Heart Failure and Risk of Dementia and Alzheimer Disease: A Population-Based Cohort Study

Background  Heart failure has been linked to cognitive impairment in several previous studies, but to our knowledge, no investigations have explored the relationship between heart failure and the risk of dementia. We sought to examine the hypothesis that heart failure is a risk factor for dementia and Alzheimer disease.

Methods  A community-based cohort of 1301 individuals 75 years or older and without dementia in Stockholm, Sweden, was examined 3 times over a 9-year period to detect patients with dementia and Alzheimer disease using the diagnostic criteria of the Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition. Heart failure was defined according to the guidelines of the Task Force on Heart Failure of the European Society of Cardiology by integrating clinical symptoms and signs with inpatient register entries and use of cardiac medications. Data were analyzed using Cox proportional hazards models with adjustment for major potential confounders.

Results  During the 6534 person-years of follow-up (mean, 5.02 years per person), 440 subjects were diagnosed as having dementia, including 333 with Alzheimer disease. At baseline, heart failure was identified in 205 subjects. Heart failure was associated with a multi-adjusted hazard ratio (HR) of 1.84 (95% confidence interval [CI], 1.35-2.51) for dementia and 1.80 (95% CI, 1.25-2.61) for Alzheimer disease. Use of antihypertensive drugs (83% of which are diuretics) seemed to reduce dementia risk due to heart failure (HR, 1.38; 95% CI, 0.99-1.94). Heart failure and low diastolic pressure (<70 mm Hg) had an additive effect on the risk for dementia (HR, 3.07; 95% CI, 1.67-5.61).

Conclusions  Heart failure is associated with an increased risk of dementia and Alzheimer disease in older adults. Antihypertensive drug therapy may partially counteract the risk effect of heart failure on dementia disorders.

Heart failure and dementia are common medical conditions in elderly persons.^1,2 Clinical observations³ report that heart failure and lower left ventricular fraction are related to poorer cognitive function. Population-based studies^4-6 show that chronic heart failure is associated with cognitive impairment and cognitive decline independent of vascular disorders and other potential confounders. Furthermore, we have previously reported^7,8 that low diastolic pressure (ie, <70 mm Hg) and a clinically significant decline in systolic pressure are associated with subsequent development of dementia and Alzheimer disease. Based on these findings, we hypothesize that (1) heart failure might be linked to dementia (because cognitive impairment could reflect a prodromal phase of dementia), and (2) heart failure might influence the relationship between low diastolic pressure and dementia given the association between low blood pressure and both heart failure and cognitive impairment in elderly persons.⁹ In the present study, we sought to examine these 2 hypotheses using 9-year follow-up data on a community cohort of older adults. To our knowledge, no prospective cohort studies have examined the relationship between heart failure and dementia risk.

The study population was derived from the Kungsholmen Project, a community-based cohort study on aging and dementia that has been fully described elsewhere.^10,11 The Figure shows a flowchart of the study population. Briefly, all registered inhabitants who were living in the Kungsholmen district of Stockholm, Sweden, and were 75 years or older in October 1987 were initially invited to participate in the project. At baseline (1987-1989), 1473 of the 1810 participants were identified as being free of dementia using a 2-phase survey. Of these 1473 participants, 172 refused to participate in the first follow-up (1991-1993) or had moved, leaving 1301 subjects for the current analysis. Of the 987 subjects who underwent the first follow-up examination, 788 remained free of dementia with 44 of these individuals refusing subsequent participation at the second follow-up evaluation (1994-1996). At the third follow-up (1997-1998), 321 of the 459 survivors without dementia from the second follow-up examination received a dementia workup with 32 individuals refusing further participation. Medical records and death certificates were available for all deceased subjects during the first (n = 314 deceased subjects), the second (n = 176), and the third (n = 106) follow-up periods.

Informed consent was obtained for all participants, with informants providing consent for cognitively impaired persons. The ethics committee at the Karolinska Institutet, Stockholm, approved all phases of the Kungsholmen Project.

