Development and validation of a neural network for NAFLD diagnosis.

Non-Alcoholic Fatty Liver Disease (NAFLD) affects about 20-30% of the adult population in developed countries and is an increasingly important cause of hepatocellular carcinoma. Liver ultrasound (US) is widely used as a noninvasive method to diagnose NAFLD. However, the intensive use of US is not cost-effective and increases the burden on the healthcare system. Electronic medical records facilitate large-scale epidemiological studies and, existing NAFLD scores often require clinical and anthropometric parameters that may not be captured in those databases. Our goal was to develop and validate a simple Neural Network (NN)-based web app that could be used to predict NAFLD particularly its absence. The study included 2970 subjects; training and testing of the neural network using a train-test-split approach was done on 2869 of them. From another population consisting of 2301 subjects, a further 100 subjects were randomly extracted to test the web app. A search was made to find the best parameters for the NN and then this NN was exported for incorporation into a local web app. The percentage of accuracy, area under the ROC curve, confusion matrix, Positive (PPV) and Negative Predicted Value (NPV) values, precision, recall and f1-score were verified. After that, Explainability (XAI) was analyzed to understand the diagnostic reasoning of the NN. Finally, in the local web app, the specificity and sensitivity values were checked. The NN achieved a percentage of accuracy during testing of 77.0%, with an area under the ROC curve value of 0.82. Thus, in the web app the NN evidenced to achieve good results, with a specificity of 1.00 and sensitivity of 0.73. The described approach can be used to support NAFLD diagnosis, reducing healthcare costs. The NN-based web app is easy to apply and the required parameters are easily found in healthcare databases.

Decreased levels of physical activity: results from a cross-sectional study in southern Italy during the COVID-19 lockdown.

BACKGROUND: During the COVID-19 pandemic, the Italian government took security measures to try to limit infections. Restrictive measures included social distancing, home confinement and the closure of all public structures like gyms and swimming pools. The impact of these limitations on health and lifestyle was inevitably negative. The purpose of this study was to establish the level of physical activity (PA), expressed as energy expenditure (MET-minute/week) in a Southern Italian population before and during the COVID-19 lockdown. METHODS: An adapted version of the International Physical Activity Questionnaire-short form (IPAQ-SF) was published on the official website of the National Institute of Gastroenterology IRCCS S. de Bellis, Castellana Grotte, Bari, Italy and on several social media in May 2020. RESULTS: Three hundred ten replies (72% women) from Apulia (60%), Calabria (28%), Campania (11%) and Sicily (1%) were included in the study. The COVID-19 lockdown had a negative effect on the vigorous PA intensity level and on walking, but not on the moderate PA intensity level. Additionally, daily time spent sitting down increased by more than 12% during the COVID-19 lockdown. CONCLUSIONS: Isolation changed PA behaviors. The decreased energy expenditure (MET-minute/week) during the lockdown had a negative impact in both genders, especially on the young adults and adults' groups.

The effect of the Mediterranean Diet on lifespan: a treatment-effect survival analysis of a population-based prospective cohort study in Southern Italy.

BACKGROUND: The Mediterranean diet (MedDiet) is associated with good health. We aimed to estimate the effect of levels of adherence to the MedDiet on lifespan by performing treatment effects survival analysis. METHODS: A sample of 5250 subjects aged ≥18 years were randomly selected from the electoral list of Castellana Grotte and Putignano (Apulian Region, Italy). Cohorts were enrolled in 2005-06 and followed-up until December 2018. The adherence to the MedDiet was measured by the relative Mediterranean score (rMED) and categorized as high, medium and low. Time-to-death (all-causes) as estimated by average treatment effect on the treated (ATET), potential outcome mean (POM) and relative efficiency of exposure were the outcomes. RESULTS: A total of 4896 subjects were included. The median follow-up time was 12.82 (inter quartile range (IQR) 12.22-13.05), 12.91 (IQR 12.21-13.27) and 12.84 (IQR 12.19-13.03) years for high, medium and low rMED subjects respectively. By December 2018, 453 (9.25%) had died. There was a strong effect of medium and low rMED {ATET, -5.10 [95% confidence interval (CI) -9.39, -0.80] and -8.91 (95%CI -13.37, -4.45), respectively}. High rMED has an important effect on mean age at death [POM 90.16 (95% CI 86.06, 94.25)]. The relative effect size for medium and low rMED subjects was a lower lifespan of 5.62% (95% CI 1.01, 10.3) and 9.90% (95% CI 5.30, 5.30), respectively. CONCLUSIONS: We observed an important benefit in additional years of survival from adherence to MedDiet in this southern Italian cohort. Further investigation corroborating our findings in other population groups in other geographic regions will be an important contribution to promoting health and longevity.

