Dose-Response Relation between Physical Activity and Blood Pressure in Youth : Medicine & Science in Sports & Exercise

Journal Logo

BASIC SCIENCES: Epidemiology

Dose-Response Relation between Physical Activity and Blood Pressure in Youth


Author Information
Medicine & Science in Sports & Exercise 40(6):p 1007-1012, June 2008. | DOI: 10.1249/MSS.0b013e318169032d
  • Free


Cardiovascular disease is a progressive disease that starts in childhood (28-31). Elevated blood pressure (BP) during childhood and adolescence increases the risk of hypertension in adulthood (3,25), when more severe forms of cardiovascular disease also occur quite frequently. It is therefore important to prevent the development of hypertension at young ages.

Physical activity (PA) is key component of the therapeutic lifestyle changes recommended for preventing and treating prehypertension and hypertension in children and youth (21,33). The American Heart Association recommends that children and youth participate in at least 60 min of moderate-to-vigorous PA daily for cardiovascular health promotion (16). The review of PA interventions completed during the development of the Centers for Disease Control and Prevention-sponsored 2005 evidence-based PA guidelines for children and youth suggests that at least 30 min of PA at a frequency of 3 times per week and intensity of 80% of maximal heart rate is required to lower BP (24). Although it is recognized that PA has important effects on BP, the specific PA guidelines and recommendations mentioned above were developed based on the best available evidence, which in the case for BP and most other health outcomes in children and youth, is not the ideal type and amount of evidence (14).

Dose-response studies are particularly useful for determining the minimal and optimal amount of PA required for good health. To our knowledge, only two studies (11,23) have examined the dose-response relation between PA and BP in children and adolescent with only one study finding a relation (11). This study found that for every 100 MET·h increase in PA, systolic BP decreased in a linear manner by 1.15 mm Hg (11). Although a dose-response relation was seen, the PA measurement unit (MET) does not provide information that can be transformed into an understandable public health message (e.g., minutes of PA per day). The self-reported PA questionnaire used in that study was also subject to recall bias (12), and likely resulted in an overestimation of actual PA levels. Objective measures of PA, such as those obtained by accelerometers, offer the advantage of being able to measure the frequency, intensity, and duration of activity.

Although PA may play an important role in the prevention of hypertension, the dose-response relation between PA and BP in the pediatric age group is unclear. Therefore, the purpose of the study was to examine the dose-response relation between objectively measured PA, BP, and hypertension in children and youth.


The study sample consisted of participants from the 2003-2004 National Health and Nutrition Examination Surveys (NHANES). NHANES is a representative cross-sectional survey of the U.S. population (7). The secondary analysis presented here was approved by the Queen's University Health Sciences Research Ethics Board. NHANES participants were identified using a complex, stratified, multistage, probability sampling design (7). The present study was limited to those aged 8 to 17 yr who completed both the home and mobile exam center (MEC) components of the survey required for the analysis. Informed consent was obtained from all the parents and participants, and the protocol was approved by the National Center for Health Statistics.

Physical activity.

PA was measured using Actigraph 7124 activity monitors (Actigraph, Ft. Walton Beach, FL, USA). Activity monitors can be used to objectively measure the duration and intensity of PA under free-living conditions (27,32). Monitors were provided to subjects during the MEC exam, were programmed to start recording at midnight of the day following the exam, and recorded activity for the following 7 d. Participants were instructed to wear the monitor during waking hours except when it would get wet. Monitors were attached to elasticized fabric belts and worn on the right hip.

The Actigraph is a uniaxial monitor that measures the intensity of movement averaged over 1-min sampling epochs. Data from each monitor were downloaded by the Centers for Disease Control and Prevention and checked for spurious data (e.g., biologically implausible values), which were subsequently removed. We did not remove additional spurious data; however, further data reduction was completed by the authors before analyses were performed. Specifically, the current study only included subjects with at least 4 d of complete monitoring, including at least one weekend day. Four days of monitoring has a test-retest reliability of 0.7 (27). A day was considered complete if it included at least 10 h of wear time (27). Wear time was determined by filtering activity monitor data and deleting sections of more than 20 min of zero counts indicating nonwear time (20).

