Painful heat attenuates electrically induced muscle pain in men and women

2.1 Design

The experiment was a single-blind cross-over study with two sessions, a test session with painful conditioning and a control session with non-painful conditioning. Half of the subjects participated in the test session first followed by the control session. In the other half, the order of the sessions was reversed. Subject gender was balanced between the groups. The second session started 30 min after the first session was finished. In each session electrical test stimuli (TS) were delivered in three blocks (before CS, during CS and 3 min after CS). In the test session a painful contact heat stimulus was delivered as a conditioning stimulus (CS) simultaneously with the electrical TS in the second block. In the control session no CS was delivered.

2.2 Subjects

Forty subjects (20 females/20 males) participated in the study. Mean age was 23.6 ± 2.7 years (male) and 25.2 ± 2.7 years (female; p = 0.08). All participants responded to an announcement at the homepage of the National Institute of Occupational Health, Norway or to flyers posted at universities and colleges in Oslo. All participants were paid 150 NOK/hour (≈$25). Inclusion criteria were age between 18 and 45 years and self-reported good health. Exclusion criteria were diagnosis of any somatic or psychiatric disease Furthermore, individuals who used prescription drugs for blood pressure, sedatives, antidepressants, or allergy medication, and smokers were excluded. Subjects were instructed not to drink any alcohol during the last 24 h before the experiment. An informed consent was obtained from each individual. The experimental protocol was approved by the regional ethical committee and was conducted according to the Helsinki Declaration.

2.3 Menstrual cycle

At inclusion, the women reported the first day of their last menstruation (=day 1). All women were tested during days 12-14 in reference to this day. Days 12-14 is referred to as the ovulatory phase by Tousignant-Laflamme and Marchand [22]. If more than 14 days had elapsed since the last menstruation, the subject was asked to report back on the first day of their next menstruation.

2.4 Procedures

2.5 Instructions to participants

The subjects were informed that the purpose of the study was to test their sensitivity to electrical stimulation of the tibialis anterior muscle. They were informed that three blocks, each consisting of 10 painful electrical stimuli, would be given to each leg in two separate sessions. They were further told that a metal element warmed to 35 °C would be attached to their opposite forearm. Before the second block in both sessions, the subjects were told that the temperature of the heating element would be increased to the pain-50 intensity. Note that the same instruction was given in both sessions, regardless of whether it was a test session or a control session. A written manuscript used by the experimenter (MG) ensured that identical oral information was given to all subjects.

2.6 Data analysis and statistics

Due to non-normal distribution, the electrical PT and TS pain ratings were log-transformed. Paired t-tests were performed to determine if the electrical PT was different between sessions, and to determine if pain ratings in response to TS at 1.1 × PT were different from ratings of TS at 1.6 × PT. Data from the ‘before CS’ block in each session was used for this analysis. Gender differences in electrical PT and pain-50 were tested with independent t-tests.

To test the inhibitory effect of CS on TS pain ratings (CPM-effect), a 2-way repeated measures analysis of variance (RM-ANOVA) was performed for each TS intensity with block (before, during, after) as one repeated measures factor and session (test vs. control) as the other repeated measures factor. Gender was entered as a between subjects factor and allowed us to determine gender differences in TS pain ratings as well as gender differences in the CPM effect. Greenhouse-Geisser correction was used when non-sphericity was detected (Maunchly’s test). The percentage of CPM-responders/non-responders was calculated. Subjects who did not have any inhibitory CPM effect during or after painful CS were defined as non-responders. Data were analyzed separately for all subjects and for CPM-responders. Correlations between the CPM effect, pain-50 and electrical pain thresholds were tested by Spearman’s bivariate analysis. PASW 18.0 was used for all statistical analyses (IBM, Chicago, Illinois, USA) and p<0.05 was considered significant.

3.1 Session differences in electrical PT

The electrical PT was equal in the two sessions; 5.43 ± 0.98 mA (mean ± standard deviation) in the first session and 5.83 ± 1.24 mA in the second session (t(39) = 0.005, p = 0.996). Both genders were included in this analysis.

