1. Introduction

Pain is a global health problem that affects at least 30% of the population in the United States.22 Epidemiological studies have shown that there is a relation between pain and alcohol use disorder (AUD),7 and this relation can be bidirectional such that extended and excessive alcohol consumption has been shown to provoke hyperalgesia due to peripheral neuropathy.24 This alcohol-induced pain condition has been widely studied in both humans and rodents.8,24 Similarly, pain also seems to have an effect on alcohol drinking. In fact, clinical studies have shown that the presence of persistent pain is directly correlated with an increase in alcohol consumption and an elevated risk of developing an AUD.32,33 Moreover, in patients with AUD, higher levels of pain have been correlated with higher levels of relapse.15 Evidence also indicates that greater levels of pain intensity and unpleasantness are associated with an increase in alcohol consumption, rates of hazardous drinking, and delay to seek for treatment.3,19

However, despite the epidemiological evidence, the effect that pain has on AUD has not been deeply studied in the preclinical setting. In addition, among the very few studies that have explored this, the results seem to be inconsistent. Some authors have shown that the presence of pain may increase alcohol intake levels in male mice.4,10,34 By contrast, Lorente et al., reported that only female rats under the pain condition showed a relapse-like behavior when evaluating the effect of pain developed during abstinence, whereas Bilbao et al., showed that male mice relapse after an abstinence period but when the pain condition is induced before the animals have any access to alcohol.2,21 Interestingly, most of these studies examined the potential effect of pain before the alcohol exposure. Based on the clinical data described above, however, pain may be a risk factor to develop AUD or to increase alcohol consumption in subjects that are not naïve to alcohol. Therefore, it is also important to further explore the effect of pain in animals with a history of alcohol exposure.

In addition, both clinical and preclinical studies suggest that the interaction between pain and alcohol use is gender or sex specific.3,7 There is evidence of sex differences in pain sensitivity and that women are more likely to develop chronic pain.20,29,31 In the case of alcohol, men have traditionally reported to drink more alcohol than women, although these differences are narrowing during the past years.11 Moreover, women with AUD show a higher prevalence of medical and psychiatric comorbidities, such as depression and anxiety. Interestingly, female rodents usually exhibit higher levels of alcohol when normalized by body weight.18,21,27,28 Altogether, these indicate the need to explore potential sex differences when studying the interaction between pain and AUD.

In this study, we aim to evaluate how inflammatory pain affects levels of alcohol drinking in male and female rats with a history of alcohol consumption, using an intermittent access 2-bottle choice paradigm. Our results show that pain does not affect alcohol intake in female rats. However, in males, pain blunts the decrease on alcohol intake when increasing the alcohol concentrations.

2. Materials and methods 2.1. Animals

All procedures were approved by Washington University and the NIH Animal Care and Use Committee in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals.

Sixty adult male and female Long Evans wild type were used (7–8 week and 260–300 g [male rats] and 170–200 g [female rats] at the beginning of the alcohol intermittent access [IA] model). Rats were individually housed on a 12/12-hour dark/light cycle (lights on at 7:00) and acclimated to the animal facility holding rooms for at least 7 days before any manipulation. The temperature for the holding rooms of all animals ranged from 21 to 24°C, whereas the humidity was between 30% and 70%. Food and water were available ad libitum throughout the experimental period.

2.2. Alcohol intermittent access model and pain induction

In this experiment, animals followed the classical alcohol IA model as previously described6,21 (Fig. 1A). In this study, rats had free access to 20% alcohol along with water 3 times a week for 24 hours, followed by 24 or 48 hours of nonaccess to alcohol. Fresh alcohol (ethanol) solution and water were always used. Alcohol bottles were introduced at 10:00 am every Monday, Wednesday, and Friday and removed 24 hours later at the same time. The bottles were weighed before and after their introduction to measure total fluid intake and alcohol intake and preference for the alcohol bottle. Rats were weighed every day before the introduction of the bottles to calculate alcohol consumption in g/kg/d. Furthermore, to ensure that the alcohol consumption was not influenced by a place preference, the order of the water and alcohol bottles was alternated each time alcohol was introduced.

