Behavior And Neural Correlates Of Resilience To Pain

Resilience,which is a personality construct reflecting capacities to persevere,keep a positive perspective and/or thrive regardless of ongoing stressors,has emerged as an important focus of research on chronic pain(CP).Recent U.S.studies of musculoskeletal pain have supported the structure and construct validity of the Pain Resilience Scale(PRS)as a pain-specific measure tapping capacities to regulate cognitions and emotions as well as behavioral perseverance despite ongoing pain.However,it is not clear whether psychometric support for the PRS extends to chronic musculoskeletal pain samples in other countries or whether PRS scores contribute to adaptation beyond the impact of general resilience.Further,although behavior studies have found more resilient persons with CP experience less pain-related dysfunction than less resilient cohorts do,the presence and nature of associated brain structure and functional differences has received scant attention.Finally,researchers have not considered interactions of pain resilience with other psychological influences within experimental research.Regarding this last limitation,terror management theory(TMT)postulates that when human realize that death is inevitable,their thought,motivation,behavior and emotion would be influenced.Previous experiments have found manipulations of mortality salience(MS)awareness can have causal effects on reported pain and behavioral pain tolerance.However,to date,researchers have yet to examine whether MS has moderating effects on associations between trait-based psychological factors including pain resilience and pain outcomes.To address these gaps,we investigated(1)the factor structure(2)construct validity,and incremental validity of the PRS in independent samples of Chinese adults with chronic musculoskeletal pain via exploratory factor analysis(EFA)and confirmatory factor analysis(CFA).In addition,towards elucidating neural correlates of pain resilience,we evaluated(3)gray matter volume(GMV)differences between more versus less resilient adults with chronic musculoskeletal pain with voxel-based morphology analysis,(4)regions of interest(ROIs)in which resting state(Rs)brain activity discriminated more from less resilient CP subgroups based on multiple kernel learning(MKL),(5)pain resilience subgroup differences in brain activity during exposure to pain empathy and pain anticipation and(6)mediating effects of GMV and Re Ho on associations between baseline pain resilience and subsequent pain intensity and interference from pain in a chronic pain sample over six-month interval.Finally,(7)the extent to which MS modulated the associations between pain resilience and pain outcomes induced by a cold pressor test(CPT).In study 1,we tested the factor structure of PRS among Chinese adults with chronic musculoskeletal pain with EFA.The sample comprised Chinese adults(417women,134 men)who completed Connor-Davidson Resilience Scale–Chinese(CDRS-C),Pain Self-Efficacy Questionnaire(PSEQ),back-translated versions of the PRS,demographics and pain Characteristics.A maximum likelihood(ML)analysis with promax rotation was applied to run EFA in SPSS 20.0.EFA yielded a two-factor Chinese version of PRS(PRS-C)featuring a cognitive/affective positivity subscale with7 items and a behavioral perseverance subscale with 3 items.We also calculated Spearman correlation coefficients to evaluate convergent validity with conceptually-related measures including general resilience based on CDRS-C scores,pain self-efficacy and discriminant validity with conceptually less-related background characteristics including age,gender,marital status,Body Mass Index(BMI),education and pain duration within this sample.Significant positive correlations between two dimensions of PRS-C(i.e.cognitive/affective positivity and behavior perseverance)and general resilience as well as pain self-efficacy supported convergent validity while low or non-significant correlations of PRS-C facets with age,gender,education,BMI,marital status,number of pain sites and pain duration evidenced discriminant validity.In sum,study 1 supported a two component factor structure of the Pain Resilience Scale among Chinese adults with chronic musculoskeletal pain.The observed structure was similar to that from initiation validation research on U.S.samples.In study 2,CFA was used to identify a best fitting PRS structure in an independent Chinese chronic pain sample.Participants(421 women,135 men)completed the PRS-C,CDRS-C,PSEQ,Chronic Pain Acceptance Questionnaire-Activity Engagement(CPAQ-AE),Coping Strategies Questionnaire-Chinese(CSQ-C),Chronic Pain Grade(CPG),Center for Epidemiologic Studies Depression Scale(CES-D),Chinese