Methods in Stroke Prevention in the Wisconsin Native American Population Free

Native American (NA) individuals are affected more frequently by risk factors for cerebral and cardiovascular disease (CVD). CVD is the leading cause of death, and stroke is the sixth leading cause of disability [1]. NA individuals are thought to be affected more frequently by the risk factors for stroke such as obesity, diabetes, alcohol abuse, cigarette smoking, and carotid artery stenosis due to carotid atherosclerosis [2]. Research studies have been published describing the high incidence of stroke risk factors in NA populations [3], but there is little research published regarding the effectiveness of intervention to reduce stroke risk factors in this population despite it being known NA individuals are of some of the highest risk for cardiovascular and cerebrovascular events. This highlights as a healthcare community why we must work to better understand and intervene upon the driving factors behind their increased level of stroke risk factors and disease. The University of Wisconsin (UW) Stroke team was approached by the Oneida Nation and asked to collaborate on a project to assess and reduce the incidence of stroke in the Oneida tribe. As a result, a partnership was formed between the Oneida Comprehensive Health Division (OCHD) and the UW Stroke team to help determine effective strategies for reducing stroke risk. While the results we find will be specific to this tribe, developing new cognitive, serum, and noninvasive ultrasound biomarkers for high-risk patients may be generalizable to the entire population of Wisconsin and beyond.

Carotid atherosclerosis is thought to contribute to increased risk for stroke due to stenosis causing restricted and turbulent blood flow to the brain and concurrent ischemia as well as carotid plaque instability and risk of microemboli release [4‒6]. B-mode ultrasound and carotid strain imaging (CSI) are ultrasound techniques that have been used to characterize plaque composition and mechanical properties [6‒16]. Total plaque area (TPA), the sum of the area of all plaques present in both the right and left carotid arteries, has been used to determine plaque burden, associations with CVD risk factors and predict future adverse CVD events [17‒21]. In addition, B-mode ultrasound has been used to characterize plaque composition by extracting plaque texture features from the ultrasound image (specifically grayscale median) [22‒38]. Lower values of grayscale median have been associated with more inflammatory cells and larger lipid cores in plaque histology specimens [27]. Extremely low grayscale values near the plaque-lumen interface (often referred to as “black” areas) have been associated with higher ulceration scores [22, 27, 39]. Inflammatory cells, larger lipid cores, and ulcerations are all findings that suggest that a plaque is more vulnerable for rupture. Carotid strain indices have been associated with cognitive performance. Higher strain indices have been shown to be associated with poorer cognitive performance; and thus, it is believed that higher strain represents plaques that are unstable and may be releasing microemboli to the brain [9, 12‒14, 16, 40‒43]. While these microemboli may not result in a transient ischemic attack (TIA) or clinical stroke, they may be damaging to the small blood vessels of the brain and contributing to brain damage associated with cognitive decline [4‒6, 40]. Thus, studying carotid plaque presence and characteristics in conjunction with other traditional stroke risk factors such as hyperlipidemia, diabetes, hypertension, and obesity may provide insight as to how these variables contribute to the high levels of risk seen in NAs.

Additionally, inflammation is thought to contribute to plaque development, and novel circulating biomarkers may play a significant role in identifying early inflammatory processes. Elevated plasma levels of various proteins such as inflammatory cytokines have been associated with obesity, type 2 diabetes, and are thought to be representative of an environment that is favorable for the development of atherosclerosis. These proteins may represent a key link between atherosclerosis, obesity, and insulin resistance and with further investigation could become targets for future therapies used to manage risk factors [44, 45].

The “Stroke Prevention in the Wisconsin Native American Population” (clinicaltrials.gov identifier: NCT04382963) study is a longitudinal interventional study conducted with the Oneida Nation of Wisconsin and UW to implement and assess an intense intervention to prevent stroke and vascular cognitive decline and test for novel biomarkers of stroke in NAs. This project is studying 120 NAs who are age 55 or greater who are high risk (greater than three risk factors for stroke) or low risk for stroke (less than three risk factors for stroke). Participants are being studied at baseline and at 2 years (Fig. 1). Ten individuals will be selected to also complete testing at 1 year. The plan is to:

• 1.

