A Microstructural Analysis of Sodium and Water Intake in Male and Female Rats

Abstract

Analysis of fluid intake microstructure is a powerful tool to provide insight into the mechanisms underlying ingestion. Foundational work demonstrated that changes in the number of licking bursts reflect changes in negative feedback related to satiety, whereas the average size of those bursts reflects changes in the hedonic value of the substance being consumed. These conclusions, however, were drawn primarily from studies of sucrose intake, and it is unclear if the interpretation can be extended to intake of water or sodium. To address this open question, we first tested and found support for the hypothesis that dehydrated rats would show a concentration-dependent decrease in sodium intake that is the result of reduced burst size, without differences in burst number. Next, we tested if a preference for cold water over tepid water was a function of burst size. We found that rats preferred 10°C water over 40°C water, and that the difference in intake was a function of both burst size and burst number. To test this in a condition with a reduced influence of satiation, we repeated this experiment using polydipsic Brattleboro rats. Like the rats in the second experiment, we found that Brattleboro rats preferred 10°C water over 40°C water, but the difference was mediated only by burst size and not burst number. Together, these findings support the conclusion that burst size can be used to measure changes in the hedonic value of sodium and water.

INTRODUCTION

Analysis of drinking (lick) microstructure is a tool which has been used for more than 30 years to aid in the understanding of liquid food and fluid consumption. Lick microstructure analysis is powerful because it allows for consumption to be analyzed beyond overall volume and provides insight into the hedonic and satiating properties of the consumed foods/fluids [1-4]. Fluid intake patterns across a test session occur in bursts, often defined as at least two licks with an interlick interval of less than 500 - 1000 ms [5-9]. The number of bursts and the average size of the licking bursts can be quantified within and across a test session. Foundational research by Davis and Smith demonstrated that the burst number associated with increasing concentrations of nutrient solutions follows an inverted U-shaped curve and that sham feeding paradigms (where a gastrointestinal fistula prevents food from leaving the stomach) result in increased burst number [3, 5, 6]. This led to the conclusion that changes in burst number represent changes in post-ingestive/satiety feedback signals. On the other hand, burst size follows the pattern of intake when the concentration of sucrose or quinine is manipulated, suggesting that burst size reflects the hedonic value of the consumed fluid [5, 6, 10, 11]. In support of this, burst size is unchanged in sham feeding [5], when there is presumably no change in the hedonic value of the consumed fluid. Together, these findings led to the conclusion that burst size represents changes in the hedonic value of the liquid, whereas burst number reflects the satiating potency of the stimulus. These original studies, however, focused mostly on ingestion of fluids with flavor profiles that signal caloric value (e.g., sucrose or other sugars), with less work done to confirm if the interpretation of these behavioral changes applied to the intake of stimuli that alter body fluid homeostasis (e.g. water or saline).

Although the foundational work was focused on intake related to energy balance, a smaller number of studies evaluated intake that helps maintain body fluid homeostasis. For example, sham-drinking by water-deprived rats demonstrates that greater water and isotonic saline consumption is driven primarily by changes in burst number, but accumulation of water in the GI tract can begin to affect burst size later in the drinking period [12, 13]. This extends the interpretation that changes in burst number reflect changes in post-ingestive/satiety feedback signals when water and sodium are ingested. As for the hedonic contributions of water and sodium intake on licking microstructure, intake of aversive (hypertonic) sodium concentrations is associated with decreased rates of licking [13], another proxy for the hedonic value of the solution [14], however, burst size was not quantified in this study. Therefore, it is not clear if changes in burst size during water and sodium consumption reflect the hedonic value of the consumed fluid.

To test if changes in burst size are a proxy for the hedonic value of water and sodium, we intentionally manipulated aspects of the consumption that most likely affect the hedonic value of the fluid and tested for changes in burst size. We did this by 1) altering the concentration of sodium provided to water-deprived rats, 2) altering the temperature of water provided to rats, and 3) altering the temperature of water provided to Brattleboro rats. For the first experiment, in which we manipulated saltiness of the solution provided, we relied on the established aversion to high concentrations of sodium in the absence of sodium-deprivation and gave water-deprived rats 0-3% sodium solutions to drink [15, 16]. Lick patterns were analyzed to test the hypothesis that concentration-related differences in sodium intake are driven by changes in burst size but not burst number. The next experiments were based on a previously identified preference for cold vs tepid water [17, 18]. Water-deprived rats show a robust preference for cold water even over calorically dense warm sucrose solutions [18] and will lever press more for access to cold water than tepid water [19]. Together, these data suggest that cold water is a hedonically positive stimulus. In our study, water-deprived rats were given a two-bottle choice test with 10°C and 40°C water to test if a preference for cold water was a function of burst size. To further probe the effect of temperature, we repeated this experiment using Brattleboro rats to provide an evaluation of the licking in a rat model that has reduced satiation during water consumption [20, 21]. Collectively, these experiments provide key tests of the applicability of previous work using sucrose solutions to experiments that measure water and sodium intake.

