24‐h Movement Behaviors and Incident Hypertension Among Adolescents: A Prospective Study From the ABCD Study

ABSTRACT

Hypertension during adolescence is a growing public health concern with long‐term cardiovascular implications. This study examined whether adherence to 24‐h movement behavior guidelines—physical activity, screen time, and sleep—was associated with the prevalence and incidence of hypertension among U.S. adolescents, and whether these associations varied by social determinants of health (SDoH). We analyzed data from 2067 participants in the Adolescent Brain Cognitive Development (ABCD) Study, followed across 21 U.S. cities from 2018–2020 (T0) to 2020–2022 (T1). Movement behaviors were dichotomized using international recommendations, and SDoH was measured using the Child Opportunity Index. Hypertension was defined as systolic or diastolic blood pressure ≥ 95th percentile for age, sex, and height. Adolescents with screen time > 2 h/day had higher hypertension prevalence (1.9%) and incidence (4.0%) than those within recommended limits. In crude analyses, adolescents meeting physical activity guidelines showed a higher incidence of hypertension than those not meeting them; however, this association was not statistically significant in adjusted models (RR = 0.87; 95% CI: 0.42–4.17), suggesting that confounding rather than a true effect may explain the crude finding. Adherence to two or more movement behavior recommendations was associated with a reduced risk of hypertension (RR = 0.31; 95% CI: 0.09–0.98), though this estimate is based on a small number of events and should be interpreted with caution. SDoH did not modify these associations. These findings underscore the importance of combined behavioral adherence in reducing hypertension risk and support the integration of lifestyle interventions for promoting cardiovascular health during adolescence.

1. Introduction

Hypertension in adolescents is becoming an increasing public health concern in the United States [1]; A global review revealed alarming pediatric hypertension rates: 13% in boys and 9.6% in girls [2], indicating a growing burden of cardiovascular risk from a young age. Early‐onset hypertension can persist into adulthood, leading to significant adverse health outcomes, including heart disease and stroke. Among the key behavioral risk factors, physical activity, sleep, and screen time have been recognized as critical determinants of blood pressure regulation in youth [3].

The 24‐h movement behavior paradigm provides an integrated framework for understanding how daily time‐use patterns, physical activity, sedentary behavior, and sleep jointly influence health outcomes [4]. These behaviors are inherently co‐dependent, as time allocated to one behavior necessarily displaces time spent in another within a fixed 24‐h day [4]. Traditional approaches that examine these behaviors in isolation may therefore overlook their combined and interactive effects [5]. In contrast, the 24‐h framework emphasizes the overall composition and balance of daily behaviors, aligning with international guidelines recommending sufficient physical activity, limited screen time, and adequate sleep. Emerging evidence suggests that adherence to multiple movement behavior recommendations confers greater health benefits than meeting individual guidelines alone. However, the lack of standardized measurements further complicates the evaluation of the effect of adherence to 24‐h movement behavior on cardiovascular health indicators among youth [3]. In addition, the long‐term impact of these combined behaviors on the development of hypertension in adolescents remains insufficiently understood [6], particularly when considering their joint effects [7].

The social environment plays a critical role in shaping health behaviors and outcomes, with social determinants of health (SDoH) [8] influencing access to healthy foods, recreational spaces, and economic opportunities, including neighborhood deprivation. These disparities may increase hypertension risk [9] by limiting opportunities to engage in health‐promoting behaviors such as regular physical activity and adequate sleep [10]. In this context, neighborhood opportunity may not only influence exposure to these behaviors but also modify their health effects. For example, the benefits of adhering to recommended movement behaviors may be attenuated in under‐resourced environments due to structural barriers or chronic stressors [11]. In this context, neighborhood opportunity may not only influence exposure to these behaviors but also modify their health effects. For example, the benefits of adhering to recommended movement behaviors may be attenuated in under‐resourced environments due to structural barriers or chronic stressors.

Examining 24‐h movement behaviors in national data in different cultural and economic contexts can improve these gaps. Using data from the Adolescent Brain Cognitive Development (ABCD) Study [12], which includes a diverse cohort of adolescents followed across 21 U.S. cities, we hypothesized that adherence to 24‐h movement behavior guidelines would be associated with lower hypertension risk. We further examined whether these associations varied across social determinants of health (SDoH), as measured by the Child Opportunity Index [13]. These findings may inform targeted public health strategies to prevent hypertension and improve cardiovascular health in youth, particularly in socially disadvantaged communities.

2. Methods

Data used to prepare this article were obtained from the ABCD Study 5.1 (doi:10.15154/z563‐zd24, https://abcdstudy.org), held in the NIMH Data Archive (NDA).

