VOLUNTARY ELEVATED ETHANOL CONSUMPTION IN ADOLESCENT SPRAGUE-DAWLEY RATS: PROCEDURAL CONTRIBUTORS AND AGE-SPECIFICITY

Abstract

Alcohol consumption is typically initiated during adolescence, with the incidence of binge drinking (production of blood ethanol concentrations [BECs] > 80mg/dl) peaking during this stage of development. Studies in outbred rats investigating the consequences of adolescent ethanol exposure have typically employed intragastric, vapor, or intraperitoneal administration to attain BECs in this range. While these procedures have yielded valuable data regarding the consequences of adolescent exposure, they are varyingly stressful, administer the full dose at once, and/or bypass digestion. Consequently, we have worked to develop a model of voluntary elevated ethanol consumption in outbred adolescent Sprague-Dawley males and females, building on our previous work (see Hosová & Spear, 2017). This model utilizes daily 30-minute access to 10% ethanol (v/v) in chocolate Boost® from postnatal day (P)28-41. Experiment 1 compared intake levels between (1a) animals given either ball-bearing or open-ended sipper tube tips for solution access, (1b) animals separated from their cage mate by wire mesh or isolated to a separate cage during solution access, (1c) animals given solution access with or without simultaneous access to banana-flavored sugar pellets, and (1d) animals that were either moderately food-restricted or fed ad libitum. Experiment 2 compared intake levels between animals given daily solution access and animals given access only on a “Monday-Wednesday-Friday” intermittent schedule. Experiment 3 compared adolescent and adult (P70-83) consumption using the finalized procedure as based on the results of Experiments 1 and 2. As in our previous work, consumptions well within the binge range were produced on some days, with high consumption days typically followed by several days of lower consumption before increasing again. Sipper tube type (1a) and simultaneous pellet access (1c) did not affect consumption, while intake was significantly higher in non-isolated (1b), food-restricted (1d), daily-access (2), and adolescent (3) animals. However, although ethanol intake was higher in food-restricted animals, the resulting BECs were equivalent or higher in non-restricted animals, likely due to a hepatoprotective effect of moderate food restriction. Post-consumption intoxication ratings correlated with BECs and were notably higher in adults than adolescents, despite the lower voluntary consumption levels of adults, confirming prior reports of the attenuated sensitivity of adolescents to ethanol intoxication relative to adults. The final model utilizes ball-bearing sipper tube tips to provide daily access to 10% ethanol in chocolate Boost® to free-feeding adolescent animals separated from their cage mate by wire mesh, with no food provided during solution access. This easy-to-implement model is effective in producing elevated voluntary ethanol consumption in adolescent, but not adult, Sprague-Dawley rats.

Introduction

Adolescence is the transition between childhood and adulthood, a time during which the brain undergoes robust changes as it matures into its adult form (see Juraska & Willing, 2017; Pfefferbaum et al., 1994; Spear, 2000, for review). Structural and functional changes that occur during this developmental transition are thought to potentially render the adolescent brain more vulnerable than the adult brain to the detrimental consequences of alcohol and other drugs of abuse (see Casey & Jones, 2010; Crews, He & Hodge, 2007; Spear, 2018, for review). At the same time, studies in laboratory animals have revealed age-specific sensitivities of alcohol that could enable greater alcohol consumption than during adulthood (Doremus-Fitzwater, Varlinskaya & Spear, 2010; Schramm-Sapyta, et al., 2009; see Spear, 2013, for review). Perhaps at least partly due to these altered sensitivities, adolescent animals given voluntary access to ethanol generally consume greater quantities than their adult counterparts do, under a variety of intake procedures (e.g., Doremus, Brunell, Rajendran & Spear, 2005; Maldonado, Finkbeiner, Alipour & Kirstein, 2008; Vetter, Doremus-Fitzwater & Spear, 2007).

This age difference observed in animals is akin to the self-reported rates of alcohol consumption in humans, with human adolescents drinking more per occasion (though on fewer days) than adults (SAMHSA, 2007). When surveyed in the 2016 National Survey on Drug Use and Health (NSDUH), 19.3% of U.S. individuals between the ages of 12-20 reported that they had consumed alcohol within the last month. Of these underage drinkers, 62.5% reported engaging in at least one episode of binge drinking within the month, which is nearly triple the percentage of drinkers 26 and over (24.2%) who reported drinking at this level (SAMHSA, 2016). Binge drinking is defined by the National Institute on Alcohol Abuse and Alcoholism (NIAAA) as consuming enough alcohol quickly enough to elevate blood ethanol concentrations (BECs) to 0.08 g/dl or greater. These BECs are usually reached by consumption of 4-5 drinks per occasion in women and men, respectively – the definition used to characterize binge drinking in the NSDUH.

Animal models of repeated exposure to ethanol at above-binge levels (i.e., > 80 mg/dl) during adolescence have provided the field with rapidly accumulating information regarding the long-term neural, cognitive, and behavioral consequences of this exposure (e.g., Broadwater & Spear, 2013; Crews et al., 2006; Lerma-Cabrera et al., 2013; O’Tousa et al., 2013; Sherrill et al., 2011; Varlinskaya, Truxell & Spear, 2014; Vetreno et al., 2014). Such adolescent exposure studies, however, have typically not employed binge-level voluntary consumption. Instead, this body of work has been largely conducted in animals exposed to alcohol using intraperitoneal injection, intragastric intubation, or vapor inhalation – routes of administration that themselves may be stressful or that do not involve absorption through the gastrointestinal tract as typical with alcohol use in human adolescents. These routes have often been used given that existing voluntary consumption models do not typically induce BECs that cross the binge threshold, unless the animals have been selectively bred for high ethanol consumption. Instead, the BECs of outbred animals undergoing traditional voluntary consumption procedures tend to, at best, remain in the low to moderate range (~10 to 45 mg/dl) (e.g., Rowland, Nasrallah & Robertson, 2005; Schindler, Tsutsui & Clark, 2014) or require many weeks of intermittent exposure (e.g., Carnicella, Ron & Barak, 2014) – well beyond the limited developmental window of rodent adolescence. While these voluntary access models have provided the field of alcohol research with important information regarding the long-term consequences of moderate ethanol consumption during adolescence, such as increased risk preference in adulthood (Schindler, Tsutsui, & Clark, 2014; Kruse, Schindler, Williams, Weber & Clark, 2017), these findings do not model the consequences of elevated consumption that is sometimes binge-like in nature. Moreover, while selectively-bred strains are critical for identifying traits that genetically correlate with the consumption phenotype under investigation, only some of the many genes that can influence ethanol intake are selected in any given line (see Mulligan et al., 2006), and hence effects observed may not reveal the diversity of contributors evident in an outbred population.

