Reduced dietary intake induces body fluid hypotonicity via alterations in water and energy metabolism

Abstract

Abstract

Maintenance of the body's fluid balance is essential for vital human functions. Although drinking water is widely recommended, the role of dietary intake in body fluids remains unclear. This study investigated whether overnight dietary restriction with adequate water consumption affects body water balance, its regulatory responses, and thermoregulatory and cognitive functions during exercise in a hot environment. Fifteen young adults experienced two conditions: sufficient meal intake (CON) and small dinner with skipped breakfast (RED), both with adequate controlled water consumption. Blood and urine samples, as well as indices of systemic circulation, thermoregulatory responses and cognitive functions, were obtained before, during and after a 60 min moderate‐intensity treadmill exercise. RED induced lower serum osmolality and urine sodium excretion compared with CON (both P < 0.05), suggesting the development of body fluid hypotonicity with impaired urinary concentrating capacity. Water‐ and electrolyte‐regulating hormone levels remained unchanged in RED (all P > 0.05). Systemic circulation and thermoregulatory responses (heart rate, blood pressure, skin blood flow and sweat rate) remained comparable between the conditions, leading to similar core body and skin surface temperature elevations (all P > 0.05). Moreover, cognitive performance (Go/No‐Go, Stroop, reaction time and digit span tasks) was not changed in RED (all P > 0.05). Notably, RED increased oxygen uptake with a reduced respiratory quotient during exercise (both P < 0.05), indicating a metabolic shift toward lipid oxidation. These findings suggest that overnight dietary restriction induces body fluid hypotonicity with altered water and energy metabolism.

Key points

• Dietary intake provides substrates essential for water and osmotic balances, yet its role in maintaining hydration remains underexplored.

• This study assessed whether overnight dietary restriction (RED), with adequate water intake, induces dehydration and affects body functions during exercise in heat.

• RED induced body fluid hypotonicity, as evidenced by lower serum osmolality and sodium excretion, without changes in body weight or urine volume before and after exercise.

• RED led to reduced urinary concentrating capacity during exercise, which may reflect changes in the reabsorption of electrolytes and water in the renal tubules.

Abstract figure legend Overnight reduced dietary intake (RED; small dinner with skipped breakfast), compared to a sufficient dietary condition (CON), induced body fluid hypotonicity during moderate exercise in a hot environment, despite equal total fluid intake. RED led to lower serum osmolality and reduced urinary sodium excretion, without differences in body weight, urine volume or systemic hydration indices. Post‐exercise increases in urine osmolality and electrolyte concentrations seen in CON were absent in RED, indicating impaired urinary concentrating ability. Additionally, RED triggered a metabolic shift toward lipid oxidation, as evidenced by a decreased respiratory quotient and elevated oxygen uptake. These results suggest that even short‐term dietary restriction, when hydration is maintained, can alter fluid–electrolyte balance and renal function, potentially representing an early stage of disrupted fluid homeostasis.

Introduction

Maintaining optimal body water content is crucial for vital human functions (Armstrong & Johnson, 2018; Sawaka et al., 2005). The effect of fluid intake on thermoregulatory and overall cognitive functions has garnered empirical attention (Adan, 2012; Grandjean & Grandjean, 2007; Kleiner, 1999; Murray, 2007; Watanabe, Kadokura et al., 2024), leading to widespread recommendations for regular hydration via ‘drinking water’ (Gandy, 2015; Kleiner, 1999; Thomas et al., 2008). In contrast, while fluid intake through meals constitutes approximately 40% of overall fluid intake (Jéquier & Constant, 2010), the fact that substrates contained in meals are essential for water and osmotic balances is often overlooked (Heyman et al., 2020; Rowntree, 1922; Soetan et al., 2010; Valentine, 2007). In modern contexts, transient fasting or insufficient dietary intake within a day is often experienced, especially in healthy young persons, for several reasons, such as skipped breakfast forced by a busy lifestyle, dieting to lose weight or loss of appetite accompanied by long‐lasting hot summers (Hill, 2002; Horikawa et al., 2011; Mandic et al., 2019; Pendergast et al., 2016; Rong et al., 2019). Such conditions may compromise the body's water balance, even when water drinking is adequately conducted. Additionally, physiological and cognitive functions may be affected, as observed when water drinking is insufficient. However, the influence of dietary intake on body fluids and the associated responses remain to be fully clarified. Given the anticipated increase in heat‐related illnesses due to rising global ambient temperatures (Gasparrini et al., 2017; Huang et al., 2011), elucidating the role of dietary intake is critical for mitigating health risks, particularly during exercise and labour in hot environments.

Some studies have reported that 3–5 day fasting‐induced diuresis and natriuresis were reversed by carbohydrate intake (Schloeder & Stinebaugh, 1970; Veverbrants & Arky, 1969). Although the mechanism remains unclear, several factors are assumed to be involved: lower plasma level of glucose that partly contributes to the maintenance of osmolality in the extracellular fluid, insulin, glucagon and atrial natriuretic peptide (ANP), and ketone levels in the plasma and renin–angiotensin–aldosterone system (RAAS) (Maoz et al., 1992; Schloeder & Stinebaugh, 1970; Spark et al., 1975; Szenasi et al., 1985). Recent speculation indicated that sodium–glucose co‐transporter 2 (SGLT‐2) at the proximal tubules in the kidney plays an important role in this phenomenon. Fasting may suppress sodium reabsorption due to lower glucose levels in the glomerular filtrate, resulting in natriuresis (Heyman et al., 2020). In a typical diet, carbohydrates constitute the largest proportion of the total caloric intake compared with other nutrients (Shan et al., 2019). Therefore, skipping meals or insufficient dietary intake may induce natriuresis and subsequent loss of extracellular water due to reduced carbohydrate intake.

Reduced electrolyte content in the body, accompanied by skipping meals or insufficient dietary intake, may also affect the body's water balance, including osmotic equilibrium and fluid distribution between the intra‐ and extracellular fluid compartments (Lobo, 2004), which are usually maintained by the volume and osmotic receptors (Bourque et al., 1994; Danziger & Zeidel, 2015; Strauss, 1957). Even when sufficient water drinking is conducted, insufficient electrolyte intake will decrease body water content. Excess water to osmotic contents decreases plasma osmolality, which returns to normal by inhibiting antidiuretic hormone secretion and water diuresis (Bie, 1980; Thornton, 2010). However, the extracellular fluid volume is maintained at lower than normal (isosmotic dehydration).

The aim of this study was to clarify the influence of insufficient dietary intake on body water balance, its regulatory responses, and thermoregulatory and cognitive functions during exercise in a hot environment. To assess the influence in the real world, insufficient dietary intake was conducted by controlling water drinking and diet starting from 24 h before the experiment. The diet consisted of a small dinner and skipped breakfast. We hypothesised that (1) overnight dietary restriction with sufficient fluid intake induces isotonic dehydration, and (2) such dehydration attenuates thermoregulatory and cognitive functions during exercise in a hot environment.

Methods

Ethics statements

All experimental procedures and protocols were approved by the Human Research Ethics Committee of Waseda University (approval number: 2023‐212) and conducted in accordance with the Declaration of Helsinki. Participants were informed of the experimental procedures but were not informed of the aim and hypotheses to prevent potential bias. The participants believed this study investigated the effects of two different dietary and water consumption patterns on physiological and psychological responses. At the end of the experiment, the participants were debriefed about the study's aim and requested not to share this information with other scheduled participants.

