Weight Loss with an AI-Powered Digital Platform for Lifestyle Intervention.

BACKGROUND: Lifestyle intervention remains the cornerstone of weight loss programs in addition to pharmacological or surgical therapies. Artificial intelligence (AI) and other digital technologies can offer individualized approaches to lifestyle intervention to enable people with obesity to reach successful weight loss. METHODS: SureMediks, a digital lifestyle intervention platform using AI, was tested by 391 participants (58% women) with a broad range of BMI (20-78 kg/m2), with the aim of losing weight over 24 weeks in a multinational field trial. SureMediks consists of a mobile app, an Internet-connected scale, and a discipline of artificial intelligence called Expert system to provide individualized guidance and weight-loss management. RESULTS: All participants lost body weight (average 14%, range 4-22%). Almost all (98.7%) participants lost at least 5% of body weight, 75% lost at least 10%, 43% at least 15%, and 9% at least 20%, suggesting that this AI-powered lifestyle intervention was also effective in reducing the burden of obesity co-morbidities. Weight loss was partially positively correlated with female sex, accountability circle size, and participation in challenges, while it was negatively correlated with sub-goal reassignment. The latter three variables are specific features of the SureMediks weight loss program. CONCLUSION: An AI-assisted lifestyle intervention allowed people with different body sizes to lose 14% body weight on average, with 99% of them losing more than 5%, over 24 weeks. These results show that digital technologies and AI might provide a successful means to lose weight, before, during, and after pharmacological or surgical therapies.

Long-Term Effects on QT Prolongation of Pretomanid Alone and in Combinations in Patients with Tuberculosis.

Concentration-QTc modeling was applied to pretomanid, a new nitroimidazooxazine antituberculosis drug. Data came from eight phase 2 and phase 3 studies. Besides pretomanid alone, various combinations with bedaquiline, linezolid, moxifloxacin, and pyrazinamide were considered; special attention was given to the bedaquiline-pretomanid-linezolid (BPaL) regimen that has demonstrated efficacy in the Nix-TB study in subjects with extensively drug-resistant or treatment-intolerant or nonresponsive multidrug-resistant tuberculosis. Three heart rate corrections to QT were considered: Fridericia's QTcF, Bazett's QTcB, and a population-specific correction, QTcN. QTc increased with the plasma concentrations of pretomanid, bedaquiline's M2 metabolite, and moxifloxacin in a manner described by a linear model in which the three slope coefficients were constant across studies, visits within study, and times postdose within visit but where the intercept varied across those dimensions. The intercepts tended to increase on treatment to a plateau after several weeks, a pattern termed the secular trend. The slope terms were similar for the three QTc corrections, but the secular trends differed, suggesting that at least some of the secular trend was due to the elevated heart rates of tuberculosis patients decreasing to normal levels on treatment. For pretomanid 200 mg once a day (QD) alone, a typical steady-state maximum concentration of drug in plasma (Cmax) resulted in a mean change from baseline of QTcN of 9.1 ms, with an upper 90% confidence interval (CI) limit of 10.2 ms. For the BPaL regimen, due to the additional impact of the bedaquiline M2 metabolite, the corresponding values were 13.6 ms and 15.0 ms. The contribution to these values from the secular trend was 4.0 ms.

Machine Learning to Identify Predictors of Glycemic Control in Type 2 Diabetes: An Analysis of Target HbA1c Reduction Using Empagliflozin/Linagliptin Data.

INTRODUCTION: Outcomes in type 2 diabetes mellitus (T2DM) could be optimized by identifying which treatments are likely to produce the greatest improvements in glycemic control for each patient. OBJECTIVES: We aimed to identify patient characteristics associated with achieving and maintaining a target glycated hemoglobin (HbA1c) of ≤ 7% using machine learning methodology to analyze clinical trial data on combination therapy for T2DM. By applying a new machine learning methodology to an existing clinical dataset, the practical application of this approach was evaluated and the potential utility of this new approach to clinical decision making was assessed. METHODS: Data were pooled from two phase III, randomized, double-blind, parallel-group studies of empagliflozin/linagliptin single-pill combination therapy versus each monotherapy in patients who were treatment-naïve or receiving background metformin. Descriptive analysis was used to assess univariate associations between HbA1c target categories and each baseline characteristic. After the descriptive analysis results, a machine learning analysis was performed (classification tree and random forest methods) to estimate and predict target categories based on patient characteristics at baseline, without a priori selection. RESULTS: In the descriptive analysis, lower mean baseline HbA1c and fasting plasma glucose (FPG) were both associated with achieving and maintaining the HbA1c target. The machine learning analysis also identified HbA1c and FPG as the strongest predictors of attaining glycemic control. In contrast, covariates including body weight, waist circumference, blood pressure, or other variables did not contribute to the outcome. CONCLUSIONS: Using both traditional and novel data analysis methodologies, this study identified baseline glycemic status as the strongest predictor of target glycemic control attainment. Machine learning algorithms provide an hypothesis-free, unbiased methodology, which can greatly enhance the search for predictors of therapeutic success in T2DM. The approach used in the present analysis provides an example of how a machine learning algorithm can be applied to a clinical dataset and used to develop predictions that can facilitate clinical decision making.