Physical Activity and Low Glycemic Index Mediterranean Diet: Main and Modification Effects on NAFLD Score. Results from a Randomized Clinical Trial.

BACKGROUND: Non-Alcoholic Fatty Liver Disease (NAFLD) is the most common chronic liver disease worldwide, and lifestyle modification is the current standard treatment. The aim of the study was to estimate the effect of two different physical activity (PA) programs, a Low Glycemic Index Mediterranean Diet (LGIMD), and their combined effect on the NAFLD score as measured by FibroScan®. METHODS: Moderate or severe NAFLD subjects (n = 144) were randomly assigned to six intervention arms during three months. Interventions arms were a control diet, LGIMD, aerobic activity program (PA1), combined activity program (PA2), and LGIMD plus PA1 or LGIMD plus PA2. The data were compared at baseline, at 45 days, and at 90 days. Analysis of variance was performed under the intention-to-treat principle. RESULTS: There was a statistically significant reduction in the NAFLD score after 45 days of treatment in every working arm except for Arm 1 (control diet). After 90 days, the best results were shown by the intervention arms in which LGIMD was associated with PA: LGIMD plus PA1 (-61.56 95% CI -89.61, -33.50) and LGIMD plus PA2 (-38.15 95% CI -64.53, -11.77). CONCLUSION: All treatments were effective to reduce NAFLD scores, but LGIMD plus PA1 was the most efficient.

Selecting the best machine learning algorithm to support the diagnosis of Non-Alcoholic Fatty Liver Disease: A meta learner study.

BACKGROUND & AIMS: Liver ultrasound scan (US) use in diagnosing Non-Alcoholic Fatty Liver Disease (NAFLD) causes costs and waiting lists overloads. We aimed to compare various Machine learning algorithms with a Meta learner approach to find the best of these as a predictor of NAFLD. METHODS: The study included 2970 subjects, 2920 constituting the training set and 50, randomly selected, used in the test phase, performing cross-validation. The best predictors were combined to create three models: 1) FLI plus GLUCOSE plus SEX plus AGE, 2) AVI plus GLUCOSE plus GGT plus SEX plus AGE, 3) BRI plus GLUCOSE plus GGT plus SEX plus AGE. Eight machine learning algorithms were trained with the predictors of each of the three models created. For these algorithms, the percent accuracy, variance and percent weight were compared. RESULTS: The SVM algorithm performed better with all models. Model 1 had 68% accuracy, with 1% variance and an algorithm weight of 27.35; Model 2 had 68% accuracy, with 1% variance and an algorithm weight of 33.62 and Model 3 had 77% accuracy, with 1% variance and an algorithm weight of 34.70. Model 2 was the most performing, composed of AVI plus GLUCOSE plus GGT plus SEX plus AGE, despite a lower percentage of accuracy. CONCLUSION: A Machine Learning approach can support NAFLD diagnosis and reduce health costs. The SVM algorithm is easy to apply and the necessary parameters are easily retrieved in databases.

Remnant cholesterol as a risk factor for cardiovascular, cancer or other causes mortality: A competing risks analysis.

