question of whether, over a 6.4-year period, participation inphysical activity by older adults reduces the rate of cognitivedecline after accounting for participation in cognitively stimu-lating activities. After statistical adjustment for cognitive ac-tivities, each additional physical activity hour per week wasassociated with a slower rate of cognitive decline. However, thisrelationship was no longer significant after adjusting for cogni-tive activities (but see Richards et al., 2003, for a report of arelationship between physical but not intellectual activities andlater-life cognition).At present, it is difficult to know which factors are most im-portant in moderating the inﬂuence of physical activity on later-life cognition and dementia. However, some possibilities thatmerit further study include the distinction between aerobic andnonaerobic physical activities; the utility of self-report versusmore objectively measured physical activities; the role ofphysical-activity duration, intensity, and frequency; the natureof the components of cognition that serve as the criterion vari-ables; the age of participants at initial and final assessment; andgenetic factors.A 6-year prospective observational study by Barnes, Yaffe,Satariano, and Tager (2003) included both self-report of physi-cal activity and objective (i.e., volume of expiratory oxygen[VO2] at peak exertion) measures of cardiorespiratory fitness in349 individuals over the age of 55. Interestingly, whereas asignificant inverse relationship was observed for the objectivefitness measures and cognitive decline, this was not the case forthe self-report activity measures. Although the explanation forthis dissociation cannot be unequivocally discerned in thesedata, it is conceivable that it is the aerobic aspects of thephysical activities—which are more reliably indexed by theobjective than self-report measures—that is more strongly re-lated to spared cognition than is those activities’ nonaerobicaspects. A number of investigators have also examined whetherpossession of e4 alleles on theapoegene moderates the rela-tionship between physical activity and cognition. Unfortunately,however, results obtained thus far are equivocal with regard towhether possession of an e4 allele has a positive effect, negativeeffect, or no effect on the relationship between physical activityand cognition (Abbott et al., 2004; Larson et al., 2006; Podewilset al., 2005; Rovio et al., 2005; Schuit, Feskens, Launer, &Kromhout, 2001). This ambiguity could be the result of anumber of different factors that may interact with genotype (suchas those factors described above) in determining the relation-ship between physical activity and cognition. As will be dis-cussed below, a number of other genes also produce proteinsrelevant to neurotransmitter systems and nerve-growth factorsthat appear to play an important role in the relationship betweenphysical activity and cognition. Such genes will be important tostudy in the future.