The field of adolescent alcohol research would benefit from an animal model of voluntary consumption that episodically reaches binge levels in outbred animals. Following the strategy set forth by Bell et al. (2006) who proposed criteria for modeling excessive ethanol consumption in rodents (and concluded that Wistar-based, selectively bred ethanol-preferring (P) rats met these benchmarks), we offer criteria for a rodent model of adolescent voluntary binge drinking. These criteria include: 1) The procedure should enable the measurement of near peak blood ethanol concentrations (BECs) through the use of a relatively short access period, given that lengthier access periods reduce the likelihood of collecting blood samples near the time that ethanol concentrations peak; 2) The self-administration of ethanol should promote individual BECs at or above the binge drinking threshold, 80 mg/dl, on at least some drinking episodes; 3) High consumption days must be evident in both sexes, as both human males and females engage in episodes of elevated drinking during adolescence (SAMHSA, 2016); 4) The procedures used should induce high voluntary intake when using outbred animals, given the genetic diversity of humans that may influence consumption levels and consequences of adolescent ethanol exposure. We believe our model of elevated adolescent alcohol intake in rats meets these criteria.

The work using this model described here stems from our prior work (Hosová & Spear, 2017), where we first investigated the utility of schedule-induced polydipsia (SIP) (see Falk, 1966, 1967; Falk, Samson, & Winger, 1972) as a method for inducing high consumption of ethanol in adolescent male and female Sprague-Dawley rats. This was accomplished by providing scheduled delivery of banana-flavored pellets and simultaneous free access to 10% ethanol in chocolate Boost® (a commercially available drink) for 30 minutes/day from postnatal day (P)28 to P41. The procedure successfully resulted in both males and females exhibiting high ethanol consumption in a variable, cyclical fashion over days (with 1-2 days of very high consumption typically followed by a few lower intake days). Average BECs generally reached the binge drinking threshold and consumptions on some days were well into the binge range. Follow-up experiments in that series showed that these high consumptions were not schedule-induced per se and were not dependent on placement in an operant chamber but were also evident in a novel or even the home cage. Consumptions were even higher, though less variable, with Boost® alone than in solutions of ethanol and Boost®. We hence concluded that the high voluntary intakes by the adolescent animals were not displays of SIP but were instead likely due to the high palatability of the chocolate Boost®. The work presented here is an extension of this earlier work. In the first series of experiments (Exps. 1a-d), we sought to answer four additional questions about which components of the daily intake procedure are important for high ethanol consumption.

In our prior work, adolescent animals were given access to the 10% ethanol in chocolate Boost® solution via ball-bearing-containing bottle tips (Hosová & Spear, 2017); these bottle tips were used in the operant chambers where our early work was conducted and were maintained throughout the series for continuity. Ball-bearing tips are less prone to solution leakage and have been reported to perhaps augment ethanol consumption when compared with standard, open-ended tubes (Doremus, Brunell, Rajendran & Spear, 2005). Experiment 1a assessed whether the use of ball-bearing tips was integral to the high consumption levels produced in this model.

During our previous study we observed that, whereas consumption levels of adolescent females were similar between the operant chamber and home or novel cages, adolescent males generally appeared to consume more in the home or novel cages than in the operant chamber. One possible contributor to this apparent sex difference could be the difference in social conditions at testing: animals were socially isolated when tested in an operant chamber, whereas in home and novel cage tests animals were only separated from their cage mate by a wire mesh divider through which the animals could still see, hear, smell, and likely touch their home cage partner. Prior work by Varlinskaya et al. (2015) has demonstrated that adolescent males appear to engage in high ethanol consumption for its socially facilitating effects, which is not the case in adolescent females. Hence, in Experiment 1b the effect of social condition during the ethanol access period was assessed in both males and females by assigning animals to one of two access conditions: separated from the cage mate within a single cage by a wire mesh divider (“Separated”), or removed from the cage mate and placed in individual cages (“Isolated”).

We next assessed the impact of providing access to banana-flavored sugar pellets during the intake sessions in Experiment 1c. In prior experiments, animals had received these pellets either via scheduled or massed delivery during ethanol-Boost® solution access sessions (Hosová & Spear, 2017 (Exps. 1a,b)). Food availability, even during short intake sessions, could potentially contribute to enhanced drinking given that rats typically consume fluids while eating to facilitate swallowing of the food (Kissileff, 1969). Because the animals in our study were moderately food restricted, it is possible that their pellet consumption may have been sufficient to promote high solution intake during the one bottle access period. Hence, this study examined whether simultaneous access to banana-flavored pellets during the drinking period was necessary for observing high voluntary ethanol consumption. Due to unexpectedly high intakes during an unforeseen power outage, the effect of lighting condition was also examined, comparing consumption between animals tested under dim (white) or red lighting conditions.

In both our previous published work as well as the earlier experiments in this series, all animals were moderately food restricted. Given that our earliest studies used SIP, animals required food deprivation, a restriction that was maintained in subsequent experiments for consistency. Food restriction has been used previously to diminish neophobia (e.g., Meisch & Thompson, 1971; Roehrs & Samson, 1981), but whether it is necessary to produce high levels of consumption using this model of adolescent drinking had not been previously explored until Experiment 1d.

The model of adolescent drinking that we have developed provides brief daily sessions of free access to ethanol to each animal, with individual consumptions typically varying across days within animals and hence providing evidence of some degree of voluntary “intermittency” in the amount of ethanol consumption (Hosová & Spear, 2017). Intermittent exposure has been reported to be important in producing long-term effects of ethanol exposure (Diaz-Granados & Graham, 2007; Sanchez, Castro, Torres & Ortega, 2014) and intermittent access schedules have often been used in voluntary ethanol consumption studies (e.g., Garcia et al., 2017; Pati et al., 2016; Priddy et al., 2017). There is even evidence to suggest that intermittent access is necessary for escalation of ethanol intake (see Rosenwasser, Fixaris, Crabbe, Brooks, & Ascheid, 2013). To determine whether intermittent solution access would influence ethanol consumption in this model, Experiment 2 compared intake levels between adolescents receiving daily access to ethanol and those only receiving access on a “Monday-Wednesday-Friday” intermittent schedule.

