Introduction

Top Abstract Introduction Methods Results Discussion References

Popular and historical accounts suggest that women are more, or less, tolerant of pain than men are (Morris, 1991; Fillingim and Maixner, 1995). Clinical studies actually indicate that women report pain more often, greater intensity of pain and/or lower thresholds for painful stimuli than men do (Woodrow et al., 1972; Hilgard and LeBaron, 1982; Keefe, 1986; Unruh, 1996). In laboratory studies, women consistently report greater sensitivity than men to pain induced by mechanical pressure (Woodrow et al., 1972; Dubreuil and Kohn, 1986; Fransson-Hall and Kilbom, 1993; Ellermeier and Westphal, 1995). In contrast, sex differences in pain thresholds with warm thermal (Feine et al., 1991; Lautenbacher and Strian, 1991; Bush et al., 1993; Lautenbacher and Rollman, 1993), cold thermal (Westcott et al., 1977; Hall and Davies, 1991) and electrical (Liberson and Liberson, 1975; Rollman and Harris, 1987) stimulation are less consistent. Thus, some controlled studies demonstrate clear sex differences in sensitivity to noxious stimuli, particularly mechanical pressure stimuli.

A few human studies also indicate sex differences in analgesic effects of drugs. A prospective study on the use of patient-controlled morphine analgesia after abdominal surgery showed that female patients required significantly less morphine (in milligrams) than males did (DeKock and Scholtes, 1991). Furthermore, Gear and colleagues recently demonstrated significantly greater pentazocine-, butorphanol- and nalbuphine-induced analgesia in women than in men against dental pain (Gear et al., 1996a, b). In addition, a single dose of ibuprofen which was effective against electrically induced experimental pain in men was ineffective in women (Walker and Carmody, 1996). Thus, these studies indicate that there may be substantial sex differences in potency or efficacy of analgesic drugs. Several experimental studies in rodents support this hypothesis. For example, male mice showed greater antinociception than females after administration of the mu opioid morphine (Kavaliers and Innes, 1987; Lipa and Kavaliers, 1990; Candido et al., 1992). Similarly, male rats showed greater antinociception than females after systemic (Baamonde et al., 1989; Islam et al., 1993; Cicero et al., 1996) or central (Kepler et al., 1989) administration of morphine or the mu opioid peptide DAMGO (Kepler et al., 1991). In contrast to the above reports, Ali and colleagues (1995) reported that morphine antinociception was significantly greater in female rats than in males, and Kepler et al. (1991) found no sex differences in DAMGO-induced antinociception when the jump test was used rather than the tail-flick test. Sex differences in kappa opioid antinociception have been examined in a single study, with male mice displaying significantly greater antinociception than females after administration of U50,488 (Kavaliers and Innes, 1987). Sex differences in delta opioid antinociception also have been examined in a single study: there were no significant sex differences in antinociception produced by DSLET in rats (Kepler et al., 1991). Thus, sex differences in opioid antinociception have been demonstrated for two of the three major classes of opioid agonists, and in most cases, males show greater opioid antinociception than females do. It is important to note that in some studies, the effects of one or more doses were examined at a single time point only (Baamonde et al., 1989; Candido et al., 1992; Ali et al., 1995), whereas in other studies, one dose was examined at multiple time points (Kavaliers and Innes, 1987; Lipa and Kavaliers, 1990); thus, in some cases it is difficult to determine whether sex differences in opioid antinociception are caused by sex differences in drug potency, time course of drug effect or both. In addition, it is not yet clear whether this sex difference in opioid antinociception is a general phenomenon that occurs with most opioid agonists, or whether it is limited to a class of opioids or even particular drugs.

The purpose of the present study was to more fully characterize sex differences in thermal antinociception produced by a variety of opioids in the rat. Dose- and time-dependent antinociceptive effects of two agonists from each major class of opioid (mu, kappa, delta) were examined on the hot-plate and warm-water tail-withdrawal assays in adult female and male rats. Additionally, because rats were tested repeatedly in the present study (each rat received three doses of a particular opioid agonist plus vehicle, one treatment/week for 4 weeks), correlational analyses were conducted to examine development of tolerance in female and male rats.

