There are large differences in life span between some species of animals. The taxonomic stratification of longevity can be seen among the mammals. Primates, generally, are the longest-lived group, although some small prosimians and New World monkeys have relatively short life spans. The murid (mouselike) rodents are short-lived; the sciurid (squirrel-like) rodents, however, can reach ages two to three times longer than the murids.

Three traits have independent correlations with life span: brain weight, body weight, and resting metabolic rate. The dependence of life span on these traits can be expressed in the form of an equation: L = 5.5E ^0.54S ^−0.34M ^−0.42. Mammalian life span (L) in months relates to brain weight (E) and body weight (S) in grams and to metabolic rate (M) in calories per gram per hour. The positive exponent for E (0.54) indicates that longevity of mammals has a strong positive association with brain size, independent of body size or metabolic rate. The negative coefficient for metabolic rate implies that life span decreases as the rate of living increases, if brain and body weight are held constant. The negative partial coefficient for body weight indicates that the tendency for large animals to be longer-lived results not from body size but rather from the high positive correlation of body weight with brain weight and its negative correlation with metabolic rate. The same kind of relation of L to E, S, and M holds for birds, but there is a tendency for birds to be longer-lived than mammals of comparable brain and body size despite their higher body temperatures and metabolic rates. The larger reptiles have life spans exceeding those of mammals of comparable size, but their rates of metabolism are about 10 times lower, so that their total lifetime energy expenditures are lower than those for mammals. The more highly cephalized animals (i.e., those with higher brain weight), especially the primates, have greater lifetime energy outputs. The total lifetime energy output per gram of tissue is about 1,200,000 calories for humans and 400,000 calories for domestic animals such as cats and dogs.

The above relations hold for the homeothermic mammals, those with nearly constant body temperature. The heterothermic mammals, which are able to enter daily torpor, or seasonal hibernation, thereby reduce their metabolic rates more than 10-fold. The insectivorous bats of temperate latitudes are the most dramatic example: although they have life spans in excess of 20 years, almost 80 percent of that time is spent in deep torpor, and, as a result, their lifetime energy expenditures are no greater than are those of other small mammals.

The longevities of arthropod species extend from a few days to several decades. The extremely short-lived insects have a brief single reproductive phase; the longer-lived spiders and crustaceans are iteroparous, with annual reproductive cycles.

The inheritance of longevity

The inheritance of longevity in animal populations such as fruit flies and mice is determined by comparing the life tables of numerous inbred populations and some of their hybrids. The longevity of sample populations has been measured for more than 40 inbred strains of mice. Two experiments concur in finding that about 30 percent of longevity variation in female mice is genetically determined, whereas the heritability in male mice is about 20 percent. These values are comparable to the heritabilities of some physiological performances in domestic animals, such as lifetime egg or milk production. The slope of the Gompertz function line indicates the rate of actuarial aging. The differences in longevity between species are the result primarily of differences in the rate of aging and are therefore expressed in differences in the slope of the Gompertz function.

Comparison of life tables between mouse strains of a single species indicates that the strain differences result primarily from differences in age-independent hardiness factors. If strains differ in hardiness, the less hardy have death rates higher by a constant multiple at all ages, as shown by the parallel Gompertz functions. It is frequently found that the first-generation (F₁) hybrids of two inbred strains live longer than either parent. There has been no direct comparison of hybrid and inbred mice with regard to the rates of their biochemical aging processes, but life table comparisons indicate that hybrid vigour (heterosis) is an increase of age-independent vigour and not a decrease in the rate of aging.

Much of the variation in survival time between mouse strains is attributable to differences in inherited susceptibility to specific diseases. An important task of gerontology is to determine the extent of such genetic influences on aging.

The inheritance of longevity in humans is more difficult to investigate because length of life is influenced by socioeconomic and other environmental factors that generate spurious correlations between close relatives. A number of studies have been published, most of them pointing to some degree of heritability with regard to length of life or susceptibility to major diseases, such as cancer and heart disease. Although there is disagreement about the degree of heritability of longevity in humans, the evidence for genetic transmission of susceptibility to coronary heart disease and related diseases is strong, as is the evidence that monozygotic (genetically identical) twins tend to have more similar life spans than do like-sex dizygotic (genetically different, fraternal) twins.