3.2 Gender differences in electrical PT and pain-50

The electrical PT did not differ significantly between genders. The electrical PT averaged across the two sessions was 6.55 ± 6.52 mA for men and 4.71 ± 6.43 mA for women (t(38) = 1.65, p = 0.106).

The intensity of the painful conditioning stimulus (pain-50) differed between genders. Pain-50 was 47.8 ± 0.6°C for men and 46.9 ± 0.8 °C for women (t(38) = 3.7, p = 0.001).

3.3 The effect of painful and non-painful CS on TS

3.3.1 CPM effects during CS

For the high TS intensity there was a significant interaction between session (painful CS vs. non-painful CS) and block (before CS, during CS, after CS), indicating stronger CPM effect during painful CS than during non-painful CS (F(1,38) = 4.43, p = 0.042). Table 1 shows the TS pain ratings before, during and after conditioning for all subjects and for the 34 responders. Fig. 1 shows the CPM-effect during CS, relative to before CS. For the low TS intensity, there was an inhibitory CPM effect during both painful CS and during non-painful CS (F(1,38) = 4.62, p = 0.038). As expected, the interaction between session and block was stronger with only responders included (F(1,32) = 7.52, p = 0.01).

Table 1

Test stimulus (TS) pain ratings (VAS) before, during and after painful and non-painful conditioning stimulus (CS). Data are mean ± standard deviation.

	Painful CS		Non-painful CS
	Low TS intensity (mm)	High TS intensity (mm)	Low TS intensity (mm)	High TS intensity (mm)
All subjects (n =40)
Male
Before CS	22.9 ± 16.6	29.1 ± 17.4	25.4 ± 17.6	30.9 ± 20.6
During CS	20.7 ± 14.2	25.9 ± 18.2	23.0 ± 17.0	29.0 ± 21.1
After CS	21.8 ± 16.9	27.5 ± 18.6	21.9 ± 15.9	28.8 ± 17.5
Female
Before CS	31.2 ± 19.8	37.8 ± 21.5	25.0 ± 15.5	33.6 ± 18.8
During CS	27.4 ± 19.5	32.8 ± 22.1	22.4 ± 15.3	30.9 ± 18.7
After CS	29.0 ± 22.9	34.9 ± 25.2	20.5 ± 13.2	29.8 ± 18.6
Responders (n = 34)
Male
Before CS	24.2 ± 17.3	30.7 ± 18.0	26.4 ± 17.6	32.1 ± 20.4
During CS	21.1 ± 14.8	26.5 ± 19.4	22.7 ± 15.6	29.2 ± 20.3
After CS	22.3 ± 18.1	27.8 ± 19.9	20.8 ± 14.8	28.4 ± 16.9
Female
Before CS	33.3 ± 20.2	39.7 ± 21.2	26.2 ± 16.4	33.2 ± 18.6
During CS	28.0 ± 20.6	33.1 ± 22.1	24.0 ± 16.1	30.9 ± 19.9
After CS	29.6 ± 23.7	35.8 ± 25.7	20.5 ± 14.1	29.2 ± 19.5

Fig. 1

Conditioned pain modulation (CPM)-effects during painful and non-painful conditioning stimulation (CS) in all subjects (n = 40; upper panels) and in CPM-responders only (n = 34; lower panels). The statistical analysis showed a significant CPM-effect during CS across both sessions, genders and intensities (see Section 3.3.1), but no effect of gender. Values are mean ± SEM.

Subjects with large inhibitory effects during painful CS also exhibited large inhibitory effects during non-painful CS. This was tested by correlation analyses, which were significant for the high TS intensity (Pearson’s r =0.68, p < 0.001), but not for the low TS intensity (Pearson’s r = 0.22, p = 0.17).

Order was not a significant between-subjects factor neither for the low TS intensity (F(1,36) = 0.017, p = 0.897) nor for the high TS intensity (F(1,36) = 0.002, p = 0.967).

