Abstract
Objective. Stressful life events are associated with increased pain severity and chronicity. However, the mechanism underlying this association remains disputed. Recent animal studies suggest that chronic stress increases pain sensitivity and persistence by enhancing peripheral and central sensitization mechanisms. To test this hypothesis in humans, the authors examined whether sensitization is enhanced in healthy women reporting more stressful life events using the topical capsaicin test.
Methods. Thirty-two healthy young women reporting varying levels of stressful life events were invited for laboratory pain testing. Capsaicin was applied topically to the volar forearm. Measurements included capsaicin-induced spontaneous pain and area of secondary hyperalgesia in the region surrounding capsaicin application. Physiological (heart rate and skin conductance) and self-reported affective (emotional valence and arousal) states were also measured.
Results. The results indicate that more stressful life events predicted a linear increase in the area of secondary hyperalgesia (β = 0.40, p = 0.023, R2 = 0.16), but not the intensity of secondary hyperalgesia nor capsaicin-induced spontaneous pain. These findings suggest that life stressors may be associated with heightened central sensitization manifested by an increased area of secondary hyperalgesia. Additionally, life stressors were related to greater sympathetic cardiac, but not to affective responses to capsaicin-induced pain.
Conclusion. This study shows that women reporting more stressful life events show a larger area of secondary mechanical hyperalgesia. These preliminary findings suggest that life stressors may facilitate pain processing by enhancing central sensitization.
Introduction
Multiple life stressors may predispose individuals to chronic pain disorders. Exposure to a stressor produces transient negative emotion and physiological alarm responses if one is able to adapt to the event through allostasis, the process of achieving stability through change [ 1 , 2 ]. However, chronic stress can lead to allostatic load, the wear-and-tear that develops due to repeated cycles of allostasis as well as the dysregulation of the allostatic systems that normally support adaptation [ 2 ]. Because pain is a stressor, individuals with multiple life stressors may respond differently to painful stimulation than those experiencing fewer stressors. A recent epidemiology study supports this hypothesis, reporting a high prevalence of chronic pain problems in individuals with biomarkers of high allostatic load (e.g., high levels of C-reactive protein and high systolic blood pressure) [ 3 ]. Previous studies also suggest life stressors are associated with increased pain sensitivity and chronicity [ 4–6 ]. One possible explanation for the association between life stressors and clinical pain is that chronic stress may increase pain severity and persistence by sensitizing the pain pathway [ 7 ].
The current pilot study investigated the link between stressful life events and capsaicin-induced experimental pain. Capsaicin produces spontaneous burning pain at the application site [ 8 , 9 ] and lower pain thresholds to mechanical stimulation in the hyperalgesic skin surrounding the application site (secondary hyperalgesia) [ 9 ]. Psychophysical and neurophysiological studies indicate that spontaneous pain is initiated by sensitization of the primary afferent nociceptors, whereas secondary hyperalgesia reflects central sensitization of neurons in the dorsal horn of the spinal cord and supraspinal pain pathways [ 9–11 ]. Importantly, central sensitization contributes to the induction and maintenance of chronic pain [ 9 ]. To avoid the potential confounds linked to ongoing clinical pain, experimental pain responses were examined in a nonclinical sample of women. Women were recruited because chronic pain [ 12 , 13 ] and psychological distress from adverse life events are more common among women [ 14 , 15 ]. We hypothesized that the number of stressful life events would predict enhanced pain sensitivity, as measured by the area and intensity of secondary hyperalgesia to mechanical stimuli, and capsaicin-induced spontaneous pain intensity and unpleasantness.
Methods
Participants
Female undergraduates were prescreened for health status ( n = 1244). Individuals reporting any psychiatric or physical health problems (38%) were ineligible to control for confounding conditions influencing pain sensitivity. To recruit individuals with varying levels of stressful life events, the Childhood Trauma Events Scale, and Recent Traumatic Events Scale were also administered in the prescreening [ 16 ]. The two scales evaluate stressful events encountered in lifetime such as death, divorce, violence, sexual abuse/assault, illness, or others prior to age 17 years (the former scale) and within the past 3 years (the latter scale). The events endorsed on both scales were summed to calculate individual stressful life event scores. Among the healthy individuals, 8% reported no trauma events and 13% reported 9 or more events, which corresponded to 1 SD above the mean. The individuals reporting no trauma or 1 SD above the mean were invited via email and those who had schedules compatible with our experiment times voluntarily signed up on the online recruitment system. These scales were re-administered in the laboratory to verify life event scores ( Table 1 ). Although we invited individuals with no or high trauma scores, verified trauma scores formed a range. Therefore, regression analysis was more appropriate for the data than the group mean comparison. Participants received the course credits for their participation.
