Design and usability evaluation of user-centered and visual-based aids for dietary food measurement on mobile devices in a randomized controlled trial.

OBJECTIVE: We designed and developed two interactive apps interfaces for dietary food measurements on mobile devices. The user-centered designs of both the IPI (interactive photo interface) and the SBI (sketching-based interface) were evaluated. Four types of outcomes were assessed to evaluate the usability of mobile devices for dietary measurements, including accuracy, absolute weight differences, and the response time to determine the efficacy of food measurements. METHODS: The IPI presented users with images of pre-determined portion sizes of a specific food and allowed users to scan and then select the most representative image matching the food that they were measuring. The SBI required users to relate the food shape to a readily available comparator (e.g., credit card) and scribble to shade in the appropriate area. A randomized controlled trial was conducted to evaluate their usability. RESULTS: A total of 108 participants were randomly assigned into the following three groups: the IPI (n=36) and SBI (n=38) experimental groups and the traditional life-size photo (TLP) group as the control. A total of 18 types of food items with 3-4 different weights were randomly selected for assessment by each type. The independent Chi-square test and t-test were performed for the dichotomous and continuous variable analyses, respectively. The total accuracy rates were 66.98%, 44.15%, and 72.06% for the IPI, SBI, and TLP, respectively. No significant difference was observed between the IPI and TLP, regardless of the accuracy proportion or weight differences. The SBI accuracy rates were significantly lower than the IPI and TLP accuracy rates, especially for several spooned, square cube, and sliced pie food items. The time needed to complete the operation assessment by the user was significantly lower for the IPI than for the SBI. CONCLUSION: Our study corroborates that the user-centered visual-based design of the IPI on a mobile device is comparable the TLP in terms of the usability for dietary food measurements. However, improvements are needed because both the IPI and TLP accuracies associated with some food shapes were lower than 60%. The SBI is not yet a viable aid. This innovative alternative required further improvements to the user interface.

Abundance of Prevotella copri in gut microbiota is inversely related to a healthy diet in patients with type 2 diabetes.

While the gut microbiota is known to be influenced by habitual food intake, this relationship is seldom explored in type 2 diabetes patients. This study aims to investigate the relationship between dietary patterns and gut microbial species abundance in 113 type 2 diabetes patients (mean age, 58 years; body mass index, 29.1; glycohemoglobin [HbA1c], 8.1%). We analyzed the gut microbiota using 16S amplicon sequencing, and all patients were categorized into either the Bacteroides enterotype (57.5%, n = 65) or the Prevotella enterotype (42.5%, n = 48) using the partitioning around medoids clustering algorithm, based on the most representative genera. Patients with the Bacteroides enterotype showed better glycemic control with a 2.71 odds of HbA1c ≤ 7.0% compared to the Prevotella enterotype (95% confidence interval, 1.02-7.87; P, 0.034). Dietary habits and the nutrient composition of all patients were assessed using a validated food frequency questionnaire. It was observed that the amounts of dietary fiber consumed were suboptimal, with an average intake of 16 g per day. Additionally, we extracted four dietary patterns through factor analysis: eating-out, high-sugar foods, fish-vegetable, and fermented foods patterns. Patients with the Bacteroides enterotype had higher scores for the fish-vegetable pattern compared to the Prevotella enterotype (0.17 ± 0.13 versus -0.23 ± 0.09; P, 0.010). We further investigated the relationship between the microbiota and the four dietary patterns and found that only the fish-vegetable dietary pattern scores were correlated with principal coordinate values. A lower pattern score was associated with the accumulated abundance of the 31 significant microbial features. Among these features, Prevotella copri was identified as the most significant by using a random forest model, with an area under the receiver operating characteristic of 0.93 (95% confidence interval, 0.88-0.98). To validate these results, we conducted a custom quantitative polymerase chain reaction assay. This assay confirmed the presence of P. copri (sensitivity, 0.96; specificity, 0.97) in our cohort, with a prevalence of 47.8%, and a mean relative abundance of 21.0% in subjects harboring P. copri. In summary, type 2 diabetes patients with the Prevotella enterotype demonstrated poorer glycemic control and deviations from a healthy dietary pattern. The abundance of P. copri, as a major contributing microbial feature, was associated with the severity in the deficiency in dietary fish and vegetables. Emphasis should be placed on promoting a healthy dietary pattern and understanding the microbial correlations.

Regimen comprising GLP-1 receptor agonist and basal insulin can decrease the effect of food on glycemic variability compared to a pre-mixed insulin regimen.