METHODS

Animals:

For Experiment 1, male and female Sprague Dawley rats (Envigo/Inotiv Laboratories, Indianapolis, IN) were singly housed in modified shoebox cages with ad libitum access to food (Teklad 2018; Harlan Laboratories) and tap water unless otherwise noted. Testing occurred in the rats’ home cages during the early part of the light phase. For Experiment 2, rats were obtained from a breeding colony at the University at Buffalo that was derived from rats from the Rat Resource and Research Center (University of Missouri, Columbia, MO). Rats were produced from breeding pairs that were heterozygous for the Brattleboro mutation. This resulted in male and female wild-type Long Evans rats (Expt. 2a), and male and female rats with homozygous Brattleboro mutations which lack functional vasopressin (Expt. 2b). Rats were genotyped between postnatal day (PD) 13–15 with day of birth considered PD 0 (as described in [20]). On PD 21, rats were weaned into same-sex-genotype groups of 2–3 rats per shoebox cage with ad libitum access to food (Teklad 2018; Harlan Laboratories) and tap water unless otherwise noted. Prior to testing, rats were moved to single housing conditions in standard shoebox cages. Testing occurred during the early part of the light phase. All temperature- and humidity-controlled colony rooms were maintained on a 12:12 h light-dark cycle (lights on at 0700 h). All experimental protocols were approved by the Animal Care and Use Committee at the University of Kentucky or the University at Buffalo, and the handling and care of the animals was in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals.

Intake Measures and Microstructure:

Total intake was calculated by weighing the water bottles before and after each test period. For Experiment 1, licking behavior was measured using a contact lickometer which connected to the customized shoebox cages (designed and built by the University of Kentucky Electronics Shop, Lexington, KY). The lickometer recorded individual licks to allow for analysis of licking microstructure as described in [22, 23]. For Experiment 2, the testing chambers were equipped with an infrared light emitting diode and a photo detector (designed and built by the Florida State University Electronics and Machine Shop, Tallahassee, FL). Each time a rat licked the sipper tube its tongue protrusion broke the beam between the diode and the detector allowing for lick counts as described in [17, 18]. For all experiments, a licking burst was defined as at least 2 licks with an inter-lick-interval (ILI) of no more than 1 s and burst size was defined as the number of licks within a burst, as previously described [4, 9].

Experiment 1: Microstructure changes associated with NaCl consumption after 24 h of fluid restriction.

Using a counterbalanced design, six female and six male rats received access to a water and NaCl (0% (0 M), 1% (0.171 M), 2% (0.342 M), or 3% (0.513 M)) bottle for 72 hours (pre-exposure). Bottles were weighed daily and then the position of the bottles was switched to prevent a side-preference. On day 4, all fluids were removed. Twenty-four hours later, food was removed from the cages, and rats were given access to a single pre-weighted bottle of NaCl (pre-exposure concentration) for 1 h. Pre-weighed water bottles were then returned to the cages for the second h of the test period. Intake and licks across this 2-h period were measured. Food was returned at the end of the testing period. This was repeated once each week for four weeks until all rats were tested with all NaCl concentrations. All testing occurred in the rat’s home cage and licks were recorded for the 2-h test period (both NaCl and water bottles) but not during the pre-exposure period. The NaCl concentrations were chosen based on previous research demonstrating aversion to high (2-3%) NaCl concentrations [15, 16, 24].

Experiment 2: Microstructure changes associated with the consumption of cold and tepid water after 24 h of fluid restriction.