For each complete day of monitoring, thresholds of 2000 (equivalent to walking ∼4.0 km·h−1) and 3000 (equivalent to walking ∼5.6 km·h−1) counts per minute were used to denote those minutes where the participants were engaged in PA of at least a low intensity (total PA) and moderate-to-vigorous intensity PA, respectively (26). The measurement unit of counts per minute is generated based on the magnitude and frequency of movement. The minutes of total and moderate-to-vigorous intensity PA were averaged across all complete days of monitoring to create the final PA variables used for analysis.

There is considerable discrepancy in the literature concerning the accelerometry threshold that should be used to classify PA intensity, and the proposed thresholds often vary by age and gender (13). The 2000 and 3000 counts per minute thresholds used in this study were selected based on available evidence and best judgement of the authors. Although the thresholds used here were developed from a study conducted entirely in adolescent females, similar thresholds have been used for males and females (1,22) and in youth ranging in age from 6 to 16 yr (22). Thresholds were based on METs achieved during regular child/youth activities (e.g., threshold for moderate-to-vigorous PA based on counts per minute for activity of 3 METs).

Blood pressure.

Measurements were completed at the MEC. Subjects were seated and rested quietly for 5 min before 3 BP measures were obtained on the right arm using a mercury sphygmomanometer. Systolic BP was recorded at the first Korotkoff sound, and diastolic BP was the point of the last audible Korotkoff sound. Mean BP values were used for all analyses. BP z-scores were calculated based on reference values for age, height, and sex (21). Hypertension was defined as having systolic or diastolic levels above the 90th percentile for age, sex, and height (considered prehypertensive or hypertensive) (21).

Body mass index.

Height and weight were measured during the MEC to the nearest 0.1 cm and 0.1 kg, respectively, and used to calculate the body mass index (kg·m−2). Subjects were classified as "at-risk of overweight" and "overweight" according to the Centers for Disease Control and Prevention age- and gender-specific cut-points for children and youth of 85th-94th percentiles and ≥95th percentile, respectively (19).

Statistical analysis.

All analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC, USA) and took into account the sample weights of the NHANES survey. Linear regression was used to examine the relationship between total PA and moderate-to-vigorous PA with systolic BP and diastolic BP. Logistic regression was used to examine the relationship between total PA and moderate-to-vigorous PA with hypertension. In all statistical models, PA was included as a continuous measure in minutes per day. Fractional polynomial models for PA were run to obtain the models that best fit the dose-response relation between PA, BP, and hypertension. The 36 fractional polynomial regression models recommended by Bagnardi et al. (2) for examining dose-response relations were used for the current analysis. This form of analysis is concerned with finding the dose-response curve with the best model fit rather than determining the significance of the linear relation between the exposure (PA) and outcome (mean BP value or hypertension). For the linear regression analyses, the model with the highest R2 and smallest error was considered to provide the best fit. For the logistic regression analyses, the model with the lowest Akaike information criterion was considered to provide the best fit. Sex, age, and race-ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, and other) were included as covariates in all statistical models. Bootstrapping was used to estimate the 95% confidence intervals (CI) for the regression analyses.


There were 2413 participants aged 8 to 17 yr in the 2003/2004 NHANES. Acceptable PA measurements were not available for 1197 participants, and acceptable BP measurements were not available for 166 participants. The current analyses included the 1170 participants (607 males and 563 females) with complete PA and BP measurements. There were no statistical differences in the body mass index or gender distribution in the participants who were included or excluded from the analyses. However, the participants included in the analyses were younger and contained a higher proportion of Hispanics.

