4. MATERIAIS E MÉTODOS
4.6 Análise estatística
Todas as análises foram realizadas no programa Statistical Package for the Social Sciences, versão 20 (IBM Corp, NY, USA). As análises abrangeram estatística descritiva e inferencial. Primeiramente foi realizado o teste de normalidade dos parâmetros da VFC, verificada através do teste Kolmogorov- Smirnoff. Para descrição dos parâmetros foram utilizados média ± desvio padrão. Para descrição dos indicadores de atividade física foram utilizadas as frequências absolutas e relativas das variáveis.
A análise de regressão logística binária foi realizada para analisar a associação entre os parâmetros da VFC que foram classificados em quartis e depois dicotomizados para a análise (1º quartil vs 2º ao 4º para o SDNN, RMSSD, pNN50 e 4º quartil vs 1º ao 3º quartis para LF/HF) com o nível de atividade física de lazer (não=0, sim=1), atividade física de deslocamento (não ativo=0, ativo=1), tempo de prática regular de atividade física (não ativo há mais de 6 meses=0, ativo há mais de 6 meses=1) com. Também foi realizada uma análise de agrupamento, onde foi realizada uma regressão logística binária entre os parâmetros da VFC com a atividade física de lazer juntamente com a prática regular de atividade física, a atividade física de deslocamento juntamente com prática regular de atividade física e as três variáveis (lazer, deslocamento e tempo de prática regular) juntamente.
Estas análises foram feitas ajustando o modelo pelas variáveis turno da coleta, obesidade e hipertensão. O modelo da associação entre atividade física de deslocamento e a VFC foi posteriormente ajustado pela atividade física de lazer. A avaliação do modelo foi realizada pelo teste de Hosmer-Lemeshow. O nível de significância adotado para todas as análises foi um valor de p<0,05 e o intervalo de confiança (IC) de 95%.
5. RESULTADOS
Original Article
ASSOCIATION BETWEEN LEISURE-TIME AND COMMUTING PHYSICAL ACTIVITIES WITH HEART RATE VARIABILITY IN MALE ADOLESCENTS
ABSTRACT
The aim of this study was to investigate the association between heart rate variability (HRV) parameters with leisure-time and commuting physical activities with in adolescent boys. The sample included 1152 male adolescents aged between 14 and 19 years. The variation of consecutive heart beats (RR intervals) was assessed and HRV parameters in time (SDNN, RMSSD, pNN50) and frequency domains (LF/HF) were calculated. Leisure-time and commuting physical activities were obtained using a questionnaire based on the Global School-based Student Health Survey. A binary logistic regression between HRV parameters and physical activity types adjusted by hypertension, obesity and period of the day was performed. Leisure-time physical activity was associated with SDNN (OR=0.573 CI=0.421 to 0.780), RMSSD (OR=0.588 CI=0.431 to 0.802), pNN50 (OR=0.597 CI=0.438 to 0.814), while LF/HF was not associated (p>0.05). These associations were stronger when adolescents were also physically active for more than six months (SDNN [OR=0.462 CI=0.317 to 0.672], RMSSD [OR=0.446 CI=0.306 to 0.651], pNN50 [OR=0.479 CI=0.328 to 0.698]). Commuting physical activity was not associated with any HRV parameter (p>0.05). Boys that practised commuting physical activity and were also physically active for more than six months presented a lower chance of having low HRV in SDNN (OR=0.619 CI=0.422 to 0.909) and RMSSD (OR=0.664 CI=0.452 to 0.974). However, these associations were lost after adjustment for leisure-time physical activities (p>0.05). Leisure-time physical activity was associated with better HRV and these associations were enhanced when adolescents were physically active for more than six months.
INTRODUCTION
Cardiovascular disease is the leading cause of mortality worldwide (Dick and Ferguson 2015). The risk factors for cardiovascular disease can start in childhood (Farias Júnior, Nahas et al. 2009, Farias Júnior, Mendes et al. 2011, Christofaro, Ritti-Dias et al. 2013, Magliano, Guedes et al. 2013) and remain in adulthood (Twisk, Van Mechelen et al. 1997).These factors cause subclinical manifestations in the cardiovascular system even in the first decade of life (Berenson 2002). The study of early cardiovascular risk factors in adolescents has provided valuable public health information. Thus, effective public policies are promoted to encourage the adoption of healthy lifestyles, preventing these adolescents from exposure to factors associated with cardiovascular risk.
