PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
J Gerontol A Biol Sci Med Sci. Author manuscript; available in PMC 2010 September 15.
Published in final edited form as:
J Gerontol A Biol Sci Med Sci. 2003 May; 58(5): M461–M467.
PMCID: PMC2939722
NIHMSID: NIHMS173300

Additive Effects of Cognitive Function and Depressive Symptoms on Mortality in Elderly Community-Living Adults

Abstract

Background

Poor cognitive function and depressive symptoms are common in the elderly, frequently coexist, and are interrelated. Both risk factors are independently associated with mortality. Few studies have comprehensively described how the combination of poor cognitive function and depressive symptoms affect the risk for mortality. Our aim was to examine whether the combination of varying levels of cognitive function and depressive symptoms affect the risk of mortality in community-living elderly adults.

Methods

We studied 6301 elderly adults (mean age, 77 years; 62% women; 81% white) enrolled in the Asset and Health Dynamics Among the Oldest Old (AHEAD) study, a prospective study of community-living participants conducted from 1993 to 1995. Cognitive function and depressive symptoms were measured using two validated measures developed for the AHEAD study. On each measure, participants were divided into tertiles representing the best, middle, and worst scores, and then placed into one of nine mutually exclusive groups ranging from best functioning on both measures to worst functioning on both measures. Mortality rates were assessed in each of the nine groups. Cox proportional hazards models were used to control for potentially confounding characteristics such as demographics, education, income, smoking, alcohol consumption, comorbidity, and baseline functional impairment.

Results

During 2 years of follow-up, 9% (548) of the participants died. Together, cognitive function and depressive symptoms differentiated between elderly adults at markedly different risk for mortality, ranging from 3% in those with the best function on both measures to 16% in those with the worst function on both measures (p < .001). Furthermore, for each level of cognitive function, more depressive symptoms were associated with higher mortality rates, and for each level of depressive symptoms, worse cognitive function was associated with higher mortality rates. In participants with the best cognitive function, mortality rates were 3%, 5%, and 9% in participants with low, middle, and high depressive symptoms, respectively (p < .001 for trend). The corresponding rates were 6%, 7%, and 12% in participants with the middle level of cognitive function (p < .001 for trend), and 10%, 13%, and 16% in participants with the worst level of cognitive function (p < .001 for trend). After adjustment for confounders, participants with the worst function on both measures remained at considerably higher risk for death than participants with the best function on both measures (adjusted hazard ratio, 3.1; 95% confidence interval, 2.0–4.7).

Conclusions

Cognitive function and depressive symptoms can be used together to stratify elderly adults into groups that have significantly different rates of death. These two risk factors are associated with an increased risk in mortality in a progressive, additive manner.

In older persons, adverse outcomes, including mortality, are often better explained by the cumulative and combined impact of multiple conditions that increase a person’s vulnerability rather than by the impact of any one condition alone (1,2). Poor cognitive function and depressive symptoms independently increase mortality in the elderly (319). However, although poor cognitive function and depressive symptoms are interrelated and often coexist (20), it is unclear how the combination of poor cognitive function and depressive symptoms contribute to increase an older person’s risk for mortality.

Many prior studies of the association of poor cognitive function and depressive symptoms with mortality have focused on one of these risk factors to the exclusion of the other (310). A few studies that have measured both cognitive function and depression (1119) have provided only limited insight into how the combination of these conditions affects mortality. Most studies have examined the effect of either cognitive impairment or depression on mortality, and viewed the other condition only as a confounder in adjustment models. In some instances, controlling for one of these two risk factors eliminates the effect of the other (21,22). Also, some studies have not considered the full range of depressive symptoms and cognitive impairment, and have assessed only the impact of major depression and advanced dementia (7,23). In addition, some studies have been limited to special subgroups such as the medically ill or hospitalized elderly adults (11,12,14).

To better understand the combined effect of poor cognitive function and depressive symptoms on mortality, we examined survival in participants in the Asset and Health Dynamics Among the Oldest Old (AHEAD) study, a population-based study of more than 6000 community-dwelling elderly adults. The aims of this study were to determine whether cognitive function and depressive symptoms are independently associated with mortality after adjustment for each other and potential confounders such as demographic characteristics, education, income, smoking, alcohol consumption, comorbidity, and baseline functional impairment. In addition to assessing statistical interaction, our analytic approach assesses how varying levels of cognitive function and depressive symptoms contribute to mortality over 2 years.

