Although each measurement is a different individual, the mean intertwin variation is zero except for the effects of age and any variables that may contribute to delayed sampling of the second twin, such as socioeconomic status, illness, or travel.