Depression in lung cancer patients: the role of perceived stigma

Authors


H. Lee Moffitt Cancer Center, 12902 Magnolia Drive MRC-PSY, Tampa, FL 33612, USA. E-mail: Brian.Gonzalez@moffitt.org

Abstract

Objective: Previous research has shown that lung cancer patients are at an increased risk for depressive symptomatology; however, little is known about the factors contributing to depression in these patients. This study focused on the possible association between perceived stigma related to a lung cancer diagnosis and depressive symptomatology. It was hypothesized that greater perceived stigma would be related to greater depressive symptomatology and that perceived stigma would account for variance in depressive symptomatology above and beyond that accounted for by relevant clinical, demographic, and psychosocial variables.

Method: A sample of 95 participants receiving chemotherapy for stage II–IV non-small cell lung cancer was recruited during routine outpatient chemotherapy visits. Participants completed a demographic questionnaire and self-report measures assessing perceived stigma, depressive symptomatology, and other psychosocial variables. A medical chart review was conducted to assess clinical factors.

Results: As hypothesized, there was a positive association between perceived stigma and depressive symptomatology, r = 0.46, p<0.001. Perceived stigma also accounted for significant unique variance in depressive symptomatology above and beyond that accounted for by relevant demographic, clinical, and psychosocial factors, β = 0.19, p<0.05.

Conclusions: Future research should aim to replicate and extend these findings in longitudinal studies and explore whether lung cancer patients' depressive symptomatology can be ameliorated by targeting perceived stigma. Copyright © 2010 John Wiley & Sons, Ltd.

Introduction

Studies of lung cancer patients report high rates of clinically significant depressive symptomatology shortly after diagnosis (21–44% [1–5]) and after treatment completion (29–44% [2, 4, 6–10]). These rates exceed those observed for individuals with many other types of cancer. For example, studies of patients with breast cancer, head and neck cancer, lymphoma, and thyroid cancer reported rates of elevated depressive symptomatology between 7 and 17% [9–11]. Another study reported rates of elevated depressive symptomatology between 7 and 23% in colorectal, breast, gynecological, and oral cancer patients [12]. Understanding the risk factors and correlates of depression among lung cancer patients can contribute to the development of effective interventions targeted specifically for this patient population.

Research on demographic correlates of depression in lung cancer patients is limited and results are mixed. For example, some studies have found that female patients [1, 4], older patients [1, 6, 13], and less educated patients [14] are at increased risk for depressive symptomatology; however, others have found no differences by sex [2, 7, 14], age [2, 7, 14], or education [2, 7]. More definitive results have been obtained about relationships between clinical factors and depressive symptomatology. Specifically, poorer performance status has consistently been linked to greater depressive symptomatology in lung cancer patients [1, 4, 6, 7, 14]. The psychosocial correlates of depressive symptomatology in lung cancer patients have only recently begun to be studied. Emerging evidence suggests that coping and social support merit further examination [6, 13]. Although dysfunctional attitudes (a measure of cognitive vulnerability to depression) and alcohol intake have not been examined in relation to depression among lung cancer patients, the existence of relationships with depression in the general population [15–18] argues for their study as potential psychosocial correlates.

Only three studies have investigated the relationship between smoking status and depressive symptomatology in lung cancer patients. Two found no relationship [7, 14]. A third study, which focused on patients who had smoked in the three months prior to lung cancer resection, found a trend suggesting that patients who continued smoking after surgery had greater depressive symptomatology than those who quit [19]. Although all three studies reported no significant relationship between depression and smoking status, the evidence of such a relationship in the general population [20–22] and the potential implications if a similar relationship was found among lung cancer patients argue for additional research on this topic.

Another factor that might be related to depressive symptomatology in lung cancer patients is perceived stigma. Modified Labeling Theory [23], which was first used to describe the effects of stigma on individuals with psychiatric disorders, posits that once society labels an individual, they are subjected to uniform responses from others. These societal responses can constrain an individual into the role to which they are being subjected, which can cause the individual to accept this role and incorporate it into their identity, a process that can often result in psychological harm. Perceived stigma, the perception that one is subjected to the uniform responses from others expected for an individual with a certain label, has been studied in people who are HIV-positive. Individuals with this illness often perceive (accurately or inaccurately) that they are subjected to uniform responses from society as a result of their medical condition. The potential for illness-related stigma in those who are HIV-positive originates from the fact that this disease is often transmitted via sexual behavior and intravenous drug use [24, 25]. A study of HIV-positive individuals found that internalized stigma accounted for unique variance in depressive symptomatology over and above relevant demographic factors, clinical factors, and social support [26].