	Methods

Top Abstract Introduction Methods Results Discussion References

Subjects. For each drug, eight male and eight female gonadally intact Sprague-Dawley rats, approximately 3 months old at the beginning of testing, were used. Rats were housed in same-sex pairs in 35.0 cm × 30.5 cm × 25.0 cm plastic tubs; males and females were located in separate, adjacent rooms, except during testing. Five days before the beginning of each experiment, rats were deprived of free access to Teklad rat chow; to maintain constant body weight throughout the experiment, rats were given 1.5 h access to food Monday through Thursday, with free access to food from Friday evening until Sunday afternoon. Access to water was ad libitum throughout the experiment. Animal quarters were maintained at 21.5 ± 2°C, on a 12:12-h light/dark cycle, with lights on at 7:00 A.M..

Surgery. For i.c.v. drug administration (DPDPE and deltorphin experiments), rats were implanted with a cannula in the lateral ventricle. Rats were anesthetized with 0.25 ml/100 g b.wt. Equithesin (active ingredients: pentobarbital sodium and chloral hydrate) i.p., and placed in a stereotaxic apparatus (Stoelting Instruments, Wood Dale, IL). In addition, 0.1 ml of 10 mg/ml of a local anesthetic, lidocaine hydrochloride, was administered intradermally at the site of incision, 1 to 2 min before incision. A 22-gauge stainless steel guide cannula (Plastics One, Roanoke, VA) was then implanted into the right or left lateral ventricle with use of the following coordinates: 0.9 mm A/P, ± 1.6 mm L/M, and 3.6 mm D/V with lambda and bregma in the same horizontal plane (Paxinos and Watson, 1986). After surgery, penicillin (11,000 U/kg) was administered i.m. to prevent infection. Cannula placement was verified 1 week postsurgery by measuring water intake after i.c.v. administration of angiotensin II (100 pmol/µl artificial cerebral spinal fluid). Rats that failed to drink at least 5 ml of water in 10 min were excluded from testing. Forty-eight hours after cannula placement was verified, antinociceptive testing began.

Apparatus. For hot-plate testing, a Hot Plate Analgesia Meter (Columbus Instruments, Columbus, OH) with the temperature set digitally at 52.0 ± 0.1°C was used. The surface of the hot plate measured 25.3 cm × 25.3 cm × 3.0 cm and was surrounded by 30-cm-high Plexiglas walls. For the tail-withdrawal procedure, a 5-gal water bath (Precision Scientifics, Inc., Chicago, IL) with the temperature set at 52 ± 1°C was used. Rats without i.c.v. cannulae were restrained in Plexiglas tubes (IITC Inc., Los Angeles, CA) with an opening at one end through which the tail hung freely. Rats with i.c.v. cannulae were similarly restrained in a cloth wrapped so that the hind legs were immobilized, but the tail hung freely.

Drugs. The mu agonists tested were fentanyl hydrochloride (National Institute on Drug Abuse (NIDA), Rockville, MD) and buprenorphine hydrochloride (NIDA). Fentanyl was dissolved in 0.9% physiological saline, and saline was the control vehicle. Buprenorphine was dissolved in lactic acid to which saline was added (pH adjusted to 6 with 1 N NaOH); the lactic acid solution served as the control vehicle for buprenorphine. The kappa agonists were U69,593 and ()-bremazocine hydrochloride (NIDA), which were dissolved in lactic acid (U69,593) or lactic acid plus ethanol (bremazocine) to which distilled water was added (pH adjusted to 5.5 with 1 N NaOH; final ethanol concentration 5.3%). The acid and acid/ethanol solutions served as the control vehicles for U69,593 and bremazocine, respectively. The delta agonists were DPDPE and deltorphin (NIDA), which were dissolved in sterile water (DPDPE) or 10% dimethyl sulfoxide (deltorphin) and prepared immediately before injecting; the sterile water and 10% dimethyl sulfoxide solutions served as control vehicles for DPDPE and deltorphin, respectively.

