PURPOSE:

To assess the relationship between diabetic polyneuropathy (DPN) and the risk of diabetic retinopathy (DR).

METHODS:

From 1997 to 2010, we identified 5031 newly diagnosed DPN patients and 20 124 controls matched for sex, age, and index year. Cox proportional hazards regression analyses were used to estimate the adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) of DR between the DPN patients and the non-DPN group. The adjusted hazard ratio was calculated and adjusted for age, sex, duration of diabetes and comorbidities of hypertension, cardiovascular disease and diabetic nephropathy.

RESULTS:

The incidence rate of DR was 5.87-fold higher in the DPN patients than in the non-DPN group (44.0 vs. 7.22 per 1000 person-years), with an adjusted HR of 5.41(95% CI = 4.92-5.94). The DPN-to-non-DPN DR incidence rate ratio decreased with age (adjusted HR = 6.63 for subgroup younger than 65 years and adjusted HR = 3.91 for subgroup aged 65 years or older). Compared with the non-DPN group, the DPN patients had a 5.63-fold risk of non-proliferative DR (adjusted HR = 5.63, 95% CI = 5.11-6.21) and a 3.67-fold risk of proliferative DR (adjusted HR = 3.67, 95% CI = 2.57-5.23).

CONCLUSION:

The patients with DPN had an increased risk of developing DR and advanced DR compared with the non-DPN group, particularly among the subgroup aged younger than 65 years.