Methods

Study Population

We studied participants in the AHEAD study, a nationally representative prospective population-based study of community-dwelling adults aged 70 years or older at the baseline interview in 1993 (N = 7447) (24). A full description of the sampling and weighting procedures used in the AHEAD study has been given previously (25). Most respondents aged 70 to 79 years (74%) were interviewed by telephone, and most respondents aged 80 years and older were interviewed in person. The overall survey response rate was 80%, and the response rate did not differ significantly for those interviewed by phone compared with those interviewed in person (26).

For the current study, we excluded the 791 participants for whom a proxy respondent was used (N = 6656). Proxy respondents were used when the subject declined an interview or was too ill to be interviewed. An additional 291 were missing cognitive information, 12 were missing depressive symptom information, and 43 were missing covariate information. Therefore, the analytic sample for this study was 6310 elderly adults (95% of self respondents).

Cognitive Function and Depressive

Symptoms Measurement

Cognitive function was measured at baseline with a 35-point scale developed for the AHEAD study (COG) (27,28). This scale shares questions with the Mini-Mental State Examination (29) and with the Telephone Interview for Cognitive Status (30). Evidence supporting the validity of this scale has been previously published (27,28,31). The scale includes questions that test memory, calculation/attention, and orientation. Memory was tested using an immediate and delayed free-recall test in which 10 common nouns were recited. The participant was asked to verbally recall as many of the nouns as possible immediately and after a 5-minute delay. Both the immediate and delayed recall conditions were scored between 0 and 10, with a score of 10 reflecting perfect recall of all 10 words. Calculation/attention was assessed using the serial 7’s test in which participants start with 100 and consecutively subtract 7 five times. Participants were given 1 point for each correct subtraction for a total of 5 points. Additional assessment included counting backwards from 20 to 10 and from 86 to 76, and naming the month, day, year, and day of the week; the object used to cut paper (scissors); the plant that lives in the desert (cactus); and the president and vice president of the United States, for a total of 10 points. An overall cognitive score was calculated as the sum of all test values for a total of 35 points, with lower scores indicating worse cognitive function (31). We grouped the individuals based on tertiles of the cognitive function scale score: worst [0–17], middle [18–22], and best [23–35].

Depressive symptoms were assessed using eight items from the Center for Epidemiologic Study Depression scale (CES-D8) (32). The CES-D has been shown to reliably measure depressive symptoms in older adults, even in people with mild cognitive impairment (33). A modified version of the CES-D has been used in other population-based studies of the elderly (21), and the CES-D8 has a similar factor structure to the CES-D20 scale (34). Specifically, participants were asked if the following feelings occurred during the prior week (yes or no): depressed mood, loneliness, sadness, happiness (reverse scored), that everything was an effort, that sleep was restless, could not “get going,” and not having a lot of energy. Each question was assigned 1 point for a total of 8 points, with a high score indicating more depressive symptoms. We grouped individuals based on approximate tertiles of CES-D8 score: worst (3–8), middle (1–2) and best (0).

Mortality

We assessed mortality over 2 years using the AHEAD survey follow-up procedures (35). Deaths were confirmed by checking the U.S. National Death Index. Survival time was defined as the number of days between the baseline interview and the date of death, or December 31, 1995, at which point surviving participants were censored.

Other Measurements

We also considered a series of variables, based on prior literature, that could confound the associations between poor cognitive function and depressive symptoms, and mortality. These included demographic characteristics (age, gender, race, and marital status); years of education; total net worth of less than $38,300 (lowest tertile); self-reported smoking and alcohol consumption; body mass index (lowest tertile); and self-reported history of comorbid medical conditions (presence of hypertension, diabetes, cancer, heart disease, lung disease, arthritis, stroke, vision problems, hearing problems); we also determined whether participants reported needing assistance in any of the five activities of daily living (ADL dependence) (36).

