Applying behavioral economics-based approaches to examine the effects of liquid sucrose consumption on motivation

Abstract

Overconsumption of sugar contributes to obesity in part by changing the activity of brain areas that drive the motivation to seek out and consume food. Sugar-sweetened beverages are the most common source of excess dietary sugar and contribute to weight gain. However, very few studies have assessed the effects of liquid sucrose consumption on motivation. This is due in part to the need for novel approaches to assess motivation in pre-clinical models. To address this, we developed a within-session behavioral economics procedure to assess motivation for liquid sucrose. We first established and validated the procedure: we tested several sucrose concentrations, evaluated sensitivity of the procedure to satiety, and optimized several testing parameters. We then applied this new procedure to determine how intermittent vs. continuous access to liquid sucrose (1 M) in the home cage affects sucrose motivation. We found that intermittent liquid sucrose access results in an escalation of sucrose intake in the home cage, without altering motivation for liquid sucrose during demand testing (1 M or 0.25 M) compared to water-maintained controls. In contrast, continuous home cage access selectively blunted motivation for 1 M sucrose, while motivation for 0.25 M sucrose was similar to intermittent sucrose and control groups. Thus, effects of continuous home cage liquid sucrose access were selective to the familiar sucrose concentration. Finally, effects of sucrose on motivation recovered after removal of liquid sucrose from the diet. These data provide a new approach to examine motivation for liquid sucrose and show that escalation of intake and motivation for sucrose are dissociable processes.

1. Introduction

The consumption of high-sugar foods is known to promote weight gain (Togo et al., 2019), impair glucose homeostasis (Togo et al., 2019), and contribute to the development of insulin resistance (Wang et al., 2014). In addition, obesogenic (high-fat/high-sugar) foods produce activations in brain areas that regulate reward and motivation (Papp et al., 2007; Tsurugizawa et al., 2008) and diet-induced changes in the activity of these brain areas in humans contribute to the development and maintenance of obesity (Boswell & Kober, 2016; Murdaugh et al., 2012) and predict food craving (Jastreboff et al., 2013) and intake (Boswell & Kober, 2016). These data suggest that high dietary sugar promotes changes in the function of brain motivation circuits that promote unhealthy eating and thereby contribute to the development and maintenance of obesity.

Rodent models can provide useful insights into the neurobiological and behavioral consequences of eating obesogenic diets. However, studies examining the effects of obesogenic diets on motivation for food have found evidence for both increases and decreases in motivation (Arcego et al., 2020; Inbar et al., 2020; Matikainen-Ankney et al., 2022), sometimes even within the same subjects (Vendruscolo et al., 2010). The majority of these studies use instrumental tasks to assess how willing rats are to work (e.g., press a lever) to obtain a given type of food after exposure to an obesogenic diet. Most studies either use fixed ratio (FR) procedures where the effort to obtain food is relatively low (e.g., 2 responses to get one sugar pellet; FR2) or progressive ratio procedures where the effort to obtain a set amount of food progressively increases across a testing session. The amount of responding during fixed ratio sessions or the final ratio achieved during progressive ratio sessions (break point) are then used as indices of motivation. Using these approaches, Vendruscolo et al. (2010) found that instrumental responding on fixed ratio schedule was reduced following liquid sucrose consumption compared to chow fed controls, but that the break point in these same rats was similar between control and experimental groups. This makes drawing firm conclusions about effects of diet on motivation challenging. In addition, progressive ratio testing provides only one metric of motivation, limiting its sensitivity (see (Stafford et al., 1998) for additional discussion).

Demand procedures based on concepts from behavioral economics provide more nuanced measures of motivation compared to fixed or progressive ratio schedules, while also providing direct measures of intake (Bauman et al., 1996; Galuska et al., 2011; Hursh & Natelson, 1981; Hursh & Silberberg, 2008). Similar to progressive ratio approaches, demand procedures rely on animals to make instrumental responses for a given reinforcer. However, these procedures measure consumption as a function of the work required to obtain a reinforcer (i. e., price). This results in a “demand curve” where consumption is stable across low prices, and declines as price increases. Examination of the demand curve and corresponding work curve provide several different measures of motivation including the preferred level of consumption when price is negligible (Q₀), the maximum price an animal is willing to “pay” to maintain Q₀ (Pmax), and the maximum number of responses made (Rmax) (Bidwell et al., 2012; Epstein et al., 2018; MacKillop et al., 2016). Thus, behavioral economic procedures provide a structured way to simultaneously quantify several different aspects of motivation (Gilroy et al., 2021; Newman & Ferrario, 2020), rather than relying on a single metric (breakpoint in the case of PR, and rate of responding in fixed ratio procedures).

Early reports using these procedures examined the essential value of solid foods across a variety of species including humans (Epstein et al., 2018), monkeys (Hursh, 1978; Hursh & Silberberg, 2008), fowl (Dawkins, 1983; Tsunematsu, 2001), and rodents (Bauman et al., 1996; Galuska et al., 2011; Hursh & Natelson, 1981; Hursh & Silberberg, 2008). More recently, studies examining effects of diet-induced obesity in rodents have found upward shifts in the demand curve that result in increases in Q₀, Pmax and Rmax (Batten et al., 2020; Kirson et al., 2022; Rasmussen et al., 2010; Townsend et al., 2015). These results are more consistent with observations in humans (Boswell & Kober, 2016; Jastreboff et al., 2013), and oppose effects of reduced motivation found with fixed ratio or progressive ratio procedures in rodents (e.g., (Vendruscolo et al., 2010; Vollbrecht et al., 2015).

Prior studies in this area used solid food reinforcers and did not evaluate potential effects of liquids high in sugar content. This is an important consideration given that superfluous calorie intake is often attributed to sugar-sweetened beverages (Martínez Steele et al., 2016), the consumption of high-sugar liquids is strongly associated with heightened obesity risk (Malik et al., 2010) and eating high-sugar liquids produces greater body weight gain and impairments in glucose homeostasis in mice (Togo et al., 2019). Thus, understanding the neurobehavioral consequences of consuming liquid diets high in sugar is essential to understanding the causes of obesity. To do this requires the development of approaches to evaluate motivation for liquid sugar reinforcers, and examination of how consumption of high-sugar liquids affects motivation. Therefore, here we first established and validated a within-session demand procedure that captures key measures of motivation (Q₀, Pmax, and Rmax) for liquid sucrose. We tested a range of sucrose concentrations, evaluated the sensitivity of our approach to hunger and satiety, and optimized several testing parameters including the range of prices needed to accurately capture these measures. We then applied this new procedure to determine how intermittent vs. continuous access to liquid sucrose (1 M) in the home cage affects sucrose motivation, and whether removal of sucrose from the diet reverses its effects.