At baseline, data on demographic features (ie, age, sex, and education) and typical symptoms and signs potentially associated with heart failure (ie, breathlessness, fatigue, and ankle swelling) were collected from the subjects by trained nurses following standardized protocols.^10,11 Global cognitive functioning was assessed with the Mini-Mental State Examination (MMSE). Weight and height were measured using a standard scale with participants in light clothing and no shoes. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. Arterial blood pressure (ie, systolic Korotkoff phase I and diastolic phase V) was measured by nurses after the subjects rested in a seated position for at least 5 minutes. Pulse rate (beats per minute) was obtained for a 1-minute interval. Information on medical history prior to baseline assessment was obtained from the computerized inpatient register system encompassing all hospitals in the Stockholm area since 1969. The system recorded up to 6 different disorders that had been diagnosed during each hospitalization. The International Classification of Diseases, Eighth Revision (ICD-8),¹² was used by the register system until 1986. Since 1987, the ICD, Ninth Revision (ICD-9),¹³ has been adopted. Data on medical drug use for the 2 weeks prior to the baseline interview were collected from the subjects and verified by inspecting drug prescriptions and containers. Medications were coded according to the Anatomical Therapeutic Chemical (ATC) classification system.⁷ Genomic DNA was prepared from peripheral blood samples that were taken at baseline, and apolipoprotein E allelic status was determined following standard procedure.

Information on history of coronary heart disease (ICD-8 and ICD-9 codes 410-414), atrial fibrillation (ICD-8 code 427.9 and ICD-9 code 427), and stroke (ICD-8 and ICD-9 codes 430-438) was derived from the inpatient register database. Diabetes mellitus was defined according to our previous study.¹⁴ Antihypertensive drugs were defined as all medications potentially used for lowering blood pressure (ATC codes C02, C03, and C07).

Heart failure was defined by integrating clinical information with inpatient register records and use of medications following the guidelines of the Task Force on Heart Failure of the European Society of Cardiology.¹⁵ Heart failure was considered to be present if (1) a subject had at least 2 of the 3 defined clinical symptoms or signs (breathlessness without using anti-asthmatic drugs, fatigue, and ankle swelling) and had inpatient register entries indicating either clinically relevant heart disease (ie, coronary heart disease and atrial fibrillation) or treatment with cardiac glycosides (ie, ATC code C01A), or (2) a subject had inpatient register entries for disorders indicating the presence of heart failure,^16,17 including heart failure or left ventricular failure or other myocardial insufficiency (ICD-8 codes 427 and 428; ICD-9 code 428.x), hypertensive heart disease (ICD-8 and ICD-9 codes 402), cardiomyopathy (ICD-8 and ICD-9 codes 425), and ill-defined heart disease (ICD-8 and ICD-9 codes 429).

At each follow-up, all participants received a comprehensive clinical examination and completed a cognitive test battery.^10,11 We used criteria from the Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition,¹⁸ to define dementia cases according to a previously validated 3-step diagnostic procedure.¹⁹ In brief, 2 examining physicians made independent preliminary diagnoses, with a third physician consulted in the case of disagreement to reach a concordant diagnosis. The diagnosis of Alzheimer disease required gradual onset, progressive deterioration, and lack of any other specific causes of dementia. Our diagnostic criteria for Alzheimer disease were similar to the international standardized criteria²⁰ for probable Alzheimer disease. For subjects who died during each follow-up period, 2 physicians made a diagnosis of dementia or Alzheimer disease by thoroughly reviewing medical records and death certificates.

The incidence rate was calculated as the number of events occurring during the entire follow-up period divided by person-years of follow-up. Cox proportional hazards models were used to estimate the hazard ratio (HR) and 95% confidence interval (CI) of developing dementia and Alzheimer disease, with follow-up time as the time scale. For subjects who did not develop dementia, the follow-up time was calculated from the date of baseline interview to the date of the last follow-up contact or death. For subjects who developed dementia, the follow-up time was estimated as the full time during which the subjects were free of dementia plus half of the follow-up time during which dementia developed. The assumption of proportional hazards was confirmed by plotting the Kaplan-Meier survival curves, revealing approximately parallel survival functions between groups with and without heart failure. We examined the combined additive effect of heart failure with use of antihypertensive drugs, high systolic pressure (ie, ≥140 mm Hg), and low diastolic pressure (ie, <70 mm Hg) on the risk of dementias. The combined effect was assessed by creating dummy variables based on the joint exposure to both heart failure and each select factor. The multiplicative interaction was also evaluated by incorporating the independent variables and their cross-product term in the same model. Age, sex, education, baseline MMSE score, stroke, diabetes mellitus, antihypertensive drug use, systolic pressure, diastolic pressure, pulse rate, and BMI were considered as covariates. Because heart failure caused substantial mortality and dementia was underdiagnosed among deceased persons, the subjects' survival status for each follow-up occasion and the interaction term of heart failure with survival status were also included in the models. Dementia in general and Alzheimer disease in particular were used as separate outcomes in all Cox regression analyses.

At baseline, 205 (15.8%) of the 1301 subjects without dementia were identified as having heart failure. Subjects with heart failure were older, less educated, and more likely to have a lower MMSE score, diabetes mellitus, low diastolic pressure, and antihypertensive therapy than those without the disease, but the 2 groups had no significant differences in sex, stroke, BMI, and pulse rate (Table 1).