Statistical Analyses

All statistical analyses used the AHEAD sampling and design weights to account for the study’s complex design. The main objective of this study was to determine the relative contributions of poor cognitive function and depressive symptoms to mortality. For our bivariate analyses, we used unadjusted survival analysis using the method of Kaplan-Meier to determine whether cognitive function tertiles and depressive symptom tertiles were associated with mortality. To describe the contributions of cognitive function and depressive symptoms as predictors of mortality, we classified participants into nine mutually exclusive exposure groups, representing each combination of the best, middle, and worst tertile on cognitive function and depressive symptoms. We then used unadjusted Cox proportional hazards regression models to examine the relationship between the nine mutually exclusive exposure groups and survival time. Next, we again used Cox proportional hazards regression to examine the relation between the nine exposure groups and survival time after adjusting for significant (p < .10) confounders. Adjusted models included age, gender, education, race, total net worth, marital status, comorbid conditions (presence of high blood pressure, diabetes, cancer, lung disease, and heart disease, as well as the total number of comorbid conditions), body mass index, and ADL dependence. For these analyses, the outcome was survival time censored at December 31, 1995, and the independent variables were the nine mutually exclusive exposure groups. Finally, to determine if there was an interaction between depressive symptoms and cognitive function, we then performed a survival analysis with the CES-D8 scale, the COG scale, and an interaction term (CES-D8 × COG).

Results

The mean age of participants at baseline was 77 years (range, 69–103 years); 62% were women, and 19% were nonwhite (Table 1). Participants had a mean of 2.6 comorbid conditions, of which the most frequent was hypertension (51%). Ten percent were dependent in 1 or more ADLs at baseline. Median follow-up was 2.0 years. Depressive symptom score and cognitive function score were modestly associated with each other (r = .22, p < .001). Participants with the worst cognitive function and depressive symptoms were older and had a higher average number of comorbid conditions as compared with participants in the best tertile on each risk factor.

Table 1
Baseline Characteristics of the 6310 AHEAD Participants

Participants in the worst cognitive function tertile at baseline had higher rates of mortality than participants in the middle and best tertiles (13%, 8%, and 5%, respectively, p < .001 for trend) (Figure 1). Similarly, participants in the worst depressive symptom tertile had higher rates of mortality than participants in the middle and best tertiles (13%, 9%, and 6%, respectively, p < .001 for trend).

Figure 1
Crude 2-year mortality rates by tertiles of cognitive function scores and depressive symptom scores in AHEAD participants (N = 6310). Symptom tertile is given on the x axis; percentage of participants is given on the y axis. AHEAD = Asset and Health Dynamics ...

When participants were categorized into the nine mutually exclusive groups based on cognitive function and depressive symptom tertiles, mortality rates varied markedly. We observed a fivefold difference across the categories (ranging from 3% in participants in the best depressive symptoms and cognitive function tertiles to 16% in participants in the worst depressive symptoms and cognitive function tertiles) (Figure 2). For all levels of cognitive function, more depressive symptoms were associated with higher mortality (p < .001 for trend for all tertiles). Similarly, for all levels of depressive symptoms, worse cognitive function was associated with higher mortality (p < .001 for trend for all tertiles).

Figure 2
Additive effects of poor cognitive function and depressive symptoms on 2-year mortality in AHEAD participants (N = 6310). Tertile of depressive symptoms is given on the x axis; tertile of cognitive function, on the y axis; and 2-year mortality rate, on ...

After adjustment for demographic characteristics, comorbid conditions, and functional status, cognitive function and depressive symptoms both remained associated with higher mortality rates (p < .001 for trend for both depressive symptoms and cognitive function) (Table 2). After adjustment, participants in the worst tertiles for both cognitive function and depressive symptoms had a threefold higher mortality rate than participants in the best tertiles for both cognitive function and depressive symptoms (hazard ratio [HR], 3.1; 95% confidence interval [CI], 2.0–4.7). Across all three depressive symptom tertiles, worse cognitive function was independently associated with higher rates of death (least depressive symptoms, p = .001 for trend; middle depressive symptoms, p < .001 for trend; most depressive symptoms, p = .06 for trend). Across all three cognitive function tertiles, more depressive symptoms were independently associated with higher rates of death (best cognitive function, p = .07 for trend; middle cognitive function, p = .12 for trend; worst cognitive function, p = .04 for trend). Using the CES-D8 and COG scales in a model of simple statistical interaction, the interaction term (CES-D8 × COG) was not significant (p = .10).