In Experiment 2a, seven female and five male wild-type rats were water deprived for 23 h. At the end of the 24-h period, rats were moved to a testing chamber equipped with two sipper tubes each mounted in an aluminum block. A Peltier device was attached to each block which could be heated or cooled to various temperatures between 10°C and 40°C, as described in [18]. Once in the testing chamber, rats were given access to water maintained at 10 and 40°C in a two-bottle choice test and intake was recorded for 30 min. Testing was repeated 48 h later with the position of the cold and tepid bottles switched to ensure no side-preferences. In Experiment 2b, the above protocol was repeated with seven female and four male Brattleboro rats with a 6 h water deprivation period, based on pilot studies to determine appropriate deprivation periods for this strain of rat [20].

Data Analysis:

All data are presented as means ± SEM throughout. In Experiment 1, the pre-testing exposure volumes were averaged across the three-day period and analyzed by a three-factor ANOVA with sex as the between-subjects variable and NaCl concentration (0-3%) and fluid bottle (water vs NaCl) as the within-subjects variable. Intake and microstructure measures during rehydration tests were analyzed with a two-factor ANOVA with sex as the between-subjects variable and NaCl concentration (0-3%) as the within-subjects variable. During hour 2 of the test, burst size for water contained missing values, therefore a mixed-effects model was fit using Restricted Maximum Likelihood (REML). For Experiment 2, intake and microstructure measures from the two test periods were averaged and analyzed with a two-factor ANOVA with sex as the between-subjects variable and temperature as the within subjects-variable. Newman-Keul’s post hoc tests were used to follow up significant main and interactive effects.

RESULTS

Experiment 1: Microstructure changes associated with NaCl consumption after 24 h of fluid restriction.

Analysis of the average daily intake across the 3 pre-exposure days revealed main effects of bottle [F(1,10) = 22.92, p < 0.001] and NaCl concentration [F(3,30) = 11.40, p < 0.001] and an interaction between bottle and concentration [F(3,30) = 23.48, p < 0.001] but no effect of sex [F(1,10) = 0.54, p = n.s] nor any additional interactions (Figure 1). Although there was no difference in intake between the water and NaCl bottles for the 0% and 1% concentrations, rats consumed significantly more water than 2% or 3% NaCl [p < 0.01].

Figure 1.

Average daily water and NaCl intake during the pre-exposure period. There was no difference in average daily intake between water and 0-1% NaCl. During 2% and 3% NaCl access, rats drank significantly less NaCl, compared to water. *Less than water, p < 0.01.

As illustrated in Figure 2A, licks for NaCl were primarily made in the first 10 min of the test session regardless of the NaCl concentration. Analysis of NaCl intake during the first hour of rehydration revealed a main effect of sex [F(1,10) = 11.70, p < 0.01], and NaCl concentration [F(3,30) = 37.65, p < 0.001], but no interaction between sex and NaCl concentration [F(3,30) = 0.81, p = n.s.] (Figure 2B). Males drank more NaCl than females [p < 0.01]. Intake for all NaCl concentrations were significantly different from each other with consumption greatest for 1% NaCl, then 0% NaCl, followed by 2% NaCl, with 3% NaCl having the lowest consumption [p < 0.05]. The number of licks showed similar patterns with a main effect of NaCl concentration [F(3,30) = 40.75, p < 0.001] but no interaction of NaCl concentration and sex [F(3,30) = 1.00, p = n.s.]. The main effect of sex failed to reach statistical significance despite a detected trend [F(1,10) = 4.30, p = 0.06] (Figure 2C). The licking microstructure analysis revealed that the decrease in intake was driven by reduced burst size. Burst size was influenced by a main effect of sex [F(1,10) = 18.83, p < 0.005], NaCl concentration [F(3,30) = 30.64, p < 0.05], and an interaction between sex and NaCl concentration [F(3,30) = 6.02, p < 0.05] (Figure 2D illustrates the main effects and Figure 2F highlights the interaction). Burst size was greater in males than females [p < 0.05] and while there was no burst size difference between water and 1% NaCl, the burst size for 2% NaCl was less than 0 and 1%, while the burst size for 3% was less than all other concentrations [p < 0.05]. Post hoc tests for the interaction revealed that in males, burst size was greatest for 1% NaCl, followed by 0% NaCl and while burst size for 2% and 3% NaCl were smaller than 0 and 1%, they did not differ from each other [p < 0.005]. In females, however, the only burst size difference was 3% NaCl was smaller than 0% and 1% NaCl [p < 0.005]. Burst number was influenced by a main effect of NaCl concentration [F(3,30) = 4.34, p < 0.05], but not by sex [F(1,10) = 0.55, p = n.s.], or by an interaction between sex and NaCl concentration [F(3,30) = 1.28, p = n.s.] (Figure 2E). Burst number for 3% NaCl was greater than 0 and 1% NaCl [p < 0.05].

