| [Objective]The present study aims to analyze the consistency of puberty measurements in discriminating timing of pubertal onset,to analyze the influence of pubertal timing and body mass index(BMI)trajectories on high blood pressure(HBP)during puberty,and to analyze the interactions between BMI trajectories and pubertal timing on HBP.[Methods]Two cohorts were used in the present study.The one was puberty cohort,which was conducted every 6 months for 30 months for 743 participants,the other was height growth cohort,which was conducted once a year for 12 years for 15,380 participants.For puberty cohort,actual age at Tanner stage ? was defined as the time when genital(boys)or breast(girls)attained Tanner ? stage,age at take-off and age at peak height velocity were calculated with Preece&Baines growth model(PBGM).The three indices were used to discriminate pubertal timing for each participant,and the pubertal timing was discriminated as early,on time or late.The actual age when their serum testosterone(boys)and estradiol(girls)and/or Tanner stage attained Tanner ? was recorded as combined age,and was set as reference age of pubertal onset.Consistency of the indices in discriminating timing of pubertal onset were analyzed with Kappa coefficient test and area under curve(AUC)by operator characteristic curve(ROC),with combined age set as reference.For height growth cohort,age at take-off and age at peak height velocity were fitted by PBGM from 12-year-follow-up height,and were used as indices to discriminate pubertal timing.Pubertal timing was discriminated as early,on time or late.BMI trajectories for up to 12 years were fitted with group-based trajectory model(GBTM),participants were grouped as lean trajectory,normal trajectory,overweight trajectory and obese trajectory.The transition of BMI for each individual was defined as catch-down growth,stable and catch-up growth.HBP was defined as systolic blood pressure(SBP)and/or diastolic blood pressure(DBP)exceeded sex-age-and height-specific 95th percentile.BP for late adolescence was defined by the last measurement of each participant,which happened between the age of 16 and 18.Population-averaged linear models,log-binomial models from generalized linear models and population attributable risks from binary regression models were used to analyze the influence of BMI trajectories and pubertal timing on HBP,and to analyze their interactions to HBP during puberty.[Results]1.Consistency of various determining methods in discriminating timing of pubertal onsetIn boys from puberty cohort,the mean age of testicular Tanner ? stage was 10.4(SD:0.7)years,the mean age at take-off height was 8.8(SD:1.1)years,and the age at peak height velocity was 11.6(SD:0.9)years.In girls from puberty cohort,the mean age of breast Tanner? stage was 9.2(SD:0.7)years,the mean age at take-off height was 7.6(SD:0.8)years,the age at peak height velocity was 9.7(SD:0.8)years.Compared with combined age,the agreement rate of the Tanner ? age and age at peak height velocity in discriminating early onset of puberty ranged from 69.13%to 100%in boys and from 68.00%to 78.75%in girls.The AUC of Tanner ? age was between 0.6300(SE:0.0283)and 0.7135(SE:0.0282).The AUC of age at peak height velocity was between 0.5429(SE:0.0264)and 0.5834(SE:0.0275).The AUC of age at take-off height was between 0.5080(SE:0.0225)and 0.5099(SE:0.0250).These trends were similar when the indices were used to discriminate late pubertal onset.2.The association between timing of pubertal onset and children's blood pressureIn boys from height growth cohort,the mean age at take-off height was 9.5(SD:1.2)years,and the age at peak height velocity was 12.7(SD:1.0)years.In girls from height growth cohort,the mean age at take-off height was 8.4(SD:1.3)years,the age at peak height velocity was 10.7(SD:1.1)years.Timing of pubertal onset was significantly correlated to children's blood pressure.In boys,early pubertal onset(discriminated by age at peak height velocity)was related to significant raise in sex-age-and height-specific BP percentiles during adolescence,with ? of 5.77(95%CI:4.93,6.61;p<0.001)in SBP and ? of 3.94(95%CI:3.31,4.58;p<0.001)in DBP,the corresponding HBP risk also increased,with RR=1.29(95%CI:0.90,1.86;p<0.001)during puberty.The influence of early puberty on HBP last to late adolescence.Compared to children of on time pubertal onset,the RRs for early and late onset groups were 1.38(95%CI:1.22,1.56;p<0.001)and 0.65(95%CI:0.56,0.75;p<0.001),respectively.In girls,early pubertal onset was related to significant raise to both SBP and DBP percentiles during puberty,with ? of 4.36(95%CI:3.37,5.35;p<0.001)and 1.99(95%CI:1.30,2.68;p<0.001),respectively.The correspondence HBP risk was 1.34(95%CI:1.24,1.46;p<0.001).However,the relationship between age of puberty onset and blood pressure became insignificant at late adolescence.3.Interactions between BMI trajectories and timing of pubertal onset on HBP at late adolescenceBMI trajectories were associated to HBP risk in children and adolescents.The HBP risk in children of overweight and obese BMI trajectories raised dramatically after pubertal onset,and remained at a relatively high level until late adolescence.In late adolescence,compared to children of normal BMI trajectory,the HBP risk in overweight and obese boys ranged between 1.93(95%CI:1.66,2.24;p<0.001)and 4.51(95%CI:3.67,5.54;p<0.001),and the HBP risk in overweight and obese girls ranged between 1.99(95%CI:1.45,2.73;p<0.001)and 6.49(95%CI:4.36,9.66;p<0.001)In boys,there were interactions between BMI trajectories and age of pubertal onset to HBP risk at late adolescence.Boys with early pubertal onset and of overweight/obese BMI trajectories were of the highest risk compared to those from on time and normal BMI trajectory group,with of 2.37(95%CI:1.91,2.93;p<0.001).Pre-pubertal overweight and catch-up growth on BMI-z-score during puberty would increase the odds of HBP risk.For the boys who was overweight before puberty and of early puberty onset,catch-down growth of BMI-z-score during puberty would decrease their odds of HBP risk to RR=2.20(95%CI:1.74,2.77;p<0.001),while the RR for catch-up growth was 10.39(95%CI:4.06,26.57;p<0.001).In girls,HBP risk at late adolescence was mainly affected by the trajectory and transition of BMI,rather than by age of pubertal onset.For girls who were overweight before puberty and had catch-up growth during puberty,the odds of HBP risk at late adolescence ranged between 5.22(95%CI:0.62,43.99;p=0.605)and 31.33(95%CI:1.94,506.39;p=0.015)compared to girls who were constantly of normal BMI.Control of pubertal BMI growth could reduce the HBP risk to the same level to reference group.[Conclusion]1.Age at peak height velocity was relatively well correlated to Tanner ? stage and combined age.It could be an alternative to describe the timing of pubertal onset in largesampled investigations.2.Early pubertal onset was correlated with higher HBP risk.It is important for early prevention of cardiovascular disease to make accurate evaluation and appropriate guidance on pubertal timing.3.In boys,there was interaction between pubertal timing and BMI trajectories on HBP risk,and the influence would last until late adolescence.In girls,the HBP risk at late adolescence was mainly affected by BMI trajectories and transitions.For those who were overweight before puberty,or had early pubertal onset,the risk of HBP could be reduced by lowering the growth rate of BMI during puberty. |
PDF Full Text Request