Results

Out of 7,568 participants, 5,975 (77.6%) were asked to and completed the phone interview assessing Criterion A traumatic experiences; of these, 272 individuals had missing survey data, resulting in an analytic data set of 5,703. Of these, 3,086 were monozygotic, 2,126 dizygotic, and 491with undetermined zygosity. Consistent with the VET Registry sample, participants were middle-aged (mean = 61.1 years, range 53-73), mostly White (93.3%), and non-Hispanic (96.9%). Most were married or widowed (79.9%) and most (71.3%) had completed at least some college education.

Of the full sample of 5,703 Veterans, 2,553 (44.8%) individuals reported at least one COPC, making up the chronic pain subsample (Table 1). 3,015 (52.9%) individuals of the full sample, and 1,581 (61.9%) of the chronic pain subsample respectively, reported experiencing at least one traumatic event consistent with PTSD criterion A of the DSM-IV.4American Psychiatric Association: Diagnostic and statistical manual of mental disorders. Within the full sample and the chronic pain subsample, individuals with a history of Criterion A trauma reported significantly higher levels of symptoms across all PCL and MPQ scales compared to individuals without a reported history of Criterion A trauma (t-test p'sd effect size values (Table 1) ranged from d = 0.43 for PPI to d = 0.70 for PPI-total, in the chronic pain subsample, suggesting a small to moderate level of practical significance, with most comparisons in the small effect size range.15Statistical power analysis for the behavioral sciences. Highest levels of symptoms were reported by individuals with a history of trauma and at least one COPC. However, individuals without a reported history of Criterion A trauma also reported symptoms across nearly the entire range of PCL scores (Figure 1) (range = 17 to 83) though with less variability.

Table 1PTSD-like symptoms (PCL) and pain indicators (MPQ) across level of trauma.

Note: PCL-Total = PTSD Checklist total score; MPQ-Total = Short Form McGill Pain Questionnaire total score; VAS = MPQ Visual Analogue Scale; PPI = MPQ Present Pain Intensity; SD = standard deviation; p = probability value; d = Cohen's d = (M1 - M2)/SD; small effect d = 0.20, medium effect d = 0.50, large effect d = 0.80. Mean levels compared using Welch two sample t-tests.

Figure 1Density plot of log-transformed PTSD Checklist (PCL) standard scores for individuals with and without trauma in the full sample.

Correlations between PCL total score and MPQ total and subscales across trauma status are presented in Table 2 for the full sample and the chronic pain subsample. The correlations for the full sample (r's = 0.48-0.62, pr's = 0.48-0.62, pr's = 0.39-0.54, pr's = 0.41-0.60, pr's = 0.42-0.59, pr's = 0.30-0.54, p

Table 2Correlations between PCL total score and MPQ and significance of interaction between trauma and PCL symptoms, controlling for age.

Note: PCL-T = PTSD Checklist total score; MPQ-Total = Short Form McGill Pain Questionnaire total score; VAS = MPQ Visual Analogue Scale; PPI = MPQ Present Pain Intensity; r = Pearson product moment correlation coefficient; β = standardized regression coefficient (beta weight); p = probability value.

Using mixed-effects regression models, we evaluated the interaction of trauma with PCL total score for each of the MPQ scales. There was no significant interaction between trauma and MPQ total score (Figure 2A). Interactions were significant for VAS (pp=.01), suggesting that the relationship between PCL and the MPQ visual analog scale and Present Pain Intensity scales varied across presence of trauma (see Figure 2B and 2C) though these differences were exceedingly small. Participants with a reported history of trauma endorsed more VAS and PPI symptoms at lower levels of the PCL but the difference in reported pain between individuals with and without a reported history of trauma was no longer present at higher levels of PCL symptoms. None of the interactions of trauma with PCL total score for the MPQ scales were significant in the chronic pain subsample, suggesting that in individuals with chronic pain conditions, the association between PCL and MPQ scales does not vary across history of trauma.

