable and they had received regular hemodialysis treatment for at least 3 months, with adequate dialysis delivery Kt/V > 1.1 and good compliance of dialysis treatment, the serum albumin was > 2.5 g/dL, hemoglobin ≥ 11 g/dL and treated with rHuEPO. Patients were not eligible for participation in the study if they had a reason to be in a catabolic state, such as hyperthyroidism, active vasculitis, malignancies, pregnant, HIV, opportunistic infections, musculoskeletal contraindication to exercise, requirement for systemic anticoagulation, participant or participated in an investigational drug or medical device study within 30 days or ve halflives or in ammations, that required intravenous antibiotics within 3 months prior to enrollment, diabetics receiving insulin therapy, New York Heart Association grade IV heart failure, and mental incapacity to consent. Dialysis protocols were performed in a random order at the same time and day of the week to minimize differences in ultra ltration volume between the four protocols. The ambient temperature in the room was 25.2-25.9 °C. Food consumption was not allowed during the dialysis procedure and all participants was wearing standardized clothing to secure the same insulating properties of clothing during the different dialysis protocols. During the different dialysis protocols, core temperature (T C ) and mean skin temperature (T sk ) were recorded. The body heat storage (S) for every minute was calculated during all conditions using the thermometric method proposed by Burton (22). The data recording lasted ve hours for each patient (1 hour before dialysis protocol and 4 hours during the dialysis protocol). The exercise program during TD + E and CD + E was performed between the 60th and the 120th minute of the dialysis protocol.
Core Temperature Measurements (t) Core temperature (T c ) was measured at the gastrointestinal tract using an ingestible telemetric pill. Data recorded continuously at 1-minute intervals, throughout the course of the experimental intervention. The telemetric pill was ingested by the patients 7-hour before arriving in the hospital (23,24).
Mean Skin Temperature Measurements (t) Skin temperature was measured at 1-minute intervals by wireless iButtons (iButton, Maxim, USA). The iButtons were programmed before their application on the skin, as outlined by the manufacturer. The iButton resolution was set at 0.06 o C and the iButton real-time clock was synchronized with a laptop computer. The iButtons were attached on the skin using water-resistant, medical-grade tape. In total four iButtons were attached on the skin, at the following anatomic locations: on the biceps, pectoralis major, rectus femoris, and gastrocnemius, and were used to calculate mean skin temperature (T sk ) using

Intradialytic Exercise Program
The patients performed cycling for 60 minutes in the supine position during the TD + EP and CD + EP protocols. Patients were asked to pedal on a bedside cycle ergometer (Model 881 Monark Rehab Trainer, Monark Exercise AB, Varberg, Sweden) at 60% of their pre-assessed maximum power capacity. The exercise regime started 1 h after the commencement of the hemodialysis session. The patients' maximum power capacity was determined by a modi ed version of the Åstrand Bicycle Ergometer Test protocol at bedside on a previous dialysis session during hemodialysis. Exercise was well tolerated by all patients, and no adverse reactions were reported.

Sample Size Estimation
The minimum required sample size was determined using data from a previous study (26) where 17 hemodialysis patients underwent CD aiming to reduce TC from 36.4 ± 0.4 °C at baseline to 36.1 ± 0.1 °C at the end of the dialysis protocol. Sample size calculations were conducted using G*Power 3.0 (27). The "Means: difference between two dependent means" method was used to calculate the power of the within effect. A two-tailed test selected. Statistical power and α error probability were set to 0.80 and 0.05, respectively. The minimum required sample size was determined by calculating the effect size d. Using the aforementioned published data, the resulting minimum required sample size was 10 participants.

Results
Ten stable chronic hemodialysis patients were eligible and consent to participated in the study. All participants completed all 4 dialysis protocols on a random order. The patient's characteristics are presented in Table 1.

