Update on Management of Type 2 Diabetes for Cardiologists

Is rating of perceived exertion a valid method for monitoring exergaming intensity in type-1 diabetics? A cross-sectional randomized trial

AIMS

The rating of perceived exertion (RPE) provides correlations with physiological measurements of exercise intensity, including metabolic equivalent (MET), oxygen consumption (V˙O2), and heart rate (HR), in real (RS) and virtual (VS) sessions. To use RPE in patients with pathology, we aimed to examine the concurrent validity of RPE in type-1 diabetes mellitus (T1DM) patients while exergaming.

METHODS

Ten T1DM patients performed two 30-min crossover sessions of moderate-intensity exercise (washout 72-196 h). The RS group performed running, and the VS group played the Kinect Adventures! video game. METs were measured by a direct gas analyzer during the sessions, and RPE was measured on the 6 - 20 point Borg scale after the sessions.

RESULTS

RS and VS showed similar RPE (13.2 ± 2.7 vs. 14.2 ± 2.4) and MET (4.6 ± 1.1 vs. 4.0 ± 0.8) values (p > 0.05). RPE vs. MET correlation-coefficients were large in RS (r = 0.64; R2 = 41; p = 0.04) and were moderate in VS (r = 0.42; R2 = 18; p = 0.22). Additionally, RS secondary values (V˙O2 and HR vs. RPE) showed high coefficients (V˙O2-r = 0.62; average HRr = 0.62; maximal HRr = 0.50, p < 0.05). VS secondary values, on the other hand, showed low-moderate coefficients (V˙O2-r = 0.42; average HRr = 0.23; maximal HRr = 0.21, p > 0.05).

CONCLUSION

The current validation showed that RPE may not be a valid and strong method for T1DM patients while exergaming. Healthcare professionals should cautiously use the 6 - 20 point RPE scale in pathological patients, specifically in T1DM while exergaming.

Do estimated metabolic equivalent and energy expenditure verify the physical effort of type-1 diabetics in resting and exercise situations? A randomized crossover trial

ABSTRACT To compare measured and estimated metabolic equivalent (MET) and energy expenditure (EE) in different situations with Type-1 diabetes (T1DM) patients. Ten T1DM patients performed three 30-minute sessions (resting, running-RS, and exergame-VS) at moderate intensity. MET and EE were measured by direct gas analyzer and estimated using the formula applying heart rate and V̇02peak. MET values (measured vs. estimated) were statistically different during RS (4.58±1.11 vs.7.59±1.36) and VS (3.98± 0.84 vs. 5.77±0.84) (p<0.001). EE values were similar: RS (147±43 vs. 246±157) and VS (129±33 vs. 184±20) (p<0.001). The error between the methods: 0.41, 1.51, and 1.07 METs and 20.1, 51.5, and 32.5 Kcals for resting, RS, and VS. Estimation could be used in resting and with caution for RS and VS.

Inhibiting Glucose Metabolism By miR-34a and miR-125b Protects Against Hyperglycemia-Induced Cardiomyocyte Cell Death

BACKGROUND

It is well-known that insulin resistance and hyperglycemia are important pathological causes for the development of diabetic cardiomyopathy (DCM). However, its precise molecular mechanisms in the pathogenesis of DCM remain unclear.

OBJECTIVES

Recent studies reveal that microRNAs (miRNA) play essential roles in the pathogenesis of DCM. This project aimed to determine the roles of miR-34a and miR-125b in hyperglycemia-induced cardiomyocyte cell death.

METHODS

Rat primary cardiomyocytes were isolated and exposed to normal and high concentrations of glucose. Cell viability was measured using MTT assay. Expressions of miR-34a and miR-125b were detected by qRT-PCR. Potential targets of miR-34a and miR-125b were predicted from www.Targetscan.org and validated from human heart tissues. A statistical significance of p<0.05 was considered.

RESULTS

The present study shows that miR-34a and miR-125b are downregulated in a human diabetic heart. Moreover, in vitro data from rat primary cardiomyocytes showed that short-term high glucose treatment stimulates miR-34a and miR-125b expressions. Under high glucose, it was found that rat cardiomyocytes displayed increased intracellular glucose metabolism, and glucose uptake and lactate production were significantly increased. It was also found that the key glucose metabolic enzymes, Hexokinase 2 (HK2) and Lactate dehydrogenase-A (LDHA), were direct targets of miR-125b and miR-34a, respectively. Overexpression of miR-125b and miR-34a could prevent hyperglycemia-induced cardiomyocyte cell death. Finally, the restoration of HK2 and LDHA in miR-125b and miR-34a overexpressed cardiomyocytes recovered the cardiomyocytes' sensitivity to hyperglycemia.

CONCLUSION

Our results proposed a molecular mechanism for the microRNA-mediated diabetic cardiovascular protection and will contribute to developing treatment strategies for diabetes-associated cardiovascular dysfunction.

Predicting complications of diabetes mellitus using advanced machine learning algorithms

OBJECTIVE

We sought to predict if patients with type 2 diabetes mellitus (DM2) would develop 10 selected complications. Accurate prediction of complications could help with more targeted measures that would prevent or slow down their development.

MATERIALS AND METHODS

Experiments were conducted on the Healthcare Cost and Utilization Project State Inpatient Databases of California for the period of 2003 to 2011. Recurrent neural network (RNN) long short-term memory (LSTM) and RNN gated recurrent unit (GRU) deep learning methods were designed and compared with random forest and multilayer perceptron traditional models. Prediction accuracy of selected complications were compared on 3 settings corresponding to minimum number of hospitalizations between diabetes diagnosis and the diagnosis of complications.

RESULTS

The diagnosis domain was used for experiments. The best results were achieved with RNN GRU model, followed by RNN LSTM model. The prediction accuracy achieved with RNN GRU model was between 73% (myocardial infarction) and 83% (chronic ischemic heart disease), while accuracy of traditional models was between 66% - 76%.

DISCUSSION

The number of hospitalizations was an important factor for the prediction accuracy. Experiments with 4 hospitalizations achieved significantly better accuracy than with 2 hospitalizations. To achieve improved accuracy deep learning models required training on at least 1000 patients and accuracy significantly dropped if training datasets contained 500 patients. The prediction accuracy of complications decreases over time period. Considering individual complications, the best accuracy was achieved on depressive disorder and chronic ischemic heart disease.

CONCLUSIONS

The RNN GRU model was the best choice for electronic medical record type of data, based on the achieved results.