This study was designed as a prospective, quasi-experimental, longitudinal intervention study. Participants were not stratified according to their baseline physiological characteristics prior to the intervention. Instead, individuals who voluntarily enrolled in the body conditioning program were assessed before and after completion of the educational course. Participants were classified into BP categories based on resting systolic BP (SBP) and diastolic BP (DBP) values obtained prior to the intervention, according to the 2017 American College of Cardiology/American Heart Association criteria, and physiological variables were subsequently analyzed according to these postintervention BP classifications. Although a nonexercise control group was not included, the within-subject repeated-measures design inherently accounts for stable individual characteristics, and baseline homogeneity across BP categories further minimizes the likelihood that the observed interaction effects were driven by pre-existing group differences. While the absence of a nonexercise control condition limits the strength of causal inference, the consistent direction of BP changes within the hypertensive categories supports a training-related effect rather than random temporal variation. The research protocol was reviewed and approved by the institutional review board of Hanseo University in Seosan, Korea (HS26-0209-01), and all participants provided informed consent prior to study enrollment in accordance with the Declaration of Helsinki.

Participants were university students aged 19–26 years who voluntarily enrolled in a standardized 15-week supervised exercise class. All participants completed the same exercise protocol under identical training conditions. Enrollment was based on participation in the academic exercise class rather than clinical recruitment, and participants generally considered themselves healthy at the time of study entry. A total of 95 students were initially enrolled in the program. Of these, 10 were excluded from the final analysis due to baseline SBP/DBP values not meeting the predefined classification criteria (n=8) and measurement errors (n=2). All remaining participants (n=85) completed both baseline and postintervention assessments, with no additional attrition during the intervention period. They were informed of the purpose and direction of the class and the outcomes measured in this study before and after class during the pre-class orientation. The appropriate sample size for this study design was 72, assuming a medium effect size (f=0.25), α=0.05, power (1−β)=0.80, 4 groups, and 2 repeated measurements using G*Power (v.3.1.9.4; Heinrich Heine University, Germany) (Faul et al., 2007). Eligibility criteria included no history of treatment or medication use affecting BP or body composition. Only students who had not engaged in a structured exercise program for at least 6 months were included. None of the participants were aware of their hemodynamic status prior to enrollment; therefore, BP, heart rate (HR), and other hemodynamic parameters were assessed during a pre-participation physical examination. Individuals with any diagnosed psychological disorder or major organ dysfunction were excluded. The longitudinal structure, with repeated measurements obtained before and after the 15-week program, enabled the assessment of within-subject changes over time and between-group differences in adaptive responses. At baseline (week 2, prior to the initiation of structured training), participants were categorized into groups based on standard SBP/DBP criteria (normal BP, <120/80 mmHg; elevated BP, 120–129/<80 mmHg; stage 1 hypertension, 130–139/80–89 mmHg; stage 2 hypertension, ≥140/≥90 mmHg) (Whelton et al., 2018).

