Stroke Survivors Who Score Below Threshold on Standard Depression Measures May Still Have Negative Cognitions of Concern
Background and Purpose— There has been an increase in screening for depression in the physically ill. We explored whether important negative cognitions may be missed by conventional approaches to screening for depression in 2 independently conducted stroke studies with similar methods.
Methods— The Auckland Regional Community Stroke (ARCOS) study was a prospective, population-based stroke incidence study conducted in Auckland, New Zealand, for 12 months in 2002 to 2003. The Stroke Outcomes Study was a prospective, hospital cohort study conducted in Leeds and Bradford, United Kingdom, for 33 months in 2002 to 2005. Symptoms of abnormal mood were assessed at 6 months in ARCOS with a single simple question, “Do you often feel sad and depressed?” and the 28-item General Health Questionnaire administered as part of a structured interview and in the Stroke Outcomes Study with the 28-item General Health Questionnaire and a single question about depressed mood taken from the Present State Examination.
Results— Mood data were available at 6 months from 770 ARCOS and 492 Stroke Outcomes Study participants. A significant proportion (up to 28%) of people who did not meet study criteria for depression reported important negative cognitions such as hopelessness, worthlessness, or suicidality. People who were older, dependent in activities of daily living, or not partnered were more likely to report negative cognitions.
Conclusions— Important negative cognitions, including suicidal thoughts, may be missed when people are screened for depression after stroke. Screening alone is not an adequate substitute for a sensitive exploration of the psychological impact of stroke on the survivor.
Clinicians and researchers are encouraged to screen to detect the substantial proportion of people1 who will experience depression at some point after stroke. For example, the National Stroke Sentinel Audit2 monitors rates of screening for depression during admission after stroke. However, depression is a complex condition that can be conceptualized in different ways, which in turn, govern how it is recognized, diagnosed, and managed.
Most commonly, depression is considered a syndrome. A clinical diagnosis can only be made after a particular constellation of signs and symptoms has been elicited through a semistructured psychiatric interview. For a DSM-IV diagnosis,3 an individual must have experienced 1 of the “core symptoms” of “depressed mood” or “a loss of interest or pleasure” nearly every day, most of the day, in a 2-week period. In addition, the same individual must experience at least 4 additional symptoms of depression, such as changes in weight or sleep or feeling restless, slowed down, fatigued, or worthless. However, if there is no “depressed mood” or “loss of interest or pleasure,” a diagnosis of depression will not be achieved regardless of how many symptoms are endorsed.
Depression can also be considered an accumulation of depressive symptoms, none of which is privileged in the way that they are in the syndromal approach. These symptoms can be grouped into 4 broad categories. The first consists of affective symptoms such as depression, anxiety, irritability, and apathy, with the second category including behavioral symptoms such as tearfulness, reassurance seeking, inertia, and social withdrawal. These symptoms are listed in most structured questionnaires such as the Centre for Epidemiological Studies Depression Scale4 or the Geriatric Depression Scale.5 The third category includes somatic symptoms such as insomnia, pains, tension, and weight loss, which have been deliberately removed from some structured questionnaires (eg, Hospital Anxiety and Depression Rating Scale6) because of a claim that these are also symptoms of physical illness that may then be misattributed to symptoms of depression. The last category consists of important negative cognitive symptoms such as hopelessness, helplessness, and suicidal thoughts. These symptoms are also listed in most structured questionnaires, (eg, the 28-item General Health Questionnaire (GHQ-28)7 and the Beck Depression Inventory,8) but not all (eg, the Hospital Anxiety and Depression Rating Scale).6 The number of symptoms endorsed is counted, and scores above a measure-specific threshold are said to indicate a likely case of depression.