Procedure. Each rat was tested with three doses of a single agonist, plus the appropriate vehicle; testing was conducted one time per week for four consecutive weeks. Because dose order might affect outcome, dose order (including vehicle) was randomized such that no two rats of the same sex received the same order of doses (sample testing order: vehicle, low dose, intermediate dose, high dose; low dose, high dose, vehicle, intermediate dose; high dose, vehicle, intermediate dose, low dose; etc.). The experimenter was blind to dose. Five minutes after injection, the rat was placed on the hot plate, and latency to lick a hindpaw or jump out of the apparatus was recorded by the digital timer on the hot plate. To prevent tissue damage, rats that did not exhibit one of the target behaviors by 45 sec were removed from the hot plate and assigned the maximal score of 45 sec. The rat was then lightly restrained for tail immersion in the waterbath. The distal 5 cm of the tail was immersed and latency to withdraw at least 4 cm of the tail from the water was recorded with a hand-held stopwatch. To prevent tissue damage, rats that did not withdraw their tails within 20 sec were removed from the waterbath and assigned the maximal score of 20 sec. After these tests, the rat was returned to its home cage until the next time point. The testing procedure above was repeated 15, 30, 60 and 90 min postinjection for fentanyl, DPDPE, deltorphin and U69,593; 15, 30, 60, 90 and 120 min postinjection for bremazocine; and 15, 30, 60, 90, 120 and 180 min postinjection for buprenorphine. The presence or absence of sedation (flattened body posture on the hot plate), Straub tail (tail stiffened and held up off surface of hot plate) and lick versus jump off the hot plate also was noted for each rat at each time point. Rats then were placed in their home cages and returned to the vivarium until the next week. All time course determinations were conducted between noon and 5 P.M.

Data analysis. To determine whether there were sex differences in response latency after vehicle administration, vehicle scores in each of the six drug experiments were separately analyzed with a two-way (sex, time) repeated measures (time) ANOVA. Because there were individual differences in vehicle scores (including some sex differences and some differences across time), drug data were corrected for these individual differences as follows: for each rat, latency to respond after vehicle administration was subtracted from latency to respond after drug administration at the same time point ("difference scores"). These difference scores then were analyzed with a three-way (sex, dose, time) repeated measures (time, dose) ANOVA. To compare the development of tolerance to each opioid among female and male rats, difference scores for the highest dose of each drug were transformed into AUC units, with use of the trapezoidal rule for unequally spaced x values (Sigmaplot, Jandel Scientific, San Rafael, CA). Only hot-plate AUC values were calculated, and only for the highest dose of each opioid, because the delta opioids produced measurable effects only on the hot-plate assay and because low doses of most opioids produced minimal effects on hot plate (such that the development of tolerance to low doses would not be readily apparent). By transforming time-effect data into AUC units, both magnitude and duration of effect could be represented in a single value; decreases in this value, which could reflect decreases in magnitude and/or duration of drug effect, would indicate development of tolerance. The Spearman rank order test was used to determine whether the magnitude of the AUC score was significantly correlated with order in which the highest dose was tested (first, second or third of the three doses of drug tested). A significant negative correlation between AUC values and order of test was taken to indicate that tolerance developed during the 4-week testing period. Significance level was P  .05 for all statistical tests.

	Results

Top Abstract Introduction Methods Results Discussion References

Vehicle effects. Table 1 shows the latency of male and female rats to respond on the hot-plate and tail-withdrawal assays after vehicle administration in each of the six drug experiments. After vehicle administration, females tended to have higher latencies than males in the U69,593, buprenorphine, DPDPE and deltorphin hot-plate experiments, and in the deltorphin tail-withdrawal experiment; this sex difference was statistically significant in the U69,593 and buprenorphine hot-plate experiments (table 1). In addition, latencies of rats changed significantly over time in the buprenorphine and bremazocine hot-plate experiments, and in the fentanyl, U69,593 and bremazocine tail-withdrawal experiments; fluctuations in latency to respond over time were different between females and males in the buprenorphine and bremazocine hot-plate experiments, and in the U69,593 and bremazocine tail-withdrawal experiments (sex by time interactions, table 1).