2.1. Study Design; Sample Size; Eligibility Criteria; Ethical Aspects

The ABCD study at abcdstudy.org represents the most extensive ongoing study in the U.S. focusing on adolescent brain development and overall health. This study spans 21 research sites nationwide and is designed to track the progress of approximately 11 877 adolescents, starting from their preteen years and continuing into their early adult life. The participants were selected using a method that involved a random sampling of both public and private elementary schools. This strategy encompasses over 20% of the U.S. population in the eligible age group. Before data collection, consent was obtained from parents or guardians, and assent of the adolescents. Trained field workers collected and managed data via a series of standardized in‐person assessments, including neuroimaging, cognitive testing, and surveys on health and lifestyle, ensuring a comprehensive dataset on adolescent development [14]. In‐person visits, up to 4–5 h [14], were meticulously structured to create a supportive and safe participant environment, ensuring high‐quality, reliable data collection. All measurements followed this sequence: physical health assessments, blood pressure and height (43 min), lifestyle questionnaires (24 min), and environmental and social inquiries (20 min). The study's methodology, including the approach to recruiting participants, the design of the questionnaires answers, the protocols for data collection, and the techniques used for processing neuroimaging data are described elsewhere [12].

This research analyzed the initial cardiovascular data set, including blood pressure readings and Fitbit assessments, as part of the 2nd ABCD Study follow‐up from 2018 to 2020. Participants (n = 3320) from 21 US cities were followed from 2018 to 2020 (T0 for this study) to 2020–2022 (T1 for this study). Participants with complete data on Fitbit‐tracked sleep, blood pressure, height, and relevant covariates were included in the analysis. The current analysis was limited to participants who had provided complete data in several areas: healthy behavior patterns, environmental information, and results from blood pressure measurements T0 and T1 and other relevant variables (as illustrated in Figure 1). To ensure the robustness of our findings, we conducted sensitivity analyses that compared the blood pressure readings and other critical variables between those who provided complete data and those who did not. The findings did not reveal significant differences in the distribution of biological sexes and all the covariates, as detailed in Table S1.

FIGURE 1.

Final sample size flowchart for each time point. * There was no blood pressure measurement at baseline, only at the second follow‐up.

The foundational ABCD study received its ethical approval from the Institutional Review Board of the University of California, San Diego. This study adhered to the moral guidelines outlined in the 1964 Declaration of Helsinki, as amended in Edinburgh, Scotland, in 2000. Our organization accessed the data through a formal agreement (specifically, the National Institute of Mental Health [NIMH] Data Archive Data Use Certification), which was established with the National Institutes of Health (NIH). Given that our analysis was based solely on de‐identified data, it was categorized as exempt from a comprehensive review by the institutional review board at UTHealth Houston (referenced as HSC‐SPH‐22‐0663).

2.2. Outcome

Data collection procedures were described previously [12, 14, 15]. Following the American Academic of Pediatrics' Clinical Practice Guideline for Screening and Management of High Blood Pressure in Children and Adolescents, systolic (SBP) and diastolic (DBP) blood pressure (mmHg) were measured using the Omron HEM‐7200 twice after 10‐min rest, with a 5‐min interval between measurements, and the lowest reading was recorded [16]. The interobserver coefficients of variation were below 5% for both BP levels. Height was measured with a portable stadiometer to the nearest 0.1 cm, with the patient barefoot and nothing on their head. Weight was measured to the nearest 0.1 kg using a digital scale. The outcome of this study was hypertension incidence, defined as SBP or DBP above the 95th percentile for age and height according to American Academic of Pediatrics [16]. Blood pressure was measured in triplicate, and the lowest value was used to define hypertension to minimize potential overestimation of prevalence and incidence; sensitivity analyses using the mean of measurements yielded comparable results. Age‐, sex‐, and height‐specific blood pressure percentiles were recalculated at each assessment using CDC Growth Charts [17] to account for developmental changes over time.

Hypertension prevalence and incidence estimation: We analyzed longitudinal data from the ABCD Study to estimate the prevalence and incidence of hypertension in adolescents. At T0 (2018–2020), blood pressure was first assessed to determine hypertension prevalence; those with existing hypertension were excluded from subsequent incidence analyses. Participants with hypertension at baseline were excluded from incidence analyses so that follow‐up was restricted to adolescents free of hypertension at study entry. Incident hypertension was identified over the 2‐year follow‐up using blood pressure measurements obtained at T1. Cumulative incidence was calculated as the proportion of new cases among participants at risk at baseline, allowing incident cases to be distinguished from preexisting hypertension. A flowchart (Figure 1) details the sample size, exclusion criteria, and follow‐up steps. Analyses were adjusted for key confounders, including age, sex, and race and ethnicity, to enhance the validity of the results.