Basic demographic characteristics of the subjects included in the analyses are presented in Table 1. Approximately one quarter of the study participants were non-Hispanic white, and 38.4% were classified as either at-risk of overweight or overweight. Subjects participated in 53.0 ± 32.1 min·d−1 (mean ± SD) and 25.8 ± 20.6 min·d−1 of total PA and moderate-to-vigorous intensity PA, respectively. Mean systolic and diastolic BP values in males were 108.5 ± 10.4 and 57.0 ± 11.3 mm Hg, respectively. The corresponding values in females were 104.1 ± 8.9 and 58.5 ± 10.6 mm Hg. Approximately 14% of the participants were hypertensive (18.1% in males, 9.4% in females). As shown in Table 2, a higher proportion of females participated in less than 30 min·d−1 of total PA compared to males (36.9% vs 14.8%). Most males (88.3%) and females (98.6%) participated in less than 60 min·d−1 of moderate-to-vigorous intensity PA.

Descriptive characteristics of subjects (n = 1170).
Distribution of study participants by minutes of PA.

Relations between PA and BP were modeled for PA values ranging from 0 min·d−1 to 2 standard deviations above the population mean. An inverse dose-response relationship was observed between total (Fig. 1A) and moderate-to-vigorous PA (Fig. 1B) with systolic BP. Total PA had no effect on systolic BP until approximately 40 min·d−1, after which systolic BP decreased as total PA increased, with no plateau in the effect. Although an inverse relation was observed between total PA, moderate-to-vigorous PA, and systolic BP, the slope of the curve was modest, indicating a minimal influence of PA on systolic BP measures. For example, the systolic BP z-score decreased by 0.09 units between 60 and 90 min·d−1 of total PA, which is equivalent to 0.93 mm Hg. A similar modest dose-response relation was seen between total PA (Fig. 1C) and moderate-to-vigorous PA (Fig. 1D) with diastolic BP. A plateau in the effect of PA on diastolic BP was seen at approximately 75 and 40 min·d−1 for total and moderate-to-vigorous PA, respectively.

Relation between total PA and systolic BP z-scores (A), moderate-to-vigorous PA and systolic BP z-scores (B), total PA and diastolic BP z-scores (C), and moderate-to-vigorous PA and diastolic BP z-scores (D). Solid lines represent the predicted relationship between PA and BP z-score for a 12-yr-old non-Hispanic white male, whereas the dashed lines represent the 95% CI.

The likelihood of hypertension decreased in a curvilinear manner with increasing minutes of total PA (Fig. 2A) and moderate-to-vigorous PA (Fig. 2B). At 30 min·d−1 of moderate-to-vigorous PA, the odd ratio (95% CI) for hypertension was 0.50 (0.28-0.64) in comparison to no PA. At 60 min·d−1 of moderate-to-vigorous PA, the odds ratio (95% CI) for hypertension was 0.38 (0.17-0.52). At 30, 60, 90, and 120 min·d−1 of total PA, the odds ratios (95% CI) for hypertension relative to no activity were 0.80 (0.36-0.97), 0.64 (0.32-0.69), 0.51 (0.22-0.58), and 0.40 (0.14-0.55), respectively.

Odds ratios for hypertension according to total PA (A) and moderate-to-vigorous PA (B). 0 min represents the referent value (odds ratio = 1.0) for PA. Solid lines represent the predicted relationships between PA and hypertension for a 12-yr-old non-Hispanic white male, whereas dashed lines represent the 95% CI.


This study focused on developing dose-response curves that best represent the relation between PA, BP, and hypertension in children and youth. A modest dose-response relation was observed between total and moderate-to-vigorous PA with mean systolic and diastolic BP values. PA did, however, have a strong effect on BP when predicting high-risk values in the hypertensive range. At 60 min·d−1 of moderate-to-vigorous PA, the PA volume corresponding to the public health recommendation in many countries (4,9,24), the likelihood of hypertension was approximately one third of that observed for no moderate-to-vigorous PA.

