Sixteen males, with no known history of kidney issues or diabetes, participated in the present study. Participants were excluded from participation if they smoked/vaped, or if they had any current medical conditions/regularly used any medications (for example, non-steroidal anti-inflammatory drugs) that could influence the study outcomes. One subject developed an illness unrelated to the study and withdrew before completing the study, leaving 15 participants (age: 27 ± 5 y; height: 1.79 ± 0.07 m; body mass: 76.4 ± 9.3 kg; BMI: 24.1 ± 3.8 kg/m2). The Loughborough University Ethics Approvals (Human Participants) Sub-Committee granted this study full ethical approval (Reference: R18-P144).

Participants visited the laboratory on five occasions: for a screening visit and two experimental trials (two visits per experimental trial). Experimental trials were performed with a randomised cross-over design and involved 24 h of either euhydration (EU; 40 mL/kg body mass water ingestion) or fluid restriction (HYP; 100 mL water ingestion) followed by an oral glucose tolerance test (OGTT). Experimental trials were separated by ≥ 7 days.

After providing verbal and written informed consent, height and nude body mass (AFW-120 K, Adam Equipment Co., UK) were measured. If necessary, participants were familiarised to blood sampling procedures. The day prior to their first experimental trial, participants recorded their food and fluid intake, ensuring to consume ≥ 40 mL/kg body mass of non-alcoholic fluid (marked bottles were supplied to aid compliance), and avoided alcohol consumption and strenuous exercise. These were all then replicated the day before the second experimental trial. Participants were instructed to cease food and fluid consumption ≥ 10 h before arrival to all experimental trial visits. Pre-trial standardisation reminders were sent to participants two days before each experimental trial, and compliance was verbally checked when participants arrived at the laboratory.

On the morning of their first experimental trial, upon waking, participants put on a step counter wristwatch (TomTom Runner 3, TomTom, Amsterdam, The Netherlands). This would provide an approximate step count, which participants were asked to match on the first day of their second experimental trial to control habitual low-intensity physical activity between trials. Participants arrived at the laboratory between 5:45 and 9:30am (standardised within subject), provided a urine sample, and had nude body mass measured (baseline measurements). Participants then commenced seated rest and answered a series of subjective feelings questionnaires, including headache, nausea, dizziness, thirst, gastrointestinal (GI) comfort, and urge to vomit (0 = no symptom, 10 = maximum symptom). After 30 min, a blood sample was taken by venepuncture of an antecubital vein (baseline blood sample).

After eating a standardised breakfast (croissant with strawberry jam), participants were provided with urine collection equipment and their food and fluid for the rest of the day and left the laboratory. Food consisted of lunch (4 h post-baseline; cheese and butter sandwich, tortilla chips, and cereal bar), evening meal (8 h post-baseline; pizza), and evening snack (12 h post-baseline; dried apricots, raisins, and protein cereal bar). If a subject did not have access to an oven at 8 h post-baseline on their first trial, they were permitted to swap the times of their evening meal and evening snack, providing that this was matched for their second trial. The diet was designed to achieve energy balance, providing an energy intake approximately equal to estimated resting energy expenditure (Mifflin et al. 1990) multiplied by a physical activity level of 1.6 with approximately 50% of energy from carbohydrate, 35% from fat, and 15% from protein. In EU, participants consumed 7 mL/kg body mass water with each meal and 4 mL/kg body mass of water between meals, totalling 40 mL/kg body mass of water intake for the day (participants were provided with a water bottle with two markings to facilitate this, and filled bottles with tap water themselves). In HYP, participants were provided with 100 mL of water for the day, which was recommended to be consumed with their evening meal.

At home, 12 h post-baseline (just before their evening snack), participants collected and then refrigerated a urine sample. The following morning, 24 h after they first arrived at the laboratory, participants returned to the laboratory and provided a urine sample, followed by a nude body mass measurement. During seated rest, a cannula was inserted, and subjective feelings questionnaires were completed. After 30 min, a blood sample was taken from the cannula (24 h post-baseline). Participants were then instructed to ingest a solution of 75 g glucose dissolved in 250 mL water as quickly as possible, ensuring not to spill any. Once this was ingested, the cup was rinsed with 50 mL water, which was also ingested by the subject. A 10 mL blood sample was withdrawn at 15, 30, 45, 60, 90, and 120 min following the initiation of drinking. After each blood sample, the cannula was flushed with  ~ 10 mL saline.