In all of our preceding work in developing this high-intake model (Hosová & Spear, 2017; Exps. 1a-d, 2), only the intake of adolescent animals was examined. Other studies, using a variety of means of assessing ethanol intake (e.g., home cage access, limited access to sweetened or unsweetened ethanol, etc.), have often found greater g/kg ethanol intakes in adolescent than adult rats (see Doremus et al., 2005; Garcia-Burgos et al., 2009; Maldonado et al., 2008). In order to determine whether the observed cyclical pattern of episodic binge-like consumption intermixed with bouts of lower consumption was adolescent-specific, adults were also tested using this intake system in Experiment 3.

Together, these studies provide critical information regarding the test conditions that promote high voluntary ethanol intake in adolescent Sprague-Dawley rats and that can be used to model adolescent drinking of alcohol in outbred rats.

Methods

Subjects

Animals used in these experiments were adolescent and adult, male and female Sprague-Dawley rats, bred in our colony at Binghamton University. On the first day after birth (postnatal day [P]1), litters were culled to 8-10 male and female pups and housed with their dams until weaning on P21. On P21, animals were pair-housed with a same-sex, same-age partner from a different litter so that cage mates could be assigned to the same test condition without placing more than one animal/sex from a given litter into any test condition. Animals were housed in a temperature controlled (20-22°C) vivarium on a 12-/1 2-h light/dark cycle (lights on at 7:00 a.m.), with ad libitum access to food (Purina Rat Chow, Lowell, MA) and tap water, except where noted. Unless otherwise specified, animals were provided access to water in their home cages through ball-bearing sipper tubes to acclimate them to the tubes used for ethanol intake. All animal maintenance and experimental procedures were conducted in accord with the guidelines established by the National Institutes of Health, using protocols approved by the Binghamton University Institutional Animal Care and Use Committee.

Food restriction

In accord with our prior work (Hosová & Spear, 2017), animals in Exps. 1a-d were food restricted from P24 (four days before the onset of solution access sessions) through to P41 (the final day of testing). Using age/weight charts and standard procedures in our laboratory (see Anderson et al., 2013), animals were gradually reduced to and then maintained at 85% of projected free-feeding body weights. Animals were weighed daily and fed 1-3 hours after the daily solution access session, with the amount of food given to each pair of cage mates adjusted as necessary to maintain target body weight trajectories.

Blood ethanol determination

Unless otherwise noted, on intake days 6 and 14 (P33 and P41, respectively), blood samples were collected for analysis of blood ethanol concentrations (BECs). Approximately 30μl was collected via lateral tail vein incision from each animal within 15 minutes of completing the drinking session. Samples were immediately frozen and stored at −80°C until assessed using a Hewlett Packard 5890 series II Gas Chromatograph (Wilmington, DE) (see Ramirez & Spear, 2010, for further details).

Experiment 1a – Ball-Bearing vs. Open-Ended Sipper Tubes

A total of 32 animals were used, n = 8 animals per group as delineated by the 2 (sex) × 2 (tube type) factorial design. From P28 to P41, males and females were given daily home cage access to the 10% ethanol-Boost® solution via either ball-bearing or open-ended sipper tubes for 30 minutes. Prior to the onset of each access period, animals were weighed, home cages were placed side by side on a countertop in a dim running room (25-30 lux), food and water were removed from the cage, and animals in each cage mate pair were separated from each other by a wire mesh divider. At that time, a clean white plastic food dish was placed on each side of the divided cage. After fifteen minutes of habituation, each animal then received 1400 mg of pre-weighed banana-flavored pellets (Bio-Serv, Pleasant Prairie, WI) dropped into the food dish, as well as access to ethanol-Boost® for 30 minutes. Testing occurred in sex-specific running rooms between 1:00-3:00 p.m., with a white noise generator on throughout habituation and solution access. After the 30 minute access period, the test solution, food dishes, and wire mesh dividers were removed from the cage, and animals were returned to the vivarium where home cage food and water access conditions were restored. Bottles were weighed before and after access, with the difference converted into grams of ethanol/kilogram body weight (g/kg) intakes.

Experiment 1b – Separated vs. Isolated

Testing was conducted identically to animals in the ball-bearing condition in Exp. 1a, except that testing was conducted earlier in the day (between 10:00 a.m. – 1:00 p.m.) and all cages were separated from each other for testing using a 24 in. wide × 12 in. high × 1 in. thick board of cork (Bangor Cork, Nazareth, PA) placed on each side of the cages to prevent animals from seeing peers in adjacent cages on the countertop. Ethanol access was provided in this experiment in test cages that were identical to the home cage but only used for the daily testing sessions, with the same test cage used for each animal or animal pair across all fourteen days of access. Animals in the separated condition were separated from each other within the same cage by a wire mesh divider as in Exp. 1a. Animals in the isolated condition were placed one per test cage, still contained to one side of their testing cage by identical wire mesh dividers. Eight animals were assigned to each of the four experimental groups defined by the 2 (sex) × 2 (test condition) factorial design. One pair of cage mates (females in the isolated condition) were removed from the study, however, because of an unrelated health issue, resulting in an n of 6-8 animals per group.

Experiment 1c – With or Without Concurrent Access to Banana-Flavored Pellets

Animals were tested identically to those in the ball-bearing condition in Exp. 1a, with a few additional considerations. Half of the animals received 1400 mg of banana-flavored pellets into the white food dish at the time of bottle placement while the other half had food dishes that remained empty. Testing was conducted in the late afternoon (between 3:00 – 5:00 p.m.). Due to an unforeseen electrical power outage, the first group of animals in this experiment were tested in the dark and in the absence of white noise on the first access day and, serendipitously, intakes were promisingly high on that day. Hence, we decided to assess the effect of lighting condition by testing animals in rooms with either the standard, previously-used dim lighting or red lighting (both approximately 25 lux), resulting in a 2 (sex) × 2 (pellet condition) × 2 (lighting condition) design. Beginning in this experiment, all animals were also assessed upon return to the vivarium after each day’s solution access period for visual signs of intoxication using the Chandler-Crews Intoxication Scale (NADIA Consortium, see https://www.med.unc.edu/alcohol/nadiaconsortium/standardized-methods). Each subject was observed in its home cage for 30 seconds for motor impairment, and was assigned a score from 1, indicating natural motor behavior, to 5, severe impairment reflected by the loss of reflexes such as the righting reflex and the eye-blink reflex. A total of 32 animals were used for Exp. 1c, n = 4 animals per group as defined by the 2 (sex) × 2 (pellet condition) × 2 (lighting condition) factorial.