Participants

Fifteen young adults [seven males and eight females; age, 20 ± 2 years; height, 162.0 ± 7.0 cm; body weight, 53.42 ± 7.91 kg; peak oxygen uptake (${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$), 42.4 ± 6.2 ml/min/kg; mean ± SD] participated in this study. The male‐to‐female ratio was balanced to represent the general population rather than to specifically investigate sex‐based differences. All participants were university students recruited through flyer distribution. They were non‐smokers, had not been diagnosed with neuromuscular or cardiovascular diseases, and were not taking any prescription medications. Participants regularly engaged in recreational activities, including at least 30 min of low‐ (such as walking) or moderate‐intensity (such as jogging) aerobic exercise 3–5 days per week. Moreover, based on their ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$, the participants were classified as performance level 2 according to the classification of subject groups in sports science research (Decroix et al., 2016), indicating that they were healthy adults.

Experimental procedures

This study's protocol was based on our previous study, investigating the effect of 24 h water restriction‐induced mild dehydration on thermoregulatory and cognitive functions (Watanabe, Kadokura et al., 2024). Participants visited the laboratory on three separate days, with at least a 48 h washout period between visits.

On the first visit, participants underwent an incremental treadmill exercise test to determine their ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ and corresponding running speed. After measuring their nude body weight and changing into a standardised clothing set (underwear, T‐shirt, gym shorts and socks), participants entered an artificially controlled environment chamber at an ambient temperature of 25°C with 50% relative humidity (TBR‐12H; Espec, Osaka, Japan). Participants were seated for 10 min, then stood on a treadmill for 2 min before starting the test. The test involved walking at 3 km/h for 3 min, followed by a gradual speed increase of 0.1 km/h every 6 s until exhaustion. The ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ was determined using the breath‐by‐breath method with a respiratory metabolism device (AE‐100i; Minato Medical Science, Osaka, Japan). After the exercise test, the participants practised a series of cognitive tests (evaluating executive function, information processing and working memory) several times.

Figure 1 shows the experimental protocol during the second and third visits. Participants completed two different trials on the visits. Participants repeated the same exercise test in each trial with comparable and adequate water consumption. The dietary condition 15–16 h before the exercise test was altered between the two trials, the order of which was randomised and counterbalanced: sufficient meals (control condition, CON) were provided in one trial and reduced meals (reduced dietary condition, RED) in the other (see ‘Dietary intervention’). Each trial was conducted at the same time on each experimental day. All experiments were completed within 2 weeks to avoid the effects of seasonal change and dietary change within participants.

Figure 1. Illustration of the experimental protocol.

ABP, arterial blood pressure; VAS, visual analog scale; RPE, rating of perceived exertion; HR, heart rate; T _ear, ear canal temperature; mean T _skin, mean skin surface temperature; SkBF, skin blood flow; SR, sweat rate.

After urine sample collection, participants changed into the same clothes as on the first visit and rested in the supine position on a bed for at least 10 min, during which the assessment of abdominal great vessels (see ‘Inferior vena cava and abdominal aorta’). Blood sampling was conducted for participants who provided consent (n = 11) (see ‘Blood and urine samples’). Then, they entered an artificially controlled environmental chamber at an ambient temperature of 30°C with 60% relative humidity and sat for 30 min while measurement devices were attached. Subsequently, they completed a series of cognitive tests similar to those conducted during their first visit. After 5 min standing rest on the treadmill, the participants performed an exercise test consisting of three sets of 20 min walking exercises (40% ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$) separated by a 2 min rest. The ${\dot{V}}_{{\mathrm{O}}_2}$ and respiratory quotient (RQ) were continuously recorded using the breath‐by‐breath method (AE‐100i; Minato Medical Science). Heart rate (HR), ear canal temperature (T _ear, as an index of core body temperature), mean skin surface temperature (T _skin), skin blood flow (SkBF_chest and SkBF_forearm), and sweat rate in the chest and forearm (SR_chest and SR_forearm) were continuously measured. After arterial blood pressure (ABP) measurement, subjective thermal and humid sensations, thermal comfort, and physical and psychological fatigue were evaluated using the visual analog scale (VAS) 2 min before the end of the baseline and each exercise set. Moreover, a rating of perceived exertion (RPE) was obtained at the same time points. After the exercise test, the participants sat for 8 min, and blood samples were collected. Subsequently, another series of cognitive tests was conducted, followed by urine sample collection. Thirst sensation was assessed using the VAS before and after the trial.

Dietary intervention

The amount of water intake from both food and beverages (i.e. total water intake) and caloric intake were the same among the participants. Body size, sex and physical activity were not considered in determining the amount. This is because the amount of frozen food could not be precisely adjusted according to the parameters. Table 1 shows the calorie, nutrient and water consumption over 24 h before arrival at the laboratory (09.00 or 10.00 h), excluding breakfast, before the trial days.

Table 1.

The calories, nutrients, and water consumed over 24 h

	CON	RED	Δ
Calories, kcal	1490	800	690
Carbohydrates, g	220.0	136.1	83.9
Protein, g	43.7	23.6	20.1
Fat, g	50.3	19.0	31.3
NaCl, g	4.7	3.8	0.9
Potassium, mg	934.2	417.1	517.1
Phosphorus, mg	411.1	155.6	255.5
Drinking water, ml	2900	3040	140
Total water intake, ml	3480	3480	0

Drinking water, amount of water intake from beverages; total water intake, amount of water intake from meals and beverages.

Participants received meal packs and water bottles before each trial. In CON, participants consumed a meal pack for lunch and dinner and nutritional supplement bars for breakfast, which were eaten 3 h before arrival at the laboratory. In RED, participants consumed the same meal pack for lunch and reduced‐calorie dinner packs and fasted afterwards. To standardise 24 h water consumption in both trials, participants were instructed to maintain a consistent and adequate amount of water consumption, accounting for the water content of the meals. They were asked to drink water and eat food at the same time schedule in the two trials. Water intake for one time point was conducted with a small amount (∼ 100 ml) and evenly distributed across the afternoon, evening and subsequent morning. They were also required to record the timing or volume of meal and water intakes in a provided log sheet to ensure uniform meal and water consumption patterns between interventions.

Assessment of hydration state

Nude body weight

Nude body weight was measured before and after each test (±20 g accuracy; DP‐7800PW‐S; Yamato, Akashi, Japan). Total sweat volume during each test was estimated by the reduction in body weight.

Blood and urine samples

An 8 ml blood sample was obtained via venipuncture and collected into a collection tube that was left undisturbed for 15 min. Subsequently, it was centrifuged at 1500 g for 15 min at 4°C to separate the serum. A 5 ml urine sample was used for the following assays.

Serum and urine osmolality (S_osmol and U_osmol) were measured using the freezing‐point depression method (One‐Ten Osmometer; Fiske, Norwood, CA, USA). Na⁺, K⁺ and Cl⁻ concentrations in the serum, and S_Na+, S_K+ and S_Cl− and U_Na+, U_K+ and U_Cl− in the urine, respectively, were determined using ion‐selective electrodes (EL SE‐1520; ARKRAY, Kyoto, Japan). Urine specific gravity (USG) was measured using reflectometry (PAL‐09S; Atago, Tokyo, Japan). Urine volume during the exercise test was determined using a graduated cylinder.

