Development of non-invasive blood glucose regression based on near-infrared spectroscopy combined with a deep learning method

Extracting micro-scale spectral features from dynamic blood glucose concentrations is extremely difficult when using non-invasive measurement methods. This work proposes a new machine-learning method based on near-infrared spectroscopy, deep belief network (DBN), and support vector machine (SVR), to improve the prediction accuracy. First, the standard oral glucose tolerance test is used to collect near-infrared spectroscopy and actual blood glucose concentration values for specific wavelengths (1200, 1300, 1350, 1450, 1600, 1610, and 1650 nm), and the blood glucose concentrations is within a clinical range of 70mg/dL~220mg/dL. Second, based on the DBN model, high-dimensional deep features of the non-invasive blood glucose spectrum are extracted. These are used to establish a support vector regression (SVR) model and to quantitatively analyze the influence of spectral sample size and corresponding feature dimensions (i.e., DBN network structure) on the prediction accuracy. Finally, based on data from six volunteers, a comparative analysis of the SVR prediction accuracy is performed both before and after using high-dimensional deep features. For volunteer 1, when the DBN-based high-dimensional deep features were used, the root mean square error (RMSE) of support vector regression (SVR) was reduced by 71.67%, the correlation coefficient (R2) and the P value of Clark grid analysis (P) were increased by 13.99% and 6.28%, respectively. Moreover, we have similar results when the proposed method was carried out on the data of other volunteers. The results show that the presented algorithm can play an important role in dynamic non-invasive blood glucose concentration prediction and can effectively improve the accuracy of the SVR model. Further, by applying the algorithm to six independent sets of data, this research also illustrates the high-precision regression and generalization capabilities of the DBN-SVR algorithm.

Age at First Gestation in Beef Heifers Affects Fetal and Postnatal Growth, Glucose Metabolism and IGF1 Concentration

This study aimed to determine the effects of age at first gestation on offspring growth performance, glucose metabolism, and IGF1 concentration. Heifers impregnated by AI from a single bull at 15 months of age (15 M, n = 20), or 27 months of age (27 M, n = 20), and multiparous cows (adult, n = 20) were used. Dams from all groups were managed in a single group during gestation and lactation. Gestational length was longer in the 15 M and 27 M than in adult dams (p = 0.009). Bodyweight at birth, at weaning and ADG during lactation were higher in calves from adult dams than in those from 27 M dams, and higher in calves from the latter than in 15 M calves (p < 0.001). Calves from 15 M dams had an increased head circumference/BW ratio compared to calves from 27 M dams, while calves from this latter group had an increased ratio compared to calves from adults (p = 0.005). Body mass index was greater in calves from adults than in those from 15 M and 27 M dams (p = 0.002). Milk production from 15 M and 27 M dams was similar but lower than that from adults (p = 0.03). Calves born from adult dams had greater blood glucose concentrations than those from 15 M and 27 M dams (p < 0.05). Serum IGF1 concentrations were higher in calves from adults than in calves from 15 M and 27 M dams (p = 0.01). This study showed that age at first gestation affects offspring postnatal growth performance, glucose metabolism and IGF1 concentration.

Delayed Rebound of Glycemia During Recovery Following Short-Duration High-Intensity Exercise: Are There Lactate and Glucose Metabolism Interactions?

Lactate constitutes the primary gluconeogenic precursor in healthy humans at rest and during low-intensity exercise. Data on the interactions between lactate and glucose metabolisms during recovery after short-duration high-intensity exercise are sparse. The aim of the present study was to describe blood glucose ([glucose]b) and lactate ([lactate]b) concentration curves during recovery following short-duration high-intensity exercise. Fifteen healthy Cameroonian subjects took part in the study and performed successively (i) an incremental exercise to exhaustion to determine maximal work rate (Pmax) and (ii) a 2-min 110% Pmax exercise after which blood lactate and glucose concentrations were measured during the 80-min passive recovery. In response to the 2-min 110% Pmax exercise, [glucose]b remained stable (from 4.93 ± 1.13 to 4.65 ± 0.74 mmol.L−1, NS) while [lactate]b increased (from 1.35 ± 0.36 to 7.87 ± 1.66 mmol.L−1, p &lt; 0.0001). During recovery, blood lactate concentrations displayed the classic biphasic curve while blood glucose concentrations displayed a singular shape including a delayed and transitory rebound of glycemia. This rebound began at 27.7 ± 6.2 min and peaked at 6.78 ± 0.53 mmol.L−1 at 56.3 ± 9.7 min into recovery. The area under the curve (AUC) of [lactate]b during the rebound of glycemia was positively correlated with the peak value of glycemia and the AUC of [glucose]b during the rebound. In conclusion, the delayed rebound of glycemia observed in the present study was associated with lactate availability during this period.