Experiment 1d – Food Restricted vs. Ad Libitum Feeding

The design of this study was a 2 Sex × 2 Feeding Condition (restricted or unrestricted) factorial. Testing was conducted under red light and without pellets or white noise as in Exp. 1c, with the singular exception that only half of the animals were food restricted from P24 through to P41. The other half of the cohort had ad libitum access to standard cage top chow pellets at all times from P28 to P41 except for less than one hour every day, during which animals were weighed, given 15 minutes of pre-test habituation, and 30 minutes of access to the 10% ethanol-Boost® solution. Unlike the previous experiments, in Exp. 1d the first blood sample collection day occurred either on P33 or P34. The decision of whether to collect samples on P33 or P34 was made based on that day’s average ethanol intakes – i.e., samples were collected on P34 when intakes on P33 were collectively low, thereby allowing us to characterize more clearly the magnitude of the BECs induced in our voluntary consumption model by increasing the probability of assessing BECs attained on a high (rather than low) consumption day. The second sample collection day remained on P41. Because the first sampling day was chosen to reflect a relatively high consumption day whereas the second was not, in the analyses across sampling day main effects or interactions involving day were not considered, though data from both collection days were still examined. A total of 32 animals were used in Exp. 1d, with an n of 8 in each of the four groups as defined by the 2 (sex) × 2 (feeding condition) design.

Experiment 2 – Daily vs. Intermittent Access to Ethanol

Testing was conducted nearly identically to the unrestricted feeding condition in Exp. 1d except that only half of the animals were given access to ethanol every day (“Daily”) whereas others were given access on a “Monday-Wednesday-Friday” schedule (“Intermittent”), from P28-P42. The testing range was extended by one day from previous experiments so that both daily and intermittent animals would drink on P42 prior to the second blood sample collection. Therefore, daily animals had a total of fifteen ethanol solution access sessions, while intermittent animals were provided a total of seven sessions. A total of 46 animals were used, with an n of 10–12 in each of the four groups as defined by the 2 (sex) × 2 (access condition) design.

Experiment 3 – Adolescent vs. Adult Consumption of Ethanol-Boost®

The design of this study was a 2 (sex) × 2 (age – adolescent, adult) factorial, with sample sizes of 10 animals per group. From P28-P41 or P70-P83, adolescent and adult animals of both sexes were given daily access to the 10% ethanol in chocolate Boost® (v/v) solution in a manner identical to the daily access animals in Exp. 2. As in Exp. 1d, the first blood sample collection day again occurred either on P33 or P34 (adults: P75 or P76), based on the average intakes on P33/75, in order to increase the probability of assessing BECs attained by animals on a high consumption day. Due to this, we again did not consider any main effects or interactions involving day in analyzes across sample collection day.

Results

Experiment 1a – Ball-Bearing vs. Open-Ended Sipper Tubes

Average 30-minute voluntary ethanol consumption across all access days (P28-P41) was 3.8 (± 0.12) g/kg, and BECs over the two blood sample collection days averaged 73.0 (± 5.44) mg/dl (range: 0-216.9 mg/dl). A 2 sex by 2 access tube type analysis of variance (ANOVA) revealed a significant effect of sex on mean intake (F(1,28)=8.436, p<0.01), with males displaying higher consumption than females overall (see Fig. 1a). There was no main effect or interaction involving tube type. A 2 sex × 2 tube type × 2 day repeated measures ANOVA on the BEC data revealed only a significant effect of sample collection day; BECs were significantly higher overall on the first collection day (P33) (84.2 ± 8.6 mg/dl) than the second (P41) (61.9 ± 6.2 mg/dl) (F(1,28)=4.498, p<0.05). A repeated measures ANOVA of average intake just on these two collection days found no significant effect (nor trend) of day or tube type on intake; thus, despite intakes being roughly equivalent on the two blood sample collection days, the resulting BECs were lower on the second day. This day effect could reflect either the development of some degree of metabolic tolerance or an age effect, although both ages fell during the mid-adolescent range within which marked age differences are not expected (see Silveri & Spear, 2000; Varlinskaya & Spear, 2006). BECs across both sample collection days were significantly correlated with intake (r=0.828, p<0.001). Due to the lack of tube effects in this experiment and the greater accuracy presumably provided by ball-bearing sipper tubes, animals in all subsequent experiments were given access to the test solution using ball-bearing tubes, with the same tube type also used for home cage ad libitum water access.

Figure 1. Exps. 1a-d: Mean Ethanol Intake Levels (g/kg).

Mean ethanol intake levels (g/kg) by adolescent animals in Exps. 1a-d. Fig. 1c is collapsed across pellet condition. In Fig. 1d, note that although higher consumption levels were observed in food restricted animals, BECs were equivalent or in fact higher in the unrestricted animals, prompting our decision to discontinue food restriction as part of this model.

* - significant main effect of sex. # - significant main effect of test condition.

Experiment 1b – Separated vs. Isolated

Average consumption of ethanol across all groups was 2.6 (± 0.14) g/kg over the fourteen days of access, with BECs on the two sampling days in the moderate range (average across groups of 40.7 (± 5.07) mg/dl; range: 0-200.3 mg/dl). A 2 sex by 2 testing condition ANOVA found a significant effect of sex on mean intake (F(1,26)=12.743, p<0.005), with male subjects again averaging higher consumption than females; there was also a trend for an effect of testing condition (p=0.086), wherein animals in the separated condition tended to have higher intakes than their counterparts tested in isolation (see Fig. 1b). Surprisingly, this trend was driven primarily by female animals, with the intakes of males similar across testing condition. Repeated measures ANOVAs used to examine the BEC and intake data across both sample collection days revealed a trend for an effect of day (p=0.057) on BECs but not intake, with BECs once again tending to be higher on the first sample collection day (P33) than the second (P41) – a trend reminiscent of the more robust day-related findings evident in Exp. 1a. BECs were again highly correlated with ethanol consumption on the two sample collection days (r=0.779, p<0.001). Overall, intake levels in this experiment tended to be lower than in Exp. 1a and in our prior work (Hosová & Spear, 2017) for reasons that were not readily apparent, although it should be noted that this experiment was the first to use cork board dividers and to conduct consumption testing in the morning. On the basis of these findings, all subsequent experiments were conducted in the afternoon and without cork board dividers, with animals provided solution access in the home cage while separated from their cage mate by a wire mesh divider.

