Anthropometric indices of nutritional status in Croatian oldest-olds: new equations to predict height and weight

Anthropometry is an essential tool in the geriatric nutritional assessment. Measuring height and weight can be quite problematic in the old age so they can be estimated using the WHO equations. The purpose of this study was to: (a) show anthropometric indices of nutritional status of the ambulatory institutionalized oldest old in Zagreb, Croatia; (b) examine the adequacy of WHO equations for predicting height and weight of oldest old persons; (c) develop population-specific equations for estimating height and weight. The sample of this cross-sectional study comprised 314 examinees (F/M= 234/80, 85-101 years). The classic anthropometric parameters (height, weight, extremities' circumferences, knee height and skinfolds) were measured. The adequacy of WHO equations for predicting height and weight in Croatian oldest old was assessed using the paired t-test. Multiple linear regression analysis was performed to derive population-specific equations for predicting height and weight. The sexual dimorphism was found in several anthropometric indices; men were wile); had larger weight and knee height, while women had larger triceps skinfold thickness. In women, most indices declined with age and in men the negative relation of knee height with age indicated a secular trend. The WHO equations for elderly acceptably predicted nothing but height in oldest old women. The presently developed equations, using the same WHO's predictor variables, proved to be more adequate for the studied population. The equations developed for estimating height and weight of non-ambulatory oldest old Croatians proved to be more appropriate than WHO equations and are therefore proposed for use in the everyday practice.