2.3. Main Exposures

Healthy Sleep duration: Participants continuously wore a Fitbit Charge HR 2 accelerometer for three weeks. This device measured total sleep time and REM sleep duration. The Fitbit Charge HR 2 combines movement data and heart rate (HR) to classify sleep stages according to its preprogrammed algorithm. Days with more than 180 min of sleep were considered valid and included in the analysis, consistent with the ABCD study's processing methods [18]. The weekly sleep summaries calculated the average total and REM sleep minutes. Previous research has demonstrated the reliability and accuracy of Fitbit devices in estimating accumulated sleep [18]. We also adopted the data extraction, filtering, and processing protocols established by the ABCD Study, including sleep thresholds. Sleep health was measured by daily sleep time and categorized into two levels: 9 to 11 h (healthy sleep), < 9 or > 11 h/night (inadequate sleep duration), based on the 24 h‐movement behavior methodology proposed [19].

Physical activity (PA) levels: During the same period, Fitbits collected daily non‐sleep steps and moderate‐to‐vigorous PA. Moderate activity is defined as the number of minutes of moderately active (3–5.9 METS) time observed in all valid minutes after exclusions. Vigorous is defined as a number of minutes of vigorously active (> 6 METS) time observed in all valid minutes after exclusions. To compile MVPA, moderate and vigorous activity were combined into one variable. Following other studies, participants were included in the final analysis if their device registered days with > 599 min of recorded wake wear time within each participant's 3‐week study period, with a minimum requirement of 1000 steps per day [20].

Screen time levels: The ABCD Youth Screen Time Survey, a self‐reported tool [21], assessed total screen time and social media usage. Participants indicated the average number of hours spent daily on six different screen activities: watching television shows or movies, streaming videos (e.g., YouTube), playing video games, texting, video chatting (e.g., Skype, Facetime), and using social media (e.g., Facebook, Instagram, Twitter). A weighted average was then calculated to represent typical daily use across the week, using the formula ((weekday average × 5) + (weekend average × 2))/7 [22]. The weighted average for each screen time modality was reported as a continuous variable, and total screen time was derived by summing these weighted averages across all activities. Thereafter, participants were classified into two groups according to the 24‐h movement guideline [19] on total screen time: ≤ 2 h/day and > 2 h/day.

24‐Hour Movement Behaviors (24 h‐MB): In this study, we considered exposure as the composition (mix) of 24‐h movement behaviors, including moderate‐to‐vigorous physical activity (mean = 65.3 min/day), screen time (mean = 10.8 h/day), and sleep duration (mean = 6.9 h/day). Below, we describe the measurement methods and classification criteria according to the 24‐h movement guidelines [19]. We selected theoretical definitions of the constructs of interest (physical activity, sedentary behavior, and sleep time) [17]. To operationalize variables for physical activity (moderate and vigorous), screen time (TV watching, playing games, computer use, studying or reading, and passive commuting), and sleep duration (essential and incidental). Using the operationalized variables categorized into groups, we applied joint effects data analysis to summarize the behaviors and further classified the variables into four groups: no behaviors meet the criteria, 1 behavior meets the criteria, 2 behaviors meet the criteria, and 3 behaviors meet the criteria, according to the 24‐h movement guidelines [19]. We also established the variable change in 24 h‐MB based on the distribution in 24 h‐MB categories at T0 and T1. Subjects were classified into the following categories: Same number of behaviors within guidelines (T0 → T1); Decrease 1+ behavior within guidelines (T0 → T1); and Increase 1+ behavior (T0 → T1).

2.4. Potential Confounders

The potential confounders for this study were divided into two groups, Contextual Factors and Individual Factors, and described below:

2.4.1. Contextual Factors

• –ABCD Study Research Cities: We used the city population density to account for relevant environmental factors [23]. The cities are: Los Angeles, California (two centers) • Miami, Florida • Tulsa, Oklahoma • Charleston, South Carolina • Portland, Oregon • Menlo Park, California • San Diego, California • Boulder, Colorado • Gainesville, Florida • Baltimore, Maryland • Ann Arbor, Michigan • Minneapolis, Minnesota • Pittsburgh, Pennsylvania • Rochester, New York • Salt Lake City, Utah • Burlington, Vermont • Milwaukee, Wisconsin • Richmond, Virginia • St. Louis, Missouri • New Haven, Connecticut.
• –Seasonality: A variable was computed by recoding the original variable “blood pressure date” into “seasonality,” as follows: winter (from 21st December to 20th March, coded as 1), autumn (from 21st September to 20th December, coded as 2), spring (from 21st March to 20th June, coded as 3), and summer (from 21st June to 20th September, coded as 4), as performed in previous studies.
• –SDoH: We used the Child Opportunity Index 2.0 (COI) [13] to quantify social determinants of health (SDoH). COI scores were georeferenced and linked to participants' residential ZIP codes to enable accurate assessment of SDOH. The COI, created by Acevedo‐Garcia et al. [13], identifies disparities in neighborhood conditions, particularly for marginalized populations. It combines 29 variables across education, health/environment, and social/economic domains. COI scores, standardized nationally, cover 72 000 U.S. neighborhoods, including major metropolitan areas. COI was categorized into very low to very high opportunity, with gaps of 80+ points indicating hoarding and narrower gaps representing sharing [13, 24].