4 Gender differences on the CPM effect

In the RM-ANOVA assessing the effect of session and block on TS pain ratings, gender was not a significant factor (F(1,38) = 0.45, p = 0.5). To determine whether non-responders contributed to the negative effect on gender, the analyses were repeated on the 34 CPM-responders. However, gender was still not a significant factor in the analysis (F(1,32) = 0.3, p = 0.6).

An additional comparison of the CPM effect between men and women was made by performing independent t-tests on the CPM effect (during CS minus before CS) for the session with painful CS. There was no difference between men and women for the low TS intensity (t(38) = 0.95, p = 0.35) or for the high TS intensity (t(38) = 0.21, p = 0.84) when all subjects were included or when only CPM-responders were included (low TS intensity: t(32) = 1.0, p = 0.31; high TS intensity: t(32) t = 0.23, p = 0.82). Taken together, this indicates that the CPM effect in the present study was not different in men and women.

3.5 Correlations between the CPM effect, pain-50 and electrical pain thresholds

Correlations between the CPM effect, the conditioning stimulus (pain-50) and electrical pain thresholds were tested on the 34 CPM-responders. No significant correlations were found between the CPM effect and pain-50 (Spearman’s rho < 0.179, p >0.31) or between the CPM effect and the electrical PT (rho < 0.282, p < 0.106).

4.1 Gender differences in pain modulation

Differences in activation of endogenous analgesic systems in men and women have been investigated in several studies during the last decade. Recently two reviews were published. Popescu et al. [13] performed a meta-analysis of 17 articles, quantifying the CPM effect in each study. They report that CPM (or DNIC) was 1.7 times less efficient in females than in males. Van Wijk and Veld-huijzen [26] report that seven studies demonstrated more efficient pain inhibition in men than in women and six studies showed no gender difference.

Two previous studies on gender differences controlled for the menstrual cycle, testing women in the follicular phase. Serrao et al. [14] found that ice water induced greater inhibition of the nociceptive withdrawal reflex in men than in women. Baad-Hansen et al. [20] did not find any difference between men and women when ice water inhibited ratings of painful intraoral capsaicin. Different definitions of the follicular phase may explain the divergent findings. The women were tested during days 8–10 of the menstrual cycle in the study by Serrao and co-workers, while they were tested during days 3–9 in the study by Baad-Hansen and co-workers, both studies defining the respective time window as the follicular phase. However, methodological variations such as the duration and strength of the applied stimuli and the anatomical region tested make it difficult to generalize across studies of pain inhibitory mechanisms [10].

The women participating in the present experiment were tested during days 12–14 relative to the self-reported first day of the last menstruation. This particular interval was selected since Tousignant-Laflamme and Marchand [22] found that women during this interval have comparable CPM efficacy to men. Since no confirmatory hormonal or temperature measurements were performed we cannot be certain that all the women were indeed tested in the ovulatory phase. Although the start date of the last menstrual period can be reliably obtained from self-report [27], it is difficult to exclude irregular cycling base on only self-report [28]. Still, it seems reasonable to take the present findings as support for the conclusion posed by Tousignant-Laflamme and Marchand [22]; that during certain periods of the menstrual cycle, the efficacy of the descending inhibitory systems is as effective in women as in men.

4.2 The effect of painful conditioning stimulation on muscle pain

The mean inhibitory CPM-effects on experimentally induced muscle pain during painful CS in the present study was somewhat lower compared to other studies that tested painful heat as CS in combination with an electrically induced test stimulus. In the present study, the maximum CPM-effect was ca 20% (obtained when only responders were analyzed on high TS intensity scores). CPM effects of 80% and 43%, respectively, were obtained in two studies measuring the electrocutaneous pain threshold [29,30]. The pain intensity (VAS) during electrical tooth stimulation was inhibited by 33% [31].

