STUDY DESIGN: A systematic search was performed using the MEDLINE, EMBASE, PsycINFO, CINAHL, and Web of Science databases to identify English-language articles published through June 2018. Articles were included if they were longitudinal studies in community-based populations, the primary exposure occurred during childhood, and the primary outcome was either a measure of subclinical CVD or a clinical CVD event occurring in adulthood. Two independent reviewers screened determined whether eligibility criteria were met.
RESULTS: There were 210 articles that met the predefined criteria. The greatest number of publications examined associations of clinical risk factors, including childhood adiposity, blood pressure, and cholesterol, with the development of adult CVD. Few studies examined childhood lifestyle factors including diet quality, physical activity, and tobacco exposure. Domains of risk beyond "traditional" cardiovascular risk factors, such as childhood psychosocial adversity, seemed to have strong published associations with the development of CVD.
CONCLUSIONS: Although the evidence was fairly consistent in direction and magnitude for exposures such as childhood adiposity, hypertension, and hyperlipidemia, significant gaps remain in the understanding of how childhood health and behaviors translate to the risk of adulthood CVD, particularly in lesser studied exposures like glycemic indicators, physical activity, diet quality, very early life course exposure, and population subgroups.
STUDY DESIGN: We assessed data from 6414 children aged 6-18 years, collected by the South East Asia Community Observatory. Child underweight, overweight, and obesity were expressed according to 3 internationally used BMI references: World Health Organization 2007, International Obesity Task Force 2012, and Centers for Disease Control and Prevention 2000. We assessed agreement in classification of anthropometric status among the references using Cohen's kappa statistic and estimated underweight, overweight, and obesity prevalence according to each reference using mixed effects Poisson regression.
RESULTS: There was poor to moderate agreement between references when classifying underweight, but generally good agreement when classifying overweight and obesity. Underweight, overweight, and obesity prevalence estimates generated using the 3 references were notably inconsistent. Overweight and obesity prevalence estimates were higher using the World Health Organization reference vs the other 2, and underweight prevalence was up to 8.5% higher and obesity prevalence was about 4% lower when using the International Obesity Task Force reference.
CONCLUSIONS: The choice of reference to express BMI may influence conclusions about child anthropometric status and malnutrition prevalence. This has implications regarding strategies for clinical management and public health interventions.