Pain Catastrophizing Scale(C-PCS),and measures of background factors.Mplus 7.0 was used to test fits of the EFA-derived PRS-C structures in study 1 versus variations from factor analyses in previous studies versus unidimensional models.The two-factor,EFA-derived PRS-C containing a 7-item cognitive/affective positivity component and a3-item behavioral perseverance component showed the best overall fit from several hypothesized alternatives.Regarding construct validity,PRS-C facets displayed significant moderate correlations with conceptually-related measures(i.e.,pain self-efficacy,general resilience,cognitive coping,activity engagement,pain severity,pain disability,pain catastrophizing and depression)and low correlations with conceptually less-related background characteristics(i.e.,pain duration,number of pain sites,gender,BMI and education).Although statistically significant associations were found with marital status,age and BMI,all corresponding strengths of relation were small in magnitude.Hierarchical multiple regression analyses evaluating the incremental validity of the PRS-C subscales indicated cognitive/affective positivity,in particular,accounted for significant unique variance in predicting pain severity,pain impairment,depression and pain self-efficacy,independent of general resilience and significant background measures.Overall,study 2 confirmed the two component factor structure of Chinese version of Pain Resilience Scale among Chinese adults with chronic musculoskeletal pain derived from study 1 and provided unambiguous incremental validity support for the PRS-C dimensions—particularly positivity in predicting adjustment to chronic pain independent of a widely-used measure of general resilience.Thus,the evaluation of pain resilience has potential utility to the identify more versus less vulnerable chronic pain subgroups as well as assessment and development of various treatment plans for chronic musculoskeletal pain within a Chinese context.In study 3 to study 5,we elucidated differences of neural structure and activity between more and less resilient participants with chronic musculoskeletal pain.The sample derived from an initial sample of community-dwelling adults with ongoing pain lasting three months or longer(103 women and 59 men).Upon arriving at their scheduled appointments,the participants experienced informed consent,selection criteria,a demographics form,magnetic resonance imaging(MRI)scan,self-report measures(CDRS-C,PRS-C,CPG,PCS-C and CES-D)in order.More and less resilient subgroup,respectively,were identified on the basis of scoring above and below median scores on two validated resilience questionnaires,the CDRS-C and PRS-C.Resilience subgroup differences on demographics(e.g.age,gender,BMI,ethnicity,relationship status,education,recruitment setting,occupational status),pain characteristics(pain duration,primary pain site,presence of other pain sites,prescription analgesics use),pain severity,pain catastrophizing and depression were also assessed in preliminary analyses using SPSS and controlled for in main analyses,if necessary.Study 3 was designed to examine the presence and nature of regional GMV differences between more versus less resilient subgroups of a large CP sample.The final sample comprised 75 women and 43 men(57 in higher resilience group and 61 in lower resilience group).Image preprocessing and Voxel-based morphology(VBM)analysis was conducted in Statistical Parametric Mapping(SPM).After controlling from resilience subgroup differences on background measures(i.e.gender,use of prescription analgesics,pain severity,depression and pain catastrophizing)as well as age and total intracranial volume(TIV),more resilient participants were observed to have significantly larger GMV in the(1)bilateral precuneus,(2)left superior and inferior parietal lobules,(3)orbital right middle frontal gyrus and medial right superior frontal gyrus,and(4)bilateral median cingulate and paracingulate gyri,even after controlling for subgroup differences on demographics and measures of pain-related distress.Together,results underscored the presence and nature of specific GMV differences underlying subjective reports of more versus less resilient responses to ongoing musculoskeletal pain.Despite multiple functions of these structures,we speculated that their roles in executive control and emotion regulation might attribute to subgroup differences in resilience.Future studies should incorporate measures of these aspects and consider longitudinal designs and randomized control trials to test causal effects of treatments on regional GMV and resilience and thus to illuminate the significance of these findings.Study 4 employed MKL models comprising