Changes in Core Body Temperature (Tc) under four dialysis protocols
The typical dialysis (TD) protocol resulted in an increase of core body temperature (T C ) compared to cold dialysis protocol (CD) (TD: 36.9 ± 0.1 °C; CD: 36.7 ± 0.2 °C). This was evident by statistically signi cant differences between CD and TD from 0 to 180 min (p < 0.05) as well as large effect sizes (d > 0. 8) observed from 0 to 150 min and a moderate effect size (d = 0.5-0.8) observed at 180 min (Fig. 2).
However, the T C remained similar during the TD + E and the TD protocols (TD: 36.9 ± 0.1 °C; TD + E: 36.9 ± 0.2 °C) except during the period that exercise training took place, where a slight increase in T C was evident (TD: 36.9 ± 0.1 °C; TD + E: 37.0 ± 0.1 °C) resulting in a medium effect size observed at the end of exercise (Fig. 2).
The T C during the CD + E protocol was lower than during the TD (TD: 36.9 ± 0.1 °C; CD + E: 36.6 ± 0.2 °C), which was evident by signi cant reductions (p < 0.05) and moderate/large effect sizes until 180 min into the protocol (Fig. 2). These differences were also observed during the exercise period of the CD + E (TD: 36.9 ± 0.1 °C; CD + E: 36.7 ± 0.1 °C).
The T C increased almost throughout the TD + E protocol compared to the CD (CD: 36.7 ± 0.2 °C; TD + E: 36.9 ± 0.2 °C), which was evident by statistically signi cant differences (p < 0.05) and moderate/large effect sizes (Fig. 2).
During the exercise period of the CD + E protocol (CD: 36.5 ± 0.1 °C; CD + E: 36.7 ± 0.1 °C), the T C was signi cantly increased compared to the CD (p < 0.05 and moderate effect sizes from 60 to 90 min), yet the low dialysate temperature used in the CD + E was able to disseminate this additional amount of heat (Fig. 2). As a result, the T C was similar across the CD and the CD + E protocols (CD: 36.7 ± 0.2 °C; CD + E: Also, T sk slightly increased during the exercise period of the CD + E protocol compares with cold dialysis (CD) protocol (CD: 31.3 ± 0.4 °C; CD + E: 31.7 ± 0.9 °C), paralleled with a p value of 0.08 and a moderate effect size at 120 (Fig. 3). However, there was no overall T sk differences between the CD and the CD + E protocols (CD: 31.3 ± 0.4 °C; CD + E: 31.3 ± 0.9 °C). Regarding comparison between TD + E protocol and CD + E resulted in somewhat increased T sk (TD + E: 31.7 ± 0.8 °C; CD + E: 31.3 ± 0.9 °C; (Fig. 3).

Changes in Body Heat Storage (S) under four dialysis protocols
The body heat storage (S) were similar during the typical dialysis (TD) and cold dialysis (CD) protocol (TD: 5.5 ± 40.5 W; CD: 19.3 ± 37.7 W; p > 0.05 and d < 0.4). Overall, the S was slightly increased during the TD + E protocol (TD: 5.5 ± 40.5W; TD + E: 11.2 ± 141.4W) and particularly during the exercise period, as indicated by moderate effect sizes compared with typical dialysis (TD) protocol (Fig. 4). Additionally S were similar between typical dialysis (TD) and cold dialysis (CD) which included a single exercise bout (TD: 5.5 ± 40.5 W; CD + E: 2.2 ± 112.5 W; p > 0.05 and d < 0.4). During the typical dialysis (TD) which included a single exercise bout (TD + E) the body heat storage (S) was somewhat increased compared to cold dialysis (CD) (CD: 19.3 ± 37.7W; TD + E: 11.2 ± 141.4W), observed via moderate effect sizes at different points (i.e., 30, 150, and 240 min) during the protocol. The slight increases in T C and T sk during the cold dialysis which included a single exercise bout (CD + E) were also evident in terms of S, where moderate/large effect sizes were observed at different time points (i.e., 30, 60, 120, 150, and 240 min) compared with cold dialysis (CD). However, as a whole, there were no major differences in S between the two protocols (CD: 19.3 ± 37.7W; CD + E: 2.2 ± 112.5W). In terms of S, the CD + E led to attenuated values compared to TD + E, particularly during the exercise phase, as evidenced by moderate effect sizes (Fig. 4).