BACKGROUND: Increasing evidence suggests that glucagon-like peptide 1 (GLP-1) receptor agonists (RA) can stabilize glycemic variability (GV) and interfere with eating behavior. This study compared the impact of insulin, GLP-1 RA, and dietary components on GV using professional continuous glucose monitoring (CGM). METHODS: Patients with type 2 diabetes underwent CGM before and after switching from a twice-daily pre-mixed insulin treatment regimen to a GLP-1 RA (liraglutide) plus basal insulin regimen. The dietary components were recorded and analyzed by a certified dietitian. The interactions between the medical regimen, GV indices, and nutrient components were analyzed. RESULTS: Sixteen patients with type 2 diabetes were enrolled in this study. No significant differences in the diet components and total calorie intake between the two regimens were found. Under the pre-mixed insulin regimen, for increase in carbohydrate intake ratio, mean amplitude of glucose excursion (MAGE) and standard deviation (SD) increased; in contrast, under the new regimen, for increase in fat intake ratio, MAGE and SD decreased, while when the protein intake ratio increased, the coefficient of variation (CV) decreased. The impact of the food intake ratio on GV indices disappeared under the GLP-1 RA regimen. After switching to the GLP-1 RA regimen, the median MAGE, SD, and CV values decreased significantly. However, the significant difference in GV between the two regimens decreased during the daytime. CONCLUSION: A GLP-1 RA plus basal insulin regimen can stabilize GV better than a regimen of twice-daily pre-mixed insulin, especially in the daytime, and can diminish the effect of food components on GV.

Gut Microbial Signatures for Glycemic Responses of GLP-1 Receptor Agonists in Type 2 Diabetic Patients: A Pilot Study.

Backgrounds: Glucagon-like peptide-1 receptor agonist (GLP-1 RA) is probably one of more effective antidiabetic agents in treatment of type 2 diabetes mellitus (T2D). However, the heterogenicity in responses to GLP-1 RA may be potentially related to gut microbiota, although no human evidence has been published. This pilot study aims to identify microbial signatures associated with glycemic responses to GLP-1 RA. Materials and Methods: Microbial compositions of 52 patients with T2D receiving GLP-1 RA were determined by 16S rRNA amplicon sequencing. Bacterial biodiversity was compared between responders versus non-responders. Pearson's correlation and random forest tree algorithm were used to identify microbial features of glycemic responses in T2D patients and multivariable linear regression models were used to validate clinical relevance. Results: Beta diversity significantly differed between GLP-1 RA responders (n = 34) and non-responders (n = 18) (ADONIS, P = 0.004). The top 17 features associated with glycohemoglobin reduction had a 0.96 diagnostic ability, based on area under the ROC curve: Bacteroides dorei and Roseburia inulinivorans, the two microbes having immunomodulation effects, along with Lachnoclostridium sp. and Butyricicoccus sp., were positively correlated with glycemic reduction; Prevotella copri, the microbe related to insulin resistance, together with Ruminococcaceae sp., Bacteroidales sp., Eubacterium coprostanoligenes sp., Dialister succinatiphilus, Alistipes obesi, Mitsuokella spp., Butyricimonas virosa, Moryella sp., and Lactobacillus mucosae had negative correlation. Furthermore, Bacteroides dorei, Lachnoclostridium sp. and Mitsuokella multacida were significant after adjusting for baseline glycohemoglobin and C-peptide concentrations, two clinical confounders. Conclusions: Unique gut microbial signatures are associated with glycemic responses to GLP-RA treatment and reflect degrees of dysbiosis in T2D patients.

Usability of Food Size Aids in Mobile Dietary Reporting Apps for Young Adults: Randomized Controlled Trial.