Figure 2.

NaCl intake during the first hour of fluid access. (A) Cumulative lick curves demonstrate that the majority of licks for NaCl occurred within the first 10 min of the test period. (B/C) Intake and licks for 1% NaCl were greater than for 0% NaCl. Access to 2% and 3% NaCl resulted in concentration-related decreases compared to 0% NaCl. In addition, males had greater intake than females. (D) Average burst size for 0% and 1% NaCl were equivalent. Access to 2% and 3% NaCl resulted in concentration related decreases in burst size. In addition, males had greater burst sizes than females. (E) Burst number for 0% and 1% NaCl were equivalent. Access to 2% and 3% NaCl resulted in a concentration-related increase in burst number. (F) Burst size as a function of sex. In males, average burst size for 1% NaCl was greater than all other concentrations. Average burst size for 2% and 3% NaCl was less than average burst size for 0% and 1% NaCl. In females, average burst size for 3% NaCl was less than average burst size for 0% NaCl. *Greater than females, p < 0.05. ⁺0% and 1% NaCl greater than 3% NaCl. Different letters denote significant differences from other conditions.

During the second hour of testing, NaCl intake was, on average, less than 0.25 ml with most rats not consuming any fluid; therefore, no analysis was performed. Water intake was influenced by sex [F(1,10) = 7.11, p < 0.05], NaCl concentration [F(3,30) = 23.34, p < 0.001], and an interaction between sex and NaCl concentration [F(3,30) = 2.95, p < 0.05] (Table 1). Males consumed more than females [p < 0.05]. Water intake in male rats was larger after 1% NaCl access than 0%, with water intake after 2% and 3% NaCl being larger than after both 0% and 1% [p < 0.05]. In females, water intake after access to 2% and 3% NaCl was greater than after access to 0% and 1% NaCl [p < 0.05]. Similar findings were observed with the licks for water, with a main effect of sex [F(1,10) = 6.54, p < 0.05] and NaCl concentration [F(3,30) = 27.34, p < 0.001], but the interaction between sex and NaCl concentration failed to reach significance, despite a detected trend [F(3,30) = 2.85, p = 0.054]. Again, males had more licks than females [p < 0.05] and licks for water were greater after access to 1% NaCl compared to 0% NaCl. Licks for water after 2% and 3% NaCl were greater than licks after 0% and 1% [p < 0.05]. Burst size for water was influenced by a main effect of NaCl concentration [F(3,27) = 16.00, p < 0.001] and an interaction between sex and NaCl concentration [F(3,27) = 4.23, p < 0.05], but the main effect of sex was not significant [F(1,10) = 3.55, p = ns]. Burst size for the water bottle associated with 0% NaCl intake was smaller than burst size for the water bottles associated with 1–3% NaCl [p < 0.001]. Burst size was greater in males, compared to females, for water intake associated with the 1% and 3% NaCl [p < 0.05]. Burst number for water was influenced by a main effect of NaCl concentration [F(3,30) = 8.16, p < 0.001] but not by sex [F(1,10) = 1.27, p = n.s.] or by an interaction between sex and NaCl concentration [F(3,30) = 1.31, p = n.s.]. Burst number for water after 0% NaCl was less than after all other concentrations of NaCl [p < 0.005].

Table 1.

Water Intake during hour 2 of the test session.

	Water
Test Solution	ml	licks	BN	BS
0%	1.17 ± 0.29	145.83 ± 58.41	4.08 ± 1.28	31.26 ± 8.91
1%	6.00 ± 0.64*	1231.67 ± 170.2*	14.50 ± 2.38*	118.47 ± 28.05*
2%	10.49 ± 1.25#	2159.75 ± 230.1#	18.25 ± 3.11*	145.11 ± 21.76*
3%	9.25 ± 1.60#	1796.33 ± 299.1#	14.58 ± 2.05*	144.21 ± 32.61*