What did this study look at? This study looked at whether a computer program could predict which people with type 2 diabetes would respond best to a particular treatment. The study treatment was a single-pill combination of two medicines, empagliflozin [em-PAH-gli-FLOW-zin] and linagliptin [LYNN-nah-GLIP-tin]. It is used to lower blood sugar (blood glucose) in people with type 2 diabetes. The researchers used machine learning to analyze data from people who received this treatment. Machine learning uses computer models to find patterns in information. The results helped to predict which people might respond best to the treatment. Who took part in this study? The researchers looked at results collected from two earlier studies of the treatment. 1363 people took part. Approximately half of them were male. Their average age was 55 years. Approximately half of them had not received any previous diabetes treatment, and approximately half (50.3%) had received metformin treatment for diabetes. What did the study show? The researchers found that two blood tests commonly used in clinical practice helped them predict who would have the best response to treatment. These tests were HbA1c levels (a measure of long-term blood glucose control), and their fasting plasma glucose (blood glucose levels when they had not eaten for 10­16 h). This study suggests that machine learning could be a useful tool to help doctors decide which treatments will work best for individuals with type 2 diabetes.

The cost of cardiovascular-disease-related death in patients with type 2 diabetes mellitus.

OBJECTIVES: To assess the magnitude of difference in all-cause healthcare resource utilization (HCRU) and costs between patients with type 2 diabetes mellitus (T2DM) who died from a cardiovascular disease (CVD)-related cause in the year preceding death vs. those who did not die during this same period. METHODS: A large US administrative claims database was used to identify patients with T2DM who died of a CVD-related cause from July 2012 to April 2015. These patients were matched 1:1 to patients with T2DM who did not die, using direct matching methods. HCRU and costs were assessed in each of the four quarters (Q4: 12-10 months; Q3: 9-7 months; Q2: 6-4 months; and Q1: 3-0 months) prior to death and compared between patient cohorts using paired t-tests and McNemar's tests. RESULTS: A final matched cohort of 7648 patients who died and 7648 patients who did not die were identified. A significantly higher proportion of patients who died utilized inpatient services vs. those who did not die (Q4: 12.6% vs. 4.6%, p < .001; Q3: 14.6% vs. 4.6%, p < .001; Q2: 17.6% vs. 5.5%, p < .001; and Q1: 65.0% vs. 10.1%, p < .001). In addition, patients who died incurred significantly higher all-cause costs (Q4: $8882 vs. $3970, p < .001; Q3: $10,462 vs. $3661, p < .001; Q2: $12,564 vs. $4169, p < .001; and Q1: $36,076 vs. $6319, p < .001). CONCLUSIONS: T2DM patients with a CVD-related death had significantly greater HCRU and costs in the year including and preceding death compared to those who did not die.

Patient Characteristics are not Associated with Clinically Important Differential Response to Dapagliflozin: a Staged Analysis of Phase 3 Data.

INTRODUCTION: This study aimed to determine if data mining methodologies could identify reproducible predictors of dapagliflozin-specific treatment response in the phase 3 clinical program dataset. METHODS: Baseline and early treatment response variables were selected and data mining used to identify/rank all variables associated with reduction in glycated hemoglobin (HbA1c) at week 26. Generalized linear modeling was then employed using an independent dataset to identify which (if any) variables were predictive of dapagliflozin-specific treatment response as compared with treatment response in the study's control arm. The most parsimonious (i.e., simplest) model was validated by meta-analysis of nine other trials. This staged approach was used to minimize risk of type I errors. RESULTS: From the large dataset, 22 variables were selected for model generation as potentially predictive for dapagliflozin-specific reduction in HbA1c. Although baseline HbA1c was the variable most strongly associated with reduction in HbA1c at study end (i.e., the best prognostic variable), baseline fasting plasma glucose (FPG) was the only predictive dapagliflozin-specific variable in the model. Placebo-adjusted treatment effect of dapagliflozin plus metformin vs. metformin alone for change in HbA1c from baseline was -0.65% at the average baseline FPG of 192.3 mg/dL (10.7 mmol/L). This response changed by -0.32% for every SD [57.2 mg/dL (3.2 mmol/L)] increase in baseline FPG. Effect of baseline FPG was confirmed in the meta-analysis of nine studies, but the magnitude was smaller. No other variable was independently predictive of a dapagliflozin-specific reduction in HbA1c. CONCLUSIONS: This methodology successfully identified a reproducible baseline predictor of differential response to dapagliflozin. Although baseline FPG was shown to be a predictor, the effect size was not of sufficient magnitude to suggest clinical usefulness in identifying patients who would uniquely benefit from dapagliflozin treatment. The findings do support potential benefit for dapagliflozin treatment that is consistent with current recommended use.