Heart rate variability (HRV), characterized as the variation of consecutive heartbeats, is an early marker of cardiovascular health in children and adolescents (Zhou, Xie et al. 2012, Farah, Barros et al. 2014). Reduced HRV was reported to be an independent predictor of mortality and a worse HRV has been associated with cardiovascular disease (Kleiger, Stein et al. 2005).The HRV is influenced by factors such as physical activity, where higher physical activity levels have been associated with better HRV (Sotiriou, Kouidi et al. 2013, Soares- Miranda, Sattelmair et al. 2014, Sharma, Subramanian et al. 2015).
In adolescents, leisure-time physical activity has been associated with better HRV parameters (Nagai and Moritani 2004, Chen, Chiu et al. 2012). However, whether commuting physical activity (walking or cycling to school), which is frequently performed by adolescents, is also associated with HRV parameters in adolescents is unknown. Given that leisure-time physical activity is commonly performed at higher intensities than commuting physical activities, it is plausible that the latter are not associated with HRV parameters in this age group (Gutin, Barbeau et al. 2000, Nagai and Moritani 2004, Chen, Chiu et al. 2012).
Physical activity when practised for a long time provides benefits to the cardiovascular system. Studies show that physical activity for a long time decreases the risk of developing non-communicable disease (Reiner, Niermann et al. 2013, Shortreed, Peeters et al. 2013). With regard to HRV, there is a gap in the literature, since there are no studies showing the effect of physical activity on heart rate variability over the years.
The hypothesis is that active teens in leisure and commuting have better HRV compared to not active ones and that the practice of physical activity for more than six months brings a positive influence to HRV. To address this issue, the aim of this study was to investigate the association between the HRV parameters with leisure and commuting physical activities in adolescents and the influence of physical activity for more than six months on HRV.
METHODS
Design and subjects
The study presents a design cross-sectional. The target population consisted in students from Pernambuco state’s public schools belonging to the 14–19 age group. For the sample selection, it was used the random sample procedure by conglomerates, stratified into two stages, the first stage school and the second class. All public schools in Pernambuco were considered eligible for inclusion in the study. In the first stage, it was adopted as a stratification criterion for realization of the draw the distribution of small (<200 students), medium (200- 499 students) and large (> 500 students) school size in each micro-region of the Pernambuco state. In the second stage, it considered the class density drawn by period (daytime and nighttime) as a criterion for drawing the classes in which the questionnaires would be applied. The draws were made by random number generation through the site www.randomizer.org.
Were included in the study the male adolescents who reported not consumption of caffeinated beverages 12 hours prior to the HRV evaluation, use of alcohol, any form of tobacco, and/or other illicit drugs, and participation in any physical exercise training 24 hours before evaluations. Were excluded those with HRV signal of low quality and those with poorly filled questionnaire. This study was approved by the Ethics Committee on Human Research of the University of Pernambuco (CAAE-0158.0.097.000-10).
Data collection
Data were collected from May to December of 2011 during the volunteers’ class period. To obtain data on physical activity level, age, ethnicity, housing area
(rural or urban) and issues related to their economic condition we used an adapted version of the Global School-based Student Health Survey (GSSHS).
This questionnaire has been widely used in epidemiological studies with adolescents and have been reported a concordance coefficient (kappa test) between 0.52 and 1.00 for the questionnaire (Tassitano, Barros et al. 2010, Tenório, Barros et al. 2010, Farah, Christofaro et al. 2015).
Assessment of physical activity Leisure-time physical activity
The leisure-time physical activity was assessed by the question: ‘Do you perform regularly some sort of physical activity in your free time, such as exercise, sports, dance or martial arts?’ The adolescents were classified as active (if the answer was yes) or not active.
