Comparison of Machine Learning Approaches Toward Assessing the Risk of Developing Cardiovascular Disease as a Long-Term Diabetes Complication.

The estimation of long-term diabetes complications risk is essential in the process of medical decision making. Guidelines for the management of Type 2 Diabetes Mellitus (T2DM) advocate calculating the Cardiovascular Disease (CVD) risk to initiate appropriate treatment. The objective of this study is to investigate the use of sophisticated machine learning techniques toward the development of personalized models able to predict the risk of fatal or nonfatal CVD incidence in T2DM patients. The important challenge of handling the unbalanced nature of the available dataset is addressed by applying novel ensemble strategies. Hybrid Wavelet Neural Networks (HWNNs) and Self-Organizing Maps (SOMs) constitute the primary models for building ensembles following a subsampling approach. Different methods for combining the decisions of the primary models are applied and comparatively assessed. Data from the 5-year follow up of 560 patients with T2DM are used for development and evaluation purposes. The highest discrimination performance (Area Under the Curve (AUC): 71.48%) is achieved by taking into account both the HWNN- and SOM- based primary models' outputs. The proposed method is superior to the Binomial Linear Regression (BLR) model justifying the need to apply more sophisticated techniques in order to produce reliable CVD risk scores.

Dietary fat intake as risk factor for the development of diabetes: multinational, multicenter study of the Mediterranean Group for the Study of Diabetes (MGSD).

OBJECTIVE: To investigate the role of dietary factors in the development of type 2 diabetes. RESEARCH DESIGN AND METHODS: In the context of the Multinational MGSD Nutrition Study, three groups of subjects were studied: 204 subjects with recently diagnosed diabetes (RDM), 42 subjects with undiagnosed diabetes (UDM) (American Diabetes Association criteria-fasting plasma glucose [FPG] > or =126 mg/dl), and 55 subjects with impaired fasting glucose (IFG) (FPG > or =110 and <126 mg/dl). Each group was compared with a control group of nondiabetic subjects, matched one by one for center, sex, age, and BMI. Nutritional habits were evaluated by a dietary history method, validated against the 3-day diet diary. In RDM, the questionnaire referred to the nutritional habits before the diagnosis of diabetes. Demographic data were collected, and anthropometrical and biochemical measurements were taken. RESULTS: Compared with control subjects, RDM more frequently had a family history of diabetes (49.0 vs. 14.2%; P < 0.001), exercised less (exercise index 53.5 vs. 64.4; P < 0.01), and more frequently had sedentary professions (47.5 vs. 27.4%; P < 0.001). Carbohydrates contributed less to their energy intake (53.5 vs. 55.1%; P < 0.05), whereas total fat (30.2 +/- 0.5 vs. 27.8 +/- 0.5%; P < 0.001) and animal fat (12.2 +/- 0.3 vs. 10.8 +/- 0.3%; P < 0.01) contributed more and the plant-to-animal fat ratio was lower (1.5 +/- 0.1 vs. 1.8 +/- 0.1; P < 0.01). UDM more frequently had a family history of diabetes (38.1 vs. 19.0%; P < 0.05) and sedentary professions (58.5 vs. 34.1%; P < 0.05), carbohydrates contributed less to their energy intake (47.6 +/- 1.7 vs. 52.8 +/- 1.4%; P < 0.05), total fat (34.7 +/- 1.5 vs. 30.4 +/- 1.2%; P < 0.05) and animal fat (14.2 +/- 0.9 vs. 10.6 +/- 0.7%; P < 0.05) contributed more, and the plant-to-animal fat ratio was lower (1.6 +/- 0.2 vs. 2.3 +/- 0.4; P < 0.05). IFG differed only in the prevalence of family history of diabetes (32.7 vs. 16.4%; P < 0.05). CONCLUSIONS: Our data support the view that increased animal fat intake is associated with the presence of diabetes.

Spontaneous bacterial peritonitis (SBP): clinical, laboratory, and prognostic features. A single-center experience.

BACKGROUND: The aim of the present study was to assess the clinical and laboratory characteristics, the course, and the factors influencing the hospital mortality and relapse rate of spontaneous bacterial peritonitis (SBP) in cirrhotic patients admitted in a single University Hospital in Greece. METHODS: The study comprises the evaluation of 81 cirrhotic patients who developed SBP during a 30-month period. RESULTS: The occurrence of SBP was independent of the etiology of liver disease and was symptomatic in 66/81 patients (82%). Encephalopathy, as presenting symptom, occurred mainly in Child C patients. Ascitic fluid culture was positive in 20 patients (25%); E. coli (60%) and Enterococcus faecalis (14%) were the most common bacteria isolated. Empirical treatment was effective in 94% of patients. Renal impairment was observed in 21 patients (26%), six of whom developed hepatorenal syndrome. Total mortality was 10% and was related to the existence of symptoms (P<0.01), ascetic fluid polymorphonuclear cell count (P<0.05), bilirubin levels (P<0.01), and kidney function at the beginning of the episode (P<0.01). The relapse rate was 24.6% and was related to the Child stage (P<0.01). CONCLUSIONS: SBP was asymptomatic in a substantial number of patients. Deterioration of renal function was frequently observed and was the main cause of death. The low (10%) in-hospital mortality seems to be related to earlier diagnosis and treatment. Relapse was associated with the severity of liver disease.