Cortical thickness of the cognitive control network in obesity and successful weight loss maintenance: a preliminary MRI study.

Cortical thickness of the cognitive control network was contrasted between obese (OB), successful weight loss maintainers (SWLM), and lean individuals. OB individuals had significant thinning, most notably in the anterior cingulate and posterior parietal cortices. SWLM individuals exhibited trends towards thicker cortex than OB individuals, which may be important in future studies.

Brain response to food stimulation in obese, normal weight, and successful weight loss maintainers.

As many people struggle with maintenance of weight loss, the study of successful weight loss maintainers (SWLM) can yield important insights into factors contributing to weight loss maintenance. However, little research has examined how SWLM differ from people who are obese or normal weight (NW) in brain response to orosensory stimulation. The goal of this study was to determine if SWLM exhibit different brain responses to orosensory stimulation. Brain response to 1-min orosensory stimulation with a lemon lollipop was assessed using functional magnetic resonance imaging among 49 participants, including SWLM (n = 17), NW (n = 18), and obese (n = 14) controls. Significant brain responses were observed in nine brain regions, including the bilateral insula, left inferior frontal gyrus, left putamen, and other sensory regions. All regions also exhibited significant attenuation of this response over 1 min. The SWLM exhibited greater response compared with the other groups in all brain regions. Findings suggest that the response to orosensory stimulation peaks within 40 s and attenuates significantly between 40 and 60 s in regions associated with sensation, reward, and inhibitory control. Greater reactivity among the SWLM suggests that greater sensory reactivity to orosensory stimulation, increased anticipated reward, and subsequently greater inhibitory processing are associated with weight loss maintenance.

Industry funded clinical trials: bias and quality.

The quality of the clinical data supporting the development and ultimately the approval for medical use of new drugs is often challenged. Many share the perception that the business goals of the pharmaceutical industry overrule the best scientific efforts to accrue critical knowledge on a new molecule, in order to inform investment of resources, regulatory approvals and appropriate use by patients. Despite this common belief, few scientists have attempted to assess objectively the quality of industry funded (IF) clinical trials by measuring it and comparing it with non-industry funded (NIF) clinical trials in a data-driven fashion. Overall, the average quality of IF clinical research has been reported to be higher than the quality of NIF clinical research.

Acquired differences in brain responses among monozygotic twins discordant for restrained eating.

We studied whether self-reported intent to exert cognitive control over eating was associated with differences in brain response to food cues, independent of genetic background. Subjects were ten pairs of identical twins in which one twin was a restrained eater and the co-twin was unrestrained, as classified by the Herman and Polivy Restraint Scale. Before and after ingestion of a milkshake, we used functional magnetic resonance imaging to measure brain response to photographs of objects, "fattening" food, and "non-fattening" food. At baseline, restrained eaters had greater activation in the left amygdala and the right thalamus in response to fattening food cues than did their unrestrained co-twins. When restrained eaters drank a milkshake, activation in response to fattening food photographs decreased across multiple brain areas, whereas activation induced by non-fattening food photographs increased. As compared to their unrestrained co-twins, restrained eaters who drank a milkshake had greater decreases in activation by fattening food images in the left amygdala and occipital lobe, and greater increases in activation by non-fattening food images in the medial orbitofrontal cortex. Because of the discordant monozygotic twin study design, the findings provide a rigorous level of support for the hypothesis that adopting an intention to restrain eating alters brain response to food cues.

Psychoactive effects of tasting chocolate and desire for more chocolate.