Given its strong association with tobacco use, lung cancer is commonly viewed as a preventable disease. Thus, lung cancer patients are often stigmatized for having caused a preventable illness because of their past or current history of tobacco use [27, 28]. Consequently, patients may blame themselves and others may blame patients for developing lung cancer. The importance of this topic is underscored by results of a qualitative study of perceived stigma in lung cancer patients, in which many patients reported feeling stigmatized because of the strong association between smoking and lung cancer [29]. These patients, including those who reported no history of smoking or had stopped smoking several years before their diagnosis, felt that they were being blamed for their disease by friends, loved ones, and even healthcare professionals [29]. In addition, a recent study found that lung cancer patients reported higher perceived stigma than breast and prostate cancer patients [30].

The primary aim of the current study was to identify psychosocial correlates of depressive symptomatology in lung cancer patients in order to provide an empirical basis for intervention development. In particular, the present study sought to test the hypothesis that perceived stigma would be associated with depression. Additionally, it was hypothesized that psychosocial factors related to depression in other populations (i.e. coping, social support, and dysfunctional attitudes) would be related to depressive symptomatology in lung cancer patients. Finally, the study sought to determine whether perceived stigma accounts for variability in depressive symptomatology in lung cancer patients above and beyond relevant demographic, clinical, and psychosocial factors.

Method

Participants

Eligible patients for the current study were: diagnosed with stage II, III, or IV non-small cell lung cancer; ⩾18 years of age; able to understand, speak, and read English; able to provide informed consent; and had no history of other cancers with the exception of non-melanoma skin cancer. In addition, eligible patients had to be receiving chemotherapy. This inclusion criterion helped to ensure that patients were at similar points in the process of dealing with lung cancer. The University of South Florida Institutional Review Board approved the project protocol prior to the initiation of the study.

Procedure

Participants were recruited between March and October, 2009. Study eligibility was determined via consultation with Moffitt Cancer Center Thoracic Oncology Program team members. Potential participants were approached by trained research personnel who were not involved in the patients' medical care during a routine outpatient visit and the study protocol was explained to them. Those eligible and interested in the study provided written informed consent. Participants were given the option of filling out the study measures during their outpatient visit or taking them home and returning them in a self-addressed stamped envelope that was provided. Participants were not compensated for their participation.

Measures

Demographic information

Demographic information was collected using a standardized self-report form. Variables assessed were: age, sex, race, ethnicity, education, income, and marital status. In addition, participants' current and past smoking status and alcohol use in the past month were assessed (Table 1).

Table 1. Demographic characteristics of the sample (N = 95)
VariableMSD
Age, years64.04(8.79)
Variablen%
Gender
 Males39(41.1)
 Females56(58.9)
Education
 ⩽High school graduate38(40.0)
 >High school graduate57(60.0)
Race
 White88(92.6)
 Non-White7(7.4)
Ethnicity
 Hispanic3(3.2)
 Non-Hispanic92(96.8)
Marital status
 Currently married58(61.1)
 Not married37(38.9)
Total household income
 ⩽$39 99922(23.2)
 ⩾$40 00044(46.3)
 Declined to answer29(30.5)
Alcohol use in past month
 No42(44.2)
 Yes53(55.8)
Cigarette use
 Never12(12.6)
 Past68(71.6)
 Current15(15.8)

Clinical information

The following clinical information was gathered via a review of patients' medical records: date of lung cancer diagnosis, disease stage, previous lung cancer treatment, planned lung cancer treatment, ECOG performance status [31], and antidepressant medication use at the time of the study visit (Table 2).

Table 2. Clinical characteristics of the sample (N = 95)
VariableMSD
Months since diagnosis (SD)18.14(30.35)
Variablen%
  1. ECOG: Eastern Cooperative Oncology Group.