Experiment 1c – With or Without Concurrent Access to Banana-Flavored Pellets

The factorial ANOVA on mean intake revealed significant main effects of sex (F(1,24)=15.083, p<0.001) and lighting condition (F(1,24)=9.332, p<0.01), with males averaging higher ethanol consumption levels than females, and red light animals consuming more ethanol on average than dim light animals (see Fig. 1c). Pellet condition did not have an effect on mean intake. Average voluntary ethanol intakes across all groups was 3.2 (± 0.12) g/kg, resulting in an overall average BEC on the two blood sample collection days of 44.5 (± 4.66) mg/dl (range: 0-104.1 mg/dl). When analyzing the data over the two blood sample collection days, several interactions with complicated patterns of effects emerged – findings that should be tempered due to the low sample sizes (n = 4 per Sex × Pellet Condition × Lighting Condition group) associated with addition of the lighting variable. Analyzing the BEC data from P33 and P41 via a repeated measures ANOVA revealed a significant main effect of sex (F(1,24)=5.7886, p<0.05, males with higher BECs than females), as well as a significant three-way interaction of day, lighting condition, and pellet condition (F(1,24)=9.9844, p<0.01). In general, BECs were lower in the pellet than non-pellet condition; this was evident in all test day by light condition groups, except for red light on sample day one (where the converse was seen) (see Table 1). A repeated measures ANOVA of ethanol intakes on the two sample collection days yielded a main effect of sex (F(1,24)=14.5113, p<0.001, males consuming more), as well as a main effect of lighting condition (F(1,24)=9.2183, p<0.01) which was tempered by another significant three-way interaction of day, lighting condition, and pellet condition (F(1,24)=5.4149, p<0.05). The pattern of this interaction is nearly identical to what was seen with the BECs, with BECs being less in the pellet than non-pellet condition under both lighting conditions, except on the second sample collection day where solution intake under dim lighting was the same regardless of pellet condition. Intakes on sample collection days and the resulting BECs were again significantly correlated (r=0.778, p<0.001). Animals averaged a rating of 1.6 (± 0.07) on the Chandler-Crews scale of visible intoxication, with these ratings being correlated significantly with their individual intake (r=0.54, p<0.001) but not associated with sex, lighting condition, or pellet condition. Given that mean ethanol intake levels were significantly higher in animals tested under red light than dim light, all subsequent testing in this series was conducted under red light. Additionally, although pellet condition did not significantly impact mean ethanol intake, BECs tended to be higher in the animals that were not provided pellet access; hence, in subsequent experiments, animals were not given pellet access during the ethanol-Boost® intake sessions.

Table 1: Ethanol Intake (g/kg) and BECs (mg/dl) from Exp. 1c.

Ethanol intake (g/kg) and BEC (mg/dl) data from Exp. 1c animals on the first (P33) and second (P41) blood sample collection day, as well as the overall mean from all fourteen days of access. “Pellet condition” refers to whether the animal was provided with simultaneous access to banana-flavored pellets during the 30 minutes of ethanol solution access. BEC = Blood Ethanol Concentration. Data presented as mean (± SEM).

Measure	Sex	Pellet Condition	P33 Mean		P41 Mean		Overall Mean
			Red Light	Dim Light	Red Light	Dim Light	Red Light	Dim Light
Ethanol Intake (g/kg)	Males	Pellets	5.3 (± 0.5)	2.3 (± 1.1)	4.8 (± 0.9)	4.1 (± 0.6)	4.0 (± 0.2)	3.4 (± 0.4)
		No Pellets	3.0 (± 1.0)	3.8 (± 1.3)	4.9 (± 0.5)	2.8 (± 0.3)	3.6 (± 0.2)	3.3 (± 0.1)
	Females	Pellets	3.6 (± 1.1)	1.5 (± 0.5)	2.6 (± 0.6)	1.3 (± 0.5)	3.4 (± 0.1)	2.3 (± 0.4)
		No Pellets	2.7 (± 1.0)	3.0 (± 0.5)	2.9 (± 0.5)	2.0 (± 0.1)	3.0 (± 0.4)	2.9 (± 0.1)
BEC (mg/dl)	Males	Pellets	60.5 (± 11.4)	19.4 (± 11.3)	57.4 (± 15.5)	51.5 (± 11.9)	58.9 (± 8.9)	35.5 (± 9.7)
		No Pellets	38.4 (± 4.8)	50.8 (± 18.5)	82.0 (± 8.5)	55.5 (± 6.5)	60.2 (± 9.4)	53.1 (± 9.1)
	Females	Pellets	47.0 (± 14.5)	21.3 (± 9.6)	31.6 (± 16.7)	16.5 (± 10.2)	39.3 (± 10.6)	18.9 (± 6.6)
		No Pellets	29.1 (± 10.8)	54.5 (± 5.6)	58.3 (± 11.5)	37.5 (± 4.2)	43.7 (± 9.2)	46.0 (± 4.6)

Experiment 1d – Food Restricted vs. Ad Libitum Feeding

Across all groups, the average amount of ethanol consumed in 30 minutes was 2.7 (± 0.12) g/kg, with average BECs across the two sample collection days of 68.0 (± 3.95) mg/dl (range: 0-150.9 mg/dl). A 2 sex by 2 feeding condition ANOVA of mean intake across all days found a significant effect of sex (F(1,28)=7.495, p<0.05), reflecting higher average consumption by the males than females, and feeding condition (F(1,28)=29.172, p<0.001), with food-restricted animals averaging greater intakes than unrestricted animals (see Fig. 1d). The repeated measures ANOVA of ethanol intake on just the two blood sample collection days again revealed significant effects of sex (F(1,28)=9.04, p<0.01) and feeding condition (F(1,28)= 9.78, p<0.01), in this case tempered by an interaction of these two variables (F(1,28)=5.2, p<0.05). Sex differences in ethanol intake were most pronounced when food deprived, with females averaging similar ethanol intakes across feeding condition, but males averaging higher intakes when food-restricted. Surprisingly, despite the significantly greater average ethanol intake of the food restricted animals, a repeated measures ANOVA of BECs did not find differences across feeding condition or sex, and if anything, BEC levels tended to be greater in the unrestricted animals (see Table 2 and Fig. 1e). Intakes on sample collection days were once again significantly correlated with the resulting BECs, albeit to a lesser extent than in previous experiments (r=0.411, p<0.001). This less robust effect was not related to the inclusion of non-deprived groups of animals: when correlational analyses were conducted separately within each deprivation condition, a higher correlation was obtained between intake and resulting BECs in animals that were not food restricted (r=0.697, p<0.001) than in animals that were (r=0.407, p<0.05). The Chandler-Crews intoxication rating scale yielded an average score over all access days of 1.3 (± 0.03); these scores correlated significantly with intake (r=0.28, p<.001), but were uninfluenced by sex or deprivation condition. Based on these results, animals tested in Exps. 2 and 3 were maintained on an ad libitum feeding schedule. This regimen was selected because, despite seeing greater ethanol intake levels in the restricted animals, BECs did not differ across groups and even tended to be higher in unrestricted animals. Ultimately, the amount of ethanol that enters the circulatory system is more relevant to this model than the amount initially consumed and entering the gastrointestinal tract.