We found that intermittent liquid sucrose access results in an escalation of sucrose intake in the home cage, without altering motivation for liquid sucrose during demand testing (1 M or 0.25 M) compared to water-maintained controls. In contrast, continuous home cage access selectively blunted motivation for 1 M sucrose, while motivation for 0.25 M sucrose was similar to intermittent sucrose and control groups. Thus, effects of continuous home cage liquid sucrose access were selective to the familiar sucrose concentration. Finally, the effects of home cage liquid sucrose on motivation recovered after removal of liquid sucrose from the diet. Taken together, these data provide a new approach to examine motivation for liquid sucrose. They also show that escalation of intake and motivation to work for sucrose are dissociable processes that are differentially affected by free home cage sucrose access.

2. Materials and methods

General methods are provided first, followed by detailed methods for each study. Procedures described below were approved by the University of Michigan Committee on the Use and Care of Animals in accordance with AAALAC and AVMA guidelines.

2.1. Subjects, apparatus, and materials

Male outbred Sprague Dawley rats (50–55 days old on arrival; Envigo, Indianapolis, IN) were individually housed and maintained on a reverse light/dark cycle throughout (12/12; lights off at 7am). All training and testing began between 9 and 10 a.m. each day (one session/day). Rats were maintained on a modified feeding schedule in which ad libitum access to standard lab chow (5LOD; LabDiet; St. Louis, MO) was provided after the completion of daily training (~11 a.m.) and testing (2–3pm). All rats had a minimum of 5 hrs of ad libitum access to food prior to food removal at the end of the day. For all studies, rats were habituated to this modified schedule for 1 week prior to the start of training. All training and testing occurred in standard operant boxes (Med Associates; St. Albans City, VT) housed within sound attenuating chambers. Each box was equipped with a house light, a tone generator, 2 levers, or 2 nose poke ports, and a retractable sipper bottle (Med Associates; St. Albans City, VT). For all studies, sucrose concentrations were based on the existing rodent literature used to study motivation (Reilly, 1999; Sclafani & Ackroff, 2003), preference (Davis & Smith, 1992; Duca et al., 2014), hedonic responses (Robinson et al., 2015), and in sensory taste systems (Sung et al., 2022).

2.2. Instrumental training and testing

For detailed reviews on instrumental responding and demand see Fanselow and Wassum (2015), Bouton et al. (2021), and Newman and Ferrario (2020), respectively.

Rats were first given free access to a bottle containing liquid sucrose (1 or 2 M; see details below) in their home cage to familiarize them with the substance (24 hrs). During this time, water was not available. Rats were then returned to ad libitum water (24 hrs), after which instrumental training began. During instrumental training sessions (30 min/session, one session/day), responding on one lever or in one nose poke port (designated active) resulted in the presentation of the sipper bottle containing sucrose and a tone (5 s), whereas response in the other lever or nose poke port (designated inactive) had no programmed consequences. Thus each response in the active port was reinforced with presentation of the sipper bottle (i.e., Fixed Ratio 1; FR1). However, the duration of bottle availability was systematically shortened across three training phases: phase 1–30 s, phase 2–15 s, and phase 3–5 s of access per bottle presentation. Rats proceeded to the next training phase when active responding during 2 out of 3 consecutive sessions did not fluctuate more than 20%. During each training session, the number of active and inactive responses and the volume of sucrose consumed were measured.

After the above training, rats underwent within session sucrose demand testing in which instrumental responding (i.e., work) and sucrose consumption were measured across several different prices. Price was defined as the number of responses needed to obtain access to the sipper bottle (fixed ratio; FR) as a function of the duration of bottle presentation and sucrose concentration (i.e., Price = FR/(duration of access per bottle presentation * sucrose concentration (M)). Rats were tested at each price for 5 min. The number of active and inactive responses and the amount of sucrose consumed were measured for each price. Individual price trials were separated by a 25 min inter-price interval (IPI) in which responding was not reinforced. During this period, the house light was turned off, the external chamber doors were opened, and bottles were removed. Thus, demand testing ranged between 3.5 and 5 hrs for 7–10 prices, respectively.

2.3. Study 1: Characterization and optimization of the within session sucrose demand procedure

Rats were pre-exposed to 2 M sucrose in the home cage and were then trained to press one lever to gain access to 2 M sucrose as described above (N = 9). The stability of behavior during within-session demand for 2 M sucrose was then evaluated across 3 consecutive testing sessions. Prices tested are shown in Table 1 and were presented in pseudorandom order (i.e., price order selected by the experimenter to occur in non-sequential order). We then conducted a series of demand tests to refine the testing procedure and assess aspects of face validity. Specifically, to refine the procedure we compared responding when prices were presented in ascending vs. pseudorandom order and determined how varying the time between each price during demand testing (10 vs 25 min) affected work and consumption during testing. Face validity of demand procedure was determined via within-subjects comparisons under hungry vs. sated testing conditions (3 sessions per condition). To induce satiety, rats were given free access to 2 M sucrose in their home cage for 1 hr immediately prior to testing demand for this same concentration of sucrose. The volume of sucrose consumed during the 1 hr pre-exposure was recorded for each rat. Lastly, sensitivity of the procedure to shifts in sucrose concentration, an initial screen was conducted in which consumption and responses for 1 M sucrose were compared to 2 M sucrose.

Table 1.

Prices, fixed ratios (FR) and access times used for within-subjects comparison of demand for 1 vs. 2 M sucrose.

Unit Price		FR	Access (sec)
2 M	1 M
0.0125	0.025	1	40
0.025	0.05	1	20
0.05	0.1	1	10
0.1	0.2	2	10
0.2	0.5	2	5
0.5	1	3	3
1	2	4	2

We next tested demand across a range of sucrose concentrations (0.25, 0.5, 1 and 2 M) in a separate cohort of rats (N = 12). Rats were pre-exposed to 1 M sucrose in the home cage and then trained to nose poke for access to 1 M sucrose as described above. Rats were then given 3 demand tests at each concentration (one concentration per day). Testing occurred in order of increasing concentration with each test separated by one test-free day. Prices used are shown in Table 2 and were tested in ascending order. In addition, free consumption of each sucrose concentration was measured in a separate 5 min session conducted in the operant boxes (one concentration tested each day).

Table 2.

Prices, corresponding fixed ratios (FR) and access times for within-subjects comparisons of demand across 4 different sucrose concentrations.