Table 2
Hazard Ratios for 2-Year Mortality* by Depressive Symptom Score and Cognitive Function Score in AHEAD Participants (N = 6310)

Discussion

This population-based study of more than 6000 elderly adults found that poor cognitive function and more depressive symptoms are independently associated with increased risks of death over 2 years. The combination of poor cognitive function and depressive symptoms increased mortality in a progressive, additive fashion. Depressive symptoms and cognitive function together defined elderly adults at markedly different 2-year death rates, ranging from 3% to 16%. After adjusting for numerous potential confounders, older adults with the worst cognitive function and depressive symptoms had 3 times the risk of death compared with older adults with the best scores for both measures.

Although many prior studies have examined the relation between either cognitive function or depressive symptoms and mortality, few studies have examined the impact of both of these risk factors in combination (310). Most prior studies that examined both risk factors found that either cognitive impairment or depressive symptoms are associated with mortality, although the magnitude of the association has varied considerably, depending on the differences in definitions of cognitive impairment and depressive symptoms, and the length of follow-up (1119). Some studies have suggested that the association between depression and mortality may not be significant after the adjustment for confounders. For example one study did not report a relationship in controlled analyses (21). Similarly, one recent review showed that not all studies report an association between depression and mortality, especially if the study adjusted for multiple confounders (22).

A few prior studies did not examine the effects of a wide range of cognitive function and depressive symptoms. However, in the community, the spectrum of cognitive function and depressive symptoms includes a wide range of functioning. Our analytic approach separated individuals with high, middle, and low depressive symptoms and cognitive function scores. This approach is useful for describing the additive effects of cognitive impairment and depressive symptoms in a clinically meaningful manner beyond an approach that uses the scores as a continuum. This separation allowed us to quantify that for each level of cognitive function, greater degrees of depressive symptoms were associated with a higher rate of mortality. Similarly, for each level of depressive symptoms, worse cognitive function was associated with a higher rate of mortality. Thus, our study adds to prior studies by demonstrating that cognitive function and depressive symptoms increase mortality risk over a wide range of symptomatology.

Further, our study included more extensive adjustment for potential confounders of the relation between poor cognitive function, depressive symptoms, and mortality than previous studies. This is important because participants with poor cognitive function or depressive symptoms generally have more comorbid illness, more functional impairment, worse socioeconomic status, more sensory impairment, and higher rates of smoking, any of which could partially confound the relationship with mortality. We have demonstrated that these two risk factors are independent of each other, as well as independent of the other confounders we assessed, in their contribution to mortality. However, it is possible that if we had controlled for a higher-level measure of functional status, such as instrumental ADLs, the relationship between cognitive function, depressive symptoms, and mortality may have been reduced as instrumental changes may be the mechanism by which these two risk factors affect mortality.

Our results are consistent with a common theme in geriatric medicine: the risk for poor outcomes is best explained by the additive effects of two or more risk factors acting together, rather than by any single risk factor acting alone (2,37). For example, Gill and colleagues (37) have described how the additive effects of physical impairment and cognitive impairment contribute to the risk for ADL decline in a progressive, additive manner. Similarly, Tinetti and colleagues (2) have demonstrated that the additive effects of risk factors encompassing multiple domains best explain the risk for geriatric syndromes such as falls, incontinence, and functional dependence. Although depressive symptoms and cognitive impairment are interrelated and often coexist, our study demonstrates that an older person’s risk for death is better explained by considering the additive effects of both risk factors, than by considering either risk factor alone.

To understand the possible mechanism explaining why poor cognitive function and depressive symptoms increase the risk of mortality, one can consider recent changes in the view of the complex relationship between depression and dementia. Until recently, much emphasis was placed on the importance of distinguishing between depression and dementia. It was commonly thought that depression could often masquerade as dementia and that some elderly adults with cognitive impairment and depression had “pseudodementia” (38). More recent conceptualizations have expanded this view by not only considering the possibility that depression could masquerade as cognitive impairment, but also by emphasizing the need to view cognitive impairment and depressive symptoms as coexisting conditions. Although there may be some elderly adults in whom cognitive impairment is entirely explained by depressive symptoms, it is probably more common that depressive symptoms are a complicating comorbidity in patients with cognitive impairment, or that depressive symptoms are an early predictor of impending cognitive loss (39).

