Depression and demoralization as distinct syndromes: Preliminary data from a cohort of advanced cancer patients
The term demoralization has been used to describe existential distress and despair of patients with advanced disease. Aim: This study sought to determine whether a cluster of symptoms interpreted as demoralization could be identified and distinguished from a cluster of depressive symptoms. Materials and Methods: As part of the Coping with Cancer Study, a federally funded multi-site study of advanced cancer patients, 242 patients were interviewed on a broad range of mental health parameters related to depression, grief, quality of life, self-efficacy, coping and religiousness/spirituality. Results: A principal components analysis revealed separate depression and demoralization/despair factors. Seven symptoms constituted the demoralization/despair factor: loss of control, loss of hope, anger/bitterness, sense of failure, feeling life was a burden, loss of meaning and a belief that life's meaning is dependent on health and were found to be internally consistent (Cronbach's a = 0.78). Only 14.8% of subjects with "syndromal demoralization" met DSM-IV criteria for Major Depression (MDD); 7.4% for Minor Depression. Of those with MDD only 28.6% had syndromal level demoralization. Prior history of MDD predicted current MDD, but not syndromal demoralization. Demoralization, not MDD, was significantly associated with the patient's reported level of inner peacefulness. When compared with MDD, syndromal demoralization was more strongly associated with wish to live and wish to die and equally predictive of mental health service use. Conclusion: The symptoms of demoralization are distinct from depressive symptoms and appear to be associated with the patient's degree of inner peacefulness.
Keywords: Demoralization, depression, inner peace.
Despair, distress and hopelessness are recognized symptoms in patients at the end-of-life. There is debate in the literature on whether these symptoms reflect a normal response to difficult circumstances, reflect clinical depression, or constitute a syndrome of despair, distress and hopelessness separate from depression.,,, The term demoralization has been suggested for this syndrome although diagnostic criteria have neither been validated through replication studies nor accepted for inclusion in standard psychiatric texts such as the Diagnostic and Statistical Manual of Mental Disorders (DSM). Below we review the history of the term and conduct an empirical test to determine if symptoms of demoralization form a coherent symptom cluster and, if so, whether that cluster is distinct from a depressive symptom cluster in advanced cancer patients.
History of the term demoralization
Despair at the end-of-life was formally recognized when Erikson postulated eight stages of ego development. During the eighth stage, the individual experiences an awareness of mortality, which prompts a process of life-review and stock-taking. The individual who has achieved ego-integrity is able to accept his or her life as it has been led, to feel a sense of connection with people and ideas that are greater than the self and to achieve a sense of peace and acceptance about the approach of death. The individual who has led a life in which s/he has not achieved a firm sense of identity, close relationships and had the opportunity to feel that his/her life has been meaningful is confronted, during this developmental stage, by feelings of despair and fear as mortality looms.
For Erikson, the crisis between ego integrity and despair was a normal part of the developmental process and did not need to be preceded by medical illness. Engel first proposed a syndrome of hopelessness and helplessness occurring in the medically ill; he called it the 'Giving up-Given up' syndrome. A few years later, Frank proposed the construct of demoralization to explain why only a small percentage of persons with psychopathological symptoms sought treatment for their emotional distress. Patients who sought help had a state of mind he called 'demoralization' that interacted with their symptoms causing them to seek help. Demoralization resulted from a "persistent failure to cope with internal or external stresses that the person and those close to him expect him to handle." The characteristic features of demoralization were impotence, isolation, and despair. Depression was thought to be a 'direct expression' of demoralization.
Over time, the term "demoralization" has taken on multiple meanings. Dohrenwend et al , developed a scale to measure non-specific distress in the general population called the Psychiatric Epidemiology Research Interview Demoralization Scale (PERI-D). This scale equates demoralization with global distress and has been used to measure distress in a variety of populations including immigrants and patients in secondary health clinics.
