Cluster analysis of obesity-related parameters of Hungarian children


Obesity is a fast-growing endemic with an already alarmingly high prevalence and serious potential health consequences; both underlining its risk. Currently existing indicators of obesity and its risk are either easily computable, but of a smaller predictive value (e.g. BMI) or accurately predict risk, but are unfit for screening of wider populations (e.g. CT, DXA). 34 healthy Hungarian children (aged 14-16 years) were medically examined to collect parameters related to obesity: anthropometric parameters, body composition data provided by bioelectric impedance analysis, laboratory blood tests and medical history. Variables derived from prepared (factor analyzed, amongst others) anthropometric data and body composition were then clustered, revealing the presence of two distinct clusters. Sexes are fortunately independently distributed among clusters. Minuscule, but statistically significant differences in the laboratory tests were found between the clusters. These – especially the physiologically consistent ones – need further attention.

BUDAMED ‘08 : 14. Magyar Orvostechnikai Konferencia és 5. Magyar Klinikai-mérnöki Konferencia