Unit Price	0.25 M		0.5 M		1 M		2 M
	FR	Access (sec)	FR	Access (sec)	FR	Access (sec)	FR	Access (sec)
0.0125	1	320	1	160	1	80	1	40
0.025	1	160	1	80	1	40	1	20
0.05	2	160	2	80	1	20	2	20
0.1	3	120	3	60	1	10	2	10
0.2	4	80	4	40	2	10	4	10
0.5	5	40	5	20	3	6	6	6
1	6	24	6	12	5	5	10	5
2	7	14	7	7	10	5	20	5
5	10	8	20	8	25	5	50	5
10	20	8	40	8	50	5	100	5

Finally, to ensure that the modified feeding schedule did not induce a stress response in rats, circulating levels of corticosterone (Cort) were measured after the completion of all behavioral testing while rats were on the modified feeding schedule and again after one week of ad lib chow access. Briefly, tail blood was collected and centrifuged (4 °C; 10,000×g; 10 min) to obtain plasma and stored at −80 °C until use. Cort analysis was conducted via ELISA (Millipore Sigma, EZRMI-13K; Burlington MA). Samples were tested in duplicate according to manufacturer specifications and read using a Varioskan Lux plate reader (Thermo-Fisher Scientific, VL0000D0; Waltham MA).

2.4. Study 2: Effects of intermittent or continuous home cage sucrose on demand for sucrose

Here we determined how consumption of sucrose in the home cage affected demand for sucrose, and the degree to which effects are long-lasting. As above, rats were pre-exposed to 1 M sucrose in the home cage and were then trained to nose poke to gain access to 1 M sucrose (N = 41). Next, baseline demand for 0.25 M and 1 M sucrose was determined (3 sessions at each concentration; prices are shown in Table 3 and were tested in ascending order). Rats were then divided into three groups counterbalanced for baseline demand and body weight: water-maintained controls (ad lib access to water in the home cage; N = 13), continuous sucrose access (ad lib access to 1 M sucrose in the home cage; N = 14), and intermittent sucrose access (ad lib 24-hrs access to 1 M sucrose every 3rd day; N = 14; similar to (Eikelboom & Hewitt, 2016)). Access conditions were maintained for 4 weeks during which time rats had ad libitum access to food. Thus, intermittent access rats had 10 intermittent access sessions across this 4-week period. Demand for 0.25 M and 1 M sucrose were measured after this manipulation (post) and again after returning all animals to home cage water for 2 weeks (recovery). To keep demand testing conditions consistent across baseline, post, and recovery testing, rats were placed back on the modified feeding schedule and liquid sucrose was removed from their home cages during post and recovery testing (6 days). As above, sucrose consumed at each price, and active and inactive nose pokes were recorded during each demand testing session. Body weight, home cage liquid and food intake were measured throughout the home cage exposure and 2-week recovery periods. During this time, water was also measured from an empty control cage to account for liquid loss due to movement of the cages and natural fluctuations in atmospheric pressure.

Table 3.

Prices, corresponding fixed ratios (FR) and access times used to examine effects of home cage sucrose consumption on demand.

Price 0.25 M	Price 1 M	FR	Access (sec)
0.05	0.0125	1	80
0.1	0.025	1	40
0.2	0.05	1	20
0.4	0.1	1	10
0.8	0.2	2	10
2	0.5	3	6
4	1	5	5
8	2	10	5
20	5	25	5
40	10	50	5

2.5. Statistics

Statistical analyses were conducted using Prism 9 software (Graph Pad; San Diego CA). Data were analyzed using standard paired and unpaired two-tailed t-tests, one-, or two-way ANOVAs as appropriate. Standard general linear models (GLM) or mixed model residual maximum likelihood (REML) were used followed by Sidak’s post-tests. For analysis of behavior during demand testing, the average behavior of each animal during the final 2 testing sessions of the same concentration of sucrose were used. These experimentally derived data were used to plot demand and work curves and to determine Q_o, the maximum preferred level of consumption when price is low. Data from this average were fit for each rat as described previously (Newman & Ferrario, 2020) and fits were confirmed visually. This curve fitting provided two main metrics: Pmax and Rmax. Pmax is the maximum price an animal is willing to “pay” to defend Q_o (Epstein et al., 2018), where Rmax is the maximum work performed regardless of price. Rmax is presented either as a raw number or as change from each animal’s baseline. For all Rmax values calculated as a change from baseline, Wilcoxon signed-rank tests were used to determine whether increases or decreases were statistically different from 0 (i.e., no change). To facilitate comparisons across sucrose concentrations, all consumption data are reported as grams of sucrose consumed. This can be converted to mL by the following: sucrose consumed (g)/concentration (g sucrose/mL). Statistical approaches and interpretation of p-values are based on Cohen’s primer theory (Cohen, 1992) and guidelines provided by the American Statistical Association (Wasserstein & Lazar, 2016). Data reported below have a statistical power ranging from 0.84 to 1 based on post-hoc G*Power analyses of the observed data (Faul et al., 2007; Faul et al., 2009).

3. Results

3.1. Study 1: Characterization and optimization of the within session sucrose demand procedure

3.1.1. Stability of demand across repeated testing

Fig. 1 shows active lever presses and sucrose consumed during the final 3 sessions of each phase of instrumental training. Rats learned to respond for 2 M sucrose and lever pressing increased as access time to the sipper decreased (Fig. 1A; one-way RM ANOVA main effect of training phase: F (1.793, 14.34) = 36.52, p < 0.0001). The total amount of sucrose consumed remained stable across training (Fig. 1B; one-way RM ANOVA no effect of training phase: F (1.684, 13.47) = 1.192, p = 0.3254). Thus, as access time per bottle presentation was shortened, rats increased their responding to maintain sucrose consumption.

Fig. 1. Instrumental training.

(A) Average number of lever presses during each phase of instrumental training. (B) Average total sucrose consumed during each phase of instrumental training. All data shown as average ± SEM.

Fig. 2A shows sucrose consumption as a function of price during each of three within session sucrose demand testing sessions. To accurately use demand approaches to assess motivation, consumption must be flat when price is low (i.e., not changing significantly across at least three prices) and then decline to zero as price increases (Hursh, 1978, 1991; Hursh & Natelson, 1981; Hursh & Silberberg, 2008; Newman & Ferrario, 2020). Behavior during the first demand test session (Fig. 2A; closed symbols), did not generally follow this pattern; there was not a clear plateau or fall off point, but rather consumption declined steadily over the session. In contrast, behavior on tests two and three (open and gray circles) followed the expected form, with relatively stable consumption at low prices, and a decline as price increased. Further, behavior during test 2 and 3 did not differ (Fig. 2A; two-way RM ANOVA no main effect of test: F (1, 7) = 0.0567, p = 0.819). In addition, the magnitude of lever pressing (i.e., work) initially increased with increasing price as rats attempted to maintain their preferred level of consumption, but sharply declined at high prices (Fig. 2B). Overall, behavior during demand testing was relatively stable across tests 2 and 3. Thus, for all subsequent studies three demand tests were conducted for each condition examined, and data are presented as the average of tests 2 and 3, as summarized in Fig. 2C.