Psychological traits and physiological state variables prior to pain testing ( n = 32)
| M | SD | Min | Max | |
|---|---|---|---|---|
| Psychological trait variables | ||||
| Life event scores | 9.2 | 7.2 | 0.0 | 25.0 |
| Psychological state variables | ||||
| PANAS-Negative Affect | 10.2 | 7.9 | 1.0 | 31.0 |
| Physiological state variables | ||||
| HR ( n = 27) | 77.4 | 9.0 | 62.0 | 96.0 |
| SCL | 6.9 | 2.3 | 4.6 | 14.1 |
| M | SD | Min | Max | |
|---|---|---|---|---|
| Psychological trait variables | ||||
| Life event scores | 9.2 | 7.2 | 0.0 | 25.0 |
| Psychological state variables | ||||
| PANAS-Negative Affect | 10.2 | 7.9 | 1.0 | 31.0 |
| Physiological state variables | ||||
| HR ( n = 27) | 77.4 | 9.0 | 62.0 | 96.0 |
| SCL | 6.9 | 2.3 | 4.6 | 14.1 |
Psychological traits and physiological state variables prior to pain testing ( n = 32)
| M | SD | Min | Max | |
|---|---|---|---|---|
| Psychological trait variables | ||||
| Life event scores | 9.2 | 7.2 | 0.0 | 25.0 |
| Psychological state variables | ||||
| PANAS-Negative Affect | 10.2 | 7.9 | 1.0 | 31.0 |
| Physiological state variables | ||||
| HR ( n = 27) | 77.4 | 9.0 | 62.0 | 96.0 |
| SCL | 6.9 | 2.3 | 4.6 | 14.1 |
| M | SD | Min | Max | |
|---|---|---|---|---|
| Psychological trait variables | ||||
| Life event scores | 9.2 | 7.2 | 0.0 | 25.0 |
| Psychological state variables | ||||
| PANAS-Negative Affect | 10.2 | 7.9 | 1.0 | 31.0 |
| Physiological state variables | ||||
| HR ( n = 27) | 77.4 | 9.0 | 62.0 | 96.0 |
| SCL | 6.9 | 2.3 | 4.6 | 14.1 |
Self-Reports
Using a Visual Analogue Scale, participants rated pain intensity (0: no pain to 10: most intense pain imaginable ) and unpleasantness (0: not at all unpleasant to 10: most unpleasant pain imaginable ). Participants’ affective responses to capsaicin-induced pain were assessed with the Self-Assessment Manikin (SAM). The SAM measures emotional valence (1: unhappy to 9: happy ) and arousal (1: calm to 9: excited ) [ 17 ]. Negative affect at baseline was evaluated with the Positive and Negative Affect Schedule (PANAS; 1: very slightly or not at all to 5: extremely ) [ 18 ].
Procedure
Texas A&M University Institutional Review Board approved the protocol. After completing the informed consent and questionnaires, heart rate (HR) and skin conductance level (SCL) sensors were attached to fingers of the non-dominant hand (for details, see [ 19 ]). Then, a grid with eight spokes radiating from the center of the capsaicin application site was drawn. Each spoke was 10cm in length and consisted of 10 stimulation sites spaced at 1cm intervals [ 20 ]. Next, a 0.3ml of a 6% capsaicin solution was topically applied for 30 minutes to the non-dominant volar forearm via a 1.5cm × 1.5cm gauze and covered with a dressing [ 21 ]. Following capsaicin application, participants rated spontaneous pain and the SAM at 5-minute intervals for the first 25 minutes. Pain ratings during capsaicin application were averaged to calculate each individual’s spontaneous pain intensity. After removing the capsaicin gauze, secondary hyperalgesia was evaluated. The area of secondary hyperalgesia was determined by applying a hand-held stiff 2.9N (300gF) von Frey filament (Stoelting, USA) starting in the area of normal sensitivity (furthest from the capsaicin site, referencing points), and sequentially stimulating the ten sites by moving towards the capsaicin site. The von Frey was applied 2s on each site until it bent with 2.9N. Participants rated pain after each von Frey application. Pain ratings to von Frey stimuli on the furthest points on all eight spokes were averaged to compute individual mechanical pain intensity. The boundaries of secondary hyperalgesia were defined as any change in pain ratings for a given site compared to the furthest site on the spoke [ 10 , 20 , 22 ]. The distance from a boundary point to the capsaicin application site was measured on each spoke to calculate the size of the hyperalgesic area (cm 2 ). Intensity of secondary hyperalgesia was calculated by subtracting subsequent pain ratings from a rating at the furthest point on each spoke and averaging the pain intensity change scores. In other words, intensity of secondary hyperalgesia reflects the average increase in pain intensity from the furthest points.