Figure 2Interaction plots and 95% confidence intervals for full sample. Models controlled for age and correlated nature of the twin data. Presented for significant interaction only (p ≤ 0.05). MPQ-Total = Short Form McGill Pain Questionnaire total score; MPQ VAS = MPQ Visual Analogue Scale; MPQ PPI = MPQ Present Pain Intensity; PCL-Total = PTSD Checklist total score; PCL-B = PTSD Re-experiencing; PCL-C = PTSD Avoidance; PCL-D = PTSD Arousal.

Correlations between MPQ Total score and PCL symptom clusters across trauma status are presented in Table 3 for the full sample and the chronic pain subsample. Within the full sample, the largest association was between MPQ Total score and PCL Arousal symptoms (r=0.60, pp'sp's

Table 3Correlation between MPQ total score and PCL symptom clusters across presence of trauma, and significance of interaction between trauma and MPQ total score, controlling for age.

Note: MPQ-T = Short Form McGill Pain Questionnaire total score; PCL = PTSD Checklist; Re-experiencing = PCL Re-experiencing symptom cluster, Avoidance = PCL Avoidance symptom cluster; Arousal = PCL Arousal symptom cluster; r = Pearson product moment correlation coefficient; β = standardized regression coefficient (beta weight); p = probability value.

All interactions between MPQ Total score and trauma history in the full sample were significant, suggesting that the relationship between MPQ and PCL symptom clusters varied across presence of trauma. Participants with a reported history of trauma reported more PCL Re-experiencing (Figure 2D), Avoidance (Figure 2E), and Arousal (Figure 2F) symptoms and the association of these symptoms with pain intensity were greater in participants with trauma. As in the full sample, in the chronic pain subsample, all interactions between MPQ Total score and trauma history were significant, suggesting that the relationship between MPQ and PCL symptom clusters varied across presence of reported trauma.DiscussionWe evaluated the relationship between multiple pain indicators and PTSD-like symptoms across reported history of trauma and history of COPCs. As expected, individuals with a reported history of Criterion A trauma reported more PTSD-like symptoms and a higher level of pain symptoms, with the highest level of symptoms reported by those with a history of trauma and COPCs. Interaction analyses showed that some aspects of the relationship between pain intensity and PTSD-like symptoms varied across history of trauma in the full sample, and the chronic pain subsample. In line with previous research, 6Arguelles LM Afari N Buchwald DS Clauw DJ Furner S Goldberg J. A twin study of posttraumatic stress disorder symptoms and chronic widespread pain.,18Fishbain DA Pulikal A Lewis JE Gao J. Chronic pain types differ in their reported prevalence of post-traumatic stress disorder (PTSD) and there is consistent evidence that chronic pain is associated with PTSD: an evidence-based structured systematic review. our results revealed a robust association between multiple indicators of pain and PTSD-like symptoms. While attenuated, the relationship between pain indicators and PTSD-like symptoms remained strong and significant for individuals with no reported Criterion A trauma history within the full sample and the chronic pain subsample, suggesting that the relationship between pain indicators and PTSD-like symptoms may not exclusively depend on a history of traumatic experiences. Similarly, the association between overall pain intensity and PTSD-like symptom clusters remained significant even in the absence of reported Criterion A trauma in both, the full and the chronic pain subsample, suggesting that while both conditions may be initiated or exacerbated by traumatic events, these seem neither necessary nor sufficient to the development of pain or the PTSD-like symptoms and may be triggered by stressful life events.As hypothesized, we found that the pain and PTSD-like symptoms relationship was maintained independent of reported Criterion A trauma history. Our results are in line with previous studies including a cross-sectional study showing that combat injury and PTSD were independent predictors of headache in 308 Iraq and Afghanistan Veterans,1Afari N Harder LH Madra NJ Heppner PS Moeller-Bertram T King C Baker DG. PTSD, combat injury, and headache in veterans returning from Iraq/Afghanistan. as well as a more recent cross-sectional study of 202 chronic pain patients showing that PTSD symptoms partially mediated the relationship between trauma exposure and multiple indicators of pain severity. 37McKernan LC Johnson BN Reynolds WS Williams DA Cheavens JS Dmochowski RR Crofford LJ. Posttraumatic stress disorder in interstitial cystitis/bladder pain syndrome: Relationship to patient phenotype and clinical practice implications. Another recent study found that physical symptoms including pain were prospectively related to PTSD symptoms in Veterans returning from deployment, even without a traumatic physical injury.36McAndrew LM Lu S-E Phillips LA Maestro K Quigley KS. Mutual maintenance of PTSD and physical symptoms for Veterans returning from deployment. Together, these findings suggest that the association between pain indicators and PTSD-like symptoms is not contingent on Criterion A exposure. It is also important to note that the attenuation of associations across trauma status may be due to the reduced variability of PTSD-like symptoms in the trauma group, which is likely to decrease the strength of associations with pain indicators in these individuals. This reduction in variability may be the sole source of differences between the trauma and no-trauma groups in our results and lends further support that the relationship between PTSD-like symptoms and pain is likely independent of Criterion A history. Given the experience of trauma is subjective, a variety of stressful life events may be perceived as traumatic and potentially serve as triggers for the development of both sets of symptoms in someone with other existing vulnerability. 31, 33, 50, 52Therefore, it is possible that the pain and PTSD-like symptoms may occur after stressful life events and not just criterion A trauma.