Serum and urine creatinine (S_Cr and U_Cr) were measured using the Jaffé method (LabAssay Creatinine; Fujifilm Wako, Tokyo, Japan). Serum and urine urea nitrogen (S_UN and U_UN) were determined using ultraviolet absorption spectrophotometry. Serum glucose (S_glu) was measured using the Mutarotase‐GOD method (LabAssay Glucose; Fujifilm Wako). Serum arginine vasopressin and aldosterone (S_AVP and S_ald), which regulate anti‐diuresis and Na⁺ retention, respectively, were measured using enzyme‐linked immune sorbent assay kits (#583 951 Arginine Vasopressin ELISA Kit and #501 090 Aldosterone ELISA Kit, respectively; Caymann, Ann Arbor, MI, USA) for the thawed samples that were stored at −80°C until the assays.

Inferior vena cava and abdominal aorta

To assess the intravascular fluid volume, the diameter of the inferior vena cava (IVC) was measured, which was corrected by that of the abdominal aorta (Ao). It has been reported that the ratio of IVC to Ao diameter (IVC/Ao index) reflects the hydration status (Jauregui et al., 2014; Kosiak et al., 2008). Ultrasound measurements were performed in the supine position using a 3.5–5 MHz convex probe (Versana Active; GE Healthcare, Chicago, IL, USA) placed beneath the xiphoid process in a longitudinal orientation. IVC diameter was measured below the confluence of the hepatic veins, ensuring parallel alignment of the anterior and posterior walls. During a normal respiratory cycle, Ao diameter was measured 5–10 mm above the coeliac trunk. For each vessel, a 25 s video clip was recorded for subsequent analysis. IVC and Ao diameters were measured at their maximum dimensions during expiration for the IVC and systole for the Ao.

Assessment of other variables

Systemic circulation and autonomic thermoregulation (HR, ABP, SkBF, SR)

HR was continuously measured using a lead II electrocardiogram (BMS‐3400; Nihon Kohden, Tokyo, Japan). Systolic and diastolic blood pressures in the right brachial artery were measured using an upper‐arm cuff device (EBP‐330; Minato Medical Science). Skin blood flow in the non‐glabrous regions of the chest (SkBF_chest) and forearm (SkBF_forearm) were assessed using laser Doppler flowmetry (ALF21; Advance, Tokyo, Japan) at a sampling rate of 100 Hz. Local sweat rate in the chest (SR_chest) and forearm (SR_forearm) were assessed using dew hygrometry (SKD‐4000; Skinos, Nagoya, Japan) at a sampling rate of 1 Hz. HR, SkBF and SR were recorded using a 16‐bit A/D converter (Power Lab 16 s; AD Instruments, Sydney, Australia), and the data were stored on a personal computer.

Body temperature (T _ear and T _skin)

T _ear was assessed as an index of core body temperature using infrared thermometry (VTB01; Vitarate, Tokyo, Japan) (Kato et al., 2023). The ear orifice was covered with medical film to reduce the direct influence of ambient ventilation. T _ear was collected on a smartphone (iPhone 11; Apple, Cupertino, CA, USA) with an application program (Thermologger; Vitarate) every 30 s using Bluetooth. T _skin of the anterior chest (T _chest), upper arm (T _arm), and lateral sides of the thigh (T _thigh) and lower limb (T _limb) were measured every 60 s using a temperature/humidity logger (iButton Hydrochron; Maxim, Dallas, TX, USA) to estimate the mean skin temperature (mean T _skin), calculated using the equation 0.3 × (T _chest + T _arm) + 0.2 × (T _thigh + T _limb) (Ramanathan, 1964).

VAS‐assessed subjective perceptions and RPE

Thirst and whole‐body thermal sensations, comfort, humid sensation, and physical and psychological fatigue were self‐recorded by the participants using a VAS (Watanabe et al., 2025; Watanabe, Kadokura et al., 2024; Watanabe, Sugi et al., 2024). The VAS comprised a statement and a 200 or 100 mm straight line. One end of the line represented ‘extremely cold, uncomfortable or dry’ (−100 mm), and the other end represented ‘extremely hot, comfortable or humid’ (100 mm). For assessing thirst sensation and physical and psychological fatigue, one end of the line represented ‘no thirst or fatigue’ (0 mm), and the other end represented ‘extremely thirsty or completely exhausted’ (100 mm).

The Borg scale was used to determine the RPE as follows: extremely light (6–8), very light (9–10), light (11–12), somewhat hard (13–14), hard (15–16), very hard (17–18) and extremely hard (19–20) (Borg, 1982).

Cognitive performance

Participants completed four cognitive tests on the first visit, and before and after the exercise test in each trial during the second and third visits. These tasks were designed to evaluate various cognitive functions, including executive function, information processing and working memory.

Go/No‐Go and incongruent Stroop tests

Go/No‐Go and incongruent Stroop tests were used to assess executive function using the PsyToolkit (v.3.4.0) (Stoet, 2010, 2017). The Go/No‐Go test is considered to be relatively easy, whereas the incongruent Stroop test is considered to be relatively difficult (Watanabe, Kadokura et al., 2024).

In the Go/No‐Go test, target stimuli (green circles) or non‐target stimuli (red circles) were displayed for 1 s following a blank screen. Participants were instructed to press the spacebar on a computer keyboard with their right index finger as quickly as possible when the target stimulus appeared. They were required to withhold their responses when a non‐target stimulus appeared. The task consisted of 50 trials with equal distributions of the target (25) and non‐target (25) stimuli. Performance was evaluated based on the mean reaction time to the target stimuli and accuracy rate across 50 trials.

In the incongruent Stroop test, colour words (red, blue, green and yellow) were displayed on a computer screen, each printed in an incongruent ink colour (red, blue, green and yellow). To increase difficulty for Japanese‐native participants, words were presented in English. Participants were required to press one of the four coloured buttons (red, blue, green or yellow) on a computer keyboard corresponding to the semantic meaning of the word as quickly as possible. Word and ink colour pairings were randomly assigned, with 100% of the trials being incongruent. Each word was displayed for 1500 ms, followed by a 500 ms blank screen. Performance was evaluated based on the mean reaction time and accuracy rate across 50 trials.

Simple reaction task

A simple reaction task assesses information‐processing speed. The participants were first presented with a control stimulus (red screen) on a monitor, followed by a target stimulus (green screen) at random intervals. The participants were instructed to click on a computer mouse with their dominant index finger as quickly as possible when the target stimulus appeared. The task comprised five trials. Performance was assessed based on the mean reaction time across five trials.

Digit span test

The digit span test assessed working memory. A sequence of eight digits was displayed, with each digit appearing every 250 ms. After the final digit, the participants were required to enter the sequence in the correct order. The test comprised five trials. Performance was assessed based on the mean maximum number of correct answers across five trials.

