Distribution-free estimation of age-related centiles

Healy, M. J. R.; Rasbash, Jon; Yang, Min

doi:10.1080/03014468800009421

Cited by 161 publications

(110 citation statements)

References 2 publications

Supporting

Mentioning

108

Contrasting

Unclassified

Order By: Relevance

“…The GROSTAT method and its appropriateness to our data was recently described by Hosseini et al 17 A grid test, which compares the observed and expected number of observations between each of the ®tted centiles, was used to assess goodness of ®t of the smooth centiles to the data. 14 An alternative test is to use the model for the smooth centiles to compute the Normal Z-scores corresponding to each of the original observations. If the curves ®t the data, these Z-scores will be Normally distributed.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Standardized percentile curves of body mass index of Iranian children compared to the US population reference

et al. 1999

View full text Add to dashboard Cite

OBJECTIVE: To present standardized percentile curves of body mass index (BMI) for Iranian children, and compare these to the US population reference. SUBJECTS: 1599 boys and 1702 girls aged 2 ± 18 y living in urban Tehran as a part of a random cluster sample survey of 1 in 1000 families throughout Iran. MEASUREMENTS: Heights (cm) and weights (kg) were collected by trained health staff. RESULTS: Standardized BMI reference curves for Iranian boys and girls were constructed. The curves are shown to ®t the data well. The development pattern of BMI for boys and girls are compared. CONCLUSIONS: The major differences observed between Iranian and the US BMI charts underline the need for population-speci®c reference data. For children over six years the 5th and 95th percentiles of our data may be used provisionally as cut-off points for de®ning thinness and obesity for Iranian children and adolescents.

show abstract

Section: Discussionmentioning

confidence: 99%

“…BMI charts for boys and girls in urban Tehran were constructed using Healy's method 14 as extended by Pan et al, 15 and implemented in GROSTAT II. 16 This method produces smooth centiles which are close to the data, and constrained to accord with neighbouring centiles; observations can be converted to Z-scores.…”

Section: Discussionmentioning

confidence: 99%

Standardized percentile curves of body mass index of Iranian children compared to the US population reference

et al. 1999

View full text Add to dashboard Cite

show abstract

“…To draw smooth growth charts, we resorted to the EMGF (extended mechanistic growth function) method (Cortinovis & Milani, 2000), an extension of the Healy et al approach (Healy et al, 1988) to the case of non-linear functions.…”

Section: Tracing Growth Chartsmentioning

confidence: 99%

Italian cross-sectional growth charts for height, weight and BMI (6–20 y)

et al. 2002

View full text Add to dashboard Cite

Objective To trace growth charts for height, weight and body mass index (BMI) that apply to the whole Italian population. Different charts were drawn for central-north and south Italy since children in central-north regions are known to be taller and leaner. Design: Cross-sectional study. Setting: A sample of schoolchildren covering 16 of the 20 Italian regions, with data collected between 1994 and 2000. Subjects: A total of 27 421 girls and 27 374 boys, aged 6 -20 y. Methods: Height and weight were measured using portable Harpenden stadiometers and properly calibrated scales, respectively. SIEDP references are presented both as centiles and as LMS curves for the calculation of standard deviation scores. According to International Obesity Task Force, SIEDP charts for BMI include the limits for overweight and obesity, ie the centiles having, at 18 y of age, the value of 25 and 30 kg=m 2 , respectively. Results: The comparison between SIEDP and Tanner et al's charts for height, still in use among most Italian paediatricians, shows that before puberty Italian children are 2 -4 cm taller than their English peers. Because of these differences, Tanner's charts fail to detect, when applied to Italian children, 50 -90% of short children aged 6 -11 y, ie with stature below the 3rd centile of their reference population. Rolland-Cachera et al's centiles for BMI are lower than those of SIEDP standards, mainly during adolescence (up to 6.6 kg=m 2 for the 97th centile), and apply poorly to Italian children. The prevalence of overweight is 27 (boys) and 19% (girls) in south Italy vs 17 (boys) and 10% (girls) in central-north Italy. Conclusions: These references intend to supply Italian paediatricians with a tool that avoids the use of outdated or inadequate charts, and thus should be suitable for monitoring their patients' growth. Sponsorship: Italian Society for Pediatric Endocrinology and Diabetes (SIEDP).

show abstract

“…He concluded that two of them (the LMS method of Cole (1989) and the log transform with offset of Rona and Altman (1977)) performed equally well, while the polynomial method of Healy, Rasbash and Yang (1988) was less good.…”

mentioning

confidence: 99%