Figure 1.:

(A) Schematic of the experimental design. (B) Mean ± SEM of total volume of liquid (mL) for the 5 weeks of acquisition for male (green) and female (orange) rats (2-way ANOVA, *P < 0.05 Bonferroni post hoc). (C) Average mean ± SEM of total volume of liquid (mL) for basal week before CFA/saline injection for male (green) and for female (orange) rats (**P < 0.01, t test). (D) Mean ± SEM of alcohol intake (g/kg/d) for the 5 weeks of acquisition for male (green) and female (orange) rats (2-way ANOVA, #P < 0.05 Tukey's post hoc). (E) Average mean ± SEM of alcohol intake (g/kg/d) for basal week before CFA/saline injection for male (green) and for female (orange) rats (*P < 0.05, t test). (F) Mean ± SEM of alcohol preference (%) for the 5 weeks of acquisition for male (green) and female (orange) rats (2-way ANOVA, #P < 0.05 Tukey post hoc). (G) Average mean ± SEM of alcohol preference (g%) for basal week before CFA/saline injection for male (green) and for female (orange) rats. ANOVA, analysis of variance; CFA, complete Freund adjuvant.

Once they reached a stable intake level (week 6), rats received 0.15 mL (for males) or 0.12 mL (for females) subcutaneous injection of the complete Freund adjuvant (CFA) or sterile saline in the plantar surface of the hind paw, without altering the intermittent access schedule. Rats then underwent 3 more weeks after the same IA schedule. Finally, during the fourth, fifth, and sixth week after CFA or saline injections, the alcohol concentration was increased up to 30%, 40%, and 50%, respectively.

2.3. Mechanical nociception assessment

To assess baseline nociception (ie, mechanical hyperalgesia) induced by CFA injections, paw withdrawal thresholds (PWTs) were obtained using an electronic Von Frey Anesthesiometer (IITC Life Science, California, USA). Animals were placed in plexiglass chambers on top of a galvanized steel mesh shelf to permit access to the rats' paws from underneath. The anesthesiometer was used to provoke a flexion reflex followed by a flinch response, and the mechanical threshold pressure in grams (PWT) was recorded. Rats were habituated to the test chambers and von Frey procedure for at least 1 hour 1 week before conducting the baseline test. On assessment days (once a week, starting the week before injection), rats were placed in the plexiglass chambers at least 2 hours after removing the alcohol bottles to mitigate potential lasting analgesic effects of alcohol. Rats were acclimated to chambers for 20 minutes before undergoing the procedure. Once acclimated, measurements of mechanical sensitivity were obtained in triplicates for each paw at 5-minute intervals, alternating between the injected and noninjected paw. Paw withdrawal thresholds were determined by averaging all 3 replicates per each testing session.

2.4. Experimental design and statistical analysis

For this experimental design, 60 rats (30 male and 30 female) were used. The experiment was replicated 3 times, including each treatment condition to confirm the reproducibility of the data.

According to the IA protocol, male and female rats were assigned to one of the treatments (saline or CFA) in a counterbalanced fashion based on the baseline alcohol intake (g/kg).

All the results are expressed in mean ± SEM. After assessing the normality of sample data using D'Agostino and Pearson tests and Shapiro–Wilk tests, statistical significance was taken as *P < 0.05, **P < 0.01, ***P < 0.001 731, and ****P < 0.0001, as determined by 2-way analysis of variance (ANOVA) for repeated measures, followed by the Tukey or Bonferroni post hoc test for intrasubjects or between-subjects comparisons, respectively, or unpaired t test. Statistical analyses were performed using GraphPad Prism 9.1.0. Data collection and analysis were performed blinded to the conditions of the experiments.

