The aim of this multicenter observational study conducted in France was to determine the prevalence of memory impairment in ambulatory patients aged ≥70 years with chronic heart failure (HF). Two hundred ninety-one cardiologists recruited 912 ambulatory patients with HF (mean age 79.2 ± 5.8 years) from January to November 2009. Memory was evaluated by the delayed-recall Memory Impairment Screen (MIS-D). Memory impairment was defined as MIS-D score ≤6 and severe memory impairment as MIS-D score ≤4. HF was diagnosed 4.4 ± 4.8 years earlier and mean left ventricular ejection fraction was 43.6 ± 12.0%. Memory impairment was found in 416 subjects (45.6%, 95% confidence interval 42.4 to 48.8) and severe memory impairment in 213 subjects (23.4%, 95% confidence interval 20.6 to 26.1), whereas cardiologists only suspected memory impairment in 109 patients (12%; before evaluation by MIS). Determinants of memory disorders included older age, lower education level, depression, history of stroke, renal failure, and less regular physical activity. The severity of memory impairment increased with increasing severity of HF (New York Heart Association classification; p <0.00001). In conclusion, memory impairment in older patients with HF is common. The use of a simple-to-use tool such as the MIS-D may identify patients at risk and enable implementation of management strategies to improve therapeutic compliance.
Both heart failure (HF) and memory disorders are conditions that primarily affect older subjects. However, the relation between HF and memory disorders remains controversial. The main objective of the Evaluation of risk Factors for cognitive Impairment in Chronic Ambulatory heart failuRe in the Elderly (EFICARE) study was to determine the usability of the Memory Impairment Screen (MIS) test by cardiologists to assess the prevalence and determinants of memory impairment in a large ambulatory population of subjects ≥70 years with HF in France. The second objective was to compare the cardiologists’ clinical global impression with the MIS for the screening of memory disorders.
Methods
EFICARE study is an observational study of ambulatory patients aged ≥70 years with symptomatic HF living in France. Private practice cardiologists were contacted by phone and invited to participate in the study. They were equally distributed by region. To ensure recruitment of a representative population from France as a whole, each cardiologist was asked to recruit the first 4 consecutive patients seen for a routine consultation who met the following criteria: men or women aged ≥70 years, HF diagnosed according to the European Society of Cardiology criteria, and hospitalized for HF in the last 12 months irrespective of left ventricular ejection fraction (LVEF) value. The study complied with the Declaration of Helsinki. The study was approved by local regulatory authorities and each participant signed an informed consent form.
Data were collected from January to November 2009. Demographic characteristics, information on concurrent co-morbidities, New York Heart Association (NYHA) classification, and prescribed medications were recorded by the cardiologists. Before the cognitive function evaluation, cardiologists were asked to rate their patients’ cognitive status subjectively as normal or impaired. Then they evaluated memory function using the French version of the delayed-recall Memory Impairment Screen (MIS-D). The MIS test, a brief, 4-word, immediate free- and cued-recall memory test, with higher scores indicating better performance, has been developed as a screening tool. The test has been modified by adding a 10-minute delayed recall to the immediate recall (MIS-D). For each word, the patient received a score of 2 for free recall and 1 for cued recall. The maximal score for MIS-D is 8. An MIS-D score ≤6 indicates memory impairment and an MIS-D recall score ≤4 suggests severe memory impairment. Before its administration to subjects, cardiologists received a training session on MIS test by phone or in person. Autonomy was assessed using the 4-item version of the instrumental activities of daily living (IADL) scale. Subjects were scored according to their highest level of functioning in each category, with a total score ranging from 0 (low function, dependent) to 4 (high function, independent).