The theory-and-practice educational class was conducted over 15 weeks, with each session lasting 2 hr and consisting of 40 min of theoretical instruction followed by 80 min of practical training. All sessions began with standardized stretching and warm-up routines, which were consistently applied throughout the study. In week 1, participants attended an orientation session covering the basic principles of exercise. A demographic survey was conducted, followed by supervised flexibility exercises (stretching from the ankle to the cervical region) and aerobic warm-up activities ranging from brisk walking to light jogging. In week 2, baseline assessments were performed prior to the intervention. Measurements included body composition, BP, HR, respiratory rate, body circumferences (upper arm, forearm, abdomen, hip, thigh, and calf), grip strength, and maximum oxygen uptake (VO_2max) estimated using the Harvard step test. During weeks 3–4, theoretical lectures addressed diet, weight control, energy metabolism, and the neuroendocrine system. Practical sessions included calisthenics (push-ups, squats, pull-ups, and lunges) and core stabilization exercises (front, back, and side planks). In week 5, following instruction on the musculoskeletal system, low-intensity machine-based resistance exercises were introduced in addition to core training. In week 6, principles of exercise training were taught, and participants performed calisthenics combined with machine-based resistance exercises at mild to moderate intensity. One-repetition maximum testing was conducted to determine appropriate training loads for advanced resistance training. During weeks 7–8, exercise intensity was progressively increased according to the principle of progressive overload, with greater repetitions and sets in machine-based exercises progressing from moderate to vigorous intensity. From weeks 9–12, a split-training routine was implemented targeting specific muscle groups: abdominal (week 9), back (week 10), shoulder (week 11), and chest (week 12). Training included machine-based exercises and free-weight exercises using dumbbells and barbells, with intensity gradually progressing from moderate to vigorous levels. In weeks 13–14, lower- and upper-limb resistance training sessions were conducted using free weights at mild to moderate intensity. In week 15, postintervention assessments identical to those conducted at baseline (week 2) were performed. Participants also completed an individualized self-training evaluation with peer assistance.

Demographic variables comprised age and sex, along with medical history, family history, smoking status, and levels of physical activity. Daily sleep duration was determined from self-reported bedtime and wake-up time, including the point at which participants resumed their usual morning activities, and was operationally defined as total sleep time. Habitual physical activity was measured with the International Physical Activity Questionnaire (IPAQ)–Short Form, a validated instrument for assessing adult physical activity (Chun, 2012). Overall physical activity was expressed as metabolic equivalent hours per week based on established scoring guidelines.

BP and HR were measured using an automated oscillometric sphygmomanometer (BioZ, Cardio Dynamics, USA). Prior to measurement, participants rested for approximately 3 min in a seated position. All measurements were conducted in a quiet, temperature-controlled laboratory environment (22°C–24°C) to minimize external influences. Participants were instructed to refrain from caffeine, alcohol, smoking, and vigorous physical activity for at least 12 hr prior to testing. Upon arrival, participants rested in a seated position for a minimum of 10 min before measurement. During this period, they were instructed to remain relaxed, avoid talking, and maintain normal breathing. BP was measured with the participant seated comfortably in a chair with back support, feet flat on the floor, and legs uncrossed. The arm was supported at heart level on a table. An appropriately sized cuff (selected based on mid-upper arm circumference) was placed snugly on the nondominant arm, with the lower edge of the cuff positioned approximately 2–3 cm above the antecubital fossa. Three consecutive measurements were obtained at 1- to 2-min intervals. If the difference between SBP and DBP readings exceeded 5 mmHg, an additional measurement was performed (Landgraf et al., 2010). The average of the two closest readings was used for statistical analysis. BP classification was stratified into four levels according to resting measurements as follows: normal BP (NBP, systolic <120 mmHg and diastolic <80 mmHg), elevated BP (EBP, systolic 120–129 mmHg and diastolic <80 mmHg), stage 1 hypertension (S1HTN, systolic 130–139 mmHg or diastolic 80–89 mmHg), and stage 2 hypertension (S2HTN, systolic ≥140 mmHg or diastolic ≥90 mmHg) (Whelton et al., 2018). Meanwhile, HR was automatically recorded simultaneously by the BP monitor during each measurement. The mean value of the valid BP measurements was used as the resting HR for analysis. All measurements were conducted at the same time of day before and after the intervention to control for diurnal variation.

Respiratory rate was measured while the participant remained seated comfortably with the back supported, feet flat on the floor, and hands resting on the thighs. Participants were instructed to breathe normally and to refrain from talking during the assessment. To minimize voluntary control of breathing, participants were not informed that their respiration was being specifically monitored at the time of measurement. Respiratory movements were assessed by visually observing thoracic excursions. One complete respiratory cycle was defined as one inhalation and one exhalation. Respiratory rate was counted for 60 sec using a stopwatch. If breathing appeared irregular, measurement was repeated after an additional rest period. Two measurements were obtained, and the average value was used for statistical analysis. All measurements were performed at approximately the same time of day before and after the intervention to reduce the influence of circadian variation.