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Antinociceptive effects of the mu opioid fentanyl in female and male rats. One dose of fentanyl or vehicle was administered s.c. one time per week for 4 consecutive weeks; rats were tested on the 52°C hot-plate (top panel) and tail-withdrawal assays (bottom panel) each week. Each point is the mean ± 1 S.E.M. of the difference scores of eight rats: each rat's vehicle latency score was subtracted from its drug latency score at each time point to correct for individual differences in nondrug responding. Results of statistical analyses are shown in table 2.

Mu opioid antinociception. Figure 1 shows that the mu agonist fentanyl time- and dose-dependently increased latency to respond on both hot-plate and tail-withdrawal tests in both sexes. There were no significant sex differences in fentanyl's antinociceptive effects (table 2); the time course and magnitude of fentanyl's effects were very similar in males and females. Two females and two males tested at the high or intermediate doses of fentanyl displayed sedation and Straub tail, beginning by 5 min postinjection and subsiding by 15 to 30 min postinjection. Figure 2 shows that another mu agonist, buprenorphine, also time-dependently increased latency to respond on both hot-plate and tail-withdrawal tests in both sexes; however, all three doses produced approximately the same effects (thus, there was no main effect of dose on either assay). There were no significant sex differences in buprenorphine's antinociceptive effects (table 2). Approximately half of the rats tested (3 female, 5 male) at the intermediate or high doses of buprenorphine displayed sedation, beginning approximately 15 min postinjection, and subsiding by 90 to 120 min postinjection.

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Antinociceptive effects of the mu opioid buprenorphine in female and male rats. One dose of buprenorphine or vehicle was administered s.c. one time per week for 4 consecutive weeks; rats were tested on the 52°C hot-plate (top panel) and tail-withdrawal assays (bottom panel) each week. Each point is the mean ± 1 S.E.M. of the difference scores of eight rats: each rat's vehicle latency score was subtracted from its drug latency score at each time point, to correct for individual differences in nondrug responding. Results of statistical analyses are shown in table 2.

Kappa opioid antinociception. Figure 3 shows that the kappa agonist U69,593 time- and dose-dependently increased latency to respond on both hot-plate and tail-withdrawal tests in both sexes. There were no significant sex differences in U69,593's effects on hot-plate. In contrast, on the tail-withdrawal test, U69,593 produced its peak effects 5 to 15 min postinjection in females but 30 min postinjection in males (sex by time interaction, table 2). Three males and one female tested at the highest dose of U69,593 jumped off the hot plate (rather than licking a hindpaw) at one or more time points. Figure 4 shows that another kappa agonist, bremazocine, time-dependently increased latency to respond on hot-plate and tail-withdrawal tests, but produced dose-dependent effects on the hot-plate test only. Bremazocine produced somewhat greater antinociception in females than in males at some doses and time points on both assays, and this sex difference was statistically significant on the tail-withdrawal assay (table 2). Six male and four female rats tested at various doses of bremazocine jumped off the hot plate (rather than licking a hindpaw) at one or more time points.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Antinociceptive effects of the kappa opioid U69,593 in female and male rats. One dose of U69,593 or vehicle was administered s.c. one time per week for 4 consecutive weeks; rats were tested on the 52°C hot-plate (top panel) and tail-withdrawal assays (bottom panel) each week. Each point is the mean ± 1 S.E.M. of the difference scores eight rats: each rat's vehicle latency score was subtracted from its drug latency score at each time point, to correct for individual differences in nondrug responding. Results of statistical analyses are shown in table 2.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Antinociceptive effects of the kappa opioid bremazocine in female and male rats. One dose of bremazocine or vehicle was administered s.c. one time per week for 4 consecutive weeks; rats were tested on the 52°C hot-plate (top panel) and tail-withdrawal assays (bottom panel) each week. Each point is the mean ± 1 S.E.M. of difference scores of eight rats: each rat's vehicle latency score was subtracted from its drug latency score at each time point, to correct for individual differences in nondrug responding. Results of statistical analyses are shown in table 2.

