The American Heart Association defines cardiovascular health (CVH) using “Life’s Essential Eight” (LE8) behaviors and factors, including diet, physical activity, sleep, nicotine exposure, BMI, blood pressure (BP), lipids, and glucose [1]. In adults, high CVH, that is, favorable levels of all 8 CVH behaviors and factors, is associated with a multitude of benefits, including a decreased incidence of cardiovascular (CV) diseases, cancer, depression, and dementia, as well as a longer, healthier lifespan [2]. However, less than 5%‐10% of the adult US population has high CVH, and its prevalence decreases with age; by age 8 years, over 20% of children no longer have high CVH [3,4]. Accumulation of nonoptimal CVH factors and behaviors early in life translates to greater CV risk and burden of CV disease later in life.

The current generation of children is the first in centuries to have a shorter projected lifespan than their parents [5,6]. One of the main drivers of this decline in health is the growing prevalence of nonideal CV risk factors and behaviors, particularly obesity, among youth. While obesity has been well-studied in childhood, the advantage of examining CVH is that it includes both the behaviors that contribute to adiposity as well as the clinical consequences, including dyslipidemia, hypertension, and diabetes that may stem from it. CVH provides a summary of the global burden of chronic disease risk in an understandable framework for patients and actionable points of intervention in clinical care.

Current definitions and measurement methods for CVH in youth are limited because of measurement burden and uncertain applicability across developmental transitions. For example, completing a gold standard dietary assessment with two 24-hour dietary recalls can take more than 60 minutes and is infeasible in clinical practice. Additionally, although there are clinical guidelines to define “abnormal” levels of clinical metrics such as non-HDL cholesterol, these do not address the potential risk associated with levels that do not meet this threshold but are quite high compared with those of age-matched peers and may portend risk. Evidence is needed to develop and refine developmentally appropriate and clinically feasible ways of measuring CVH, particularly in infancy and early childhood [2].

Comprehensive measurement of CVH and its determinants in children demands rigorous collection of clinical, behavioral, and psychosocial data—a combination previously unavailable in any contemporary, diverse pediatric cohort. Established pediatric studies such as the US’s Environmental Influences on Child Health Outcomes cohort, the Young Finns Study, and Australia’s Childhood Determinants of Adult Health study have served as benchmarks in the field. However, these cohorts have not collected data on modern contextual factors contemporaneously with CVH; for example, the Young Finns and CDAH began data collection over 30 years ago [7,8], prior to the widespread use of cellphones. Although the Environmental Influences on Child Health Outcomes cohort features a large, contemporary sample with detailed socioenvironmental assessments, its comprehensive CVH data collection is restricted to BMI, while critical factors such as lipids and glucose are only available in a subcohort [9].

These cohorts and others have contributed valuable data to the International Childhood Cardiovascular Cohort (i3C) Consortium [10]. The i3C Consortium has provided the first evidence that childhood cardiovascular risk factors correlate with subclinical cardiovascular disease in early adulthood [11,12]. Subsequent analysis of i3C found that cardiovascular risk factors in childhood acted on adult cardiovascular disease both indirectly and directly [13], providing a compelling rationale for additional research. The current cohort was therefore established to address significant gaps in the literature and enhance our understanding of CVH and related social and environmental risk factors in youth populations.

The overarching goal of the Young Hearts (YH) study is to identify trajectories of CVH and its component factors and behaviors from birth through age 20 years among a large, diverse cohort of children and their families. Using an innovative study design for a pragmatic cohort using technology-enabled data collection, we plan to recruit children (or their guardians) from 6 city-wide health systems in Chicago. We will use 2 vehicles to collect data across all 8 components of the American Heart Association’s LE8: we will capture CVH behaviors via detailed e-surveys, and we will link to available clinical CVH factors in the electronic health record (EHR) of participating health systems. All participants will be followed longitudinally for 2 years to better understand the changing trajectories of pediatric and adolescent CVH. Specifically, we plan to accomplish the following specific aims.