During the 6534 person-years of follow-up (mean, 5.02 years per person; maximum, 10.48 years), 440 subjects were diagnosed with dementia, including 333 with Alzheimer disease. As shown in Table 2, subjects with heart failure had a higher incidence rate of dementia and Alzheimer disease than those without the disease. Multi-adjusted Cox regression analysis suggested that heart failure was significantly associated with an elevated risk of dementia and Alzheimer disease. Low diastolic blood pressure (<70 vs ≥70 mm Hg) was significantly associated with increased HRs of dementia and Alzheimer disease independent of heart failure, follow-up survival status, and all other covariates. High systolic blood pressure (≥140 vs <140 mm Hg) and use of antihypertensive drugs were not significantly related to incident dementia and Alzheimer disease.

We repeated the analyses among subjects who survived to the time when dementia status was determined (n = 705; 383 dementia cases and 302 Alzheimer disease cases) yielding similar patterns of results. The multi-adjusted HRs related to heart failure were 1.70 (95% CI, 1.24-2.34) for dementia and 1.61 (95% CI, 1.11-2.34) for Alzheimer disease. Low diastolic blood pressure was associated with multi-adjusted HRs of 1.58 (95% CI, 1.13-2.21) for dementia and 1.84 (95% CI, 1.28-2.67) for Alzheimer disease.

We also conducted stratified analysis by heart failure and antihypertensive drug use (Table 3). Compared with neither heart failure nor use of antihypertensive drugs, heart failure was significantly associated with an increased risk of dementia and Alzheimer disease. Patients with heart failure who used antihypertensive drugs had a relatively lower risk of developing dementias. We also examined the combined additive effect of heart failure and blood pressure. Increased risk of dementia and Alzheimer disease was significantly associated with heart failure and marginally associated with high systolic blood pressure. Subjects with both heart failure and high systolic blood pressure showed a tendency toward decreased risk of dementias relative to individuals with heart failure but with systolic blood pressure lower than 140 mm Hg. By contrast, heart failure and low diastolic blood pressure had an additive effect on the risk of dementia and Alzheimer disease in addition to the significant independent effects for each factor. In the multiplicative models, heart failure showed no significant multiplicative interaction with antihypertensive drug use, systolic blood pressure, or diastolic blood pressure on the risk of dementia and Alzheimer disease.

We further examined the independent effect of heart failure on the risk of dementia and Alzheimer disease by adjusting for pulse pressure and apolipoprotein E genotypes as well as for coronary heart disease and atrial fibrillation, yielding results very similar to those reported in Table 2 (data not shown).

In this follow-up study of a large-scale community cohort of older adults, in which women accounted for 75% of the cohort, we found (1) that heart failure is associated with an increased risk of dementia and Alzheimer disease, (2) that low diastolic pressure is related to subsequent development of dementias independent of heart failure, (3) that use of antihypertensive drugs might partially counteract the risk of dementias due to heart failure, and (4) that heart failure and low diastolic blood pressure additively increase the risk of dementia and Alzheimer disease.

The diagnosis of heart failure, even in a clinical setting, is fairly difficult. Most population-based epidemiological studies have used a strategy of combining information sources such as medical records, clinical symptoms and signs, and drug prescriptions.^21,22 In accordance with the guidelines of the Task Force on Heart Failure of the European Society of Cardiology,¹⁵ we defined heart failure by integrating the clinical symptoms and signs (eg, breathlessness, fatigue, and ankle swelling) with objective evidence of major cardiac dysfunction from inpatient medical records and use of cardiac medications. The operational approach used in defining heart failure in our study was similar to the one that had been previously validated in the Rotterdam study.²¹ Using this approach, we found that the overall prevalence rate of heart failure was 15.8% in our population. The figure is comparable with the report²¹ on the Rotterdam population, in which heart failure was detected in approximately 13% of people aged 75 to 94 years.

