Methods

The REGARDS study is a longitudinal study consisting of 30,239 community-dwelling black and white adults 45 years or older enrolled between January 2003 and October 2007. The study aimed to examine the causes for excess stroke mortality in the Stroke Belt of southeastern U.S. (North Carolina, South Carolina, Georgia, Tennessee, Alabama, Mississippi, Arkansas, and Louisiana) and among blacks. Initial eligibility criteria included potential participants having a name, phone number and address in the Genesys database. Exclusion criteria included race other than black or white, active treatment for cancer, medical conditions that would prevent long-term participation, inability to understand survey questions as judged by the telephone interviewer, residence in or inclusion on a waiting list for a nursing home, or inability to communicate in English. Demographic information and medical history data were collected via a computer-assisted telephone interview (CATI) system followed by an in-home physical examination 3 ^– 4 weeks later that included blood pressure measurements, electrocardiogram (ECG) recording, and blood draw. Participants were followed every 6 months with cognitive assessments and suspected stroke events were identified. Verbal consent was obtained initially on the telephone then written informed consent was obtained during the in-home physical exam. For the purpose of this analysis, we excluded participants with baseline history of stroke, missing data on AF and no follow-up cognitive assessments, resulting in a sample of 25,980 participants included in the analysis ( Figure 1 ). The study methods and procedures were reviewed and approved by institutional review boards at all participating institutions, and by an external observational study monitoring board selected by the funding agency.

Consort diagram of REGARDS participants.

Baseline AF was determined by the presence of AF or atrial flutter on ECG or self-reported physician diagnosis. Self-reported AF was defined as a positive response to the question, “Has a physician or a health professional ever told you that you had atrial fibrillation?” . REGARDS ECG tracings were read and coded at a central reading center by analysts who were masked to other REGARDS data (Epidemiological Cardiology Research Center, Wake Forest School of Medicine, Winston-Salem, NC, USA). AF was identified from ECG tracings using the Standard Minnesota ECG Classification , then verified by a physician.

To characterize cognitive function and cognitive decline, cognitive assessments were performed at baseline and repeated over time. The six-item screener (SIS) was added to the baseline assessment in December 2003 and was conducted annually thereafter. The SIS is a brief screening measure that assesses global cognitive function with 3-item temporal orientation and 3-item delayed recall. The SIS score ranges from 0 to6; a score ≤4 suggests cognitive impairment. We assessed incident impairment based on time to first SIS score ≤4. As a sensitivity analysis, we also examined incident impairment based on the time until 2 consecutive assessments with SIS score ≤4. A three-test battery which consists of the Consortium to Establish a Registry for Alzheimer Disease (CERAD) Word List Learning (WLL) (range, 0–30), Word List Delayed Recall (WLD) (range, 0–10) and animal naming test, was introduced in 2006 and conducted at 2-year intervals. The WLL score is the total number of words immediately recalled on a 10-item, three-trial word list learning task. The WLD task involves recalling the number of words recalled after a filled 5-minute delay. The animal naming score is the total number of spontaneously named animals in 60 seconds. The Letter F Fluency test was introduced into follow-up telephone assessments in 2008 and administered at 2-year intervals along with WLL, WLD, animal naming, and the rest of the short battery. This fluency measure required participants to name as many words as they could that begin with the letter ‘F’ in 60 seconds. The NINDS/CSN 5-minute protocol was recommended by the National Institute of Neurological Disorders and Stroke-Canadian Stroke Network Vascular Cognitive Impairment Harmonization Standards for use in studies calling for very brief assessments, epidemiological studies, and/or telephone administration. This protocol includes subsets of the Montreal Cognitive Assessment (MoCA) consisting of a 5-word memory registration, 5-word delayed memory recall (5 points), 6-item temporal orientation (6 points), and 1-letter F phonemic fluency (1 point if >10 words that begin with the letter “F” generated in 60 seconds), was implemented into the follow-up telephone assessment at 2-year intervals. ^, During telephone administration, the spatial orientation items (place and city) were modified such that the participant was asked his or her street address and city (confirmed by the interviewer via a pre-populated field in the computer script). Higher scores on cognitive assessments reflect better performance.

