Our study has several important findings. First, a strong SOC is associated with better pain-specific outcomes, especially for constructs relevant to coping such as better pain self-efficacy and less pain catastrophizing. Second, a strong SOC is also associated with better HRQoL, specifically self-reported general health, vitality, and social functioning. Third, the association of SOC remained significant for most outcomes (except pain intensity and impairment) even after controlling for important covariates, including the presence of major depression. Indeed, the strength of association of SOC with pain coping and HRQoL outcomes was 40-80% that of major depression. Fourth, SOC did not change over the 12 months of the trial, indicating that SOC was a stable trait in our study population.
Our findings are consistent with previous studies that have found an association between SOC and HRQoL. A systematic review published in 2007 found that a stronger SOC was associated with better HRQoL . Our results in chronic pain patients complement findings in other medical populations confirming the association between SOC and HRQoL as measured by the SF-36, including coronary heart disease patients , patients following a myocardial infarction , nursing home residents , family caregivers of older adults living in the community , and middle-aged persons from the general population .
Similar to our study, a cross-sectional study of 232 participants recruited from a “Neck and Back” unit in Norway found that stronger SOC was associated with greater pain self-efficacy . Prior studies have found mixed results for associations of SOC with pain intensity and functional impairment. Our findings are in line with those of Benz et al. , who found that SOC was unrelated to pain severity in a prospective cohort study of 355 patients with hip and knee osteoarthritis. In contrast, at least two studies have found associations of SOC with pain severity. A small sample of 73 patients who received laparoscopic cholecystectomy found that a strong SOC was significant, but weak predictor of less intense postoperative pain at one week . Likewise, in another small study assessing the features of fibromyalgia syndrome in 40 women with non-metastatic breast cancer, SOC was inversely correlated with BPI severity and BPI interference ; however, this study did not adjust for sociodemographic or clinical variables (including depression).
Although low SOC does not inherently imply a depressive mood, it has been associated with depression in prior research . In contrast to most previous studies, we were able to control for a clinical diagnosis of major depression, rather than simply a continuous measure of sub-threshold depressive symptoms. We found that SOC remained statistically significant for all outcomes (except for 2 pain outcomes), even after controlling for major depression. More specifically, after we adjusted for major depression, the T-value was attenuated, but remained statistically significant for five of the seven outcomes.
Because of the strong association between SOC and depression, some researchers have even questioned the extent to which they are conceptually distinct from one another [14, 43]. For instance, in a recent national Finnish-based survey study, higher SOC scores were related to a lower risk of cardiovascular disease and mortality, but this association disappeared after adjustment for depressive symptoms . One partial explanation for the independent effects of SOC found in our study may be that the measure we used – the SOC-3 -- has been found to have a lower correlation with depression compared with other SOC instruments (e.g., the widely used 13-item of the SOC) .
Similar to Antonovsky’ s original tenet of SOC, we found that SOC was a stable trait in the present study. Many of the outcomes (except for those associated with pain, which was treated in the trial), changed very little over the 12 months. Therefore, we are unable to conclude that baseline SOC is not a predictor of change in HRQoL outcomes, mainly because SOC and these HRQoL outcomes remained stable over time. This lack of change over time was not surprising since the intervention did not specifically target these outcomes. According to Antonovsky [7, 9], SOC remains stable throughout adulthood and is thereafter only minimally affected by traumatic life events. Several population-based studies supported this notion (i.e., SOC was found to be stable over time) [46–49]. However, more recent longitudinal studies derived from a more diverse group of non-pain related study participants and ranging from two to five years follow-up have found that SOC may vary over time [35, 44, 50, 51]. To our knowledge, only one study has examined the extent to which SOC varied over time in patients with pain. In patients receiving laparoscopic cholecystectomy for gallstone disease, SOC was unstable over a 6 month period, changing more than 10% in 37% of the patients over 6 months . Therefore, more longitudinal research is needed to examine the stability of SOC in different populations.
This study has several limitations. First, our sample was comprised entirely of US Veterans who received primary care from a single VAMC. Thus, our findings may be less generalizable to non-VA settings. However, in comparison to past studies with veterans, our sample included more women (18%) as well as higher educational attainment, employment rates and income. Second, there is the potential for selection bias because this study included only patients who were willing to enroll in a clinical trial. Third, the Cronbach’s α for the 3-item measure of SOC in the present study was .58, a value that is less than ideal. However, a systematic review of studies using the SOC-3 found that the Cronbach’s α ranged from .35 to .61 . Schumann et al.  created a German version of SOC-3 and reported a Cronbach’s α of .45. Very brief scales typically have lower Cronbach’s α than longer scales. The fact that our principal analyses used the SOC as a binary measure based upon a validated cutpoint partly mitigates the moderate internal reliability of the 3-item SOC as a continuous measure. As Eriksson and Lindstrom  articulated, the intercorrelations between the SOC-3 and the original long SOC measures are satisfactory. The present study does not have validation data to compare the SOC-3 to the longer versions. Future research could explore the SOC-3 with longer versions of the SOC in English based samples. Fourth, our intervention focused on optimizing medications focused on pain. Future research should develop and implement an intervention that targets aspects of SOC and HRQoL to determine if SOC was a predictor of change in HRQoL.
Knowledge of the role of SOC in adapting to and coping with stress may equip health professionals for developing patient-centered care that incorporate SOC concepts to assist individuals suffering from musculoskeletal pain. Intervention strategies might be developed to help patients suffering from musculoskeletal pain; to identify their internal and external resources in order to strengthen their belief that their life is meaningful (i.e., the problems and demands of life are viewed as challenges instead of burdens), comprehensible (i.e., the circumstances that an individual will encounter in the future is viewed as predictable, ordered, and explicit), and manageable (i.e., the patient has the capacity at their disposal to deal with their pain). Thus, clinicians could help patients shift their focus away from crises, but instead view the encounters as balanced and meaningful .