3. Results 3.1. Basal alcohol intake is higher in female rats

Alcohol and water drinking behaviors were evaluated on every consumption day for the 5 weeks before saline or CFA injection. Basal levels were considered as the average of the measurements from the past 3 consumption days.

For the total volume of liquid consumed, calculated as the total volume of water and total volume of alcohol, our results show differences between male and female rats. In the individual consumption sessions (Fig. 1B), the ANOVA for repeated measurements detected a main effect of sex (F [4.343, 251.9] = 3.404, P = 0.0005) and of time (F [4.343, 251.9] = 3.404, P = 0.0080), but there was not a significant interaction between sex and time (F [14, 812] = 0.7400). Next, the Bonferroni post hoc analysis for multiple comparisons revealed a significant difference of total volume of liquid between males and females in sessions 6 to 10. Moreover, the t test showed significant differences in the average of basal week of total volume of liquid between males and females (P = 0.0056) (Fig. 1C).

When analyzing alcohol intake (calculated as g/kg/d) for the individual consumption sessions (Fig. 1D), the ANOVA for repeated measurements detected a main effect of time (F [14, 812] = 4.393, P < 0.0001) but not of sex (F [1, 58] = 2.239, P = 0.1400) or a significant interaction of sex and time (F [14, 812] = 0.9476, P = 0.5065). Interestingly, in the average of basal week, female rats showed significantly higher levels of alcohol intake compared with males (Fig. 1E), as it was reported by the t test (P = 0.0304).

Preference for the alcohol solution was also analyzed for the acquisition period, calculated as the percentage of alcohol volume relative to the total volume consumed. For the individual consumption sessions (Fig. 1F), the ANOVA for repeated measurements detected a main effect of time (F [14, 812] = 10.12, P < 0.0001) and a significant interaction of the sex and time (F [14, 812] = 1.757, P = 0.0409), but it did not detect differences of the sex (F [1, 58] = 0.6577, P = 0.4207). When analyzing intrasubject differences, the Tukey post hoc analysis for multiple comparisons revealed a significant difference from the first consumption session in several consumption sessions in males (sessions 3, 6–12, 14 and 15) and in the last consumption session in females (session 15). Finally, no significant differences were found between males and females in preference for the weekly average of basal (Fig. 1G), as reported by the t test (P = 0.6293).

Based on the aforementioned differences on basal week levels of alcohol intake, subsequent male and female data were analyzed separately.

3.2. Inflammatory pain does not alter total intake of 20% alcohol in male or female rats

To explore the effect of inflammatory pain on alcohol intake in our IA procedure, alcohol intake and preference were monitored for 3 weeks after saline or CFA injection and compared with their respective basal levels before the injections (Fig. 2A). Moreover, changes in total volume of liquid consumed were also evaluated to ensure that changes in alcohol intake or preference were not due to an alteration in overall animal drinking behavior.

Figure 2.:

Effect of inflammatory pain on 20% alcohol drinking. Represented are data from basal week and from weeks 1, 2, and 3 after saline (SAL, in black) or CFA (in red) injection for males (full symbols) and females (empty symbols). (A) Schematic of the experimental design. (B) Mean ± SEM of total volume of liquid (mL) for the individual sessions in males. (C) Weekly average mean ± SEM of total volume of liquid (mL) in males (2-way ANOVA, #P < 0.05 Tukey post hoc). (D) Mean ± SEM of total volume of liquid (mL) for the individual sessions in females (2-way ANOVA, ##P < 0.01 Tukey post hoc). (E) Weekly average mean ± SEM of total volume of liquid (mL) in females (2-way ANOVA, ###P < 0.001 Tukey post hoc). (F) Mean ± SEM of alcohol intake (g/kg/d) for the individual sessions in males (2-way ANOVA, #P < 0.05 Tukey post hoc). (G) Weekly average mean ± SEM of alcohol intake (g/kg/d) in males. (H) Mean ± SEM of alcohol intake (g/kg/d) for the individual sessions in females (2-way ANOVA, *P < 0.05 Bonferroni post hoc). (I) Weekly average mean ± SEM of alcohol intake (g/kg/d) in females. (J) Mean ± SEM of alcohol preference (%) for the individual sessions in males. (K) Weekly average mean ± SEM of alcohol preference (%) in males (2-way ANOVA, #P < 0.05 Tukey post hoc). (L) Mean ± SEM of alcohol preference (%) for the individual sessions in females. (M) Weekly average mean ± SEM of alcohol preference (%) in females. ANOVA, analysis of variance; CFA, complete Freund adjuvant.