Disease stage
 II–III32(33.7)
 IV63(66.3)
ECOG performance status
 020(21.0)
 162(65.3)
 2–313(13.7)
Diagnosis of past major depression
 No82(86.3)
 Yes13(13.7)
Taking antidepressant medication at time of study visit
 No79(83.2)
 Yes16(16.8)
Had surgery for this cancer
 No50(52.6)
 Yes45(47.4)
Had radiation therapy for this cancer
 Never54(56.9)
 Before current course of chemo33(34.7)
 Currently8(8.4)

Stigma

Perceived stigma was assessed using the Social Impact Scale (SIS, [32]). This 24-item measure yields a total score as well as subscale scores for respondent's perceptions about the extent they believe they are experiencing social rejection, financial insecurity, internalized shame, and social isolation as a result of their illness. The validity of this scale has been demonstrated in previous studies with HIV-positive patients as well as cancer patients [32, 33]. The SIS total scale demonstrated good internal consistency reliability (Cronbach's α = 0.95) in the current study, as did the SIS subscales (Cronbach's α⩾0.81).

Avoidant coping

We chose to measure avoidant coping because of its frequently found association with psychological distress, including in studies with cancer patients [34, 35]. Avoidant coping was assessed using the 6-item Cognitive Avoidance Subscale of the Coping Responses Inventory [36]. Instructions asked participants to report how often they had engaged in certain responses in connection with the problem of their cancer and its treatment. The validity of this subscale has been demonstrated in prior research [37]. It demonstrated adequate internal consistency reliability (Cronbach's α = 0.72) in the current study.

Social support

Social support was assessed using the 5-item ENRICHD Social Support Instrument (ESSI; [38]). The ESSI has been shown to have good convergent validity with other measures of social support [38]. The ESSI demonstrated good internal consistency reliability (Cronbach's α = 0.91) in the current study.

Dysfunctional attitudes

Dysfunctional attitudes were assessed using the 40-item Dysfunctional Attitudes Scale (DAS, [39]). Designed to measure cognitive distortions consistent with the Cognitive Theory of Depression [40, 41], the validity of the DAS has been demonstrated in several previous studies [15, 39]. The DAS demonstrated good internal consistency reliability (Cronbach's α = 0.89) in the current study.

Depressive symptomatology

Participants' depressive symptomatology in the past week was assessed using the 20-item Center for Epidemiological Studies Depression Scale (CES-D; [42]. The CES-D has been shown to be a valid measure of depressive symptomatology in people with cancer [43]. The CES-D demonstrated good internal consistency reliability (Cronbach's α = 0.84) in the current study.

History of depression

Participants' history of Major Depressive Disorder (MDD) prior to their lung cancer diagnosis was assessed using the mood episode section of the Structured Clinical Interview for DSM-IV (SCID; [44]). The SCID is a widely used semi-structured interview designed to identify the presence or absence of Axis I mental disorders according to Diagnostic and Statistical Manual of Mental Disorders IV (DSM-IV; [45]) criteria. Before administering the SCID, study staff reviewed video-taped instructional materials from the measure's authors, practiced administration and rating under supervision, and were periodically observed to ensure their fidelity to established procedures.

Statistical analyses

Independent samples t-tests, ANOVAs, and chi-square tests were performed, as appropriate, to examine relationships of demographic and clinical variables with depressive symptomatology. Correlational analyses were conducted to determine the relationships between depressive symptomatology and each of the following: perceived stigma, social support, avoidant coping responses, and dysfunctional attitudes. A hierarchical regression analysis was conducted to determine whether perceived stigma accounted for unique variance in depressive symptomatology not accounted for by other psychosocial variables and by demographic and clinical variables related to depressive symptomatology. Variables were included in the model in the following steps: (1) the presence/absence of a lifetime (before lung cancer diagnosis) history of MDD; (2) any demographic or clinical variable related to depressive symptomatology; (3) social support, coping responses, and dysfunctional attitudes regardless of the significance of their relationship to depressive symptomatology; and (4) perceived stigma. Additional hierarchical regression analyses were conducted to determine which components of perceived stigma contributed unique variance to depressive symptomatology. The four subscales of the SIS were added as the independent variable in each of these analyses on the fourth step.