Fig. 2. Sucrose consumption and work across three within session sucrose demand test sessions.

(A) Average sucrose consumed as a function of price (i.e., demand) during each test session. (B) Average number of active responses as a function of price during each test session. (C) Average consumption (closed circles) and active responses (open circles) during test 2 and 3. Behavior was stable across the second and third tests and followed the expected pattern: increasing work to maintain the preferred level of consumption at lower prices, and a drop in consumption and work as price increased.

3.1.2. Varying price order and inter price interval (IPI)

Next, we assessed whether the order of price presentation or the time between prices (inter-price interval, IPI) significantly affected behavior in the task. To examine effects of price order, within session demand testing was repeated with prices presented in ascending price order (3 sessions) and compared back to behavior during initial pseudorandom testing above. Fig. 3 shows consumption (A) and work (B) averaged across test 2 and 3 using pseudorandom or ascending price order. There were no robust effects of order on consumption as a function of price (Fig. 3A). Specifically, although there was an interaction between price and order (Fig. 3A; two-way REML ANOVA price × order: F (1.97,12.58) = 5.90, p = 0.016), no main effect of order was detected (Fig. 3A: two-way REML ANOVA no main effect of order: F (1, 7) = 4.78, p = 0.065) and post hoc tests showed there was only a difference in consumption at one price point (price = 0.1; Sidak’s post-test, p < 0.05), whereas all other points did not differ (p = 0.3–0.9). Similarly, when active responses were examined, there was a significant effect of order (Fig. 3B; two-way REML ANOVA main effect of price order: F (1, 7) = 9.46, p = 0.018), but again there was only a difference at one price point (price = 1; Sidak’s post-test, p < 0.05). This was not the same price point that differed for consumption. Thus, we concluded that there were no robust differences across the procedures and that both are able to similarly capture changes in consumption and work as a function of price.

Fig. 3. Refinement of the within session demand procedure; effects of price order and inter-price interval (IPI).

(A) Average sucrose consumption when prices were presented pseudorandomly (closed circles) or in ascending order (open circles). (B) Average number of active responses when prices were presented pseudorandomly (closed circles) or in ascending order (open circles). (C) Average sucrose consumed when prices were tested 25 min (closed squares) or 10 min (open squares) apart. (D) Average number of active responses when prices were tested 25 min (closed squares) or 10 min (open squares) apart. Main effect of IPI *p < 0.01.

We next determined how shortening the time between each price affected behavior. When the time between each price was shortened to 10 min, consumption was reduced compared to the longer IPI (Fig. 3C; two-way REML ANOVA main effect of IPI: F (1, 8) = 12.46, p = 0.0077). In addition, although the shape of the demand curve did not differ statistically between groups (two-way REML ANOVA no price × IPI interaction: F (2.12, 13.2) = 1.46, p = 0.27), for the 10 min IPI there is no point at which the demand curve was flat, and therefore it would not possible to assess Q₀ or identify a clear price where consumption declines (as it is declining steadily across all prices in the 10 min IPI group). For the 25 min group consumption was stable across prices 2–4 (two-way RM ANOVA no main effect of price 2–4: F (0.28, 1.55) = 2.12, p = 0.20), thus allowing for a better estimation of the demand metrics described above. In addition, overall work was reduced and the peak work was shifted when the shorter IPI was used (Fig. 3D; two-way REML ANOVA price × IPI interaction: F (2.21, 13.70) = 5.36, p = 0.017; main effect of IPI: F (1, 8) = 9.03, p = 0.017). This is likely due to satiety effects that are inherent to studying food reinforcers. Thus overall, the 25 min IPI appears to be better suited to assessing changes in consumption as a function of price (i.e., demand). Of course, there are likely experiments where the procedure parameters would need to be adjusted to best suit the experimental conditions (e.g., if one were testing younger animals or using mice, perhaps a shorter IPI could be used).

3.1.3. Varying hunger and sucrose concentration

Based on the data above, prices were presented in ascending order with a 25-min IPI for the remaining experiments. We next determined whether within session demand for sucrose was sensitive to changes in hunger state and sucrose concentration. Fig. 4A–C shows consumption of 2 M sucrose and corresponding work in rats tested hungry or after selective satiety induced by free access to 2 M sucrose immediately prior to testing. During this 1 hr pre-feeding period, rats consumed an average of 3.017 g of sucrose (±0.528 SEM; data not shown). As expected, selective satiety significantly reduced consumption and work during testing (Fig. 4A; two-way REML ANOVA main effect of testing condition: F (1, 8) = 12.80, p = 0.0072; Fig. 4B; two-way REML ANOVA main effect of testing condition: F (1, 8) = 41.32, p = 0.0002). This was also associated with an overall flattening of the demand curve. Consistent with these findings, total sucrose consumption was also reduced in the sated vs. hungry condition (Fig. 4C; paired two-tailed t-test: t (7) = 6.481, p = 0.0003).

Fig. 4. Assessment of face validity of the within session sucrose demand procedure.

(A) Average sucrose consumed as a function of price (i.e., demand) when rats were tested hungry (closed circles) or selectively sated on 2 M sucrose (open circles). (B) Average number of active responses when rats were tested hungry (closed circles) or selectively sated on 2 M sucrose (open circles). (C) Total sucrose consumed during testing conducted in hungry or sated conditions. (D) Average consumption of 2 M (closed squares) or 1 M (open squares) sucrose as a function of increasing price and preferred level of consumption (Q_o)(E) Average active responses for access to 2 M (closed squares) or 1 M (open squares) sucrose as a function of increasing price. (F) Total sucrose consumed at each concentration tested. Main effect of testing condition **p < 0.01, ***p < 0.001; paired t-test ^###p < 0.001.

