High blood pressure and sedentary behavior in adolescents are associated even after controlling for confounding factors

Abstract

Objective. The aim of this study was to determine whether high blood pressure (HBP) is associated with sedentary behavior in young people even after controlling for potential confounders (gender, age, socioeconomic level, tobacco, alcohol, obesity and physical activity). Methods. In this epidemiological study, 1231 adolescents were evaluated. Blood pressure was measured with an oscillometric device and waist circumference with an inextensible tape. Sedentary behavior (watching television, computer use and playing video games) and physical activity were assessed by a questionnaire. We used mean and standard deviation to describe the statistical analysis, and the association between HBP and sedentary behavior was assessed by the chi-squared test. Binary logistic regression was used to observe the magnitude of association and cluster analyses (sedentary behavior and abdominal obesity; sedentary behavior and physical inactivity). Results. HBP was associated with sedentary behaviors [odds ratio (OR) = 2.21, 95% confidence interval (CI) = 1.41–3.96], even after controlling for various confounders (OR = 1.68, CI = 1.03–2.75). In cluster analysis the combination of sedentary behavior and elevated abdominal obesity contributed significantly to an increased likelihood of having HBP (OR = 13.51, CI 7.21–23.97). Conclusions. Sedentary behavior was associated with HBP, and excess fat in the abdominal region contributed to the modulation of this association.

Key Words::

• Abdominal obesity
• adolescents
• high blood pressure
• physical activity
• sedentary behaviors

Introduction

Arterial hypertension is commonly diagnosed in the general, mainly adult, population. (1,2). However, more recently, studies have pointed to high rates of high blood pressure (HBP) in pediatric populations worldwide (3–6). There is substantial scientific literature identifying increased body fat as crucial to the development of HBP in the pediatric population (7).

Sedentary behavior has been studied as an important determinant of cardiovascular health in pediatric populations. This type of behavior can be characterized as the high number of hours that the subject remains engaged in low-energy expenditure activities such as watching television, computer use or playing video games during the day, which could contribute to the increased levels of blood pressure. Martínez-Gómez et al. (8), in a study of Spanish adolescents, found that young people with sedentary behavior had higher blood pressure values. However, this direct relationship between HBP and high sedentary behavior seems to be affected by other factors, such as obesity (9).

In addition, physical activity may be a factor that can interfere with this relationship, since regular physical activity may be associated with lower blood pressure levels in adolescents (10). Thus the aim of this study was to determine whether blood pressure is associated with sedentary behavior in adolescents even after controlling for potential confounders. Furthermore, one of the objectives was verifying cluster analysis (sedentary behavior plus obesity and sedentary behavior plus insufficient physical activity) for the relationship between sedentary behavior and HBP.

Methods

Ethics

This study was previously approved by the Ethics and Research Committee of the State University of Londrina and was conducted in accordance with the Declaration of Helsinki. The inclusion of all adolescent participants in this study was approved by their parents or guardians (protocol: 203/10).

Sample

The study was conducted with adolescents of both genders in the city of Londrina in southern Brazil. Londrina is the third largest city in southern Brazil and has a human development index of 0.842. Sample size was estimated using an HBP prevalence of 17.3% (11), error of 2.8, Z = 1.96 and design effect of 50%, which indicated a minimum sample of at least 1018 adolescents (predicted losses and refusals were estimated at 20%, resulting in a sample of 1222 adolescents).

In a meeting with the chief of the regional board of education, which is the entity responsible for all schools in the city of Londrina, it was noted that the six largest schools located in the central part of the city received students from all regions (north, south, east, west and central areas) of the city and, therefore, that these schools had student populations that were representative of students throughout the entire city. All classrooms were invited to participate and five visits were performed in each classroom, trying to minimize possible losses due to absences of the students. Data collection was carried out by trained staff from March to September 2011. The questionnaires were administered in the classroom under supervision of the head researcher. After fieldwork, data were collected from 1231 adolescents, with ages ranging from 14 to 17 years old.