Data Analyses
For this pilot study, participants were recruited over a semester. Subsequently, effect sizes ( R2 ) were noted for future studies to calculate a priori sample size. Among the 33 participants who completed pain testing, one was removed from analysis as an outlier based on Cook’s D (= 0.59) [ 23 ]. Notably, this outlier had the highest stressful life event score with minimal pain response ( Figure 1 ).
Relationship between the number of stressful life events and the area of secondary hyperalgesia.
Results
Baseline Characteristics
Of 32 participants, the majority were Caucasian (81%). The average age was 18.7 years ( SD = 0.9). The baseline psychophysiological characteristics are shown in Table 1 . Stressful life event scores were not correlated with baseline PANAS-Negative Affect nor with baseline autonomic measures of HR and SCL ( p s > 0.488). To note, technical difficulties lead to missing HR data in five cases and pairwise deletion was used to maximize the sample size.
Secondary Hyperalgesia
Multivariate regression analysis was conducted to examine whether stressful life event scores would predict the two correlated outcome variables: area and intensity of secondary hyperalgesia (Pearson r = .49, p = .006). With the use of Wilks’ criterion, the overall model in predicting the two outcome variables from stressful life event scores approached significance (Wilks’ Λ = 0.81, F (2, 29) = 3.30, p = .051). Separate regression analysis revealed that more stressful life events predicted larger areas of secondary hyperalgesia (β = 0.40, SE β = 1.7, p = 0.023, R2 = 0.16), but did not predict the intensity of secondary hyperalgesia ( p = 0.771). Area and intensity of secondary hyperalgesia were on average 82.8cm 2 (range: 6.2–252.3 cm 2 ) and 0.6 (range: -0.5–6.4), respectively.
On average, participants reported mild pain ( M = 1.6, SD = 1.1) to mechanical stimulation on normal skin located 10cm from the capsaicin site and these intensity ratings on normal skin were unrelated to life event scores ( p = 0.748).
Capsaicin-Induced Spontaneous Pain
Multivariate regression analysis was conducted to examine whether stressful life event scores would predict the capsaicin-induced spontaneous pain intensity and unpleasantness (Pearson r = 0.76, p < 0.001). The two outcome variables were transformed with a square root for normalization. With the use of Wilks’ criterion, the overall model in predicting the two outcome variables from stressful life event scores was not significant ( p = 0.991). On average, capsaicin induced moderate pain intensity ( M = 6.3, SD = 2.1) and unpleasantness ( M = 6.9, SD = 2.2). Interestingly, capsaicin-induced spontaneous pain ratings were uncorrelated with the area and intensity of secondary hyperalgesia ( p s > 0.467).
Affective and Physiological Responses to Capsaicin
Multivariate regression analysis was conducted to examine whether stressful life event scores would predict affective responses such as perceived arousal and negative emotional valence to capsaicin (Pearson r = 0.46, p = 0.008). With the use of Wilks’ criterion, the overall model in predicting the two response variables from stressful life event scores was not significant ( p = 0.909). On average, participants reported moderate levels of emotional arousal ( M = 4.0, SD = 2.2) and negative affect ( M = 4.8, SD = 1.8) on the SAM during capsaicin application.
More stressful life events were associated with greater sympathetic cardiac responses to capsaicin-induced pain (ΔHR = HR at post - pre; Pearson r = 0.40, p = 0.046), which were not correlated with any pain outcome variables ( p s > 0.486). Stressful life events were unrelated to SCL changes by capsaicin-induced pain ( p s > 0.477).
Discussion
We tested the hypothesis that stressful life events would be associated with enhanced pain sensitivity using topical capsaicin to evoke peripheral and central sensitization. Women endorsing more life stressors showed larger areas of secondary hyperalgesia without changes in perceived pain intensity to capsaicin or to a mechanical stimulation. Expansion of the area constitutes an abnormal pain sensation that accompanies clinical pain syndromes (e.g., fibromyalgia, neuropathic pain) [ 9 , 24 ]. The expanded area reflects a central sensitization mechanism involving receptive field expansion that enables mechanical stimulation of non-injured skin to produce pain [ 10 , 24 ]. Consequently, life stressors may enhance pain sensitivity, and this heightening of pain sensitivity may lower the threshold for experiencing pain related symptoms and increase vulnerability for chronic pain conditions associated central sensitization. Indeed, stress may contribute to the transition from acute localized pain into chronic widespread pain by enhancing the spread of central sensitization at spinal segmental levels and higher brain regions [ 25 ] or to the exacerbation of chronic widespread pain [ 26 ].
