Andersen, Kjeld*; Lolk, Annette*; Kragh-Sørensen, Per*; Petersen, Niels Erik†; Green, Anders‡
Epidemiologic studies of a possible relation between depression and Alzheimer disease have produced mixed results. Case-control studies have in general reported that depression is associated with Alzheimer disease.1–7 Some studies suggested that depression was a prodromal sign,8,9 whereas others found that depression might be both a prodromal symptom of Alzheimer disease and a risk factor.5,7 One study reported no association between depression and Alzheimer disease.10 Most studies relied on self-reported depression. In one study, a geriatric psychiatrist and a neuropsychologist validated the self-report,5 in another the participants were diagnosed by 2 geriatric psychiatrists,9 and in one third the diagnosis was based on register information.6 All studies used the criteria for dementia stated in the Diagnostic and Statistical Manual of Mental Disorders, third edition (DSM-III) or the criteria for Alzheimer disease stated in a report from the National Institute of Neurologic and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association in 1984 (NINCDS-ADRDA).11
Similarly, some longitudinal studies have reported increased risk for Alzheimer disease among persons with a history of depression,12–15 others reporting depression as a prodromal symptom,12,13,16–18 and 1 study reported no association.19 In most studies, the diagnosis of depression was made according to DSM-III-R20 or the International Classification of Diseases, 10th Revision (ICD-10).21 In 2 studies, the final diagnosis of depression was made according to a “DSM-III-R-guided” approach based on information collected at baseline.12,16 One study relied on general practitioners’ diagnosis of depression13 and 1 on hospital register information.15 Like in the case-control studies, the diagnosis of dementia was made according to accepted criteria at the time of the study, (eg, DSM-III-R and NINCDS-ADRDA).
It has been reported that agreement between self-reported depression and diagnosis made by psychiatrists is limited (κ = 0.20). Thus, studies based on self-report may miss many cases, thereby diluting possible associations between depression and Alzheimer disease.22
The purpose of the present study was to examine the risk of Alzheimer disease among persons who had a history of depression compared with persons who had no history of depression, hypothesizing that depression would increase the subsequent risk of Alzheimer disease. We had access only to self-reported depression, thus risking a type 2 error, wrongly accepting a false null hypothesis of no association between depression and Alzheimer disease because of the dilution mentioned here.
The data originate from a population-based cohort study on dementia conducted from 1992 to 1999 on prevalence, incidence, and risk factors for dementia.23,24 The study was approved by the Biomedical Research Ethics Committee for the Counties of Vejle and Funen, and all participants signed an informed consent on participation. Potential study participants were randomly selected from the Central Office of Civil Registration, where all persons in Denmark are listed according to a 10-digit registration number. We selected 5237 persons who fulfilled 2 criteria: age between 65 and 84 years and living in the municipality of Odense, Denmark. At baseline (1992–1994), 156 were diagnosed as having Alzheimer disease and 3086 did not have dementia (including 21 who were subsequently rediagnosed as not demented) (Fig. 1).
At the 2-year follow up (1994–1996), 102 were diagnosed as incident cases of Alzheimer disease and 2313 did not have dementia (including 6 who were subsequently rediagnosed as not demented). Finally, at the 5-year follow up (1997–1999), there were 82 new cases of Alzheimer disease, whereas 1688 did not have dementia. At this final examination, 30 of the 102 incident Alzheimer disease cases at the 2-year follow up had died, 14 declined to participate, and 1 had moved. Of the remaining 57 incident Alzheimer disease cases, 7 were diagnosed as mixed cases with combined Alzheimer disease and vascular dementia, leaving 50 cases of Alzheimer disease. Two cases with dementia of other types were diagnosed as also having Alzheimer disease, resulting in 52 cases, bringing the total number of eligible Alzheimer disease cases up to 134 (82 + 52 cases).
We applied a 2-phase procedure for the diagnosis of Alzheimer disease. The first phase was a screening, which aimed to identify persons with and without dementia. In the second phase, we determined whether persons who potentially had dementia actually were demented.
