An artificial neural network to predict resting energy expenditure in obesity.

BACKGROUND & AIMS: The resting energy expenditure (REE) determination is important in nutrition for adequate dietary prescription. The gold standard i.e. indirect calorimetry is not available in clinical settings. Thus, several predictive equations have been developed, but they lack of accuracy in subjects with extreme weight including obese populations. Artificial neural networks (ANN) are useful predictive tools in the area of artificial intelligence, used in numerous clinical fields. The aim of this study was to determine the relevance of ANN in predicting REE in obesity. METHODS: A Multi-Layer Perceptron (MLP) feed-forward neural network with a back propagation algorithm was created and cross-validated in a cohort of 565 obese subjects (BMI within 30-50 kg m-2) with weight, height, sex and age as clinical inputs and REE measured by indirect calorimetry as output. The predictive performances of ANN were compared to those of 23 predictive REE equations in the training set and in two independent sets of 100 and 237 obese subjects for external validation. RESULTS: Among the 23 established prediction equations for REE evaluated, the Harris & Benedict equations recalculated by Roza were the most accurate for the obese population, followed by the USA DRI, Müller and the original Harris & Benedict equations. The final 5-fold cross-validated three-layer 4-3-1 feed-forward back propagation ANN model developed in that study improved precision and accuracy of REE prediction over linear equations (precision = 68.1%, MAPE = 8.6% and RMSPE = 210 kcal/d), independently from BMI subgroups within 30-50 kg m-2. External validation confirmed the better predictive performances of ANN model (precision = 73% and 65%, MAPE = 7.7% and 8.6%, RMSPE = 187 kcal/d and 200 kcal/d in the 2 independent datasets) for the prediction of REE in obese subjects. CONCLUSIONS: We developed and validated an ANN model for the prediction of REE in obese subjects that is more precise and accurate than established REE predictive equations independent from BMI subgroups. For convenient use in clinical settings, we provide a simple ANN-REE calculator available at: https://www.crnh-rhone-alpes.fr/fr/ANN-REE-Calculator.

Cereal processing influences postprandial glucose metabolism as well as the GI effect.

OBJECTIVE: Technological processes may influence the release of glucose in starch. The aim of this study was to compare the metabolic response and the kinetics of appearance of exogenous glucose from 2 cereal products consumed at breakfast. METHODS: Twenty-five healthy men were submitted to a randomized, open, crossover study that was divided into 2 parts: 12 of the 25 subjects were included in the "isotope part," and the 13 other subjects were included in the "glycemic part." On test days, subjects received biscuits (low glycemic index [GI], high slowly available glucose [SAG]) or extruded cereals (medium GI, low SAG) as part of a breakfast similar in terms of caloric and macronutrient content. The postprandial phase lasted 270 minutes. RESULTS: The rate of appearance (RaE) of exogenous glucose was significantly lower after consumption of biscuits in the first part of the morning (90-150 minutes) than after consumption of extruded cereals (p ≤ 0.05). Conversely, at 210 minutes, it was significantly higher with biscuits (p ≤ 0.01). For the first 2 hours, plasma glucose and insulin were significantly lower after biscuits during the glycemic part. C-peptide plasma concentrations were significantly lower at 90, 120, and 150 minutes after ingestion of the biscuits (p ≤ 0.05). CONCLUSION: The consumption of biscuits with a high content of slowly digestible starch reduces the appearance rate of glucose in the first part of the morning and prolongs this release in the late phase of the morning (210 minutes). Our results also emphasize that modulation of glucose availability at breakfast is an important factor for metabolic control throughout the morning in healthy subjects due to the lowering of blood glucose and insulin excursions.

Daily intake of conjugated linoleic acid-enriched yoghurts: effects on energy metabolism and adipose tissue gene expression in healthy subjects.

Conjugated linoleic acid (CLA) is a group of positional and geometric isomers of conjugated dienoic derivatives of linoleic acid. The present study was designed to determine whether 14-week CLA supplementation as triacylglycerols (3.76 g) with a 50 : 50 combination of the two main isomers (35 % cis-9, trans-11 and 35 % trans-10, cis-12) added to flavoured yoghurt-like products was able to alter body composition in healthy subjects and to alter the expression of several key adipose tissue genes (PPAR gamma, lipoprotein lipase (LPL), hormone-sensitive lipase (HSL) and uncoupling protein 2 (UCP-2)). Forty-four healthy subjects were randomly assigned to consume daily either a CLA-supplemented yoghurt-like product or a placebo yoghurt for 98 d. There were no significant effects of CLA supplementation on body weight, fat mass or free fat mass. Basal energy expenditure expressed as kg free fat mass increased significantly in the CLA group (123.3 (SEM 2.5) kJ/kg free fat mass per d on day 98 v. 118.7 (SEM 2.3) kJ/kg free fat mass per d on day 0, P = 0.03). PPAR gamma mRNA gene expression increased significantly with CLA supplementation (53 (SEM 20) %, P < 0.01) and a significant reduction in mRNA levels of HSL was observed ( - 42 (SEM 7) %, P = 0.01). The levels of UCP-2 and LPL mRNA were not affected. The present results suggest that a 98 d supplementation diet with a 50 : 50 mixture of the two CLA isomers cis-9, trans-11 and trans-10, cis-12 in a dairy product was unable to alter body composition, although a significant increase in the RMR has been induced. Moreover, changes in mRNA PPAR gamma and HSL in adipose tissue were recorded.

Modified quantitative insulin sensitivity check index is better correlated to hyperinsulinemic glucose clamp than other fasting-based index of insulin sensitivity in different insulin-resistant states.

Fasting-based index estimates of insulin sensitivity were compared with euglycemic hyperinsulinemic clamp (IS clamp) measurements in 148 subjects: normal controls (n = 46), and obese (n = 12), polycystic ovary syndrome (n = 16), first-degree relatives of type 2 diabetic (n = 17), impaired glucose tolerance (n = 28), and type 2 diabetic (n = 29) patients. The fasting-based indexes tested included log homeostasis model assessment (HOMA), the quantitative insulin sensitivity check index (QUICKI), the revised QUICKI, and a new revised QUICKI using fasting plasma glycerol. In the population studied, at 40 mU/m(2).min (n = 30) revised QUICKI (r = 0.86; P < 0.0001) and QUICKI-glycerol (r = 0.87; P < 0.0001) gave higher correlations with the IS clamp than QUICKI and log HOMA (r = 0.78 and r = -0.78; P < 0.001). For subjects tested at 75 mU/m(2).min (n = 118), comparable correlations were found for all indexes (r > 0.80; P < 0.0001). When studied in subgroups, revised QUICKI and QUICKI-glycerol give significantly higher correlations with the IS clamp than other indexes for lean control subjects studied at 40mU/m(2).min and impaired glucose tolerance subjects. We confirmed, in a large patient population with a wide range of insulin sensitivities, that no single test is superior in all groups of patients. However, QUICKI and revised QUICKI are good indexes that offer correlations similar to or higher than values obtained with log HOMA. Such indexes are simple tools to estimate insulin sensitivity appropriate for epidemiological studies.