This study is the first to evaluate for differences in depression and anxiety in a large sample of women who did and did not have pain in their breast prior to breast cancer surgery. Regardless of pain status, 37.5% of these women reported a CES-D score above the cutoff for clinically significant levels of depressive symptoms. This prevalence is higher than previous reports of depressive symptoms that ranged between 18% to 26% (Parker et al., 2007
). In terms of anxiety, regardless of pain status, almost 70% of the sample reported clinically meaningful levels of anxiety. This finding is significantly higher than the 50% reported in previous studies (Ozalp et al., 2003
; Parker et al., 2007
). The differences in prevalence rates across studies may be related to differences in the measures used to assess anxiety and depressive symptoms, timing of the measures, and sample characteristics. Taken together, these findings suggest that a relatively high percentage of women are experiencing significant amounts of psychological distress prior to surgery for breast cancer. Future studies need to incorporate a clinical interview to refine the diagnoses of anxiety and depression
Our hypothesis regarding the relationship between depressive symptoms and pain was supported. After controlling for the effects of age, menopausal status, and ethnicity, women with pain reported significantly higher CES-D Total, Somatic, and Depressed Affect subscale scores and significantly lower Positive Affect subscale scores. The difference in total CES-D scores between the two pain groups represents not only a statistically significant but a clinically meaningful difference in depressive symptoms (i.e., d=1.2) (Osoba, 1999
). In addition, 49.5% of the women with pain met the cutoff for clinically significant levels of depressive symptoms compared to only 32.8% in the no pain group (p=.003). Of note, in the pain group, the combination of high DA with low PA subscale scores is characteristic of the expression of a depressive disorder (Watson et al., 1988
). Furthermore, when specific conditions on SCQ were evaluated, a higher percentage of women with pain reported a history of depressive illness (31%) compared to the no pain group (18%, p=.009). These women with a history of depressive illness, high depressive symptoms scores, and pain prior to surgery may represent a group of women who are at risk for worse postoperative outcomes.
Only one study was found that reported CES-D subscale scores for oncology patients (Musick et al., 1998
). Compared to Musick and colleagues' large sample of cancer patients (n = 1371) with a variety of cancer diagnoses that included both women and men (DA=1.2, S=0.9, PA=2.7, ID=0.9), women in our study reported higher DA (4.9) and S (5.6) subscale scores, better PA (9.1) scores, and similar ID scores (0.3). Similar differences exist between our findings and mean subscale scores reported for a community sample of insured women (n = 218) (DA=3.3, S=4.5, PA=2.1, I.D=0.4) (Gatz and Hurwicz, 1990
) and older persons (n = 174) (DA=2.0, S=3.1, PA=5.1, I.D=0.1) (Baune et al., 2007
). In contrast, CES-D subscale scores for women in our study are similar for the S and PA subscales of the CES-D reported by a sample of patients with rheumatoid arthritis (n=415; DA=2.6, S=5.8, PA=9.1, I.D=0.9) where 83% of the patients were women (McQuillan et al., 2003
). Interestingly, the CES-D subscale scores for our sample are lower than scores reported by a community sample (n = 680) of married women (DA=5.6, S=8.1, PA=6.3, I.D=0.7)(Ross and Mirowsky, 1984
). However, this inconsistent finding might be explained by the fact that Ross and colleagues (1984)
removed two items from the CES-D (“crying” and “life is a failure”) before the factor structure of the subscales was tested.
Our hypothesis regarding differences in the severity of both state and trait anxiety was not supported. While previous studies of acute and chronic pain have shown positive associations between anxiety and pain (Katz et al., 2005
; Poleshuck et al., 2006
), it is not entirely clear why pain status influenced depression but not anxiety. One possible explanation is that high levels of anxiety associated with the diagnosis and impending surgery in both groups blunted any differential effects of pain on this symptom.
As expected, women with pain prior to surgery scored lower on the physical well-being subscale of the QOL-PV compared to women without pain. However, the total and psychological, spiritual, and social subscale scores were not different between the two pain groups. It is not entirely clear why between group differences in psychological well-being scores were not found given the fact that these women differed on depression scores. One possible explanation is that the psychological well-being subscale of the QOL-PV evaluates additional aspects of psychological well-being that are not evaluated using either the CES-D or the STAI-S (e.g., coping) and that these characteristics are not affected by pain.
Several limitations need to be acknowledged. First, self-report measures were used to evaluate depression and anxiety. Future studies need to incorporate a diagnostic interview like the Structured Clinical Interview for DSM-IV Disorders to confirm a clinical diagnosis of anxiety or depression. Second, based on recent reports that neuroticism is associated with depression and anxiety (Middeldorp et al., 2005
), future studies should include a personality measure in order to assess the relative contribution of various aspects of an individual's personality to pain, anxiety, and depression. A possibility exists that the rates of anxiety and depression reported in this study may be underestimations because one of the major reasons for refusal was that women reported that they were too overwhelmed by their cancer experience to participate in this study. Finally, because the majority of women were Caucasian and well-educated, findings from this study may not generalize to all women who undergo surgery for breast cancer.
Findings from this study suggest that a significant percentage of women, regardless of pain status, experience anxiety and depressive symptoms. Future studies need to evaluate for changes over time in anxiety and depressive symptoms and the factors that influence the trajectories of these symptoms.