Others have used the term demoralization to describe reactions to illness, thereby incorporating ideas from Engel's proposal of a Giving Up-Given Up complex. This use has been established in the literature only on the basis of clinical observation and case reports.,,,, In a narrative review of the demoralization literature, Clark and Kissane found that demoralization is commonly described as a reaction to a stressful illness or medical event (e.g, cancer diagnosis). It is experienced as existential despair, hopelessness, helplessness and loss of meaning and purpose in life. Contrary to the original definition of Frank, Clark and Kissane's review concluded that demoralization is distinct from depression; depressed patients experience anhedonia, whereas demoralized patients experience subjective incompetence.
Although several possible diagnostic criteria have been proposed for demoralization, it has not been defined in the Diagnostic and Statistical Manual of Mental Disorders (DSM) published by the American Psychiatric Association. Few studies address the distinction between proposed symptoms and criteria for demoralization and depression, or examine potential differences between these syndromes with respect to etiology, course, or treatment response., In one of the few studies to address these issues, Mangelli et al , applied Diagnostic Criteria for Psychosomatic Research (DCPR) derived from variables identified in psychosomatic research to 807 consecutive outpatients recruited from different medical settings (gastroenterology, cardiology, endocrinology and oncology). The DCPR diagnostic criteria are 1) a feeling state characterized by the patient's consciousness of having failed to meet his or her own expectations (or those of others) or being unable to cope with some pressing problems; the patient experiences feelings of helplessness, hopelessness or giving up, 2) the feeling state should be prolonged and generalized (at least one month in duration) and 3) the feeling should closely antedate the manifestations of a medical disorder or exacerbate its symptoms. All three criteria are required for diagnosis. The symptoms must antedate the manifestations of a medical disorder because in this definition, which is based on Engel's idea of the Giving Up-Given Up syndrome, demoralization is thought to be a physiological state often precipitated by a loss, that predisposes a person to become ill. Using this definition, Mangelli found a high prevalence (30.4%) of demoralization in the medically ill. Whereas there was considerable overlap between demoralization and major depression, 69% of those classified as demoralized did not satisfy criteria for major depression. The predictive validity and treatment implications of the syndrome of demoralization as defined by the DCRP have not been established.
Another tool to measure demoralization was recently proposed by Kissane et al . Using factor analysis, he identified five relatively distinct dimensions in 100 patients with cancer: loss of meaning, dysphoria, disheartenment, helplessness and a sense of failure. A subgroup of patients with high demoralization did not meet DSM-IV criteria for major depression suggesting that demoralization is a clinical syndrome distinct from major depression.
The initial aims of our study were to determine whether symptoms consistent with previously reported definitions of demoralization constitute a distinct symptom cluster, or factor and to determine the extent to which these symptoms of demoralization overlap with those of Major Depressive Disorder (MDD). The third part of our study focuses on clinical correlates of the emergent factors to determine if they have different risk factors and out comes that further distinguish them from each other. We hypothesized that if there were a demoralization factor distinct from a depression factor, that it would be associated with the patient's difficulty accepting his/her cancer diagnosis, a sense of inner peacefulness, a wish to live and die. We also hypothesized that lifetime (i.e., prior history) of MDD would be associated with current MDD, but not demoralization, in the advanced cancer patient.
Selection of criteria for demoralization
To select diagnostic criteria we 1) reviewed the literature to identify symptoms associated with the idea of demoralization, 2) selected symptoms thought to best represent the concept of demoralization and 3) used principal components analysis to determine how the chosen symptoms clustered together and whether they were separable from symptoms of depression. Below we describe how and why we selected particular symptoms in our definition of demoralization.
Since the idea of demoralization has changed over time with no consensus definition, we initially included all definitions in the literature to capture all its facets. As suggested by the factor analysis of Kissane et al , we included the symptoms of loss of meaning, dysphoria, disheartenment, helplessness and a sense of failure. From the DCPR, we included the criteria of failing to meet expectations, being unable to cope, helplessness, hopelessness and giving up. We also included the original symptoms of impotence, isolation and despair suggested by Frank and the idea of distress suggested by Dohrenwend et al .
