It is known that being overweight is often associated with impaired glucose tolerance and increased risk of developing type 2 diabetes [ 21 , 34 , 35 ]. This study shows that the group of patients who developed diabetes following steroid therapy not only weighed less, when compared to individuals with type 2 diabetes receiving or not receiving steroids, but that despite treatment with steroids, which in itself may cause weight gain, obesity was not a distinctive feature. Furthermore, type 2 diabetes is typically associated with a strong family history [ 36 – 39 ]. By contrast, patients with NOSID in this study had less family history of diabetes. If NOSID was simply type 2 diabetes uncovered opportunistically (due to concurrent illness or steroid treatment), a similar prevalence of family history and obesity would be expected in all groups. These two findings are more consistent with the notion that patients with NOSID have less risk factors for diabetes. They only become diabetic under the stress of steroid treatment.