A power analysis using Power and Precision 2.0 indicated that for a correlational analysis with a Type I error rate of 0.05 (two-tailed), the sample size of the current study (N = 95) yielded power equal to 0.85 for detecting a medium-sized effect (r = 0.30). A second power analysis for the hierarchical regression analysis described above indicated that if Steps 1–3 accounted for 40% of the variance in depressive symptomatology, power was equal to 0.82 with a Type I error rate of 0.05 (two-tailed) to detect a 5% increase in variance accounted for by stigma on Step 4.

Results

Of the 839 patients initially screened for this study via medical record review, 680 were ineligible based on their medical characteristics (e.g. history of other cancer, not receiving chemotherapy). The remaining 159 patients were approached for participation; of these patients, an additional 17 were deemed ineligible before consent, 33 refused to participate, and 109 agreed to participate (77% of those eligible). The 109 patients who agreed to participate did not differ in terms of age, gender, or race from the 33 patients who declined to participate, p's⩾0.48. Of those patients who agreed to participate, eight withdrew from the study or never completed the study measures, and six were found to be ineligible after they participated because they were not receiving chemotherapy. Thus, analyses were conducted on 95 participants who had evaluable data.

Participants ranged in age from 42 to 83 years (M = 64.04; SD = 8.79). The majority were high school graduates (60%), married (61.1%), and White (92.6%). On average, participants were 18.14 months (SD = 30.35) from their lung cancer diagnosis. Forty-five (47.4%) participants had surgery for this cancer and 41 (43.2%) had been treated with radiation therapy. Thirteen participants (13.7%) met criteria for a diagnosis of past MDD, and 16 (16.8%) were taking antidepressant medications at the time of the study visit. Twelve participants (12.6%) were never smokers, 68 (71.6%) were past smokers, and 15 (15.8%) were current smokers at the time of the study visit (see Tables 1 and 2 for complete demographic and clinical information).

Participants' mean total score on the SIS (M = 42.90) lies approximately at the midpoint of its range from 0 to 80, and the standard deviation of scores (SD = 11.87) suggests considerable variability in participants' reporting of perceived stigma. The mean CES-D score was 14.39 (SD = 8.26), suggesting that on average participants exhibited low to moderate depressive symptomatology; 38% of participants (n = 36) met the CES-D cutoff for clinically significant depressive symptoms (score⩾16).

Relationships of demographic and clinical variables to depressive symptomatology

Analyses were conducted to identify demographic and clinical variables related to depressive symptomatology. No demographic variables were significantly associated with depressive symptomatology, p's>0.05. Among clinical variables, those patients with a diagnosis of past MDD reported significantly greater depressive symptomatology than those without a diagnosis of past MDD, t(93) = 2.37, p = 0.02. Greater time since diagnosis was also significantly related to depressive symptomatology, r(93) = 0.20, p = 0.05.

Relationship of perceived stigma and psychosocial variables to depressive symptomatology

Correlational analyses were conducted to test the hypothesis that perceived stigma would be related to depressive symptomatology (Table 3). As hypothesized, greater overall perceived stigma as well as more social rejection, financial insecurity, internalized shame, and social isolation were significantly related to greater depressive symptomatology. Correlational analyses were also conducted to test the hypothesis that social support, avoidance coping, and dysfunctional attitudes would be related to depressive symptomatology. As hypothesized, poorer social support, more avoidant coping, and more dysfunctional attitudes were significantly related to greater depressive symptomatology.

Table 3. Relationship of depressive symptomatology with perceived stigma and psychosocial variables (N = 95)
VariableDepression (CES-D)p-value
  1. CES-D: Center for Epidemiologic Studies Depression Scale; SIS: Social Impact Scale; ESSI: ENRICHD Social Support Instrument; CRI CA: Coping Responses Inventory Cognitive Avoidance Subscale; DAS: Dysfunctional Attitudes Scale.