Data analyses

Blood and urine sample analyses

For all participants, urine Na⁺ and osmotic excretion were calculated as follows (Post, after the exercise test; UV, urine volume in ml; SI, the sampling interval between pre‐ and post‐exercise tests in min):

$$\begin{array}{ccc}& & \mathrm{Urine}\mathrm{N}{\mathrm{a}}^{+}\mathrm{excretion}\left(\mathrm{mmol}/\min \right)\hfill \\ & & =\mathrm{UV}\times \mathrm{Post}\_{{\mathrm{U}}_{\mathrm{Na}}}^{+}/1000/\mathrm{SI}\hfill \end{array}$$

$$\begin{array}{ccc}& & \mathrm{Urine}\mathrm{osmotic}\mathrm{excretion}\left(\mathrm{mOsm}/\min \right)\hfill \\ & & =\mathrm{UV}\times \mathrm{Post}\_{\mathrm{U}}_{\mathrm{osmol}}/1000/\mathrm{SI}\hfill \end{array}$$

For participants with blood sampled, clearance values of the kidney for osmoles (C _osmol), free water (C _H2O), creatine (C _Cr) and Na⁺ (C _Na+), and fractional excretion of Na⁺ (FE_Na+) were calculated as follows (Pre, before the exercise test):

$$\begin{array}{ccc}\hfill {\mathrm{C}}_{\mathrm{osmol}}\left(\mathrm{mL}/\min \right)& =& \mathrm{Post}\_{\mathrm{U}}_{\mathrm{osmol}}/1000\times \mathrm{UV}/\hfill \\ & & \left[\left(\mathrm{Pre}\_{\mathrm{S}}_{\mathrm{osmol}}+\mathrm{Post}\_{\mathrm{S}}_{\mathrm{osmol}}\right)/2\right]/\mathrm{SI}\hfill \end{array}$$

$${\mathrm{C}}_{\mathrm{H}2\mathrm{O}}\left(\mathrm{mL}/\min \right)=\mathrm{UV}/\mathrm{SI}-{\mathrm{C}}_{\mathrm{osmol}}$$

$$\begin{array}{ccc}\hfill {\mathrm{C}}_{\mathrm{Cr}}\left(\mathrm{mL}/\min \right)& =& \mathrm{Post}\_{\mathrm{U}}_{\mathrm{Cr}}\times \mathrm{UV}/\hfill \\ & & \left[\left(\mathrm{Pre}\_{\mathrm{S}}_{\mathrm{Cr}}+\mathrm{Post}\_{\mathrm{S}}_{\mathrm{Cr}}\right)/2\right]/\mathrm{SI}\hfill \end{array}$$

$$\begin{array}{ccc}\hfill {{\mathrm{C}}_{\mathrm{Na}}}^{+}\left(\mathrm{mL}/\min \right)& =& \mathrm{Post}\_{\mathrm{U}}_{\mathrm{Na}}+\times \mathrm{UV}/\hfill \\ & & \left[\left(\mathrm{Pre}\_{\mathrm{S}}_{\mathrm{Na}}++\mathrm{Post}\_{\mathrm{S}}_{\mathrm{Na}}+\right)/2\right]/\mathrm{SI}\hfill \end{array}$$

$$\mathrm{F}{{\mathrm{E}}_{\mathrm{Na}}}^{+}={{\mathrm{C}}_{\mathrm{Na}}}^{+}/{\mathrm{C}}_{\mathrm{Cr}}\times 100$$

U_Na ⁺/U_K ⁺ was also calculated for the samples before and after the exercise test.

${\dot{V}}_{{\mathrm{O}}_2}$, HR, ABP, SkBF, and SR

${\dot{V}}_{{\mathrm{O}}_2}$ was normalised by the nude body weight obtained on the first day and expressed as the percentage of ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ (%${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$). Mean arterial blood pressure (MAP) was calculated from the systolic and diastolic blood pressures. Cutaneous vascular conductance (CVC) was derived from SkBF/MAP. SkBF, CVC and SR were expressed as the difference (Δ) from the baseline value. HR, ΔSkBF, ΔCVC, ΔSR, T _ear, mean T _skin and %${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ were averaged over 1 min before the end of a 30 min rest, the end of exercise at each set, and the end of an 8 min rest as the baseline, 1st, 2nd and 3rd exercise set, and recovery values, respectively.

Statistical analyses

Data are presented as the mean ± SD unless stated otherwise. A paired t test was used to compare total sweat volume, IVC/Ao index, C _osmol, C _H2O, C _Cr, C _Na+, FE_Na+, osmotic excretion, Na⁺ excretion and total urine volume between the two dietary conditions. Two‐way ANOVA with repeated measures (dietary condition × time) was used to analyse the other variables. When a significant interaction or main effect was detected, post hoc multiple comparisons were performed to identify pairwise differences using paired t tests with Bonferroni correction. A Pearson product–moment correlation coefficient was calculated to determine the relationship between osmotic excretion and total urine volume and between Na⁺ excretion and total urine volume. Statistical analyses were conducted using SPSS software (SPSS Statistics 27; IBM Corp., Armonk, NY, USA), with a significance set at P < 0.05. Effect sizes were calculated as eta squared (η²) for two‐way ANOVA outcomes and Cohen's d values for paired t tests (Lakens, 2013). The effect size was interpreted conservatively as small (η² = 0.01, Cohen's d = 0.20), medium (η² = 0.06, Cohen's d = 0.50) or large (η² = 0.14, Cohen's d = 0.80) (Lakens, 2013). Except for the P values, the statistical results are presented in the figures or tables.

Results

Hydration state

Nude body weight and sweat volume

A significant time effect was observed in body weight, without any effect of dietary condition or interaction (dietary condition, P = 0.686; time, P < 0.001; interaction, P = 0.998; Table 2). Post hoc analysis revealed a significant decrease in body weight after the experiment (P < 0.001). There was no significant difference in sweat volume between CON (531 ± 226 ml) and RED (531 ± 177 ml; P = 0.998, Cohen's d = 0.001).

Table 2.