Discussion
The present study, sought to examine for the rst time the separate and combined effects of cold dialysis and intradialytic exercise training on the thermoregulatory responses (core temperature, skin temperature, and body heat storage) of stable hemodialysis patients using for the rst-time whole-body direct temperature assessment tools. Our results demonstrated that the TD and TD + E protocols are associated with increased body heat storage leading to moderate increases in core temperature (as high as 0.4 °C).
Such changes in body heat storage and core temperature are known to cause peripheral vasodilation (28,29) and may offset the vasoconstrictive response to hypovolemia (7) which is responsible for intradialytic hypotension causing patient discomfort and increased mortality (8,9). Therefore, the present detailed thermoregulatory assessment results con rm previous evidence suggesting that TD represents a challenge for the thermoregulatory system of patients with end-stage renal disease especially in developing countries where dialysis units ambient conditions are inadequately controlled (1,2). In contrast, the low temperature of the dialysate during the CD and the CD + E protocols prevented the rise in body heat storage and core temperature, even during the period that exercise training took place.
It has been well-established that intra-dialytic exercise leads to bene ts in physical performance and quality of life in hemodialysis patient (14)(15)(16). Our results demonstrated that during exercise phase and especially when the dialysis temperature was at 37 °C, body heat storage slightly increased (Fig. 1). Studies show that during the intradialytic exercise an increased in skin blood ow limits the cardiovascular adjustment needed for work, because skin circulation participated both in hemodynamic control and thermoregulation (14).
Even though the uniqueness of our study as well as the laborious methodology and highly skilled personnel required a number of limitations unfortunately still remain. It is important to denote that we were unable to obtain peripheral/skin blood ow data assessing the blood redistribution occurred during the 4 different scenarios. Therefore, our inferences regarding peripheral vasoconstriction/vasodilation stem from skin temperature measurements. Nevertheless, it is well known that skin temperature is very well correlated with changes in the cutaneous circulation (28,29). Another issue to consider is the effect of xed reduction in dialysate temperature. As a consequence, the dialysis treatment must be adapted to the patient's individual condition and response to treatment. Indeed, recent reports support the link between cold dialysis and low intradialytic hypotension episodes in the hypotensive HD patients (30).
However, the used of ingestible telemetric pill (temperature sensor) represents a valid index of body core temperature (Tc) that is convenient and shows excellent utility for prove valuable in eld studies, in investigations requiring frequent measurements over long periods, or if the subject needs to be entirely free during the observations (31).

Conclusions
In conclusion, the results of the present study demonstrate that typical dialysis and intradialytic exercise are accompanied by a moderate level of hyperthermia, which can be offset by cold dialysis. Based on these ndings, we observed that hemodialysis sessions which incorporate cold dialysis alone or supplemented with intradialytic exercise can prevent the rise of body heat storage. The latter has been shown to be a major factor for developing intradialytic hypotension, which is due to both hemodynamic responses (hypovolaemia stress) and thermoregulatory responses (thermal stress) during hemodialysis. Programme "H2020 MSCAS-RISE-Muscle Stress Relief" (grant agreement No. 645648). Funders did not have any involvement in aspects related to study design, data acquisition, data analysis or interpretation.   Core Body temperature (Tc) during 4 different dialysis protocols. *Statistical signi cant differences between TD vs CD protocol. ^ Statistical signi cant differences between TD vs CD +E protocol. + Statistical signi cant differences between TD +E vs CD protocol. # Statistical signi cant differences between CD +E vs CD protocol during exercise period.