Methods of determination of incident stroke have previously been reported. During telephone follow-up, participants or their proxies were asked about events that required hospitalization, as well as physician evaluations for stroke-like symptoms. For potential strokes, medical records were requested and adjudicated by a team of stroke experts. Stroke events were defined by the 1989 World Health Organization (WHO) classification or classified as clinical strokes if imaging was consistent with a stroke and categorized as hemorrhagic or ischemic.

Baseline age was calculated from the day the baseline telephone interview was conducted and self-reported birth date. Sex, race (black or white), education (<high school, high school graduate, some college, college graduate and above) and region (stroke belt, stroke buckle or non-stroke belt) were also self-reported. Urban and or rural status was defined by residence based on percentage of the census track residing inside of urban areas/clusters; the status was rural if ≤25% urban, mixed between 25 and 75% urban, and urban if ≥75% urban. Behaviors included in this analysis were current cigarette smoking (pack-years) and exercise habits (none, 1–3 days per week and 4 days per week). Coronary artery disease was defined as self-reported physician diagnosis of myocardial infarction (MI), self-reported coronary revascularization, or evidence of MI on ECG. Hypertension was defined as systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg on average of two measurements or self-reported use of hypertension medication. Dyslipidemia was defined as total cholesterol ≥240 mg/dL, LDL ≥160, HDL ≤40 or self-reported lipid-lowering drug use. Left ventricular hypertrophy (LVH) assessed by ECG was defined using Sokolow-Lyon criteria. Diabetes was determined by fasting glucose ≥126 mg/dL, non-fasting glucose ≥200 mg/dL or self-reported use of diabetes medications. Estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI and MDRD equations. C-reactive protein (CRP) was measured in plasma samples collected at baseline using a validated, high sensitivity, particle enhanced immunonephelometric assay. Waist circumference was measured over skin or lightweight clothing at the midpoint between the lowest rib on the right side and the top of the iliac crest using a cloth tape measure at the end of expiration. Sex-specific waist circumference criteria was assigned to participants in three categories: low (<80 cm for women, <94 cm for men), moderate (80–88 cm for women, 94–102 cm for men) or high (>88 cm for women, >102 cm for men) based on the National Institutes of Health and international guidelines. Body mass index (BMI) was calculated from height and weight measurements and categorized participants as underweight <18.5 kg/m ² , normal 18.5 to 24.5 kg/m ² , overweight 25 to 29.9 kg/m ² , or obese ≥30 kg/m ² . The Center for Epidemiologic Studies Depression (CESD) 4-item version was used to evaluate depressive symptoms.

Baseline characteristics and cardiovascular risk factors were compared between participants with and without AF using student t-tests for continuous variables and chi-square tests of association for categorical variables. The relationships between AF and baseline cognitive performance were assessed in multivariable general linear regression models. The relationships between baseline AF and longitudinal cognitive performance was examined using multivariable linear mixed models, including the interaction between AF and time to assess the association of AF and cognition over time. We report these results as least squares mean and standard errors, adjusted as above, as well as adjusted for incident stroke. Survival analyses using Cox proportional hazards models were also employed to estimate the association between AF and time to first assessment of incident impairment (as defined above). All models relating AF with cognitive performance and impairment were assessed in incremental models, as follows: model 1 was adjusted for age, sex, race, region and education; model 2 was further adjusted for hypertension, dyslipidemia, history of heart disease, LVH, diabetes, obesity, current smoking and depressive symptoms; and model 3 (longitudinal analyses) was further adjusted for incident stroke in longitudinal models only. Effect modification by sex, race and incident stroke were considered. Statistical significance for all analyses was set at p<0.05. All analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC).

Results

A total of 25,980 (mean age = 64±9.3 years; 56% women; 39% black) participants had baseline assessments of AF and cognitive data. Of these, 2,168 (8.3%) had AF at baseline of which 1784 (6.9%) were identified by self-reported physician diagnosis and 129 (0.5%) by ECG. Table 1 shows the demographic characteristics, comorbidities and socio-demographic factors among those with and without AF. Those with AF were significantly older, more likely to be white, and displayed more symptoms of depression compared to those without AF. Participants with AF at baseline were more likely to have a greater prevalence of cardiovascular risk factors and educational attainment <high school completion.

Stay updated, free articles. Join our Telegram channel

Join