Delta opioid antinociception. Cannula placement was verified in all rats after i.c.v. administration of angiotensin II; all rats drank at least 5 ml of water in 10 min (data not shown). However, one male rat in the deltorphin experiment showed evidence of infection during the week after surgery, and so was not tested with deltorphin. Figure 5 shows that the delta agonist DPDPE produced time-dependent increases in latency to respond on the hot-plate test, and to a very limited extent on the tail-withdrawal test. In females, all three doses of DPDPE produced approximately the same hot-plate effect, which was significantly less than that produced by the highest dose of DPDPE in males (sex by dose interaction, table 2). Additionally, peak antinociception occurred 5 min postinjection in females, whereas it occurred 15 to 30 min postinjection in males. Two males tested at the high dose of DPDPE, two females tested at the intermediate dose and one male tested at the low dose displayed sedation, all beginning by 5 min postinjection and subsiding by 30 min postinjection. Figure 6 shows that another delta agonist, deltorphin, also time- and dose-dependently increased latency to respond on the hot-plate test, and the highest dose of deltorphin produced significantly greater hot-plate antinociception in males than in females (sex by dose interaction, table 2). On the tail-withdrawal test, deltorphin produced only minimal increases in latency to respond (primarily in males). All rats tested with deltorphin (or deltorphin vehicle) displayed hyperactivity during the first test day, beginning by 5 min postinjection and subsiding by 30 to 60 min postinjection; excessive locomotor activity was not noted after the first day of testing.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Antinociceptive effects of the  opioid DPDPE in female and male rats. One dose of DPDPE or vehicle was administered i.c.v. one time per week for 4 consecutive weeks; rats were tested on the 52°C hot-plate (top panel) and tail-withdrawal assays (bottom panel) each week. Each point is the mean ± 1 S.E.M. of the difference scores of eight rats; each rat's vehicle latency score was subtracted from its drug latency score at each time point to correct for individual differences in nondrug responding. Results of statistical analyses are shown in table 2.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Antinociceptive effects of the delta opioid deltorphin in female and male rats. One dose of deltorphin or vehicle was administered i.c.v. one time per week for 4 consecutive weeks; rats were tested on the 52°C hot-plate (top panel) and tail-withdrawal assays (bottom panel) each week. Each point is the mean ± 1 S.E.M. of the difference scores of eight female (or seven male) rats: each rat's vehicle latency score was subtracted from its drug latency score at each time point to correct for individual differences in nondrug responding. Results of statistical analyses are shown in table 2.

Development of tolerance. Figure 7 shows scatter plots of hot-plate AUC values for the highest dose of each opioid agonist as a function of order of testing of that dose (first, second or third of three doses tested) in individual female and male rats. There was evidence of the development of tolerance in at least one sex in several experiments: there was a significant negative correlation between AUC value and order in which the highest dose was tested, for U69,593 (females: r = 0.85, P = .002), bremazocine (males: r = 0.72, P = .04) and deltorphin (females: r = 0.79, P = .02); that is, rats tested with the highest dose of these opioids at week 3 or 4 (after they had received two lower doses of the same opioid) showed less antinociception than rats tested with the highest doses at week 1 or 2 (first time receiving drug). Because female and male rats were not necessarily tested with the same order of doses, only limited sex comparisons can be made. The regression lines shown in figure 7 suggest possible sex differences in the development of tolerance in the deltorphin experiment, in which females showed a significant negative correlation between AUC value and order of test, whereas males showed a significant positive correlation (r = 0.78, P = .03).

View larger version (25K): [in this window] [in a new window]   Fig. 7.   Correlation between effect produced by the highest dose of each opioid agonist and order in which that dose was tested (1st, 2nd or 3rd of the 3 doses tested) in individual female (open circles, dashed regression lines) and male (closed circles, solid regression lines) rats tested on the hot-plate assay. Area-under-the-curve values were calculated from each individual rat's corrected time-effect curve (drug latency-vehicle latency) so that antinociception across the entire time course could be represented as a single value. For each drug, eight females and eight males were tested (except deltorphin, n = 8 females, 7 males); some individual data points overlap completely and thus are not visible.