Higher education level was defined as above high school achievement. Regular physical activity was defined as moderate daily physical activity. Lifestyle adjustments for HF were defined as low sodium diet (diet containing ≤6 g of salt per day), regular physical activity, and following a HF management program. Depression was defined as use of antidepressant medications or diagnosis of depression made by the subjects’ psychiatrist or general practitioner. Renal failure was defined as a creatinine clearance of <60 ml/min as calculated with Cockcroft formula. Hypertension was defined as a history of hypertension or blood pressure ≥140/90 mm Hg.
Results are presented as means and SD for quantitative variables and as the number of subjects and percentage for categorical variables. Demographic and clinical characteristics were compared among the 3 groups of subjects, that is, MIS-D score >6 (MIS-D [7-8] ), MIS-D ≤6 and >4 (MIS-D [5-6] ), and MIS-D ≤4 (MIS-D [0-4] ), using analysis of variance for quantitative data and chi-square test or Fisher’s exact test for categorical variables. Correlation of MIS-D and IADL was tested using Kendall τ statistic. Univariate analysis was performed for each variable using a logistic regression model adjusted for age and gender in which the reference group was composed of subjects with MIS-D [7-8] .
All variables associated with memory impairment (p <0.15) were entered into multivariate multinomial logistic regression models to identify those associated independently with memory impairment. A stepwise backward selection method of the clinical variables was used to obtain the final model. Finally, in subjects of MIS-D [0-6] group (i.e., cognitively impaired), we analyzed demographic and clinical characteristics according to the cardiologists’ subjective cognitive status assessment. In a sensitivity analysis, we performed a hierarchical model to take into account the correlation between patients seen by the same cardiologist, comparing MIS-D [5-6] group with MIS-D [7-8] group and MIS-D [0-4] group with MIS-D [7-8] group, including the same variables as the final multivariate model. Statistical analysis was performed with the R statistical software. In all analyses, the 2-sided α level of 0.05 was used for significance testing.
Two hundred eighty-nine cardiologists (259 men [90%], 30 women [10%], mean age 51.7 ± 7.5 years) agreed to participate and recruited 1,054 patients. Seventy-four subjects were excluded because of protocol deviations (date of inclusion before January 1, 2009 [n = 1], last hospitalization >1 year ago [n = 59], or no data provided [n = 14]) leaving 980 patients. Compared with excluded subjects, the 980 included subjects had higher systolic (136.8 ± 18.3 vs 130.6 ± 18.0 mm Hg, p = 0.0007) and diastolic blood pressures (79.1 ± 10.3 vs 76.1 ± 10.3 mm Hg, p = 0.004) and lower LVEF (43.6 ± 12.0% vs 45.6 ± 11.6%, p = 0.01). They also used beta blockade more frequently (80% vs 66%, p = 0.0005) and thiazide diuretics less often (6.8% vs 15.7%, p = 0.002). Cardiologists successfully evaluated with MIS-D 93% (n = 912) of the 980 included subjects. Errors with MIS-D were mainly in scoring the test: 2 points per free recall items and 1 point for cued recall item that had not been free recalled. There were no significant differences in general characteristics between the 912 subjects included in the analysis and those with missing MIS-D (results not shown).
Results
General characteristics of the population (n = 912) at baseline are listed in Table 1 . Mean age was 79.1 ± 5.8 years and 590 subjects (65%) were men. Compared with patients with MIS-D [7-8] , subjects with MIS-D [0-6] were older, more often women, and had a lower education level. They were also more likely to have renal failure, a history of stroke, depression, and at least 1 fall in the last 12 months and were less likely to be physically active. As expected, level of functional impairment (IADL score) was strongly associated with MIS-D, with the lowest score in subjects with MIS-D [0-4] (Kendall’s correlation coefficient τ = 0.38, p <0.0001).
