Demographic survival analysis. Treatments that increase lifespan can do so by either reducing 'risk', which results in a shift of the survival curve downward and to the right, or by reducing senescence, which results in a decrease in the slope of the mortality curve. (a) The mortality rate of Australian prisoners of war (POWs) in Japanese concentration camps during World War II compared with that of civilians in Australia between 1944 and 1945 (adapted from Finch et al., Science 1990, 249:902-905). The mortality rate is the same for both populations, as illustrated by the curves having the same slope, but the POWs experienced a much greater intrinsic risk than their civilian counterparts, resulting in an upward shift of the mortality curve. (b,c) Dietary restriction (DR) and temperature extend lifespan to similar degrees, but demographic analysis shows that they do so by very different mechanisms in Drosophila melanogaster. Dietary restriction extends lifespan by shifting the survival curve (b), whereas temperature extends lifespan by changing the slope of the curve (c). Strikingly, shifting flies from high-food to low-food conditions results in an immediate reversal of the survival curve to low-food rates (broken line in (b)), implying that there is some inherent 'risk' associated with the high-food diet and that dietary restriction can protect against this.