A glikációs index lehetséges magyarázata a hemoglobinglikáció biokinetikus modellje alapján

Összefoglaló. Bevezetés: A HbA1c integrált retrospektív mutatója az elmúlt időszak vércukrának, rendszeres vizsgálata a cukorbetegek anyagcserekontrolljának megítélésében elengedhetetlen. Helyes értékelése azonban nem egyszerű, mert a HbA1c és a vércukor közötti összefüggés nem lineáris. A mérést közvetlenül megelőző hyperglykaemiás epizódok hatása a HbA1c szintjére nagyobb, mint azoké, amelyek régebben történtek. A jelenségre a glikáció biokinetikus modellje ad magyarázatot. Célkitűzés: A mért és a biokinetikus modell alapján számított HbA1c közötti egyezés, illetve diszkordancia vizsgálata. Módszer: A vizsgálatokat 157, 1-es és 2-es típusú cukorbeteg 1793, laboratóriumban mért éhomi vércukor- és 511 HbA1c-adatából végeztük. A különbséget a glikációs index segítségével számítottuk, amely a mért és a számított HbA1c-érték aránya. Eredmények: Egyezést mindössze a vizsgált betegek kevesebb mint egyötödödében találtunk, 60%-ban az index értéke alacsony (<0,95) és 21%-ban magas (>1,05) volt. Az adatok részletes analízise szerint jó anyagcserekontroll esetében gyakoribb a vártnál magasabb, mért HbA1c-érték, mint a biokinetikus egyenlet által számítotté, és rosszabb kontroll (magasabb átlagos vércukor) esetében ez fordítva van. Egyezés esetén a regressziós egyenlet együtthatói gyakorlatilag azonosak a modell alapján számított értékekkel. Következtetés: Vizsgálataink felvetik azt a lehetőséget, hogy a biokinetikus modell magyarázatot adhat a vércukor és a HbA1c közötti diszkordanciára. Orv Hetil. 2021; 162(41): 1652–1657. Summary. Introduction: HbA1c is an integrated retrospective marker of previous blood glucose concentrations and its regular measurement is indispensable in the assessment of glycaemic compensation of diabetic patients. However, its proper interpretation is not simple becasuse the relationship between HbA1c and average glycemia is not a linear one. Hyperglycemic episodes occuring immediately before the measurement have greater impact on the HbA1c level as compared with those taking place earlier. Objective: Assessment of concordance and discordance between measured and according to the biokinetic model calculated values of HbA1c. Method: The calculations were made from averages of 1793 fasting blood glucose and 511 HbA1c of 157, type 1 and 2 diabetic patients. The glycation index is the quotient between measured and calculated HbA1c. Results: Agreement was found in less than one fifth of the 157 patients; in 60% the value of glycation was low (<0.95) and in 21% high (>1.05). Analysis of the glycation index according to the level of glycemic compensation revealed that in patients with good compensation, the measured HbA1c value was more often higher than the expected and in patients with unsatisfactory compensation the opposite was true. Conclusion: These results raise the possibility that the discordance between average glycemia and measured HbA1c can be explained by the biokinetic model. Orv Hetil. 2021; 162(41): 1652–1657.

Prevention of Hyperglycemia

Hyperglycemia is the elevation of blood glucose concentrations above the normal range. Prolonged uncontrolled hyperglycemia is associated with serious life-threatening complications. Hyperglycemia arises from an imbalance between glucose production and glucose uptake and utilization by peripheral tissues. Disorders that compromise pancreatic function or affect the glucose counter-regulatory hormones cause hyperglycemia. Acute or serious illness or injury may also bring about hyperglycemia, as can many classes of drugs. Metformin lowers blood glucose levels by inhibiting the production of glucose by the liver whilst enhancing uptake of circulating glucose and its utilization in peripheral tissues such as muscle and adipose tissue. Metformin suppresses hepatic gluconeogenesis by inhibiting mitochondrial respiration and causing a reduction of cellular ATP levels. Metformin may also modulate the gut-brain-liver axis, resulting in suppression of hepatic glucose production. Metformin also opposes the hyperglycemic action of glucagon and may ameliorate pancreatic cell dysfunction associated with hyperglycemia. Metformin is therefore recommended for use in the prevention of hyperglycemia, including drug-induced hyperglycemia, in at risk patients. The benefits of metformin in the prevention of hyperglycemia are unmatched despite its contraindications.

A Systematic Review of Intravenous β-Hydroxybutyrate Use in Humans – A Promising Future Therapy?