The modest nature of the dose-response relation between PA and BP was not unexpected considering the discrepancy in the literature surrounding the effects of PA on BP in children and youth. In a previous study of adolescents with dyslipidemia, a linear dose-response relation was found between self-report PA and systolic BP; PA was not related to diastolic BP (11). That study did not consider whether the relation between PA and BP was nonlinear in nature. Of the two studies in children and youth that examined the relation between PA and BP using an objective measure of PA (e.g., accelerometers) (1,5), only one (1) reported a significant correlation. Neither of these studies used an analysis strategy that adequately explored the dose-response relation between PA and BP (1,5).

A recent meta-analysis of randomized controlled trials reported a similar inconsistency in the effects of PA interventions on lowering BP in children and adolescents (18). The authors suggested that the null effect of PA on BP may have been in part related to the inclusion of normotensive subjects in the intervention trials (18). Of the 16 groups included in the meta-analysis, only two were limited to hypertensive subjects (18). A similar observation was made by the authors of an adult based meta-analysis on PA and BP. The adult meta-analysis found PA to have a threefold greater effect on lowering BP in hypertensive compared to normotensive individuals (10). The findings from these meta-analyses suggest that PA has a greater effect on high-risk BP values. Our results support this contention. The slopes of the dose-response curves between PA and mean BP values were modest, whereas the slope of the dose-response curves between PA and hypertension were quite steep. That PA has a greater effect on high-risk BP values is consistent with the effects of PA on other cardiovascular risk factors including blood lipids/lipoproteins (17) and insulin resistance (15). Perhaps it is only intuitive that PA does not positively impact cardiovascular risk factor values that already fall within the normal healthy range.

In studies of adults, it has been reported that the intensity of PA does not significantly contribute to whether or not a PA intervention is able to successfully reduce BP levels (10). Conversely, the 2005 evidence-based PA guidelines for children and youth suggest that an intensity corresponding to 80% of the maximum heart rate is required in a PA intervention to reduce BP levels in hypertensive children and adolescents (24). In the current study, the nature of the dose-response curve between PA and hypertension was similar for both total and moderate-to-vigorous intensity PA. However, considerably fewer minutes of moderate-to-vigorous intensity PA were required to lower hypertension risk. For instance, the likelihood of hypertension was reduced by approximately 50% at 30 min·d−1 of moderate-to-vigorous PA versus 90 min·d−1 of total PA. Thus, in the development of strategies aimed at preventing and treating hypertension in children and youth, both time and intensity of PA are important factors to consider.

Our results provide support for the 2005 evidence-based PA guidelines (24), which recommend that school-aged children and youth accumulate at least 60 min of moderate-to-vigorous intensity PA daily. Specifically, we observed a plateau in the relationship between moderate-to-vigorous PA and hypertension risk at approximately 60 min·d−1. The absence of a plateau in the relationship between total PA and hypertension risk also corroborates the Canadian PA guidelines for children and youth, which recommends that all children and youth, regardless of their current PA level, increase overall PA by 90 min·d−1 ( Given the volume of PA recommended in the evidence-based PA guidelines, and the dose-response relation between PA and hypertension observed here, it is very concerning that only 11.7% of the male and 1.4% of the female participants in this study were moderately or vigorously active for at least 60 min·d−1 on average. These findings emphasize the need for the implementation of public health interventions and policies that will substantially improve the PA levels of most children and adolescents. In the development of strategies to increase PA participation, it is important to consider the roles of the family, school, and community environments have on PA participation in children and youth (6,8).

The strengths of this study include the use of a large and representative study sample, the use of an objective PA measure, and the statistical modeling approach that allowed us to accurately characterize the dose-response relation between PA, BP, and hypertension. The primary study limitation was the cross-sectional and observational nature, which prevents us from making causal inferences about the relation between PA and BP. Although objective in nature, PA accelerometers are not perfect measures of PA, and their main limitation is the inability to capture certain types of PA such as swimming and bicycling (12). Another issue with accelerometers is the discrepancy in thresholds used by different investigators to denote those minutes where the participants were physically active, as discussed in more detail in the Methods. Had different threshold been employed in the present study, the slopes of the dose-response curves may have changed. In addition, we considered all minutes of PA, not just those minutes that occurred within bouts of at least a few minutes in length. Had we limited the physically active minutes to those that occurred in bouts, the total minutes of PA in the study participants would have been reduced substantially, and the reduction in BP values or hypertension risk for a given increase in PA would have been larger. Finally, the slight differences in characteristics between the participants used in our analyses and the overall NHANES sample limit the generalizability of the results.