2.4.2. Individual Factors

• –Race and ethnicity: was self‐reported using the American Community Survey, and participants were grouped into five categories: Asian, Black, Hispanic, Other, or White [12]. The Other category included adolescents identified as multiracial or who belonged to an ethnicity too small to create a separate category.
• –Biological sex: biological sex (male; female) according to what is described in the birth certificate.
• –Age years: Continuous calculation (data collection date—date of birth).
• –Parent education was categorized as follows: a high school diploma or less, college or technical diploma, Bachelor's degree, and Graduate degree.
• –Puberty status: maturational status was evaluated annually using sex‐specific pubertal developmental scales [25], with both parent and child reports averaged to increase reliability. Scores ranged from 1 (prepubescent) to 5 (postpuberty). Given that maturational status may change over time, participants were categorized into no changes, a one‐category change, or a two‐to‐three‐category change over the follow‐up period.
• –Waist circumference: Waist circumference was used as an indicator of abdominal obesity due to its stronger predictive value for hypertension compared to other anthropometric measures, reflecting the central distribution of adiposity that is closely linked to elevated blood pressure [26]. was measured at the midpoint between the lowest rib cage and the top of the iliac crest with a nonelastic tape to the nearest 0.1 cm. The intraobserver technical errors of measurement were between 0.53 and 1.75 cm, and interobserver reliability was greater than 94.9% 23 for this circumference.

2.5. Statistical Analysis

• –Descriptive analyses were performed using mean (continuous variables) and percentage (categorical variables) and respective 95% confidence intervals (95% CI). We calculated the cumulative incidence and 95% confidence interval (CI) of hypertension outcomes for total and principal exposures at the two‐year follow‐up. Based on Zou's methods [27], we estimated the magnitude of these associations was subsequently expressed in unadjusted and adjusted relative risk (RR) and 95% CI; we applied the multilevel Poisson regression models using mixed effects intercept to estimate the effect of 24‐h movement behaviors on hypertension incidence [28, 29]. The context variable used was the center. Additionally, we employed multilevel linear regression models with fixed effects to analyze the relationship between blood pressure levels and neighborhood noise levels, treated as a continuous variable.
• –For the adjusted analysis, we developed a conceptual framework [30] previously separated into five levels (the association of these levels is not shown): (1) contextual factors; (2) ethnicity, sex, and age (years); (3) parental education; (4) waist circumference; (5) mainly exposures (24‐h movement behaviors). In this model, variables were controlled for those at the same level and those in a higher level [30]. p‐values ≤ 0.20 were adopted in the univariate analysis [30] (as necessary to include a variable in the multivariate analysis and, then, it was entered through the hierarchical model method following the levels above) or when there was more than 10% modification in RR of any variable already in the model.
• –Multilevel analyses were performed with two objectives: (1) to test the associations between BP categories and two separate measures of individual exposures; and (2) to test the extent to which city‐specific characteristics and contextual variables mediate the associations between hypertension categories and 24‐h movement behaviors. We also conducted an exploratory analysis, testing for effect modification by including an interaction term for sex and time, which provided evidence suggesting sex differences in maturational. Significance was set at p‐values < 0.05, and Stata 18 (Stata Corp., College Station, TX, USA) was used for all statistical calculations.

3. Results

The sociodemographic, behavioral, and clinical characteristics of the study population are presented in Table 1. The mean age of participants increased from 11.9 years at the second follow‐up (2018/2020) to 14.1 years at the fourth follow‐up (2020/2022). The distribution of biological sex remained stable, with a slight increase in the proportion of males. The ethnic composition was predominantly White Caucasian (T0: 57%; T1: 56.7), followed by Hispanic and African American participants, with no significant changes between follow‐ups. The Child Opportunity Index indicated that most participants were in the high/very high opportunity categories. However, there was a slight decrease by T1, with corresponding increases in the very low/low and moderate categories.

TABLE 1.

Sociodemographic, behavioral, and clinical characteristics by biological sex, ABCD Study.