Finally, depression may also be conceptualized as a group of clinically important negative cognitions such as worthlessness or hopelessness that may be present without the traditional “core symptoms” of depressed mood or loss of interest or pleasure, although they often coexist with somatic symptoms. This presentation of so-called depression without sadness9 is thought to be more common in the elderly. Although more attention has been given to depression as a risk for mortality in the elderly,10–12⇓⇓ there is evidence that specific negative cognitions may be relevant, independent of any associated depressive diagnosis, for mortality9,13,14⇓⇓ and subsequent physical and mental morbidity.13,15,16⇓⇓
Screening for depression has been largely oriented toward measuring depressive symptom burden in its own right or as a means of identifying people likely to meet syndrome-based diagnostic criteria. Recently a variety of short screening tools17–19⇓⇓ has been developed, the briefest of which involve just 1 or 2 questions that ask respondents directly about the experience of depression. For example, the National Institute for Health and Clinical Excellence in the United Kingdom20 recommends that patients seen in primary care are asked 2 screening questions: “During the last month, have you been feeling down, depressed, or hopeless?” and “During the last month, have you often been bothered by having little interest or pleasure in doing things?” Not all depression screening tools contain negative cognitions for endorsement or only include (as in the Patient Health Questionnaire 2-item21) reference to 1 such symptom as part of a composite question.6,17–19⇓⇓⇓
There has been relatively little exploration of which mood-related symptoms are missed when we screen for depression after stroke. We were especially interested in negative cognitions in physical illness, and we therefore examined the relation between these negative cognitions and responses to a single question about depression and to a standardized validated questionnaire in 2 independently conducted (with similar methods) epidemiologic stroke cohorts. Our question was whether using these approaches to identifying depression might lead to neglect of potentially-important negative cognitions in patients who would not meet traditional criteria for a depression diagnosis.
The Auckland Regional COmmunity Stroke (ARCOS) study has been described elsewhere.22 In brief, ARCOS used a prospective, population-based register to ascertain all cases of acute new or recurrent stroke that occurred among adults in the usually resident population of Auckland during a 12-month period from 2002 to 2003. Each participant was interviewed as soon as possible after stroke. All survivors were reinterviewed at 6 months when depressive symptoms were assessed with the Yale question, “Do you often feel sad or depressed?”23 and symptoms of abnormal mood were assessed in cognitively competent participants (those scoring ≥7 on the Hodkinson Mental Test24) with the GHQ-28 (standard 0, 0, 1, 1 scoring).7
The Stroke Outcomes Study (SOS2) was a prospective, observational cohort study of the impact of depressive symptoms on outcomes for stroke patients. Patients from 2 large hospitals were recruited after new or recurrent acute stroke, from Trusts in Leeds and Bradford in the United Kingdom, between July 2002 and March 2005.25 The patients were screened with the Mini-Mental State Examination26 and interviewed within 2 to 6 weeks (baseline assessment) after their index stroke episode, depending on their cognitive and physical state. They were interviewed again at 6 months. Participants were asked to complete a range of outcome measures that provided information on functional status and psychological well-being. The measures of mood included the GHQ-287 and the Present State Examination (PSE).27
The GHQ-28 is a broad measure of psychiatric morbidity that includes a subscale for severe depression. The GHQ-28 is probably the most extensively tested scale for reliability, validity, and sensitivity to change across the world.28 Each symptom/question has 4 possible answers: “not at all,” “no more than usual,” “rather more than usual,” and “much more than usual,” Scores range from 0 (the best possible) to 28 (the worst possible) on standard scoring (0,0,1,1).7 We used the recommended threshold of >11 to indicate caseness in those with comorbid physical illness.29 A negative cognition or suicidal thought was considered endorsed if a response was other than “not at all” on the GHQ-28 depression subscale.
The PSE consists of standardized questions administered by a trained interviewer and can yield output consistent with the diagnostic criteria of the DSM3 and ICD.30 It includes question 18, “How have you felt in your mood in the past month? Do you keep reasonably cheerful or have you been depressed or low-spirited recently?” which is equivalent to the Yale question.
Depressive symptoms were assessed in ARCOS and SOS2 in cognitively competent participants (no proxy data are reported) who were able to respond for themselves by trained nursing and medical interviewers. The Auckland ethics committee approved the ARCOS study; the Leeds Teaching Hospital Trust and the Bradford Hospitals Trust local research ethics committees approved the SOS2 study, and written, informed consent was obtained from all participants.
Demographic data from ARCOS and SOS2 cohorts and associations between negative cognitions and demographic characteristics were compared by χ2 or 2-sample t tests, where appropriate. To assess the presence of important cognitions in participants who did not answer the single Yale or PSE depression question positively, we present GHQ-28 item responses for those who were positive GHQ-28 cases and those who were not a case on the GHQ-28.
