To validate the short-form of the MNA (MNA-SF) and the cut-off point of 31 cm for calf circumference (CC) in older people in Latin America.

A cross-sectional study was conducted with 5,722 community-dwelling older subjects (range: 60-102 years) in Latin America's five main cities: Sao Paulo (Brazil), Santiago (Chile), Havana (Cuba), Mexico DF (Mexico) and Montevideo (Uruguay). All participants underwent an interview, which included anthropometric measurements, completing the MNA and obtaining socio-demographic, nutrition and health information. The short-form of the MNA consists of only six questions from the original 18. It has two versions: one using body mass index (BMI) and the second using CC as a surrogate. Cohen's Kappa was calculated to assess the agreement between the MNA and the MNA-SF; diagnostic tests were performed, and Receiver-operating characteristic (ROC) curves were developed. Criterion-related validity was assessed in the Chilean sample.

Both version of the MNA-SF showed high sensitivity and specificity with the MNA, showing good accuracy (0.88), although higher values were estimated for malnutrition and risk of malnutrition in the total sample by sex. The cut-off point of 31 cm for CC showed high sensitivity (74.6-94.4%), specificity (72.6-100%) and good area under the ROC curve (0.87-0.95) when compared with BMI. There was good agreement between MNA and both version of MNA-SF for identifying persons with malnutrition or a risk of malnutrition in the five cities (Kappa coefficient: 0.6193-0.7852).

Both versions of the MNA-SF are population-screening tools of easy and fast application, with good accuracy for assessing malnutrition and risk of malnutrition in Latin American older people.