| General Characteristic | MIS-D [7-8] (n = 496) | MIS-D [5-6] (n = 203) | MIS-D [0-4] (n = 213) | p Value ∗ |
|---|---|---|---|---|
| Mean age (yrs) | 78.1 ± 5.6 | 79.6 ± 5.7 | 81.0 ± 5.7 | <0.0001 |
| Men | 352 (71.0) | 113 (55.6) | 125 (58.7) | 0.002 |
| Higher education | 92 (18.7) | 19 (9.4) | 11 (5.2) | <0.00001 |
| Mean body mass index (kg/m²) | 26.5 ± 4.4 | 26.4 ± 3.9 | 25.9 ± 4.2 | 0.27 |
| Low sodium diet | 411 (84.9) | 167 (82.2) | 167 (78.4) | 0.02 |
| At least 1 lifestyle adjustment for HF | 421 (89.8) | 174 (87.9) | 173 (82.4) | 0.009 |
| Therapeutic education program for HF | 86 (18.2) | 27 (13.6) | 20 (9.5) | 0.06 |
| Regular physical activity | 260 (53.6) | 74 (36.5) | 62 (29.1) | <0.00001 |
| Cognitive impairment according to cardiologists before performing MIS-D | 20 (4.1) | 17 (8.7) | 72 (33.9) | <0.00001 |
| Atrial fibrillation | 155 (31.8) | 79 (39.3) | 79 (37.4) | 0.21 |
| Renal failure (eGFR <60 ml/min) | 281 (67.9) | 135 (78.5) | 154 (87.0) | 0.009 |
| Diabetes mellitus | 116 (23.9) | 57 (28.8) | 57 (26.9) | 0.11 |
| Chronic obstructive pulmonary disease | 92 (18.8) | 35 (17.6) | 38 (17.8) | 0.99 |
| Depression | 60 (12.5) | 34 (17.3) | 54 (25.7) | 0.0002 |
| Anemia (hemoglobin <11 g/L) | 54 (11.0) | 35 (17.7) | 29 (13.7) | 0.20 |
| Previous stroke | 51 (10.4) | 19 (9.5) | 43 (20.3) | 0.001 |
| Dysthyroidism | 45 (9.2) | 17 (8.5) | 16 (7.6) | 0.30 |
| Fall experienced in the last 12 months | 26 (5.4) | 12 (6.2) | 31 (14.8) | 0.006 |
| Cancer | 36 (7.4) | 10 (5.0) | 17 (8.0) | 0.39 |
| Malnutrition | 15 (3.1) | 6 (3.1) | 15 (7.2) | 0.12 |
| IADL | ||||
| 4 | 329 (67.1) | 100 (49.8) | 46 (21.9) | <0.00001 |
| 3 | 98 (20.0) | 48 (23.9) | 31 (14.8) | |
| 2 | 33 (6.7) | 25 (12.4) | 37 (17.6) | |
| 1 | 24 (4.9) | 21 (10.4) | 64 (30.5) | |
| 0 | 6 (1.2) | 7 (3.4) | 32 (15.2) |
∗ Overall difference among the 3 groups of MIS-D scores adjusted for age and gender.
The prevalence of abnormal delayed MIS (MIS-D [0-6] ) was 45.6% (n = 416) with 95% confidence interval 42.4 to 48.8. MIS-D [5-6] was found in 203 subjects (22.2%, 95% confidence interval 19.5 to 25.5) and MIS-D [0-4] in 213 subjects (23.4%, 95% confidence interval 20.6 to 26.1; Figure 1 and Table 1 ). Meanwhile, cardiologists spontaneously suspected cognitive impairment in 109 subjects (12%) before the cognitive evaluation (8.7% of subjects with MIS-D [5-6] and 34% of subjects with MIS-D [0-4] ; Figure 1 ).