Our findings are consistent with animal studies; repeated exposure to same or variant stressors in childhood or adulthood causes persistent hyperalgesia [ 27–30 ]. Early life stress leads to mild muscle hyperalgesia in adult rats and this stress-induced hyperalgesia is further exacerbated by sound stress, suggesting the incremental effect of stressors on pain sensitization [ 31 ]. In an animal model of chronic wide spread pain, the sympathoadrenal stress axis plays a key role in stress-induced persistent hyperalgesia [ 28 ]. Exposure to unpredictable sound stress over four days produces prolonged elevation of plasma epinephrine levels, which mediate the development and maintenance of persistent hyperalgesia. Similarly, human pain pathways may be sensitized by persistent activation of the sympathoadrenal axes in exposure to multiple life stressors.
Conversely, another laboratory has observed reduced temporal summation of heat pain in adults reporting childhood physical or sexual abuse, but no difference in pain threshold and tolerance [ 32 ]. Temporal summation reflects homosynaptic central sensitization and produces progressive increase in pain intensity to identical stimuli [ 9 ]. In contrast, similar to our findings, slower recovery of capsaicin-induced muscle hyperalgesia and greater temporal summation of muscle pain have been observed in combat veterans with PTSD, but no differences in pain thresholds [ 33 ]. Perhaps, hypersensitivity to pain may depend on the nature, frequency, and variant types of stress exposure, which may produce different levels of allostatic load and reduce the adaptive capacity to modulate central pain processing.
Our results suggest that stressful life event scores predicted a linear increase in the area of secondary hyperalgesia. However, the relationship may be non-linear when including extreme levels of life stress. Indeed, a previous study indicates that childhood abuse is associated with decreased experimental pain sensitivity [ 32 ]. Here, one individual reporting the highest life event score showed a minimal response in the area of secondary hyperalgesia. Because of small sample size, this single data point with high leverage, high discrepancy, and high influence was dropped to evaluate the overall pattern in predicting areas of secondary hyperalgesia from levels of life stressors. Therefore, additional research is needed to determine whether a linear or nonlinear pattern would be observed in a large sample that includes individuals reporting extremely high levels of stressful life events.
Central sensitization is induced and maintained by input from the peripheral afferents, and also modulated by supraspinal processes. Neuroimaging studies show that differences in brain activation patterns correspond to secondary hyperalgesia phenotypes [ 34 , 35 ]. Thus, psychological responses during (e.g., dissociation, catastrophizing) and after (e.g., anxiety and depression) exposure to stressful life events may contribute to changes in secondary hyperalgesia by altering descending modulation and supraspinal neural processing [ 36 ]. Recent support for this view comes from evidence that dissociation at the time of childhood trauma predicts the magnitude of central sensitization during adulthood on the temporal summation of second pain test (in review). In contrast, measures of current negative affect and the intensity of the affective responses at the time of trauma, such as fear, horror, or helplessness, were not predictive. In the present study we did not measure psychological traits, however we found that negative affect at the baseline (PANAS) and affective responses to capsaicin (SAM-valence and perceived arousal) were unrelated to stressful life events scores. Although this finding should be considered preliminary, it suggests that the effect of stressful events on central sensitization is not mediated by current negative affect and perceived arousal. Importantly, evidence indicates that secondary hyperalgesia can be reduced by written disclosure of trauma [ 19 ] and brief cognitive-behavioral therapy [ 37 ], suggesting that psychological interventions can attenuate central sensitization and may reduce vulnerability to persistent pain.
Finally, limitations of the current study should be acknowledged. Because participants were female college students, additional study is needed to examine the generality of the current findings to men and a community sample. Furthermore, the small sample size and unknown menstrual cycle may limit our ability to draw a definitive conclusion. We are currently conducting a replication study with a larger sample, both genders, assessment of the menstrual cycle for women participants, and an additional measure of central sensitization (i.e., temporal summation of second pain). Preliminary analyses of this ongoing study consistently indicate life stressors as being linked to pain hypersensitivity.
In conclusion, the current study suggests women with greater life stressors may be vulnerable to chronic pain conditions associated with central sensitization as manifested by increased areas of secondary hyperalgesia.
Acknowledgments
The authors wish to thank Jim Grau, Heather Bortfeld, and Holly Foster for their comments on earlier drafts of this manuscript.
Funding sources: The Texas A&M Department of Psychology and the College of Liberal Arts provided partial funding for equipment and personnel. This article was partly based on a dissertation by Suzannah K Creech who received a fellowship from the Texas A&M Department of Women Studies.
Conflicts of interest: The authors have no conficts of interest to disclose.
References
