Comparison of measurement methods for capillary basement membrane thickness. Collapsed-ellipse technique.

Both the grid method of Siperstein et al. (tSIP) and the minimum-points method of Williamson et al. (tWIL) for measurement of capillary basement membrane thickness are inaccurate for assessing mean true membrane thickness of a give section (tTRUE) for various reasons, including errors in selectivity, sensitivity, and geometry. In general, it is agreed that tSIP greater than tTRUE greater than tWIL, but estimates of tTRUE beyond this have not been made. In this study, a collapsed-ellipse method for approximating tTRUE is presented that measures thickness by areas (tEA). One hundred-twenty capillaries from the forearm skin of 12 diabetic subjects and 12 age-matched controls were measured to examine these concepts. We found that, whereas tWIL was up to 63% below tTRUE, tEA was less than 30% too low. Although tSIP, tWIL, and tEA did not distinguish between diabetic and normal subjects, tEA and tWIL measurements had highly predictable and small errors, and tSIP had unpredictable ranges of error, especially when tSIP was low.