^*Greater than 0%

^#Greater than 0 and 1%

Finally, total intake and licking microstructure was analyzed across the entire 2-h testing period. As illustrated by Figure 3A, licks for both NaCl and water were primarily within the first 10 minutes of bottle access (NaCl = 0–10 min, water = 60–70 min). The only exception was when rats received water after having access to water (0% NaCl) there was minimal intake in the 2^nd h of testing. Analysis of total intake revealed a main effect of sex [F(1,10) = 10.32, p < 0.01], and NaCl concentration [F(3,30) = 8.70, p < 0.001], but not an interaction between sex and NaCl concentration [F(3,30) = 0.58, p = n.s.] (Figure 3B). Males drank more than females [p < 0.01]. Total intake when the NaCl solution was 1% or 2% was greater than total intake when the NaCl solution was 0% or 3% [p < 0.005]. Licks showed similar patterns with a main effect of sex [F(1,10) = 5.82, p < 0.05] and NaCl concentration [F(3,30) = 9.31, p < 0.001] but without an interaction of NaCl concentration and sex [F(3,30) = 0.59, p = n.s.] (Figure 3C). The licking microstructure analysis revealed that the increase in total intake for the test period when rats had water + 1% NaCl was driven by a larger burst size. Burst size was influenced by a main effect of sex [F(1,10) = 23.28, p < 0.001], NaCl concentration [F(3,30) =12.84, p < 0.001], and an interaction between sex and NaCl concentration [F(3,30) = 5.06, p < 0.01] (Figure 3D illustrates the main effects and Figure 3F highlights the interaction). Burst size was greater in males than females [p < 0.05]. Burst size for total intake for the test period when rats had water + 1% NaCl was greater than in all other test conditions and when rats had water + 3% NaCl burst size was less than in all other test conditions [p < 0.05]. The interaction revealed that these difference in burst size were exclusively driven by the males [p < 0.01]. Burst number was influenced by a main effect of NaCl concentration [F(3,30) = 5.10, p < 0.01], but not by sex [F(1,10) = 1.33, p = n.s.], or by an interaction between sex and NaCl concentration [F(3,30) = 1.87, p = n.s.] (Figure 2E). Burst number for total intake for the test period when rats had water + 2% NaCl or water + 3% NaCl was greater than when the NaCl solution was 0% [p < 0.01].

Figure 3.

Total intake and lick microstructure analysis. (A) Cumulative lick curves demonstrate that licks for both NaCl and water were primarily within the first 10 min of bottle access. (B/C) Intake and licks for total intake when the NaCl solution was 1% and 2% were greater than total intake when the NaCl solution was 0% or 3%. In addition, males had greater total intake and total licks than females. (D) Total average burst size when the NaCl solution was 1% was greater than all other test conditions. Total average burst size when the NaCl solution was 3% was less than all other test conditions. In addition, males had greater burst sizes than females. (E) Total burst number when the NaCl solution was 0% and 1% were equivalent. Total burst number for when the NaCl solution was 2% or 3% was greater than when the NaCl solution was 0%. (F) Total burst size as a function of sex. In males, total average burst size when the NaCl solution was 1% was greater than all other conditions and when the NaCl solution was 3% total average burst size was less than all other conditions. There was no difference in total average burst size for females. *Greater than females, p < 0.05. Different letters denote significant differences from other conditions.

Experiment 2: Microstructure changes associated with the consumption of cold and tepid water after 24 h of fluid restriction.

As illustrated in Figure 4A, in wild-type rats (Expt. 2a) licks for tepid water were concentrated in the first 10 min of the test session, while animals licked for cold water across the first 20 min of the test session. Water intake was influenced by a main effect of temperature [F(1,10) = 11.45, p < 0.01] but not by sex [F(1,10) = 0.50, p = n.s.] or by an interaction between temperature and sex [F(1,10) = 4.80, p = n.s.] (Figure 4A). In the two-bottle choice test, rats consumed more 10°C water, compared to 40°C water [p < 0.01]. A similar effect was observed with there being more licks for 10°C water, compared to 40°C water [F(1,10) = 21.32, p < 0.001] but there was no effect of sex [F(1,10) = 1.66, p = n.s.] or an interaction between temperature and sex [F(1,10) = 2.53, p = n.s.] (Figure 4B). Burst size was also influenced by a main effect of temperature [F(1,10) = 36.27, p < 0.01] but not by sex [F(1,10) = 0.92, p = n.s.] or by an interaction between temperature and sex [F(1,10) = 3.51, p = n.s.] (Figure 4C). Average burst size for 10°C water was greater than the average burst size for 40°C water [p < 0.01]. Finally, burst number was influenced by a main effect of temperature [F(1,10) = 15.49, p < 0.01] but not by sex [F(1,10) = 1.77, p = n.s.], or by an interaction between temperature and sex [F(1,10) = 2.90, p = n.s.] (Figure 4D). Burst number for 10°C water was greater than burst number for 40°C water [p < 0.01].