	Discussion

Top Abstract Introduction Methods Results Discussion References

The present results indicate that there are sex differences in opioid antinociception, but these differences may be assay-, dose- and/or time-dependent. This study confirms the literature regarding antinociception induced by mu, kappa and delta opioid agonists in male rats: all three classes of opioids produced antinociception on the hot-plate assay (e.g., VonVoigtlander et al., 1983; Schmauss and Yaksh, 1984; Ossipov et al., 1995), whereas mu and kappa (but not delta) opioid agonists produced antinociception on the warm-water tail-withdrawal assay (VonVoigtlander et al., 1983; Adams et al., 1993; Walker et al., 1994; Ossipov et al., 1995). All opioid agonists tested in the present study also produced antinociception in female rats on at least one assay. Thus, this study replicates the literature demonstrating that opioid antinociception can be induced in female rodents (Kavaliers and Innes, 1987; Kepler et al., 1991; Candido et al., 1992; Cicero et al., 1996) and extends this finding to several opioid agonists previously unexamined in females.

In some experiments in the present study, there were sex differences in latency to respond to noxious stimuli under nondrug conditions. In four of six experiments, females tended to have higher hot-plate (and in one case, tail-withdrawal) latencies than males did after vehicle administration; this sex difference was statistically significant in two of those experiments. Sex differences in response to noxious stimuli under base-line or vehicle conditions have been reported previously, but are not consistent across studies. For example, similar to the present study, Forman and colleagues (1989) reported that Sprague-Dawley female rats had longer hot-plate and tail-withdrawal latencies than males did; however, explicit sex comparisons were not conducted so it is not clear whether the differences were statistically significant. Female rats also had significantly longer latencies than males on a tail-flick test (Islam et al., 1993). In contrast, female rodents were more sensitive to the noxious stimulus than males were using electric shock (Beatty and Beatty, 1970; Kepler et al., 1991) or injection of formalin into the paw (Aloisi et al., 1994). In many other cases, no significant sex difference in base-line sensitivity to noxious stimuli has been found, on the tail-flick test (Kepler et al., 1991; Candido et al., 1992; Menendez et al., 1994), the tail-withdrawal (immersion) test (Menendez et al., 1994) or the hot-plate test (Kavaliers and Innes, 1987; Ali et al., 1995). It is likely that both type and intensity of the noxious stimulus influence whether sex differences are observed, as appears to be the case in humans as well (Lautenbacher and Rollman, 1993; Ellermeier and Westphal, 1995). Given the occasional base-line sex difference obtained in the present study, we chose to correct for differences in base-line sensitivity to the noxious stimuli before analyses of drug effects; this correction is commonly applied in studies with all male subjects, but has not always been applied in studies comparing females with males (e.g., Kavaliers and Innes, 1987; Lipa and Kavaliers, 1990). Failure to correct for even small sex differences in base line may lead to different conclusions. For example, in the present study, female rats had significantly higher latencies than males did at all doses of buprenorphine and U69,593; however, females also had higher latencies than males after administration of vehicle in these two experiments. Thus, analysis of the drug results without correction for the base-line sex difference would lead to the conclusion that buprenorphine and U69,593 produced significantly greater antinociception in females than in males. Although this correction may be theoretically important, we do not know whether individual differences in base-line sensitivity to noxious stimuli contribute to individual differences in drug-induced analgesia in humans; thus, the functional significance of using proportional rather than absolute values of sensitivity to noxious stimuli is unclear.