Hydrated state assessed nude body weight and blood and urine samples

		Time			Two‐way ANOVA
Variable	Dietary condition	Before	After		Dietary condition	Time	Interaction
Nude body weight (kg)	CON	53.31 ± 7.65	52.78 ± 7.46	df	1, 14	1, 14	1, 14
	RED	53.28 ± 7.58	52.75 ± 7.43	F	0.17	109.4	<0.001
				P	0.686	<0.001	0.998
				η2	<0.001	0.682	<0.001
Serum
Osmolality (mOsm/kgH2O)	CON	278 ± 8	281 ± 3	df	1, 10	1, 10	1, 10
(n = 11)	RED	273 ± 4	279 ± 4	F	8.764	24.569	1.931
				P	0.014	<0.001	0.195
				η2	0.105	0.283	0.037
Urea nitrogen (mg/dl)	CON	9.3 ± 1.8	9.5 ± 1.9	df	1, 10	1, 10	1, 10
(n = 11)	RED	8.2 ± 1.6	8.4 ± 1.5	F	5.327	6.157	0.027
				P	0.044	0.032	0.873
				η2	0.110	0.012	<0.001
Glucose (mg/dl)	CON	125 ± 35	136 ± 64	df	1, 10	1, 10	1, 10
(n = 11)	RED	94 ± 18	121 ± 13	F	3.781	9.016	1.311
				P	0.080	0.013	0.279
				η2	0.083	0.109	0.020
Aldosterone (pg/dl)	CON	86 ± 37	132 ± 71	df	1, 10	1, 10	1, 10
(n = 11)	RED	110 ± 46	155 ± 121	F	1.389	5.183	<0.001
				P	0.266	0.046	0.992
				η2	0.024	0.160	<0.001
Sodium (mmol/l)	CON	137 ± 4	137 ± 4	df	1, 10	1, 10	1, 10
(n = 11)	RED	135 ± 5	137 ± 4	F	0.380	18.966	4.329
				P	0.551	0.001	0.064
				η2	0.015	0.062	0.017
Potassium (mmol/l)	CON	3.7 ± 0.2	4.1 ± 0.2	df	1, 10	1, 10	1, 10
(n = 11)	RED	5.1 ± 4.2	4.1 ± 0.2	F	1.121	0.312	1.103
				P	0.315	0.589	0.318
				η2	0.026	0.010	0.034
Chlorine (mmol/l)	CON	102 ± 3	105 ± 4	df	1, 10	1, 10	1, 10
(n = 11)	RED	100 ± 6	105 ± 5	F	0.492	48.895	2.126
				P	0.499	<0.001	0.175
				η2	0.012	0.340	0.015
Urine
Osmolality (mOsm/kgH2O)	CON	179 ± 89	355 ± 197*†	df	1, 14	1, 14	1, 14
(n = 15)	RED	209 ± 118	227 ± 93	F	1.618	10.356	9.190
				P	0.224	0.006	0.009
				η2	0.030	0.153	0.103
Urea nitrogen (mg/dl)	CON	233.7 ± 139.8	370.9 ± 225.6*	df	1, 10	1, 10	1, 10
(n = 11)	RED	284.8 ± 193.5	232.2 ± 108.4	F	0.403	0.630	9.705
				P	0.540	0.446	0.011
				η2	0.016	0.018	0.092
Creatine (mg/dl)	CON	24.7 ± 17.5†	60.0 ± 24.4*	df	1, 10	1, 10	1, 10
(n = 11)	RED	48.0 ± 34.2	45.0 ± 32.5	F	0.242	2.041	10.13
				P	0.633	0.184	0.010
				η2	0.005	0.087	0.122
Sodium (mmol/l)	CON	50 ± 32	142 ± 115*†	df	1, 14	1, 14	1, 14
(n = 15)	RED	65 ± 47	63 ± 37	F	3.579	9.094	21.007
				P	0.079	0.009	<0.001
				η2	0.047	0.148	0.159
Potassium (mmol/l)	CON	14.1 ± 8.3†	44.6 ± 34.5*	df	1, 14	1, 14	1, 14
(n = 15)	RED	22.6 ± 13.1	34.4 ± 20.5*	F	0.021	15.889	6.669
				P	0.887	0.001	0.022
				η2	<0.001	0.285	0.055
Chlorine (mmol/l)	CON	53 ± 34	174 ± 134*†	df	1, 14	1, 14	1, 14
(n = 15)	RED	77 ± 60	87 ± 59	F	2.837	9.681	20.188
				P	0.114	0.008	<0.001
				η2	0.030	0.197	0.145
Urine specific gravity	CON	1.005 ± 0.003	1.010 ± 0.006*†	df	1, 14	1, 14	1, 14
(n = 15)	RED	1.006 ± 0.004	1.005 ± 0.003	F	1.300	4.307	10.882
				P	0.273	0.057	0.005
				η2	0.028	0.068	0.138
UNa+/UK+ ratio	CON	4.35 ± 2.7	3.03 ± 1.18	df	1, 10	1, 10	1, 10
(n = 11)	RED	3.20 ± 1.80	2.23 ± 1.17	F	6.788	5.912	0.335
				P	0.026	0.035	0.575
				η2	0.066	0.190	0.005

Data are shown as mean ± SD. ^* P < 0.05 vs. before and ^† P < 0.05 vs. RED. The results indicated that RED induced body fluid hypotonicity before exercise, as assessed by lower serum osmolality, without altering urine concentration. Moreover, body fluid hypotonicity was still observed after exercise, particularly owing to changes in urinary sodium excretion.

Blood and urine samples

Blood sample

A significant dietary condition and time effects were observed in S_osmol and S_UN without interaction effect (dietary condition, both P ≤ 0.044; time, P ≤ 0.032; interaction, P ≥ 0.195; Table 2). Post hoc analysis revealed that S_osmol and S_UN were lower in RED than in CON and increased after the exercise. Moreover, a time effect was observed in S_Na+ and S_Cl− without any effect of dietary condition or interaction (dietary condition, both P ≥ 0.449; time, P ≤ 0.001; interaction, P ≥ 0.064; Table 2), while no significant effects were observed in S_K+ (dietary condition, P = 0.315; time, P = 0.589; interaction, P = 0.318; Table 2). Post hoc analysis revealed that S_Na+ and S_Cl− increased after the exercise test. A time effect was observed in S_glu and S_ald without any effect of dietary condition or interaction (dietary condition, both P ≥ 0.080; time, P ≤ 0.046; interaction, P ≥ 0.279; Table 2). S_AVP for all samples obtained in CON and RED was lower than the detection level (23.4 pg/ml). Therefore, the data are not presented.

Urine sample

No significant difference in total urine volume was observed between CON and RED (P = 0.083; Fig. 2A ). A significant interaction effect was observed on U_osmol, U_UN, U_Cr, U_Na+, U_K+, U_Cl− and USG (dietary condition, all P ≥ 0.079; time, P ≤ 0.446; interaction: P ≤ 0.022; Table 2). Post hoc analysis revealed that U_Cr and U_K+ were lower in RED than in CON before exercise (both P ≤ 0.046) but none of the other variables (all P ≥ 0.136). After exercise, U_osmol, U_UN, U_Cr, U_Na+, U_Cl− and USG increased only in CON (all P ≤ 0.043), with values of U_osmol, U_Na+, U_Cl− and USG being greater than those observed in RED (all P ≤ 0.033). In contrast, U_K+ increased in CON and RED (both P ≤ 0.034), with comparable values between the conditions (P = 0.267).

Figure 2. Grouped data of the total urine volume after exercise (A), Na⁺ excretion (B) and osmotic excretion (C) under two conditions: sufficient (CON, white circle) and reduced dietary intake (RED, black circle), with comparable and adequate water intake in both conditions.

The box plot represents the 25th, median and 75th percentiles, with individual data points and the minimum and maximum values indicated as horizontal lines outside the box. Correlation between Na⁺ excretion and urine volume, as well as between osmotic excretion and urine volume, in CON and RED. The results indicate that RED led to decreased urine Na⁺ excretion without altering the total urine volume or osmotic excretion. Nevertheless, a relationship was observed where urine volume was influenced by the magnitude of urine Na⁺ excretion suppression.

Urine Na⁺ excretion was lower in RED than in CON (P = 0.028; Fig. 2B ), with no significant differences in C _osmol, C _H2O, C _Cr, C_Na+, FE_Na+ or osmotic excretion (all P ≥ 0.070; Table 3 and Fig. 2C ). Significant dietary condition and time effects were also observed in the U_Na+/U_K+ ratio (dietary condition, P = 0.026; time, P = 0.035; interaction, P = 0.575; Table 2). Post hoc analysis revealed that U_Na+/U_K+ was lower in RED than in CON, while U_Na+/U_K+ decreased after the exercise test.

Table 3.

Calculated clearance of osmolality, free water, creatin and sodium (C _osmo, C _H2O, C _Cr and C _Na+) as well as fractional excretion of sodium (FE_Na+)

	Dietary condition		t test
Variable	CON	RED	P value	Cohen's d
C osmo	0.88 ± 0.34	0.82 ± 0.42	0.719	0.112
C H2O	0.05 ± 0.63	0.39 ± 0.59	0.253	0.366
C Cr	71.5 ± 28.2	64.7 ± 24.72	0.422	0.252
C Na+	0.77 ± 0.54	0.56 ± 0.44	0.070	0.483
FENa+	1.12 ± 0.76	1.00 ± 0.89	0.313	0.151

Data are mean ± SD. Values were calculated from the relevant data obtained before and after the exercise, as explained in the Methods. These results indicate that the clearance capacity was preserved in RED.