Figure 4.

Wild-type rats prefer cold water over tepid water. (A) Licks for tepid water were concentrated in the first 10 min of the test session, while licks for cold water were concentrated in the first 20 min of the test (B/C) Intake and licks for 10°C water was greater than 40°C water. (D) Average burst size for 10°C water was greater than 40°C water. (E) Burst number for 10°C water was greater than 40°C water. *Less than 10°C water, p < 0.05.

In Brattleboro rats (Expt. 2b), licks for tepid water were concentrated in the first 10 min of the test session, while licks for cold water were sustained throughout the test session (Figure 5A). Water intake was influenced by a main effect of temperature [F(1,9) = 16.22, p < 0.01] and sex [F(1,9) = 5.83, p < 0.05] but not by an interaction between temperature and sex [F(1,9) = 1.29, p = n.s.] (Figure 5A). In the two-bottle choice test, Brattleboro rats consumed more 10°C water, compared to 40°C water [p < 0.01] and males consumed more fluids than females [p < 0.05]. Licks for water revealed a main effect of temperature [F(1,9) = 9.75, p < 0.05] but no effect of sex [F(1,9) = 3.27, p = n.s.] or an interaction between temperature and sex [F(1,9) = 0.43, p = n.s.] (Figure 5B). Again, more licks were made for 10°C water, compared to 40°C water [p < 0.01]. Burst size was also influenced by a main effect of temperature [F(1,9) = 9.63, p < 0.05] but not by sex [F(1,9) = 0.29, p = n.s.] or by an interaction between temperature and sex [F(1,9) = 0.32, p = n.s.] (Figure 5C). Again, average burst size for 10°C water was greater than the average burst size for 40°C water [p < 0.01]. Unlike WT rats, burst number in the Brattleboro rats was not influenced by temperature [F(1,9) = 1.26, p = n.s.], sex [F(1,9) = 0.44, p = n.s.] or an interaction between temperature and sex [F(1,9) = 0.2, p = n.s.] (Figure 5D).

Figure 5.

Brattleboro rats prefer cold water over tepid water. (A) Licks for tepid water were concentrated in the first 10 min of the test session, while licks for cold water were sustained throughout the test session. (B/C) Intake and licks for 10°C water was greater than 40°C water. Males had greater intake than females, regardless of water temperature. (D) Average burst size for 10°C water was greater than 40°C water. (E) There was no difference in burst number between 10°C and 40°C water. *Less than 10°C water, p < 0.05. +Greater than females, p <0.05.

DISCUSSION

The present results provide important support for the use of lick microstructure analysis as a tool in the study of the physiological and neural controls of intake that helps maintain body fluid homeostasis. Foundational studies by Davis and Smith in the 1990s set the groundwork for our understanding of licking microstructure and enabled our ability to study ingestive behavior beyond total intake. These efforts established that licking behavior in the rat occurs at a consistent rate of about 6-7 Hz with interlick intervals less than 250 ms [5], which is dependent on the motor output of a central pattern generator in the hindbrain [25, 26]. Pauses of 500–1000 ms are considered terminations in the licking burst, with more recent studies in both rats and mice concluding that 1000 ms is the most appropriate pause criterion [6, 27]. Burst number reflects the effect of satiating feedback, which is supported by studies demonstrating increases in burst number during sham feeding paradigms (which eliminate post-ingestive feedback). While the original studies establishing the connection between burst number and post-ingestive processes used sucrose [5, 6], later studies reproduced this finding when water and sodium were being consumed [12, 13]. Burst size, on the other hand, reflects hedonic evaluation of the solution. This conclusion is based on studies demonstrating that burst size increases as sucrose or maltose concentration increases while burst size decreases when solutions are adulterated with quinine [5, 6, 10]. In further support of the role of burst size in hedonic evaluation, sham feeding does not influence burst size [5]. No studies, however, addressed whether changes in burst size are a proxy for hedonic evaluation when water and sodium were the fluid of interest despite those conclusions being used [9, 20, 23, 28-39]. These experiments, therefore, have addressed a major gap in the literature, and provide support for the conclusion that when rats are consuming water or sodium, changes in burst size reflect the hedonic value of the consumed liquid.