In males, when examining total volume of liquid consumed, the ANOVA for repeated measurements did not detect differences in treatment (F [1, 28] = 0.9119, P = 0.3478) in the case of individual consumption sessions. However, a main effect of time (F [6.400, 179.2] = 3.030, P = 0.0064) was detected, as well as a significant interaction of treatment and time (F [10, 280] = 2.299, P = 0.0132). When comparing with the first basal session, the Tukey post hoc analysis for multiple comparisons did not reveal any significant difference in saline or CFA treated males (Fig. 2B). When evaluating the weekly average of total volume of liquid consumed in males (Fig. 2C), the ANOVA for repeated measures detected a main effect of time (F [3, 84] = 3.421, P = 0.0209) but did not detect differences in treatment (F [1, 28] = 0.7878, P = 0.3823) or in the interaction between time and treatment (F [3, 84] = 1.661, P = 0.1817). When evaluating differences from basal week, the Tukey post hoc analysis for multiple comparisons revealed a significant difference in week 3 in saline-treated males (P = 0.0287) but not in the CFA group.

Similarly, when comparing the individual sessions of total volume of liquid consumed in females (Fig. 2D), the ANOVA for repeated measurements did not detect differences in treatment (F [1, 28] = 0.9164, P = 0.3466) but a main effect of time (F [5.939, 166.3] = 3.768, P = 0.0016) and an interaction between time and treatment (F [10, 280] = 2.483, P = 0.0073) were observed. Subsequent Bonferroni post hoc analysis for multiple comparisons revealed a significant difference from the first basal session vs session 1 after CFA injection only in CFA-treated females (P = 0.0020) and no intrasubject differences were detected in saline treated females. In the case of the weekly average (Fig. 2E), the ANOVA for repeated measures also detected a main effect of time (F [3, 84] = 4.499, P = 0.0056), but not of treatment (F [1, 28] = 0.6193, P = 0.4379) or in the interaction between time and treatment (F [3, 84] = 2.662, P = 0.0533). In addition, the Tukey post hoc analysis for multiple comparisons revealed a significant difference from basal week (Fig. 2A) in week 2 in saline-treated females (P = 0.0004), but no differences from basal week were detected in CFA-treated females.

When assessing overall consumption of alcohol in males (calculated as g/kg/d) for the individual sessions (Fig. 2F), the ANOVA for repeated measures detected a main effect of time (F [6.531, 182.9] = 4.563, P = 0.0002) but did not detect differences in treatment (F [1, 28] = 0.01580, P = 0.9009) or in the interaction between time and treatment (F [11, 308] = 1.348, P = 0.1972). When comparing with the first basal session (session −3), the Tukey post hoc analysis for multiple comparisons revealed a significant difference from the first basal session (session −3) in session 5 only in CFA-treated males (P = 0.0456), and no intrasubject differences were detected in saline-treated males. When analyzing the weekly (Fig. 2G), the ANOVA for repeated measures did not detect differences in treatment (F [1, 28] = 0.01277, P = 0.9108), time (F [3, 84] = 2.242, P = 0.0893), or in the interaction between treatment and time (F [3, 84] = 1.070, P = 0.3665).