Variables	T0 (2018/2020)		T1 (2020/2022)
	n = 2067		n = 1702
	Mean or %	(95% CI)	Mean or %	(95% CI)
Age (years)	11.9	(11.8; 12.0)	14.1	(14.0; 14.2)
Biological sex birth (%)
Female	48.5	(46.2; 50.7)	47.2	(45.4; 49.0)
Male	51.5	(49.3; 53.8)	52.8	(51.0; 54.6)
Race and ethnicity (%)
White Caucasian	57.0	(54.7; 59.2)	56.7	(54.9; 58.4)
African American	10.6	(9.3; 12.0)	9.9	(8.9; 11.0)
Hispanic	18.6	(16.9; 20.5)	20.5	(19.1; 22.0)
Don't know	2.9	(2.3; 3.8)	2.3	(1.9; 3.0)
Mixed or other ethnicity a	10.9	(9.6; 12.4)	10.5	(9.5; 11.7)
Child opportunity index a
Very low/Low	23.9	(22.0; 25.9)	25.6	(24.0; 27.1)
Moderate	13.3	(11.8; 14.9)	15.6	(14.4; 16.9)
High/Very high	62.8	(60.6; 65.0)	58.8	(57.1; 60.6)
Waist circumference (cm)	72.6	(72.1; 73.2)	81.6	(80.9; 82.3)
Screen time
> 2 h/day	68.6	(66.5; 70.6)	64.2	(61.8; 66.5)
≤ 2 h/day	31.4	(29.4; 33.5)	35.8	(33.5; 38.2)
Sleep time
< 9 or > 11 h/night	77.8	(75.9; 79.6)	75.1	(72.9; 77.2)
9–11 h/night	22.2	(20.4; 24.1)	24.9	(22.8; 27.1)
Physical activity
< 60 min/d	80.7	(78.9; 82.4)	91.1	(91.1; 92.4)
≥ 60 min/d	19.3	(17.6; 21.1)	8.9	(8.9; 10.4)
Total 24‐h movement guideline adherence
0 Guideline	44.6	(42.4; 46.8)	42.0	(39.6; 44.4)
1 Guideline	39.5	(37.4; 41.7)	41.8	(39.4; 44.3)
2 Guidelines	14.3	(12.8; 15.9)	14.5	(12.9; 16.4)
3 Guidelines	1.6	(1.1; 2.3)	1.6	(1.1; 2.4)
Systolic Blood Pressure (mm Hg)	102.5	(102.1; 103.0)	108.6	(108.2; 109.0)
Diastolic Blood Pressure (mm Hg)	60.3	(59.9; 60.6)	63.0	(62.6; 63.3)
Hypertension	1.5 b	(1.1; 2.2)	2.9 c	(2.4; 3.5)

Note: Significant (α < 0.05) associations are in bold.

^a Childhood Opportunity Index is a Social Determinants of Health indicator.

^b Prevalence.

^c Incidence.

The prevalence of hypertension at T0 was 1.5% (95% CI, 1.1%–2.2%), with a cumulative incidence of 2.9% (95% CI, 2.4%–3.5%) over the two‐year period. Clinical trends also indicated increases in waist circumference and blood pressure from T0 to T1. Regarding behavioral patterns, daily screen time ≥ 2 h declined over time, while the proportion of adolescents with < 2 h per day increased. Sleep duration remained largely unchanged, with most participants reporting either less than 8 or more than 11 h of sleep per night. Physical activity levels declined, with a growing number of participants engaging in < 60 min of daily moderate‐to‐vigorous activity by T1. Adherence to 24‐h movement behavior guidelines remained low at T1: 44.6% of adolescents met none of the three recommendations, 39.5% met one, 14.3% met two, and only 1.6% met all three, indicating substantial room for improvement in guideline compliance and highlighting clear opportunities for intervention.

Table 2 details hypertension's prevalence and cumulative incidence based on exposure between T0 and T1. Participants with screen time of > 2 h per day had higher prevalence and incidence of hypertension compared to those with ≤ 2 h per day. Adherence to the sleep guideline showed similar hypertension prevalence and incidence for those sleeping < 9 or > 11 h per night and those sleeping 9–11 h per night. Participants who did not meet the physical activity recommendation had a lower prevalence of hypertension (1.2%) than those who met the recommendation, and cumulative incidence was also higher among those meeting the guideline in crude analyses. However, this association was attenuated after adjustment and was no longer statistically significant. When examining total guideline adherence, findings revealed that adolescents who did not meet recommendations for any of the 3 behaviors had a lower prevalence but higher incidence of hypertension than those who met recommendations for one or two behaviors. No cases of hypertension were observed among participants meeting all three movement guidelines.

TABLE 2.

Prevalence and cumulative incidence (%) and respective confidence interval 95 (95% CI) of the outcomes per 100 individuals according to exposure.

Exposures	Hypertension (2nd Follow‐up; 2018/2020)		Hypertension (4th Follow‐up; 2020/2022)
	Prevalence	(95% CI)	Incidence	(95% CI)
Screen time
> 2 h/day	1.9	(1.5; 2.4)	4.0	(3.0; 5.4)
≤ 2 h/day	1.0	(0.6; 1.7)	1.7	(0.9; 3.2)
Sleep time
< 9 or > 11 h/night	1.5	(1.0; 2.3)	3.1	(2.3; 4.3)
9–11 h/night	1.6	(0.7; 3.4)	3.2	(1.9; 5.4)
Physical activity
< 60 min/d	1.2	(0.8; 1.9)	3.1	(2.3; 4.1)
≥ 60 min/d	2.7	(1.5; 4.7)	4.7	(2.3; 9.6)
Number of 24‐h movement behaviors guideline adherence
0	1.0	(0.5; 1.9)	4.2	(2.9; 5.9)
1	2.1	(1.3; 3.4)	3.2	(2.1; 4.8)
2	1.8	(0.7; 4.2)	0.8	(0.2; 3.3)
3	0	(no observations)	0	(no observations)

Note: Significant (α < 0.05) associations are in bold.