These alternative frameworks for considering the relation between poor cognitive function and depressive symptoms suggest three plausible mechanisms by which these two risk factors could both increase the risk for mortality. First, it is possible that it is primarily poor cognitive function that predicts mortality, and depressive symptoms only seem to predict mortality because depressive symptoms predict subsequent cognitive impairment (20). Second, it is possible that it is primarily depressive symptoms that predict mortality, and poor cognitive function only seems to predict mortality because participants with cognitive function are likely to be depressed as a result of their poor cognitive function. Third, it is possible that poor cognitive function and depressive symptoms each increase the risk for mortality independently and that both have true etiologic roles in increasing the risk for mortality.

Although our results do not definitively distinguish between these three possibilities, we believe they are most consistent with the third hypothesis. First, poor cognitive function and depressive symptoms contribute almost equally to the risk for mortality. If depressive symptoms only predicted mortality because they were a predictor of future poor cognitive function, one would expect the relation between depressive symptoms and mortality to have been smaller for the worst levels of cognitive function. Conversely if poor cognitive function acted via depressive symptoms, the effect would have been smaller in the groups with the worst depressive symptoms. Instead the relation between depressive symptoms and mortality is consistent across all levels of cognitive function, and the relation between cognitive function and mortality is consistent across all levels of depressive symptoms. This finding is most consistent with the hypothesis that both depressive symptoms and cognitive impairment contribute to an increased risk for mortality.

The primary strength of this study is that it is representative of a population-based sample of more than 6000 older adults in whom cognitive function and depressive symptoms were measured with validated scales at baseline. Further, we adjusted for many confounders that could be associated with cognitive function and depressive symptoms, making it less likely that our findings could be explained by greater comorbid illness and functional impairment. However, a few methodologic considerations deserve comment. First, although the scales used to measure cognitive function and depressive symptoms are not widely used, they underwent extensive validity testing and are strongly correlated with commonly used instruments (27,28,34). Second, since no clinical measurements of cognitive impairment and depressive symptoms were taken in the AHEAD study, we cannot determine to what extent our findings represent specific clinical diagnoses such as dementia or major depressive disorder. However, since only a small number of people would have met clinical criteria for dementia or mild cognitive impairment, the scales used allow us to consider the full range of function, in line with the goals of this study. Third, another potential limitation is that participants who were cognitively impaired were more likely to have proxy interviews and therefore to be excluded from this study. This could potentially underestimate the true effect of poor cognitive function on mortality risk. Fourth, we did not have standardized cause of death information. Fifth, although the AHEAD study assessed a wide range of potential predictors of mortality, we were not able to control for all potential confounders of mortality. For example, future studies of this topic may wish to include psychological covariates such as social support as they may play an important role in the development of depressive symptoms and are also correlated with both poor cognitive function and mortality (40).

In summary, poor cognitive function and depressive symptoms are independent predictors of mortality over 2 years, and these associations remain after adjustment for multiple potential confounders. In participants with the best, middle, and worst cognitive function, more depressive symptoms are associated with increased mortality, and in participants with low, middle, and high depressive symptom counts, poor cognitive function is associated with higher levels of mortality. Thus the combination of depressive symptoms and worse cognitive function increases mortality risk in a progressive, additive manner. Cognitive function and depressive symptoms together stratify elderly adults into groups with widely differing 2-year rates of mortality, ranging from 3% to 16%. Our results highlight the need to consider both of these measures of mental well-being as important indicators of vulnerability in community-living elderly adults.

