As Wei and colleagues point out, their study was a baseline analysis of self-selected, mostly white, men. Persons taking insulin, whose disease is presumably at a more advanced stage, were not included. The least physically fit patients (those in the lowest 20% of the fitness categories) also had the most preexisting risk factors and history of cardiovascular disease. Finally, the study did not examine the benefits, if any, of initiating a program of physical activity.

Notwithstanding these limitations, the study is welcomed because it specifically addresses people with diabetes. In many epidemiologic studies and clinical trials of cardiovascular disease, persons with diabetes have been systematically excluded to simplify analysis of the results, leaving us with the problem of how to apply their results to persons with diabetes. Logic suggested that diabetes would only enhance cardiovascular risk, but until recently, evidence was lacking.

Fortunately, subanalysis of existing trials and trials specifically directed at persons with diabetes are changing this situation. Subanalysis has shown with certainty that hypertension, hyperlipidemia, and smoking are synergistic risk factors in persons with diabetes and that treatment of these conditions will benefit patients with diabetes at least as much as those without it (3). Similarly, we can be sure that aspirin prophylaxis is beneficial in diabetes (4). Finally, the United Kingdom Prospective Diabetes Study has clearly shown that aggressive treatment of hypertension is markedly beneficial in preventing cardiovascular events and death in persons with type 2 diabetes (5). We can now aggressively treat these risk factors and hyperglycemia in patients with diabetes and have confidence that the patients are benefiting.

The value of an exercise program in the prevention of type 2 diabetes in high-risk persons has been suggested. In the Malmö Preventive Trial (6), a prospective diet and exercise program reduced conversion from abnormal glucose tolerance to frank diabetes by one third. Wei and associates have presented similar data from their cohort (7). Both of these studies included only persons of European ancestry. However, James and colleagues' study (8) of the risk for type 2 diabetes among African Americans also demonstrated a marked reduction of risk in physically active men. Thus, a program of increased physical exercise should help prevent progression to frank diabetes in high-risk persons: those with a family history of diabetes, history of gestational diabetes, and obesity, especially in combination with high-risk ethnicity or abnormal fasting or postprandial glucose levels. This question is currently being addressed in the multicenter Diabetes Prevention Program (9).

The Malmö Preventive Trial also supports the benefit of exercise in type 2 diabetes (6). Diabetes control was significantly improved in the experimental participants who developed diabetes during the study. Epidemiologic data from persons without diabetes also support the benefit of physical activity in preventing cardiovascular events. Particularly interesting in this regard is Bijnen and associates' study (10), which demonstrated that physical activity prevented cardiovascular death in elderly men; this result suggested that a program of increasing physical activity would benefit patients with type 2 diabetes, even those who are older (10). These studies also suggest that the amount of exercise needed to achieve a cardioprotective benefit is modest. James and associates found that modest exercise, such as walking three times a week, achieved the desired benefit (8). The exercise program of the Diabetes Prevention Program aims at weekly expenditure of 2000 calories (9).

Internists are on the front line in the care of persons with diabetes. Compelling evidence indicates that the long-term complications of diabetes can be prevented, reduced, or delayed by a comprehensive program of aggressive treatment of hyperglycemia and associated cardiovascular risk factors. Wei and associates' findings (1) remind us that a comprehensive program of exercise should be an integral part of diabetes care.

Moreover, one of the most powerful predictors of the development of diabetes in genetically susceptible persons is weight gain in adulthood (11). Obesity is increasing dramatically throughout the developed world, and half of the adults in the United States are overweight (12). The average caloric intake has not changed over the past two decades in the United States; we can only conclude that this increase in obesity is due to a reduction in physical activity (13, 14). Thus, increasing physical activity is the key to the prevention of weight gain in adults.

Children are not immune to this trend. Type 2 diabetes among children in ethnic minority groups is increasing at an alarming rate (15). Physicians who care for adolescents and young adults must now be alert to the possibility of type 2 diabetes and to the health risks of childhood obesity (16). They must also work with pediatric and family medicine associates to increase awareness of the health risks of a sedentary lifestyle, particularly among nonwhite persons. Diabetes is a complex and challenging disease. Our task today is to establish an environment that will both enhance patients' lives and prevent complications. To accomplish this, physicians must work with professional organizations to enhance public awareness of the long-term health risks of overweight and the benefits of a sustained exercise program (17).

Physicians also need to be more honest. Americans work harder and longer than ever, physicians included. General admonishments to get more exercise are as unlikely to work as general advice to eat less or to stop smoking. Specific programs need to be prescribed, and follow-up is essential. This will require that we be convinced and that we convince our patients that the trade-off between time spent in exercise and the other demands on their time is worthwhile. In "The Will to Believe," William James states that at some point we have to accept that the evidence will never be perfect and move forward (18). The data supporting the health benefits of physical activity are overwhelming. We must move from demanding more data to learning how to apply what we already know.