Re Ho,ALFF and f ALFF(i.e.,F3method)to identify ROIs that distinguished best between Rs activity of more versus less resilient adults with CP.Participants with 70 women and 39 men were recruited(55 in higher resilience group and 54 in lower resilience group).Participants engaged in structural and functional magnetic resonance imaging(MRI)scans wherein MKL conducted in PRo NTo toolbox assessed Rs activity based on amplitude of low frequency fluctuations(ALFF),fractional amplitudes of low frequency fluctuations(f ALFF),and regional homogeneity(Re Ho)modalities to identify ROIs most salient for discriminating more versus less resilient subgroups after controlling for covariates.For each participant,116 features of functional maps were extracted from 116 ROIs based on the Automated Anatomical Labeling(AAL)atlas as an MKL source(ROI-MKL).Features were selected and combined to form a single mixed-kernel matrix through a multi-kernel strategy.A nested cross-validation scheme was used to obtain unbiased estimates of classification performance.We proposed a framework based on multi-kernel support vector machines that combined ALFF,f ALFF and Re Ho features to integrate different feature vectors and examined performance differences between multimodal versus unimodal approaches in distinguishing Rs of more and less resilience subgroups.Results showed that multi-modal classification based on combined ALFF,f ALFF and Re Ho features achieved a substantially higher classification accuracy rate compared to classification based on single modalities.The 10 Rs ROIs with greater relative classification power included the left olfactory cortex,right inferior occipital gyrus(IOG),left pre-central gyrus,left caudate nucleus,bilateral heschl gyrus(HG),right inferior temporal gyrus(ITG),left cuneus,left supplementary motor area(SMA),right supra Marginal gyrus(SMG),and bilateral inferior parietal lobule(IPL).In general,brain regions with the best discriminative power in the comparison of more resilient chronic pain patients versus less resilient cohorts included regions implicated in pain processing,reward,executive function,goal-directed action,emotion regulation and resilience to mood disorders.Identification of objective biomarkers could assist in clinical decisions for individual patients with chronic pain.Such results are an intriguing pattern that awaits replication in future investigations with larger samples and should be seen as preliminary evidence to drive future studies of intervention to improve resilience.In study 5,we tested pain resilience subgroup differences in brain activity during exposure to pain empathy and pain anticipation.The final sample comprised 72 women and 42 men in pain empathy task(i.e.sixty images of hands or feet in painful situations represented in two runs)and 74 women and 43 men in pain anticipation task(sixty images of white triangle/rectangle-cues for type of somatosensory stimulation represented in two runs).SPM 12 and DPABI were applied for image preprocessing,general linear model(GLM)in first level and a mixed design factorial Group(higher resilience vs.lower resilience)× Contrast(pain stimuli vs.non-pain stimuli)(SPM―flexible factorial‖ model)ANCOVA in second level analysis.We found that the entire sample showed increased bilateral supra Marginal gyrus(SMG)activation to pain picture compared to non-pain picture.Lower resilient group showed increased activation in response to pictures in left median cingulate and paracingulate gyri and bilateral supplementary motor area(SMA)compared to higher resilient group.For the brain responses to pain anticipation vs.non-pain anticipation,lower resilient group showed increased right insula and left superior temporal gyrus(STG)activation to stimuli anticipation compared to higher resilient group.In addition,higher resilient group showed decreased left inferior frontal gyrus activation to pain anticipation than to non-pain anticipation.These findings may due to that low resilient group displayed higher pain-related fear,pain sensitivity and engaged more neural resources in pain empathy and pain anticipation,enlightening new treatment strategies for chronic pain.Overall,these findings not only provide important information on the neural basis underlying the association between resilience and stimuli representation/anticipation,they also give a perspective into chronic pain management,such as altering pain empathy/anticipation may be critical for the therapeutic efficiency.Study 6 aimed to identify the mediating effects of GMV and Re Ho on associations between baseline pain resilience and subsequent pain intensity and interference from pain in a chronic pain sample over six-month interval.The final sample