The prospective assessment of autonomic nerve function by pupillometry in adolescents with type 1 diabetes mellitus.

The study aimed to compare the longitudinal assessment of automatic nerve function by computerized infrared pupillometry and standard cardiovascular tests in adolescents with diabetes. Adolescents (n = 150) were assessed at two time points (T1 and T2). The median time interval between assessments was 1.5 (range 0.9-3) years. At T1 the median age was 14.5 (range 8.3-19.5) years and the median duration was 6.5 (range 1.1-16) years. The pupillary variables assessed included the resting pupil diameter, the maximum constriction velocity, and the reflex amplitude of constriction. Heart rate reflexes were assessed in response to deep breathing, the Valsalva manoeuvre, and on standing from a lying position (30/15 ratio). Between visits there was a significant decrease in maximum constriction velocity (6.0 mm s-1 vs 6.3 mm s-1, p = 0.0001) and resting pupil diameter (6.2 mm vs 6.3 mm, p = 0.001). At reassessment pupillary abnormalities increased from 32 (21%) to 45 (30%), with 17 (54%) of the initial abnormalities persisting. Adolescents with abnormally slow maximum constriction velocity compared to those with normal maximum constriction velocity had a higher glycated haemoglobin (HbA1c%) at T2 (p = 0.02) and between assessments (p = 0.01). Cardiovascular test abnormalities did not increase between visits and the persistence of initial abnormalities was low (21%). In summary, pupillometry appears a more sensitive test of automatic nerve dysfunction in adolescents with diabetes than assessment of cardiovascular reflexes.