Premium
Anthropometric and Laboratory Markers of Nutritional Status in a Large Sample of Older Australians: the ALSA study
Author(s) -
Whitehead C.H.,
Giles L.C.,
Andrews G.R.,
Finucane P.
Publication year - 2000
Publication title -
australasian journal on ageing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.63
H-Index - 34
eISSN - 1741-6612
pISSN - 1440-6381
DOI - 10.1111/j.1741-6612.2000.tb00150.x
Subject(s) - anthropometry , grip strength , malnutrition , sarcopenia , gerontology , demography , medicine , body mass index , physical therapy , sociology
Objectives : bL To determine which combination of available measures of nutritional status can adequately describe the nutritional status of a large sample of older patients. bL To describe the prevalence of malnutrition in this large sample of older community dwelling Australians and provide a set of values that could be used as reference data for future studies. Design : A descriptive study based on the baseline cohort of the Australian Longitudinal Study of Ageing (ALSA) using a principal component analysis and correlations between the variables. The prevalence of malnutrition is estimated using currently available definitions of malnutrition. Method : 1,404 community living older people (aged 70–103 yrs, mean 77 yrs) underwent measurement of body mass index (BMI), relaxed arm girth, arm muscle area, triceps, abdominal and supraspinale skinfold thicknesses and grip strength together with albumin and total lymphocyte count. All pairwise correlations were calculated between the variables separately for each gender and the relationship between all the variables along with age further explored by principal component analyses. Results : In both sexes, strong correlations were noted amongst BMI, skinfold thicknesses and arm girth (r values 0.43 to 0.81). Grip strength, albumin and lymphocyte count did not appear to be notably correlated with each other or with any of the other variables (r values 0.00 to 0.37). The principal component analyses indicated that the majority of the variation in the data could be explained in as few as three dimensions. The prevalence of malnutrition varied using available definitions from 0.7% to 17.9%. Conclusions : We conclude that the strength of correlations between variables is such that nutritional status could be measured with as few as three variables. BMI, albumin and possibly lymphocyte count seem an ideal combination. The prevalence of malnutrition depends greatly on which definition is chosen and the clinically significant level of malnutrition is unclear without more longitudinal prognostic data.