  Identify 100 high-risk NAs and examine traditional and novel risk factors for stroke.

• 2.

  50 of the 100 high-risk NAs will be randomized to receive standard medical treatment and care for their risk factors. The other 50 will be randomized to receive standard medical care and treatment AND work with a health wellness coach to modify their stroke risk factors.

• 3.

  Identify 20 low-risk NAs and examine traditional and novel risk factors for stroke.

• 4.

  Determine if working with a health wellness coach to manage/modify stroke risk factors changes outcomes between the high-risk group with standard care/treatment and the high-risk group with standard care/treatment and health wellness coaching.

• 5.

  Characterize the differences in traditional and novel risk factors between the high-risk and low-risk group.

• 6.

  Characterize the changes from baseline to year two in the high-risk groups and low-risk group. Primary outcomes are changes in Montreal Cognitive Assessment – Vancouver Island Coastal First Nations (MoCA First Nations) (version 8.1) and TabCAT score at the beginning and end of the study as well as vascular cognitive decline and change in stroke risk factors. The overall effect of health coaching and the effect of each specific component in health wellness coaching will be evaluated by examining the measures in the American Heart Association (AHA) Seven Simple Rules for blood pressure, cholesterol, blood sugar, physical activity, body mass index (BMI), and smoking status. Secondary outcomes are change in plaque area, pulsatility index, correlation of plaque grayscale area texture features (grayscale median value) to stroke risk factors, and change in novel serum biomarkers for stroke (Table 1).

Implement and assess an intense intervention to prevent stroke and vascular cognitive decline. A population-wide and individual intervention will be conducted as part of this project. The population-wide intervention includes development of an educational website (with information on stroke risk factors, how to assess yourself for stroke risk factors, how to modify stroke risk factors, and information about the study), health events, community stroke risk education, and reminders to physicians about the need for tight control for stroke risk factors. The individual intervention consists of 100 high risk for stroke participants randomized to group 1 (standard of care/treatment) or group 2 (working with a health wellness coach standard of care/treatment). In addition, 20 low-risk individuals are being recruited to serve as controls for the study. The goal of health coaching is to achieve strict management of cholesterol levels, blood glucose, blood pressure, weight, diet, and exercise. Data on stroke risk factors are being collected at baseline, end of year one, and end of year two. Change in cognition over time and individual risk are outcomes to be assessed.

Test for novel biomarkers of stroke. Novel inflammatory biomarkers of stroke risk including dipeptidyl peptidase (DPPIV) and galectin 3 (Gal-3) previously identified and described by our group will be examined. Blood/plasma will be used to identify/quantify circulating proteins. Novel biomarkers will be examined for differences between the high-risk and low-risk groups.

Participants are recruited on a voluntary basis using an educational website with information on stroke risk factors and prevention, at events in the community staffed by wellness coaches and tribal members as well as through information provided at clinic appointments. The Oneida Nation is a tribe made up of approximately 17,272 NA individuals. Roughly 18% are living below the federal poverty line. 95% of the population has at least a high school diploma. 79% of the population uses the OCHD for their primary healthcare, and 69% of individuals have had a routine checkup in the past year. 54% of patients associated with the OCHD are considered obese, 21% smoke, and 20% report having binge drank in the last month [46].

Inclusion to the study requires participants to be actively receiving their health care at OCHD, age 55 or greater. The age criteria were determined based on evidence cited by the AHA that stroke risk begins to double at the age of 55 [47]. Participants are assigned to a group (group 1 high-risk standard of care/treatment, group 2 high-risk standard of care/treatment and health wellness coaching, and group 3 low-risk controls) after consenting and undergoing a health assessment. Participants are excluded from participation if they have dementia, inability to participate in health assessment and exercise programs, inability to provide informed consent, prior carotid procedure, or presence of a medical condition that would preclude participation in follow-up over the next 2 years.

Participants are provided with a written document to read and review outlining the goals of the study and what is involved in terms of time commitment, scope of the interviews, study test procedures, collection of blood samples, and participation in health and wellness coaching. Individuals who agree to participate are consented by a research study coordinator.

