Background/Purpose:

Acute pain is a vital adaptive and protective mechanism. Conversely, chronic pain is a persistent, maladaptive response that outlasts the normal healing period of an injury or insult, and may result in drastic deterioration of quality of life, sometimes for the rest of life. In a longitudinal brain imaging study of people with subacute back pain, we recently identified brain markers by fMRI that predict transition to chronic pain. Here we investigate, in the same study, behavioral biomarkers that may be predictive of pain chronification.

Methods:

62 patients with new onset subacute back pain (less than 3 mo duration, no back pain for at least a year prior to symptom onset) were seen 6 times over one year.  At each visit, pain intensity was determined using a 100mm visual analog scale, behavioral questionnaires completed and fMRI brain scans obtained.   For this analysis, the subacute group was divided into persisting (SBPp) and recovering (SBPr) groups using a greater than 20 % change criterion, from the baseline visit to the one year visit. Multivariate logistic regression was utilized to evaluate individual behavioral parameters and their association with pain persistence.

 Results:

There were 32 males and 30 females, with mean age at study onset of 43±11 yrs, mean pain duration 10 weeks, and mean initial VAS pain intensity was 64±16.   By the end of one year, there were 36 SBPr and 26 SBPp.  No significant group differences in VAS pain intensity were evident at baseline. However, when evaluated over time, SBPp and SBPr segregated as early as their second visit (within 2 weeks) and the divergence persisted until the final visit at one year. The Neuropathic Pain Scale (NPS), the McGill Pain Questionnaire affective (MPQa), Pain detect (pDetect) and smoking status were significantly different between SBPr and SBPp groups at baseline (t-test). Smoking status at time of entry into the study was the strongest predictor of SBPr and SBPp groups at one year from symptom onset, odds ratio = 5, CI 1.6-16, p<0.007, accuracy = 0.67. NPS, MPQa and pDetect were correlated with each other, and their logistic multiple regression was not significant. Thus, all three parameters reflect interrelated properties of back pain.  In a multiple regression logistic model when we include all four parameters, only smoking status remains significant with a stronger odds ratio = 13.5, CI = 2.5-75, p=0.003, accuracy =0.84; thus after correcting for pain characteristics the effects of smoking on pain chronification are even stronger. 

Conclusion:

There is a strong association between smoking at the time of the onset of back pain symptoms and longer term pain chronification.  As this observation was not a prespecified hypothesis of the study, we label the observed result as an association that needs to be validated in a systematic study.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predictors-of-persistence-in-people-with-subacute-low-back-pain/