comprised 92 women and 50 men.In the baseline,structural and resting MRI data as well as PRS-C and CPG were acquired.Six months after their scans,participants were contacted for a brief,structured follow-up phone interview in which pain severity and impairment were reassessed.Correlations of demographic and pain characteristics,baseline pain measures with follow-up impairment and pain severity were assessed via Spearmen correlation analysis and showed that follow-up pain outcomes significantly related to gender,presence of other pain sites,prescription analgesics use,baseline pain severity and impairment which served as covariates in the following analyses.Next,individual normalized indicators of brain regions having significant associations with follow-up outcomes from GMV and Re Ho analyses were identified via SPM regression models and DPABI with a Gaussian random field(GRF)correction.We found that follow-up pain impairment had significant positive associations with GMV values in the left precuneus,left pre-central gyrus and left temporal pole:superior/middle temporal gyrus and Re Ho in the right caudate nucleus.Follow-up pain severity had a significant positive association with GMV values in the left fusiform gyrus.Significant brain areas identified from GMV and/or Re Ho analyses were labeled as regions of interest(ROIs)and resultant average ROI values were extracted using Response Exploration(REX)and treated as mediators.Subsequently,partial correlations were calculated between follow-up outcome measures,baseline resilience facet scores,and identified GMV and Re Ho areas.We found that GMV in the precuneus and temporal pole as well as Re Ho in the caudate nucleus significantly correlated with baseline resilience facets and follow-up pain outcomes.Finally,mediating effects of ROIs on potential relations of baseline behavioral perseverance and cognitive/affective positivity with follow-up levels of impairment and/or pain severity were performed using model 4 of the Hayes Process macro for SPSS with a bootstrapping(5000)procedure.For the mediation analysis,we found the significant mediation effect of left precuneus GMV in the association between baseline cognitive/affective positivity and follow-up pain impairment.Specifically,higher baseline cognitive/affective positivity scores not only directly predicted lower levels of pain impairment at the 6-month follow-up but also indirectly predicted via reduced left precuneus GMV.In addition,we also discovered the significant mediation effect of right caudate Re Ho in the association between baseline behavior perseverance and follow-up pain impairment.Specifically,higher baseline behavior perseverance not only directly predicted lower levels of follow-up pain impairment but also indirectly predicted via reduced right caudate Re Ho.In conclusion,this study may be the first to identify brain mechanism underlying baseline facets of pain resilience–pain impairment follow up associations.Although these structures have multiple functions,the roles of precuneus in pain sensitivity,pain catastrophizing,physical activities and emotion regulation and involvement of caudate in pain avoidance,executive function may explain the mediating effect of them between two facets of pain resilience at baseline and pain impairment at follow-up.These results supported for neural correlates underlying pain resilience with longitudinal data suggest that we can test hypotheses that specific brain structure,resting state activity effects identified can be used to evaluate future level of pain impairment due to CP.In study 7,moderating effects of mortality salience on the associations between pain resilience and pain outcomes were tested in an experiment featuring 170 healthy college students(86 women,84 men)randomly divided into an MS condition in which they answered two open-ended short-essay questions about death and University Students Personal Death Anxiety Scale versus a control condition in which they answered two parallel questions about watching television and Satisfaction to Variety Show Scale.The participants experienced background information,PRS-C,practice CPT,MS manipulation,delay task,second CPT,pain tolerance and pain intensity and measures in order.Pain tolerance was measured as the total time that participants immersed their non-writing hand in circulating ice water maintained at 2 Celsius for a4-minute maximum time.Subjective pain intensity was assessed with a three-item questionnaire assessing average and worst pain during the CPT as well as current pain.MS group differences on demographics,PRS-C and dependents variables(i.e.pain tolerance and pain intensity)were assessed via t-tests,chi square tests and analysis of