There is evidence suggesting an association between heart failure and cognitive impairment in elderly persons. Clinical studies have shown that older age and a lower left ventricular fraction are related to poorer cognitive performance.³ Patients with end-stage heart failure awaiting heart transplantation are more likely to have cognitive impairment, which could improve after cardiac surgery.²³ A cross-sectional survey on an Italian community sample found that chronic heart failure is associated with cognitive impairment.⁴ Furthermore, a systematic review of 5 case-control studies concluded that congestive heart failure is associated with a specific pattern of cognitive impairment that includes attention and memory deficits.⁵ In addition, some longitudinal findings have also suggested that heart failure is predictive of cognitive decline.⁶

To date, there are no community-based prospective studies that specifically address the relationship between heart failure and risk of dementia. We found that heart failure was associated with a more than 80% increased risk of developing dementia and Alzheimer disease after controlling for major potential confounders such as vascular comorbidities and follow-up survival status. Low diastolic blood pressure was previously found to be associated with an increased risk of subsequent dementia and Alzheimer disease in this population, independent of major potential confounding factors, including pulse pressure.⁷ The present findings confirm that the risk effect of low diastolic blood pressure on the development of dementia is independent of heart failure. Furthermore, heart failure had an additive effect with low diastolic blood pressure on the risk of dementia and Alzheimer disease. More important, our data demonstrate that the risk effect of heart failure on dementia seems to be partially reduced by the use of antihypertensive drugs (83% of which are diuretics). Because only baseline drug use was taken into account in our analysis, the effect of antihypertensive drug use might have been underestimated. Owing to its relevant implications for clinical practice and dementia prevention, the effect modification of antihypertensive drug therapy warrants further clarification. Finally, the tendency for heart failure to lead to a relatively decreased dementia risk for individuals with a higher systolic blood pressure is consistent with a recent report²⁴ showing that high systolic blood pressure (≥130 mm Hg) is associated with a lower risk of cognitive impairment among patients with heart failure.

The major strengths of our study include the relatively long-term follow-up of a large-scale community cohort and the comprehensive diagnostic assessments of dementia and Alzheimer disease. However, a few limitations deserve mention. First, data available in our project allowed us to define heart failure following a well-recognized definition, but the accuracy of heart failure diagnosis remains a concern owing to the lack of objective assessments on cardiac function by means such as radiography of the chest and echocardiography.¹⁵ However, potential misclassifications are unlikely to differ between those who did or did not develop dementia. Second, we analyzed only baseline heart failure in relation to dementia. Some subjects might have developed heart failure during the follow-up period, resulting in an underestimation of the association between heart failure and dementia. Third, we were not able to distinguish between systolic and diastolic heart failure, although different types of heart failure may be linked to dementia and Alzheimer disease through the similar mechanism of reduced cerebral perfusion. Finally, caution is needed when generalizing our findings to younger populations because our study cohort consisted of individuals with a minimum age of 75 years at entry.

In light of the main findings, it is worth considering the potential biological pathways linking heart failure to dementia, and to Alzheimer disease in particular. First, the substantially impaired cerebral circulation resulting from chronic heart failure may play a part in the neurodegenerative process, especially in very old people.^25,26 Second, heart failure is a risk factor for clinical and silent brain infarcts, leukoaraiosis, and cerebral thrombosis, and vascular comorbidities such as atrial fibrillation and diabetes mellitus are common in patients with heart failure. These vascular conditions have been linked to dementia and, to some extent, to Alzheimer disease.^14,27-29 Third, the impaired cerebrovascular reactivity due to chronic heart failure could further aggravate hypoperfusion-related cerebral ischemia and neurodegeneration.²⁶ Proper use of antihypertensive drugs may partially counteract the effect of heart failure by improving cerebral perfusion and by preventing vascular lesions in the brain.⁷

In summary, the major findings that both heart failure and low diastolic pressure increase the risk of dementia and Alzheimer disease in a very old population support the view that cerebral hypoperfusion is a condition that could lead to dementia, including Alzheimer disease. In addition, antihypertensive therapy mainly with diuretics might reduce the risk effect of heart failure on dementia. These findings have relevant implications for clinical practice as a previous study has shown that nearly half of heart failure patients from general primary care practices do not receive diuretics.³⁰ Our findings also highlight an urgent need for cognitive function and dementia to be added to the endpoints of future observational studies and clinical trials involving heart failure and antihypertensive therapy.

Correspondence: Chengxuan Qiu, MD, PhD, Stockholm Gerontology Research Center, Olivecronas väg 4 (Box 6401), S-113 82 Stockholm, Sweden (chengxuan.qiu@ki.se).

Accepted for Publication: December 28, 2005.

Author Contributions: Drs Fratiglioni and Qiu had full access to all of the data in this study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Financial Disclosure: None.

Funding/Support: This study was funded by research grants from the American Alzheimer's Association, the Loo and Hans Ostermans Foundation (Stockholm, Sweden), and the Swedish Research Council.

Acknowledgment: We thank all members of the Kungsholmen Project Study Group for their collaboration in data collection and management, Miia Kivipelto, MD, PhD, for valuable comments on the manuscript, and Stuart MacDonald, PhD, for helpful editorial comments.