Our results also showed that some of the interplay between pain intensity and PTSD-like symptoms varied across history of trauma in the full sample, and the chronic pain subsample, but that these differences were small. The interaction between PTSD-like symptoms and trauma history were significant but small for the MPQ total score and the current pain ratings and not for the affective or the sensory symptoms, which evaluated the specific aspects of the pain experience. Within the chronic pain subsample, none of the interaction terms between PTSD-like symptoms and trauma history were significant, suggesting that in those with chronic pain conditions, the relationship between pain intensity and PTSD-like symptoms did not vary across history of trauma. This attenuation in the chronic pain sample could be due to the decreased range of pain intensity scores and PTSD-like symptom scores in individuals with chronic pain. Evaluation of the interaction of trauma and pain intensity total scores however, revealed that the relationship between the total pain intensity scores and PTSD-like symptom clusters (B, C, D) did vary across trauma exposure in the full sample as well as in the chronic pain subsample. Taken together, these results show that the effect of PTSD-like symptoms on pain intensity scales does not vary in a meaningful way across trauma and that relationship between pain intensity and PTSD-like symptoms is largely independent of trauma history as defined by Criterion A.

The finding that trauma did not moderate the relationship between PTSD-like symptoms and some of the pain intensity symptoms may also reflect the fact that Veterans in the trauma group reported trauma consistent with Criterion A, while Veterans in the non-trauma group most likely experienced a significant stressful life event that results in PTSD-like symptoms. As a result, our findings suggest that the relationship between PTSD-like symptoms and pain is maintained both in the context of trauma or events that are sufficiently stressful to cause PTSD-like symptoms, in the absence of Criterion A traumatic events. Because the trauma evaluated in this study could have occurred at any time (pre, during, or post deployment) we are not able to directly evaluate the trajectory of the trauma-pain relationship. It is worth noting, however, that the current relationship between PTSD-like symptoms and pain intensity symptoms may have endured because of trauma occurring many decades prior in combat. Even though the association between PTSD-like symptoms and pain intensity was maintained in the absence of trauma, trauma does appear to make some contribution to the overall relationship between PTSD-like symptoms and measures of pain intensity. Future research should continue to examine the different qualities of the pain experience and its association with psychopathology, and the role of traumatic experiences on this relationship.

Our results also demonstrate that endorsement of PTSD-like symptoms can occur in the absence of Criterion A trauma history required for a diagnosis of PTSD. Furthermore, symptoms comprising the three PTSD symptom clusters, including re-experiencing and avoidance symptoms, were present in individuals with and without Criterion A, and the association between these PTSD-like symptoms and pain indicators was robust and maintained across all three clusters. The PCL assesses PTSD symptom information from a potentially wide range of stressful life experiences. Thus, re-experiencing symptoms (e.g., intrusive memories, feelings) of PTSD are presented in the context of “stressful experiences” rather than Criterion A trauma as are several of the PCL items targeting PTSD avoidance symptoms (e.g., avoiding thinking or talking about stressful experience or specific activities). The remaining avoidance items (e.g., emotional numbing, social isolation), and arousal items (e.g., irritability, jumpiness) of the PCL are presented without reference to stressful experiences and may represent a measure of general distress including sleep difficulty, anger, irritability, difficulty concentrating, alertness, and jumpiness. While hyperarousal and avoidance symptoms are shared by other anxiety disorders, reexperiencing symptoms are unique to PTSD. Our finding suggests that the association between pain indicators and PTSD-like symptoms is likely not limited only to symptoms characterized by general distress, such as social isolation or irritability, but may instead reflect the emergence of key PTSD-like symptoms in response to stressful life events and less-severe traumatic experiences.