BACKGROUND AND AIMS: Cardiovascular diseases (CVDis) are leading causes of morbidity and mortality. Even after the introduction of pharmacological therapy to lower Cholesterol, there is still a residual risk that may be ascribed to remnant cholesterol (RC). We aimed, by analyzing two prospective cohort studies, to estimate the effect of RC on risk and hazard of cardiovascular deaths (CVDs), while accounting for competing risks such as cancer (CDs) and other-causes deaths (OCDs). METHODS AND RESULTS: Cohorts were enrolled in 1992 and 2005. Personal data history was recorded. A fasting venous blood sample was obtained, and RC was calculated at baseline. Cause of Death was coded by using ICD-10th version. Follow-up ended on December 31, 2017. Flexible parametric competing-risks models were applied, with age at death as time-axis. In total, 5729 subjects were enrolled. There were 861 (15.1%) deaths: 234 CVDs (27.2%), 245 CDs (28.5%), 271 OCDs (31.5%) and 111 unknown causes of death (12.8%). RC exposure was a strong risk factor only for CVDs (Risk 2.54, 95% Confidence Interval 1.21; 5.34; Trend 1.26 (1.00; 1.58) for ≥1.29 mmol/L). CONCLUSIONS: RC is a strong independent risk factor for cardiovascular mortality. Competing risk analysis is demonstrably a useful tool to disentangle associations among different competing events with a common risk factor.

The Effect of Three Mediterranean Diets on Remnant Cholesterol and Non-Alcoholic Fatty Liver Disease: A Secondary Analysis.

BACKGROUND: Elevated fasting remnant cholesterol (REM-C) levels have been associated with an increased cardiovascular risk in patients with metabolic syndrome (Mets) and Non-Alcoholic Fatty Liver Disease (NAFLD). We aimed to estimate the effect of different diets on REM-C levels in patients with MetS, as well as the association between NAFLD and REM-C. METHODS: This is a secondary analysis of the MEDIDIET study, a parallel-arm Randomized Clinical Trial (RCT). We examined 237 people with MetS who underwent Liver Ultrasound (LUS) to assess the NAFLD score at baseline, 3-, and 6-months follow-up. Subjects were randomly assigned to the Mediterranean diet (MD), Low Glycemic Index diet (LGID), or Low Glycemic Index Mediterranean diet (LGIMD). REM-C was calculated as [total cholesterol-low density lipoprotein cholesterol (LDL-C)-high density lipoprotein cholesterol (HDL-C)]. RESULTS: REM-C levels were higher in subjects with moderate or severe NAFLD than in mild or absent ones. All diets had a direct effect in lowering the levels of REM-C after 3 and 6 months of intervention. In adherents subjects, this effect was stronger among LGIMD as compared to the control group. There was also a significant increase in REM-C levels among Severe NAFLD subjects at 3 months and a decrease at 6 months. CONCLUSIONS: fasting REM-C level is independently associated with the grade of severity of NAFLD. LGIMD adherence directly reduced the fasting REM-C in patients with MetS.

Effects of Some Food Components on Non-Alcoholic Fatty Liver Disease Severity: Results from a Cross-Sectional Study.

Background: The high prevalence of non-alcoholic fatty liver disease (NAFLD) observed in Western countries is due to the concurrent epidemics of overweight/obesity and associated metabolic complications, both recognized risk factors. A Western dietary pattern has been associated with weight gain and obesity, and more recently with NAFLD. Methods: This is a baseline cross-sectional analysis of 136 subjects (79 males) enrolled consecutively in the NUTRIATT (NUTRItion and Ac-TiviTy) study. Study subjects had moderate or severe NAFLD diagnosed by using Fibroscan-CAP. Food Frequency Questionnaire was used to obtain information about food intake. Statistical analysis included descriptive statistics and a multivariable logistic regression model. Results: The mean age was 49.58 (±10.18) with a mean BMI of 33.41 (±4.74). A significant inverse relationship was revealed between winter ice-cream intake and NAFLD severity (O.R. 0.65, 95% C.I. 0.95-0.99); chickpeas intake and NAFLD severity (O.R. 0.57, 95% C.I. 0.34-0.97), and not industrial aged-cheeses type (O.R. 0.85, 95% C.I. 0.74-0.98). A statistically significant positive association also emerged between rabbit meat (O.R. 1.23, 95% C.I. 1.01-1.49), industrial type aged cheeses (O.R. 1.17, 95% C.I. 1.01-1.35), milk-based desserts (no winter ice cream) (O.R. 1.11, 95% C.I. 1.01-1.21), fats (O.R. 1.12, 95% C.I. 1.01-1.25), and NAFLD severity. Conclusion: The fresh foods from non-intensive farming and high legume intake that characterize the Mediterranean diet would seem to be beneficial for patients with NAFLD.

