Pupil size in insulin-dependent diabetes. Relationship to duration, metabolic control, and long-term manifestations.

The pupil area was measured in complete darkness by infrared TV-videopupillography in 109 insulin-dependent diabetics, aged 25-43 yr, diabetes duration 0-35 yr, and 39 control subjects, aged 26-41 yr. In darkness the pupil was 19.6% +/- 4.2 (SEM) smaller in diabetics than in controls (2 P less than 10(-5). There was an inverse relationship between diabetes duration and pupil area (Kendall's coefficient of correlation, tau = -0.33, 2 P less than 10(-4). There was also an inverse correlation between pupil size and vibratory perception threshold (tau = 0.32, 2 P less than 10(-3). Long-term diabetics (duration greater than or equal to 15 yr) with proliferative retinopathy had a 28.4% +/- 8.1 (SEM) smaller pupil than those without (2 P = less than 10(-3). Likewise, long-term diabetics with nephropathy had a 19.8% +/- 9.1 smaller pupil than those without nephropathy (2 P = 0.035). In the long-term diabetics there was an inverse relationship between the level of blood glucose throughout the years and pupil area (tau = -0.49, 2 P less than 10(-3). Also, high blood glucose levels throughout the years were correlated to severity of retinopathy (tau = 0.43, 2 P less than 10(-3) and nephropathy (tau = 0.30, 2 P = 0.024). There was no correlation between biomicroscopic changes in the iris and the diminished pupil area. Pupil area in light was measured in 85 patients and 31 controls. In continuous light only the long-term diabetics had a smaller pupil size than the controls. Both the absolute and relative change in pupil size from darkness to light was less in the diabetic group. Measuring the pupil size in darkness is a simple, noninvasive and reproducible method that may yield information about autonomic nervous involvement in diabetes.