Dementia is a leading contributor to the global burden of disease due to progressive and unrelenting deterioration of the capacity of an individual to live independently. Vascular etiology is the second most common cause of dementia1 and cardiovascular disease (CVD) results in faster decline.2 Mild cognitive impairment (MCI), which may indicate early dementia, occurs frequently in acute coronary syndrome (ACS) patients.3 Prevalence rates vary widely (9-85%), depending on timing and assessment method used. A recent report indicated 63% had MCI on screening4 4-wk post-ACS. Data based on full neurocognitive assessment when recovery has progressed (4-mo post-ACS) are required for confirmation and to inform inpatient screening that reliably predicts MCI. Early detection of MCI enables timely intervention to potentially slow the progression of MCI to dementia and enable customizing of education and support.
METHODS
We recruited consecutive ACS inpatients from two university hospitals in metropolitan Sydney, Australia. Included patients were free from dementia and proficient in English. Inpatient screening for MCI was undertaken using the Montreal Cognitive Assessment (MoCA)5 and the Hopkins Verbal Learning Test (HVLT).6 Together, these tests are sensitive to the cognitive decline common in vascular cognitive impairment. Mild cognitive impairment was classified as MoCA score <26 and/or HVLT >= -1 z-score. Participants were then assessed 4-mo post-ACS using a full neurocognitive battery conducted by qualified neuropsychologists in an accredited testing center (see Supplemental Digital Content Table 1, available at: http://links.lww.com/JCRP/A432). Testing determined if patients met clinical criteria for MCI, as well as classification of amnestic and nonamnestic MCI subtypes. Classification of MCI was determined by >=1 test from any domain falling below expected levels (age and education standardized, 1.5-SD decrement). This would represent a substantial clinical deviation from expected levels and unlikely to be overly sensitive. While requiring impairment on >=2 tests of the same domain would be more stringent,7 it would be less pragmatic in terms of testing time. Cases with multiple-domain MCI are more likely to progress to dementia than single-domain MCI (especially if nonamnestic single domain), which is more likely to remain stable or revert to normal range cognition. Testing also documented impairment in specific cognitive domains. Previous work in ACS patients attending cardiac rehabilitation (CR) indicated reducing MCI rates from 63% (4 wk) and 25% (4 mo) and 15% loss to follow-up. We therefore estimated a sample of 80 would ensure sufficient patients with MCI at 4 mo for the analyses planned.
RESULTS
Of the ACS patients screened during admission, 24/81 (30%) opted to participate in full neurocognitive testing at 4 mo, providing the basis for this study. Participants who did not undergo further testing cited difficulties with travel (n = 23), work (n = 11), and lack of interest (n = 13), or were lost to follow-up (n = 10). Participants who undertook full testing did not differ in age, sex, education, employment, baseline MCI screens, or depressive symptoms.
Participant mean age was 64.8 +/- 11.1 yr, 88% were men, 54% were married/partnered, 54% were tertiary educated, and 46% were employed. ST-segment elevation myocardial infarction occurred in 42% and modifiable CVD risk factors were common including hypertension (71%) and risky alcohol intake (17%). At 4 mo few had depressive symptoms (3/24) and 42% had participated in CR.
At baseline (inpatient), MCI was identified in 13/24 (54%) using the MoCA, 5/24 (21%) using the HVLT, and 16/24 (67%) using these tests combined (Table). At 4 mo, 19/24 (79%) were classified as MCI, with 17% having amnestic MCI, 42% having nonamnestic MCI, and 42% having multi-domain MCI. The cognitive domains most commonly found to be impaired were processing speed (38%), executive function (33%) and visuospatial skills (21%), learning (17%), and memory (17%), whereas the domains attention/working memory were rarely impaired (8%).
Of the participants who screened positive for MCI as inpatients, 88% (14/16) were classified as having MCI at 4 mo in full neurocognitive assessment. Of the participants who screened negative for MCI as inpatients, 63% (5/8) were classified as having MCI at 4 mo (Table). As such, the use of the combined MoCA and HVLT screening during hospitalization had moderate accuracy (71%), sensitivity (74%), and specificity (60%) for predicting MCI diagnosis at 4 mo. Ten patients had attended CR at 4 mo, four of whom went from cognitively intact to MCI, five continued to have MCI, and one no longer had MCI.
DISCUSSION
This exploratory study reports results of full neurocognitive testing in ACS patients at 4-mo follow-up. Mild cognitive impairment was evident in >75% of study participants, much higher than reported in a previous review3 or in CR samples.4 Mild cognitive impairment has a high risk of progressing to dementia and reported association with worse prognosis following ACS.1 This study indicates that many patients screen positive for MCI as inpatients and MCI continues to emerge out to 4 mo post-ACS; therefore, screening is recommended as inpatients and during recovery. Combined MoCA and HVLT screens may be suitable for this screening need, but the emergence of many patients having MCI at 4 mo suggests that a lower threshold for testing as inpatients and retesting may be warranted. Further research, with a larger sample size, is required to delineate the characteristics of those whom are false negatives on MCI screening as inpatients, to determine the long-term cognitive trajectories and extent/severity of impairment of those with MCI following ACS.
This exploratory study was limited by the small sample size engaged in 4-mo neurocognitive testing and emphasizes the challenges associated with widespread cognitive assessment. Full neurocognitive testing takes approximately 3-4 hr and occurs in purpose-built clinics creating difficulties in travel, time, and scheduling for participants. A reduced, targeted, or even virtual set of neurocognitive tests may improve feasibility in this older population.8 Importantly, however, this study showed that multiple cognitive domains may be impaired post-ACS, including executive functioning and processing speed. Tests for these domains should be included in any reduced assessment, as these domains are typically linked to vascular cognitive impairment and can impede patient insight and motivation and limit the capacity for self-care behaviors.9 Adding to these difficulties, <30% completed >=1 session of CR, the most effective intervention to support secondary prevention in ACS.10 Direct, systematic referral to CR for patients screening positive for MCI as inpatients is recommended to promote prevention11 and the study results strengthen the recommendation for CR in this population. Education strategies using a variety of methods to promote learning, cognitive interventions, and referral to CR are needed for this population.
REFERENCES