Similarly, in females and for the levels of alcohol intake in individual sessions (Fig. 2H), the ANOVA for repeated measures detected a main effect of time (F [6.086, 170.4] = 4.661, P = 0.0002) but did not detect differences in treatment (F [1, 28] = 0.7946, P = 0.3803) or in the interaction between time and treatment (F [11, 308] = 1.628, P = 0.0898). In this case, the Bonferroni post hoc analysis for multiple comparisons revealed a significant difference between saline-treated and CFA-treated females in session 1 after injection (P = 0.0213) but did not detect intrasubject differences in either saline-treated or CFA-treated rats. Similar results were found for the weekly average of alcohol intake (Fig. 2I). Thus, the ANOVA for repeated measures did not detect differences in treatment (F [1, 28] = 0.8084, P = 0.3763), time (F [3, 84] = 1.878, P = 0.1396), or in the interaction between treatment and time (F [3, 84] = 0.9295, P = 0.4302).

For the preference of the individual consumption sessions in males (Fig. 2J), the ANOVA for repeated measures revealed a main effect of time (F [5.893, 165.0] = 6.022, P < 0.0001), but not of treatment (F [1, 28] = 0.3554, P = 0.5559), or in the interaction between treatment and time (F [11, 308] = 0.6955, P = 0.7428). However, when comparing with the first basal session, the Tukey post hoc analysis for multiple comparisons did not detect significant differences in saline or CFA-treated males. Next, when assessing the average of preference values weekly (Fig. 2K), the ANOVA for repeated measures detected a significant effect of time (F [3, 84] = 5.916, P = 0.0010), but no differences were detected with respect to treatment (F [1, 28] = 0.3605, P = 0.5531) or in the interaction between treatment and time (F [3, 84] = 0.8634, P = 0.4634). Subsequently, Tukey post hoc analysis for multiple comparisons revealed significant differences in week 2 when compared with basal week only in saline-treated males (P = 0.0131), but not in the CFA-treated group.

For females (Fig. 2L), similar results were found for the preference of the individual consumption sessions and the ANOVA for repeated measures did not detect differences in treatment (F [1, 28] = 0.7665, P = 0.3888) or in the interaction between treatment and time (F [11, 308] = 1.173, P = 0.3053), but it detected a main effect of time (F [6.168, 172.7] = 5.670, P < 0.0001). In addition, when comparing the postinjection session with the first basal session, the Tukey post hoc analysis for multiple comparisons did not detect significant differences in saline-treated or CFA-treated females. In the case of the average of preference values weekly (Fig. 2M), the ANOVA for repeated measures did not detect differences in treatment (F [1, 28] = 0.7221, P = 0.4027), time (F [3, 84] = 2.448, P = 0.0693), or in the interaction between treatment and time (F [3, 84] = 1.718, P = 0.1694).

3.3. Inflammatory pain affects alcohol intake in a dose-dependent manner in male rats

Finally, we explored how inflammatory pain affects the intake when rats are exposed to increasing concentrations of alcohol following the above described IA protocol. To this end, as described in Figure 3A, rats underwent 3 additional weeks of our IA procedure in which they were exposed to 30%, 40%, and 50% alcohol, respectively (Fig. 3A). Alcohol intake and preference were monitored throughout this time and compared with the levels on the last week of exposure to 20% alcohol (week 3 post-CFA or saline).

Figure 3.:

Effect of inflammatory pain on alcohol consumption of different alcohol concentrations. Represented are data from the last week of 20% and from weeks of 30%, 40%, and 50% for saline (SAL, in black) and CFA (in red) groups and for males (full symbols) and females (empty symbols). (A) Schematic of the experimental design. (B) Mean ± SEM of total volume of liquid (mL) for the individual sessions in males. (C) Weekly average mean ± SEM of total volume of liquid (mL) in males. (D) Mean ± SEM of total volume of liquid (mL) for the individual sessions in females. (E) Weekly average mean ± SEM of total volume of liquid (mL) in females. (F) Mean ± SEM of alcohol intake (g/kg/d) for the individual sessions in males (2-way ANOVA, #P < 0.05 Tukey post hoc). (G) Weekly average mean ± SEM of alcohol intake (g/kg/d) in males (2-way ANOVA, #P < 0.05, ##P < 0.01, ####P < 0.0001 vs 20%, †††P < 0.001 vs 50%, Tukey post hoc). (H) Mean ± SEM of alcohol intake (g/kg/d) for the individual sessions in females. (I) Weekly average mean ± SEM of alcohol intake (g/kg/d) in females. (J) Mean ± SEM of alcohol preference (%) for the individual sessions in males (2-way ANOVA, #P < 0.05 Tukey post hoc). (K) Weekly average mean ± SEM of alcohol preference (%) in males (2-way ANOVA, ####P < 0.0001 vs 20%,††P < 0.05, ††P < 0.01, †††P < 0.001 vs 50%, Tukey post hoc). (L) Mean ± SEM of alcohol preference (%) for the individual sessions in females (2-way ANOVA, #P < 0.05 Tukey post hoc). (M) Weekly average mean ± SEM of alcohol preference (%) in females (2-way ANOVA, #P < 0.05, ###P < 0.001, ####P < 0.0001 vs 20%, †P < 0.05 vs 50%, Tukey post hoc). ANOVA, analysis of variance; CFA, complete Freund adjuvant.

We then looked at individual consumption sessions in males for the total volume of liquid during these last weeks of experiments (Fig. 3B). The ANOVA for repeated measurements did not detect differences in treatment (F [1, 28] = 1.238, P = 0.2754), time (F [2.337, 65.43] = 1.533, P = 0.2209), or on the interaction between time and treatment (F [11, 308] = 1.333, P = 0.2047). Similarly, for the weekly average of total volume consumed (Fig. 3C), the ANOVA for repeated measurements did not detect differences in treatment (F [1, 28] = 0.5184, P = 0.4775), time (F [3, 84] = 0.8849, P = 0.4524), or in the interaction between time and treatment (F [3, 84] = 0.8324, P = 0.4798).

In the individual consumption sessions in females for the total volume of liquid (Fig. 3D), the ANOVA for repeated measures revealed a significant interaction for treatment and time (F [11, 308] = 1.889, P = 0.0401). However, it did not detect differences in treatment factor (F [1, 28] = 1.664, P = 0.2076) or in time (F [6.860, 192.1] = 1.243 P = 0.2816). Moreover, the Tukey post hoc analysis for multiple comparisons did not detect differences between saline and CFA groups in any of the sessions or intrasubject differences when comparing with the first session when rats were exposed to 20% alcohol. In the case of the weekly average (Fig. 3E), no differences were detected by the ANOVA for repeated measures in treatment (F [1, 28] = 1.664, P = 0.2076), time (F [3, 84] = 1.077, P = 0.3634), or on the interaction between time and treatment (F [3, 84] = 0.3495, P = 0.7896).

When analyzing alcohol intake in males for the individual sessions (Fig. 3F), the ANOVA for repeated measures revealed a main effect of time (F [6.219, 174.1] = 9.286, P < 0.0001), but not of treatment (F [1, 28] = 0.2651, P = 0.6107) or in the interaction between treatment and time (F [11, 308] = 1.186, P = 0.2959). In addition, when comparing with the first session of the 20% dose week, the Tukey post hoc analysis for multiple comparisons revealed a significant difference for the first session of the 50% in both saline (P = 0.0421) and CFA groups (P = 0.0497). However, when comparing the weekly average of alcohol intake in males (Fig. 3G), the ANOVA for repeated measures revealed a main effect of dose (F [3, 84] = 20.20, P < 0.0001) but not treatment (F [1, 28] = 0.2645, P = 0.6111) or a significant interaction between treatment and dose (F [3, 84] = 0.8702, P = 0.4600). Interestingly, the Tukey post hoc analysis for multiple comparisons revealed a significant decrease (as compared with 20% alcohol concentration), for the 30% (P = 0.0404), 40% (P = 0.0142) and 50% (P < 0.0001) doses in saline-treated males. However, in CFA-treated males, the post hoc only detected differences for the 40% (P = 0.0041) and 50% (P < 0.0001) doses when compared with 20%, but not for the 30% dose (P = 0.3263). Moreover, in CFA-treated males the post hoc analysis also detected a significant difference between weeks when rats were exposed to 30% and 50% (P = 0.0003) alcohol concentrations.