Characteristics of the Study Populations
Complete mood data were available at 6 months from 770 ARCOS and 492 SOS2 participants. The demographic characteristics of the study populations are shown in Table 1. ARCOS and SOS2 participants were similar in age (69±13 years), sex, and marital status. More ARCOS participants were completely independent in activities of daily living, whereas more SOS2 participants were dependent on others for help with activities of daily living and met the criteria for abnormal mood. These differences probably reflect that fact that ARCOS recruited a community sample whereas the majority of SOS2 subjects were recruited in hospital.
Abnormal Mood at 6 Months in the Study Group
Overall, 56 (7%) ARCOS participants and 61 (12%) SOS2 participants met the study criteria for caseness on the GHQ-28 (score >11). Consequently, GHQ-28 mean scores (±SD) were significantly lower in ARCOS (3.57±4) participants than in SOS2 participants (4.61±6; t=3.39, P<0.001). One hundred fifty-nine (21%, ARCOS) and 129 (26%, SOS2) participants responded positively to the single simple question (see Table 1). Only 40 (5%, ARCOS) and 50 (10%, SOS2) participants met the criteria for caseness on the GHQ-28 and responded positively to the single simple question (see Table 2).
Table 3 shows the proportion of participants who endorsed negative cognitions or suicidal thoughts on the GHQ-28 but did not meet the criteria for caseness on the GHQ-28 (score >11) and answered negatively to the Yale question or question 18 of the PSE. A significant proportion (up to 28%) of people who did not meet study criteria for depression on 1 or more measures reported important negative cognitions, such as hopelessness, worthlessness, or suicidality. Being older was associated (P<0.05) with thoughts of being worthless, that life was hopeless or not worth living, or wishing themselves dead for ARCOS and SOS2 participants. Being dependent in activities of daily living was associated with feeling that life was hopeless or not worth living (ARCOS) and wishing themselves dead and away from it all (ARCOS and SOS2). Not being married or partnered was associated with thoughts of being worthless, hopeless, and nerves being bad (SOS2) or life wasn’t worth living or wishing themselves dead (ARCOS; data not presented).
These 2 independently conducted stroke studies have demonstrated that a significant proportion (up to 28%) of patients with negative cognitions including suicidal thoughts may be missed by researchers and clinicians who rely on screening methods such as a single question or a standardized self-report measure, alone or in combination, to detect depressive illness. Symptoms listed in the GHQ-28 depression subscale are important in their own right, because they are markers of significant distress, even in those who do not have sufficient other symptoms to warrant a psychiatric diagnosis.
We suggest that clinicians and patients will benefit from reviewing responses to individual questions on mood rating scales together, regardless of whether scores are below a prespecified cutpoint or whether people screen positive or negative for depression on a single question. Reviewing individual item responses may provide the opportunity for frank but sensitive discussions between clinician and patients about endorsed negative cognitions, including suicidal thoughts. This process may also facilitate establishment of a “watchful waiting” period, as recommended in the National Institute for Health and Clinical Excellence guidelines31 to prevent unnecessary treatment of mild symptoms of abnormal mood that may resolve with no intervention.
A strength of our data is the consistency of the phenomenon in 2 different studies that used similar screening questions but different methods of assessment. Both studies made strenuous efforts to avoid bias in sampling and follow-up, and a standardized assessment of mood in participants with stroke was used (GHQ-287), which is an established measure for assessing depressive symptoms after stroke,32,33⇓ although this was interviewer administered in one study (ARCOS) and self-completed in the other (SOS2). It should also be noted that these were post hoc analyses. As a result, we have been unable to reinterview participants to explore further the meaning of negative cognitions in those who answered questions about depressed mood negatively.
The likelihood of experiencing depression after stroke is high,1 and many patients do not seek help for or even recognize abnormal mood symptoms of clinical concern.34 Screening for depression is therefore desirable, but our conclusion is that alone it is not likely to be an adequate substitute for a sensitive exploration of the psychological impact of severe disease on the survivor. Clinically, a single question about depression, or a high score on a depression questionnaire, may be useful and is certainly better than asking nothing, but it is unlikely to elicit important information about certain aspects of the patient’s mental state, such as a sense of worthlessness or hopelessness or thoughts that life is not worth living. We need more research to understand the origins of these cognitions and their impact on outcomes.
Sources of Funding
The ARCOS study was funded by the Health Research Council of New Zealand, and the SOS2 Study was funded by the Stroke Association, UK. During the completion of this work, Maree Hackett was a recipient of a University of Auckland Senior Health Research Scholarship (2003–2005) and a National Health and Medical Research Council Public Health (Australia) Fellowship (2006–2009), and Craig Anderson was a recipient of a National Health and Medical Research Council Research Fellowship (2008–2012). None of these funding bodies had a role in the conduct or reporting of this report.