The focus of early work on intake related to calorie consumption made it questionable when applying this type of analysis to consumption of water or sodium intake. That is not to say that there was no basis for using burst analyses on water or sodium intake. Indeed, sham feeding studies previously established a relationship between changes in burst number and satiety when water and sodium solutions were consumed [12, 13]. Furthermore, the rate of licking, which is affected by changes in burst size, is lower when aversive concentrations of sodium are consumed [13], however, the report describing this did not include a direct measure of burst size. Moreover, to the best of our knowledge, there have been no studies that purposefully manipulated the hedonic value of water to test for differences in burst size. The present experiments addressed these open questions by demonstrating that intake of aversive concentrations of sodium are associated with reduced burst size and a preferred water choice (cold water) was associated with an increase in burst size. Importantly, in an animal model with impaired fluid satiety, the changes in burst size associated with cold water consumption persisted. Together, these results support usage of the long-standing interpretation of lick microstructure that changes in burst size reflect changes in hedonic feedback when the liquid is water or sodium.

We first tested the hypothesis that reduced intake of less preferred concentrations of sodium would be mediated by a reduction in burst size. Rats were first pre-exposed to the testing solution for three days prior to the intake testing, to prevent neophobia during the test session. During this time, rats showed no preference between water and 1% sodium, however, rats showed a strong preference for water compared to 2% and 3% sodium. The avoidance of 2–3% sodium is consistent with other studies [15, 16, 24], and taste reactivity testing has shown aversive responses (gaps, chin rubs, face washing, forelimb flailing, and head shakes) associated with 3% (~0.5M) sodium consumption [24, 40, 41]. On test day, in response to 24 h of water deprivation, rats consumed more 1% (0.171 M) NaCl than water, consistent with previously reported preferences for 0.03–0.15 M NaCl [15]. The increased intake of 1% NaCl was associated with an increase in burst size in male rats, but unexpectedly, no change in burst size in female rats. As expected, rats consumed significantly less 2% and 3% sodium, compared to water or 1% sodium. Importantly, this reduction was mediated by a decrease in burst size in both males and females. This suggests that when the sodium concentration is aversive, the aversive oral feedback manifests as a change in burst size. This further supports the previous research demonstrating that intake of hypertonic (0.22–0.3 M) sodium concentrations is associated with decreased rates of licking, another proxy for hedonic feedback [13]. While this was accompanied by an increase in burst number, it was not sufficient to compensate for the reduced burst size. The critical comparison for testing our hypothesis was the drinking of varied concentrations of NaCl, while holding the need state of the animals constant. It is important, however, to consider that these experiments were conducted while the rats were water deprived, and in physiological need of water. It is, therefore, possible that the comparisons across the salt concentrations made here could look different in non-deprived rats.

In the first experiment, we also evaluated water intake subsequent to the primary measure of sodium intake. This was done by giving rats access to water after having access to the first test solution for an hour. When water was returned to rats that had ample opportunity to drink sodium solutions, we found that rats given 2%-3% NaCl in the first hour drank more water in the second hour than rats that had water or 1% NaCl for the first hour. This is not surprising given that there was likely still motivation to drink because of the water deprivation. What is, perhaps, more interesting is that rats given 1% or 2% NaCl overconsumed fluid during the entire test, whereas rats in the water-only and 3% NaCl conditions drank similar amounts of fluid during the entire test. We observed that licking was invigorated when water was presented after all concentrations of salt, and this was driven by increases in both burst size and burst number. Without additional experiments it is unclear if this is due to a contrast effect between the salty solution and plain water or if the salt consumed in the first hour drove osmotic thirst. Indeed, rats maintained on a high sodium diet were found to increase daily water intake, and, similarly to this study, the increase was a function of increased burst number and burst size [42]. Furthermore, gastric preloads of NaCl result in water intake that varies as a function of the NaCl concentration [43]. Irrespective of the precise mechanism driving the increased intake, it could be similar to what is found in human studies showing differences in intake of various beverages during exercise [44, 45] in that preferred solutions generated more overall intake than did less preferred solutions, despite need being the same across the subjects. The present experiments were not designed to test these possibilities directly, but future experiments applying licking microstructure to experimental questions such as these could be revealing.