In summary, PA is a better predictor of hypertension than mean BP levels. Meeting the guidelines of 60 min·d−1 of moderate-to-vigorous intensity PA greatly reduces the likelihood of hypertension in the pediatric population. Unfortunately, few children and youth achieve this PA level.

This study was supported by an operating grant provided by the Canadian Institutes of Health Research (MOP 84478). Ian Janssen is supported by a New Investigator Award from the Canadian Institutes of Health Research and an Early Researcher Award from the Ontario Ministry of Research and Innovation.


1. Andersen LB, Harro M, Sardinha LB, et al. Physical activity and clustered cardiovascular risk in children: a cross-sectional study (The European Youth Heart Study). Lancet. 2006;368(9532):299-304.
2. Bagnardi V, Zambon A, Quatto P, Corrao G. Flexible meta-regression functions for modeling aggregate dose-response data, with an application to alcohol and mortality. Am J Epidemiol. 2004;159(11):1077-86.
3. Bao W, Threefoot SA, Srinivasan SR, Berenson GS. Essential hypertension predicted by tracking of elevated blood pressure from childhood to adulthood: the Bogalusa Heart Study. Am J Hypertens. 1995;8(7):657-65.
4. Biddle SJ, Sallis JF, Cavin N. Policy framework for young people and health-enhancing physical activity. In: Young and Active? Young People and Health-Enhancing Physical Activity: Evidence and Implications, SJ Biddle, JF Sallis, and N Cavin. London, U.K.: Health Education Authority; 1999, pp. 3-16.
5. Brage S, Wedderkopp N, Ekelund U, et al. Objectively measured physical activity correlates with indices of insulin resistance in Danish children. The European Youth Heart Study (EYHS). Int J Obes Relat Metab Disord. 2004;28(11):1503-8.
6. Canadian Paediatric Society. Healthy active living for children and youth. Paediatrics & Child Health. 2002;7(5):339-45.
7. Centers for Disease Control and Prevention (CDC). National Health and Nutrition Examination Survey Data, Hyattsville, MD: Department of Health and Human Services, Centers for Disease Control and Prevention; 2003-2004.
8. Council on Sports Medicine and Fitness and Council on School Health. Active healthy living: prevention of childhood obesity through increased physical activity. Pediatrics. 2006;117(5):1834-42.
9. Department of Health and Ageing, Australian Government. National physical activity guidelines for Australians. Commonwealth of Australia: Canberra, Australia, 1999. Available from: (accessed on January 14, 2008).
10. Fagard RH. Exercise characteristics and the blood pressure response to dynamic physical training. Med Sci Sports Exerc. 2001;33(6 Suppl):S484-92. discussion S493-4.
11. Gidding SS, Barton BA, Dorgan JA, et al. Higher self-reported physical activity is associated with lower systolic blood pressure: the Dietary Intervention Study in Childhood (DISC). Pediatrics. 2006;118(6):2388-93.
12. Goran MI. Measurement issues related to studies of childhood obesity: assessment of body composition, body fat distribution, physical activity, and food intake. Pediatrics. 1998;101(3 Pt 2):505-18.
13. Guinhouya CB, Hubert H, Soubrier S, Vilhelm C, Lemdani M, Durocher A. Moderate-to-vigorous physical activity among children: discrepancies in accelerometry-based cut-off points. Obesity (Silver Spring). 2006;14(5):774-7.
14. Janssen I. Physical activity guidelines for children and youth. Appl Physiol Nutr Metab. 2007;32:S109-21.
15. Kang HS, Gutin B, Barbeau P, et al. Physical training improves insulin resistance syndrome markers in obese adolescents. Med Sci Sports Exerc. 2002;34(12):1920-7.
