Mild Dehydration Led to Difficulty Falling Asleep and Increased Sleep Duration

Hypohydration has been shown to decrease physical performance [1], cognitive function [2], and overall quality of life. Losing more than 1–2% in body weight can be classified as a mildly dehydrated state [3]. According to Armstrong et al., 25–33% of the adult population are mildly dehydrated in the USA and Europe [4,5,6,7]. Mild dehydration can be achieved by consuming less than 1.5 L of water in a day [4,5,6,7] or when fluid balance is negative, which can be induced through exercise or a low water intake.

Similar to hydration, sleep can heavily influence physical and cognitive performance as well as alter mood and motor function [8]. The Centers for Disease Control and Prevention recommends that the average adult sleep at least 7 h per night; however, at least one-third of the adult population does not get the recommended amount of sleep [9]. Sleep is an essential part of human health, and sleep deprivation has been linked to depression [10], cardiovascular disease [11], and Alzheimer’s [12].

Previous studies have investigated the effects of dehydration on sleep. Mildly dehydrated women were not as alert, increased feelings of sleepiness, and experienced confusion following fluid intake restriction overnight [13]. Another study induced mild dehydration by fluid restriction found that there were no significant interactions between dehydration and sleep quality and quantity [14]. Aside from mild dehydration through water restriction, during Ramadan, which involves food and water fasting, individuals become dehydrated [15], and experience increased incidence of sleep disturbances and decreased sleep time [15,16,17].

Currently, the effects of dehydration on sleep are inconclusive. The effects of rehydration after dehydration and the dynamic effects of varying hydration states over days on sleep quality is not well understood despite that this may be the real-life daily experience of individuals with effects of a major health and performance factor, sleep. Thus, the purpose of this study was to quantify the effects of varying hydration states over multiple days, including dehydration and rehydration protocols to achieve hypohydrated and euhydrated states on measures of sleep health including sleep duration, sleep quality, perceived ease of falling asleep, sleep calmness, number of dreams, and perceived ease of falling asleep. The current study will aid in understanding how sleep measurements are impacted by different hydration states.

Participants

Eighteen participants were recruited (mean ± standard deviation [SD]; age, 23 ± 4 years; height, 175.8 ± 5.7 cm; weight, 80.1 ± 9.7 kg; 26.0 ± 3.5 BMI) and briefed on the study and the procedures of the experiment, also providing written and informed consent to participate in the study. Each participant was given a medical history questionnaire that excluded those with kidney disease or chronic illnesses that altered fluid-electrolyte balance, the use of medications affecting kidney function (e.g., diuretics or antihypertensives), tobacco use, or conditions such as diabetes. This questionnaire and exclusion criteria were verified by a physician. This study conducted convenience sampling recruited from the university campus in males between 21 and 35 years. Previous research in this community and population has demonstrated that the selected age ranges were compliant with experimental instruction. Sample size calculation was determined by body mass loss, an indicator of each participants’ dehydration level. Previous research conducted in which the standard deviation of body mass loss was 0.8%, the between subject variability day to day was around 50%. Assuming a minimum detectable body mass loss of 1.5%, an effect size of 0.2, and a type I error probability of 5%, the minimum sample size required to detect a significant difference (p < 0.05) was calculated as n = 12. The final sample size of this study (n = 18).

The research conducted was part of a larger study which investigated other research questions, resulted in multiple manuscripts, and was approved by the university Institutional Review Board [18,19,20]. The following data, tables, and figures are original to this manuscript to avoid duplication of data.

Procedures

At the beginning of the study and each day, participants were told to abstain from caffeine, alcohol, and exercise. Day 1 was a baseline day (baseline) in which participants came to the laboratory and fasted upon awakening between 7:00 am and 8:00 am with their normal hydration status (Fig. 1). Baseline measures aimed to assess habitual fluid intake behaviors and their corresponding hydration states. At the laboratory, sleep was assessed with the Karolinska Sleep Diary (KSD), a self-reported sleep questionnaire, that has a strong correlation with polysomnography [21]. The KSD has 11 questions that are used to measure multiple aspects of perceived sleep quality, sleep quantity, ease of falling asleep, sleep calmness, number of dreams, and ease of falling asleep [21]. Each question was presented on a 5-point Likert scale with a score of “1” associated with inadequate sleep measures and “5” associated with positive sleep measures. Additionally, subjects reported fatigue on the visual analog scales (VAS) during the evening before sleeping. The VAS consisted of a statement and a 100-mm straight line marked at both ends using opposite adjectives. One end of the line represented a low rating (fatigue), and the other end would have a high rating (energized). Fatigue was presented on a page-sized at 5 cm × 20 cm.