Ours is not the first study to report such findings 47Posttraumatic stress disorder: An empirical evaluation of core assumptions. and our results are consistent with Franklin et al. who recently reported no difference in PCL symptom severity or clinically-indicated level of PTSD in a sample of outpatient Veterans with and without Criterion A trauma.19Franklin CL Raines AM Hurlocker MC. No trauma, no problem: Symptoms of posttraumatic stress in the absence of a criterion a stressor. Presence of PTSD symptoms without Criterion A has also been documented in civilians who reported greater symptom severity subsequent to stressful life events such as the death or illness of a loved one rather than a Criterion A trauma 21Gold SD Marx BP Soler-Baillo JM Sloan DM. Is life stress more traumatic than traumatic stress?. as well as a sample of women Veterans in which stressful life events were associated with both pain phenotypes and PTSD symptoms. 33Lehavot K Goldberg SB Chen JA Katon JG Glass JE Fortney JC Simpson TL Schnurr PP. Do trauma type, stressful life events, and social support explain women veterans’ high prevalence of PTSD?.What, then, are PTSD-like symptoms without trauma? What does it mean for someone to endorse items on the PCL at a level predictive of a PTSD diagnosis, without meeting Criterion A? Several possibilities emerge. First, these PTSD-like symptoms may reflect general distress independent of trauma, consistent with overall demoralization and general vulnerability to psychopathology. Additionally, it is possible that the relationship between pain intensity and PTSD-like symptoms may not result from the trauma itself, but rather reflect an association with the psychological distress characteristic of PTSD and as well as other internalizing psychopathology, such as anxiety or depression disorders. Given the fact that over half of the symptoms on the PCL make a specific reference to “stressful experiences” it is unlikely, however, that all individuals endorsing PTSD-like symptoms in the clinical range are reporting general distress. This is further supported by the presence of reexperiencing items in individuals with no Criterion A trauma suggesting that symptoms specific to PTSD were also present without life-time Criterion A trauma. Alternatively, PTSD-like symptoms in the absence of Criterion A may be symptoms consistent with a post-traumatic reaction but present in response to events that failed to meet the threshold set by Criterion A. If the latter is true, this may suggest that Criterion A is not necessary for the emergence of PTSD-like symptoms and that various stressful life events may elicit similar lasting response, as is consistent with a growing body of research. 19Franklin CL Raines AM Hurlocker MC. No trauma, no problem: Symptoms of posttraumatic stress in the absence of a criterion a stressor.,25Hyland P Karatzias T Shevlin M McElroy E Ben-Ezra M Cloitre M Brewin CR. Does requiring trauma exposure affect rates of ICD-11 PTSD and complex PTSD?.,32Did the DSM-5 improve the traumatic stressor criterion?: association of DSM-IV and DSM-5 criterion a with posttraumatic stress disorder symptoms. This does not suggest that stressful life events that do not provoke feelings of fear or threat contribute to the emergence of PTSD-like symptoms, but that individuals who react with fear, threat to survival, or horror to events below the Criterion A threshold may develop PTSD-like symptoms.While broadening the definition of trauma further stands to erode the integrity of the diagnosis and the ability to identify posttraumatic mechanisms that differentiate PTSD from other psychopathology46Traumatic Events, Criterion Creep, and the Creation of Pretraumatic Stress Disorder.,47Posttraumatic stress disorder: An empirical evaluation of core assumptions., our current findings suggest that closer attention to the role of stressful life events in the context of both PTSD-like symptoms as well as chronic pain is warranted in both the research and clinical context. Specifically, a better understanding of whether PTSD symptoms in the absence of Criterion A trauma are conceptually distinct from PTSD symptoms in the presence of Criterion A trauma is important to clarify the relationship between trauma and PTSD-like symptoms. Research has shown that despite similar overall PTSD symptom severity, individuals with Criterion A endorse more PTSD-specific items, than those without.19Franklin CL Raines AM Hurlocker MC. No trauma, no problem: Symptoms of posttraumatic stress in the absence of a criterion a stressor. Replicating these findings in non-clinical or civilian samples would help to identify symptoms with high discriminant validity. Further research into whether individuals with PTSD symptoms with and without Criterion A trauma differ in their course, prognosis, treatment response, or outcomes and how these potential trajectories relate to the development and maintenance of chronic pain are important targets for future studies in both Veteran and civilian samples of men and women.Our results suggest that the relationship between pain and PTSD-like symptoms extends well beyond the clinical threshold represented by a DSM PTSD diagnosis, and should be the focus of ongoing research and clinical efforts. Because psychological responses to traumatic experiences can take many forms, future studies on the association between pain and trauma should consider this relationship both in the context of general emotional disturbance and across a wider range of psychopathology, in addition to PTSD, including other mood and anxiety disorders. Given the subjective nature of trauma, a variety of stressful life events such as divorce, illness, moving, unemployment, poverty may serve as triggers for the development of both pain and PTSD-like symptoms