Table 2: Ethanol Intake (g/kg) and BECs (mg/dl) from Exp. 1d.

Ethanol intake (g/kg) and BEC (mg/dl) data from Exp. 1d animals on the first (P33) and second (P41) blood sample collection day, as well as the overall mean from all fourteen days of access. “Feeding condition” refers to whether the animal was moderately food-restricted to 85% of projected free-feeding body weight or allowed ad libitum food access outside of testing. Note that while restricted animals averaged greater ethanol intake, BECs were equivalent if not greater in the unrestricted animals, suggesting a hepatoprotective effect of food restriction. BEC = Blood Ethanol Concentration. Data are presented as mean (± SEM).

Measure	Sex	Feeding Condition	P33 Mean	P41 Mean	Overall Mean
Ethanol Intake (g/kg)	Males	Restricted	5.3 (± 0.5)	3.5 (± 0.3)	3.4 (± 0.2)
		Unrestricted	2.7 (± 0.5)	3.1 (± 0.3)	2.5 (± 0.2)
	Females	Restricted	3.9 (± 0.6)	1.9 (± 0.4)	2.9 (± 0.2)
		Unrestricted	3.0 (± 0.4)	2.4 (± 0.3)	2.1 (± 0.1)
BEC (mg/dl)	Males	Restricted	66.7 (± 8.1)	62.3 (± 8.2)	64.5 (± 5.6)
		Unrestricted	63.8 (± 7.4)	95.1 (± 14.8)	79.4 (± 9.0)
	Females	Restricted	68.7 (± 7.2)	50.0 (± 16.1)	59.3 (± 8.9)
		Unrestricted	67.9 (± 12.4)	70.0 (± 9.4)	68.9 (± 7.5)

Figure 1e. Correlation between Ethanol Intake (g/kg) and BEC (mg/dl) in Exp. 1d.

Correlation between ethanol intake (g/kg) and resulting BEC (mg/dl) in Exp. 1d animals on the first and second blood sample collection day. On the first sampling day, food-restricted animals’ ethanol intake was not correlated with the resulting BEC [r=0.09, p=0.75; Y= 62.81+1.05*(intake)], while in the unrestricted animals the relationship was highly significant [r=0.77, p<0.001; Y= 17.34+16.99*(intake)]. On the second sampling day, intake and BEC were significantly correlated in the food-restricted animals [r=0.64, p<0.01; Y= 6.94+17.91*(intake)] as well as the unrestricted [r=0.79, p<0.001; Y= −11.86+34.27*(intake)].

Experiment 2 – Daily vs. Intermittent Access to Ethanol

Across all groups, the mean amount of ethanol consumed in 30 minutes was 1.8 (± 0.09) g/kg, with an average BEC across both sample collection days of 44.3 (± 4.84) mg/dl (range: 0-135.5 mg/dl). A 2 sex by 2 access condition ANOVA of mean ethanol intake across all days revealed a main effect of access condition (F(1,42)=14.7764, p<0.001), tempered by a significant interaction of access condition and sex (F(1,42)=12.6299, p<0.001). Females in both groups displayed very similar levels of mean ethanol consumption, whereas daily-access males consumed considerably more ethanol than did intermittent-access males or than their daily-access female counterparts (see Fig. 2a). 2 sex by 2 access condition repeated measures ANOVAs of ethanol intakes and of BECs on the two sampling days revealed a significant main effect of access condition: F(1,42)=7.6512, p<0.01; F(1,40)=4.6464, p<0.05, respectively, with daily-access animals consuming more (2.2 ± 0.20 g/kg) and achieving higher mean BECs (52.5 ± 6.27 mg/dl) than intermittent-access animals (mean intake: 1.5 ± 0.24 g/kg; BECs: 35.4 ± 7.08 mg/dl). Ethanol intakes on blood sample collection days were significantly correlated with the resulting BECs (r=0.764, p<0.001). The mean Chandler-Crews intoxication rating across all animals was 1.4 (± 0.06) – intoxication ratings that were correlated with intake (r=0.37, p<0.01) but not significantly related to other variables. The findings from Experiment 2 indicate that producing high levels of voluntary ethanol consumption within our model is better accomplished through daily rather than intermittent solution access.

Figure 2. Exps. 2 and 3: Mean Ethanol Intake Levels (g/kg).

Mean ethanol intake levels (g/kg) by adolescent animals in Exp. 2 (Fig. 2a) and 3 (Fig. 2b).

* - significant effect of sex. # - significant effect of test condition.

Experiment 3 – Adolescent vs. Adult Consumption of Ethanol-Boost®

Across all groups, the mean amount of ethanol consumed was 1.8 (± 0.10) g/kg and the average BEC across both sample collection days was 58.7 (± 4.27) mg/dl (range: 0.9-167.3 mg/dl). A 2 (sex) × 2 (age) factorial ANOVA on mean ethanol intake revealed a significant effect of age (F(1,36)=18.2495, p<0.001), reflecting higher mean consumption of ethanol by adolescent than adult animals (see Fig. 2b). Adolescent intake appeared to vary across day to a greater extent than adult intake (see Fig. 3), in the same pattern of 1-2 high consumption days followed by 1-2 low consumption days as in our previous work (Hosová & Spear, 2017). However, a factorial ANOVA on coefficient of variance (CV; calculated as standard deviation/mean) did not yield any significant effects of age or sex on the extent of variability in consumption. Surprisingly, no effect of sex on mean intake emerged in this experiment, unlike the rest of the series. Given that adolescent male rats usually consume more ethanol than their female counterparts, while the opposite is typically seen in adult rats, we conducted one-way ANOVAs separately at each age; these analyses, however, still did not uncover an effect of sex on mean intake. The analyses conducted only on the blood sample collection days revealed no significant age- or sex-based differences in consumption on these two days (adolescent mean: 2.0 (± 0.21) g/kg, adult mean: 1.7 (± 0.16) g/kg), with BECs likewise not differing (adolescents: 54.5 (± 6.34) mg/dl, adults: 62.9 (± 5.73) mg/dl). Ethanol intakes on blood sample collection days again correlated significantly with BECs (r=0.722, p<0.001) and this did not differ between age or sex groups. Animals averaged an overall rating of 1.9 (± 0.13) on the Chandler-Crews scale of visible intoxication, with the 2 (sex) × 2 (age) factorial ANOVA revealing a significant main effect of age (F(1,36)=8.92, p<0.01), with adolescents averaging significantly lower ratings (1.4 ± 0.07) than adults (2.5 ± 0.26). Thus, although (as discussed above) adolescent animals consumed significantly more ethanol in the analysis conducted across all access days than did adults, adolescent animals exhibited significantly lower intoxication scores when assessed after each of the access sessions.

