Every 6 months, beginning at the 18-month or 24-month visit, participants were administered a structured questionnaire in which they were asked whether, since their last clinic visit, they had experienced any skin rashes (and the percentage of time any skin rashes were present), any bruises on their skin more often than in previous years, and more dry skin than usual (a “dummy” question used to identify non-steroid-related positive responders). If they reported bruises they were queried as to the percentage of time the bruising had been present, the number of bruises that were usually present at one time, the usual location of the bruises, and the occurrence of trauma as a possible cause of the bruising. They were also questioned with regard to exposure to the sun and the wearing of sunscreen. Every 3 months, beginning at the 3-month visit, participants reported changes in their skin since the last clinic visit, including skin rashes, easy bruising, acne, slow healing of cuts or sores, or any other skin changes. Furthermore, they were asked to grade the severity of their skin changes, as follows: 0, not present; 1, mild (present and noticeable or annoying but does not interfere with normal activities); 2, moderate (interferes with normal activities or requires an outpatient visit to a physician, clinic, or emergency department); or 3, severe (prevents normal activities, requires major medical/surgical intervention or hospitalization, or is life-threatening). The participants were followed from randomization to a common ending date (April 30, 1999) 4.5 years after the initiation of the trial.
LHS II randomized 1,116 participants. Data from the structured skin questionnaire were available on 1,086 participants. All analyses, including baseline characteristics, are limited to this slightly smaller cohort.