in someone with existing vulnerability. Even without a PTSD diagnosis, these PTSD-like symptoms may make a clinically meaningful impact and assessing and treating PTSD-like symptoms among chronic pain patients, regardless of Criterion A history may encourage integrative treatments and lead to better treatment outcomes. Because individual differences may affect the individual severity of trauma sufficient for a traumatic response, future research into individual differences such as anxiety sensitivity8Asmundson GJ Coons MJ Taylor S Katz J. PTSD and the experience of pain: Research and clinical implications of shared vulnerability and mutual maintenance models. as a predisposing factor contributing to the development of both chronic pain and PTSD is encouraged.Chronic pain and PTSD share many physiological, cognitive, and behavioral characteristics. For example, both chronic pain and PTSD are characterized by biases in attention towards threatening stimuli, appraisal tendencies, heightened startle reaction, trait fear,44Ploghaus A Narain C Beckmann CF Clare S Bantick S Wise R Matthews PM Rawlins JNP Tracey I Exacerbation of pain by anxiety is associated with activity in a hippocampal network. hypervigilance, emotional numbing, avoidance,51Fear-avoidance and its consequences in chronic musculoskeletal pain: A state of the art. and stress response dysregulation 8Asmundson GJ Coons MJ Taylor S Katz J. PTSD and the experience of pain: Research and clinical implications of shared vulnerability and mutual maintenance models. suggesting that the two conditions share similar features. Negative affect and anxiety vulnerability have also been linked to chronic pain conditions and PTSD, 16Cottam WJ Condon L Alshuft H Reckziegel D Auer DP. Associations of limbic-affective brain activity and severity of ongoing chronic arthritis pain are explained by trait anxiety.,27Kain ZN Mayes LC Caldwell-Andrews AA Karas DE McClain BC. Preoperative anxiety, postoperative pain, and behavioral recovery in young children undergoing surgery.,44Ploghaus A Narain C Beckmann CF Clare S Bantick S Wise R Matthews PM Rawlins JNP Tracey I Exacerbation of pain by anxiety is associated with activity in a hippocampal network. suggesting that differences in personality features that predispose individuals to chronic pain and PTSD symptoms may be partially responsible for their comorbidity.12Beckham JC Crawford AL Feldman ME Kirby AC Hertzberg MA Davidson J Moore SD. Chronic posttraumatic stress disorder and chronic pain in Vietnam combat veterans. The role of acute central sensitization, a heightened response in neurons and circuits in nociceptive pathways, has also been proposed as mediating links between pain and PTSD.42Moeller-Bertram T Strigo IA Simmons AN Schilling JM Patel P Baker DG. Evidence for acute central sensitization to prolonged experimental pain in posttraumatic stress disorder. Furthermore, findings from imaging studies have independently established connections between chronic pain and PTSD with activity in the hypothalamic-pituitary-adrenal axis, the amygdala,17Functional neuroimaging of anxiety: a meta-analysis of emotional processing in PTSD, social anxiety disorder, and specific phobia.,26Jiang Y Oathes D Hush J Darnall B Charvat M Mackey S Etkin A. Perturbed connectivity of the amygdala and its subregions with the central executive and default mode networks in chronic pain. as well as structural and functional alterations in the anterior cingulate cortex, a brain region involved in attention and emotion.13Bliss TV Collingridge GL Kaang B-K Zhuo M. Synaptic plasticity in the anterior cingulate cortex in acute and chronic pain.,28Kennis M Rademaker AR van Rooij SJ Kahn RS Geuze E. Resting state functional connectivity of the anterior cingulate cortex in veterans with and without post-traumatic stress disorder.,56Yoshino A Okamoto Y Onoda K Yoshimura S Kunisato Y Demoto Y Okada G Yamawaki S. Sadness enhances the experience of pain via neural activation in the anterior cingulate cortex and amygdala: An fMRI study.This study had several limitations. The use of self-report data for pain intensity and PTSD-like symptom burden assessment may have led to underreporting of symptoms. Despite this, overall prevalence and correlations are consistent with previous studies suggesting that our data are representative.14Brennstuhl MJ Tarquinio C Montel S. Chronic Pain and PTSD: Evolving Views on Their Comorbidity. Additionally, our results are based on symptoms measures and self-reported physician diagnosis of COPCs which may not adequately capture their chronicity, or the length of time someone has had the condition. Research involving the history and duration of chronic pain may be beneficial. While the total sample and the chronic pain subsample did overlap, constructing our groups in this way allowed us to evaluate the impact of COPCs on the relationship between PCL and MPQ. Future research should examine these associations in participants meeting the diagnostic criteria for chronic pain conditions and PTSD and fully examine the potential confounding role of other internalizing disorders such as major depression. Given that we were specifically interested in any history of trauma, we did not examine the impact of multiple traumatic events or types of trauma exposure. It is worth noting that the restricted variability of PCL scores in individuals without a history of Criterion A trauma may account for the reductions in the correlations we report between PCL and MPQ scores rather than any meaningful contribution of trauma history to the relationship. Additional research to examine the potential impact of types and quantity of traumatic and stressful life events on the relationship between pain and PTSD could further our understanding of the pain-PTSD interplay.