Once data were corrected for individual differences in latency to respond under nondrug conditions, an examination of sex differences by class of opioid showed that there were sex differences in antinociception produced by both kappa and both delta agonists, but not by either of the mu opioids tested, buprenorphine and fentanyl. The dose range for buprenorphine turned out to be too narrow to fully answer the question; all three doses produced approximately the same intermediate effects; and thus, it is possible that there would be sex differences at lower or higher doses of buprenorphine. In contrast, the three doses of fentanyl examined produced effects that spanned from minimal to maximal antinociception on both assays, and there were no significant sex differences in magnitude or duration of fentanyl's antinociceptive effects. The lack of sex differences with either fentanyl or buprenorphine contrasts with previous studies in which the mu opioid morphine was examined in rats (Baamonde et al., 1989; Kepler et al., 1989; Islam et al., 1993; Cicero et al., 1996) or mice (Kavaliers and Innes, 1987; Lipa and Kavaliers, 1990; Candido et al., 1992). In all of these studies, peak or total morphine antinociception was greater in males than in females. In contrast, Kepler and colleagues (1991) found that the mu opioid peptide DAMGO produced greater antinociception in male than in female rats on the tail-flick test, but not on the shock-jump test; and Ali and colleagues (1995) reported that morphine antinociception was greater in female rats than in males on the hot-plate test. The previous studies examining morphine antinociception encompass a variety of antinociceptive assays: writhing (Baamonde et al., 1989; Cicero et al., 1996), tail-flick (away from radiant heat stimulus) (Kepler et al., 1989, 1991; Candido et al., 1992; Islam et al., 1993; Cicero et al., 1996), jump (away from electric shock stimulus) (Kepler et al., 1989, 1991) and hot-plate (50°C, Kavaliers and Innes, 1987; 52°C, Ali et al., 1995; 58°C, Cicero et al., 1996). The present study differed from some of these previous studies in the antinociceptive assays used, and differs from all of the previous studies in the particular mu opioids examined; thus, sex differences in mu opioid antinociception may depend on the assay and particular mu opioid examined. Additionally, it is important to note that sex differences in opioid antinociception have been assessed using single-dose, single-time point determinations (Ali et al., 1995), multiple-dose, single-time point determinations (Baamonde et al., 1989; Candido et al., 1992), single-dose, multiple-time point determinations (Kavaliers and Innes, 1987; Lipa and Kavaliers, 1990) and multiple-dose, multiple-time point determinations (Kepler et al., 1989, 1991; Cicero et al., 1996; present study), which may contribute to the lack of agreement across studies. For example, had we examined only the lowest dose of fentanyl in either assay across the time course, or the intermediate dose of fentanyl at its peak effect in the tail-withdrawal assay (15 min postinjection), we may have concluded that fentanyl produced significantly greater antinociception in males than in females (fig. 1).

For both kappa opioid agonists tested, significant sex differences were observed on the tail-withdrawal assay. Whereas the sex difference in effect of U69,593 was primarily one of time course (antinociception peaked earlier in females than in males), the sex difference in the effect of bremazocine (antinociception greater in females than in males) was not time- or dose-dependent. In the only previous study in which sex differences in kappa opioid antinociception were examined, Kavaliers and Innes (1987) also reported sex differences produced by the kappa agonist U50,488 in deer mice: 1.0 mg/kg U50,488 produced significantly greater antinociception in males than in females, 30 and 60 min postinjection (but not 15, 90 or 120 min postinjection). Our U69,593 results generally agree with those of Kavaliers and Innes (1987): male rats showed greater U69,593-induced antinociception than females did 30 to 60 min postinjection. In addition, Kavaliers and Innes examined several doses of U50,488 only at 30 min postinjection; all doses produced greater effects in males than in females, which also agrees with our results at the 30-min time point. However, Kavaliers and Innes did not find a sex difference in time course of U50,488's effects, as we did with U69,593. Moreover, we found the opposite sex difference with bremazocine (females showed greater antinociception than males did), and our significant sex differences were on the tail-withdrawal assay, whereas theirs were on the hot-plate assay (albeit at 50°C). Thus, the precise nature of sex differences in kappa opioid antinociception may depend on the particular kappa agonist, species and stimulus intensity examined.

For both delta opioid agonists examined in the present study, sex by dose interactions were observed on the hot-plate test; whereas lower doses of DPDPE and deltorphin produced equivalent or greater antinociception in females than in males, the highest doses produced greatest antinociception in males. It is possible that doses higher than 100 nmol would have produced greater effects in females; however, the fact that some males and females showed brief but marked sedation after administration of 56 to 100 nmol DPDPE suggests that these doses were near the limit of what could be administered without causing substantial motor impairment. Other investigators have reported that high doses of delta agonists may produce adverse motor effects in rats (Stewart and Hammond, 1993; Ossipov et al., 1995). Thus, it is unlikely that testing a higher dose in females would have produced greater antinociceptive effects without adverse motor effects. The only previous study in which sex differences in the antinociceptive effects of a delta opioid agonist were examined showed no significant sex differences in peak or total antinociception after i.c.v. administration of DSLET to rats (Kepler et al., 1991). In addition to testing different opioid peptides, however, our study differed from that of Kepler and colleagues in the assays used (hot-plate vs. tail-flick and jump), and probably the stage of the estrous cycle during which females were tested (all of theirs were tested during vaginal estrus, whereas we did not assess cycle); such procedural differences could contribute to the discrepant results.