Commuting physical activity
Commuting physical activity was assessed by the question: ‘During the past seven days, how many days have you walked or cycled to and from school?’ The adolescents whose responded that they got to and from school on foot or by bicycle on three days or more were considered active
Time of regular practice of physical activity
The time of regular practice of physical activity i.e physical activity for more than six months (PA>6 months) was assessed with the question ‘It is considered physically active young man who accumulate at least 60 minutes of daily physical activity on 5 or more days of the week. In relation to their practice habits of physical activity, you say:”. Adolescents were considered PA>6 months if they answered ‘I have been physically active for more than 6 months’.
Covariates
Adolescents were weighed without shoes and coats on an automatic scale, and the height was measures using a stadiometer. Waist circumference was measured in the standing position at the level of the umbilicus using a constant tension tape. Overweight was determined by body mass index above the 85th percentile for their age (Cole, Bellizi, et al. 2000). Abdominal obesity was
determined by waist circumference above the 80th percentile for their age (Taylor, Jones et al. 2000).
Blood pressure was measured using the Omron HEM 742 (Christofaro, Fernandes et al. 2009) (Omron, Shangai, China) after the adolescents rested and remained seated with legs uncrossed for 5 minutes. All blood pressure measurements were performed 3 times in the right arm placed at heart level in a seated position. The mean value of the least 2 measurements was used for analysis. High blood pressure was defined as systolic and/or diastolic blood pressure equal or higher than the reference sex, age, and height-specific 95th percentile (Falkner and Daniels 2004).
Heart rate variability
Heart rate variability was obtained by analysis of RR intervals using a heart rate monitor (POLAR, RS 800 CX; Polar Electro OyInc, Kempele, Finland). Adolescents were in a supine position for 10 minutes after approximately 30 minutes of rest. Analyses were performed using Kubios HRV software (Biosignal Analysis and Medical Imaging Group, Joensuu, Finland) by a single evaluator blind to the other variables, following the recommendations of the Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology (1996).
The time domain variables, such as standard deviation of all RR intervals (SDNN), root mean square of the squared differences between adjacent normal RR intervals (RMSSD), and the percentage of adjacent intervals over 50 ms (pNN50). The frequency domain was analysed using the spectral analysis of HRV. Stationary periods of the tachogram, of at least five minutes, were broken down into bands of low (LF) and high (HF) frequencies, using the autoregressive method with a model order of 12 according to Akaike’s information criterion. Frequencies were considered physiologically significant when they ranged between 0.04 and 0.4 Hz. Oscillations between 0.04 and 0.15 Hz were considered as a LF component, whereas oscillations between 0.15 and 0.4 Hz represented the HF component. For analysis, we used the LF/HF component as an indicator of the sympathovagal balance on the heart. The reliability of the measurements was assessed using the intraclass correlation coefficient (ICC) and the values ranged from 0.70 to 0.91 (Farah, Lima et al. 2014).
Statistical analyses
All statistical analyses were performed using the software SPSS/PASW version 20.0 (IBM Corp, Armonk, New York). Data normality was verified by the Kolmogorov–Smirnoff test. The HRV parameters data are presented in mean, standard deviation and confidence intervals (CI) of 95%. The HRV parameters were was classified in quartiles and then dichotomized for analysis (1st quartile vs 2nd to 4th to SDNN, RMSSD, pNN50 and 4th quartile vs 1st to 3rd quartiles for LF/HF).
The binary logistic regression analysis was performed to analyse the association between the leisure-time and commuting physical activities with HRV. Was also conducted a cluster analysis, where a binary logistic regression between the parameters of HRV with leisure-time physical activity along with the regular practice of physical activity (LPA+PA>6), commuting physical activity along with regular practice of physical activity (CPA+PA>6) and the three variables together (CPA+LPA+PA>6). Adjusted for the period of the day, obesity and hypertension. The Hosmer–Lemeshow test was used to evaluate the goodness of fit. The significance level for all analysis was p<0.05.
RESULTS
A total of 1212 male adolescents were enrolled in the study, 60 boys were excluded due to low signal quality (stationary periods of the tachogram lengths lower than 5 minutes). Thus, the final analysis consists of data from 1152 male adolescents with a mean age of 16.6 ± 1.2 years. When asked which leisure activity they preferred, 43.8% of adolescents said playing sports. Table 1 describes the characteristics of the sample and the percentage of activity in each type of physical activity. Table 2 presents the criteria used for the stratification of HRV parameters in statistical analysis.