Body composition was assessed using a bioelectrical impedance analysis device, the InBody 320 (Biospace Co., Korea), which provides measurements of body weight and skeletal muscle mass. Standing height was measured with a BMS330 stadiometer (Biospace Co.). The bioelectrical impedance analysis device evaluates segmental impedance by detecting voltage differences across both the upper and lower body. Prior to assessment, participants were instructed to remove all metal accessories and any items that could interfere with electrical conductivity. Each participant stood barefoot on the measurement platform, grasped the hand electrodes, and remained still for approximately 3 min while the assessment was conducted. To minimize measurement error, participants were asked to refrain from food intake for at least 4 hr, alcohol consumption for 48 hr, and any form of exercise for 10 hr prior to testing. Additionally, they were instructed to empty their bladder approximately 30 min before the assessment. In this study, the body composition variables assessed were body weight, muscle mass, fat mass, body mass index (BMI), fat percentage, waist-to-hip ratio (WHR), and basal metabolic rate (BMR).

All circumferences were measured using a nonelastic anthropometric tape to the nearest 0.1 cm. Measurements were taken on the dominant side of the body with participants in a standardized standing position of the upper arm, forearm, abdomen, hip, thigh, calf (Kalkışım et al., 2022). Each site was measured twice, and the average value was recorded. If the two measurements differed by more than 0.5 cm, a third measurement was obtained.

Handgrip strength, an indicator of muscular strength, was measured for both the left and right hands using a dynamometer (T.K.K 5401, Tachometer, Takei, Japan). Participants stood with their arms slightly abducted and exerted maximal force for 3–5 sec. Two trials were conducted for each hand, and the higher value was recorded (Lee et al., 2022). Meanwhile, VO_2max was estimated using the HR ratio method described by Uth et al. (2004), according to the formula: VO_2max=15.3×(HR_max/HR_rest), expressed in mL/kg/min. HR_rest was measured under standardized resting conditions prescribed above, and HR_max was predicted using age-based equations: HR_max=208–(0.7×age). All tests were conducted at approximately the same time of day before and after the intervention to control for diurnal variation.

All statistical analyses were performed using IBM SPSS Statistics ver. 23.0 (IBM Co., USA). Graphical representations were generated using GraphPad Prism ver. 11, (GraphPad Software Inc., USA). Data are presented as mean±standard deviation. Normality of demographic, anthropometric, and clinical variables was assessed using the Kolmogorov–Smirnov test. Before conducting the primary analyses, baseline values were examined using the Kruskal–Wallis test to identify any initial group differences. Variable comparisons were performed using a repeated-measures (4×2) analysis of covariance (ANCOVA) with covariates, muscle mass and BMR. Because random allocation was not performed and no separate nonexercise control group was included, the study is classified as quasi-experimental. However, repeated measurements were conducted before and after the intervention, allowing within-subject comparisons over time and between-group comparisons of differential adaptations. This longitudinal repeated-measures structure enabled the evaluation of time effects, group effects, and time-by-group interactions while statistically adjusting for relevant baseline covariates. Effect sizes were calculated using Cohen d (η²), defined as the mean difference between groups divided by the pooled standard deviation (Cohen, 1992). For a more detailed evaluation, changes across time were expressed as delta (Δ) percentages. SBP and DBP were predefined as the primary outcomes of this study. Other physiological variables were considered secondary or exploratory outcomes. Bonferroni-adjusted post hoc comparisons were applied for significant interaction effects. Statistical significance was established at P≤0.05.