The screening procedure has been described in detail elsewhere.25 Briefly, it included administration of Cambridge Cognitive Examination (CAMCOG), a subsection of The Cambridge Mental Disorders of the Elderly Examination (CAMDEX).26 CAMCOG assesses a range of cognitive functions (memory, orientation, praxis, perception, abstract thinking, language, and calculation); the maximum score is 107. The scores on cognitive tests depend on age and education,27 suggesting that a cutoff should be graded according to the person's age and level of education. We consequently defined the cutoff as a difference between the actual CAMCOG score obtained at the screening and a predicted CAMCOG score of at least −4.41 points (equivalent to −1.25 standard deviation), which resulted in a sensitivity of 88% and a specificity of 89%.25 The predicted CAMCOG score was based on age, social class, and premorbid intellectual level and was calculated from a regression equation, which is described in detail elsewhere.25
Persons for whom the screening indicated possible dementia were further examined with the remaining part of CAMDEX, which includes 1) standardized psychiatric interview, incorporating questions regarding present state (both psychologically and physically), 2) personal and medical history, 3) family history, and 4) a physical and neurologic examination. Furthermore, 7 neuropsychologic tests were administered for a more elaborate examination of the cognitive function.28
Diagnostic Criteria for Alzheimer Disease
Alzheimer disease was diagnosed according to the criteria for probable Alzheimer disease established by NINCDS-ADRDA.11 The diagnosis was made by consensus between the participating doctors and the neuropsychologist. Clinically diagnosed cases had a blood test and computed tomography scan. The severity of dementia was assessed according to the Clinical Dementia Rating.29 Persons fulfilling the NINCDS-ADRDA criteria, and subsequently assessed as questionably demented according to the Clinical Dementia Rating (a score of 0.5), were diagnosed as very mildly demented.
History of Depression
At baseline, participants were interviewed about previous depressive episodes. An introductory question was “Has a medical doctor ever told you that you suffered from a depression?” If the answer was “yes,” further questions about number of episodes, age at each episode, treatment, and hospitalization were asked. When the participants reported that they had been hospitalized as a result of depression, we traced and reviewed records from the Department of Psychiatry at Odense University Hospital to ascertain the diagnosis. For persons who reported depressive episodes without hospitalization, the diagnosis was based entirely on the self-reported information obtained at the interview.
Differences between participants and nonparticipants were analyzed with chi-squared test, chi-squared test for trend, and t test. Logistic regression analysis was used to calculate the odds ratio (OR) and 95% confidence interval (CI) for persons with Alzheimer disease compared with persons who did not have dementia; cases with vascular dementia or dementia of other types were excluded from the analyses. History of depression was the independent variable of interest. The Odense Study is part of the EURODEM analyses of risk factors for dementia. In the EURODEM studies, a logistic regression model with age, age squared, sex, and intellectual level has been used.30 We used the same basic model and subsequently included other possible confounders (myocardial infarction and diabetes mellitus). Both have been reported as a risk factor for depression31,32 and may be associated with poor cognitive performance.33,34
The participation rate at baseline was higher for men (69%) than for women (60%). There was no difference at the 2-year follow up (participation rate for eligible persons was 81% for both sexes), whereas more eligible women (79%) than eligible men (75%) participated at the 5-year follow up (Table 1). At baseline, more nonparticipants had a history of depression than participants. At the 2- and 5-year follow up, more nonparticipants than participants had a history of myocardial infarction and of diabetes mellitus. There were no notable differences between participants and nonparticipants at any of the examinations regarding hypertension, thyroid diseases, Parkinson's disease, head trauma, or family history of dementia (data not shown).
The mean age of nonparticipants at baseline was 74.0 years. For persons who did not participate in the 2-year and 5-year follow up, the mean baseline age was 74.3 years. Among nonparticipants at the 2-year follow up, 12% had reported a history of depression at baseline; at the 5-year follow up, this figure was 13%.
Finally, we compared the 52 incident Alzheimer disease cases who were reexamined at the 5-year follow up with the 50 incident Alzheimer disease cases who were not reexamined. There were only minimal differences in age, sex, or history of depression (data not shown).
A History of Depression and the Risk of Developing Alzheimer Disease
At baseline, a total of 362 persons reported a history of depression, and 21 of these were diagnosed as having Alzheimer disease. Persons with a history of depression had an increased risk of being diagnosed with Alzheimer disease at baseline compared with persons who had no history of depression (OR = 1.7; CI = 1.0–2.7) (Table 2). For persons with at least 2 episodes of depression, the risk was even higher (2.7; 1.1–6.6). Persons who had been diagnosed with depression less than 10 years before study initiation had a higher risk (2.4; 0.9–6.5) than persons diagnosed 10 years or more previously (1.6; 0.8–2.9). The same pattern was found at the 2-year follow up and at the 5-year follow up in all analyses (data not shown).
In a subanalysis, we excluded persons who had reported a history of depression but denied having received antidepressant medication, and we restricted the analysis to persons who reported that they had been hospitalized as a result of depression. These restrictions did not change any of the results markedly (data not shown). Finally, adjustment for myocardial infarction and for diabetes mellitus did not change any of the results notably (data not shown).