Fig. 1

Study timeline. Body mass, urine and blood analyses were measured, and Karolina Sleep Diary was given or returned at; baseline (BASE), euhydration (EUH), dehydration (DEH), and ad libitum (AD)

Then, first-morning spot urine samples were taken to assess urine specific gravity (USG) with a refractometer (model 300CL, Atago Co, Tokyo, Japan), urine color utilizing a validated chart [22], and body mass with minimal clothes (undergarments only) was measured (model DS44L, Ohaus Inc, Florham Park, NJ). Additionally, blood was drawn from an antecubital vein to assess plasma osmolality.

Following baseline measurements, participants left the lab, continued normal activities (excluding physical activity outside the lab, caffeine intake, and alcohol consumption). Participants were instructed to perform 24-h urine collection, fill out and return the KSD, and arrive at the laboratory between 0700 and 0800. Also, participants were instructed to consume 500 ml of water in the evening to arrive at the laboratory on the morning of day 2 in the euhydrated state (EUH). Once participants arrived in the morning, the 24-h urine sample was collected and the KSD, first-morning spot urine sample, body mass, and blood draw were performed. The 24-h urine was used to assess urine osmolality using the freezing-point depression method (model OsmoPRO, Advanced Instruments, Inc., Norwood, MA). Following the morning visit of day 2, participants did not consume fluids and ate dry foods into the morning of day 3 to arrive at the laboratory in a dehydrated state (DEH). On the morning of day 3, participants arrived at the lab and completed the same measurements on day 2 along with the 24-h urine sample and KSD. Following the morning of day 3, subjects were able to drink and eat without restrictions. Participants came on the morning of day 4 and the same assessments were performed to determine their hydration status following ad libitum drinking (AD).

Percent body mass loss was calculated based on body mass at EUH as a baseline value for each visit. The interpretation of hydration state was made at EUH after instructions for hydration were provided given the BASELINE (i.e., first visit to the laboratory) was not reflective of a starting point euhydrated state given many individuals may not maintain hydration state habitually. Hydration measurements, the KSD, and the VAS for fatigue were analyzed with repeated measures ANOVA with Bonferroni post hoc comparison. The assumption of local sphericity was examined with Mauchly’s test of sphericity. Data were reported as mean ± SD. All statistical analyses were performed using SPSS software (v.25. IBM Corporation, Armonk, NY). Significance was set a priori at p ≤ 0.05.

Table 1 shows hydration measurements (percent body mass loss, first-morning USG and first-morning urine color, and 24-h urine osmolality) that the intervention was successful (i.e., euhydrated and dehydrated).

Table 2 represents KSD values and the difference between the days. In DEH (7.5 ± 1.3 h), subjects slept significantly longer than BASE (6.4 ± 1.4 h, p = 0.001), EUH (6.7 ± 1.4 h, p = 0.006), and AD (6.9 ± 1.0 h, p = 0.024) (Fig. 2). Participants subjectively reported more difficulty falling asleep, DEH (3 ± 1) from AD (4 ± 1) (p < 0.05) (Fig. 3). Although nonsignificant, EUH (4 ± 1) was close to significance different from DEH (3 ± 1) (p = 0.056). Other questionnaire items such as “How long did it take you to fall asleep?”, “How did you sleep?”, “How refreshed did you feel after awakening?”, How calmly did you sleep?”, “Did you sleep through the allotted time?”, “Number of awakenings?”, “How easy was waking up?”, and “How much did you dream?” were not significant between different hydration states (p > 0.05).

Fig. 2

Hours of sleep were described using mean ± standard error (SE). Hours of sleep were documented on day 1 (baseline), day 2 morning (EUH), day 3 morning (DEH), and day 4 morning (AD). All data was collected on weekdays. * Significance comparing DEH versus BASE (p = 0.001), DEH versus EUH (p = 0.006), and DEH versus AD (p = 0.024)

Fig. 3

The rating of ease of falling asleep was described using mean ± SD. Subjects rated on a scale of 1 to 5 how easy falling asleep was on day 1 (baseline), day 2 morning (EUH), day 3 morning (DEH), and day 4 morning (AD). All data was collected on weekdays. *Significance comparing DEH versus BASE, DEH versus EUH, and DEH versus AD (p < .05)

Figure 4 represents VAS evening fatigue values between each day. In DEH (21.93 ± 13.07), subjects were significantly more fatigued than in EUH (30.23 ± 14.03, p = 0.049). There were no significant differences between AD (32.82 ± 15.42) and DEH (p = 0.64) or AD and EUH (p = 0.61). The BASE evening was not measured.

