Although the effect of age on body composition has been intensively discussed during the past 20 years, we do not have a uniform definition of sarcopenia. A suitable definition of low, lean body mass should be based on magnetic resonance imaging (MRI) estimates of muscle mass. Using recent MRI data of a population of 446 healthy free-living Caucasian volunteers (247 females, 199 males) age 18-78 years, a low skeletal muscle mass and sarcopenia were defined as a skeletal muscle mass >1 and >2 s.d. below the mean value obeserved in younger adults at age 18-39 years. The cutoffs for low muscle mass according to the skeletal muscle index (skeletal muscle mass/(height)(2)) or the appendicular skeletal muscle mass index (skeletal muscle mass of the limbs/(height)(2)) were 6.75 or 4.36 kg/m(2) for females and 8.67 or 5.54 kg/m(2) for males, respectively. On the basis of these cutoffs, prevalences of sarcopenia in the group of adults at >60 years are calculated to be 29% in females and 19.0% in males. Faced with different sarcopenic phenotypes (that is, sarcopenia related to frailty and osteopenia; sarcopenic obesity related to metabolic risks; cachexia related to wasting diseases), future definitions of sarcopenia should be extended to the relations between (i) muscle mass and adipose tissue and (ii) muscle mass and bone mass. Suitable cutoffs should be based on the associations between estimates of body compositions and metabolic risks (for axample, insulin resistance), inflammation and muscle function (that is, muscle strength).