We read with interest the study by Furukawa et al. [1]. The authors reported a significant and independent association of subclinical hypothyroidism (SCH) with nephropathy (defined as albumin:creatinine ratio ≥300 mg/g creatinine) in patients with diabetes mellitus (DM). In particular, those with diabetic nephropathy had more than two-fold higher prevalence of SCH than those without. Estimated glomerular filtration rate (eGFR) was lower in patients with SCH compared with euthyroid subjects.Some comments might be of interest. Although many studies have shown an association between overt -in most cases-hypothyroidism and renal dysfunction, what appears to be of outmost importance is whether restoration of thyroid function with L-thyroxine could be beneficial and to what extent, especially in patients with DM. It is well-known that SCH affects many cardiovascular parameters involved in the pathogenesis of diabetic nephropathy, such as lipids and blood pressure [2]. Only two studies have reported observations on diabetic nephropathy in correlation with SCH [3,4]. The first showed that 18 months of L-thyroxine replacement therapy in SCH women significantly increased eGFR to the levels of euthyroid controls (from 69 to 73 mL/min). However, whether this change is a mere reflection of a reversible and non-clinically significant haemodynamic effect of SCH on renal clearance, rather than a true morbidity, remains to be elucidated [3]. The other study included patients with chronic kidney disease (CKD) (32% men, 39% DM) and showed that L-thyroxine significantly attenuated the rate of decline in eGFR after 12 and 24 months, both in stage 2 and 3-4 of CKD. However, it had no effect on proteinuria [4]. These studies, as well as those in overt hypothyroidism, included patients with different degrees of renal function, were of small size and short duration and used different criteria of renal dysfunction assessment.Important issues to be taken into account when considering the benefit (if any) in patients with SCH and nephropathy are baseline eGFR or proteinuria levels, as well as, TSH concentrations, before L-thyroxine substitution. The baseline mean TSH levels in the aforementioned studies were 13.2 [3] and 8.8 mIU/L [4], respectively. It is not known whether L-thyroxine would be beneficial in lower TSH levels (e.g. <7 mIU/L). In a recent prospective study, we did not find any significant effect of L-thyroxine replacement therapy on lipid profile, coagulation markers, systemic inflammation and glucose homoeostasis in SCH patients (mean baseline TSH: 6.8 mIU/L). It only reduced blood pressure, especially in those with TSH >7 mIU/L [5]. The duration of L-thyroxine therapy, DM and nephropathy and the multi-factorial origin of the latter may also play a critical role in the effect of L-thyroxine on renal function.In conclusion, it would be prudent to assess thyroid function in patients with diabetic nephropathy. In cases of SCH, L-thyroxine might be of benefit, on an individual basis. Large prospective trials assessing the effect...