Validation of the History of Depression
We traced all 48 persons who had been admitted to the Department of Psychiatry, Odense University Hospital. For 47, the diagnosis of depression was confirmed; in 1 case, the person had been admitted as a result of possible depression but was discharged with a diagnosis of hysterical neurosis (ICD-8 criteria). Exclusion of this person from the analyses did not change the results (data not shown).
We found an increased risk of Alzheimer disease for persons with a history of depression compared with persons without a history of depression. The odds ratios ranged from 1.6 to 1.9 (Table 2).
Case-control studies have yielded odds ratios ranging from 1.1 to 5.0,1–10 and follow-up studies have yielded odds ratios ranging from 2.1 to 14.12–15 In general, the follow-up studies have reported higher estimates of the risk than case-control studies. Because most case-control studies have relied on self-report, a possible explanation for this pattern could be the dilution of the association between depression and Alzheimer disease because of errors in self-reported depression.22 The general comparison between persons with and without a history of depression provides results most similar to the case-control studies. Among persons reporting 2 or more depressive episodes, odds ratios were in the upper end of the range reported for case-control studies and in the lower end reported for follow-up studies. This is likely a reflection of our study design. Although we used a longitudinal design, following persons who did not have dementia at baseline, our ascertainment of depression is similar to that in case-control studies. Thus, self-report of history of depression is likely to yield conservative estimates irrespective of whether the design is cross-sectional or longitudinal. Although we were able to confirm the diagnosis in 47 of the 48 persons admitted to the local psychiatric department, lending support to the self-report made by the participants, we had no mechanism for confirming the self-reports of depression.
Persons with a history of depression, and especially 2 or more episodes, had a higher risk of Alzheimer disease. This pattern was less obvious at the 5-year follow up. The onset of Alzheimer disease may precede the time of diagnosis by several years, which suggests that incident cases after a relatively short follow-up period of 2 years might include preexisting (prevalent) cases. Because the information on risk factors was collected at baseline, the baseline and the 2-year follow-up analyses might be expected to yield similar results. At the 5-year follow up, the prevalent cases would be a smaller proportion of the apparent new cases. If so, depression would not truly be increasing the risk of Alzheimer disease. Alternatively, the weaker association at the 5-year follow up could be the result of the imprecise estimates.
We restricted the analyses to persons reporting having received antidepressant medication or being hospitalized because of depression, which did not change the results. However, this measure of exposure is also limited by possible recall bias, inconsistency in decisions made by various general practitioners over many years, changes over time in the propensity to be hospitalized, and changes over time in the indication for treatment with antidepressants. The group reporting having had depressive episodes is therefore likely to be a heterogeneous group. We did not have enough information to examine this lack of precision in exposure classification.
In summary, we found that history of depression was associated with an increased risk of developing Alzheimer disease. Although the present study was longitudinal in design, the data collection of the risk factor was most comparable to cross-sectional studies and the results were also most comparable to the cross-sectional studies.
1.French LR, Schuman LM, Mortimer JA, et al. A case-control study of dementia of the Alzheimer type. Am J Epidemiol. 1985;121:414–421.
2.Shalat SL, Seltzer B, Pidcock C, et al. Risk factors for Alzheimer's disease: a case-control study. Neurology. 1987;37:1630–1633.
3.Kokmen E, Beard CM, Chandra V, et al. Clinical risk factors for Alzheimer's disease: a population-based case-control study. Neurology. 1991;41:1393–1397.
4.Speck CE, Kukull WA, Brenner DE, et al. History of depression as a risk factor for Alzheimer's disease. Epidemiology. 1995;6:366–369.
5.Steffens DC, Plassman BL, Helms MJ, et al. A twin study of late-onset depression and apolipoprotein E epsilon 4 as risk factors for Alzheimer's disease. Biol Psychiatry. 1997;41:851–856.
6.Cooper B, Holmes C. Previous psychiatric history as a risk factor for late-life dementia: a population-based case-control study. Age Ageing. 1998;27:181–188.
7.Wetherell JL, Gatz M, Johansson B, et al. History of depression and other psychiatric illness as risk factors for Alzheimer disease in a twin sample. Alzheimer Dis Assoc Disord. 1999;13:47–52.
8.Broe GA, Henderson AS, Creasey H, et al. A case-control study of Alzheimer's disease in Australia. Neurology. 1990;40:1698–1707.
