Arm measurements as determinants of further survival in centenarians.

Being indicators of nutritional and functional status, anthropometric characters may have great prognostic significance for survival at extremely advanced ages. For ethical and practical reasons however it is advisable to use characters such as arm measurements easily measurable even in bedridden subjects. This study compares the influence of some upper arm measurements and of Body Mass Index (BMI) on survival of the 77 subjects aged 98 years and over (98+) recruited within the MALVA project, one of the first Italian population-based studies on extremely old people. Adopting methods for multiple imputation of missing values, Gompertz regression models adjusted for gender and age were estimated for each anthropometric character or combination of characters, i.e. BMI; mid-upper arm circumference (MUAC)+elbow breadth (EB)+triceps skinfold thickness (TSF); corrected arm muscle area (CAMA). Being underweight and having a low CAMA and a low MUAC/high EB were positively associated with an increased risk of death, while no significant association was found with the condition of being overweight/obese and the triceps skinfold thickness. When anthropometric variables were included in regression models along with covariates relating to nutritional and functional status, BMI and MUAC, but not CAMA, emerged as protective factors. It is suggested that MUAC can be recommended in evaluating the health status of extremely old people and that measuring EB may help to estimate the non-boney component of the arm.