Reducing NAFLD-screening time: A comparative study of eight diagnostic methods offering an alternative to ultrasound scans.

BACKGROUND & AIMS: The use of ultrasound scan (US) in non-alcoholic fatty liver disease (NAFLD) screening overloads US waiting lists. We hypothesized and tested a hybrid two-step method, consisting of applying a formula, to exclude subjects at low risk, before US. METHODS: The sample included 2970 males and females (937 with NAFLD) diagnosed by US. We selected eight formulas: Fatty Liver Index (FLI), Hepatic Steatosis Index (HIS), body mass index (BMI), waist circumference (WC), Abdominal Volume Index (AVI), waist-to-height ratio (WHtR), waist/height0.5 (WHT.5R) and Body Roundness Index (BRI), and calculated their performance in the two-step method evaluating percentage reduction of the number of liver US (US reduction percentage), percentage of false negative and percentage of NAFLD identified. RESULTS: The US reductions percentage were 52.2% (WHtR), 52.1% (HIS), 51.8% (FLI), 50.8% (BRI), 50.7% (BMI and WHt_5R), 46.5% (WC) and 45.2% (AVI). The false negative percentage were 8.5% (WHtR), 7.9% (BRI), 7.3% (WHt_5R), 7.2% (BMI), 6.7% (HIS), 6.6% (FLI), 5.6% (WC) and 5.2% (AVI). The best percentage of NALFD identified was obtained using AVI (83.6%) before US, then WC (82.2%), FLI (79%), HIS (78.9%), BMI (77.3%), WHt_5R (76.9%), BRI (74.8%) and WHtR (73%). CONCLUSION: The best formula to use in two-step diagnostic NAFLD screening was AVI, which showed a low false negative rate and a higher percentage of identified NAFLD. Other studies evaluating the economic advantages of this screening method are warranted.

Irisin Serum Levels in Metabolic Syndrome Patients Treated with Three Different Diets: A Post-Hoc Analysis from a Randomized Controlled Clinical Trial.

BACKGROUND: Irisin, a hormone-like myokine, regulates energy homeostasis and mediates the benefits of physical activity on health. METHODS: To estimate the effect of different diets on irisin concentrations in subjects with the Metabolic Syndrome (MetS). METHODS: Subjects with MetS were derived from a population survey; 163 subjects were enrolled and randomized to a: Low Glycaemic Index (LGID), Mediterranean (MD) or Low Glycaemic Index Mediterranean (LGIMD) Diet, and the groups were compared, also with 80 controls without MetS. Sociodemographic, medical and nutritional data were collected and fasting blood samples drawn. Subjects underwent LUS and bioimpedentiometry. Generalized Estimating Equations were performed. RESULTS: At baseline, lower irisin concentrations were observed in MetS subjects. Mean irisin levels increased in all diet groups but only the LGID group reached statistical significance, as well as showing an interaction between LGID and time at the sixth month examination (4.57, 95% CI −1.27, 7.87). There was a positive effect of Vegetable Proteins (0.03, 95% CI −0.01,0.06) and Saturated Fatty Acids (0.04, 95% CI 0.01, 0.07) on irisin concentrations. In the LGIMD, a positive effect on Fat-Free Mass (0.38, 95% CI 0.19, 0.57) and a negative effect on the Body Mass Index (−0.75, 95% CI −1.30, −0.19) were observed. CONCLUSIONS: There seems to be a link between diet and muscle physiology. We showed that patients following a LGID had higher levels of irisin, a promising biomarker of muscle activity.