Table 1. Characteristics of adolescents (n=1152). Variables Values Age (years) 16.6 ± 1.2 Weight (kg) 63.7 ± 12.6 Height (cm) 171.6 ± 7.1 WC (cm) 76.7 ± 9.5
Body mass index 21.6 ± 3.8
SBP (mmHg) 121.6 ± 12.4
DBP (mmHg) 67.8 ± 8.6
Heart rate (bpm) 71.7 ± 11.8
Race (% non-white) 72
Place of residence (% urban) 79
Period of the day (% matutinal) 42
Physical activity leisure (% actives) 78.2
Physical activity commuting (% actives) 51.3
Time of regular practice of physical activity (% actives for more the six months) 46.4
WC: waist circumference; SBP: systolic blood pressure; DBP: diastolic blood pressure.
Table 2. Criterion used for stratification of HRV parameters in statistical analysis (n=1152). HRV parameters Mean ± SD Quartile 1st 2nd 3rd 4th Time Domain SDNN (ms) 61,9 ± 23,7 44,5 58,4 76,5 157,0 RMSSD (ms) 54,7 ± 29,5 33,2 49,7 68,9 204,2 pNN50 (%) 29,5 ± 20,5 11,0 28,1 45,1 84,0 Frequency domain LF/HF 1,4 ± 1,1 0,7 1,1 1,8 7,9
SD: standard deviation; SDNN: standard deviation of the RR interval; RMSSD: root mean square of the squared differences between adjacent normal RR intervals; pNN50: the percentage of adjacent intervals over 50 ms; LF/HF: sympathovagal balance.
The association between leisure-time and commuting physical activities with HRV parameters are shown in Figure 1. There were significant associations between leisure-time physical activity and HRV parameters in the time domain (SDNN [OR = 0.573 95%CI = 0.421 to 0.780], RMSSD [OR= 0.588 95%CI = 0.431 to 0.802], pNN50 [OR = 0.597 95%CI = 0.438 to 0.814]). Commuting physical activity was not associated with any HRV parameter. Total volume of commuting physical activity was not associated with any HRV parameter (SDNN [p = 0.937], RMSSD [p = 0.664], pNN50 [p = 0.323]).
The cluster of leisure-time physical activity and PA>6 months (Table 3) revealed significant associations with HRV in the time domain (p<0.05). The cluster of commuting physical activity and PA>6 months was significant (p<0.05) only in the SDNN and RMSSD parameters. However, these associations were lost after an additional adjustment for leisure-time physical activity.
Figure 1. Association between the types of physical activity and the parameters the heart
rate variability. SDNN – standard deviation of the RR interval, RMSSD – root mean square of the squared differences between adjacent normal RR intervals, pNN50 - the percentage of adjacent intervals over 50 ms, LF/HF – sympathovagal balance (ratio between the low and high frequency components). OR: odds ratio; CI: confidence interval. Models adjusted by hypertension, obesity and period of the day.
Table 3. Association between the cluster the types of physical activity and HRV parameters.
SDNN (1st quartile) RMSSD (1st quartile) pNN50 (1st quartile) LF/HF (4th quartile)
OR CI (95%) OR CI (95%) OR CI (95%) OR CI (95%)
LPA + PA>6 0.462 0.317 – 0.672 0.446 0.306 – 0.651 0.479 0.328 – 0.698 1.194 0.808 – 1.765
CPA + PA>6 0.619 0.422 – 0.909 0.664 0.452 – 0.974 0.768 0.522 – 1.129 1.237 0.848 – 1.806
CPA + PA>6+LPA 0.735 0.491 – 1.101 0.785 0.524 – 1.176 0.917 0.611 – 1.378 1.131 0.760 – 1.683 LPA: leisure-time physical activity; PA: physical activity; CPA: commuting physical activity. Models adjusted by hypertension, obesity and period of the day.
DISCUSSION
The main results of this study were: (i) leisure-time physical activity is associated with better HRV; (ii) the associations between leisure-time physical activity were enhanced when physical activity is practised for more than 6 months; (iii) commuting physical activity is not associated with HRV in adolescents.