To further test the face validity of the within-session behavioral economics procedure, demand for 1 M or 2 M sucrose were assessed in separate test sessions (Fig. 4D–F). The overall shape of the demand curve (Fig. 4D) was not strongly affected by the concentration of sucrose available. However, the demand curve was shifted down and to the right when rats were tested using 1 M vs. 2 M sucrose (Fig. 4D; two-way REML ANOVA main effect of sucrose concentration: F (1, 8) = 7.627, p = 0.0246). Specifically, the preferred level of consumption (Q_o) was lower for 1 M than 2 M, but somewhat counter-intuitively consumption remained elevated out to a slightly higher price for 1 M than 2 M sucrose (Avg Pmax ± SEM: 2 M 0.3716 ± 0.053, 1 M 0.7445 ± 0.153; paired two-tailed t-test: t (4) = 2.420, p = 0.072). This is also reflected in the work curve, where responding was greater at higher prices for 1 M than for 2 M sucrose (Fig. 4E; two-way REML ANOVA main effect of sucrose concentration: F (1, 8) = 5.807, p = 0.0425). Notably however, Rmax, the maximum work performed, was the same across sucrose concentrations (Avg Rmax ± SEM: 2 M 62.12 ± 6.174, 1 M 55.87 ± 8.609; paired two-tailed t-test: t (4) = 1.450, p = 0.2207). Overall, demand for 1 M sucrose was slightly greater than for 2 M sucrose, however 1 M sucrose did not support increased responding and was associated with a lower preferred level of consumption than 2 M sucrose. Increased demand for 1 M sucrose was also reflected by a slight increase in total sucrose intake during 1 M vs. 2 M testing (Fig. 4F; paired two-tailed t-test: t (7) = 2.027, p = 0.0823). That rats continue to respond at higher prices for a lower concentration of sucrose is consistent with the idea that rats are working to defend their preferred level of consumption, regardless of sweetness per se.

The data above establish a basic within session procedure to assess demand for sucrose. However, given the somewhat counter-intuitive pattern of behavior observed in response to 1 vs. 2 M sucrose (Fig. 4D–F), we conducted a within-subjects study in a new cohort of rats to examine behavior across 4 different sucrose concentrations (0.25, 0.5, 1, and 2 M). To improve sensitivity, we also increased the number of price points examined (from 6 to 10; see Table 2).

Fig. 5 shows instrumental responding (A) and sucrose consumption (B) during initial acquisition using 1 M sucrose. Rats showed good discrimination between the active and inactive ports (Fig. 5A; two-way RM ANOVA main effect of port: F (1, 11) = 475.5, p < 0.0001) and increased their responding as access time per bottle presentation was reduced (Fig. 5A: two-way RM ANOVA main effect of training phase: F (3.624, 39.86) = 26.98, p < 0.0001). This resulted in relatively stable sucrose consumption across training phases (Fig. 5B).

Fig. 5. Instrumental training.

(A) Average number of nose pokes during each phase of instrumental training. (B) Average total sucrose consumed during each phase of instrumental training.

We next conducted within session demand testing using 4 different sucrose concentrations. Inactive responses remained low (data not shown; Avg ± SEM; 0.25 M: 0.3292 ± 0.09159; 0.5M: 0.2875 ± 0.1119; 1 M: 0.2333 ± 0.07381; 2 M: 0.3750 ± 0.1154) and were unaffected by price or concentration (two-way RM ANOVA no price × concentration interaction: F (1.843, 20.27) = 1.961, p = 0.1688; no main effect of price: F (1.244, 13.68) = 1.916, p = 0.1894; no main effect of concentration: F (1.198, 13.18) = 0.3940, p = 0.5780). Fig. 6A–E shows demand, work, and corresponding parameters for 0.25, 0.5, 1 and 2 M sucrose. Consistent with results above, Q_o increased with increasing sucrose concentration while consumption remained stable out to higher prices for lower sucrose concentrations (Fig. 6A; two-way RM ANOVA price × concentration interaction: F (5.950, 65.45) = 21.86, p < 0.0001; main effect of price: F (2.551, 28.06) = 98.28, p < 0.0001; main effect of concentration: F (2.016, 22.18) = 66.22, p < 0.0001). Indeed, while the higher prices tested for 0.5, 1, and 2 M sucrose resulted in a fairly sharp reduction in consumption, the consumption curve for 0.25 M declined more modestly at the highest prices tested. This suggests that additional price points should be included to gain a more sensitive measure.

Fig. 6. Comparison of demand and free consumption across sucrose concentrations.

(A) Average demand across 4 sucrose concentrations. (B) Average number of active responses as a function of price across 4 sucrose concentrations. (C) Total sucrose consumed during demand testing at each concentration. (D) Pmax for each concentration of sucrose tested; the inset shows example consumption and active response data (circles) with the corresponding fit lines for one rat at 0.25 M. (E) Rmax for each concentration of sucrose tested. (F) Total sucrose consumed during 5-min of free access. Sidak’s post-test:*p < 0.05; **p < 0.01, 1 M vs. 2 M ***p < 0.001.

Overall, consumption was maintained across higher prices for lower concentrations of sucrose (0.25 M and 0.5 M) compared to higher concentrations (1 M or 2 M). In addition, work curves were shifted to the right for lower vs. higher concentrations of sucrose, with the greatest number of responses occurring for 0.25 M sucrose (Fig. 6B; two-way RM ANOVA price × concentration interaction: F (4.583, 53.38) = 24.32, p < 0.0001; main effect of price: F (2.148, 23.63) = 13.62, p < 0.0001; main effect of concentration: F (2.148, 23.63) = 13.62, p < 0.0001). Despite showing sustained consumption for lower concentrations of sucrose, total sucrose consumption was greatest for high sucrose concentrations (Fig. 6C; one-way RM ANOVA: F (2.016, 22.18) = 66.22, p < 0.0001), with rats showing greater total sucrose consumption at 1 and 2 M sucrose compared with both 0.5 and 0.25 M sucrose. This is largely due to the high levels of intake at low prices for 1 and 2 M sucrose.

We also fit the consumption data and extracted measures of Pmax and Rmax for each rat at each concentration. For 0.25 M sucrose, the majority of rats (8/12) continually increased their responding in an attempt to defend their preferred level of consumption. Thus, rather than consumption being flat and then falling to zero it declined linearly with increasing price, while responding continually ramped up (Fig. 6D inset). Thus, although good fits were obtained from the model (gray line), it is not possible to get an accurate measure of Pmax or Rmax for subjects with this pattern of behavior. Nonetheless, rats are working more and out to higher prices in an attempt to maintain their preferred level of consumption.

Consumption and work for the other concentrations of sucrose tested did follow the expected form, and thus reliable measures of Pmax and Rmax were obtained from each animal. These data are shown in Fig. 6D and E, respectively. Pmax and Rmax declined with increasing sucrose concentration, as expected from the summary demand and work curves (Fig. 6D; one-way REML ANOVA: F (0.6242, 4.786) = 10.27, p = 0.0309; Fig. 6E: one-way REML ANOVA : F (1.143, 9.525) = 7.992, p = 0.0166).

