Linear spline multilevel models for summarising childhood growth trajectories: A guide to their application using examples from five birth cohorts

Childhood growth is of interest in medical research concerned with determinants and consequences of variation from healthy growth and development. Linear spline multilevel modelling is a useful approach for deriving individual summary measures of growth, which overcomes several data issues (co-linearity of repeat measures, the requirement for all individuals to be measured at the same ages and bias due to missing data). Here, we outline the application of this methodology to model individual trajectories of length/height and weight, drawing on examples from five cohorts from different generations and different geographical regions with varying levels of economic development. We describe the unique features of the data within each cohort that have implications for the application of linear spline multilevel models, for example, differences in the density and inter-individual variation in measurement occasions, and multiple sources of measurement with varying measurement error. After providing example Stata syntax and a suggested workflow for the implementation of linear spline multilevel models, we conclude with a discussion of the advantages and disadvantages of the linear spline approach compared with other growth modelling methods such as fractional polynomials, more complex spline functions and other non-linear models.

[1]  B L De Stavola,et al.  BMI peak in infancy as a predictor for later BMI in the Uppsala Family Study , 2009, International Journal of Obesity.

[2]  Mercedes Onis,et al.  Assessment of differences in linear growth among populations in the WHO Multicentre Growth Reference Study , 2006, Acta paediatrica (Oslo, Norway : 1992). Supplement.

[3]  Jeremy M. G. Taylor,et al.  A Stochastic Model for Analysis of Longitudinal AIDS Data , 1994 .

[4]  Debbie A Lawlor,et al.  Multivariate multilevel spline models for parallel growth processes: application to weight and mean arterial pressure in pregnancy , 2012, Statistics in medicine.

[5]  M S Kramer,et al.  Promotion of Breastfeeding Intervention Trial (PROBIT): a randomized trial in the Republic of Belarus. , 2001, JAMA.

[6]  Roderick J. A. Little,et al.  Statistical Analysis with Missing Data: Little/Statistical Analysis with Missing Data , 2002 .

[7]  E. Eugster.  Childhood Body-Mass Index and the Risk of Coronary Heart Disease in Adulthood , 2008 .

[8]  J M Taylor,et al.  Does the covariance structure matter in longitudinal modelling for the prediction of future CD4 counts? , 1998, Statistics in medicine.

[9]  Richard H. Steckel,et al.  Stature and the Standard of Living , 1995 .

[10]  Graham Watt,et al.  Height and risk of death among men and women: aetiological implications of associations with cardiorespiratory disease and cancer mortality , 2000, Journal of epidemiology and community health.

[11]  P Royston,et al.  The use of fractional polynomials to model continuous risk variables in epidemiology. , 1999, International journal of epidemiology.

[12]  M. Baines,et al.  A new family of mathematical models describing the human growth curve. , 1978, Annals of human biology.

[13]  Tim J Cole.  Modeling postnatal exposures and their interactions with birth size. , 2004, The Journal of nutrition.

[14]  Kate Tilling,et al.  Changes in Ponderal Index and Body Mass Index across Childhood and Their Associations with Fat Mass and Cardiovascular Risk Factors at Age 15 , 2010, PloS one.

[15]  G. Robinson.  That BLUP is a Good Thing: The Estimation of Random Effects , 1991 .

[16]  Debbie A Lawlor,et al.  Maternal education inequalities in height growth rates in early childhood: 2004 Pelotas birth cohort study , 2012, Paediatric and perinatal epidemiology.

[17]  Peter McCarron,et al.  Social Inequalities in Height: Persisting Differences Today Depend upon Height of the Parents , 2012, PloS one.

[18]  Debbie A Lawlor,et al.  Describing differences in weight and length growth trajectories between white and Pakistani infants in the UK: analysis of the Born in Bradford birth cohort study using multilevel linear spline models , 2013, Archives of Disease in Childhood.