Subjects had been diagnosed with HF on average 4.4 ± 4.8 years earlier ( Table 2 ). The most frequent underlying cardiovascular diseases were hypertension (59%), coronary heart disease (53%), dilated cardiomyopathy (37%), and atrial fibrillation (35%). Mean LVEF was 44 ± 12.0% and 24% patients had preserved ejection fraction (LVEF ≥50%). None of the treatments for HF was associated with memory impairment. Subjects had an average of 4.4 ± 1.3 different cardiac medications and 43% of them had at least 5 different cardiovascular drugs.
| Cardiovascular Characteristic | MIS-D [7-8] (n = 496) | MIS-D [5-6] (n = 203) | MIS-D [0-4] (n = 213) | p Value ∗ |
|---|---|---|---|---|
| Mean duration of HF disease (yrs) | 4.24 ± 5.05 | 4.62 ± 5.14 | 4.62 ± 4.00 | 0.50 |
| Mean time since last hospitalization (mo) | 4.35 ± 3.20 | 4.64 ± 3.23 | 4.14 ± 3.10 | 0.30 |
| Hypertension | 381 (77.1) | 162 (80.2) | 165 (77.5) | 0.67 |
| Coronary heart disease | 263 (53.2) | 102 (50.5) | 117 (55.2) | 0.46 |
| Dilated cardiomyopathy | 189 (38.3) | 72 (35.6) | 75 (35.5) | 0.98 |
| Severe valvulopathy | 94 (19.0) | 42 (20.8) | 56 (26.4) | 0.42 |
| NYHA classification | ||||
| Class I | 40 (8.1) | 6 (3.0) | 3 (1.4) | <0.00001 |
| Class II | 293 (59.4) | 117 (57.6) | 106 (49.8) | |
| Class III | 149 (30.2) | 76 (37.4) | 86 (40.4) | |
| Class IV | 11 (2.2) | 4 (1.9) | 18 (8.4) | |
| Mean systolic blood pressure (mm Hg) | 129.7 ± 18.0 | 131.5 ± 16.7 | 132.0 ± 19.2 | 0.54 |
| Mean diastolic blood pressure (mm Hg) | 75.7 ± 10.0 | 77.2 ± 10.0 | 76.1 ± 11.1 | 0.27 |
| Mean LVEF (%) | 43.4 ± 12.2 | 44.1 ± 12.1 | 43.4 ± 11.5 | 0.50 |
| Mean heart rate (beats/min) | 70.7 ± 14.2 | 74.2 ± 13.7 | 73.2 ± 16.2 | 0.05 |
| Cardiovascular treatment | ||||
| β Blockers | 396 (80.5) | 167 (82.2) | 164 (77.0) | 0.53 |
| Angiotensin-converting enzyme inhibitors | 331 (67.0) | 124 (61.4) | 135 (63.4) | 0.64 |
| Antiplatelet therapy | 262 (54.5) | 110 (55.8) | 123 (59.4) | 0.37 |
| Vitamin K antagonists | 200 (41.2) | 87 (43.5) | 85 (40.7) | 0.85 |
| Angiotensin II receptor blockers | 120 (24.4) | 59 (29.1) | 61 (28.6) | 0.37 |
| Aldosterone antagonists | 134 (27.2) | 49 (24.3) | 38 (17.9) | 0.11 |
| Calcium channel blockers | 73 (15.1) | 29 (14.6) | 34 (16.1) | 0.90 |
| Thiazide diuretics | 29 (5.9) | 17 (8.4) | 16 (7.5) | 0.56 |
| Loop diuretics | 429 (87.2) | 171 (84.2) | 190 (89.2) | 0.40 |
| Nitroglycerin (glyceryl trinitrate) | 91 (18.8) | 30 (15.2) | 44 (20.1) | 0.37 |
| Digoxin | 67 (13.6) | 33 (16.3) | 41 (19.3) | 0.43 |
∗ Overall difference among the 3 groups of MIS-D scores adjusted for age and gender.
The severity of the HF (NYHA classification) increased with lowering MIS-D score ( Table 2 and Figure 2 ). In contrast, we did not find any relation between MIS-D groups and the underlying cause of HF, the level of blood pressure, or LVEF when analyzed as a continuous variable or with the threshold of 50%. There was a marginal association with heart rate. There were no differences in HF treatments in the 3 groups.