Acknowledgments

This work was supported by a grant from the National Institute on Aging (R01AG19827). Dr. Mehta was supported by a training grant from the National Institute on Aging (T32-AG00212-08) and by a pilot grant from the Center for Aging in Diverse Communities at UCSF. Dr. Yaffe was supported by a Patient-Oriented Research Development award from the National Institute on Aging (K23-AG00888). Dr. Covinsky was supported by an independent scientist award from the Agency for Healthcare Research and Quality (K02 HS00006-01). Drs. Covinsky and Yaffe are Paul Beeson Faculty Scholars in Aging Research. Dr. Langa was supported by a Career Development Award from the National Institute on Aging (K08 AG19180) and a New Investigator Grant from the Alzheimer’s Association. Dr. Whooley is supported by Research Career Development Awards from the Department of Veterans Affairs Health Services Research and Development Service, the Paul Beeson Physician Faculty Scholars in Aging Research Program, and the Robert Wood Johnson Foundation Generalist Physician Faculty Scholars Program.

References

1. Williams ME, Hadler NM. Sounding board: the illness as the focus of geriatric medicine. N Engl J Med. 1983;308:1357–1360. [PubMed]
2. Tinetti ME, Inouye SK, Gill TM, Doucette JT. Shared risk factors for falls, incontinence, and functional dependence: unifying the approach to geriatric syndromes. JAMA. 1995;273:1348–1353. [PubMed]
3. Schulz R, Beach SR, Ives DG, Martire LM, Ariyo AA, Kop WJ. Association between depression and mortality in older adults: the Cardiovascular Health Study. Arch Intern Med. 2000;160:1761–1768. [PubMed]
4. Frisoni GB, Fratiglioni L, Fastbom J, Viitanen M, Winblad B. Mortality in nondemented subjects with cognitive impairment: the influence of health-related factors. Am J Epidemiol. 1999;150:1031–1044. [PubMed]
5. Stump TE, Callahan CM, Hendrie HC. Cognitive impairment and mortality in older primary care patients. J Am Geriatr Soc. 2001;49:934–940. [PubMed]
6. Neale R, Brayne C, Johnson AL. Cognition and survival: an exploration in a large multicentre study of the population aged 65 years and over. Int J Epidemiol. 2001;30:1383–1388. [PubMed]
7. Dewey ME, Saz P. Dementia, cognitive impairment and mortality in persons aged 65 and over living in the community: a systematic review of the literature. Int J Geriatr Psychiatry. 2001;16:751–761. [PubMed]
8. Bosworth HB, Schaie KW, Willis SL. Cognitive and sociodemographic risk factors for mortality in the Seattle Longitudinal Study. J Gerontol Soc Sci. 1999;54:S273–S282. [PubMed]
9. Gale CR, Martyn CN, Cooper C. Cognitive impairment and mortality in a cohort of elderly people. BMJ. 1996;312:608–611. [PMC free article] [PubMed]
10. Swan GE, Carmelli D, LaRue A. Performance on the digit symbol substitution test and 5-year mortality in the Western Collaborative Group Study. Am J Epidemiol. 1995;141:32–40. [PubMed]
11. Inouye SK, Peduzzi PN, Robison JT, Hughes JS, Horwitz RI, Concato J. Importance of functional measures in predicting mortality among older hospitalized patients. JAMA. 1998;279:1187–1193. [PubMed]
12. Arfken CL, Lichtenberg PA, Tancer ME. Cognitive impairment and depression predict mortality in medically ill older adults. J Gerontol Med Sci. 1999;54A:M152–M156. [PubMed]
13. Whooley MA, Browner WS. Association between depressive symptoms and mortality in older women. Study of Osteoporotic Fractures Research Group. Arch Intern Med. 1998;158:2129–2135. [PubMed]
14. Covinsky KE, Kahana E, Chin MH, Palmer RM, Fortinsky RH, Landefeld CS. Depressive symptoms and 3-year mortality in older hospitalized medical patients. Ann Intern Med. 1999;130:563–569. [PubMed]
15. Rozzini R, Sabatini T, Frisoni GB, Trabucchi M. Association between depressive symptoms and mortality in elderly people. Arch Intern Med. 2001;161:299–300. [PubMed]
16. Fried LP, Kronmal RA, Newman AB, et al. Risk factors for 5-year mortality in older adults: the Cardiovascular Health Study. JAMA. 1998;279:585–592. [PubMed]