All participants will complete a baseline ultrasound examination to assess atherosclerotic load (plaque presence, thickness area, and stability assessed by strain indices), neurocognitive testing, health assessment with a physician, health resource conversation with a wellness coach and blood analysis for glucose, cholesterol, and key proteins felt to be biomarkers of stroke [45, 48‒50] (Table 2). High-risk participants are randomly placed into two groups using simple randomization. One group will receive standard of care therapy which is a letter informing them of their lab results, results of the health assessment, carotid ultrasound results, and recommendations for management of their stroke risk factors. This letter may also be sent to their primary care provider if indicated by the participant at the time of consent. The other group will receive the same plus intensive health wellness coaching. At the end of the 2-year follow-up, all groups will be reassessed for atherosclerotic plaque progression or regression and its stability, serum biomarker response to therapy interventions, successful risk factor modification, vascular cognitive decline, and incidence of stroke and TIA. Groups will be compared for change in both risk factors and outcomes.

Power analysis was based on change in MoCA raw score. We anticipate no change from baseline in MoCA First Nations score in the intensive health wellness coaching group but anticipate a decline of 2 units in the standard of care group after 2 years. In a previous study, it was found that the standard deviation of change in MoCA scores over 3.5 years was at 2.6 units with a test-retest reliability coefficient of 0.92 [51]. With a sample size of 50 subjects in each treatment arm, our statistical power to detect the 2-unit difference is 96.7% using the 2.6-unit standard deviation using a two-sided two-sample t test with a significance level of 0.05. Given our follow-up sample will be taken at 2 years, we would anticipate that the reliability coefficient would actually be higher than 0.92 and thus reduce the presumed standard deviation of change scores. Thus, we believe our estimate here to be a conservative estimate of the statistical power of the study for measuring cognitive decline. While the study was powered based on cognitive decline, additional outcome measures will also be collected.

Ten high-risk coached participants will be compared at baseline and at year one, and all study participants will be compared at baseline and 2 years into the program for stroke, TIA, vascular cognitive decline, and correction of risk factors. These include changes in carotid atherosclerosis, blood pressure, BMI, hemoglobin A1C, smoking, glucose, cholesterol, and protein markers. Blood genomic signatures of patients with and without history of stroke will be analyzed from the RNA-seq data to identify differentially expressed genes and confirmed by using real-time PCR. The goal of this research was to devise a panel of clinical biomarkers in people with high stroke risk and target therapies to prevent onset of stroke. Data analysis will study between groups, recidivism, rates of compliance with risk factor interventions, incidence of stroke or TIA, presence of atherosclerotic plaque and unstable atherosclerotic plaque, presence of vascular cognitive decline, presence of major modifiable clinical risk factors, blood pressure, serum glucose or hemoglobin A1C, smoking, obesity measures, and whether these interventions do modify these risk factors and whether they affect outcomes. The study design allows analysis to take place both at baseline and at completion of the study. At baseline, the study compares the 100 high-risk patients to the control population to determine if the unique new biomarkers of vascular cognitive decline, carotid plaque instability, and protein levels are indeed differentially expressed in the high-risk patients compared to controls. At the completion of this study, the analysis will also allow us to determine differences between groups [4, 52], each outcome and risk factor change over time, as well as did intervention improve both outcomes and risk factors. Specifically, which factors are modifiable by interventions. During the period of this study, it is possible that actual stroke rates will not be high enough to see a significant difference. However, previous studies [4] have shown that vascular cognitive decline changes can be significantly altered in 1-year time; therefore, most significant analysis will be on outcomes of dementia/vascular cognitive decline, and additional primary outcomes of TIA, clinical risk factors, allowing us to study whether successful coaching changes those risk factors and which are most important for essential outcomes (vascular cognitive decline, TIA, and stroke). The additional primary outcomes are individual change in numerical values as well as the number of patients in each group who are able to demonstrate a change in modifiable risk factors that are included in the National Guidelines [52] (systolic blood pressure, diastolic blood pressure, hemoglobin A1C, total cholesterol, LDL-C, HDL-C, BMI, and smoking status) (Table 1). χ² or Fisher exact tests and Mann-Whitney U tests will be performed to examine differences between groups [52], primary outcome variable will be MoCA First Nations total raw score. If significant, secondary analysis will examine Tablet-based Cognitive Assessment Tool (TabCAT) favorites (rote verbal learning and memory), match (processing speed), flanker (executive functions), and line orientation (visuospatial abilities). Expected outcome: There will be more change in group 2 (intensive health coaching) toward the National Guideline Standards for acceptable clinical values for cardiovascular risk factors. For the secondary outcomes, correlation of ultrasound plaque grayscale median values with clinical risk factors and changes in ultrasound biomarkers (TPA, pulsatility index) [1, 5, 6, 22, 27, 37, 39, 45], as well as multiple inflammatory cytokines we feel to be unique serum biomarkers [45, 48] of symptomatic carotid atherosclerotic disease will be evaluated. In addition, compliance rates will also be assessed.

