While the above models provide useful and interesting data, few analyses have utilized actual population and cost data. Pratt et al performed a cross-sectional stratified analysis of the 1987 National Medical Expenditures Survey. For those without physical limitations, the average annual direct medical costs were $1,019 for those who were regularly physically active and $1,349 for those who reported being inactive. Medical care use was also lower for physically active people than for inactive people. The mean net annual benefit of physical activity was $330 per person. The authors estimated that increasing participation in regular physical activity among the > 88 million inactive Americans over the age of15 years could reduce annual national medical costs by as much as $77 billion. Similarly, Luepker et al reported an analysis of 5-year Medicare costs in 3,393 women and 2,495 men aged > 65 years who were categorized as to activity level based on the National Heart, Lung, and Blood Institute Cardiovascular Health Study. When placed in quar-tiles based on the number of kilocalories expended per week, healthy nonsedentary seniors in the two least active quartiles averaged $4,300 less in 5-year costs than did healthy sedentary seniors. Savings were improved to $6,180 in the most active quartile. These studies were based on a questionnaire of exercise habits rather than on the results of an exercise test.