- Received October 29, 2009.
- Revision received November 29, 2009.
- Accepted December 14, 2009.
- ↵Hackett ML, Yapa C, Parag V, Anderson CS. The frequency of depression after stroke: a systematic review of observational studies. Stroke. 2005; 36: 1330–1340.
- ↵Intercollegiate Stroke Working Party. National Sentinel Stroke Clinical Audit. London: Royal College of Physicians; 2009.
- ↵American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. Washington DC: American Psychiatric Association; 1994.
- ↵Goldberg DP, Williams P. A User’s Guide to the General Health Questionnaire. Windsor: NFER-Nelson; 1988.
- ↵Dewey ME, Davidson IA, Copeland JRM. Expressed wish to die and mortality in older people: a community replication. Age Ageing. 1993; 22: 109–113.
- ↵Sone T, Nakaya N, Ohmori K, Shimazu T, Higashiguchi M, Kakizaki M, Kikuchi N, Kuriyama S, Tsuji I. Sense of life worth living (Ikigai) and mortality in Japan: Ohsaki Study. Psychosom Med. 2008; 70: 709–715.
- ↵Everson SA, Goldberg DE, Kaplan GA, Cohen RD, Pukkala E, Tuomilehto J, Salonen JT. Hopelessness and risk of mortality and incidence of myocardial infarction and cancer. Psychosom Med. 1996; 58: 113–121.
- ↵Arroll B, Khin N, Kerse N. Screening for depression in primary care with two verbally asked questions: cross sectional study. BMJ. 2003; 327: 1144–1146.
- ↵Mahoney J, Drinka TJK, Abler R, Gunter-Hunt G, Matthews C, Gravenstein S, Carnes M. Screening for depression: single question versus GDS. J Am Geriatr Soc. 1994; 1006–1008.
- ↵Mitchell AJ, Coyne JC. Do ultra-short screening instruments accurately detect depression in primary care? a pooled analysis and meta-analysis of 22 studies. Br J Gen Pract. 2007; 57: 144–151.
- ↵National Institute for Clinical Excellence. Depression: management of depression in primary and secondary care. 2004;www.nice.org.uk/CG023. Last accessed April 2009.
- ↵Anderson C, Carter K, Hackett M, Feigin V, Barber PA, Broad JB, Bonita R, on behalf of the Auckland Regional Community Stroke (ARCOS) Study Group. Trends in stroke incidence in Auckland, New Zealand, during 1981 to 2003. Stroke. 2005; 36: 2087–2093.
- ↵Watkins C, Daniels L, Jack C, Dickinson H, van den Broek M. Accuracy of a single question in screening for depression in a cohort of patients after stroke: comparative study. BMJ. 2001; 323: 1159.
- ↵Hodkinson HM. Evaluation of a mental test score for assessment of mental impairment in the elderly. Age Ageing. 1972; 1: 233–238.
- ↵Hill KM, West RM, Hewison J, House AO. The Stroke Outcomes Study 2 (SOS2): a prospective, analytic cohort study of depressive symptoms after stroke. BMC Cardiovasc Dis. 2009; 1: 22.
- ↵Wing J, Cooper J, Sartorius N. Measurement and Classification of Psychiatric Symptoms: An Instruction Manual for the PSE and Catego Program. Cambridge, England: Cambridge University Press; 1974.
- ↵Bowling A. Measuring Disease. Buckingham: Open University Press; 1995.
- ↵Bridges KW, Goldberg DP. The validation of the GHQ-28 and the use of the MMSE in neurological in-patients. Br J Psychiatry. 1986; 148: 548–553.
- ↵World Health Organization. The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva: World Health Organization; 1992.
- ↵National Institute for Health and Clinical Excellence. Depression (amended): management of depression in primary and secondary care. London: National Institute for Health and Clinical Excellence; 2007: N1237.
- ↵O'Rourke S, MacHale S, Signorini D, Dennis M. Detecting psychiatric morbidity after stroke: comparison of the GHQ and the HAD Scale. Stroke. 1998; 29: 980–985.
- ↵World Health Organization. The World Health Report 2001: Mental Health: New Understanding, New Hope. Geneva: World Health Organization; 2001.