Gil HW, et al. J Korean Med Sci. 2014 Nov;29(11):1441-1449. https://doi.org/10.3346/jkms.2014.29.11.1441

Chronic kidney disease (CKD) leads to cognitive impairment and emotional changes. However, the precise mechanism underlying the crosstalk between the kidneys and the nervous system is not fully understood. Inflammation and cerebrovascular disease can influence the development of depression in CKD. CKD is one of the strongest risk factors for cognitive impairment. Moreover, cognitive impairment occurs in CKD as patients experience the dysregulation of several brain functional domains due to damage caused to multiple cortical regions and to subcortical modulatory neurons. The differences in structural brain changes between CKD and non-CKD dementia may be attributable to the different mechanisms that occur in CKD. The kidney and brain have similar anatomical vascular systems, which may be susceptible to traditional risk factors. Vascular factors are assumed to be involved in the development of cognitive impairment in patients with CKD. Vascular injury induces white matter lesions, silent infarction, and microbleeds. Uremic toxins may also be directly related to cognitive impairment in CKD. Many uremic toxins, such as indoxyl sulfate, are likely to have an impact on the central nervous system. Further studies are required to identify therapeutic targets to prevent changes in the brain in patients with CKD.

Background: Indoxyl sulfate (IS), an organic anion uremic toxin, promotes the progression of renal dysfunction. Some studies have suggested that IS inhibits osteoclast differentiation and suppresses parathyroid hormone (PTH)-stimulated intracellular cAMP production, decreases PTH receptor expression, and induces oxidative stress in primary mouse calvaria osteoblast cell culture. However, the direct effects of IS on osteoblast apoptosis have not been fully evaluated. Hence, we investigated whether IS acts as a bone toxin by studying whether IS induces apoptosis and inhibits differentiation in the cultured osteoblast cell line MC3T3-E1. Methods: We assessed the direct effect of IS on osteoblast differentiation and apoptosis in the MC3T3-E1 cell line. We examined caspase-3/7 activity, apoptosis-related proteins, free radical production, alkaline phosphatase activity, and mRNA expression of type 1 collagen and osteonectin. Furthermore, we investigated the uptake of IS via organic anion transport (OAT). Results: We found that IS increased caspase activity and induced apoptosis. Production of free radicals increased depending on the concentration of IS. Furthermore, IS inhibited the expression of mRNA type 1 collagen and osteonectin and alkaline phosphatase activity. The expression of OAT, which is known to mediate the cellular uptake of IS, was detected in in the MC3T3-E1 cell line. The inhibition of OAT improved cell viability and suppressed the production of reactive oxygen species. These results suggest that IS is transported in MC3T3-E1 cells via OAT, which causes oxidative stress to inhibit osteoblast differentiation. Conclusions: IS acts as a bone toxin by inhibiting osteoblast differentiation and inducing apoptosis.

In this study, we propose a personalized glucose prediction model using deep learning for hospitalized patients who experience Type-2 diabetes. We aim for our model to assist the medical personnel who check the blood glucose and control the amount of insulin doses. Herein, we employed a deep learning algorithm, especially a recurrent neural network (RNN), that consists of a sequence processing layer and a classification layer for the glucose prediction. We tested a simple RNN, gated recurrent unit (GRU), and long-short term memory (LSTM) and varied the architectures to determine the one with the best performance. For that, we collected data for a week using a continuous glucose monitoring device. Type-2 inpatients are usually experiencing bad health conditions and have a high variability of glucose level. However, there are few studies on the Type-2 glucose prediction model while many studies performed on Type-1 glucose prediction. This work has a contribution in that the proposed model exhibits a comparative performance to previous works on Type-1 patients. For 20 in-hospital patients, we achieved an average root mean squared error (RMSE) of 21.5 and an Mean absolute percentage error (MAPE) of 11.1%. The GRU with a single RNN layer and two dense layers was found to be sufficient to predict the glucose level. Moreover, to build a personalized model, at most, 50% of data are required for training.

Lee JW, et al. J Korean Med Sci. 2015 Oct;30(10):1517-1521. https://doi.org/10.3346/jkms.2015.30.10.1517