Figure 3. Exp. 3: Voluntary Consumption Across Day.

Voluntary consumption (g/kg) of 10% ethanol in chocolate Boost® (v/v) by adolescent (Fig. 3a) and adult (Fig. 3b) Sprague-Dawley rats of both sexes in Exp. 3.

Discussion

Collectively, the work presented here supports our goal of establishing an easy-to-implement model of adolescent voluntary ethanol consumption that follows a cyclical pattern of high, binge-like intake separated by several days of reduced intake in outbred rats. Across this series, we found that sipper tube type (1a) and simultaneous sugar pellet access (1c) did not affect ethanol consumption, whereas intake was significantly higher in separated (1b), food-restricted (1d), daily-access (2), and adolescent (3) animals. Assessment of BECs generally revealed similar effects, except that food restriction appeared to lower BEC levels, perhaps due to hepatoprotective effects. These findings support the use of an adolescent consumption model that utilizes ball-bearing sipper tube tips to provide daily access to 10% ethanol in chocolate Boost® to free-feeding adolescent Sprague-Dawley rats separated from their cage mates by wire mesh during the access period, with no food provided during the 30 minute access period. This model of ethanol consumption produces BECs in the binge range on some consumption days in the animals, and a cyclical pattern of alternating high- and lower consumption days as reported in our earlier work with this model (Hosová & Spear, 2017).

The high consumption by adolescents of the ethanol-Boost® solution was apparent even without food deprivation (Exp. 1d); hence, the avidity of adolescent rats for this high-sugar, high-calorie solution does not appear to be driven by caloric need, but rather by its palatability. This highly palatable solution is reminiscent of the types of ethanol solutions preferred by human adolescents as well. Human adolescents are the main consumers of so-called “alcopops”, beverages that contain ethanol in a typically sweet vehicle (e.g., wine coolers [wine in fruit juice], Mudshakes™ [vodka in chocolate milk], and so on). Studies have found that the sweet taste of alcopops, along with targeted marketing efforts, have made these beverages highly popular with adolescents (Forsyth, 2001; Lanier, Hayes & Duffy, 2005; Waiters et al., 2001; cited in Kraus, Metzner, & Piontek, 2010), an age demographic that also tends to report drinking for effect while simultaneously disliking the taste of alcohol (Hughes et al., 1997; White & Hayman, 2004). In fact, alcopops are increasingly being reported as the first type of alcoholic beverage consumed by adolescents (e.g., Copeland et al., 2007) and are associated with an earlier age of first drunkenness (Kraus, Metzner, & Piontek, 2010). Not surprisingly, their popularity declines linearly with age across adolescence and into early adulthood (Copeland et al., 2007; Siegel et al., 2011). While use of ethanol in a high-calorie sweetened vehicle is reminiscent of intake patterns of drinking youth, it should be noted that there is some evidence that long-term effects of adolescent sucrose exposure may mask the long-term effects of ethanol exposure. For example, Vendruscolo et al. (2010) found reduced ethanol consumption (but not intravenous self-administration) in adult male Wistar rats that had been given continuous free access to sucrose (with or without ethanol) during mid-adolescence. Such findings point to the importance of including a Boost®-only control group in studies using this model to assess the long-term effects of adolescent ethanol exposure.

When food restricted animals were compared with free-feeding animals in Exp. 1d, the food restricted animals generally consumed more ethanol than did the unrestricted animals. However, analysis of the BEC data revealed that this did not result in greater amounts of ethanol in circulation. Specifically, on the first blood sample collection day, intakes were markedly higher in the restricted animals and yet the BECs were equivalent between groups. On the second sample collection day, when intakes happened to be nearly equal between restricted and unrestricted animals, the resulting BECs were substantially lower in the restricted group. The disparity between g/kg intakes and BECs cannot be explained by the between-group differences in average body weight. As would be expected, unrestricted animals had greater body weights (1^st collection day: 140.8 ± 3.2 g, 2^nd collection day: 193.3 ± 7.0 g) than restricted animals (1^st collection day: 112.6 ± 2.7 g, 2^nd collection day: 156.1 ± 3.2 g). Yet, it was the larger, non-deprived adolescents (who did not drink as much ethanol relative to their body weight as their deprived counterparts did) that had higher concentrations of ethanol in their circulatory system. The likely explanation for this finding is that food restriction enhances ethanol metabolism. Vučević et al. (2013) found evidence of just such an effect in male Wistar rats: blood samples taken from the tip of the tail 15 minutes after administration of 2 g/kg ethanol revealed higher BECs in the ad libitum fed males (mean BEC: 196.2 ± 15.4 mg/dl) than in those restricted to 60-70% of their usual caloric intake (mean BEC: 107.9 ± 16.9 mg/dl). Interestingly, relatively severely food-deprived animals (40-50%) displayed the highest BECs of all (mean: 258.4 ± 18.6 mg/dl) (Vučević et al., 2013). The authors suggested that these results were due to a hepatoprotective effect of moderate caloric restriction, with this modest restriction increasing the efficiency of the liver and leading to more rapid ethanol metabolism (Vučević et al., 2013). Collectively, their findings in conjunction with our data indicate that modest to moderate food restriction directly enhances ethanol metabolism to lower relative BECs, compared with those of non-restricted animals.

