Diagnostic test of predicted height model in Indonesian elderly: a study in an urban area

Aim In an anthropometric assessment, elderly are frequently unable to measure their height due to mobility and skeletal deformities. An alternative is to use a surrogate value of stature from arm span, knee height, and sitting height. The equations developed for predicting height in Indonesian elderly using these three predictors. The equations put in the nutritional assessment card (NSA) of older people. Before the card which is the first new technology in Indonesia will be applied in the community, it should be tested. The study aimed was to conduct diagnostic test of predicted height model in the card compared to actual height. Methods Model validation towards 400 healthy elderly conducted in Jakarta City with cross-sectional design. The study was the second validation test of the model besides Depok City representing semi urban area which was undertaken as the first study. Result Male elderly had higher mean age, height, weight, arm span, knee height, and sitting height as compared to female elderly. The highest correlation between knee height and standing height was similar in women (r = 0.80; P < 0.001) and men (r = 0.78; P < 0.001), and followed by arm span and sitting height. Knee height had the lowest difference with standing height in men (3.13 cm) and women (2.79 cm). Knee height had the biggest sensitivity (92.2%), and the highest specificity on sitting height (91.2%). Conclusion Stature prediction equation based on knee-height, arm span, and sitting height are applicable for nutritional status assessment in Indonesian elderly. (Med J Indones 2010; 19: 199-204)