Our sample consisted of male Veteran participants, and future research should expand these findings to women, evaluate these relationships in female combat Veterans, as well as representative community samples. Although chronic pain and PTSD are more prevalent in women, they are nonetheless highly prevalent and problematic in men, making the current evaluation important in improving our understanding of their relationship. Our overall sample contained individuals with chronic pain conditions and allowed our results to remain more generalizable than a chronic pain-free sample. The current study does not account for predisposing factors that may contribute to the development of either chronic pain or PTSD and prospective studies evaluating participants prior to trauma exposure may further elucidate the pain, trauma, and PTSD symptom interplay. Studies evaluating pre-deployment health in Veterans, for example, would be well-equipped to address this important question. Despite these limitations, this study substantially extends the work to date by specifically examining the role of trauma on the association between pain and PTSD-like symptoms.

Future research incorporating both chronic pain and PTSD together would be of great benefit. Using several of these measures collectively may represent an endophenotype for vulnerability for maladaptive stress response which may be expressed in either a physical chronic pain or anxiety and stress-related disorders such as PTSD. Similarly, the role of diverse types of traumatic and stressful experiences (for example, with and without physical injury) may further clarify the relationship between chronic pain, trauma, and PTSD. Finally, exploring these experiences in longitudinal samples to investigate temporal progression of these conditions and establish the chronicity of pain, is important to understanding of their interplay across the lifespan.

In summary, our findings suggest that while a reported history of Criterion A trauma may increase the presentation of both pain indicators and PTSD symptoms, the association between pain and PTSD-like symptoms is largely independent of Criterion A trauma. Future research should evaluate the relationship between pain and trauma across a wider range of psychopathology, including, but not limited to PTSD. Because of the complicated interplay between pain, psychopathology, and trauma, clinical intervention aimed at evaluating and addressing both physical and emotional distress in chronic pain patients is important in improving patient outcomes.