Fig. 6F shows sucrose consumption during a 5 min free access test for each concentration. Sucrose consumption increased with increasing concentration, peaked at 1 M, and then declined significantly at the highest concentration tested (Fig. 6F; one-way RM ANOVA: F (2.021, 22.24) = 25.44, p < 0.0001; Sidak’s post-test 2 M vs. 1 M p < 0.001). Interestingly, intake of 2 M sucrose was much lower during this free consumption test than during demand testing, where they consistently consumed 2–4 times more sucrose at low price than when it was free. Across all other concentrations, consumption in the 5-min free access test was similar to the maximum amount of sucrose consumed in the 5-min price trials during demand testing. This suggests that rats are indeed working to maximize their sucrose consumption in each of the 5-min price periods. Based on these results, 0.25 and 1 M sucrose were used to examine effects of home-cage sucrose consumption on demand in subsequent studies. Additionally, the price range tested for 0.25 M was expanded to better capture reductions in responding at high price and thus improve our ability to assess Rmax and Pmax (Table 3).

Lastly, rats across all experiments were maintained on a modified feeding schedule in which food was only available for part of their dark cycle each day (see methods). While rats had ad lib access to food during this time, it’s possible that this manipulation could increase circulating levels of Cort (Yau & Potenza, 2013), a stress hormone that can also impact feeding behavior (Kumar & Leibowitz, 1988). However, when we compared circulating Cort levels when rats were maintained on the modified feeding schedule vs. after ad lib feeding for one week, no differences were found (mean ng/mL ± SEM: modified feed 30.78 ± 4.25, ad lib fed 31.60 ± 3.27). Thus, the modified feeding schedule used here does not produce a peripheral stress response that could alter behavior in the demand task.

3.2. Study 2: Effects of intermittent or continuous home cage sucrose on demand for sucrose

3.2.1. Baseline demand and work for 0.25 and 1 M sucrose

Having established a reliable within session demand procedure, we then used this approach to determine how consumption of sucrose in the home cage affects demand and work for 1 M and 0.25 M sucrose, and the degree to which effects are long-lasting. As above, rats readily learned to respond for access to 1 M sucrose during initial acquisition (Fig. 7A and B), showing strong discrimination between active and inactive ports, and increasing their responding as access time decreased in order to maintain stable levels of sucrose consumption (Fig. 7A; two-way RM ANOVA main effect of port: F (1, 40) = 441.0, p < 0.0001; main effect of session: F (9, 360) = 71.02, p < 0.0001; session × port interaction: F (9, 318) = 82.55, p < 0.0001).

Fig. 7. Acquisition and baseline demand for 1 M and 0.25 M sucrose.

(A) Average number of nose pokes during each phase of instrumental training. (B) Average total sucrose consumed during each phase of instrumental training. (C) Baseline demand for 1 M and 0.25 M sucrose. Inset shows total sucrose consumed at each concentration. (D) Baseline number of active responses as a function of price for 1 M and 0.25 M sucrose. (E) Pmax for 1 M and 0.25 M sucrose: right panels show representative data and fits for a single rat. (F) Rmax for 1 M and 0.25 M sucrose: right panels show representative data and fits for a single rat. Paired two-tailed t-test, ****p < 0.0001.

Fig. 7C and D shows consumption and responses, respectively, during baseline within session demand testing of 1 M and 0.25 M sucrose. Consistent with initial studies above, the demand curve for 0.25 M was shifted down and to the right compared to demand for 1 M sucrose (Fig. 7C). Thus, the preferred level of consumption was lower for 0.25 M vs. 1 M, but rats maintained their consumption of 0.25 M out to higher price points compared to 1 M. As above, total sucrose intake was lower for 0.25 M vs. 1 M (Fig. 7C inset; paired two-tailed t-test: t (40) = 22.57, p < 0.0001). The work curve for 0.25 M was again shifted up and to the right compared to 1 M sucrose (Fig. 7D), consistent with maintained consumption at high prices. Furthermore, testing additional prices for 0.25 M sucrose revealed the descending limb of the demand curve and enabled accurate measurement of Pmax and Rmax at this concentration (Fig. 7E and F). The apparent differences in the sucrose demand and work curves resulted in a greater Pmax (Fig. 7E; paired two-tailed t-test, t (38) = 5.640, p < 0.0001) and Rmax (Fig. 7F: paired two-tailed t-test, t (40) = 4.370, p < 0.0001) for 0.25 M vs. 1 M sucrose. Together these data show that while the preferred level of consumption for 1 M was greater than for 0.25 M, rats were more motivated to maintain their consumption of 0.25 M sucrose and worked harder for access to it than 1 M.

3.2.2. Home cage sucrose exposure

Following baseline demand testing, rats were counterbalanced into one of three groups: water-maintained controls (water), intermittent sucrose, or continuous sucrose access groups (1 M). Fig. 8 summarizes food and liquid intake and changes in body weight across the 4-week access period. The continuous access group consumed significantly less chow across the exposure period compared to the water group (Fig. 8A; two-way REML ANOVA main effect of group: F (1, 25) = 358.1, p < 0.0001), consuming on an average 10–15 g less of chow per day than water maintained controls (Fig. 8B; unpaired two-tailed t-test: t (25) = 19.27, p < 0.0001). While food intake was stable across days in water and continuous access groups, rats in the intermittent access group showed a saw tooth pattern, with chow intake dropping dramatically on days when liquid sucrose was available and increasing again when it was not (Fig. 8C; arrows indicate days when sucrose was present). These data are summarized in Fig. 8D where chow intake is significantly reduced when sucrose is present (paired two-tailed t-test: t (12) = 9.925, p < 0.0001).

Fig. 8. Food and sucrose intake and weight gain during home cage sucrose exposure.

(A) Daily chow intake for water and continuous access groups. (B) Average chow consumed per day across 4-weeks in water and continuous access groups. (C) Daily chow intake prior to, during, and following intermittent access session 1, 3, 6, 9, and 10 as indicated by arrows. (D) Average chow consumed when sucrose was (on) or was not (off) present in the home cage for the intermittent access group. (E) Amount of 1 M sucrose consumed by intermittent and continuous access groups during the first and last access session. (F) Total sucrose consumed by intermittent and continuous access groups across the 4-week home cage exposure. (G) Absolute body weight (g) for all groups. (H) Total weight gain across the 4 weeks of access in each group. Unpaired t-tests, ****p < 0.0001; Sidak’s post-test water and intermittent vs. continuous #p < 0.05.

