Child growth is characterised by increases in height, and increases in maturational status. Functional data analysis provides a tool to separate these two sources of variation (registration) and differentiates between the variation in maturational tempo (temporal, or “phase” variation) and the variation in height (amplitude variation). We extended this concept by combining Principal Component Analysis (PCA) and Maximum Likelihood Principle. The combination of PCA and Maximum Likelihood Principle provides a new, powerful and automatic tool for growth modelling that includes estimates of future growth, adult stature and developmental tempo. Preliminary results indicate that this approach can be used for automatised screening purposes. The website www.willi-will-wachsen.com may be used for automatic growth modelling.
In 1983, an attractive technique was developed for accurate non-invasive determinations of the lower leg length (knemometry). This device is easier to handle, it has a measurement error between 0.09 mm (Valk et al. 1983) and 0.16 mm (Hermanussen et al. 1988a), and is observer independent (Wales and Milner 1987). Initially, intervals of three weeks were recommended for short-term growth measurements, but it soon became obvious that intervals of one week were technically feasible, and provided evidence for non-linear progress of growth at short-term (Hermanussen et al. 1988b, Wit et al. 1987). Besides the original device, similar techniques for lower leg measurements were inaugurated, a modified Danish knemometer by Michaelsen et al. (1991), a Polish knemometer (Hulanicka et al. 1999), a “knee height measuring device” by Cronk et al. (1989), a portable knemometer by Davies et al. (1996) for measuring sitting individuals, a miniaturised portable knemometer (1.5 kg) for field work (Hermanussen, unpublished), and an Italian device (Benso, personal communication 1992). Mikro-knemometry is available for lower leg length measurements in rodents (Hermanussen et al. 1992, Hermanussen et al. 1995), and a particularly small version (mini-knemometer) for the measurement of infants, or neonates and premature infants within their incubators (Hermanussen, Seele 1997, Kaempf et al. 1998). The latter devices operate with a precision that reduces the amount of variance due to measurement error, to five percent of the total variance of 24 hour measurement series, and thus visualise length changes within periods of less than one day.
National standards for growth in height and length belong to the traditional inventory of the paediatric practice. Yet, national standards are world-wide only available for few major ethnic groups. Most children and adolescents from the smaller countries, and many ethnic minorities, particularly the children from developing countries, must still be referred to "international" usually US-based growth references. In view of this dilemma, we introduced a method to generate synthetic reference standards for height, weight and BMI (Hermanussen and Burmeister 1999). Snythetic references are based on data obtained at certain design age groups (at birth, and at the ages of 2,6,14, and 18 years in males; respectively at birth, and at the ages of 2,6,14, and 18 years in females) and generate age-specific linear regression coefficients. These coeffients are used to calculate mean values, standard deviations and centiles for height, weight and BMI of all age groups from birth to maturity. Synthetic growth charts are used in several countries, and have recently been generated and published for German born Turkish children and adolescents (Redlefsen et al. 2007).