Perceived stigma (SIS Total)0.46<0.001
 Social rejection (SocRej)0.290.004
 Financial insecurity (FinIns)0.43<0.001
 Internalized shame (IntSha)0.270.010
 Social isolation (SocIso)0.58<0.001
Social support (ESSI)−0.330.001
Avoidant coping (CRI CA)0.36<0.001
Dysfunctional attitudes (DAS)0.48<0.001

Regression analysis of depressive symptomatology

A hierarchical regression analysis was conducted to determine whether perceived stigma accounted for significant variability in depressive symptomatology above and beyond that accounted for by relevant variables (Table 4). Diagnosis of past MDD was specified for inclusion in the model prior to conducting analyses, time since original lung cancer diagnosis was included in the model because of its relationship with depressive symptomatology, and psychosocial variables were included based on the pre-specified model. Diagnosis of past MDD accounted for 5% of the variance and time since original lung cancer diagnosis accounted for 4% of the variance in depressive symptomatology. Social support, avoidant coping, and dysfunctional attitudes together accounted for an additional 35% of the variance in depressive symptomatology. Perceived stigma was entered into the model on the final step. As hypothesized, perceived stigma accounted for a statistically significant amount of additional variance (3%) in depressive symptomatology (p = 0.043).

Table 4. Summary of hierarchical regression analysis for variables predicting depressive symtpomatology (CES-D) (N = 95)
VariableβΔR2
  • Overall F (6, 88) = 12.43, p<0.001; CES-D: Center for Epidemiologic Studies Depression Scale; MDD: Major Depressive Disorder; ESSI: ENRICHD Social Support Instrument; CRI CA: Coping Responses Inventory Cognitive Avoidance Subscale; DAS: Dysfunctional Attitudes Scale; SIS: Social Impact Scale.

  • *

    *p<0.05;

  • **

    **p<0.01.

Step 1 0.05
 History of MDD0.25* 
Step 2 0.04
 Time since diagnosis0.11 
Step 3 0.35
 Social support (ESSI)−0.11 
 Avoidant coping (CRI CA)0.27** 
 Dysfunctional attitudes (DAS)0.33** 
Step 4 0.03
 Perceived stigma (SIS)0.19* 

Four additional hierarchical regression analyses were conducted to determine which, if any, of the subscales of the SIS accounted for significant variance in depressive symptomatology above and beyond that accounted for by relevant demographic, clinical, and psychosocial variables. In these analyses, the first three steps were identical to the above-described hierarchical regression; however, the final step consisted of entry of one of the SIS subscales rather than the SIS total score. Results indicated that the Social Rejection and Internalized Shame did not account for additional significant variability in depressive symptomatology; however, Financial Insecurity (β = 0.19, p<0.036) and Social Isolation (β = 0.33, p<0.001) did account for additional significant variability (3 and 7% of the variance in depressive symptomatology, respectively).

Discussion

Rates of clinically significant depressive symptomatology (38%) were similar to those found in previous studies in patients with lung cancer (21–44% [1–5]). As expected, greater perceived stigma related to a lung cancer diagnosis was significantly related to higher levels of depressive symptomatology. More avoidant coping, poorer social support, and more dysfunctional attitudes were also significantly related to greater depressive symptomatology. Additional analyses indicated that perceived stigma contributed unique variance in depressive symptomatology above and beyond that accounted for by clinical variables (time since lung cancer diagnosis and history of past MDD), and psychosocial factors (avoidant coping, social support, and dysfunctional attitudes).

In addition to examining the relationship of perceived stigma to depressive symptomatology, we also examined the relationships of perceived stigma subscales (social rejection, financial insecurity, internalized shame, and social isolation) to depressive symptomatology. Univariate analyses indicated that all four subscales were related to depressive symptomatology. In contrast, in multivariate results only two subscales (financial insecurity and social isolation) contributed unique variance to depressive symptomatology above and beyond that contributed by control variables. The reason for this differential pattern of relationships is unclear. It may be that the financial insecurity and social isolation subscales are tapping into issues that extend beyond the stigma of illness and more broadly reflect reactions to disease. Another possibility is that scales assessing social rejection and internalized shame overlap more with the other psychosocial predictors of depression that were controlled for in multivariate analyses (i.e. social support, avoidant coping, and dysfunctional attitudes).