17. Ramos LR, Simoes EJ, Albert MS. Dependence in activities of daily living and cognitive impairment strongly predicted mortality in older urban residents in Brazil: a 2-year follow-up. J Am Geriatr Soc. 2001;49:1168–1175. [PubMed]
18. Kelman HR, Thomas C, Kennedy GJ, Cheng J. Cognitive impairment and mortality in older community residents. Am J Public Health. 1994;84:1255–1260. [PubMed]
19. Smits CH, Deeg DJ, Kriegsman DM, Schmand B. Cognitive functioning and health as determinants of mortality in an older population. Am J Epidemiol. 1999;150:978–986. [PubMed]
20. Yaffe K, Blackwell T, Gore R, Sands L, Reus V, Browner WS. Depressive symptoms and cognitive decline in nondemented elderly women: a prospective study. Arch Gen Psychiatry. 1999;56:425–430. [PubMed]
21. Blazer D, Burchett B, Service C, George LK. The association of age and depression among the elderly: an epidemiologic exploration. J Gerontol Med Sci. 1991;46:M210–M215. [PubMed]
22. Schulz R, Drayer RA, Rollman BL. Depression as a risk factor for non-suicide mortality in the elderly. Biol Psychiatry. 2002;52:205–225. [PubMed]
23. Zheng D, Macera CA, Croft JB, Giles WH, Davis D, Scott WK. Major depression and all-cause mortality among white adults in the United States. Ann Epidemiol. 1997;7:213–218. [PubMed]
24. Myers GC, Juster FT, Suzman RM. Asset and Health Dynamics Among the Oldest Old (AHEAD): initial results from the longitudinal study. Introduction. J Gerontol B Psychol Sci Soc Sci. 1997;52:v–viii. (spec No) [PubMed]
25. Heeringa S. Technical Description of the Asset and Health Dynamics Survey Sample Design. Ann Arbor: Institute for Social Research, University of Michigan; 1995.
26. Soldo BJ, Hurd MD, Rodgers WL, Wallace RB. Asset and Health Dynamics Among the Oldest Old: an overview of the AHEAD Study. J Gerontol B Psychol Sci Soc Sci. 1997;52(spec):1–20. [PubMed]
27. Ofstedal MB, McAuley GF, Herzog AR. Documentation of Cognitive Functioning Measures in the Health and Retirement Study. Ann Arbor: Survey Research Center, University of Michigan; 2001.
28. Herzog AR, Rodgers WL. Cognitive performance measures in survey research on older adults. In: Schwarz N, Park DC, Knauper B, Sudman S, editors. Cognition, Aging and Self-Reports. Philadelphia, PA: Psychology Press; 1999.
29. Folstein MF, Folstein SE, McHugh PR. “Mini-Mental State”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189–198. [PubMed]
30. Brandt J, Spencer M, Folstein M. The Telephone Interview for Cognitive Status. Neuropsychiatry Neuropsychol Behav Neurol. 1988;1:111–117.
31. Herzog AR, Wallace RB. Measures of cognitive functioning in the AHEAD Study. J Gerontol B Psychol Sci Soc Sci. 1997;52(spec):37–48. [PubMed]
32. Radloff L. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1:385–401.
33. Andresen EM, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale) Am J Prev Med. 1994;10:77–84. [PubMed]
34. Steffick DEWR, Herzog AR, Ofstedal MB, Fonda S, Langa K. Documentation of Affective Functioning Measures in the Health and Retirement Survey. Ann Arbor: University of Michigan; 2000.
35. Health and Retirement Study, the University of Michigan. 2002. [Accessed November 26, 2002]. Available at: http://www.umich.edu/~hrswww/docs/impute.html.
36. Rodgers W, Miller B. A comparative analysis of ADL questions in surveys of older people. J Gerontol B Psychol Sci Soc Sci. 1997;52 (spec):21–36. [PubMed]
37. Gill TM, Williams CS, Mendes de Leon CF, Tinetti ME. The role of change in physical performance in determining risk for dependence in activities of daily living among nondisabled community-living elderly persons. J Clin Epidemiol. 1997;50:765–772. [PubMed]
38. Wells CE. Pseudodementia. Am J Psychiatry. 1979;136:895–900. [PubMed]
39. Reifler B. A case of mistaken identity: pseudodementia is really pre-dementia. JAGS. 2000;48:593–594. [PubMed]
40. Harris T. Recent developments in understanding the psychosocial aspects of depression. Br Med Bull. 2001;57:17–32. [PubMed]