Fig. 8E shows sucrose intake during the first and last exposure for continuous and intermittent access groups. Rats in the intermittent access group escalated their sucrose intake, while rats in the continuous access group did not (Fig. 8E; two-way REML ANOVA group × day interaction: F (1, 24) = 45.35, p < 0.0001, Sidak’s post-test first vs. last intermittent: p < 0.01, continuous p > 0.5). Rats in the intermittent group had a total of 10 days of sucrose exposure across the 4-week period. Thus, it’s not surprising that they consumed less total sucrose than rats in the continuous group (Fig. 8F; unpaired two-tailed t-test: t (26) = 13.65, p < 0.00001). This resulted in significantly greater body weight (Fig. 8G; two-way RM ANOVA group × time interaction: F (18, 342) = 7.812, p < 0.0001) and weight gain in the continuous access compared to the water and intermittent access groups (Fig. 8H; one-way ANOVA: F (2, 38) = 9.702, p = 0.0004, Sidak’s post-test water vs continuous p < 0.05; continuous vs intermittent p < 0.001).

3.2.3. Post-home cage and reversal sucrose demand for 0.25 and 1 M sucrose

Fig. 9 shows demand for 0.25 M sucrose after the 4-week exposure period. The preferred level of consumption was similar across all three groups, but consumption was maintained out to higher prices for all three groups compared to baseline (Fig. 9A; two-way RM ANOVA price × group interaction: F (27, 702) = 2.221, p = 0.0004). Although this was not reflected by significant shifts in Pmax (data not shown), active responses were increased in all three groups compared to baseline (Fig. 9B; two-way RM ANOVA significant price × group interaction: F (27, 702) = 3.328, p < 0.0001). This resulted in slight increases in the total amount of sucrose consumed during post vs. baseline testing in both groups exposed to sucrose (Fig. 9C; one-way ANOVA: F (3, 78) = 5.247, p < 0.01; Sidak post-test intermittent p < 0.01 and continuous p < 0.05 vs. baseline). Finally, the change in Rmax from baseline was also increased in all three groups (Fig. 9D gray bars; Wilcoxon signed-rank test: water mean rank = 82.41, confidence interval (CI) = 97.75, p < 0.05; intermittent mean rank = 84.77, CI = 98.71, p < 0.001; continuous mean rank = 80.88, CI = 98.71, p < 0.01). Thus, it appears that time-off the task (or relative novelty of 0.25 M sucrose) and not home cage sucrose consumption per se resulted in elevations in work and an increase in demand for 0.25 M sucrose.

Fig. 9. Behavior during within session demand for 0.25 M sucrose at baseline and after home cage sucrose exposure.

(A) Consumption of 0.25 M sucrose as a function of increasing price in water, continuous sucrose and intermittent sucrose groups. (B) Active responses as a function of increasing price in all three groups. (C) Total sucrose intake during post-exposure demand testing. (D) Rmax expressed as a change from baseline following 4 weeks of home cage sucrose exposure (post) and after an additional 2 weeks without sucrose exposure (recovery). Sidak’s post-tests *p < 0.05; **p < 0.01 vs. baseline.

All three groups were tested again two weeks later with no additional home cage sucrose exposure. Consistent with potential novelty effects, the increase in Rmax for 0.25 M sucrose was maintained in water and intermittent groups but returned to towards baseline in the continuous access group (Fig. 9D open bars; Wilcoxon signed-rank test: water mean rank = 75.65, CI = 98.83, p < 0.01; intermittent mean rank = 52.55, CI = 98.71, p < 0.01; continuous mean rank = 6.093, CI = 98.71, p = 0.27).

When these same rats were tested using 1 M sucrose, a very different pattern emerged (Fig. 10). While the preferred level of consumption was similar across all groups at the end of the 4 week period, demand and work declined at lower prices in continuous access vs. water and intermittent access groups (Fig. 10A: two-way RM ANOVA significant main effect of group: F (3, 78) = 7.751, p = 0.0001; Fig. 10B: two-way RM ANOVA significant main effect of group: F (3, 78) = 3.78, p < 0.05). Despite clear diet-induced effects on consumption, Pmax did not differ between groups (one-way ANOVA: F (2, 34) = 1.72, p = 0.19; data not shown). Similar to demand, the work curve was slightly shifted right-ward in the water group, and nearly flattened in the continuous access group (Fig. 10B). This was also reflected by an increase in Rmax from each rat’s baseline value in the water group, and a reduction in Rmax from baseline in the continuous access group (Fig. 10C closed bars; Wilcoxon signed-rank test: water mean rank = 42.65, CI = 96.14, p < 0.05; continuous mean rank = −44.36, CI = 98.83, p < 0.001). In contrast, behavior in the intermittent group was comparable to that seen during baseline (Fig. 10C closed bars; Wilcoxon signed-rank test: intermittent mean rank = 1.314, CI = 98.71, p = 0.76). Thus, despite enhancements in demand and work for 0.25 M compared to baseline and comparable to the water group, motivation for 1 M sucrose was reduced in rats given continuous sucrose access, but unaffected by intermittent access. When rats were tested again 2-weeks later, enhancements in active responses in the water group were still present, whereas some rats in the continuous access group showed a return to baseline (Fig. 10C open bars; Wilcoxon signed-rank test: water mean rank = 67.99, CI = 97.75, p < 0.05; continuous mean rank = −18.04, CI = 98.71, p = 0.43; Fig. 10D, two-way RM ANOVA price × group interaction: F (27, 702) = 1.697, p < 0.05).

Fig. 10. Behavior during within session demand for 1 M sucrose at baseline and after home cage sucrose exposure.

(A) Consumption of 1 M sucrose as a function of increasing price in controls, continuous sucrose and intermittent sucrose groups. (B) Active responses as a function of increasing price in all three groups. Inset shows comparisons of active responses between individual groups and baseline. (C) Rmax expressed as a change from baseline following 4 weeks of home cage sucrose exposure (post) and after an additional 2 weeks without sucrose exposure (recovery). (D) Active responses as a function of increasing price in all three groups after the 2-week recovery period. Inset shows comparisons of active responses between individual groups and baseline.