Table 3 presents the relative risk (RR) of hypertension at T1, adjusted for potential confounders, including ABCD Study Research Cities, seasonality, sex, age, ethnicity, waist circumference, and social determinants of health (SDoH). Participants who met the screen time component of the 24‐h movement guidelines had a significantly lower risk of hypertension compared to those who did not meet the guideline, though this association was not consistent across different SDoH levels. For sleep duration, no significant differences in hypertension risk were observed between those who met and did not meet the guideline recommendations across all SDoH categories. Adherence to the physical activity guidelines also showed no significant impact on hypertension risk, with similar relative risks for those meeting versus not meeting the recommended 60 min or more of daily activity. Adherence to 2 movement behavior guidelines was associated with lower hypertension risk in the overall sample. This pattern was not consistent across SDoH strata, and stratified estimates should be interpreted cautiously because of wide confidence intervals and sparse data. No hypertension cases were observed among participants meeting all 3 guidelines; however, this finding should be interpreted with caution and not as evidence of zero risk.

TABLE 3.

Relative risk (RR) and respective confidence interval 95% (95% CI) by multilevel Poisson according to each main exposure.

Exposures	Hypertension Incidence RR (T1; 2020–2022)
	RR (95% CI)	RR (95% CI)	RR (95% CI)	RR (95% CI)
	Full Sample	SDoH Very low or Low	SDoH Moderate	SDoH High or Very High
Screen time
> 2 h/day	Ref	Ref	Ref	Ref
≤ 2 h/day	0.47 (0.22; 0.98)	0.11 (0.09; 1.20)	0.59 (0.11; 3.21)	0.58 (0.22; 1.49)
Sleep time
< 9 or > 11 h/night	Ref	Ref	Ref	Ref
9–11 h/night	1.72 (0.71; 4.17)	2.38 (0.29; 9.57)	1.05 (0.10; 1.82)	1.52 (0.42; 5.43)
Physical activity
< 60 min/d	Ref	Ref	Ref	Ref
≥ 60 min/d	0.87 (0.42; 4.17)	0.56 (0.08; 3.96)	1.67 (0.31; 2.98)	0.85 (0.34; 2.12)
Number of 24‐h movement behaviors guideline adherence
0	Ref	Ref	Ref	Ref
1	0.92 (0.58; 1.46)	1.01 (0.41; 2.48)	0.81 (0.25; 2.58)	0.96 (0.51; 1.78)
2	0.31 (0.09; 0.98)	0 (no observations)	1.23 (0.12: 2.24)	0.31 (0.07; 1.36)
3	0 (no observations)	0 (no observations)	0.00 (no observations)	0.00	(no observations)

Note: Significant (α < 0.05) associations are in bold. This analysis was adjusted for potential confounders: ABCD Study Research Cities, Seasonality, sex, age ethnicity, waist circumference, puberty status, and social determinants of health (only for the whole sample).

Table 4 illustrates that changes in adherence to 24‐h movement behavior guidelines were associated with hypertension risk in adolescents. Sustained adherence to recommended physical activity levels was associated with the lowest risk of hypertension, highlighting the long‐term protective benefits of regular moderate‐to‐vigorous activity. In contrast, inconsistent physical activity patterns were associated with a higher risk, underscoring the importance of maintaining healthy behaviors over time. For sleep, those who improved to meet the recommended duration of 9–11 h per night demonstrated substantial protective effects, while deviations from sleep guidelines were linked to increased hypertension risk. Although the associations for screen time were less pronounced, consistent adherence to the guideline of two hours or less daily appeared modestly protective compared to prolonged screen use. These findings highlight the cumulative benefits of sustained adherence to movement behavior guidelines, reinforcing the importance of promoting consistent healthy behaviors to mitigate hypertension risk during adolescence.

TABLE 4.

Hypertension cumulative incidence per 100 individuals and respective confidence interval 95% (95% CI) according to behavioral change from T0 to T1.