Demographic information will be collected via questions regarding potential participant’s basic information (i.e., age, gender), languages spoken, and educational history. The health assessment and history are collected via a targeted medical interview using a health assessment tool and completion of the Stroke Risk Assessment from the American Stroke Association (ASA) [53]. Through the medical interview, the number of stroke risk factors a participant has is determined and they are placed in either the low-risk control group, the high-risk standard of care group, or high-risk health coaching group. The stroke risk factors used for categorization include hypertension or current elevated blood pressure, diabetes mellitus, current smoker, body mass ≥30, history of TIA/stroke or coronary artery disease. Study participants also have their blood pressure measured at the time of the medical interview and height and weight collected for calculation of BMI.

Blood samples are collected from participants in the outpatient laboratory at OCHD by credentialed medical laboratory scientists. Participants are required to fast for a minimum of 8 h prior to laboratory draws. The study laboratory panel consists of fasting lipid panel (total cholesterol, HDL, LDL non-LDL, triglycerides), hemoglobin A1C, and novel serum biomarkers. All blood samples are processed and reported by the UW Health core laboratory (Madison, WI). Samples collected for assessment of stroke biomarkers are processed at the principal investigator’s research laboratory (UW-Madison, Madison, WI).

The comprehensive carotid ultrasound (CCU) is performed at baseline, year one, and year two. The CCU examination consists of bilateral imaging with B-mode, color Doppler, and pulse wave Doppler. The B-mode images acquired are sweeps from the proximal common carotid artery (CCA) to the carotid bulb, bifurcation, and in to the internal (ICA) and external carotid (ECA) arteries. The CCA, ICA, and ECA are imaged in transverse and longitudinal planes. Color Doppler images are acquired in the transverse and longitudinal planes of the CCA, carotid bulb, ICA, and ECA. Pulse wave Doppler is used to measure the peak systolic velocity, end diastolic velocity, resistive index (RI), pulsatility index (PI), and systolic to diastolic ratio (S/D). Velocities are recorded in the CCA, ICA (proximal, mid, and distal vessel), ECA, and vertebral arteries. Plaque presence is recorded based on location (CCA, bulb, ICA, and ECA). The vertebral artery is evaluated for velocity, flow direction, and qualitative properties of the Doppler waveform (i.e., presence of reversal flow associated with a partial subclavian steal).

These methods have been previously described [42]. Carotid strain imaging will be performed at baseline. Briefly, strain indices are measured using radio-frequency (RF data) from ultrasound imaging. RF data are acquired using the Siemens Acuson S2000 ultrasound system and the 18L6 transducer (Siemens Ultrasound, Malvern, PA, USA) at a center frequency of 11.4 MHz and a sampling rate of 40 MHz [42]. All imaging for RF data collection is performed in the longitudinal plane, and data are acquired from the mid-CCA, carotid bulb, and ICA bilaterally. Plaque regions and vessel walls are segmented over two cardiac cycles [42]. Interframe displacements are estimated from a digitized RF data loop and summed over a cardiac cycle. The strain tensor is calculated from a gradient of the accumulated displacement over several time points in the cardiac cycle [42].

