Blood Pressure, Heart Rate Variability, and Adiposity in Caribbean Pre-pubertal Children

Abstract

Childhood obesity prevalence has increased over the last 30 years. The Heart Rate Variability (HRV) studies performed in adults suggest a possible relation between abnormal autonomic regulation and hypertension in the situation of overweight or obesity. Objective: The aims of this study were to explore the early relationships between adiposity and blood pressure and HRV in pre-pubertal children. Methods: Data were collected during the medical examination of the follow-up at 7 years of the TIMOUN mother-child cohort in Guadeloupe. Body Mass Index z-score (zBMI), sum of tricipital and subscapular skinfold thickness, percentage of fat mass, and Waist-to-Height Ratio were measured. A global corpulence score was computed using a Principal Component Analysis (PCA). Systolic Blood Pressure (SBP) and HRV parameters (cardiac holter monitoring) were collected under 2 conditions (calm and tachycardial period). Relations between HRV, SBP, each adiposity indicator and the corpulence score were studied with restricted cubic splines models, and linear regression models. The age at adiposity rebound (AR) was estimated from the individual growth curves. Results: 575 children were included in the SBP study (mean age: 7.7 years, from 85 to 99 months). SBP was linearly correlated with the corpulence score and the zBMI. An increase of 1 in the zBMI was associated with an increase of 2.3 (±0.28) mmHg in SBP. The effect-size of zBMI on SBP was higher in children with early age at AR. Compared to children with normal BMI, children with a zBMI <™2SD had their RMSSD, SDNN, LF and HF indicators in tachycardial conditions significantly reduced by -30, -21, -37, and -48%, respectively. In boys with a zBMI >2SD, we observed a global increase in all HRV parameters (under tachycardial conditions), particularly the LF [β = 0.43 (±0.18)]. Conclusion: In pre-pubertal period a positive correlation between adiposity excess and SBP was observed with significant changes of HRV in boys, arguing for an early abnormal autonomic regulation and for early preventive intervention in the infancy period, particularly in case of overweight or obesity. Thinness was associated with a reduction in almost all the HRV parameters studied, when compared to normal corpulence, suggesting a decrease in autonomic influence.

Keywords: BMI; adiposity; blood pressure; children; fat mass; heart rate variability (HRV); pre pubertal; systolic blood pressure.

Figure 1

Flow chart of the samples used for analysis.

Figure 2

Association between SBP (mmHg) and BMI z-score (A) and corpulence score (B). Restricted cubic splines regression models, adjusted for sex of the child, age at measurement, maternal place of birth and education, TV and videogames times, sport time, indicator of obesogenic food consumption. All non-linear associations were non-significant (p > 0.05). Interpretation: The reference value for the adiposity indicator (indicated with a vertical dotted line) was 0 for the z-BMI and the median level for the PCA based corpulence indicator. The y-axis represents the mean difference in SBP (mmHg) associated with the adiposity indicator compared to the mean level of SBP associated with the reference level of the adiposity indicator. The black points in the bottom of the graphs represent the number of children observed in the study at each level of the adiposity indicator. For the PCA-based adiposity indicator, the blue points on the spline regression curve represent the knots, set at the 5th, 25th, 50th, and 75th percentiles of the adiposity indicator. For the BMI z-score, the knots correspond to the WHO thresholds defining underweight and overweight and obesity (respectively −2SD, +1SD, +2 SD). The bluish surface around the estimated regression curve represents the 95% confidence interval of the values of variation in SBP estimated from the spline regression model.

Figure 3

Association between SBP (mmHg) and BMI z-score, according to sex. Restricted cubic splines regression models, adjusted for age at measurement, maternal place of birth and education, TV and videogames times, sport time, indicator of obesogenic food consumption. The y-axis represents the mean difference in SBP (mmHg) associated with BMI z-scores compared to the mean level of SBP associated with a 0 BMI z-score. All non-linear associations were non-significant (p > 0.05). Blue line corresponds to boys (1) and red lines to girls (2).

Figure 4

Association between HRV parameters (Y-axis) and BMI z-score (X-axis). Restricted cubic splines regression models, adjusted for age at measurement, maternal place of birth and education, TV and videogames times, sport time, indicator of obesogenic food consumption.