This study is the first to report the effects of four common neuropsychiatric symptoms, comorbid vascular conditions, and subcortical ischemic vascular disease on mortality in a unique large cohort of individuals over age 50 with and without cognitive impairment. We found that depressed mood was significantly associated with elevated mortality risk after adjustment for age, gender, education level, race, cognitive status, and the presence of lacunes, as well as multiple vascular risk factors. Other significant correlates of mortality included cognitive status, older age, lower education level, male gender, and cerebrovascular lacunes. Hypertension, congestive heart failure, history of myocardial infarction, arrhythmia, and coronary artery bypass graft surgery were significantly associated with mortality in the unadjusted analyses. Race and other comorbid medical conditions did not predict mortality.
Our findings of the detrimental effect of depressed mood on mortality are consistent with previous reports of clinical depression associated with social, occupational, and physical impairment and suicide and non-suicide-related mortality (1
). For example, in some studies psychiatric disorders, particularly depression, have been found to predict mortality in various medical conditions (2
). Depression and other psychiatric problems were identified as predictors of mortality in the Baltimore Epidemiologic Catchment Area Study (15-year follow-up) (40
Among depressed patients, anhedonia and hopelessness have been related to mortality (41
). Stroke survivors have a greatly elevated risk for clinically significant depressive symptoms even 2 or more years after the index stroke, independent of functional disability, cerebrovascular risk factors, and prior depressive symptoms (43
). Poststroke depression has not been accurately diagnosed and treated, particularly when silent subcortical ischemic infarctions are involved. With the advantage of MRI, researchers should focus investigations on the association of specific cerebral regions with depressive manifestations and treatment response. Methodological issues such as previous history of depression and the type of depressive manifestation should be considered in evaluating specific risk factors for depression and mortality (6
Depression complicates the outcomes of medical and cognitive diseases. It can cause more severe functional impairment, retard the rehabilitation process, complicate outcomes, and increase mortality risk. It has been hypothesized that depression can lead to worse outcomes for several reasons. First, there may be some direct biological effects, as suggested in cardiovascular data, such as exaggerated platelet aggregation and reduced heart rate variability (44
). Second, depression-related symptoms may prevent patients from participating in their medical care. The presence of anhedonia, apathy, hopelessness, or anergia also may lead to behaviors that discourage family members and caregivers from providing necessary assistance, causing delay or even preventing these patients from receiving optimal care. Third, cognitive symptoms affect how depressed patients perceive and interpret clinical information, which can contribute to further stress and increase depressive symptoms (46
). There are also complex interactions between depression and cognitive impairment in relation to mortality. Depression rates increase with increasing cognitive impairment, and depression and cognitive impairment can have additive effects on mortality (47
). Lastly, the individual symptoms may have an additional deleterious effect because depressed, inactive, withdrawn patients may be susceptible to unexpected complications, such as thromboembolic events and infections (45
), or poor long-term outcomes of diabetes (48
Despite growing evidence regarding the negative impact of depression on mortality, data have begun to illuminate the effects of apathy, anergia, or anhedonia on mortality (7
). In our sample, the presence of depressed mood increased the risk of dying on average by 90%–120%, while anergia and our global behavioral measure of neuropsychiatric symptoms increased the risk of dying on average by 30%–80%, depending on the multivariate model (). The presence of apathy or anhedonia also increased the risk of dying by 20%–70% but did not reach statistical significance after adjustment for covariates (). In older adult populations, the deleterious effect of depressive or other neuropsychiatric symptoms may be obscured by a higher expected mortality rate and multiple competing potential causes of death (46
). An alternative explanation is “selective survival,” where those susceptible to the adverse outcomes of depression died as young adults (42
As expected, the presence of cognitive impairment or dementia, older age, and male gender were associated with a greater risk of mortality and shorter survival ( and Figures and ). This is consistent with other studies reporting on predictors of mortality (51
). We found male gender to be a predictor of mortality, which is consistent with other reports of men having greater mortality rates and shorter life spans than women, attributed in part to a greater prevalence of cardiovascular disease (53
). Depression and hopelessness in men are associated with increased risks of all-cause and cause-specific mortality (54
Similar to other elderly cohorts, our participants had multiple comorbid physical illnesses, particularly cardiovascular and cerebrovascular diseases, which can result in depression as a result of their direct effect on the brain or pathophysiology or of existing disability. In fact, we recruited participants on the basis of subcortical ischemic vascular disease and in the absence of cortical strokes, which is unique to our sample compared with previous prospective naturalistic studies of medically ill participants. Survival was consistently shorter among participants with lacunes and/or with depressed mood in all cognitive groups, with the shortest time to death among participants with both lacunes and depressed mood (Figures and ). Lacunes were associated with an approximate doubling of the mortality risk (). Future analyses of predictors of mortality could provide more refined estimates of the associations of lacunes with mortality by including lacunar size, number, and localization as independent variables.
As described in , hypertension, congestive heart failure, history of myocardial infarction, arrhythmia, and coronary artery bypass graft surgery were significantly associated with mortality. This is consistent with existing data on the effects of serious cardiovascular disease on mortality in medically ill populations (55
). However, associations of neuropsychiatric symptoms with mortality in this cohort were not altered by adjustment for these vascular factors. Greater depressive symptoms are associated with elevated risks of all-cause and, more specifically, cardiovascular disease mortality in men (56
). A large proportion of older women also report levels of depressive symptoms that are significantly related to elevated risks of cardiovascular disease mortality and all-cause mortality, even after controlling for established cardiovascular disease risk factors. Whether early recognition and treatment of subclinical depression will lower cardiovascular disease risk remains to be determined in clinical trials (57
There are limitations to our study. First, our analyses are based on the symptoms reported by caregivers of patients with dementia, and not patients' reports. We combined participants with major and minor depression, as well as those with and without cognitive impairment. Forty percent of the sample had evidence of cerebrovascular disease as indicated by the presence of lacunes on MRI, which is not outside of the range in older adults with cognitive impairment or psychiatric symptoms (10
). Although given the complexity of the operations of multiple clinics and research programs involved in the study, we are not able to estimate the total number of participants screened and excluded, our sample was probably reflective of the population of older adults with comorbid medical and cognitive conditions with cerebrovascular disease who are seen in tertiary referral memory clinics, a group that tends to have higher rates of depressed mood and related neuropsychiatric symptoms.
Our study suggests that early diagnosis of depressed mood as well as other neuropsychiatric symptoms of anergia, apathy, and anhedonia, especially if these are present together, may lead to identification of groups at high risk for early mortality. If present together, cognitive impairment, depressed mood, and subcortical vascular disease increase the risk of mortality. However, we need to improve our understanding of the impact of neuropsychiatric symptoms on morbidity and mortality in individuals with cerebrovascular disease and cognitive impairment in order to develop approaches to the early diagnosis, treatment, and prevention of these significant mortality risk factors. Successful treatment of depression and vascular risk factors is likely to have a positive effect that also applies to rehabilitation, quality of life, and cardiovascular mortality (59
). Identifying and modifying key risk factors are crucial to reducing the morbidity and mortality of cerebrovascular disease. Future studies should evaluate the potential use of MRI for early detection of groups at high risk for depression, cognitive impairment, and mortality. Potential intervention targets should include treatment of depression and dementia as well as reduction of vascular risk factors in order to reduce mortality in this vulnerable population.