Blood pressure

Blood pressure was measured twice, using an oscillometric device (Omron HEM 742) that had previously been validated for use in adolescents (12). All subjects were seated at rest for 5 min before the first measurement, and the interval between the first and second evaluations was 2 min. After the two assessments (in the same day), the mean values of systolic and diastolic blood pressure were calculated, and those who had blood pressure values above the 95th percentile, according to the criteria established by the National High Blood Pressure Education Program (13), were classified as having HBP.

Sedentary behavior

Sedentary behavior was assessed through daily use of three types of electronic devices: television, computer and video games. Adolescents were asked how many hours per day they watched television, used the computer and played video games. The total number of hours on these three devices was summed and adolescents who reported at least 2 h per day were classified as having sedentary or high sedentary behavior according to the American Academy of Pediatrics (14).

Body mass index

The adolescents were weighed with a digital scale accurate to 0.1 kg and their height was measured using a portable stadiometer accurate to 1 cm. Body mass index (BMI) was calculated by dividing the weight (kg) by the square of the height (m²).

Obesity

The waist circumference was determined by measuring the minimum circumference between the last rib and the iliac crest using an inextensible tape, in millimeters (mm). Participants in the sample were classified as with or without abdominal obesity, according to their age and gender, following the criteria proposed by Taylor et al. (15).

Physical activity

The practice of physical activity was assessed by the questionnaire developed by Baecke et al. (16). This questionnaire evaluates physical activities through three different domains (school, leisure time and sports), and the sum of these three domains is described as total physical activity. The questionnaire has a Likert scale with the following response options: never, rarely, sometimes, often and always. When considering physical activities in sports, the number of hours per week and the number of months per year spent engaged in a given sport are also taken into account. In the present study, the total amount of physical activity performed during leisure time was considered, e.g. cycling, walking and sports practiced outside school. In this study, two domains were utilized (leisure and sports). The cut-off points for physical activity classification were arbitrarily (16) defined using quartiles (quartile 4, 3 and 2 = sufficiently active and quartile 1 = insufficiently active).

Smoking and alcohol consumption

To assess smoking in adolescents, they were asked whether they smoked and the number of cigarettes they smoked per day. Adolescents were considered smokers if they reported smoking at least one cigarette per day during the week in which the study was conducted.

Alcohol consumption was assessed through questions based on the questionnaire of the Brazilian Center for Psychotropic Drugs (CEBRID) (17), and the number of weekly drinks (each drink corresponded to 250 ml) was investigated. Participants were considered as high alcohol consumers if they reported drinking alcohol on at least 1–2 days per week with an intake of 1–2 servings per day (18).

Sociodemographic variables

We collected the following variables: gender, age and economic class. The Economic Classification Criteria of Brazil in 2011, from the Brazilian Association of Research Companies (ABEP) (19), was employed for the definition of the economic class of adolescents. We took into account the presence of material goods and a maid at home, and the educational level of the household head. Based on the score obtained from the sum of the scores assigned to the items examined, adolescents were grouped into the following economic classes: low, medium and high.

Statistical analysis

The sample characterization was shown as mean and standard deviation. To examine associations between HBP and high sedentary behavior and possible confounding variables, we used the chi-squared test. Those variables that were associated with high blood pressure (p < 0.05) were included in the multivariate model. Possible confounding variables (age, gender, socioeconomic status, smoking and alcohol consumption) and variables that had a p-value < 0.20 were also included in the multivariate model in order to control for confounding effects. Multivariate analysis was used to conduct binary logistic regression in which the objective was to determine whether HBP was associated with sedentary behavior even after the introduction of the variables reported above. Later, it was verified whether the clusters of high sedentary behavior and obesity and high sedentary behavior and insufficient physical activity were associated with elevated blood pressure values. The statistical program used was BioEstat (version 5.0) and the significance was set at 5%.

