Retinopathy in patients with diabetic ophthalmoplegia

Abstract

Purpose: To review the clinical characteristics, prevalence, and severity of retinopathy in diabetics with cranial nerve (CN) 3, 4, and/or 6 palsies, and to determine the relationship between type and duration of diabetes mellitus (DM), presence of retinopathy, and occurrence of CN palsy.

Design: Retrospective, comparative cohort study.

Participants: Chart reviews of 2229 patients with CN 3, 4, and/or 6 palsies were performed at the Bascom Palmer Eye Institute (BPEI) from January 1991 through December 1997 and at the Dean A. McGee Eye Institute (DMEI) from January 1994 through July 2001. A total of 306 patients qualified for the study group. The Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR) was used as a control.

Methods: Demographic and clinical data were extracted to determine characteristics of patients with diabetic ophthalmoplegia. The subsets of data regarding type of DM and level of diabetic retinopathy in the study population were compared with the WESDR control data for statistical analysis.

Main outcome measures: The prevalence of diabetic retinopathy in patients with diabetic ophthalmoplegia.

Results: Of 2229 patients at both institutions with ocular motor CN palsy, 306 (13.7%) were associated with DM. The frequency of CN involvement was 6 (50.0%), 3 (43.3%), and 4 (6.7%). There was a total of 12 patients (3.9%) with consecutive palsies and 8 patients (2.6%) with simultaneous palsies (5 unilateral and 3 bilateral). At both institutions, the prevalence of retinopathy controlling for duration of DM was lower in both insulin-dependent DM (IDDM) and non-IDDM (NIDDM) type II diabetics as compared with controls (BPEI, P = 0.009 and P = 0.005; DMEI, P = 0.004 and P = 0.29). When data from both locations were combined, the difference was even more significant (IDDM, P = 0.001 and NIDDM, P = 0.006). There were no significant differences between the two institutions in gender, type or duration of DM, age at presentation, or frequency of CN involvement.

Conclusions: Diabetic ophthalmoplegia most commonly involves CN 3 and 6, with relative sparing of CN 4. Multiple cranial nerves are affected simultaneously in 2.6% of cases, and consecutive palsies occurred in 3.9% of cases. Type II diabetics with ocular motor CN palsy have significantly less diabetic retinopathy than do controls. This may imply a different pathophysiologic mechanism for these two microvascular complications of DM.