Patients were recruited from 8/1/2002 to 8/25/2005, as part of an ongoing multi-institutional longitudinal evaluation (MH63892, CA106370) of advanced cancer patients and their primary, informal (unpaid) caregivers in the Coping With Cancer Study. Participating sites included the Yale Cancer Center (New Haven, CT), the Veterans Affairs Connecticut Healthcare System Comprehensive Cancer Clinics (West Haven, CT), Memorial Sloan-Kettering Cancer Center (New York, NY), Simmons Comprehensive Cancer Center (Dallas, TX) and the Parkland Hospital Palliative Care Service (Dallas, TX). Approval was obtained from the human subjects committees of all participating centers; all enrolled patients provided written informed consent.
Inclusion criteria for the study were: 1) diagnosis of advanced cancer (presence of distant metastases and failure of first-line chemotherapy); 2) diagnosis at a participating site; 3) age greater than 20 years; 4) identified unpaid, informal caregiver; and 5) adequate stamina to complete the interview. Patient-caregiver dyads in which either the patient or caregiver met criteria for dementia or delirium (by neuro-behavioral cognitive status exam) or did not speak either English or Spanish were excluded. Potentially eligible patients were identified by oncology staff at each site. Trained research staff approached each identified patient to offer participation in the study by telephone call or hospital visit. Once the patient's written informed consent was obtained, medical records and clinicians were consulted to confirm eligibility. Of the 538 eligible patients who were approached for inclusion into the study, 338 patients were enrolled. Of these, 242 patients provided a response to the questions that form the basis for this study. The most common reasons for nonparticipation included "not interested" (N=94), "caregiver refuses" (N=27) and "too upset" (N=18). Compared with participants, nonparticipants were more likely to be white (73.3% of nonparticipants vs. 62.1% of participants, P =0.011) and more likely to be older (mean age, 60.9 yrs vs. 57.1 yrs, P =0.002). Nonparticipants reported more distress (mean, 2.76 vs. 2.40, P =0.005) on a 5-point Likert scale whose extremes ranged from 1 ("minimal/nonexistent") to 5 ("distraught"). They did not differ significantly from participants by gender or education.
Potential indicators of demoralization came from the following scales: helplessness, from the SCID, McGill Quality of Life Questionnaire (MCG) and the Brief COPE (BC); hopelessness, from the Complicated Grief Assessment (CGA); despair, from the MCG; sense of failure, from the SCID and the MCG; loss of meaning, from the CGA and the MCG; inability to cope, from the CGA, the Yale Evaluation of Suicidality (YES), the General Self-Efficacy Scale (GSES) and the Brief Cope; anger/bitterness, from the CGA were added because they would reflect lack of acceptance of the patient's terminal illness; and brooding, self-pity or pessimism from the SCID was added because it reflected hopelessness. (Appendix A)
Depressive symptoms and current and lifetime
MDD were assessed using the Structured Clinical Interview for the DSM-IV (SCID) Axis I Modules. The SCID has high test-retest reliability for these lifetime diagnoses, with an overall weighted K of 0.68.
The NIA/Fetzer Multidimensional Measure of Religiousness/Spirituality for Use in Health Research was included as an assessment of general spiritual experiences. In this questionnaire, patients were asked to rate the statement "I feel deep inner peace or harmony," on a scale of 1-6, where 1="Many times a day" and 6="Never or almost never". For analyses, the responses were stratified such that those with responses >3 ("Most days") were coded "1"="Peaceful"; otherwise responses were coded "2". Steinhauser et al , showed that a one-item assessment of peacefulness was strongly correlated with emotional and spiritual well being, faith and purpose subscales and had broad applicability across different definitions of spirituality.
Other clinical correlates
The wish to live and wish to die items came from the Yale Evaluation of Suicidality (YES) scale, the inability to accept the cancer diagnosis was assessed using the CGA, terminal illness acknowledgment and mental health service use were assessed using measures we have used successfully in other studies.,
We conducted a principal components analysis on the candidate demoralization items and the SCID depression items, in order to 1) confirm that demoralization and depression were separate constructs and 2) identify the items most closely associated with the underlying construct of demoralization. The items that loaded most highly (greater than 40) on the demoralization factor were selected as the demoralization scale. Cronbach's a for the items in the resulting scale was calculated. The demoralization summary score was compared with a continuous score for depression based on the total number of SCID depression items endorsed. In order to compare syndromal demoralization with the diagnosis of depression, we created a dichotomous variable using the cut-point associated with the top 10% of demoralization scores. The upper 10% was chosen because that was roughly the point prevalence of MDD in this sample and we sought to compare similar levels of symptom severity between depression and demoralization. We used logistic regression analyses and correlations to determine the associations of demoralization and MDD with past history of depression, inner peace, wish to live and wish to die and mental health service use.