Surprisingly, different results were found in females. For example, for the alcohol intake in the individual sessions (Fig. 3H), the ANOVA for repeated measurements did not detect differences in treatment (F [1, 28] = 0.7631, P = 0.3898), time (F [4.369, 122.3] = 1.243, P = 0.2953), or in the interaction between time and treatment (F [11, 308] = 1.279, P = 0.2357). In a similar manner, for the weekly data in females (Fig. 3I), the ANOVA for repeated measures did not detect differences in treatment (F [1, 28] = 0.7635, P = 0.3897), dose (F [3, 84] = 0.6163, P = 0.6063), or in the interaction between treatment and dose (F [3, 84] = 0.7540, P = 0.5231).

When evaluating the alcohol preference in the individual sessions in male groups (Fig. 3J), the ANOVA for repeated measures detected a main effect of time (F [4.999, 140.0] = 34.99, P < 0.0001), but not treatment (F [1, 28] = 0.2605, P = 0.6137) or a significant interaction between the treatment and time (F [11, 308] = 1.326, P = 0.2085). Furthermore, when comparing with the first session when rats were exposed to 20% alcohol concentration, the Tukey post hoc analysis for multiple comparisons revealed a significant decrease of the alcohol preference for sessions 10, 11, 13, 14, and 16 to 18 postinjection in the saline group (P < 0.05) and for sessions 10, 11, and 13 to 18 in the CFA group (P < 0.05). When examining the preference that males exhibited weekly (Fig. 3K), the ANOVA for repeated measures detected a main effect of dose (F [3, 84] = 64.08, P < 0.0001), but not for treatment (F [1, 28] = 0.2605, P = 0.6138) or the interaction between dose and time (F [3, 84] = 1.426, P = 0.2410). Moreover, when comparing with the dose of 20% alcohol, the Tukey post hoc analysis for multiple comparisons revealed a significant decrease when for the 30%, 40%, and 50% doses in both saline-treated and CFA-treated animals (P < 0.0001). When compared with the 50% alcohol dose, the post hoc analysis revealed a significant difference with the 30% dose in the saline group (P = 0.0048) and for the 30% and 40% doses in the CFA group (P < 0.05).

Similarly, for the alcohol preference in the individual sessions in females (Fig. 3L), the ANOVA for repeated measures revealed a main effect of time (F [3.576, 100.1] = 16.13, P < 0.0001), but not treatment (F [1, 28] = 0.2923, P = 0.5930) or a significant interaction between the treatment and time (F [11, 308] = 1.222, P = 0.2710). However, when comparing with the first session when rats were exposed to 20% alcohol concentration, the Tukey post hoc analysis for multiple comparisons only revealed a significant difference for day 9 postinjection in the CFA group (P = 0.0401). Finally, for the weekly average (Fig. 3M), the ANOVA for repeated measures did not detect significant differences for treatment (F [1, 28] = 0.2921, P = 0.5931) or for the interaction of treatment and dose (F [3, 84] = 0.5527, P = 0.6477), but it revealed a main effect of dose (F [3, 84] = 26.44, P < 0.0001). When compared with the 20% concentration, the Tukey post hoc analysis for multiple comparisons revealed a significant decrease for the 30%, 40%, and 50% (P < 0.05) doses and between the 50% and 30% doses (P < 0.05) in both saline-treated and CFA-treated females.