TPA and grayscale median value are extracted from ultrasound images of plaque using the LifeQ Medical Plaque Analysis software (Nicosia, Cyprus) [7, 22]. These methods have been previously described [7, 22]. Briefly, each plaque is identified, segmented, and measured using the LifeQ software. For each individual plaque image, the plaque area and grayscale median value are calculated. All plaque areas for each participant are summed to provide the TPA [7].

The cognitive testing performed is based on the MoCA [54‒59], a widely used tool to assess for vascular cognitive decline [56, 57]. The MoCA First Nations adaption for use with indigenous populations (including the Oneida of the Thames) from the coast of British Columbia (Vancouver Island Coastal First Nations) (version 8.1) will be used as this is a novel and promising screening tool for use with NA populations. Additional measures include the TabCAT, devised for use in diverse settings and for longitudinal monitoring of cognition in adults [54]. Domains assessed include memory, executive function, language, and visuospatial skills in the evaluation of cognitive impairment. Subtests have demonstrated neuroanatomical validity: favorites performance has been associated with medial temporal lobe volumes and fornix fractional anisotropy, and Match with frontal, parietal, and basal ganglia volumes and with corpus callosum fractional anisotropy [58, 59]. Specific subtests were carefully selected to assess the impact of cerebrovascular disease in the brain (executive dysfunction and processing speed) and rule out underlying neurodegenerative processes (Alzheimer’s disease). Cognitive testing is conducted by study personnel who are trained by a neuropsychologist specifically to conduct cognitive testing and score the testing battery in a standardized format per study protocol. Once cognitive testing data are collected, it is scored and entered into a secure online study database. All processes are conducted with oversight from a laboratory neuropsychologist.

Self-report questionnaires of cultural identity/background (acculturation) and emotional state will be collected. The Native American Acculturation Scale (NAAS) is a 20-item multiple-choice questionnaire that assesses an individual’s level of acculturation [60]. Scores are obtained by calculating the average responses of the 20 items assigned. Items included load onto six distinct domains: language (5 items), identity (2 items), friendships (3 items), behaviors (4 items), generational/geographic background (5 items), and attitudes (1 item). Scores range from a low of 1, representing low acculturation (or high NA identity) to a high of 5, indicating high acculturation (or high mainstream American identity). A score of 3 indicates biculturalism. The NASS higher levels of acculturation will likely translate to more reliable and valid neuropsychological scores, and this will be better aligned with published normative data of the mainstream US population.

Participants will complete a self-report mood questionnaire. The Kessler Screening Scale for Psychological Distress (K6, Kessler et al., 2002) [61] is a widely used and robust screener to identify the presence of psychological distress. The questionnaire consists of 6 items and respondents’ rate including how often (in the past month) they felt nervous, hopeless, restless or fidgety, so sad that nothing could cheer them up, that everything was an effort, and worthless. Mitchell and Beals [62] studied the applicability and utility of the K6 in American Indian communities of two closely related Northern Plains tribes and a Southwestern tribe and found the K6 to be an appropriate measure to assess for mood disorders in these populations.

Patients randomized to the high-risk health wellness coaching group establish a relationship with a study-specific wellness coach employed by OCHD. The wellness coach works with participants throughout the 2-year duration of the study on health and wellness goals. Participants are contacted by health coaches on a weekly basis for coaching sessions via their preferred contact method (face-to-face in person or virtually, by phone call, or by email) to set goals and increase adherence to lifestyle modifications and medical interventions. Participants meet with their coaches at the initial study visit and write health goals in a health journal with guidance from the wellness coach. The participant, along with the coach, list ways they can work towards completing these health goals. At each follow-up meeting, the coach reviews progress toward achieving the health goals and assists the participants with ways to continuously work toward these goals and journal their efforts. Examples of health and wellness goals include eating healthy food, losing weight, increasing physical activity. Coaches who lead participants in study-related coaching sessions underwent a specific online “stroke risk factors” training course that included reading material and videos. The online curriculum consisted of 10 modules: overview of the study, overview of stroke as a disease, overview of clinical research, dyslipidemia, hypertension, diabetes, medications and avoiding stroke, smoking, diet and exercise, teaching children about stroke. Coaches also participate in discussion regarding stroke risk factors with the principal investigator of the study.