The strengths of this study include the large sample size, since there are no epidemiological studies with a sample of this size analysing HRV. We assessed blood pressure, which is an important confounding variable closely linked to HRV. The study examined the relationship of HRV and different physical activity, which has not previously been done.
Sports practice is the main frequent form of leisure-time physical activity in adolescent boys (Azevedo, Araujo et al. 2007). The association between leisure- time physical activity and cardiovascular function (Buchheit, Platat et al. 2007, DeFina, Haskell et al. 2015), including HRV parameters, was previously described in adolescents (Cayres, Vanderlei et al. 2015). In this study, adolescents active in leisure-time also presented better HRV. However, this study expands the current knowledge, showing that adolescents active during leisure time and who practised physical activities for more than six months presented an even better HRV.
Although the mechanisms underlying this association were not investigated, these results probably reflect that long-term physical activity involvement alters both the functional and structural cardiovascular parameters.
Regular physical activity brings changes that are evident at rest, such as a decrease in HR, improvements in the neurohumoral modulation caused by the decrease in catecholamine levels(Vanhees, Fagard et al. 1984) and angiotensin II (Vanhees, Fagard et al. 1984, Fernandes, Hashimoto et al. 2011) and an increase in nitric oxide levels (Hambrecht, Adams et al. 2003). Another aspect noted with the practice of physical activity is related to the modulation of cardiorespiratory capacity which may slow the reduction of parasympathetic activity (De Meersman and Stein 2007, Fernandes, Vaz Ronque et al. 2013, DeFina, Haskell et al. 2015).These adaptations alter the autonomic nervous system, increasing vagal and decreasing sympathetic activity to the heart (da Silva, Pereira et al. 2014, Panda and Krishna 2014).
Commuting activity such as walking or cycling was not associated with HRV parameters and the main hypothesis for that is the low volume and intensity of these activities. Active commuting is opted for mainly when distances are no longer than 20 min (Silva, Vasques et al. 2011, Silva, Lopes Ada et al. 2014) and is commonly performed at low to moderate intensity. Therefore the volume and intensity of commuting physical activities are probably not enough to promote adaptations in cardiac autonomic modulation (Rennie, Hemingway et al. 2003, Gutin, Howe et al. 2005, Buchheit, Platat et al. 2007, Sandercock, Hardy- Shepherd et al. 2008). Interestingly, although the commuting physical activity along with the PA>6 months was associated with the HRV parameters, the observed associations were lost after adjustment for leisure-time physical activity. This fact can be explained by not having statistical power due to the percentage of adolescents who are active in three contexts.
Regarding practical applications, the present study shows that physical activity performed during leisure time – most likely being more intense – can provide improvements in the autonomic system, especially when taken over a longer period. As health promotion actions, physical activity of at least moderate intensity should be encouraged from the earliest ages, aiming to provide better cardiovascular conditions in the population.
Some limitations of this study should be considered. The cross-sectional design is the main concern, as it limits the establishment of causal relationships. The sample included only male adolescents, and extrapolation for female
adolescents’ is limited. The leisure-time physical activity included all leisure physical activities together and it was not possible to determine the influence of different types of physical activities. Commuting physical activity considered only the route to and from school, and whether other commuting physical activities affect HRV cannot be assessed.
CONCLUSION
Leisure-time physical activity is associated with better HRV, and these associations are enhanced when the adolescent is physically active for more than six months. On the other hand, commuting physical activity was not associated with HRV parameters in adolescents.
REFERENCES
1. Dick B, Ferguson BJ. Health for the world's adolescents: a second chance in the second decade. The Journal of adolescent health : official publication of the Society for Adolescent Medicine. 2015;56(1):3-6.
2. Farias Júnior JCd, Mendes JKF, Barbosa DBM, Lopes AdS. Fatores de risco cardiovascular em adolescentes: prevalência e associação com fatores sociodemográficos. Revista Brasileira de Epidemiologia. 2011;14:50-62.
3. Farias Júnior JCd, Nahas MV, Barros MVGd, Loch MR, Oliveira ESAd, De Bem MFL, et al. Comportamentos de risco à saúde em adolescentes no Sul do Brasil: prevalência e fatores associados. Revista Panamericana de Salud