3.4. Mechanical nociception hypersensitivity is unaltered in complete Freund adjuvant–injected rats throughout the experimental procedure

The von Frey test showed that mechanical PWTs were lower in both male (Fig. 4A) and female (Fig. 4B) rats under pain condition until the end of the experimental procedure. The ANOVA for repeated measures detected significant differences in treatment (males: F [1, 28] = 55.07, P < 0.0001; females: F [1, 28] = 35.88, P < 0.0001) and time (males: F [3.691, 103.4] = 5.459, P = 0.0007; females: F [2.575, 72.11] = 13.60, P < 0.0001) variables and in the interaction between time and treatment (males: F [5, 140] = 8.917, P < 0.0001; females: F [5, 140] = 10.95, P < 0.0001). The Bonferroni post hoc test confirmed that basal mechanical nociception was not different between groups (males and females: P > 0.9999). However, it revealed a significant decrease on days 5 to 35 after injection (males and females: P < 0.05) in comparison with the saline-treated rats in both male and female groups. Moreover, when comparing with basal levels, the Tukey post hoc analysis for multiple comparisons revealed a significant decrease of the mechanical nociception for days 5 to 35 postinjection only in CFA-treated males and females (P < 0.05).

Figure 4.:

Mean ± SEM of paw withdrawal threshold (PWT) (g) before and after saline (SAL, in black) or CFA (in red) injection for males (full symbols) and females (empty symbols). (A) Mean ± SEM of PAW in males (2-way ANOVA, *P < 0.05, **P < 0.01, ****P < 0.0001, Bonferroni post hoc, #P < 0.05, ##P < 0.01, ###P < 0.001, Tukey post hoc). (B) Mean ± SEM of PAW in females (2-way ANOVA, **P < 0.01, ***P < 0.001, ****P < 0.0001, Bonferroni post hoc, ###P < 0.001, ####P < 0.0001, Tukey post hoc). CFA, complete Freund adjuvant.

Figure 5.:

Graphical representation of the obtained results.

4. Discussion

Pain and AUD are 2 major health problems that can interfere with each other.3,7,8,15,19,24,32,33 Epidemiological findings suggest that pain may constitute a risk factor for heavy drinking, AUD, and relapse15,19,32,33 and that can differentially affect men and women.3,7 However, there is a lack of preclinical research on this topic that could help us understand the specific effects of pain on alcohol-related behaviors and the potential sex differences. Here, we conducted a thorough study in which we assessed the effect of chronic inflammatory pain on alcohol drinking in male and female rats with a history of alcohol exposure. Our results show that pain does not alter alcohol drinking behavior in females. In male rats, however, the presence of inflammatory pain blunts the decrease of alcohol intake when higher concentrations of alcohol are available (Fig. 5).

In our paradigm, after 5 weeks of exposure to alcohol following the IA model, inflammatory pain by injection of CFA in the hind paw was induced in half of our animals. Interestingly, during the following 3 weeks that the rats had access to 20% alcohol, we did not detect differences in total weekly alcohol intake or in the weekly preference for the alcohol bottle, when compared with the control (no pain) group. Interestingly, in females, we detected a decrease of alcohol intake only in the first session after CFA injection. However, this difference was not persistent when looking at the average of that week. Based on the schedule of our paradigm, this 1 time point event may not be a representative of the overall effect of pain on alcohol drinking behaviors. To the best of our knowledge, this is the first study in which the induction of pain is produced after the acquisition period using a 2-bottle choice model in rodents. A recent study in mice showed that capsaicin (an acute model of inflammatory pain) did not alter intake of 10% alcohol when mice were exposed to only 1 bottle with alcohol for 2 hours a day.14 Moreover, Fucich et al.9 reported that rats in pain generated from traumatic brain injury had a higher breakpoint for 10% alcohol in a progressive ratio session after being trained in a self-administration paradigm. These different results are likely due to differences in species, the model of pain, or the alcohol paradigm. In addition, the previously mentioned studies were only performed in male animals.

The interaction between pain and substance abuse has been deeply explored in the last few decades. Previous data from our laboratory show that rats under inflammatory pain (using the C