BACKGROUND: Young adults are more likely to use self-managed dietary reporting apps. However, there is scant research examining the user experience of different measurement approaches for mobile dietary reporting apps when dealing with a wide variety of food shapes and container sizes. OBJECTIVE: Field user experience testing was conducted under actual meal conditions to assess the accuracy, efficiency, and subjective reaction of three food portion measurement methods embedded in a developed mobile app. Key-in-based aid (KBA), commonly used in many current apps, relies on the user's ability to key in volumes or weights. Photo-based aid (PBA) extends traditional assessment methods, allowing users to scroll, observe, and select a reduced-size image from a set of options. Gesture-based aid (GBA) is a new experimental approach in which the user makes finger movements on the screen to roughly describe food portion boundaries accompanied by a background reference. METHODS: A group of 124 young adults aged 19 to 26 years was recruited for a head-to-head randomized comparison and divided into 3 groups: a KBA (n=42) control group and PBA (n=41) and GBA (n=41) experimental groups. In total, 3 meals (ie, breakfast, lunch, and dinner) were served in a university cafeteria. Participants were provided with 25 dishes and beverages for selection, with a variety of food shapes and containers that reflect everyday life conditions. The accuracy of and time spent on realistic interaction during food portion estimation and the subjective reaction of each aid were recorded and analyzed. RESULTS: Participants in the KBA group provided the highest accuracy in terms of hash brown weight (P=.004) and outperformed PBA or GBA for many soft drinks in cups. PBA had the best results for a cylindrical hot dog (P<.001), irregularly shaped pork chop (P<.001), and green tea beverage (660 mL; P<.001). GBA outperformed PBA for most drinks, and GBA outperformed KBA for some vegetables. The GBA group spent significantly more time assessing food items than the KBA and PBA groups. For each aid, the overall subjective reaction based on the score of the System Usability Scale was not significantly different. CONCLUSIONS: Experimental results show that each aid had some distinguishing advantages. In terms of user acceptance, participants considered all 3 aids to be usable. Furthermore, users' subjective opinions regarding measurement accuracy contradicted the empirical findings. Future work will consider the use of each aid based on food or container shape and integrate the various advantages of the 3 different aids for better results. Our findings on the use of portion size aids are based on realistic and diverse food items, providing a useful reference for future app improvement of an effective, evidence-based, and acceptable feature. TRIAL REGISTRATION: International Standard Randomized Controlled Trial Registry ISRCTN36710750; http://www.controlled-trials.com/ISRCTN36710750.

Association of unhealthy dietary behaviors with renal function decline in patients with diabetes.

OBJECTIVE: Balanced nutrition is important for patients with diabetes, and nutrition might well influence diabetes-related complications, although there is limited evidence for this supposition at present. Consequently, we investigate the association between dietary behaviors and renal function decline among patients with diabetes. RESEARCH DESIGN AND METHODS: From 2011 to 2013, a total of 2797 patients with type 2 diabetes participated in the Diabetes Shared Care Program at Chang Gung Memorial Hospital. All received nutritional consulting by dieticians and an eight-item list of unhealthy dietary behaviors, which included the excessive intake of carbohydrates, fats, protein, fruit, pickled foods, dessert and alcohol, as well as inadequate dietary vegetable. Estimated glomerular filtration rate (eGFR) decline ≥40% was defined as a surrogate end point for kidney damage. Independent dietary risk factors predicting poor renal outcomes were assessed. RESULTS: Stable mean glycated hemoglobin (A1c) (7.78% to 7.75%, p=0.151), improved cholesterol (174.04 to 170.13 mg/dL, p<0.001) and low-density lipoprotein (104.19 to 98.07 mg/dL, p<0.001) were found in patients throughout 2 years of therapy. However, significant eGFR decline was noted (94.20 to 88.08 mL/min/1.73 m2, p<0.001). A total of 125 subjects had eGFR decline ≥40% and 2672 had stable renal progression.In regression analysis, 625 stable renal patients (selected via propensity score matching) and 125 subjects with eGFR decline ≥40% demonstrated excessive pickled foods to be predictive of poor renal outcomes (OR 1.861, 95% CI 1.230 to 2.814, p=0.003). CONCLUSIONS: Our study suggests that excessive pickled foods deteriorate renal function more than other unhealthy dietary behaviors in patients with diabetes.

Design and Usability Evaluation of Mobile Voice-Added Food Reporting for Elderly People: Randomized Controlled Trial.

