The Effect of Hydration Status on Glycemic Control and Appetite Regulation

The aim of this study is to investigate whether hydration status affects blood sugar control and appetite regulation. This will be a randomised crossover trial with two trial arms separated by 5-35 days. Each trial arm consists of five days in total, with the only difference between the trial arms being the amount of fluid consumed after the heat tent on day 4 (as described below).

Days 1-3 of each trial arm are the monitoring phase. Participants will be given a chest mounted activity monitor (ActiHeart TM), some body mass scales, a food diary with food weighing scales and three urine sample pots. Participants will be required to provide a urine sample upon waking, and note their weight before eating. The activity monitor will need to be worn from midnight on day 1 in order to gauge three full 24-h periods of activity. The food diary will also start at this point; participants will be required to weigh and record all their food and beverages (including recording the leftovers). On day 3 of the monitoring phase, participants will be required to continue their normal daily routine, however they will be provided with a set amount of water to consume (40 mL/kg lean body mass) in order to standardize baseline hydration status. No food or fluid will be allowed to be consumed after 2200 h on day 3.

On day 4, participants will take the same measures (weight and urine sample) and will arrive at the university in a fasted state (i.e. no food or fluid). A blood sample will be drawn from an antecubital vein followed by a peripheral quantitative computer tomography (pQCT) scan of the thigh. After the blood sample and scan, participants may eat food from a list of low water-content foods only for the rest of the day. At approximately midday (flexible depending on the participants' availability, but matched on both trial arms), participants will be weighed again, and placed in a heat tent (40-50 degrees Celsius) until they lose 1 % of their pre-heat tent weight or for a maximum of one hour (whichever comes first). On the second trial arm, participants will be matched by percentage of body mass loss, even if this requires spending more or less time in the heat tent, but still no more than one hour.

After the heat tent, participants will receive a salty meal in order to facilitate fluid absorption/retention. This will be matched in both trial arms in order to reduce confounding. In the dehydration trial arm, participants will be provided with 3 mL/kg lean body mass of plain water to consume for the rest of the day. In the rehydration trial arm, participants will be given 40 mL/kg lean body mass plus 150 % of their body mass losses from the heat tent. This will be metered out throughout the day to maximise rehydration. Thus the amount of water consumed on day 4 after the heat tent is the only discernible difference between the trial arms, with everything else matched. In both trial arms only plain water (i.e. no caffeine or alcohol) will be allowed to be consumed, and only foods from the dry food list can be eaten. This will be matched on the subsequent trial arm. No food or fluid after 2200 h on day 4.

On the fifth day of each trial arm, participants will be required to attend the laboratory at approximately 0730 h (± 1 h). A urine sample will be collected, along with body mass. Another pQCT scan of the thigh will be taken and participants' metabolic rate will be measured using the Douglas bag method. As part of the trial, participants are able to opt-in for a muscle biopsy. If they have opted-in, a muscle biopsy will be taken after the pQCT scan. This will involve a local anesthetic (Lidocaine), making a small incision using a scalpel, then obtaining a muscle sample using the Bergstrom technique. Participants will then be fitted with a cannula, and their hand will be placed in a 60 degree Celsius hotbox for five minutes, after which a baseline blood sample will be drawn. Participants will be given a 75 g glucose drink to consume within five minutes. Whilst remaining rested, a researcher will draw a 10 mL blood sample every 15 minutes for 120 minutes, with approximately 10 mL saline infused after each blood sample drawn. Metabolic rate will be measured at 0 (pre-OGTT), 60 and 120 minutes. For the participants who opted-in for the muscle biopsy, another one will be taken after 120 minutes.

Participants will then be presented with some computer based appetite assessing psychological inventories to assess subconscious food preferences. These tests are pilot data for future research. Following this, participants will be presented with a pasta meal, with the instruction to eat until satisfied for a maximum of 30 minutes. After 30 minutes, a blood sample will be drawn, followed by 10 minutely samples for 60 minutes. Visual analogue scales for subjective appetite ratings will be taken pre-meal, immediately, post-meal, then at 30 and 60 minutes post-meal. The cannula will then be removed.

The sample size required for this trial is 16 participants, calculated using pilot data which assessed blood glucose control after being dehydrated and rehydrated. Based on that pilot study, hydration status resulted in a marked difference in the rate of return to fasted glycaemia. This primary conclusion was supported by a magnitude of effect 45 minutes post-oral glucose tolerance test (D = 1.1 mmol/l). The standard deviation at this time point in the control (normal hydration) group was also 1.1 mmol/l, resulting in an effect size (d) of ≈1. To provide a 95 % power (beta level) to detect this effect at an alpha level (p) of less than or equal to 0.05 using a 2--tailed paired t--test will require ~16 participants.

If any participant does not complete the study (i.e. they have missing data), their data will not be included in the analysis. In other words, the analysis plan will be per protocol (completers only). The statistical techniques used will be linear modelling with post-hoc t-tests corrected for multiple comparisons where relevant. However, until the final dataset has been obtained, the statistical tests may be subject to change (e.g. if data are non-parametric).