All blood samples were  ~ 10.5 mL in volume. From this, 1 mL was dispensed into a tube containing K2EDTA (1.75 mg/L, Teklab, Durham, UK) and was used to determine changes in plasma volume from baseline (Dill and Costill 1974), using haemoglobin concentration (cyanmethaemoglobin method) and haemotocrit (microcentrifugation; Hawksley Microhematocrit Centrifuge, Hawksley, Worthing, UK). Another 5 mL was dispensed into a pre-chilled K2EDTA tube (1.6 mg/L, Sarstedt Ltd, Leicester, UK) and 4.5 mL was dispensed into a room temperature tube containing a clotting catalyst (Sarstedt Ltd, Leicester, UK). After ≥ 20 min, these two tubes were centrifuged (2200 g, 15 min, 4 °C), with the resultant plasma/serum stored at − 80 °C until analysis.

Plasma was used to determine glucose concentrations and serum was used to determine creatinine and uric acid concentrations (ABX Pentra C400; Horiba Medical, Northampton, UK), as well as osmolality by freezing point depression (Osmomat Auto, Cryoscopic Osmometer, Gonotec, Berlin, Germany). The intra-assay CV for plasma glucose, serum creatinine, and serum uric acid were 0.4, 1.6, and 0.5%, respectively. All urine samples had osmolality determined (Osmocheck; Vitech Scientific, Horsham, UK), before being aliquoted and stored at − 80 °C. They were later thawed to measure uNGAL (Human NGAL ELISA Kit, BioPorto, Hellerup, Denmark; CV: 10.0%) and uKIM-1 (KIM-1 Human ELISA Kit, Enzo Life Sciences, Lausen Switzerland; CV: 6.1%) concentrations, using commercially available ELISA kits. Plasma insulin was measured using a commercially available ELISA kit (Mercodia, Uppsala, Sweden; CV: 2.4%).

Manipulating hydration status can alter the concentrations of biomarkers by concentrating/diluting blood and urine. Therefore, serum/plasma biomarkers were corrected for changes in plasma volume and urine biomarkers were corrected for urine osmolality, with results presented in both uncorrected and corrected forms. For plasma glucose and insulin concentrations, incremental (iAUC) and total (tAUC) area under the curve (Narang et al. 2020), Matsuda insulin sensitivity index (ISI; Matsuda and DeFronzo 1999), and the homeostatic model of insulin resistance (HOMA-IR2; HOMA2 calculator v2.2.3, University of Oxford) were also determined. As blood samples were not collected from two participants, and were haemolysed in one subject, all blood measures are presented as n = 12.

Data were analysed using IBM SPSS statistics (version 27). Shapiro–Wilk tests were used to check data for normality of distribution. Paired t tests or Wilcoxon signed-rank tests, depending on normality, were used to analyse data containing one factor (trial). Data containing two factors (trial × time) were analysed using a two-way repeated-measures analysis of variance. If data violated the assumption of sphericity, the Greenhouse–Geisser adjustment was used. Significant effects, which were defined as P ≤ 0.05, were followed up with Holm–Bonferroni corrected t tests or Holm–Bonferroni corrected Wilcoxon signed-rank tests, as appropriate. Parametric data are presented as (mean ± SD), and non-parametric data are presented as (median [interquartile range]).

At the time of the study, there were no published data available to inform the effect size of manipulating hydration status, in the absence of exercise, on osmolality-corrected uNGAL and uKIM-1 concentrations. Therefore, in line with studies that have found significant effects when assessing the effects of manipulating hydration status during exercise on biomarkers of renal injury (Chapman et al. 2020; Juett et al. 2021), 16 participants were recruited.