Pearson product–moment correlation coefficient analysis revealed a significant correlation between Na⁺ excretion vs. urine volume in RED (r = 0.679, P = 0.005) but not in CON (r = 0.057, P = 0.840; Fig. 2D ). In contrast, no significant correlations were observed between osmotic excretion and urine volume in both trials (CON, r = 0.063, P = 0.824; RED, r = 0.474, P = 0.074; Fig. 2E ).

IVC/Ao index

No significant difference was observed in the IVC/Ao index between CON (1.1 ± 0.1) and RED (1.1 ± 0.2) (P = 0.160, Cohen's d = 0.383).

Energy metabolism (%${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ and RQ)

Significant dietary condition and time effects were observed in %${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$, with no interaction effect (dietary condition, P = 0.003; time, P < 0.001; interaction, P = 0.088; Fig. 3A ). Post hoc analysis revealed that %${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ was higher in RED than in CON. Additionally, %${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ increased from the resting baseline to the 1st exercise set (P < 0.001), remained stable until the 3rd exercise set (all P = 1.000) and decreased during the recovery (P < 0.001) but remained higher than the baseline value (P = 0.003).

Figure 3. Grouped data of oxygen uptake (%${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$, A) and respiratory quotient (RQ) during sufficient (CON, white circle) and reduced dietary intake (RED, black circle) conditions, with comparable and adequate water consumption in both conditions.

The box plot represents the 25th, median and 75th percentiles, with individual data points and the minimum and maximum values indicated as horizontal lines outside the box. The results indicate that RED altered the energy source from glycogen to fat, as assessed by increased %${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ and decreased RQ.

A significant interaction effect was observed in RQ (dietary condition, P = 0.032; time, P = 0.006; interaction, P = 0.025; Fig. 3B ). Post hoc analysis revealed that RQ was lower in RED than in CON during the exercise (all P ≤ 0.049).

Other variables

No significant interaction or condition effects were observed in the variables associated with systemic circulation and autonomic thermoregulatory responses (HR, MAP, SkBF, CVC and SR), body temperature (T _ear and mean T _skin), VAS‐assessed subjective perception or cognitive performance (all P > 0.050; Figs 4, 5, 6 and Table 4).

Figure 4. Grouped data of heart rate (HR, A); mean arterial pressure (MAP, B); differences in the skin blood flow of the chest (SkBF_chest, C) and forearm (SkBF_forearm, D); differences in cutaneous vascular conductance of the chest (CVC_chest, E) and forearm (CVC_forearm, F); and differences in sweat rate of the chest (SR_chest, G) and forearm (SR_forearm, H) during sufficient (CON, white circle) and reduced dietary intake (RED, black circle) conditions, with comparable and adequate water consumption in both conditions.

The box plot represents the 25th, median and 75th percentiles with individual data points, and the minimum and maximum values are indicated by the horizontal lines outside the box.

Figure 5. Grouped data of ear canal (T _ear, A) and mean skin surface temperature (T _skin, B) during sufficient (CON, white circle) and reduced dietary intake (RED, black circle) conditions.

The box plot represents the 25th, median and 75th percentiles with individual data points, and the minimum and maximum values are indicated by the horizontal lines outside the box.

Figure 6. Grouped data of the visual analog scale (VAS)‐assessed thermal (A) and humid sensations (B), thermal comfort (C), physical (D) and psychological fatigue (E), and rating of perceived exertion (RPE, F) during sufficient (CON, white circle) and reduced dietary intake (RED, black circle) conditions.

The box plot represents the 25th, median and 75th percentiles with individual data points, and the minimum and maximum values are indicated by the horizontal lines outside the box.

Table 4.

Cognitive performance

		Time			Two‐way ANOVA
Variable	Dietary condition	Before	After		Dietary condition	Time	Interaction
Go/No‐Go task
Correct answer rate (%)	CON	98.9 ± 1.5	99.2 ± 1.7	df	1, 14	1, 14	1, 14
	RED	99.3 ± 1.2	99.9 ± 0.5	F	1.882	1.556	0.318
				P	0.192	0.233	0.582
				η2	0.042	0.033	0.004
Response time (ms)	CON	328.9 ± 28.3	315.0 ± 27.3	df	1, 14	1, 14	1, 14
	RED	318.7 ± 16.8	313.2 ± 31.3	F	1.181	4.629	1.35
				P	0.296	0.049	0.265
				η2	0.013	0.09	0.017
Incongruent Stroop task
Correct answer rate (%)	CON	94.9 ± 4.5	94.1 ± 3.9	df	1, 14	1, 14	1, 14
	RED	95.9 ± 2.8	94.1 ± 4.6	F	0.546	2.573	0.604
				P	0.472	0.131	0.450
				η2	0.004	0.075	0.010
Response time (ms)	CON	585.2 ± 56.4	583.7 ± 66.9	df	1, 14	1, 14	1, 14
	RED	592.7 ± 60.2	573.8 ± 72.2	F	0.018	2.571	2.381
				P	0.895	0.131	0.145
				η2	<0.001	0.046	0.033
Digit span test
Number of correct answers	CON	6 ± 2	6 ± 1	df	1, 14	1, 14	1, 14
	RED	7 ± 1	6 ± 1	F	0.617	0.035	2.224
				P	0.445	0.854	0.158
				η2	0.007	0.001	0.039
Simple reaction test
Response time (ms)	CON	308.7 ± 24.0	297.5 ± 23.8	df	1, 14	1, 14	1, 14
	RED	303.6 ± 21.2	302.9 ± 21.1	F	0.001	1.671	1.786
				P	0.978	0.217	0.203
				η2	<0.001	0.036	0.028

Data are shown as mean ± SD. The results indicated that RED did not impair cognitive function.

HR, MAP, SkBF, CVC and SR

A significant time effect was observed in HR, without any effect of dietary condition or interaction (dietary condition, P = 0.140; time, P < 0.001, interaction, P = 0.599; Fig. 4A ). Post hoc analysis revealed that HR consistently increased from the resting baseline to the 3rd exercise set (all P < 0.001) and decreased during recovery (P < 0.001) but remained higher than the baseline (P < 0.001). In contrast, no significant effects were observed on MAP (dietary condition, P = 0.064; time, P = 0.559; interaction, P = 0.637; Fig. 4B ).

A significant time effect was observed on ΔSkBF and ΔCVC in the chest and arm, without any effect of dietary condition or the interaction (dietary condition, all P ≥ 0.228, time, P < 0.001, interaction, P ≥ 0.192; Fig. 4C–F ). In the chest, post hoc analysis revealed that ΔSkBF_chest and ΔCVC_chest increased from baseline to the 1st exercise set (both P < 0.001), remained stable until the 3rd exercise set (all P ≥ 0.721) and returned to the baseline level during recovery (both P ≥ 0.397). In the forearm, ΔSkBF_forearm and ΔCVC_forearm increased from baseline to the 1st exercise set, remained stable until the 3rd exercise set and decreased during recovery but remained higher than the baseline (both P < 0.001).

Similarly, a significant time effect was observed on ΔSR in the chest and forearm, without any effect of dietary condition or interaction (dietary condition, both P ≥ 0.347; time, P < 0.001; interaction, P ≥ 0.202; Fig. 4G and H ). ΔSR_chest increased from baseline to the 2nd exercise set (both P < 0.001), remained stable until the 3rd exercise set (P = 1.000) and decreased during recovery (P < 0.001) but remained higher than the baseline (P < 0.001). ΔSR_forearm also increased from baseline to the 2nd exercise set (both P < 0.001), remained stable until the 3rd exercise set (P = 1.000) and decreased during recovery (P < 0.001) but remained higher than baseline (P < 0.001).

