Abstract
Purpose:
To quantitatively determine the degree to which neuroretinal function of the adult retina is affected non-uniformly by type 2 diabetes (T2DM) in patients without clinical signs of diabetic retinopathy, and in patients with nonproliferative diabetic retinopathy (NPDR). In addition, to develop and utilize an index of non-uniformity of function.
Methods:
Photopic multifocal electroretinograms (mfERG) were recorded from 103 retinal locations of the right eyes of 32 T2DM patients without retinopathy (NoRet group), 30 T2DM patients with NPDR (NPDR group), and 47 healthy control subjects (Cntl group). To quantify non-uniformity of neuroretinal function, we calculated the following for local mfERG implicit times (IT) and N1-P1 amplitudes (AMP) measured using the template scaling method (Hood & Li, 1997). First, the Cntl group’s 103 local IT and AMP averages were calculated. Second, each subject’s local IT and AMP values were divided by the appropriate local Cntl average and converted to decibels. Third, the standard deviations of these values were computed within each eye to generate an “index of non-uniformity” (ION) for IT and AMP. The ION values of the subject groups were compared using two-tailed t-tests. Sensitivities and specificities were established by analyzing receiver operating characteristics.
Results:
The NoRet group’s IT ION (mean±2SEM: 0.081±0.021 dB) was larger than the Cntl group’s (0.058±0.005 dB; P<0.05). The NoRet group’s AMP ION (0.633±0.147 dB) was also larger than the Cntl group’s (0.471±0.047 dB; P<0.05). As expected, the NPDR group’s IT ION (0.103±0.026 dB) and AMP ION (0.710±0.139 dB) were also larger than those of the Cntl group (P<0.005 for both). However, IONs of the NoRet and NPDR groups were not significantly different. At 89.4% specificity, 31.3% of the NoRet eyes and 63.3% of the NPDR eyes had abnormal IT IONs. Abnormal AMP IONs occurred in 31.3% of the NoRet eyes and 43.3% of the NPDR eyes at 89.4% specificity.
Conclusions:
Type 2 diabetes does not affect neuroretinal function uniformly across the central retina, even in eyes without clinical signs of diabetic retinopathy. This suggests that averaging mfERG waveforms over different retinal locations, especially relatively large areas, is not an appropriate approach for sensitively detecting altered neuroretinal function. The ION measure could prove to be useful in the study of additional retinal diseases.
Keywords: 499 diabetic retinopathy •
509 electroretinography: clinical •
498 diabetes