At completion of the study, we hope to identify risk factors for CVD and cerebrovascular disease that particularly affect this population in addition to characterizing if there are novel stroke biomarkers prevalent in this population that can inform stroke risk across all populations. By studying the population with the highest risk, it is more likely we will identify the widest range of risk factors for a population, and this can inform the risk factors we identify and modify in other populations. However, the population in this study is unique. While they fit under the umbrella of NA, they are only from the Wisconsin region. Variables for stroke risk in this study population may be present in other NA populations, and the findings of this study could influence future research in similar populations, but we must be careful in extension of our findings over other populations, even those who identify as NA.

It is essential when studying populations that are more frequently affected by a disease that we look beyond the traditional risk factors. In this case, we are using modern vascular atherosclerotic imaging techniques and novel vascular inflammatory proteins to see if there are novel biomarker measures that better quantify and help understand the stroke risk in this population. If we do not attempt to identify new ways to understand risk of disease, we cannot expect advancement in the identification, management, and prevention of the disease which becomes of utmost importance when collaborating with populations at extremely high risk.

The authors would like to thank Ms. Panouchi Fu for assistance in manuscript preparation.

This protocol was approved by the University of Wisconsin – Madison Health Sciences Institutional Review Board, approval number 2019-1550. All participants in the study provided their written informed consent prior to participation in any study activities.

H. Cress, S. Wilbrand, U. Wesley, G. Morel Valdés, T. Hess, M. Metoxen, A. Riesenberg, C. Vandenberg, C. Blohowiak, J. Kennard, and D. Danforth: none. C. Mitchell: Elsevier, author textbook chapters, and W. L. Gore & Associates contracted research grants to University of Wisconsin-Madison, consulting Acoustic Range Estimates and funded by R01 HL147866. T. Varghese: Research agreement with Siemens Ultrasound and funded by R01 HL147866. J. Maybock and R. Dempsey: funded by R01 HL147866.

This work was supported by the Partnership Education and Research Committee (PERC) Opportunity Grants Program, Wisconsin Partnership Program, University of Wisconsin School of Medicine and Public Health, Award ID: 4355.

H. Cress: data curation (equal), formal analysis (equal), investigation (equal), writing – original draft (lead), and writing – review and editing (lead). C. Mitchell and S. Wilbrand: conceptualization (equal), data curation (equal), formal analysis (equal), funding acquisition (equal), investigation (equal), methodology (equal), project administration (equal), resources (equal), software (equal), supervision (equal), writing – original draft (equal), and writing – review and editing (equal). U. Wesley: conceptualization (equal), data curation (equal), formal analysis (equal), funding acquisition (equal), investigation (equal), methodology (equal), project administration (equal), resources (equal), software (equal), supervision (equal), writing – original draft (supporting), writing – review and editing (equal). G. Morel Valdés: conceptualization (equal), data curation (equal), formal analysis (equal), investigation (equal), methodology (equal), resources (equal), software (equal), supervision (equal), writing – original draft (equal), and writing – review and editing (equal). T. Hess: methodology (equal), formal analysis (equal), and writing – review and editing (equal). T. Varghese: conceptualization (equal), data curation (equal), formal analysis (equal), investigation (equal), methodology (equal), resources (equal), software (equal), supervision (equal), writing – original draft (supporting), and writing – review and editing (equal). J. Maybock, A. Riesenberg, and D. Danforth: data curation (equal), project administration (equal), writing – original draft (supporting), and writing – review and editing (equal). M. Metoxen and J. Kennard: conceptualization (equal), methodology (equal), writing – original draft (supporting), and writing – review and editing (equal). C. Vandenberg and C. Blohowiak: data curation (equal), writing – original draft (supporting), and writing – review and editing (equal). R. Dempsey: conceptualization (lead), data curation (equal), formal analysis (equal), funding acquisition (lead), investigation (lead), methodology (lead), project administration (lead), resources (equal), software (equal), supervision (equal), writing – original draft (equal), and writing – review and editing (equal).

Data availability information are not applicable as this manuscript outlines the protocol and methods for this study and does not report data.