16. Kavey RE, Daniels SR, Lauer RM, Atkins DL, Hayman LL, Taubert K. American Heart Association guidelines for primary prevention of atherosclerotic cardiovascular disease beginning in childhood. J Pediatr. 2003;142(4):368-72.
17. Kelley GA, Kelley KS. Aerobic exercise and lipids and lipoproteins in children and adolescents: a meta-analysis ofrandomized controlled trials. Atherosclerosis. 2007;191(2):447-53.
18. Kelley GA, Kelley KS, Tran ZV. The effects of exercise on resting blood pressure in children and adolescents: a meta-analysis of randomized controlled trials. Prev Cardiol. 2003;6(1):8-16.
19. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC Growth Charts for the United States: methods and development. Vital Health Stat. 2002;11(246):1-190.
20. Masse LC, Fuemmeler BF, Anderson CB, et al. Accelerometer data reduction: a comparison of four reduction algorithms on select outcome variables. Med Sci Sports Exerc. 2005;37(11 Suppl):S544-54.
21. National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics. 2004;114(2Suppl 4th Report):555-76.
22. Puyau MR, Adolph AL, Vohra FA, Butte NF. Validation and calibration of physical activity monitors in children. Obes Res. 2002;10(3):150-7.
23. Raitakari OT, Taimela S, Porkka KV, et al. Associations between physical activity and risk factors for coronary heart disease: the Cardiovascular Risk in Young Finns Study. Med Sci Sports Exerc. 1997;29(8):1055-61.
24. Strong WB, Malina RM, Blimkie CJ, et al. Evidence based physical activity for school-age youth. J Pediatr. 2005;146(6):732-7.
25. Sun SS, Grave GD, Siervogel RM, Pickoff AA, Arslanian SS, and Daniels SR. Systolic blood pressure in childhood predicts hypertension and metabolic syndrome later in life. Pediatrics. 2007;119(2):237-46.
26. Treuth MS, Schmitz K, Catellier DJ, et al. Defining accelerometer thresholds for activity intensities in adolescent girls. Med Sci Sports Exerc. 2004;36(7):1259-66.
27. Trost SG, Pate RR, Freedson PS, Sallis JF, Taylor WC. Using objective physical activity measures with youth: how many days of monitoring are needed? Med Sci Sports Exerc. 2000;32(2):426-31.
28. Twisk JW. Physical activity guidelines for children and adolescents: a critical review. Sports Med. 2001;31(8):617-27.
29. Twisk JW, Kemper HC, van Mechelen W. The relationship between physical fitness and physical activity during adolescence and cardiovascular disease risk factors at adult age. The Amsterdam Growth and Health Longitudinal Study. Int J Sports Med. 2002;23 Suppl 1:S8-14.
30. Twisk JW, Kemper HC, van Mechelen W. Tracking of activity and fitness and the relationship with cardiovascular disease risk factors. Med Sci Sports Exerc. 2000;32(8):1455-61.
31. Twisk JW, Van Mechelen W, Kemper HC, Post GB. The relation between "long-term exposure" to lifestyle during youth and young adulthood and risk factors for cardiovascular disease at adult age. J Adolesc Health. 1997;20(4):309-19.
32. Van Coevering P, Harnack L, Schmitz K, Fulton JE, Galuska DA, Gao S. Feasibility of using accelerometers to measure physical activity in young adolescents. Med Sci Sports Exerc. 2005;37(5):867-71.
33. Williams CL, Hayman LL, Daniels SR, et al. Cardiovascular health in childhood: A statement for health professionals from the Committee on Atherosclerosis, Hypertension, and Obesity in the Young (AHOY) of the Council on Cardiovascular Disease in the Young, American Heart Association. Circulation. 2002;106(1):143-60.


©2008The American College of Sports Medicine