BACKGROUND: Advances in voice technology have raised new possibilities for apps related to daily health maintenance. However, the usability of such technologies for older users remains unclear and requires further investigation. OBJECTIVE: We designed and evaluated two innovative mobile voice-added apps for food intake reporting, namely voice-only reporting (VOR) and voice-button reporting (VBR). Each app features a unique interactive procedure for reporting food intake. With VOR, users verbally report the main contents of each dish, while VBR provides both voice and existing touch screen inputs for food intake reporting. The relative usability of the two apps was assessed through the metrics of accuracy, efficiency, and user perception. METHODS: The two mobile apps were compared in a head-to-head parallel randomized trial evaluation. A group of 57 adults aged 60-90 years (12 male and 45 female participants) was recruited from a retirement community and randomized into two experimental groups, that is, VOR (n=30) and VBR (n=27) groups. Both groups were tested using the same set of 17 food items including dishes and beverages selected and allocated to present distinct breakfast, lunch, and dinner meals. All participants used a 7-inch tablet computer for the test. The resulting data were analyzed to evaluate reporting accuracy and time efficiency, and the system usability scale (SUS) was used to measure user perception. RESULTS: For eight error types identified in the experiment, the VBR group participants were significantly (P<.001) more error prone owing to the required use of button-tapping actions. The highest error rates in the VOR group were related to incomprehensible reporting speech (28/420, 6.7%), while the highest error rates in the VBR group were related to failure to make required button taps (39/378, 10.3%). The VOR group required significantly (P<.001) less time to complete food reporting. The overall subjective reactions of the two groups based on the SUS surpassed the benchmark and were not significantly different (P=.20). CONCLUSIONS: Experimental results showed that VOR outperformed VBR, suggesting that voice-only food input reporting is preferable for elderly users. Voice-added apps offer a potential mechanism for the self-management of dietary intake by elderly users. Our study contributes an evidence-based evaluation of prototype design and selection under a user-centered design model. The results provide a useful reference for selecting optimal user interaction design. TRIAL REGISTRATION: International Standard Randomized Controlled Trial Registry ISRCTN17335889; http://www.isrctn.com/ISRCTN17335889.

Impact of Carbohydrate on Glucose Variability in Patients with Type 1 Diabetes Assessed Through Professional Continuous Glucose Monitoring: A Retrospective Study.

INTRODUCTION: The aim of this study was to objectively analyze the correlation between dietary components and blood glucose variation by means of continuous glucose monitoring (CGM). METHODS: Patients with type 1 diabetes mellitus (T1DM) who received CGM to manage their blood glucose levels were enrolled into the study, and the components of their total caloric intake were analyzed. Glycemic variation parameters were calculated, and dietary components, including percentages of carbohydrate, protein and fat in the total dietary intake, were analyzed by a dietitian. The interaction between parameters of glycemic variability and dietary components was analyzed. RESULTS: Sixty-one patients with T1DM (33 females, 28 males) were enrolled. The mean age of the participants was 34.7 years, and the average duration of diabetes was 14 years. Glycated hemoglobin before CGM was 8.54%. Participants with a carbohydrate intake that accounted for < 50% of their total caloric intake had a longer DM duration and a higher protein and fat intake than did those with a carbohydrate intake that accounted for ≥ 50% of total caloric intake, but there was no between-group difference in total caloric intake per day. The group with a carbohydrate intake that accounted for < 50% of their total caloric intake also had lower nocturnal continuous overlapping net glycemic action (CONGA) 1, - 2 and - 4 values. The percentage of protein intake had a slightly negative correlation with mean amplitude of glycemic excursions (MAGE) (r = - 0.286, p < 0.05) and a moderately negative correlation with coefficient of variation (CV) (r = 0.289, p < 0.05). One additional percentage of protein calories of total calories per day decreased the MAGE to 4.25 mg/dL and CV to 0.012 (p < 0.05). The optimal dietary protein percentage for MAGE < 140 mg/dL was 15.13%. The performance of predictive models revealed the beneficial effect of adequate carbohydrate intake on glucose variation when combined with protein consumption. CONCLUSIONS: Adequate carbohydrate consumption-but not more than half the daily total calories-combined with protein calories that amount to approximately 15% of the daily caloric intake is important for glucose stability and beneficial for patients with T1DM.

Amino acids and wound healing in people with limb-threatening diabetic foot ulcers.

BACKGROUND: Amino acids are associated with wound healing in traumatic wounds and burns, although their effects on healing in patients with diabetic foot ulcers (DFUs) are limited. This study aimed to evaluate and identify specific amino acids associated with healing outcomes of patients with DFUs. METHODS: Sixty-two out of 85 patients who completed the in-hospital treatment for limb-threatening DFUs were enrolled. All ulcers had epithelialization without clinical evidence of infection at discharge. The patients and their families were instructed on foot-care techniques and committed to regular follow-up for 1 year. Baseline characteristics, PEDIS wound classification, laboratory data and serum amino acid levels were used to analyze their predictive power. RESULTS: Fifty-seven patients completed the study in which 38 had healed and 19 had unhealed ulcers. The unhealed group had higher incidence of coronary artery disease and larger wound size. Most patients received endovascular therapy (81.6% healed group; 78.9% unhealed group) before enrollment. Following adjustments for clinical factors, the serum levels of arginine (326.4 µmol/L vs. 245.0 µmol/L, P = 0.045), isoleucine (166.7 µmol/L vs. 130.1 µmol/L, P = 0.019), leucine (325.8 µmol/L vs. 248.9 µmol/L, P = 0.039), and threonine (186.7 µmol/L vs. 152.0 µmol/L, P = 0.019) were significantly higher in the healed group. CONCLUSIONS: The amino acids associated with wound healing in DFUs differ from those reported for traditional traumatic wounds. These findings affirm the necessity for future large-scaled studies for the application of these amino acids in DFU healing, either as prognostic predictors or supplemented regimens.