First, we plan to identify trajectories of CVH and its component factors and behaviors from birth to young adulthood within a large, diverse cohort of participants (target enrollment >7000) across the city of Chicago. Then, we will identify predictors of CVH trajectories, including a broad range of markers of resilience and vulnerability, including sociodemographic characteristics, economic stability, residential and neighborhood environment, social context and support, neuropsychological skills, parental CVH, and family health history. Last, we will develop an accurate and generalizable CVH growth curve for use in clinical care to target prevention efforts to children at high risk of declining CVH and subsequent CVD.

The YH study is being conducted by a team of researchers at Northwestern University, in partnership with 7 health systems across Chicago, IL, USA. Children (or legal guardians of children) ages 0 to 18 years will be recruited from across Chicago. Participants will be asked to complete a developmentally appropriate and clinically feasible set of online surveys and questionnaires at 3 e-visits, spaced approximately 12 months apart, for a total of approximately 2 years of follow-up. These surveys and questionnaires will cover CVH behaviors, as well as markers of resilience and vulnerability. EHRs will be used to identify recent measurements of cardiovascular risk factors and related comorbidities, including BP, lipid levels, and familial hypocholesterolemia; these will be linked to their survey responses. The CVH behaviors and factors will be combined to calculate the primary CVH measure and age- and sex-specific CVH z-scores. Leveraging the repeated outcome assessments, additional analyses will examine within-child changes in CVH z-scores, behaviors, and factors over the 2-year period.

This is a fully remote study—there will be no in-person contact of study staff with participants. Northwestern University’s Institutional Review Board (IRB) approved the study protocol. Data use agreements for EHR data transfers were set up with each of the 7 participating health systems, Northwestern Medicine, Ann & Robert H Lurie Children’s Hospital of Chicago (Lurie), AllianceChicago (Alliance), Advocate Aurora Health, Rush University Medical Center (Rush), University of Illinois, and University of Chicago (UChicago); each health system’s IRB reviewed and approved the study protocol for site-specific distribution of study materials to eligible patients. Researchers interested in collaborating with YH or proposing an ancillary study should contact the corresponding author.

All children aged 0 to 18 years will be eligible to enroll in the study. Eighteen-year-olds, as legal adults, will provide their informed consent electronically. Each child aged 0 to 17 years will need a parent or legal guardian (hereafter “guardian”) to participate with them; informed consent for minors will be provided electronically by guardians, as well as assent from participants aged 12‐17 years. As participating children aged 16 and 17 years turn 18 years old before their second or third e-visit and gain legal independence, we will obtain electronic informed consent forms from them directly; the same applies to the collection of assent from pre-teens who turn 12 during the course of the study. At enrollment of their first participating child, guardians will be given the option to enroll additional children; this will facilitate future sibling studies.

Recruited participants (and guardians) will provide informed consent for linkage to their EHR. A unique study ID will be used to identify EHRs for YH participants. These IDs will ensure that no identifiers are shared between institutions and that all study data is not identifiable. All data will be recorded and stored in HIPAA-compliant, password-protected databases that can only be accessed by approved study personnel. No protected health information will be shared with employers, insurers, or nonresearch personnel.

Participants in the study will be provided with a US $20 incentive for each completed e-visit, for a maximum total compensation of US $60 per child. Compensation will be distributed via gift card, using the Hyperwallet (pre-2023) or Amazon (2023 onwards) platforms. Additionally, all participants who complete their baseline e-visit are entered into a raffle to win a US $150 Amazon gift card. The raffle will be conducted for every 1000 participants who complete their baseline e-visit until the end of recruitment. Each participant is included in exactly one drawing. These raffles will continue through the follow-up surveys, with one raffle occurring for every 1000 participants who complete their 1st and 2nd follow-up surveys.

We are working with several partners, including the 7 health systems listed above, to achieve our recruitment goal of at least 7000 children aged 0 to 18 years from across Chicago. Our recruitment efforts will be divided into 2 strategies: population-based and site-based recruitment.