T _ear and T _skin

A significant time effect was observed on T _ear and mean T _skin without any effect of dietary condition or interaction (dietary condition, both P ≥ 0.133: time, P < 0.001; interaction, P ≥ 0.070; Fig. 5). Post hoc analysis revealed that T _ear increased from baseline to the 3rd exercise set (all P ≤ 0.013), remained stable during recovery (P = 0.218) and remained higher than baseline (P < 0.001). Mean T _skin increased from baseline to the 2nd exercise set, remained stable during recovery (P = 1.000) and remained higher than baseline (P < 0.001).

VAS and RPE

A significant time effect was observed on thermal and humid sensations, thermal comfort, and physical and psychological fatigue, without any effect of dietary condition or interaction (dietary condition, all P ≥ 0.134; time, P < 0.001; interaction, P ≥ 0.135; Fig. 6A–E ). Post hoc analysis revealed that thermal and humid sensations and thermal discomfort increased from baseline to the 2nd exercise set (all P ≤ 0.029), remained stable until the 3rd exercise set (all P ≥ 0.123) and returned to baseline level during recovery (all P ≥ 0.360). Physical fatigue increased from baseline to the 3rd exercise set (all P < 0.001), remained stable until recovery (P = 0.579) and remained higher than that at baseline (P < 0.001). Additionally, psychological fatigue increased from baseline to the 2nd exercise set (all P = 0.033), remained stable until recovery (both P ≥ 0.072) and remained higher than baseline (P = 0.011).

A significant time effect was observed on RPE without any effect of dietary condition or interaction (dietary condition, P = 1.000; time, P < 0.001; interaction, P = 0.778; Fig. 6F ). RPE increased from the 1st to the 3rd exercise set (all P < 0.001).

A significant time effect was observed on thirst sensation, without any effect of the dietary condition or interaction (CON, 45 ± 23 and 67 ± 20; RED, 45 ± 17 and 72 ± 20 before and after exercise, respectively) (dietary condition, F _1,14 = 0.097, P = 0.760, η² = 0.003; time, F _1,14 = 38.473, P < 0.001, η² = 0.338; interaction, F _1,14 = 0.637, P = 0.438, η² = 0.003). The intensity of the thirst sensation increased after completion of the experiment.

Cognitive performance

A significant time effect was observed on the response time of the Go/No‐Go task without condition and interaction effects (dietary condition, P = 0.296; time, P = 0.049; interaction, P = 0.265; Table 4), whereas no significant effects were observed on other variables of the cognitive tasks (dietary condition, all P ≥ 0.192; time, P ≥ 0.131; interaction, P ≥ 0.145; Table 4). After the exercise, post hoc analysis revealed a faster response time for the Go/No‐Go task.

Discussion

Main findings

This study tested the hypothesis that overnight dietary restriction, which consisted of small dinner portions and skipped breakfast, induced isotonic dehydration despite adequate water consumption, and subsequently impaired thermoregulatory responses and cognitive function. Contrary to the hypotheses, (1) overnight dietary restriction resulted in body fluid hypotonicity; (2) led to reduced urinary concentrating capacity during exercise, as evidenced by the inability to augment urine osmolality, creatinine and sodium concentration; and (3) neither thermoregulatory response nor cognitive performance was impaired. These results suggest that despite adequate water consumption, transient dietary restriction can alter fluid balance, reflecting early signs of impaired homeostasis, but does not compromise thermoregulatory responses or cognitive performance, at least during relatively short‐duration exercises.

Hydration status and fluid balance

As shown in Table 1, total water consumption from meal and water bottles was equivalent between the two trials, whereas total caloric intake was 46% lower in RED than in CON. Among the nutrients, carbohydrates were 38% lower in RED than in CON. Regarding electrolytes, NaCl intake was lower in RED than in CON by 20%, whereas potassium and phosphorus intakes were lower in RED by 55% and 20%, respectively. Given that the absolute daily NaCl intake in both conditions was relatively low (CON, 4.7 g; RED, 3.8 g) compared with the average in many populations (∼11 g/day) (World Health Organization, 2025), and that the between‐condition difference was modest, the impact of sodium restriction is probably limited. Moreover, considering that the physiological requirement of NaCl is approximately 2 g/day as recommended by the World Health Organization, both conditions in this study exceeded this minimum requirement. Therefore, sodium restriction may have played only a minor role; rather, it is more plausible that reduced carbohydrate intake or other RED‐related physiological factors were the main trigger of the observed changes.

Before the exercise

RED resulted in lower S_osmol while maintaining body water content and intravascular fluid volume, as assessed by nude body weight and IVC/Ao index. These results indicate that overnight dietary restriction with adequate water consumption led to body fluid hypotonicity (Table 2). However, no significant differences between CON and RED in variables representing renal functions (serum and urinary electrolytes, non‐electrolytes and hormones; Table 2) and autonomic nerve activity (i.e. HR and MAP; Fig. 4A and B , respectively) were observed between CON and RED. In addition, S_AVP was below the detection threshold under both conditions.

Although nude body weight and IVC/Ao index were comparable between the two trials, subtle shifts in the effective circulating volume might have triggered hormonal or autonomic responses. For instance, reduced blood volume unloads cardiopulmonary volume receptors, leading to decreased vagal afferent input to the nucleus tractus solitarius and increased renal sympathetic nerve activity, which, in turn, suppresses ANP secretion, activates the RAAS, and promotes renal sodium and water retention (Antunes‐Rodrigues et al., 2004; DiBona, 2000; Navar, 2014).

Fasting reportedly induces natriuresis (Maoz et al., 1992), but its influence on ANP‐related mechanisms is complex because fasting appears to exert opposing regulatory effects on ANP dynamics in different tissues (Gower et al., 2000; Sarzani et al., 1995). For example, in the gastric antrum, ANP gene expression is markedly suppressed during fasting, presumably because of reduced gastric distension (Gower et al., 2000). In contrast, fasting decreases the expression of the clearance receptor NPr‐C in adipose tissue, which may enhance circulating ANP activity by limiting its degradation (Sarzani et al., 1995). This tissue‐specific bidirectional regulation may fine‐tune ANP‐mediated natriuresis and fluid balance during dietary restriction. Additionally, SGLT‐2 contributes to Na⁺ reabsorption in the proximal tubules, and reduced circulating glucose levels may attenuate Na⁺ and water reabsorption (Heyman et al., 2020).

Although these mechanisms may have contributed to the observed shift in body fluid hypotonicity, no significant differences were observed in S_ald, A_AVP or U_Na+ between CON and RED before exercise (Table 1). These results suggest that, whereas major sodium‐ and water‐regulatory hormones were similarly engaged in both conditions, the observed lower S_osmol in RED may reflect subtle shifts in water–solute balance that are not captured by the measured variables. Body fluid hypotonicity cannot be fully explained by sodium and water regulatory mechanisms, and its physiological basis remains to be elucidated. In addition, the reason for the increase in U_K⁺ in RED is unclear, as S_K⁺ levels did not differ significantly between conditions (Table 2), suggesting a possible shift in renal K⁺ handling under dietary restriction.