In addition to the species, assay and other procedural differences between the present and previous studies, there is one additional factor that may be particularly important when comparing opioid antinociception in female and male rodents: tolerance. In some studies, animals were tested acutely (one treatment condition per animal: Candido et al., 1992; Ali et al., 1995), whereas in other studies each animal was tested with two or more doses, at intervals of 3 (Baamonde et al., 1989), 4 or more days (Kavaliers and Innes, 1987) or 7 days (Kepler et al., 1989, 1991; Islam et al., 1993; present study). Repeated testing with opioids has been shown to induce tolerance, that is, decreasing drug effect on each subsequent test (Ferguson et al., 1969; Martin et al., 1976; Solomon et al., 1987). In fact, our results indicate that for some drugs examined in the present study, there was a significant negative correlation between drug effect and order of testing of the highest dose, which suggests that tolerance developed because of repeated testing. Furthermore, opioid tolerance may develop at different rates in female and male rats; although the present study was not specifically designed to investigate sex differences in tolerance development, the correlational analyses suggest that tolerance may have developed differentially in females vs. males (e.g., to deltorphin, fig. 7). In a previous study, morphine antinociception decreased more rapidly in male than in female rats administered 15 mg/kg of morphine daily; that is, males showed greater morphine antinociception than females did on day 1, but females showed greater antinociception than males did on day 13 (Badillo-Martinez et al., 1984). Thus, sex differences in opioid antinociception may depend on whether opioids are administered acutely or chronically, and it is likely that the particular drug testing paradigm used contributes to the discrepancies among studies of sex differences in opioid antinociception. Espejo and colleagues (1994) noted that both acute and repeated testing paradigms are important for examining the effects of opioids on nociception, to provide clinically relevant information on both acute and chronic effects of opioid analgesics.

Possible mechanisms underlying sex differences in opioid antinociception include sex differences in opioid pharmacokinetics, opioid receptor pharmacology and gonadal hormones. Previous studies have shown sex differences in brain levels of morphine 45 to 60 min postinjection (males had higher brain levels of morphine than females did; Candido et al., 1992; Craft et al., 1996), although no sex difference in blood levels of morphine has been reported (Cicero et al., 1996; Craft et al., 1996). The fact that sex differences in opioid antinociception have been observed after i.c.v. administration (Kepler et al., 1989, 1991), however, suggests that factors other than pharmacokinetics may be involved. Several studies have demonstrated that morphine antinociception may be gonadal hormone-dependent in male and female rodents (e.g., Kepler et al., 1989; Islam et al., 1993; Ali et al., 1995); however, other investigators report no significant difference in opioid antinociception between sham-gonadectomized and gonadectomized rats (e.g., Kepler et al., 1991; Cicero et al., 1996). Additionally, opioid receptor number may change over the estrous cycle in female rodents and may differ between males and females (Hammer, 1990; but see Candido et al., 1992). It has not yet been determined whether sex differences in opioid receptor pharmacology underlie sex differences in opioid antinociception. Further investigation of these and other potential mechanisms is necessary.

In summary, there were no sex differences in mu opioid antinociception in the present study, whereas sex differences in the antinociceptive effects of kappa and delta opioids were observed. However, these sex differences were assay-, dose- and/or time-dependent. The onset and peak of antinociceptive effects of some opioids varied between males and females; for example, peak effects of U69,593 and DPDPE tended to occur earlier in females than in males. In addition, there were sex differences in potency of bremazocine, DPDPE and deltorphin: bremazocine produced greater tail-withdrawal antinociception in females than in males at the doses examined, whereas the highest doses of the two delta opioids produced greater hot-plate antinociception in males than in females. Thus, it is important to consider complete time- and dose-effect relationships when investigating the variable of sex on opioid antinociception. Methodological differences among studies that may contribute to lack of agreement regarding sex differences in opioid antinociception include particular opioid agonist examined, type of assay, time point/dose observed and acute vs. repeated testing.