It has been accepted by the pediatric community that blood pressure levels in children are meaningful and have consistently shown correlation with adult bp levels in many studies. In addition, obesity levels in children have increased resulting in corresponding increases in pediatric bp levels. The standard protocol for screening high blood pressure in children is to measure bp on 3 separate visits;if bp is above the 95th percentile on each visit, then a child is designated as hypertensive.
In specific aim 1 of this application, we consider alternative possibly more efficient adaptive screening rules that will be useful in identifying hypertensive children in clinical trials and managing them in clinical practice.
In specific aims 2 and 3 of this application, we propose to refine the relationship between adult and pediatric bp levels using data from the Bogalusa Heart study and the Fels study which have long-term follow-up data of pediatric cohorts into young and mid-adulthood. Childhood bp levels will be characterized both over the long-term based on average levels throughout childhood and adolescence as well as over the short-term based on 3 repeated visits 6 months apart to approximately mimic serial bp readings obtained in clinical practice. Logistic regression models (aim 2) and autoregressive linear mixed models with measurement error (aim 3) will be used for this purpose. Finally, in aim 4, SAS and SPSS macros will be developed to enable researchers to obtain bp percentiles and z-scores adjusted for age, sex and height in batch mode. This will supplement the currently available SAS macro which computes bp percentiles and z- scores on an individual level which is useful in clinical practice.
Analysis of longitudinal cardiopulmonary data Relevance: An important goal of this application is to identify better methods for screening children for hypertension and evaluating treatment efficacy among hypertensive children. A second important goal is to estimate the effect of elevated childhood blood pressure on subsequent adult levels of bp. A third important goal is to propose better method of analysis for longitudinal bp based on the Fels data which includes data on 826 children seen at e 10 visits from early childhood to age 65. These methods should be applicable to other long-term longitudinal bp datasets as well as longitudinal analyses of pulmonary function data and other cardiopulmonary measures.
|Rosner, Bernard; Cook, Nancy R; Daniels, Stephen et al. (2013) Childhood blood pressure trends and risk factors for high blood pressure: the NHANES experience 1988-2008. Hypertension 62:247-54|
|Carey, Vincent J; Wang, You-Gan (2011) Working covariance model selection for generalized estimating equations. Stat Med 30:3117-24|
|Lee, Mei-Ling Ting; Whitmore, G A; Rosner, Bernard A (2010) Threshold regression for survival data with time-varying covariates. Stat Med 29:896-905|
|Rosner, Bernard; Cook, Nancy; Portman, Ron et al. (2009) Blood pressure differences by ethnic group among United States children and adolescents. Hypertension 54:502-8|
|Falkner, Bonita; Gidding, Samuel S; Portman, Ronald et al. (2008) Blood pressure variability and classification of prehypertension and hypertension in adolescence. Pediatrics 122:238-42|
|Rosner, B; Cook, N; Portman, R et al. (2008) Determination of blood pressure percentiles in normal-weight children: some methodological issues. Am J Epidemiol 167:653-66|
|Rosner, Bernard; Glynn, Robert J (2007) Interval estimation for rank correlation coefficients based on the probit transformation with extension to measurement error correction of correlated ranked data. Stat Med 26:633-46|
|Cook, Nancy R; Rosner, Bernard A; Chen, Wei et al. (2004) Using the area under the curve to reduce measurement error in predicting young adult blood pressure from childhood measures. Stat Med 23:3421-35|
|Tishler, Peter V; Carey, Vincent J; Reed, Terry et al. (2002) The role of genotype in determining the effects of cigarette smoking on pulmonary function. Genet Epidemiol 22:272-82|
|Schmid, C H (2001) Marginal and dynamic regression models for longitudinal data. Stat Med 20:3295-311|
Showing the most recent 10 out of 29 publications