Fig. 4

The rating of fatigue was described using mean ± SE. Subjects rated fatigue on a 100 mm straight line where 0 was associated with feelings of fatigue while 100 was associated with feeling energized. This was measured on day 2 evening was (EUH), day 3 evening (DEH), and day 4 evening (AD). Day 1 (BASE) was not measured, and all data was collected on weekdays. * Difference between DEH versus EUH (p < .05)

The aim of this study was to examine varying hydration states over multiple days and effects on sleep measures. According to the KSD, during each hydration state, the subjects did not experience any statistically significant differences in sleep quality, number of dreams, or number of awakenings. Therefore, the main findings of this study are that sleep time significantly increased during mild dehydration (DEH) than on all other days (baseline, EUH, AD), subjects reported falling asleep easier on AD compared to DEH, and evening fatigue was greater in DEH than in EUH.

Inducing mild dehydration through fluid restriction increased fatigue perception in men [23] and women [13, 24]. This can be explained by the dopaminergic and noradrenergic systems located in the brainstem responsible for attention, motivation, and fatigue, which is influenced by dehydration [25]. The findings of this study suggest that with increased fatigue experienced, sleep time increased in this study. Unlike the relationship between fatigue and hydration, scientific literature still lacks data that examines the relationship between the perceived difficulty of falling asleep—to our knowledge, this is the first study that found significance in this item of the KSD regarding dehydration.

A previous study used polysomnography, the gold standard of sleep measurement, and found no significance in mild dehydration on sleep latency and sleep duration [14]. Similar to our findings, there was no significance in sleep latency; however, we found that sleep duration increased while mildly dehydrated. Although the findings were done through the KSD, the questionnaire has been validated in previous studies with polysomnography [26, 27]. The conflicting data on sleep duration could be attributed to not assessing the percent body mass loss across baseline and dehydration conditions. Analyzing percent body mass loss helps demonstrate the degree to which individuals are dehydrated. As a result, the absence of the data demonstrates an unknown dehydration status. The level of dehydration may impact sleep measurements such as sleep duration and quality; however, conclusions cannot be definitely drawn and could possibly be the difference in the results.

In contrast, fasting for Ramadan has been found to dehydrate individuals [15], increase sleep latency [16, 17], and decrease sleep duration by \(\sim\) 1 h [17]. As opposed to dehydration through fluid restriction, Ramadan requires individuals to refrain from consuming both food and water. Although the link between dehydration and sleep is significant [28], there is also a relationship between food and sleep [29]. Food restriction during Ramadan results in hunger [30]. Hunger is primarily influenced by ghrelin, an endocrine hormone, and increases in ghrelin have also been associated with sleep loss [29]. Therefore, the relationship between sleep loss and dehydration during Ramadan is unclear.

A limitation of this study was that the participants recruited were male college-aged individuals. The protocol required participants to report to the laboratory between 7:00 am and 8:00 am, potentially influencing habitual living patterns, therefore, increasing sleep time during DEH. Despite increased sleep time, fatigue increases with DEH were consistent with previous literature. Furthermore, female subjects were not recruited and need to be tested.

A major limitation in sleep studies is the preliminary pursuit of testing effects of states such as hypohydrated or euhydrated state on measures of sleep that are self-report and include subject measures of sleep quality that cannot be collected blinded to an obvious physical state induced by behavioral interventions (e.g., not drinking fluids over the course of multiple hours to induce dehydration). This study used self-report and subjective measures of sleep gathered through a validated, but still limited in interpretation, KSD tool. Our data provide preliminary evidence to pursue more invasive and objective measures of sleep.

Authors

1. Alan T. Ky
2. Gabrielle E. W. Giersch
3. Yasuki Sekiguchi
4. Leslie Dunn
5. Douglas J. Casa
6. Lawrence E. Armstrong
7. Elaine C. Lee