Results

The average age of the adolescents who participated in the study was 15.55 years. The prevalence of HBP observed in this study was 12% [95% confidence interval (CI) 10.2–13.8%] considering the overall sample (n = 1231: 515 boys and 716 girls). Boys had a higher prevalence of HBP (19.2%, 95% CI 15.8–22.6%) than girls (6.8%, 95% CI 5.1–8.7%) (p < 0.001). Sedentary behavior above 2 h a day was identified in approximately 90% of adolescents. Table I presents the means of continuous variables according to the classification of blood pressure (HBP or normal blood pressure). With the exception of tobacco, alcohol and physical activity, all variables had significant differences.

Table I. General characteristics of the sample according to blood pressure level.

Variable	Normal (n = 1083) Mean ± SD	HBP (n = 148) Mean ± SD	p
Age (years)	15.53 ± 1.06	15.73 ± 1.07	0.302
Weight (kg)	58.24 ± 10.68	73.83 ± 19.36	< 0.001
Height (cm)	166.16 ± 9.05	170.62 ± 9.22	< 0.001
BMI (kg/m²)	21.04 ± 3.21	25.25 ± 5.94	< 0.001
WC (cm)	71.12 ± 8.05	81.46 ± 13.22	< 0.001
Tobacco (cigarettes/day)	0.26 ± 1.80	0.52 ± 2.80	0.130
Alcohol (days/week)	0.32 ± 0.64	0.32 ± 0.69	0.963
Sedentary behaviors (h/week)	32.41 ± 14.57	39.07 ± 16.29	< 0.001
Physical activity (Baecke's score)	6.01 ± 2.67	6.27 ± 2.83	0.263

HBP, high blood pressure; BMI, body mass index; WC, waist circumference.

Table II shows observed associations between HBP and independent variables. Table III shows the multivariate analysis between blood pressure and sedentary behavior. In the crude analysis, there was a significant association between HBP and high sedentary behavior. Subsequently, one of the confounders was inserted and it was observed that the association remained significant, but a major change in odds ratio (OR) and in the confidence interval was noted when the obesity variable was inserted.

Table II. Association between high blood pressure (HBP) and independent variables in adolescents.

Variable	Total (n = 1231) n	HBP (n = 148) n (%)	Chi-squared test p
Gender
Male	515	99 (19.2)	< 0.001
Female	716	49 (6.8)
Age (years)
14–15	569	53 (9.3)	0.009
16–17	662	95 (14.3)
Socioeconomic level
Low	68	8 (11.7)	0.816
Medium	683	84 (12.2)
High	480	56 (11.6)
Tobacco
Yes	42	7 (16.6)	0.484
No	1189	141 (11.8)
Alcohol
Yes	255	33 (12.9)	0.690
No	976	115 (11.7)
Abdominal obesity
Yes	216	79 (36.5)	< 0.001
No	1015	69 (6.7)
Sedentary behaviors
≥ 2 h	1107	123 (12.4)	< 0.001
< 2 h	124	25 (8.1)
Physical activity
Insufficient	924	102 (11.7)	0.082
Sufficient	307	46 (14.9)

Table III. Association between high blood pressure and sedentary behaviors in adolescents.

Multivariate model	OR	CI (95%)	p
Unadjusted model	2.21	(1.41–3.96)	0.001
Model 1 (adjusted by gender)	2.01	(1.27–3.17)	0.003
Model 2 (Model 1 + age)	2.17	(1.37–3.44)	0.001
Model 3 (Model 2 + socioeconomic level)	2.18	(1.37–3.45)	0.001
Model 4 (Model 3 + tobacco)	2.16	(1.35–2.47)	0.001
Model 5 (Model 4 + alcohol)	2.16	(1.36–2.45)	0.001
Model 6 (Model 5 + abdominal obesity)	1.70	(1.04–2.77)	0.032
Model 7 (Model 6 + physical activity)	1.68	(1.03–2.75)	0.037

Model 1, unadjusted model adjusted by gender; Model 2, Model 1 adjusted by age; Model 3, Model 2 adjusted by socioeconomic level; Model 4, Model 3 adjusted by tobacco; Model 5, Model 4 adjusted by alcohol; Model 6, Model 5 adjusted by abdominal obesity; Model 7, Model 6 adjusted by physical activity.