[19]  Yoav Ben-Shlomo,et al.  Immediate Postnatal Growth Is Associated With Blood Pressure in Young Adulthood: The Barry Caerphilly Growth Study , 2008, Hypertension.

[20]  Esben Budtz-Jørgensen,et al.  Life course path analysis of birth weight, childhood growth, and adult systolic blood pressure. , 2009, American journal of epidemiology.

[21]  Debbie A Lawlor,et al.  Estimating Trajectories of Energy Intake Through Childhood and Adolescence Using Linear-Spline Multilevel Models , 2013, Epidemiology.

[22]  Debbie A Lawlor,et al.  Socioeconomic differences in childhood growth trajectories: at what age do height inequalities emerge? , 2010, Journal of Epidemiology & Community Health.

[23]  Yu-Kang Tu,et al.  Why evidence for the fetal origins of adult disease might be a statistical artifact: the "reversal paradox" for the relation between birth weight and blood pressure in later life. , 2005, American journal of epidemiology.

[24]  Neil M Davies,et al.  Associations of growth trajectories in infancy and early childhood with later childhood outcomes. , 2011, The American journal of clinical nutrition.

[25]  P Royston,et al.  Calculation of unconditional and conditional reference intervals for foetal size and growth from longitudinal measurements. , 1995, Statistics in medicine.

[26]  Risto Lehtonen,et al.  Multilevel Statistical Models , 2005 .

[27]  Yoav Ben-Shlomo,et al.  Is infant weight associated with childhood blood pressure? Analysis of the Promotion of Breastfeeding Intervention Trial (PROBIT) cohort. , 2011, International journal of epidemiology.

[28]  Russell V. Lenth,et al.  Statistical Analysis With Missing Data (2nd ed.) (Book) , 2004 .

[29]  D. Rubin,et al.  Statistical Analysis with Missing Data. , 1989 .

[30]  H Goldstein,et al.  Multilevel time series models with applications to repeated measures data. , 1994, Statistics in medicine.

[31]  Lawrence C. Marsh.  Spline Regression Models , 2001 .

[32]  Cécile Proust-Lima,et al.  Robustness of the linear mixed model to misspecified error distribution , 2007, Comput. Stat. Data Anal..

[33]  Fiona Steele,et al.  Multilevel models for longitudinal data , 2007 .

[34]  Mark Woodward,et al.  Height, wealth, and health: an overview with new data from three longitudinal studies. , 2009, Economics and human biology.

[35]  T. Sørensen,et al.  [Childhood body mass index and the risk of coronary heart disease in adulthood]. , 2007, Ugeskrift for laeger.

[36]  E. Levitan,et al.  Invited commentary: dietary fiber, estradiol, and cholesterol. , 2011, American journal of epidemiology.

[37]  N. Cameron,et al.  The Human Growth Curve, Canalization and Catch-Up Growth , 2002 .

[38]  Cesar G Victora,et al.  Cohort Profile: The 2004 Pelotas (Brazil) Birth Cohort Study , 2010, International journal of epidemiology.

[39]  P. Raynor,et al.  Born in Bradford, a cohort study of babies born in Bradford, and their parents: Protocol for the recruitment phase , 2008, BMC public health.

[40]  D. Lawlor,et al.  Cohort Profile: The ‘Children of the 90s’—the index offspring of the Avon Longitudinal Study of Parents and Children , 2012, International journal of epidemiology.

[41]  Yoav Ben-Shlomo,et al.  SITAR--a useful instrument for growth curve analysis. , 2010, International journal of epidemiology.

[42]  D. Lawlor,et al.  Cohort Profile: The Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort , 2012, International journal of epidemiology.

[43]  Risto Lethonen.  Multilevel Statistical Models (3rd ed.) , 2005 .

[44]  J. Eriksson,et al.  Trajectories of growth among children who have coronary events as adults. , 2005, The New England journal of medicine.