The principal components analysis revealed two distinct symptom clusters: one, a depression cluster, comprised almost entirely of SCID MDD items and the other a demoralization symptom cluster comprised of the following seven symptoms. Four concern feelings over the last two days: (1) lack of control over one's life, (2) that life was a burden rather than a gift, (3) that life has been worthless, and (4) that life lacks meaning and purpose. The three other symptoms are: (5) feeling anger/bitterness about the cancer diagnosis, (6) that without their health life is empty, and (7) that the future holds no meaning [Table - 1]. The standardized Cronbach's a was 0.78 for the seven demoralization symptoms. Demoralization as defined by these seven symptoms appeared to be relatively distinct from major depression when we compared subjects with a SCID diagnosis of MDD with those subjects who scored in the top 10 percent on our demoralization scale, whom we defined as having syndromal demoralization [Table - 2]. Only 14.8% of subjects with syndromal demoralization met DSM-IV criteria for Major Depression (MDD); conversely, of those with MDD only 28.6% had syndromal demoralization. A Fisher's exact test for major depression and demoralization was not statistically significant (two sided P =0.06) and the Phi coefficient was low (0.14), indicating that MDD and demoralization are two separate constructs. Demoralization was also distinct from Minor Depression, with only 7.4% of demoralized subjects meeting criteria for minor depression [Table - 3].
Logistic regression revealed that prior history of MDD predicted current MDD (O. R.=4.6, 95% C. I.=1.85-9.36, P =0.0006), but not demoralization (O. R. = 0.82, 95% C. I.=0.23-2.89, P =0.75). Further regression analyses demonstrated that syndromal demoralization was significantly associated with a decreased level of inner peace (OR=0.63, 95% C.I.=0.47-0.82, P =0.0008), whereas MDD was not (OR=0.91, 95% C.I.=0.66-1.24, P =0.55).
For wish to live (In light of current circumstances, how strong would you say your wish to live has been? 1=strong to 4=have none), the Spearman correlation coefficient was -0.33 ( P <0.0001) (demoralization) versus -0.17 ( P <0.002) (depression) and for wish to die (In light of your circumstances, how strong has your wish to die been) the Spearman correlation coefficient was 0.35 ( P <0.0001) (demoralization) versus 0.18 ( P <0.0008) (depression). Demoralization was also somewhat more closely associated with inability to accept the cancer diagnosis (Spearman correlation coefficient -0.30 ( P <0.003) (demoralization) versus -0.24 ( P <0.001) (depression) and likelihood of having a post-cancer discussion with the clinician about mental health services (Spearman correlation coefficient 0.25 ( P <0.0001) (demoralization) versus 0.15 ( P <0.005) (depression). Our demoralization scale was associated with similar mental health service use as MDD with a Spearman correlation of 0.20 ( P <0.002) (demoralization) versus 0.15 ( P <0.005) (depression) [Table - 4].
In these analyses we identified a cluster of symptoms that might be interpreted as demoralization. In our data, this cluster of symptoms is distinct from major and minor depression and is significantly associated with wish to die, inversely correlated with wish to live and inner peace. The concept of demoralization has both criterion validity (e.g, predicted lack of inner peace) and clinical utility in that it is more associated with clinical concerns of a wish to die and lack of a wish to live and was associated with discussion of mental health concerns as a cancer patient and associated with similar mental health service use as MDD.