OR, odds ratio; CI, confidence interval.

A cluster analysis was performed involving sedentary behavior and waist circumference and their relation to HBP. After logistic regression analysis, it was observed that in the raw data relating to high sedentary behavior and normal waist circumference, there was no association between this cluster and HBP in adolescents (OR = 1.72, 95% CI 0.95–3.09). However, when considering low sedentary behavior combined with abdominal obesity of adolescents, there was a significant association (OR = 6.80, 95% CI 2.81–16.45), and this association increased when analyzed in conjunction with high sedentary behavior and obesity (OR = 13.51, 95% CI 7.21–23.97).

After adjustment for potential confounders (gender, age, economic status, smoking, alcohol and physical activity), the associations remained virtually the same. The Hosmer–Lemeshow test identified that the models were adequately fit and the same model could explain up to 80% of the occurrence of associations observed with HBP in adolescents (Figures 1 and 2).

Figure 1. Cluster of obesity and sedentary behaviors (SB) and blood pressure. WC, waist circumference; CI, confidence interval.

Figure 2. Cluster of physical activity (PA) and sedentary behaviors (SB) in blood pressure. Ins., insufficiently; CI, confidence interval.

When we analyzed the cluster involving physical activity and sedentary behavior in the crude analysis, no association was found between low sedentary behavior and insufficient physical activity (OR = 0.71, CI 0.30–1.66). When considering high sedentary behavior and sufficient physical activity, there was a significant correlation between those variables and HBP (OR = 2.39, CI 1.10–5.17). Surprisingly, when considering high sedentary behavior and insufficient physical activity, there was no significant correlation with HBP (OR = 1.57, CI 0.76–3.23). After adjusting for confounders, there was no significant association between clusters and HBP in adolescents.

Discussion

This study aimed to explore the possible association between HBP and sedentary behavior. It was observed that even after controlling for confounding variables, the association between HBP and high sedentary behavior remained significant. These findings are worrisome because about 90% of the study sample reported having high sedentary behavior according to the protocol recommended by the American Academy of Pediatrics. The high prevalence of sedentary behavior in the present study is in line with other studies conducted in different parts of the world (20,21), showing that this problem is worldwide.

The association found in our study is in agreement with the findings of Martínez-Gómez et al. (8) who, after evaluating 210 Spanish adolescents, found that those who had a high level of sedentary behavior were more likely to have cardiovascular risk factors, including HBP. In another study from the same group of researchers, it was observed that the time children spent in front of the television was associated with HBP, whereas the youth who were classified in the lowest tertiles had lower values of blood pressure (21). Goldfield et al. (22) also corroborate the findings of this study after verifying an association between adolescents who spent much of their time sitting around playing video games and HBP.

One possible explanation for this association is the fact that the more time is spent in sedentary behaviors with low energy expenditure, the greater the likelihood of having less time for vigorous activity and subsequently the release of nitric oxide causing a decrease in arterial pressure (23). Excessive time spent being sedentary can contribute to elevated blood pressure in different ways: one is due to the relationship between longer screen time and increased consumption of high-calorie foods, which contributes directly to increased caloric intake (24), increasing body weight and hence higher blood pressure values (25). In the present study, the association between HBP and sedentary behavior after controlling for confounding variables tested one by one remained significant, but there was a clinically significant change in the relationship with the abdominal obesity variable. This finding suggests that sedentary behavior is associated with HBP in adolescents, but that excess fat helps to modulate this relationship. These results are worrying, since the inactivity linked to excess body fat may be associated with changes in left ventricular structure (26), causing problems in the future. Some physiological factors may help to elucidate the results of our study. Abdominal obesity contributes to increased sympathetic nervous system activity derived by stimulation of peripheral α₁- and β-adrenergic receptors. Moreover, excess adiposity causes endothelial dysfunction, contributing to the release of adipokines that are associated with greater insulin resistance, causing increased vasoconstriction and subsequently contributing to hypertension (27).