In Exp. 1b, animals that were placed in individual cages during solution access tended to drink less than those that were only separated from their cage mate by wire mesh during the intake session. Our decision to conduct this experiment was based on preliminary evidence (Hosová & Spear, 2017) that male adolescents might consume more ethanol if they are simultaneously able to interact, albeit in a limited fashion, with their cage mate than when completely removed from the presence of their cage mate. Unexpectedly, however, this trend for greater intake in the separated condition was driven by the female animals in Exp. 1b, whereas the males consumed similar amounts of ethanol under both testing conditions. This finding was surprising as Varlinskaya et al. (2015) have shown that adolescent males, but not females, appear to engage in high ethanol consumption for its socially facilitating effects. However, overall ethanol intakes were relatively low in Exp. 1b, possibly due to testing earlier in the day; the limited number of animals engaging in high ethanol consumption per se in this experiment may have contributed to the lack of an effect of social isolation. Additionally, although isolated males were alone in an individual cage, they could very likely still hear and smell the other males present in the testing room, which could have circumvented any potential deficit in their intake as a result of being confined to an individual cage.

Exps. 1a and c addressed a number of other methodological issues including sipper tube type, lighting condition, and food access during testing. Earlier work in our lab found that adolescent rats voluntarily consumed significantly more ethanol when access was provided through ball-bearing sipper tube tips rather than through open-ended tips, with only the former group displaying a significant preference for a sweetened ethanol solution over sweetened water (Doremus, Brunell, Rajendran & Spear, 2005). This effect of sipper tube type on mean intake was not seen in Exp.1a in the present series, possibly due to the seemingly higher palatability of ethanol in Boost® than the sweetened water solution used previously. Indeed, we have observed that the dose of ethanol required to produce a conditioned taste aversion to plain chocolate Boost® in adolescent rats is substantially higher than the dose required to induce an aversion to sweetened water (i.e. “supersac”, 3% sucrose and 0.125% saccharin; unpublished observation). Furthermore, in our previous series we found that adolescent consumption of plain chocolate Boost® was significantly higher than the consumption of the ethanol-Boost® solution (Hosová & Spear, 2017). Together, these observations suggest that adolescent Sprague-Dawley rats find the chocolate Boost® to be highly palatable, more so than supersac, with this high palatability perhaps masking effects of sipper tube type on voluntary consumption. In Exp. 1c, there was no effect of pellet condition on mean ethanol intake across access days, although no-pellet animals tended to have higher BECs than their pellet-provided counterparts. This could have stemmed from either (1) an effect of concurrent food digestion on ethanol metabolism or (2) delayed onset of solution consumption by the pellet animals, perhaps owing to prioritization of pellet consumption over ethanol consumption – possibilities that could be addressed in the future through the use of consumption-tracking technology, such as lickometers. In this experiment, however, light was observed to influence consumption, with ethanol intake of animals tested under red light being greater than when tested under dim lighting conditions. This effect, while in need of replication due to power limitations in Exp. 1c, seems to be in accord with what is known about white light suppressing the behavior of nocturnal animals (e.g., ‘masking’; see Redlin, 2001, for review).

A final and critical methodological variable was examined in Exp. 2: the impact of providing intermittent rather than daily access to the ethanol-Boost® solution. Since the seminal work of Wise (1973), researchers in the field of voluntary ethanol consumption have repeatedly demonstrated that intermittent access paradigms often induce substantially greater intake levels than do continuous access paradigms (e.g., Carnicella, Amamoto & Ron, 2009; Hwa et al., 2011; Loi et al., 2010; Simms et al., 2008). Indeed, along with selective breeding and drinking-in-the-dark paradigms, providing rodents with intermittent access to ethanol solutions has become something of a gold standard for facilitating intake (see Becker & Ron, 2014, for discussion). However, in our work here, daily-access adolescent males were found to consume significantly more ethanol, and had significantly higher BECs, than their intermittent-access counterparts, while adolescent females remained impartial to access schedule. While unexpected, this finding may be a result of our model’s short access period. Most continual-access paradigms, whether free- or forced-choice, provide solution access for 24 hours and do not typically engender high intake levels. Limiting the solution availability – whether only to specific days (intermittent access) or specific portions of the circadian cycle (drinking in the dark) – seems critical for successfully inducing high intake. Therefore, it may be that the daily-but-ultra-short 30 minute access period, in conjunction with the high palatability of the ethanol-Boost® solution, contributes to the cyclical pattern of episodic binge-level intake and corresponding elevated BECs followed by several days of lower intake that is typical of our adolescent animals.

In Exp. 3, adolescents consumed significantly more ethanol than the adults did over the fourteen days of solution access. This finding is consistent with age differences that have been observed using other intake procedures in adolescent versus adult rats (Doremus, Brunell, Rajendran & Spear, 2005; Vetter, Doremus-Fitzwater & Spear, 2007) and are akin to age differences in per-occasion ethanol intake observed in human adolescents (SAMHSA, 2016). Animals were assessed for visible intoxication after every intake session, and analysis of these data revealed a significant age difference; specifically, the adult animals consistently scored higher on the Chandler-Crews intoxication scale than their adolescent counterparts, despite the younger animals consuming more ethanol per session on average. These findings are reminiscent of previous reports of an attenuated sensitivity of adolescents to the intoxicating effects of ethanol relative to adults (see Spear, 2013, for review), including a reduced sensitivity to ethanol’s motor discoordinating effects. With regard to sex-based differences, we saw the typical sex-based difference in mean ethanol intake in the adult animals, with adult females consuming more ethanol than the males (e.g., Vetter-O’Hagen, Varlinskaya & Spear, 2009). No sex difference in intake was evident, however, in adolescents, findings that contrast with earlier studies in the present series as well as typical data showing that male adolescent rats consume larger quantities of ethanol than females (e.g., Hosová & Spear, 2017; Vetter-O’Hagen, Varlinskaya & Spear, 2009); whether this finding was spurious or a function of some aspect of the final test circumstances remains to be seen. No age or sex differences in BECs were observed in Exp. 3, with intakes (and their consequent BECs) coincidentally being the same across groups on the two sampling days, despite the age difference in mean intake. Taken together, the data suggest that using this model of voluntary ethanol consumption produces greater average intake in adolescents than adults, successfully modeling similar age differences seen in per-occasion ethanol use in humans. While these consumption differences present challenges for using this model to compare the subsequent consequences of such exposure during adolescence versus in adulthood, this adolescent consumption model provides an exciting, easy-to-implement technique for studying immediate and long-term effects of voluntary ethanol intake that intermittently reaches binge-like levels in short, daily access periods during adolescence.

Highlights for Hosová & Spear 2019.

• Adolescent intake was not affected by sipper tube type (1a) or food access (1c)

• Adolescent intake was greater in non-isolated (1b) and food-restricted (1d) animals

• Food restriction may have had a hepatoprotective effect, resulting in blunted BECs

• Intake levels were higher under daily than intermittent ethanol solution access

• Adolescents drank more ethanol but were rated as less intoxicated than adults