Our symptom cluster is similar to that proposed by Kissane et al , who proposed a five-factor scale: loss of meaning and purpose, dysphoria, disheartenment, helplessness and sense of failure. Each of Kissane's factors is defined by 4-6 questions for a total of 24 items. Some factors, such as loss of meaning and sense of failure, overlap with symptoms in our unitary demoralization scale. Our scale differs from his in that we propose symptoms of life feeling more of a burden than a gift, of loss of control over one's life, (different from one of Kissane's questions about loss of emotional control in general) and of life's meaning being dependent on health. Kissane's factor of dysphoria is made up of five questions (hurt easily, angry, guilty, irritable, regret) one of which, (angry) overlaps with our proposed symptom of angry/bitter, although our symptom reflects the patient's reaction to his/her diagnosis, whereas Kissane's is more general (I am angry about a lot of things). Similarly, Kissane asks a general question about hopelessness (I feel hopeless) as one component of helplessness, whereas our scale asks about hopelessness more specifically in the context of health (I feel that the future holds no meaning or purpose without my health). The dependence on health of three of our symptoms (angry/bitter, hopelessness and meaning) identifies a kind of brittle thinking: the patient's life evaluation depends on health. This kind of thinking is different from general anger, loss of hope, or meaning. The large overlap between Kissane's and our scale provides evidence of independent validation of the construct of demoralization.
Both Kissane's results and those presented in this report suggest that demoralization is separate from depression. Among our subjects with syndromal demoralization 85% did not meet DSM-IV criteria for MDD. Kissane identified 7-14% of patients who were seriously demoralized but not clinically depressed. Our lower rate of overlap with MDD may be because our shorter scale has less potential for confounding with the construct of MDD. Kissane's demoralization items include questions about feeling guilty, hurt easily and loss of emotional control in general (I no longer feel emotionally in control), which may create more overlap with MDD. Furthermore, the distinction in our data between demoralization and minor depression suggests that demoralization does not exist on a continuum with depression but is a separate construct.
When compared to MDD, our demoralization construct was more strongly associated with wish to live and wish to die. Our work supports the results of Chochinov et al, who found that several existential issues including hopelessness, sense of dignity and sense of burden to others were significantly associated with will to live. Past work by Breitbart et al , has also suggested that depression is not the only predictor of desire for hastened death and that hopelessness is independently associated with this desire. Breitbart also found only a modest correlation between depression and hopelessness at the end-of-life (r=0.29), indicating that depression and hopelessness are distinct constructs.
Although the symptoms of demoralization were found to be distinct from DSM-IV criteria for MDD, it is still too early to conclude for certain whether the emergent factor is end-of-life despair or whether the symptoms of demoralization define a depressive state in the context of advanced cancer. It appears that there are problems with the application of DSM-IV criteria for MDD to advanced cancer patients. Aside from the problems associated with the confounding of medical illness with depressive illness, we have found only about 7% of our sample of advanced cancer patients met criteria for MDD. Clinically this seems a large underestimate and, based on our own experience, we find most clinicians use the symptoms that we call demoralization as a basis for deciding whether or not to intervene (e.g., to use antidepressants and other psychosocial interventions). Future research is needed to determine whether demoralization represents existential, end-of-life despair or whether the criteria for what depression really is in advanced cancer patients need to be redefined.
One strength of this work is that we were able to compare demoralization to MDD diagnosed with the SCID, which diagnoses depression using DSM-IV criteria. Another strength is that our questionnaires included a complicated grief assessment, thereby enabling us to differentiate demoralization from symptoms suggestive of complicated grief, although the validity of this construct in dying patients has not been determined. Interestingly, three out of the seven demoralization questions were derived from the ten-item complicated grief assessment suggesting a substantial overlap between a dying patient's grief and demoralization. If demoralization is the despair that Erikson has proposed and (as suggested by our data) is a failure to resolve the final life crisis and find peace with how one has lived one's life, then the individual may be grieving for the aspirations that he can no longer fulfill.
This research was supported in part by the following grants: MH56529 (HGP) and from the National Institute of Mental Health and CA106370 (HGP) from the National Cancer Institute; the American Foundation for Suicide Prevention (LCV); the Center for Psycho-oncology and Palliative Care Research, Dana-Farber Cancer Institute (HGP, LCV).
[Table - 1], [Table - 2], [Table - 3], [Table - 4]