An additional analysis was performed by clustering variables in this study with the hypothesis that sedentary behavior, when linked with excess fat deposited in the abdominal region, could further increase the risk of HBP in adolescents. This hypothesis was confirmed: adolescents with obesity and a sedentary lifestyle were 13 times more likely (OR = 13.51, CI 7.21–23.97) to have HBP compared to adolescents who did not have any of these risk factors. This association was significant in the crude analysis that controlled for several confounding factors. In line with what was mentioned above, screen time contributed to the consumption of unhealthy foods in adolescents, since several commercials for such food are aired on television and the internet (28), contributing to an increase in weight. Furthermore, Shi and Mao (29) observed that excessive use of a computer and television was associated with the consumption of soft drinks in adolescents. This would present two factors in addition to obesity. The first is that high doses of caffeine (as contained in some of these soft drinks) can contribute to increased arterial pressure (30), and the second is that pediatric populations obtain less sleep and increase their screen time, successively sleeping less, and are thus more likely to have increased blood pressure on account of their having too few hours of sleep (31).

In our study we also conducted a cluster analysis considering the sum of sedentary behavior and physical activity. The only significant association was that sedentary but sufficiently active adolescents were more likely to develop HBP compared to adolescents with low sedentary behavior and sufficient physical activity. Our findings are in line with those of Ekelund et al. (9) who, after cluster analysis between sedentary behavior adjusted for physical activity and other factors, found no differences in clustered metabolic risk in pediatric populations. Unlike our study, de Moraes et al. (32) compared two epidemiological studies (one conducted in Europe and one in Brazil) and found that sufficiently active male adolescents with low sedentary behavior have lower chances of having HBP. Differences between these findings can be explained as follows. First, compared to the Brazilian study, the instrument for the assessment of physical activity was different between studies (Baecke and IPAQ, respectively); and secondly, the assessment of physical activity in the Europe study was also conducted using an accelerometer, which provides more reliable measures of physical activity, especially considering its intensity (light, moderate or vigorous).

One limitation of this study is its cross-sectional design, which does not allow for a cause-and-effect analysis. Another limiting factor is that physical activity was assessed by questionnaire, which can create some problems for the classification of physical activity, especially its intensity. In addition, blood pressure was measured in a single day, which can overestimate the prevalence of HBP. Furthermore, genetic factors such as family history of hypertension may also exert an influence on high levels of blood pressure in adolescents, and the lack of such data in this study may be considered another limitation. Positive factors include the evaluation of not only watching television, but also using computer and video games, providing a more detailed analysis of this behavior. The second factor is that few studies have conducted an analysis of physical inactivity and HBP (22) and shaped clusters (32). Furthermore, it is noteworthy that not only day-school students, but also night-school students were included in this study, thus avoiding sample homogeneity, since adolescents who work and study at night may have different lifestyles from other adolescents.

Conclusion

Based on the findings of this study, sedentary behavior may be associated with HBP levels in adolescence, and this behavior, when associated with abdominal obesity, contributes significantly to the increased chances of having HBP. Actions that promote not only the reduction of sedentary behavior, but also increased physical activity and reduced body weight should be encouraged, beginning at an early age, since behaviors acquired at that stage of life tend to be maintained in adulthood (33–35).

Authors’ contributions

DGDC was the researcher responsible for the collection, analysis and interpretation of data, and also for drafting the manuscript. SMA, AEM and JSC were involved in the analysis and interpretation of data as well as the critical review of the paper. JRC and RAF were involved in critically reviewing the manuscript for important intellectual content.

Declaration of interest: The authors of this paper have no conflict of interest to declare.