The population-based recruitment strategy will use targeted approaches in specific neighborhoods across Chicago to engage potential participants. This includes heavy placement of advertisements in areas with high concentrations of targeted socioeconomic, race, and ethnic groups. Additionally, we will partner with community organizations and other studies to distribute brochures. For instance, we will collaborate with the Patrick M. Magoon Institute for Healthy Communities, providing them with flyers for distribution. Of these, the majority will be given to their Mobile Health Program, which offers medical services in underserved neighborhoods throughout Chicago. The remaining flyers will be given to their volunteer coordinator to display at food pantries and other community events. Other efforts will include coordinating with leadership and staff at the Northwestern branch of the All of Us Research Program, a nationwide research program to advance precision medicine, and the Early Start program, a nonprofit organization promoting child development in Illinois, to distribute flyers to promote participation in YH.

This study will involve the linkage of participant survey data with existing EHRs, so we will collaborate with participating health systems to conduct site-based recruitment. This strategy, recruiting current patients at each site to participate in YH, will maximize data availability on CV factors. The mode of recruitment at each site will vary, but is primarily electronic, including emails, SMS texting, and MyChart messages; a small number of postcards will be sent as well. Each site will send 2 or 3 additional electronic reminders after their initial contact, inviting participation in YH, by either email, SMS messaging, or MyChart messages.

The YH study will be a longitudinal e-cohort study of children and adolescents in the Chicagoland area. After recruitment and consent, participants will complete electronic surveys at 3 e-visits (baseline and 1 and 2 years after baseline), each of which takes approximately 30 min to complete via the web-based portal through the Eureka Research Platform (University of California, San Francisco) [14]. Participants aged 18 years and older will complete all surveys about themselves. Participants aged 12‐17 years, once assented, will be asked to complete a set of age-appropriate surveys. Participants aged 0‐17 years will have guardians fill out surveys on their behalf. EHRs available from 7 participating health systems will be linked to participant survey responses. Figure 1 describes the recruitment of participants by site. Data collected in surveys and from the electronic medical record are described below.

Each e-visit consists of answering 6 questionnaires in total; all questions must be answered to open subsequent questionnaires. If any questionnaires are incomplete, the web app sends reminders on days 3, 7, 14, and 30 via email, text if a phone number exists with SMS capabilities, and notifications through the mobile app (only for participants who download the app). Although participation in YH is done fully remotely, study staff will be accessible by phone or email to all potential and consented participants during regular work hours, Monday through Friday from 9 AM to 5 PM. Recruitment messages sent to eligible patients will include the shared study email and a direct phone number for a designated staff member.

The Eureka Research Platform was used to enroll, obtain consent, and collect survey data. Eureka is a National Institutes of Health–funded (5U2CEB021881) direct-to-participant eConsent and data collection system. The Platform uses a HIPAA-compliant design template and technology developed in the Health eHeart Study that has been used for web-based recruitment of over 260,000 consented research participants [15]. Additional information about the platform is available [14,16].

Once participants have completed their consent forms, Eureka sends reminders for them to complete their first e-visit (baseline surveys) 4 times in the first 30 days. These reminders are also scheduled for the 2nd and 3rd e-visits (1 and 2 years post-baseline). After the initial 30-day period, study staff will contact participants who have not completed their e-visits via text, email, or phone calls. These contacts notify participants that the e-visits are incomplete and remind them of the financial incentive for completing them.

To encourage participant engagement between e-visits, the study staff, together with input from the study investigators, write newsletters about the scientific content of the study. For example, the first newsletter introduced study participants to the concept of CVH scores and their components. Other topics for past newsletters include sleep, healthy eating, and physical activity. Future topics will cover physical activity and body composition, BP, cholesterol, and blood glucose. These newsletters are written for lay audiences and emphasize the role of each topic within childhood and adolescence, providing targeted content that is age-appropriate. In addition, after the initial study visit, participants and their guardians are notified via email of their child’s component LE8 behavioral CVH scores (diet, physical activity, sleep, and nicotine exposure).