4. Discussion

Here we report a new within-session demand procedure to assess motivation for liquid sucrose. We first determined the factors that result in reproducible demand and work curves. Varying the order of price presentation (ascending vs. pseudo-random) did not strongly affect behavior (Fig. 3A and B). However, reducing the duration of the inter-price-interval (IPI) from 25 min to 10 min blunted both the demand and the work curves (Fig. 3C and D). This is likely due to satiety effects when different prices are tested too closely in time. Longer durations were not examined, however, using a 25 min IPI allowed for the assessment of up to 10 prices (5 min at each price) within one day while avoiding satiety effects. Regarding face validity, both demand and work were sensitive to changes in hunger state and to sucrose concentration. Specifically, pre-feeding with 2 M sucrose blunted demand and work for 2 M sucrose compared to when rats were tested hungry (Fig. 4A–C). In addition, the preferred level of consumption (Q₀) for 1 M sucrose was lower compared to Q₀ for 2 M sucrose (Fig. 4D). A similar pattern was observed when additional sucrose concentrations were examined in separate rats (Fig. 6A), with Q₀ increasing as a function of increasing sucrose concentration. These data are largely consistent with free sucrose consumption where intake within a 5-min period increases across 0.25–1 M sucrose (Fig. 6F). Interestingly, although the Q₀ for 2 M and 1 M sucrose were similar, free consumption of 2 M sucrose was lower than 1 M sucrose (Fig. 6F). Given these data, we choose to use 1 M and 0.25 M sucrose concentrations to examine the effects of home-cage sucrose ingestion on motivation (discussed further below).

One final note regarding this within session demand procedure is that across several independent cohorts, rats maintained their preferred level of consumption out to higher prices for lower concentrations of sucrose. Thus, lower sucrose concentrations supported higher Pmax and Rmax values (e.g., Fig. 7E and F). It may seem counter-intuitive that less sweet, and less calorically dense sucrose solutions produced stronger motivation. However, solutions with relatively low sucrose concentrations are metabolized into glucose and absorbed into the blood stream more quickly than relatively high concentrated solutions (Davidson & Leese, 1977). Thus, it’s possible that this difference in metabolism and absorption account for the maintenance of responding for 0.25 M vs. 1 M sucrose. Alternatively, this difference could result from rats from working toward a caloric goal. The contribution of calories vs. sweetness to maintained responding and preferred levels of consumption (Q₀) should be examined in future studies. In addition, other factors that we did not control for could contribute to differences in consumption and motivation (e.g., potential differences in viscosity between solutions). Nonetheless, the parameters used here allowed for reliable, within session evaluation of motivation for liquid sucrose that can be employed in a range of studies.

We next examined motivation for liquid sucrose (0.25 M and 1 M) before (baseline) and after free liquid sucrose consumption in the home cage (1 M continuous or 1 M intermittent). During the 4-week home cage access period, intermittent access rats escalated their intake from the first to the last access session, while continuous access rats maintained stable intake (Fig. 8E), consistent with previous work (Eikelboom & Hewitt, 2016). During demand testing, prior continuous home-cage liquid sucrose consumption selectively blunted motivation for this same concentration of sucrose (1 M), without reducing motivation for a lower concentration of sucrose (0.25 M) compared to baseline. In fact, motivation for 0.25 M sucrose was enhanced in control and experimental groups, suggesting that “time-off” from the task and/or the relative novelty of this solution was sufficient to support stronger motivation during testing. Consistent with this interpretation, motivation for 1 M sucrose was increased in water-maintained controls and remained unchanged in the intermittent access group relative to the pre-exposure baseline. Thus, for controls, both concentrations of sucrose are relatively “novel” as they only have access to these solutions during initial training and demand testing sessions, whereas for the intermittent and continuous access groups 0.25 M is relatively novel, while 1 M is not.

Reductions in motivation for 1 M sucrose following its continuous home-cage consumption are not likely due to effects of diet manipulation on satiety or taste (Sung et al., 2022), as alterations in these processes would also be expected to reduce motivation for 0.25 M sucrose, not enhance it as was found here. Thus, overall effects of home cage sucrose consumption found here do not appear to be due to fundamental shifts in sucrose motivation, but are more likely related to habituation to 1 M sucrose from continual consumption. This interpretation is also consistent with the observed recovery back to baseline after removal of sucrose from the home cage.

To our knowledge, this is the only study to examine effects of home-cage liquid sucrose consumption on motivation for this same reinforcer. Other studies have examined effects of high-fat diet on motivation for a variety of foods including liquid sucrose (e.g., (Matikainen-Ankney et al., 2022; Orsini et al., 2022; Robertson & Rasmussen, 2017). Results from these studies vary, with reductions, no change and enhancements in motivation reported. These inconsistencies may be due to the complexity of parameters involved (e.g., duration of diet exposure, type of food reinforcer used, hunger state, etc.), and likely reflect true biological variance. However, one interesting feature is that alterations in food-motivation (decreases or increases) do not appear to require obesity per se, but rather are due to consumption of calorie-dense foods (e.g., Orsini et al., 2022 and results here). This is consistent with the idea that alterations in brain motivational systems are causal to obesity, rather than consequential.

A few studies have examined effects of home-cage liquid sucrose consumption on food motivation, though sucrose exposure, reinforcers evaluated, and procedures vary, as do the effects on motivation. For example, Suárez-Ortiz et al. (2018)) found that in female rats 28 days of continuous home cage access to 10% sucrose enhanced break point for chocolate flavored sugar pellets, whereas Vendruscolo et al. (2010) found that continuous home cage access to 5% sucrose in male rats reduced fixed ratio responding for saccharin. In addition, limited daily sucrose exposures that resulted in binge-eating did not alter food-motivation (Suarez-Ortiz et al., 2018), similar to null effects of intermittent access found here. However, it should be noted that common obesity is distinct neurobiologically, behaviorally, and psychologically from binge-eating disorder which is not tightly linked to obesity (Vainik et al., 2019; Wilson et al., 2007).

Overall, additional work is needed to better understand how different attributes of food (nutrition, calories, taste) interact with internal states and the dietary environment to alter feeding behavior and motivation. However, the use and further development of novel approaches to examining motivation like those presented here and in other recent work (e.g., Matikainen-Ankney et al., 2022) will help refine our understanding of the complex effects of dietary environment and obesity on food motivation. Of course, the specific procedure used (e.g., number of prices tested, inter-price interval), will likely need to be optimized to experimental conditions of each study (e.g., if using adolescent animals, females, mice, etc.). Furthermore, while the approach described here provides several different metrics of motivation that may increase sensitivity to detect motivational shifts, it does not yet incorporate more complex aspects of feeding behavior (e.g., choice between outcomes, decision-making or responding in complex food environments). Incorporating these aspects in future may provide further insights.

Abbreviations

Q_o

Preferred level of consumption

Rmax

Maximum responses

Pmax

Maximum price “paid” to maintain Q_o

IPI

Inter-price interval

Data availability

Data will be made available on request.