9.Alexopoulos GS, Young RC, Meyers BS. Geriatric depression: age of onset and dementia. Biol Psychiatry. 1993;34:141–145.
10.The Canadian Study of Health and Aging: risk factors for Alzheimer's disease in Canada. Neurology. 1994;44:2073–2080.
11.McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology. 1984;34:939–944.
12.Devanand DP, Sano M, Tang MX, et al. Depressed mood and the incidence of Alzheimer's disease in the elderly living in the community. Arch Gen Psychiatry. 1996;53:175–182.
13.Buntinx F, Kester A, Bergers J, et al. Is depression in elderly people followed by dementia? A retrospective cohort study based in general practice. Age Ageing. 1996;25:231–233.
14.Palsson S, Aevarsson O, Skoog I. Depression, cerebral atrophy, cognitive performance and incidence of dementia. Population study of 85-year-olds. Br J Psychiatry. 1999;174:249–253.
15.Kessing LV, Olsen EW, Mortensen PB, et al. Dementia in affective disorder: a case-register study. Acta Psychiatr Scand. 1999;100:176–185.
16.Chen P, Ganguli M, Mulsant BH, et al. The temporal relationship between depressive symptoms and dementia: a community-based prospective study. Arch Gen Psychiatry. 1999;56:261–266.
17.Alexopoulos GS, Meyers BS, Young RC, et al. The course of geriatric depression with ’reversible dementia': a controlled study. Am J Psychiatry. 1993;150:1693–1699.
18.Reding M, Haycox J, Blass J. Depression in patients referred to a dementia clinic. A three-year prospective study. Arch Neurol. 1985;42:894–896.
19.Henderson AS, Korten AE, Jacomb PA, et al. The course of depression in the elderly: a longitudinal community-based study in Australia. Psychol Med. 1997;27:119–129.
20.Diagnostic and Statistical Manual of Mental Disorders, third edition revised. Washington, DC: American Psychiatric Association; 1987.
21.ICD-10 Classification of Mental and Behavioural Disorders. Geneva: World Health Organization; 1992.
22.Eaton WW, Neufeld K, Chen LS, et al. A comparison of self-report and clinical diagnostic interviews for depression: diagnostic interview schedule and schedules for clinical assessment in neuropsychiatry in the Baltimore epidemiologic catchment area follow-up. Arch Gen Psychiatry. 2000;57:217–222.
23.Andersen K, Lolk A, Nielsen H, et al. Prevalence of very mild to severe dementia in Denmark. Acta Neurol Scand. 1997;96:82–87.
24.Andersen K, Nielsen H, Lolk A, et al. Incidence of very mild to severe dementia and Alzheimer's disease in Denmark. The Odense Study. Neurology. 1999;52:85–90.
25.Lolk A, Nielsen H, Andersen K, et al. CAMCOG as a screening instrument for dementia:the Odense study.Cambridge Cognitive Examination. Acta Psychiatr Scand. 2000;102:331–335.
26.Roth M, Tym E, Mountjoy CQ, et al. CAMDEX. A standardised instrument for the diagnosis of mental disorder in the elderly with special reference to the early detection of dementia. Br J Psychiatry. 1986;149:698–709.
27.Huppert FA, Brayne C, Gill C, et al. CAMCOG—a concise neuropsychological test to assist dementia diagnosis: socio-demographic determinants in an elderly population sample. Br J Clin Psychol. 1995;34:529–541.
28.Nielsen H, Lolk A, Kragh-Sørensen P. Normative data for eight neuropsychological tests, gathered from a random sample of Danes aged 64 to 83 years. Nord Psyk. 1995;47:241–255.
29.Berg L. Clinical Dementia Rating (CDR). Psychopharmacol Bull. 1988;24:637–639.
30.Launer LJ, Andersen K, Dewey ME, et al. Rates and risk factors for dementia and Alzheimer's disease. Results from EURODEM pooled analyses. Neurology. 1999;52:78–84.
31.Fielding R. Depression and acute myocardial infarction: a review and reinterpretation. Soc Sci Med. 1991;32:1017–1028.
32.Carney C. Diabetes mellitus and major depressive disorder: an overview of prevalence, complications, and treatment. Depress Anxiety. 1998;7:149–157.
33.Breteler MM, van Swieten JC, Bots ML, et al. Cerebral white matter lesions, vascular risk factors, and cognitive function in a population-based study: the Rotterdam Study. Neurology. 1994;44:1246–1252.
34.Skoog I. Status of risk factors for vascular dementia. Neuroepidemiology. 1998;17:2–9.
© 2005 Lippincott Williams & Wilkins, Inc.