Exposures	Hypertension (T1; 2020/2022)
	Incidence	(95% CI)		Relative Risk	(95% CI)
SB changes (T0 → T1)
Always ≤ 2 h/d	2.6%	1.8%	3.7%	Ref
≤ 2 h/d → > 2 h/d	1.6%	0.7%	3.5%	0.56	0.26	1.22
> 2 h/d → ≤ 2 h/d	3.7%	1.8%	7.6%	0.88	0.40	1.95
Always > 2 h/d	2.7%	1.5%	4.8%	0.84	0.43	1.65
Sleep time changes (T0 → T1)
< 9 or > 11 h/night	3.0%	2.0%	4.4%	Ref
9–11 h/night → < 9 or > 11 h/night	6.8%	2.6%	16.7%	2.73	0.87	8.59
< 9 or > 11 h/night → 9–11 h/night	1.0%	0.3%	4.0%	0.32	0.07	1.48
Always 9–11 h/night	2.0%	0.3%	12.9%	0.68	0.08	5.54
PA changes (T0 → T1)
Always ≥ 60 min/d	2.5%	1.6%	3.8%	Ref
≥ 60 min/d == >< 60 min/d	2.3%	0.9%	6.0%	0.80	0.30	2.10
< 60 min/d → ≥ 60 min/d	5.5%	2.6%	11.1%	1.40	0.58	3.40
Always < 60 min/d	2.8%	1.1%	7.2%	0.78	0.29	2.09
Total 24‐h movement guideline adherence changes (T0 → T1)
Same # behaviors within guidelines	2.2%	1.5%	3.2%	Ref
Decrease 1+ behavior within guidelines	4.4%	2.8%	6.9%	2.39	0.88	6.46
Increase 1+ behavior within guidelines	2.1%	1.1%	3.9%	0.87	0.32	2.37

Note: Significant (α < 0.05) associations are in bold.

4. Discussion

This study aimed to investigate the relationship between 24‐h movement behaviors and the incidence of hypertension in adolescents, and to assess whether social determinants of health (SDOH) modified these associations. The findings indicate that adhering to more components of the 24‐h movement behavior guidelines significantly reduces the risk of hypertension in adolescents. These findings provide evidence that comprehensive lifestyle interventions must simultaneously target physical activity, screen time, and sleep to effectively reduce the incidence of hypertension in adolescents. While SDoH did not consistently modify these associations, the observed reduction in hypertension risk associated with meeting multiple movement guidelines suggests a universal benefit across different socioeconomic backgrounds [31].

The results align with previous research linking screen time to cardiovascular risk factors in youth, reinforcing the role of excessive screen use as a significant contributor to hypertension [32]. The higher incidence of hypertension among adolescents exceeding screen time recommendations supports the growing body of evidence that sedentary behaviors negatively impact blood pressure regulation. Sedentary behavior is associated with numerous negative cardiovascular outcomes. Prolonged sitting has been shown to impair vascular function, contribute to poor lipid profiles, and increase insulin resistance, all of which elevate cardiovascular risk [33]. Furthermore, regular interruptions to sedentary periods can improve systolic blood pressure and vascular function [33], highlighting the importance of breaking up sedentary time for cardiovascular health improvement in the pediatric population [34]. Conversely, the observed association between adherence to physical activity guidelines and increased hypertension incidence was unexpected [35]. This paradox may reflect differences in the type, intensity, or context of the physical activity performed, with lower‐intensity or non‐sustained efforts offering limited cardiovascular benefit. Another possibility is behavioral compensation, whereby adolescents who engage in structured physical activity may also spend more time in sedentary behaviors outside of exercise. Additionally, this finding may be influenced by reverse causality [36], as adolescents with elevated blood pressure could have initiated lifestyle modifications, including increased physical activity, prior to the T1 assessment, in response to early clinical advice. The absence of significant differences in hypertension risk based on sleep duration also diverges from some studies suggesting a relationship between sleep patterns and blood pressure, underscoring the complexity of these interactions.

Adhering combined movement behavior guidelines (physical activity, screen time, sleep) significantly reduced the risk of high blood pressure, regardless of social or economic background. Unexpectedly, the analysis did not reveal a clear pattern of effect modification by social determinants of health (SDoH), contrary to previous studies indicating that neighborhood deprivation and socioeconomic factors can exacerbate health disparities [37]. Our findings suggest that adherence to movement behavior guidelines may be associated with consistent protective effects across SDoH strata, indicating that these benefits may extend beyond social disparities [37]. Nevertheless, this interpretation warrants caution, as individual‐level characteristics such as gender, age, and socioeconomic status have consistently been shown to be associated with health behavior patterns, underscoring the need for tailored prevention and intervention efforts targeting groups at greatest risk. Alternatively, the Child Opportunity Index used in this study might not capture the full range of SDoH nuances, particularly those operating at the individual rather than neighborhood level. Further research using more granular measures of SDoH, including individual‐level socioeconomic indicators and demographic moderators, is necessary to better understand their potential impact on hypertension risk and to identify subgroups for whom targeted behavioral interventions may be most beneficial [38].