variance(ANCOVA)using SPSS.Analyses indicated MS condition participants showed greater pain tolerance based on longer CPT immersion time compared to controls though pain intensity did not differ between the two conditions.Next,both of the PRS-C facets(behavior perseverance,cognitive-affective positivity)were positively correlated with pain tolerance and negatively correlated with pain intensity within the entire sample via spearman correlation coefficients analysis.Finally,Model 1 of Process macro version 3.4.1 in SPSS was used to conduct moderator analyses for mortality salience,pain resilience and pain tolerance/intensity models with a bootstrapping procedure.We found that MS moderated the association between behavior perseverance(i.e.one facet of pain resilience)and pain tolerance.The simple slope analysis indicated that the relationship between behavior perseverance and pain tolerance was stronger among MS group participants than it was among controls.In sum,this study is the first,to our knowledge,to examine the moderator effect of mortality salience on the association between pain resilience and pain tolerance.These findings imply that MS had a causal impact in enhancing the association between behavior perseverance and pain tolerance in the sample.This study provides empirical foundations for testing whether interventions used to increase awareness that one’s lifespan is limited helps more resilient people to not take their life for granted and increase their ability to bear pain.Extensions also need to evaluate how well current results apply to adults with chronic pain with longitudinal design.In sum,several important findings were revealed in the thesis.First,support for a stable two component(7-item cognitive/affective positivity and 3-item behavioral perseverance)structure,construct validity,and incremental validity of the PRS-C across two independent adult Chinese chronic pain samples(Studies 1 and 2)supports the use of this measure to identify individual differences in resilience to pain within clinical pain samples in a Chinese cultural context.Clinicians and researchers can also use the PRS-C as an outcome measure to evaluate the effectiveness of psychological pain management interventions.With both of the PRS-C and CDRS-C,more and less resilient subgroup were identified on the basis of scoring above and below median scores on the two validated resilience questionnaires.Study 3 to 6 support for neural correlates underlying pain resilience with VBM,MKL,flexible factor and mediation effects analysis.VBM analysis revealed that more resilient participants displayed significantly larger GMV in precuneus,parietal lobules,frontal gyrus and cingulate,regions involved in emotion regulation,attentional control and executive function.In addition,MKL conducted to Rs activity discovered several ROIs most salient for discriminating more versus less resilient subgroups including olfactory cortex,occipital gyrus,pre-central gyrus,caudate nucleus,temporal gyrus,cuneus,supplementary motor area,supra Marginal gyrus and parietal lobule.These brain regions implicated in reward,executive function,goal-directed action and emotion regulation.For the pain empathy and pain anticipation,the brain activation difference between two resilience subgroups showed that low resilient group displayed higher pain-related fear,pain sensitivity and engaged more neural resources in pain empathy and pain anticipation which was revealed by higher activity in supra Marginal gyrus,cingulate gyri,supplementary motor area,insula and temporal gyrus.Study 6 indicated precuneus and caudate,regions associated with pain sensitivity,emotion perception and executive function mediated the association between baseline cognitive/affective positivity or behavior perseverance and follow-up pain impairment.As such,study 3 to 6 support for neural correlates underlying pain resilience suggests that we can test hypotheses that specific brain structure,Rs,and activation effects identified in this research can be used to evaluate 1)current resilience levels of people with CP as well as 2)future levels of impairment due to CP.In study 7,we found that reminders of death prolonged pain tolerance and that mortality salience enhanced the association between behavior perseverance and pain tolerance.Study 7 provides empirical foundations for testing whether interventions used to increase awareness that death is inevitable promotes more resilient people to not take their life for granted and increase their ability to bear pain.Taken together,the present thesis provides evidence of structure and functional brain regions underpin pain specific resilience and provide potential new assessment and treatment goals for chronic pain.