The purpose of this study was to characterize the psychoactive effects of tasting chocolate and to evaluate the contribution of the main chocolate components to the desire to consume more of it. A total of 280 participants, (F-155; M=125) ranging in age from 18-65, completed the study. Participants were randomly assigned to taste 12.5 g of either white chocolate ("control") or one of four chocolate ("cocoa") samples varying in sugar, fat and percent cocoa content, then answered the question: "Do you want more of this chocolate?" and "If yes, how many more pieces of this chocolate would you like to eat?" They completed pre- and post-consumption surveys, consisting of 30 questions derived from the Addiction Research Center Inventory (ARCI) subscales, Morphine-Benzedrine Group (MBG), Morphine (M) and Excitement (E). Significant decreases in post-pre consumption changes in MBG subscale were observed between the control sample and the 70% cocoa (p=0.046) or the 85% cocoa sample (p=0.0194). Proportionally more men than women wanted more of the tasted chocolate (p=0.035). Participants were more likely to want more of the tasted chocolate if they displayed a greater change in the MBG scale, and if their chocolate sample had high sugar and cocoa content, as assessed by multiple logistic regression. Our results suggest that multiple characteristics of chocolate, including sugar, cocoa and the drug-like effects experienced, play a role in the desire to consume chocolate.

Differential functional magnetic resonance imaging response to food pictures in successful weight-loss maintainers relative to normal-weight and obese controls.

BACKGROUND: Prior research indicates that successful weight-loss maintainers (SWLs) work harder than people of normal weight to maintain their weight loss, including greater dietary restriction of fat and higher physical activity levels. However, little work to date has examined how SWLs differ biologically from normal-weight (NW) and obese controls. OBJECTIVE: The objective was to compare the brain responses of SWLs to food pictures with those of NW and obese controls. DESIGN: Blood oxygen level-dependent responses to high- and low-energy food pictures were measured in 18 NW controls, 16 obese controls, and 17 SWLs. RESULTS: Group differences were identified in 4 regions, which indicated significant change in activation in response to the food pictures. SWLs showed greater activation in the left superior frontal region and right middle temporal region than did NW and obese controls-a pattern of results confirmed in exploratory voxel-wise analyses. Obese controls also showed greater activation in a bilateral precentral region. CONCLUSIONS: These results suggest that SWLs show greater activation in frontal regions and primary and secondary visual cortices-a pattern consistent with greater inhibitory control in response to food cues and greater visual attention to the food cues. A greater engagement of inhibitory control regions in response to food cues as well as a greater monitoring of foods may promote control of food intake and successful weight-loss maintenance.

Temporal relationship between depressive symptoms and cognitive impairment: the Italian Longitudinal Study on Aging.

The temporal relationship between depression and cognitive decline has not been extensively investigated in prospective population-based studies, and most of these have only looked in one direction. We estimated the bidirectional temporal relationship between depressive symptoms and cognitive function in older subjects, excluding subjects with a clinical diagnosis of dementia or mild cognitive impairment (MCI). In a total of 2,963 individuals from the Italian Longitudinal Study on Aging, depressive symptoms, global cognitive function, and episodic memory were measured. Dementia, Alzheimer's disease, vascular dementia, and MCI were classified using current clinical criteria. Depressive symptoms at baseline were associated with an accelerated global cognitive function decline and an accelerated rate of episodic memory delayed recall decline in a 3.5-year follow-up. Finally, an accelerated increase with time of depressive symptoms during the same follow-up period was not associated with global cognitive function and episodic memory (immediate and delayed recall). In older subjects non-cognitively impaired, depressive symptoms at baseline predicted change over time of global cognitive decline and episodic memory delayed recall. Global cognitive function and episodic memory at baseline were not associated with the course of depressive symptoms during the follow-up.

Inhaled insulin is associated with prolonged enhancement of glucose disposal in muscle and liver in the canine.

Diabetic patients treated with inhaled insulin exhibit reduced fasting plasma glucose levels. In dogs, insulin action in muscle is enhanced for as long as 3 h after insulin inhalation. This study was designed to determine whether this effect lasts for a prolonged duration such that it could explain the effect observed in diabetic patients. Human insulin was administered via inhalation (Exubera; n = 9) or infusion (Humulin R; n = 9) in dogs using an infusion algorithm that yielded matched plasma insulin kinetics between the two groups. Somatostatin was infused to prevent insulin secretion, and glucagon was infused to replace basal plasma levels of the hormone. Glucose was infused into the portal vein at 4 mg/kg/min and into a peripheral vein to maintain the arterial plasma glucose level at 160 mg/dl. Arterial and hepatic sinusoidal insulin and glucose levels were virtually identical in the two groups. Notwithstanding, glucose utilization was greater when insulin was administered by inhalation. At its peak, the peripheral glucose infusion rate was 4 mg/kg/min greater in the inhalation group, and a 50% difference between groups persisted over 8 h. Inhalation of insulin caused a greater increase in nonhepatic glucose uptake in the first 3 h after inhalation; thereafter, net hepatic glucose uptake was greater. Inhalation of insulin was associated with greater than expected (based on insulin levels) glucose disposal. This may explain the reduced fasting glucose concentrations observed in humans after administration of certain inhaled insulin formulations compared with subcutaneous insulin.