In most testing paradigms, females show a greater sodium preference than male rats, which was not replicated in our study [46]. There was no sex effect on sodium preference or intake during the 72-h pre-exposure period, neither in the daily average (Figure 1) nor across the individual days (data not shown). This is surprising given the plethora of reports demonstrating a higher sodium preference in females, compared to age matched males, during two-bottle access for a wide range of sodium concentrations (0.3–1.0 M)[15, 46-52]. During the first hour of rehydration testing, when access was limited to sodium, males consumed more sodium and had larger burst sizes than females. While this again is unexpected based on the well-established finding that females consume more sodium and have a stronger preference than males, to our knowledge, none of those studies were conducted in dehydrated rats. Given that dehydration-induced water intake is greater in males, compared to females [53-56], these data suggest that the sex difference extends to sodium intake during rehydration as well.

Previous work suggests that rats prefer cold water over tepid water in brief-access tests, which minimize post-ingestive feedback, and short term (30 min) preference tests [18]. We, therefore, leveraged this preference to test the hypothesis that a preference for cold water would be mediated by changes in burst size. As expected, wild-type rats consumed more cold water, compared to the tepid water. In support of our hypothesis, this increase was associated with a greater burst size for cold, compared to tepid water. Burst number, however, showed the same pattern. This suggests that both hedonic and satiating signals may contribute to the increased intake of cold water. To test this hypothesis under conditions with minimized satiating feedback, we repeated the experiment using Brattleboro rats. This rodent model was selected because of previous work suggesting that Brattleboro rats, which lack a functional vasopressin gene and are a model of hereditary hypothalamic diabetes insipidus [57], have altered systems that are important for fluid intake satiation. Specifically, GLP-1 has been demonstrated to play an important role in fluid intake satiation [28, 58, 59] and Brattleboro rats have altered responses to GLP-1 receptor agonist treatment [20, 21]. Moreover, the differences in ad libitum water intake between Brattleboro and wild-type rats is a function of burst number, not burst size [21]. In the present experiments, we found that Brattleboro rats were like wild-type rats with respect to the preference for cold versus tepid water. The difference was mediated by a larger mean burst size for cold, compared to tepid water, with no differences in burst number. Together, this provides strong evidence to support the conclusion that burst size is a proxy for hedonic control when the fluid of interest is water. Future studies using taste reactivity testing during access to cold water in wild-type and Brattleboro rats could be useful in further exploring this preference for cold water. One limitation is our small sample of male rats in the Brattleboro experiment which may have prevented the analysis from revealing any temperature by sex interactions. While we believe this is unlikely given the intake and microstructure patterns were of similar magnitude and direction in the males and females, future studies should further address the effect of sex on the preference for cold water.

Despite previous reports concluding that changes in burst size for sodium and water reflect changes in the hedonic value of the fluid [9, 20, 23, 28-39], the conclusion was based on an application of previous studies without any confirmation that the application was warranted. The present series of studies now provides important evidence in support of that conclusion. These studies complement earlier work extending the conclusion that changes in burst number during water and sodium intake reflect changes in post-oral satiety feedback [12, 13]. Furthermore, these studies are the first investigations into the underlying mechanisms of drinking microstructure to include female subjects. Given the well reported sex differences in ingestive behaviors [46, 60-65], inclusion of females in these fundamental studies is extremely important. Of note, the burst sizes in Experiment 2 are considerably smaller than those in Experiment 1. We believe this to be driven by the altered temperatures of the water in Experiment 2. While this difference could be attributed to the different types of lickometers (contact vs photobeam) used in the two studies, previous work using the photobeam system used here with room temperature solutions found larger burst sizes, more similar to those in Experiment 1 [17, 18]. Future studies could directly compare the effect of room temperature vs cold and warm water on burst size dynamics to get a better understanding of how temperature influences the hedonic evaluation of the water. The findings presented here fill an important gap in the understanding of lick microstructure and support the conclusion that burst size is a proxy to measure changes in the hedonic value of water and sodium.

Highlights.

• NaCl intake corresponded with changes in burst size but not burst number.

• Wild-type and Brattleboro rats preferred cold water to tepid water.

• The preference for cold water was mediated by an increased burst size.

• The data indicate that burst size reflects the hedonic value of water and NaCl.