The YH Community Advisory Board (CAB) was established in May 2023. Recruitment efforts targeted professionals working with young children via LinkedIn messages and guardians of YH participants through email outreach. Ensuring diversity within the CAB is critical, as it brings varied insights and perspectives that enhance decision-making and maximize the study’s impact. As such, guardians were randomly selected for CAB membership based on criteria including race, income, gender, and country of origin. The CAB has provided critical feedback on key areas such as survey enhancement, recruitment strategies, retention efforts, community education activities, and the dissemination of study results. Notably implemented suggestions from the CAB include more comprehensive study descriptions and reassurance of incentives prior to surveys, incorporating graphics and engaging images in youth surveys, and strategically placing flyers in key community locations to boost cohort diversity.

To address the survey response rates among guardians of our younger participants, we collaborated with Northwestern’s Center for Community Health in hosting a one-time Youth Shared Resource Panel. Our Shared Resource Panel was comprised of 8 members, aged 14‐17 years old. The youth panel provided specific feedback on how to increase survey participation among minors, who respond to surveys at a lower rate than their guardians. Two key suggestions emerged from the panel: first, improving the youths’ understanding of the importance of the research and CVH before asking that they complete the surveys; and second, requiring guardians to enter their children’s cellphone numbers so that survey notifications are sent via SMS rather than email.

After the raw data have been processed as described above, the EHR data will be linked to the e-Survey data. For each component of clinical CVH, we will identify the measurement that took place closest in calendar time to the e-visit and record it as the YH measurement. All measurements of clinical CVH components will be available for study analyses as appropriate. This will facilitate our definition of CVH as it relies on using only the information gathered under current clinical guidelines (as detailed above) for our study to remain pragmatic and immediately clinically translatable.

For each of the health behavior and clinical age-specific CVH metrics, we will combine survey responses or data from EHRs to create one composite measure per metric that can be used to describe where a child falls relative to the population distribution at their current age. The age- and sex-specific population distributions are all assumed to be normal, so each child will be assigned to what amounts to a z-score and a percentile that corresponds to the age-specific population distribution. This idea has been detailed previously by our team [48]. These z-scores will incorporate all CVH factors and behaviors, providing a global measure of CVH in infants, children, and adolescents that can be used to study associations between CVH and outcomes across all ages. This z-score will be calculated based on the average score across all 8 components of CVH, with age-specific scoring detailed in Table S1 in Multimedia Appendix 1.

As one of our goals is to create a reference standard for age-specific CVH z-scores, our sample size calculation centered around our ability to precisely estimate particular percentiles of a standard normal distribution. Thus, we chose to estimate the sample sizes necessary to precisely estimate the 50th percentile (median/mean), 16th percentile (1st SD), and 2.5th percentile (2nd SD) of a standard normal distribution at one particular age. Estimating these particular quantiles as accurately as possible should provide reassurance that 50% of children fall below the median, and that 68% and 95% of children fall within 1 and 2 SDs of the mean, respectively.

After investigating a range of sample sizes, we concluded that a sample size of 400 children at each age would give us sufficient precision to estimate the median, 16th percentile, and 2.5th percentile (Table S3 in Multimedia Appendix 1). Larger sample sizes are needed as the percentile nears 0, as we would expect fewer children to fall there. Therefore, we aimed to recruit a minimum of 7000 infants, children, and adolescents.

We plan to use generalized linear models to study the associations between novel cardiovascular risk and resilience factors and global metrics of pediatric CVH, including adapted Life’s Essential 8 scores (defined in Table S1 in Multimedia Appendix 1) and CVH z-scores. Initial analyses will use only data collected at the baseline visit, studying cross-sectional associations. Further analyses will use generalized mixed linear models using repeated measures of CVH z-scores on the same children, to create pediatric CVH growth curves showing the distribution of CVH z-scores over childhood. We will also use group-based trajectory modeling to identify children who have a similar underlying trajectory of CVH z-scores. Once children have been assigned to a trajectory group, we will use multinomial regression models to identify risk and resilience factors that are associated with group membership. Age-specific multiple imputation (M=50) of missing data is planned for all analyses; Rubin’s rules will be used to combine estimates across datasets.