Sir,

The study by Garcia-Arumi et al1 was of engaging interest. The authors conclude that erythropoietin (EPO) has a neuroprotective role in diabetic macular oedema (DME). In addition to the limitations of this study, lucidly highlighted by the authors, we demur on some of the methodological constraints of this study.

Diabetes being a chronic multisystem disease with end-organ damage, it would be natural to expect raised serum EPO, especially with coexisting nephropathy and anaemia.2 Earlier reports show that the correction of anaemia (and also supplementation of EPO) decrease the effects of retinopathy with structural improvement, possibly through the improved oxygenation of the macula.2, 3 It would be fallacious then to compare the acute imbalance of retinal vein occlusion (RVO) cases, with DME. Also, EPO level in the DME group was found to be 430 mU/ml (41–3000), and in the RVO group, it was 91.3 mU/ml (30–417). It would have been interesting to note the near normal levels of EPO in some of the DME cases despite the presence of DME. This would probably imply the role of other factors in DME. In this regard, the authors may have overlooked important factors for the production of EPO, such as nephropathy status, Hb levels, and coexisting vascular disease.3

The authors have shown ingenuity in their study; however, we reckon that it would have been better served with a comprehensive outlook of the risk factors being studied.