Depressive symptoms, vascular risk factors and mild cognitive impairment. The Italian longitudinal study on aging.

AIMS: We evaluated the impact of depressive symptoms on the rate of incident mild cognitive impairment (MCI) after a 3.5-year follow-up, and we assessed the interaction between depressive symptoms and vascular risk factors for incident MCI. METHODS: A total of 2,963 individuals from a sample of 5,632 65- to 84-year-old subjects were cognitively and functionally evaluated at the 1st and 2nd surveys of the Italian Longitudinal Study on Aging, a prospective cohort study with a 3.5-year follow-up. MCI and dementia were classified using current clinical criteria. Depressive symptoms were measured with the Geriatric Depression Scale. RESULTS: Among the 2,963 participants, 139 prevalent MCI cases were diagnosed at the 1st survey. During the 3.5-year follow-up, 105 new events of MCI were diagnosed. We did not observe any significant association between depressive symptoms and incident MCI (RR = 1.25, 95% CI = 0.85-1.84, chi(2) = 1.30, p < 0.25). No sociodemographic variables or vascular risk factors modified the relationship between depressive symptoms and incident MCI. CONCLUSION: In our population, depressive symptoms were not associated with the rate of incident MCI. Our findings did not support a role of sociodemographic variables or vascular risk factors in the link between depressive symptoms and incident MCI.

Less activation in the left dorsolateral prefrontal cortex in the reanalysis of the response to a meal in obese than in lean women and its association with successful weight loss.

BACKGROUND: We previously found that obese men have less activation in the left dorsolateral prefrontal cortex (LDLPFC) in response to a meal than do lean men, which indicates an association between this altered neuronal response and the pathophysiology of obesity. OBJECTIVES: The objectives of the study were to extend this finding in obese women and to investigate activity in this region in women with a history of severe obesity who have successfully lost weight (ie, formerly obese women, sometimes called postobese women). DESIGN: We reanalyzed previously collected data to compare postmeal (after receiving a liquid meal) with premeal (after a 36-h fast) regional cerebral blood flow, a marker of neuronal activity, by using (15)O-water positron emission tomography in 10 lean [26 +/- 6% body fat (BF)], 9 obese (39 +/- 3%BF) and 8 formerly obese (28 +/- 4%BF) right-handed women. Data were analyzed by using a 2-level, random-effect analysis of variance. RESULTS: The regional cerebral blood flow in the LDLPFC differed in response to the meal across the 3 groups (P < 0.001, uncorrected for multiple comparisons). Post hoc group comparisons showed that obese women had significantly less activation in this area than did lean and formerly obese women. No significant difference between formerly obese and lean women was found. CONCLUSIONS: These results extend our previous findings, indicating that obese women have less activation in the LDLPFC in response to a meal than do lean or formerly obese women. Neuronal activity in this region did not differ significantly between the latter 2 groups. Longitudinal studies are needed to determine whether these differences in neuronal activity change with or predict weight change.

Incident occurrence of depressive symptoms among patients with mild cognitive impairment - the Italian longitudinal study on aging.

AIMS: We estimated the prevalence and incidence rates of depressive symptoms and the role of vascular risk factors and sociodemographic variables on the occurrence of depressive symptoms in mild cognitive impairment (MCI). METHODS: In the Italian Longitudinal Study on Aging, 2,963 individuals from 5,632 65- to 84-year-old subjects were evaluated at the 1st and 2nd survey, with a 3.5-year follow-up. Dementia and MCI were classified using current criteria. Depressive symptoms were measured with the Geriatric Depression Scale. RESULTS: Among the 2,963 participants, 139 prevalent MCI cases were diagnosed with a depressive symptoms prevalence rate of 63.3%. During the 3.5-year follow-up, we estimated an incidence rate of depressive symptoms of 29.6 per 100 person-years. No sociodemographic variables or vascular risk factors modified the incidence of depressive symptoms in cognitively stable MCI patients or in MCI patients who reverted to normal cognition. CONCLUSION: In our population, there was a high prevalence and incidence of depressive symptoms in MCI. Our findings do not provide support for a possible role of vascular risk factors in the development of depressive symptoms in MCI, although these findings were based on relatively few cases of MCI, limiting the capacity to draw definitive conclusions.