To the best of our knowledge, this is the first study to examine the relationship of perceived stigma to depressive symptomatology in lung cancer patients. This study's findings provide quantitative evidence consistent with qualitative evidence provided by Chapple and colleagues (2004), which suggested that lung cancer patients experience significant stigma from others as a result of their illness. Moreover, this study extends the qualitative findings by providing evidence for a link between perceived stigma and depressive symptomatology. Documenting this link among lung cancer patients is important for several reasons. First, it adds further evidence to the growing body of research suggesting a connection between illness-related stigma and depressive symptomatology. As noted earlier, studies have found that patients with other stigmatizing conditions (e.g. HIV infection) who report more perceived stigma also report greater depressive symptomatology [26]. Second, it adds to knowledge about factors contributing to depressive symptomatology among lung cancer patients, a group that is particularly likely to experience depressive symptoms [7, 46].

This study's findings are consistent with the Cognitive Theory of Depression, which states that one's experiences may contribute to depressive symptomatology by activating maladaptive schemas [40, 41]. When activated, these maladaptive schemas perpetuate dysfunctional attitudes that are evident in automatic cognitive responses to stimuli. These automatic cognitive responses propagate negative views of oneself, the experiences one undergoes, and one's outlook on the future. These negative thought patterns, termed the Cognitive Triad, are the most conscious manifestations of the depressive state and are theorized to result in affective and somatic depressive symptoms [40]. Findings from the current study suggest that this chain reaction could be activated in lung cancer patients who perceive they are being stigmatized because of their illness. Some lung cancer patients may in fact misperceive that they are being stigmatized because of their illness; however, the effect of misperceived stigma would likely be similar to that of actual stigmatization.

With regard to clinical implications, the findings suggest psychotherapeutic approaches that might be employed to alleviate or prevent depressive symptoms among lung cancer patients. Most approaches to reducing stigma focus on educating the public about inaccuracies of stereotypes and replacing these inaccuracies with facts [46]. Although potentially useful in the long-term, these efforts are unlikely to address the perceived stigma currently experienced by many lung cancer patients. Other approaches, such as cognitive therapy, may be helpful in counteracting the negative effects of stigma (e.g. dysfunctional attitudes) on the stigmatized individual [46]. Specifically, a modified cognitive therapy approach that focuses on altering thoughts and feelings associated with perceived stigma may prove effective in reducing depressive symptomatology. One strategy might involve pointing out to patients that, although it is true that smoking causes many cases of lung cancer, self-blame is a maladaptive coping strategy after lung cancer is diagnosed. To help patients move beyond the self-blame, they might be encouraged to consider exempting beliefs. For example, emphasizing the addictiveness of cigarette smoking and the deception in tobacco industry advertisements could help patients view themselves as having been wronged rather than as being a wrong-doer. Though self-blame can be targeted and reduced, it may remain in some patients. These patients could be encouraged to acknowledge the potential for culpability and then move on to more productive uses of their energies. These and other focuses within the broader framework of Cognitive Therapy for depression could help patients move past self-blame and other consequences of their illness.

This study has several limitations. First, its cross-sectional nature limits the conclusions that can be drawn. Although results can be interpreted as suggesting that perceived stigma contributes to depression, the possibility that depression contributes to stigma cannot be ruled out. Second, the study's sample was relatively homogeneous with respect to race and ethnicity, which limits the ability to generalize to the broader lung cancer patient population. Third, the lung cancer patients in this study were receiving chemotherapy; thus, this study's findings may not generalize to lung cancer patients receiving other types of treatments or receiving no treatment. Last, although use of antidepressant medication was not related to depressive symptomatology in this study, participants' use of psychotherapy and related services was not assessed.

Because this is the first quantitative study to identify the relationship between perceived stigma and depressive symptomatology among lung cancer patients, it will be important to see if these findings can be replicated in future research. Beyond this, there is a need for longitudinal research that would allow for examination of the temporal relationships between perceptions of stigma and depressive symptomatology. In addition, future studies should examine the potential relationship between the use of psychotherapy services and depressive symptomatology. Future studies should also aim to recruit samples of lung cancer patients that are more diverse with regard to race, ethnicity, and socioeconomic status. Additionally, future research should explore whether perceived stigma can be reduced in lung cancer patients and, if so, whether it results in reductions in depressive symptomatology.

Acknowledgements

Special thanks to Pamela Reiersen and Amber Isley for their assistance with participant recruitment and data collection.

Ancillary