An argument for selecting a muscular test stimulus in the present study was that data from animal studies indicate that endogenous inhibitory systems are likely to influence muscle nociception to a stronger degree than skin nociception [24]. Comparing the present data with data from the above mentioned studies indicate that endogenous inhibitory systems influence skin nociception more strongly than muscle nociception. A recent study showed that the pressure pain threshold from the tibialis anterior was inhibited by 66.3% during cold pressor pain [32]. This show that weak inhibition of muscle nociception is not a general phenomenon and indicate that differences in outcome measures, body locations and experimental procedures may have a strong influence on these divergent findings. A paired study during which both muscular and skin nociceptive test stimuli are tested with the same conditioning stimulus is necessary in order to shed more light on this issue.

It has been uncommon to test several TS intensities in CPM studies. The two TS intensities tested in the present study were both perceived as mildly painful. It is possible that stronger CPM effects would have been detected if stronger TS intensities had been used. Oono et al. [31], who tested three TS intensities (mean ratings: 30, 41 and 59 of 100) reported inhibitory CPM effects for the two strongest TS intensities, but not for the weakest. There was also a tendency towards stronger CPM effect for the high TS intensity vs. the low TS intensity in the present study. Taken together, these and previous findings indicate that the CPM effect depends on TS intensity as well as on CS intensity.

Three minutes after CS the CPM effect was reduced, although it was still significant. This is in accordance with several previous studies [31,33,34,35] and indicates that activation of descending pain inhibitory mechanisms by a painful stimulus depends on an ongoing noxious input.

4.3 Nonspecific inhibitory effects

The present experiment demonstrated an inhibitory effect on TS also in the control session when a non-painful CS was applied. Although weaker than in the test session, this indicates that pain inhibitory mechanisms were activated by factors not related to nociceptive systems. Previous studies have shown inhibitory effects during non-painful heat (≈42 °C) [19,29,30,36,37] or during vibration [9,38].

In the present control session (non-painful CS) the thermode temperature remained at baseline level (35 °C) throughout the session (before, during, and after CS). Since no CS was actually given, pain inhibition activated by non-painful somatosensory input seems to be an unlikely explanation. A recent study found that habituation produced inhibitory effects of similar magnitude as a non-painful conditioning stimulus [39]. Habituation may explain some of the present effects. However, a higher TS responses ‘after CS’ than ‘during CS’ (for the high TS intensity) speaks against this explanation. A second explanation lays in the instructions given to the subjects. Before the second block in both sessions, subjects were told that they should expect that the forearm thermode temperature would increase to the pain-50 intensity. The wording was: “In the first series you will receive only electrical stimulations. In the second series you will receive electrical stimulations and heat stimulation at the pain-50 intensity. In the third series you will receive only electrical stimulations”. Since pain-50 occurred in the test session only and not in the control session, one may speculate that the mere anticipation of a painful heat to the forearm resulted in inhibition of the TS [40,41]. Distraction from the TS while anticipating the noxious heat may also have contributed to reducing TS pain ratings [42].

4.4 Gender differences in pain perception

The present data indicated that women were more sensitive than men to suprathreshold heat pain. Increased pain perception in women is in accordance with several previous studies and seems to hold across several pain modalities [1,15,18,43,44,45,46]. Whereas most previous studies are based on pain threshold measurement, the present findings extend these findings by showing increased pain sensitivity in women also for suprathreshold stimuli.

There was no differences between men and women in pain perception when comparing ratings of suprathreshold intramuscular stimulation (1.1 × PT and 1.6 × PT). Neither was there a difference when comparing intramuscular electrical PT, although women showed a tendency towards lower PT than men. In previous studies using electrocutaneous stimulation women reported greater pain than men [14,45]. The differences between the present study and the latter two studies may depend on type of stimulated tissue (muscle vs. skin).

4.5 Conclusion

The present findings indicate that painful contact heat inhibits electrically induced muscle pain and that inhibition was not different between men and women when women were tested in the time interval 12–14 days after their last menstruation. This interval corresponds to the ovulatory phase of the menstrual cycle. A weaker inhibition of muscle pain was seen during non-painful CS, indicating that nonspecific inhibitory effects were triggered. Also the nonspecific inhibitory effects were similar in men and women.