Therapeutic ketosis is traditionally induced with dietary modification. However, owing to the time delay involved, this is not a practical approach for treatment of acute conditions such as traumatic brain injury. Intravenous administration of ketones would obviate this problem by rapidly inducing ketosis. This has been confirmed in a number of small animal and human studies. Currently no such commercially available product exists. The aim of this systematic review is to review the safety and efficacy of intravenous beta-hydroxybutyrate. The Web of Science, PubMed and EMBASE databases were searched, and a systematic review undertaken. Thirty-five studies were included. The total beta-hydroxybutyrate dose ranged from 30 to 101 g administered over multiple doses as a short infusion, with most studies using the racemic form. Such dosing achieves a beta-hydroxybutyrate concentration &gt;1 mmol/L within 15 min. Infusions were well tolerated with few adverse events. Blood glucose concentrations occasionally were reduced but remained within the normal reference range for all study participants. Few studies have examined the effect of intravenous beta-hydroxybutyrate in disease states. In patients with heart failure, intravenous beta-hydroxybutyrate increased cardiac output by up to 40%. No studies were conducted in patients with neurological disease. Intravenous beta-hydroxybutyrate has been shown to increase cerebral blood flow and reduce cerebral glucose oxidation. Moreover, beta-hydroxybutyrate reduces protein catabolism and attenuates the production of counter-regulatory hormones during induced hypoglycemia. An intravenous beta-hydroxybutyrate formulation is well tolerated and may provide an alternative treatment option worthy of further research in disease states.

Presence of Adult Companion Goats Favors the Rumen Microbial and Functional Development in Artificially Reared Kids

Newborn dairy ruminants are usually separated from their dams after birth and fed on milk replacer. This lack of contact with adult animals may hinder the rumen microbiological and physiological development. This study evaluates the effects of rearing newborn goat kids in contact with adult companions on the rumen development. Thirty-two newborn goat kids were randomly allocated to two experimental groups which were reared either in the absence (CTL) or in the presence of non-lactating adult goats (CMP) and weaned at 7 weeks of age. Blood and rumen samples were taken at 5, 7, and 9 weeks of age to evaluate blood metabolites and rumen microbial fermentation. Next-generation sequencing was carried out on rumen samples collected at 7 weeks of age. Results showed that CTL kids lacked rumen protozoa, whereas CMP kids had an abundant and complex protozoal community as well as higher methanogen abundance which positively correlated with the body weight and blood β-hydroxybutyrate as indicators of the physiological development. CMP kids also had a more diverse bacterial community (+132 ASVs) and a different structure of the bacterial and methanogen communities than CTL kids. The core rumen bacterial community in CMP animals had 53 more ASVs than that of CTL animals. Furthermore, the number of ASVs shared with the adult companions was over 4-fold higher in CMP kids than in CTL kids. Greater levels of early rumen colonizers Proteobacteria and Spirochaetes were found in CTL kids, while CMP kids had higher levels of Bacteroidetes and other less abundant taxa (Veillonellaceae, Cyanobacteria, and Selenomonas). These findings suggest that the presence of adult companions facilitated the rumen microbial development prior to weaning. This accelerated microbial development had no effect on the animal growth, but CMP animals presented higher rumen pH and butyrate (+45%) and ammonia concentrations than CTL kids, suggesting higher fibrolytic and proteolytic activities. CMP kids also had higher blood β-hydroxybutyrate (+79%) and lower blood glucose concentrations (-23%) at weaning, indicating an earlier metabolic development which could favor the transition from pre-ruminant to ruminant after the weaning process. Further research is needed to determine the effects of this intervention in more challenging farm conditions.

Advances in Biosensors for Continuous Glucose Monitoring Towards Wearables

Continuous glucose monitors (CGMs) for the non-invasive monitoring of diabetes are constantly being developed and improved. Although there are multiple biosensing platforms for monitoring glucose available on the market, there is still a strong need to enhance their precision, repeatability, wearability, and accessibility to end-users. Biosensing technologies are being increasingly explored that use different bodily fluids such as sweat and tear fluid, etc., that can be calibrated to and therefore used to measure blood glucose concentrations accurately. To improve the wearability of these devices, exploring different fluids as testing mediums is essential and opens the door to various implants and wearables that in turn have the potential to be less inhibiting to the wearer. Recent developments have surfaced in the form of contact lenses or mouthguards for instance. Challenges still present themselves in the form of sensitivity, especially at very high or low glucose concentrations, which is critical for a diabetic person to monitor. This review summarises advances in wearable glucose biosensors over the past 5 years, comparing the different types as well as the fluid they use to detect glucose, including the CGMs currently available on the market. Perspectives on the development of wearables for glucose biosensing are discussed.