Evaluating Mobile Health Apps for Customized Dietary Recording for Young Adults and Seniors: Randomized Controlled Trial.

BACKGROUND: The role of individual-tailored dietary recording in mobile phone health apps has become increasingly important in management of self-health care and population-based preventive service. The development of such mobile apps for user-centered designing is still challengeable and requires further scientific evidence. OBJECTIVE: This study aims to conduct a randomized trial to assess the accuracy and time efficiency of two prototypes for dietary recoding utilization related to the input method of food intake. METHODS: We first present an innovative combinatorial concept for dietary recording to account for dish variation. One prototype was a self-chosen tab app that featured choosing each food ingredient to synthesize an individual dish, whereas the other was an autonomous exhaustive list app that provided one selection from a comprehensive list of dish items. The concept included commercially available choices that allowed users to more accurately account for their individual food selection. The two mobile apps were compared in a head-to-head parallel randomized trial evaluation. Young adults (n=70, aged 18-29) and older adults (n=35, aged 55-73) were recruited and randomized into two groups for accuracy and response time evaluation based on 12 types of food items in use of the developed self-chosen tab and autonomous exhaustive list apps, respectively. RESULTS: For the trials based on the self-chosen tab (53 participants) and autonomous exhaustive list groups (52 participants), the two prototypes were found to be highly accurate (>98%). The self-chosen tab app was found to be more efficient, requiring significantly less time for input of 11 of 12 items (P<.05). The self-chosen tab users occasionally neglected to select food attributes, an issue which did not occur in the autonomous exhaustive list group. CONCLUSIONS: Our study contributes through the scientific evaluation of the transformation step into prototype development to demonstrate that a self-chosen tab app has potentially better opportunity in effectiveness and efficiency. The combinatorial concept offers potential for dietary recording and planning which can account for high food item variability. Our findings on prototype development of diversified dietary recordings provide design consideration and user interaction for related further app development and improvement. TRIAL REGISTRATION: ISRCTN Registry ISRCTN86142301; http://www.isrctn.com/ ISRCTN86142301 (Archived by WebCite at http://www.webcitation.org/74YLEPYnS).

The impact of nutritional status on treatment outcomes of patients with limb-threatening diabetic foot ulcers.

AIM: This study aimed to investigate the nutritional status of patients with limb-threatening diabetic foot ulcers (DFUs) and its impact on treatment outcomes. METHODS: A total of 478 consecutive patients (mean age, 65.4years) treated for limb-threatening DFUs were enrolled. Nutritional status assessment using the Mini Nutritional Assessment (MNA) and Geriatric Nutritional Risk Index (GNRI) was performed by three qualified dieticians within 48hours of admission. Limb-preservation outcomes were stratified into major lower extremity amputation (LEA) (above the ankle, n=33), minor LEA (distal to ankle, n=117) and no amputation (non-LEA, n=328). RESULTS: Most patients were identified as being at risk of malnutrition (70.5%) or malnourished (14.6%) (mean MNA score, 20.6±3.4). MNA scores decreased with increasing severity of LEA (mean, 21.1, 20.0, and 17.9, respectively; P for linear trend <0.001), associated inversely with the tendency to require LEA (P for linear trend was 0.001), and associated independently with both major and minor LEA outcomes (adjusted odds ratio [aOR]=0.80, 95% confidence interval [CI], 0.65-0.99, P=0.042 and aOR=0.89, 95% CI, 0.80-0.99, P=0.032, respectively). The predictive value was sustained in patients younger than age 65years. Though GNRI results had similar associations with outcomes, its predictive value was limited in minor LEA and younger population. CONCLUSIONS: Patients' nutritional status was shown to have significant influence on limb-preservation outcomes for limb-threatening DFUs. Nutritional assessment of this patient population using the MNA is recommended.