During and after exercise

After exercise, the increase in S_osmo under both conditions can be attributed primarily to fluid loss through sweating (Table 2). Despite stable S_Na+, S_K+ and S_Cl−, as well as comparable amounts of total sweat volume and urine output after exercise between the conditions (Table 2 and Fig. 2A ), the increases in U_osmol and urine electrolyte concentrations observed in CON were absent in RED (Table 2). These results indicate that body water and electrolyte regulation in RED differed from that in CON. Among electrolytes, sodium plays a crucial role in maintaining the body's water balance by redistributing intracellular and extracellular fluids. In RED, U_Na+ remained unchanged from and after exercise (Table 1), and urine Na⁺ excretion was lower than in CON (Fig. 2).

Dynamic exercise enhances the central blood volume through respiratory and muscle pumps, stimulating ANP secretion and suppressing activation of the RAAS (De Paoli Vitali et al., 1991; Follenius & Brandenberger, 1988; Mannnix et al., 1990). However, under heat stress, RAAS activation often overrides the effects of ANP, thereby promoting sodium retention (De Paoli Vitali et al., 1991; Melin et al., 2001). This occurs via reduced central volume and renal perfusion owing to vasodilatation and sweat loss. Although aldosterone promotes Na⁺ reabsorption in renal distal tubules (Heyman et al., 2020; Maoz et al., 1992; Schloeder & Stinebaugh, 1970; Spark et al., 1975; Szenasi et al., 1985), no significant difference in S_ald was observed between CON and RED, suggesting that RASS does not explain lower Na⁺ excretion in RED.

Although ANP was not measured in this study, comparable nude body weight suggests no substantial difference in central volume between conditions; however, subtle differences in volume‐sensitive hormonal responses remain possible, as previously discussed. Indeed, S_AVP was below the detection threshold under both conditions, indicating the minimal activation of ADH. Although the urine concentrating ability was impaired under RED, the total urine volume during exercise did not differ between the conditions. If AVP were a dominant factor, increased urine output would be expected in RED. The absence of this difference suggests that AVP (or ADH) does not play a central role in water regulation.

The most plausible alternative mechanism is metabolic adaptation. In RED, a reduced RQ and increased %${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ at the same workload (Fig. 3) indicated a shift towards enhanced lipid oxidation to compensate for reduced carbohydrate utilization (Péronnet & Massicotte, 1991). This metabolic shift promotes ketone body production, which can lead to ketoacidosis‐induced systemic pH reduction (Laffel, 1999). In such an acidic state, hydrogen ions are transported out of the cell via the Na⁺/H⁺ exchanger, facilitating Na⁺ uptake into the cell. This mechanism may explain the differences in urinary electrolyte concentrations between RED and CON.

Furthermore, a significant correlation was observed between urine Na⁺ excretion and volume in RED but not in CON (Fig. 2D ), suggesting that RED modulates sodium processing and urine regulation. These findings highlight the dynamic nature of fluid balance regulation under dietary restriction and emphasise the critical role of electrolytes in maintaining water homeostasis.

Effects on systemic circulation and thermoregulatory responses

Despite the regulatory alterations in body water and electrolytes in RED, the responses of HR, MAP, SkBF and SR were comparable to those in CON, resulting in no difference in T _core and mean T _skin responses between the conditions (Figs 4 and 5). This suggests that systemic circulation and autonomic thermoregulation were maintained even with overnight dietary restrictions and adequate water consumption. Given the relatively slight hypotonic shift of body fluid and sufficient body water content before exercise, this maintenance is plausible. Furthermore, although energy substrates are related to thermoregulatory responses, during RED, a certain amount of energy substrates were provided. Additionally, glycogen is stored in the liver and muscles (Adeva‐Andany et al., 2016). Therefore, the systemic circulation and thermoregulatory function could be preserved through energy availability and water content in RED.

Cognitive performance and subjective thermal and exertional perceptions

RED did not impair subjective thermal perceptions, fatigue (Fig. 6) or cognitive performance (Table 4). The faster response time in the Go/No‐Go task after exercise was observed in both conditions, indicating a general exercise‐induced enhancement rather than a fasting‐specific influence (Basso & Suzuki, 2017; Chang et al., 2012). Brain function underlying subjective perceptions and cognitive functions relies on water and electrolyte balance to regulate neuronal excitability, synaptic transmission and network activity (Jéquier & Constant, 2010). Theoretically, it is plausible that body fluid hypotonicity could alter osmotic gradients and impair neuronal and glial functions, thereby affecting sensory and cognitive processes. However, in this study, the brain appeared to maintain normal processing, probably because the hypotonic shift induced by RED was minimal. Furthermore, as the dietary restriction duration was short and meals were consumed, glycogen stores may not be depleted to levels that impair brain functions. Additionally, compensatory mechanisms, such as increased availability of ketone bodies (Laffel, 1999), could help sustain the function.

Limitations

This study has several limitations. First, the duration of dietary restriction was relatively short, and longer durations may result in different outcomes, particularly regarding thermoregulatory and cognitive functions. Second, given that the absolute daily NaCl intake in both conditions was relatively low compared with typical population averages, the results might differ under more common intake levels. Third, although participants were required to maintain consistent water consumption, individual variations in hydration behaviours may have influenced the results. Fourth, this study focused on young, healthy individuals, and the findings may not be generalisable to older adults, patients or athletes. Fifth, sweat sodium content was not measured to assess sodium reabsorption capacity. Finally, the indices of total plasma/blood volume and/or body water were not directly measured; however, the absence of changes in nude body weight, IVC/Ao index, total urine volume and autonomic thermoregulation in RED may indirectly reflect their maintenance to some extent.

Conclusion

Despite the preservation of thermoregulatory and cognitive responses even with adequate water consumption, we found that overnight dietary reduction, especially reduced carbohydrate intake, led to subtle shifts toward body fluid hypotonicity and altered body water and electrolyte regulatory mechanisms during relatively short‐duration exercises. These changes potentially reflect early signs of impaired homeostasis, which regulates the water–electrolyte balance. These results suggest that adequate dietary intake, in addition to hydration, may contribute to maintaining physiological homeostasis under heat stress and should not be overlooked in strategies to prevent heart illness and stroke, particularly during long‐term physical activity.

Competing interests

Y.K. is an employee of Toshiba Corporation. The authors declare that they have no additional conflict of interest related to the subject matter or materials discussed in this study. All data in the present study are presented clearly and honestly without fabrication, falsification or inappropriate data manipulation.

Author contributions

H.W.: Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualisation. T.O.: Investigation, Formal analysis, Data curation. Y.K.: Writing – review & editing, Methodology. K.S.: Supervision, Resources. K.N: Writing – review & editing, Supervision, Methodology, Investigation, Funding acquisition, Conceptualisation.

Funding

This work was supported by Toshiba Corporation and the Japan Society for the Promotion of Science KAKENHI (Grant‐in‐Aid for Scientific Research A; JP 19H01128) awarded to K.N. We also thank Editage (www.editage.com) for English language editing.

Supporting information

Peer Review History

Biographies

Hironori Watanabe is a researcher at Waseda University. His research explores human thermoregulatory systems by examining respiratory, circulatory and neural dynamics. This study was conducted with Professor Nagashima. Their relationship is reminiscent of a Jedi apprentice and Master from Star Wars, united by a shared mission, and they aim to create a new wave in thermophysiology. Dr Watanabe is committed to applying scientific knowledge to real‐world solutions that advance public health and environmental adaptability.

Kei Nagashima has conducted integrative physiology. By bridging neuroscience, behavioural biology and clinical physiology, he aimed to deepen our understanding of how organisms sense and adapt to thermal challenges. The insights into heat acclimation mechanisms and neural coding of thermal pleasantness have potential applications in thermal comfort design, clinical rehabilitation and sports physiology.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.