Congestive heart failure (CHF) is a common complication of most diseases of the heart. Its prevalence increases exponentially from age 60 years [1], such that CHF is now one of the leading causes of hospitalisation, morbidity and mortality in Western societies [2]. The findings of several surveys indicate that physical, social, work, and leisure activities are significantly impaired amongst CHF subjects [3]. Psychological distress is also frequent, although data on this important aspect of the quality of life of patients remain sparse and difficult to interpret. Depression rates, for example, seem to be high (up to 58%) both in inpatient and outpatient settings [4, 5]. Another important, but neglected, aspect of the quality of life of patients with CHF is cognitive functioning. Early reports indicated that up to 80% of patients with severe CHF display deficits in memory and other cognitive abilities [6]. The consequences of these deficits are not clear, but it is conceivable that patients with cognitive impairment have even higher morbidity and mortality rates. For example, Cline et al. [7] reported that 10 of their 22 patients were unable to name the medication they were receiving for the treatment of CHF, 11/22 could not state the doses, and 14/22 failed to remember when to take their tablets. All subjects were surveyed 30 days after receiving detailed verbal and written information about their treatment regimen. Failing to take the prescribed medication, in such cases, may increase the frequency of clinical complications associated with CHF.

A recent meta-analysis confirmed that cognitive impairment, particularly attentional deficits, are prominent amongst CHF patients [8]. In addition, it seems that people with severe forms of CHF (functional class II/III) have greater difficulty than mild cases (functional class I) to perform tasks such as digit span and trail making [9]. These results indicate that the attentional deficits associated with CHF may become more pronounced with increasing severity of illness, although it is unclear whether clinical treatment can reverse these deficits. We designed the present study to clarify this question.

The results of the present study confirm that clinically decompensated CHF is associated with attentional deficits and generalised cognitive impairment. The clinical significance of these findings is even more apparent when one considers that the control group selected for the study was 10 years older than the CHF group. In fact, approximately half of the patients had CAMCOG and MMSE scores below the cutoff points used to determine the presence of cognitive impairment. Patients with CHF had considerable difficulties performing even a relatively simple attentional task such as the Digit Span.

Attentional deficits are expected amongst patients with a severe and debilitating medical illness such as CHF. In fact, subjects' performance in other cognitive tasks is likely to be influenced by their presence, which may partly explain the finding of generalised cognitive impairment amongst patients. Gorkin et al. [9] showed that subjects with more severe forms of CHF (functional class II/III) have greater difficulty than controls (functional class I) on digit span and trail making A tests. They also found that performance on these tasks is associated with decreased functional capacity, as measured by the 6-minute walk test. Nonetheless, it is important to note that the cognitive performance of patients may be influenced by factors not directly related to the presence of CHF, such as depression, anxiety, alcohol abuse, concomitant medical problems, and use of certain medications.

Our results also showed that the 6-week mortality rate of patients with CHF recruited into the study was high (22%) and was the main source of dropout from the study. Interestingly, the most robust predictor of dropout was the baseline score on the Digit Symbol Test (the result on the Letter Cancellation Test was marginally non-significant). The clinical significance of this finding requires further prospective evaluation.

The findings of the present study indicate that the clinical treatment of CHF is associated with significant improvement in attentional scores. More importantly, attentional scores of patients after treatment of CHF were similar to those of elderly controls, which suggests that these deficits may be, at least partly, reversible. Other studies had already shown that cognitive performance improves with surgery [6], although, to our knowledge, this is the first report showing that attentional deficits can be reversed with clinical treatment. From a clinical point of view, this finding is comparable to that of patients with other metabolic encephalopathies such as lupus erythematosus [19]. It remains unclear at this point whether such deficits interfere with the subject's ability to provide informed consent – this issue requires further investigation as patients with advanced heart failure may be candidates for experimental therapies and high-risk interventions such as cardiac transplantation.

The mechanisms that contribute to the development of cognitive impairment among patients with CHF remain unclear. Zucallà et al. [20] observed a linear relationship between MMSE scores and left ventricular ejection fraction rates for values lower than 40%. Similarly, Putzke et al. [21] noted that Trail B, Digit Symbol Test and Stroop scores were all significantly associated with cardiac output. Cerebrovascular disease is another likely candidate, as many patients with CHF have widespread cardiovascular problems and are at increased risk for strokes. Data derived from the Rotterdam study [22], for example, indicate that white matter disease is associated with subjective memory impairment and lower scores on tests of cognitive function. Similarly, patients with strokes are more likely to develop cognitive deficits and dementia. Other cardiovascular problems such as low [23] and high blood pressure [24], both frequent amongst patients with CHF, are associated with cognitive impairment. Cognitive impairment in CHF may also be due to the abnormal hormonal response that characterises the disease, although no direct evidence is currently available to support this hypothesis. There is also the possibility that the cognitive deficits in CHF are secondary to deficits in blood supply to the brain. Our SPECT results indicated the presence of a trend in that direction, with 5/6 subjects showing a non-significant increase in blood flow to the brain after clinical treatment. Alcohol abuse is another factor that may contribute to the development cardiomyopathy (and consequently lead to CHF) and cognitive impairment. Hence, in clinical settings it is not uncommon for patients with a positive history of alcohol abuse or dependence to present with dementia and CHF at the same time. The results of the present study indicate that there was an excess of patients with CHF who made regular use of alcohol, although none of them showed signs suggestive of the diagnosis of alcohol abuse or dependence. Finally, the medications commonly used for the treatment of CHF have an as yet unknown effect on cognition and may, from a theoretical point of view, contribute to some of the deficits observed amongst patients.

Some limitations of the study should be considered before its results can be generalised. Only patients with CHF were retested and their superior performance may partly reflect learning rather than a direct effect of treatment. However, the effect of learning on attentional scores tends to be small [18] and is unlikely to have significantly influenced our results. In addition, we must acknowledge that the scores of subjects on tests such as the Digit Symbol and Letter Cancellation may not necessarily indicate the presence of attentional deficits, as performance on these tests may be unrelated influenced by psychomotor speed difficulties. There is also a possibility that the improved scores were a direct result of the medication used for the treatment of CHF, rather than change in the clinical status of patients. At present, this seems to be an unlikely explanation to our findings, although we are unable to dismiss this hypothesis due to lack of information on this subject. In addition, the assessor was not blind to the clinical status of subjects, which may have interfered with the scoring of tests. In practice, however, this is improbable as most tests were performed independently by study participants. Finally, patients with CHF and elderly controls were not adequately matched for age, although any performance bias due to age should favour CHF patients – our findings were in the opposite direction. However, this age difference may have influenced the comparisons between the scores of patients after treatment and controls. Unfortunately, we were unable to adjust the analysis for the effect of age, as patients and controls had significantly different variances for the tests under consideration.

The results of the present study indicate that patients with CHF functional class IV show a pattern of generalised cognitive impairment that includes attentional deficits. We also showed that cognitive performance in this population is correlated with left ventricular ejection fraction, and that scores on attention tasks improve with clinical treatment. Future studies should now attempt to clarify the mechanisms that underlie cognitive functioning in CHF, as well as its long-term effects not only on attention and psychomotor speed abilities but also in general measures of cognitive functioning.