Our study has some limitations and several strengths. Limitations include attrition and potential selection bias, as reflected by the reduction in sample size from baseline to the final follow‐up. Although we used the lowest blood pressure reading rather than the average of repeated measures, repeated assessments showed no meaningful differences, suggesting that this choice is unlikely to have materially affected prevalence or incidence estimates; however, regression to the mean may still have influenced longitudinal changes in blood pressure. In addition, screen time was assessed by questionnaire because, although Fitbit can estimate total sedentary time, it cannot distinguish sedentary domains; this distinction is important because studying or reading may have different implications than recreational screen‐based behaviors such as television viewing, computer use, or video gaming. Dichotomizing continuous movement behaviors may also have reduced variability and statistical precision, and residual confounding cannot be completely excluded despite adjustment for multiple covariates. At the same time, the study's strengths enhance its rigor and relevance to adolescent health research. Sensitivity analyses confirmed that the included sample was comparable to those excluded, suggesting that missing data were likely random. High blood pressure, a major cardiovascular risk factor in adolescence [16], underscores the importance of early intervention. The diverse geographic origins of the cohort improve the generalizability of our findings. Furthermore, the cohort design allows for the temporal sequencing of risk factors, such as environmental noise and sleep patterns, with hypertension, providing temporal evidence. The use of objective sleep measures minimizes bias, while multilevel models adjusted for key confounders ensure reliable associations. Together, these strengths significantly contribute to the study's credibility and its potential to inform the prevention and management of hypertension in adolescents.

5. Perspective

Adherence to combined 24‐h movement behavior guidelines is associated with lower hypertension risk during adolescence, independent of neighborhood‐level social determinants of health, suggesting that cardiovascular benefit reflects the joint contribution of multiple behaviors rather than any single component. This finding aligns with evidence that physical activity and sedentary behavior are interrelated; high levels of both may coexist, whereas the most adverse pattern, low physical activity combined with high sedentary time, often clusters with other unhealthy behaviors, indicating that isolated promotion of physical activity may be insufficient to improve cardiovascular health in pediatric populations. Reducing screen‐based sedentary behavior may be particularly important given its potential effects on vascular function and behavioral patterns. Future studies should incorporate compositional analytical approaches to better account for the codependent nature of daily behaviors and use repeated or ambulatory blood pressure measurements to improve outcome assessment. Additional longitudinal studies in diverse and higher‐risk populations are needed to enhance generalizability. Interventions should emphasize integrated behavior change while addressing structural and socioeconomic barriers that limit adherence to healthy movement patterns.

Author Contributions

Augusto César F. De Moraes: writing – review and editing, writing – original draft, methodology, formal analysis, conceptualization. Ethan T. Hunt; Jacob Szeszulski; Gabrielli T. De Mello: review and editing, writing – original draft and final version approved. Marcus V. Nascimento‐Ferreira: writing – review and editing, writing – original draft, visualization, formal analysis. Pamela A. Madden: resources, supervision, methodology, writing – review and editing.

Funding

Augusto César Ferreira De Moraes was a Scholar of the ABCD START Program, supported by the National Institutes of Health under award number R25DA059073. Augusto César Ferreira De Moraes, John Virostko, and Marcus Vinícius Nascimento‐Ferreira received funds from the IC2 Institute of The University of Texas at Austin (#2024). Ethan Hunt received the Start‐Up Fund from The University of Texas Health Science Center at Houston, School of Public Health in Austin. Marcus Vinícius Nascimento‐Ferreira received funds from the Brazilian Government from the National Council for Scientific and Technological Development (CNPq; proc. 402 391/2021–7) by the Federal University of Tocantins (PROPESQ Universal No. 088/2022). This is a multisite, longitudinal study designed to recruit more than 10 000 children age 9–10 and follow them over 10 years into early adulthood. The ABCD Study is supported by the National Institutes of Health and additional federal partners under award numbers U01DA041048, U01DA050989, U01DA051016, U01DA041022, U01DA051018, U01DA051037, U01DA050987, U01DA041174, U01DA041106, U01DA041117, U01DA041028, U01DA041134, U01DA050988, U01DA051039, U01DA041156, U01DA041025, U01DA041120, U01DA051038, U01DA041148, U01DA041093, U01DA041089, U24DA041123, U24DA041147. A full list of supporters is available at https://abcdstudy.org/federal‐partners.html.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Sensitivity analysis of the sample comparing the variables: covariates by included vs. excluded adolescents, ABCD Study baseline (T0; 2018/2020).

Data Availability Statement

The data used in this study are available through the Adolescent Brain Cognitive Development (ABCD) Study. Interested researchers can access the data by visiting the ABCD Study website (https://abcdstudy.org/) and submitting a data access request to the National Institute of Mental Health (NIMH) Data Archive (https://nda.nih.gov/abcd). Access is subject to approval by NIMH and compliance with their data use agreements.