The final sample of 85 participants exceeded this requirement, indicating adequate statistical power to detect time-by-group interaction effects. There were no meaningful differences in age across groups (overall mean 22.14±1.96 years). As shown in Table 1, there were no significant differences among groups in age, sex distribution, disease status, family history, smoking status, physical activity level, IPAQ score, sleep duration, fat mass, BMI, body fat percentage, or WHR (all P>0.05). These findings indicate baseline homogeneity across the four BP categories (NBP, EBP, S1HTN, and S2HTN). However, significant between-group differences were observed for muscle mass and BMR, with moderate effect sizes.

Pre- and postintervention changes in hemodynamic variables are presented in Table 2. After adjustment for baseline muscle mass and BMR, repeated-measures ANCOVA revealed significant time× group interactions for both SBP and DBP. For SBP, significant main effects of time and group were observed, along with a robust interaction effect (P=0.001, η²=0.684), indicating that training responses differed according to baseline BP status. As illustrated in Fig. 1A, reductions were most pronounced in the S2HTN group, followed by S1HTN, whereas normotensive groups showed minimal change. In within-group analyses, SBP decreased by 7.2 mmHg (95% confidence interval [CI], −11.2 to −3.2) in the S1HTN group and by 17.2 mmHg (95% CI, −21.1 to −13.3) in the S2HTN group, whereas changes in the NBP and EBP groups were not statistically significant. Significant time×muscle mass and time×BMR interactions suggest that baseline body composition partially moderated SBP responses. For DBP, although the main effect of time was not significant, a significant group effect and time×group interaction were detected (Fig. 1B), with the greatest reductions observed in S2HTN. Within-group analysis indicated that DBP decreased by 7.3 mmHg (95% CI, −12.9 to −1.7) in the S2HTN group, whereas changes in the NBP, EBP, and S1HTN groups were not statistically significant. No significant moderating effects of muscle mass or BMR were identified. HR showed no significant main effects but demonstrated a significant time×group interaction (Fig. 1C), reflecting heterogeneous adaptations across BP categories. In contrast, respiratory rate remained relatively stable, with no significant main or interaction effects (Fig. 1D).

Changes in body composition variables are presented in Table 3. Repeated-measures ANCOVA adjusted for baseline muscle mass and BMR revealed no significant main effects of time, no significant group effects, and no significant time×group interactions for body weight, fat mass, BMI, body fat percentage, or WHR (all P>0.05). Although small directional changes were observed—such as modest reductions in fat mass in the EBP and S2HTN groups and slight increases in body weight in some normotensive participants—these changes were minimal in magnitude and did not differ significantly across BP categories. Percentage changes across variables generally remained within a narrow range, indicating relative stability of structural body composition over the 15-week intervention period.

As shown in Table 4, no significant main effects of time or group, nor significant time×group interactions, were detected for abdomen, hip, upper arm, forearm, thigh, or calf circumference (all P>0.05). Although minor directional changes were observed—such as slight increases in upper and forearm girth across several groups and small reductions in abdominal circumference in S2HTN —these changes were modest in magnitude and did not significantly differ between BP categories after covariate adjustment. Percentage changes were generally small and inconsistent across groups, indicating the absence of systematic structural remodeling over the 15-week intervention period.

Physical fitness outcomes are presented in Table 5. Repeated-measures ANCOVA adjusted for baseline skeletal muscle mass and BMR revealed no significant main effects of time or group, and no significant time×group interactions for left or right grip strength (all P>0.05). Although mean grip strength values demonstrated modest increases across most groups, with percentage improvements generally small and comparable between categories, these changes did not differ significantly according to baseline BP status. Furthermore, no significant time×muscle mass or time×BMR interactions were identified, indicating that baseline body composition did not meaningfully moderate strength adaptations. In contrast, VO_2max demonstrated a significant group effect and a significant time×group interaction, indicating heterogeneous responses across BP classifications. As shown in Table 5, improvements were primarily observed in the NBP and S2HTN groups, whereas the S1HTN group exhibited a slight decline and the EBP group showed minimal change. Despite these differential patterns, no significant moderating effects of baseline muscle mass or BMR were detected.