Colestipol was begun at a dose of 5 g three times a day with meals and increased to 10 g three times a day after 10 days, unless side effects delayed the increase. Psyllium hydrophic mucilloid (Metamucil) was provided if dietary bran was insufficient to control constipation. Niacin was started at 125 mg twice a day and gradually increased to 500 mg four times a day (with meals and at bedtime) at one month and 1 g four times a day at two months. If the LDL cholesterol level did not fall below 3.1 mmol per liter (120 mg per deciliter) after three months, the dose of niacin was increased to 1.5 g (three tablets) four times a day, but no further.

Colestipol was given as described above. Lovastatin was begun at a dose of 20 mg twice a day (in the morning and at bedtime). If the LDL cholesterol level did not fall below 3.1 mmol per liter after three months, the dose of lovastatin was increased to 40 mg twice a day.

Patients assigned to conventional therapy (the control regimen) received placebos for colestipol and for lovastatin, given as described above, unless their base-line LDL cholesterol level exceeded the 90th percentile for age. We felt obliged to provide such patients (43 percent of the group) with colestipol instead of its placebo. For purposes of blinding, the lovastatin placebo dose for a patient assigned to conventional therapy was doubled each time the lovastatin dose was doubled for a patient assigned to receive lovastatin and colestipol.

Lovastatin was begun at a dose of 20 mg twice a day (in the morning and at bedtime). If the LDL cholesterol level did not fall below 3.1 mmol per liter after three months, the dose of lovastatin was increased to 40 mg twice a day.

At base line, the average percentage of stenosis caused by the worst lesion in each of nine proximal segments was 34 percent. On average, after 2 1/2 years of conventional therapy, this index of stenosis increased by 2.1 percentage points. By contrast, it decreased by 0.7 point during treatment with lovastatin and colestipol and by 0.9 with niacin and colestipol (P for trend <0.003). Thus, at the end of the study, on average, these nine lesions were almost 3 percentage points less severe among patients treated intensively rather than conventionally. This difference represents almost 1/10 of the amount of disease present at base line (34 percent stenosis). The minimum diameter, an alternative index of the severity of disease, in the nine proximal lesions averaged 1.91 mm for all patients. It decreased (worsened) by 0.050 mm with conventional therapy but increased (improved) by 0.012 mm with lovastatin and colestipol and by 0.035 with niacin and colestipol (P for trend <0.01).