Application of n-of-1 clinical trials in personalized nutrition research: a trial protocol for Westlake N-of-1 Trials for Macronutrient Intake (WE-MACNUTR) =========================================================================================================================================================== * Yunyi Tian * Yue Ma * Yuanqing Fu * Ju-Sheng Zheng ## Abstract **BACKGROUND** Diet and nutrition play essential roles in human health. Personalized dietary recommendations or nutritional advice tailored to each individual can help with more effective disease prevention. N-of-1 trials can provide a pragmatic clinical means of addressing individual postprandial blood glucose variation in response to different food ingredients or nutrients. **OBJECTIVE** To investigate the individual postprandial glucose response to diets with diverse macronutrient proportions at both individual level and population level and the potential of the novel single-patient (n-of-1) trial for the personalization of diet. **DESIGN** Westlake N-of-1 Trials for Macronutrient Intake (WE-MACNUTR) is a multiple crossover feeding trial. Individual response to different dietary patterns in terms of postprandial glucose response is the primary outcome. Secondary outcomes include individual phenotypic response and the effects of dietary ingredients on the composition and structure of gut microbiota. **SETTING** Participants experience three successive 12-day dietary intervention pairs including a 6-day wash-out period before each isocaloric dietary intervention. Two different type of diets (a 6-day high-fat, low-carbohydrate (HF-LC) diet and a 6-day low-fat, high-carbohydrate (LF-HC) diet) are assigned to an individual in a randomized sequence using block randomization with a fixed block size of two. This feeding trial takes place in Hangzhou, China. **PARTICIPANTS** Target enrolment is 30 healthy individuals aged between 18 and 65 years. Exclusion criteria are inability or unwillingness to approved informed consent; other serious medical conditions; food allergy; and no access to a smart phone or computer with an internet connection. **DISCUSSION** This trial addresses the feasibility of n-of-1 approach for personalizing dietary intervention to individuals. The results will help provide personalized dietary recommendation on macronutrients in terms of postprandial blood glucose response. Well-designed n-of-1 trial is likely to become an effective method of optimizing individual health and advancing health care. **TRIAL REGISTRATION** [ClinicalTrials.gov](http://ClinicalTrials.gov) identifier: [NCT04125602](http://medrxiv.org/lookup/external-ref?link_type=CLINTRIALGOV&access_num=NCT04125602&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) ## Introduction Diet and nutrition are key to maintain human health. Previous studies have showed much interest in the metabolic effects of different ratios of dietary fat to carbohydrate intake1-5. Some evidence suggests that a high-fat, low-carbohydrate (HF-LC) diet can improve glycemic control by reducing glycated hemoglobin and fasting glucose, while others support the beneficial effect of a low-fat, high-carbohydrate (LF-HC) diet with particular focus on the quality of carbohydrate1,6-10. One important interpretation for the inconsistent results is the individualized or personalized response to the dietary macronutrient intake, which is also called ‘personalized nutrition’11. Personalized nutrition was born in the context that a conventional ‘one-size-fits-all’ approach usually fail to meet individual’s need for the nutritional requirements12. The general aim of personalized nutrition is to improve health using nutritional, genetic, phenotypic and other information about individuals to develop targeted nutritional advice, services or other products13-15. Although specific dietary recommendation has been made for pregnant women, infants, children, adults or elderly, they are still subgroup recommendation, which is far from the stage of ‘personalization’ or ‘precision’. Application of ‘n-of-1’ clinical trial or ‘single subject’ study represents a new direction in personalized nutrition research (**Figure 1**). It can capture intra-individual variability in health behaviors over time, aiming to identify individual response to a given intervention in a controlled trial, which provides great opportunity to answer the personalization potential of different diets, nutrients or nutrition supplements16-18. The idea of n-of-1 has been applied in special education, psychotherapy, psychology and pharmaceutical studies for decades to test the individual response to specific drugs or treatments19-23. Stunnenberg et al. reported the efficacy of mexiletine on reducing muscle stiffness in patients with nondystrophic myotonia (NDM) using a series of n-of-1 trials24. However, there is no published ‘n-of-1’ study in nutrition field so far. Westlake N-of-1 Trials for Macronutrient Intake (WE-MACNUTR) is a novel clinical trial that uses the macronutrient intake as an exemplar for the new progress in the field. ![Figure 1.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2020/04/20/2020.02.15.20023259/F1.medium.gif) [Figure 1.](http://medrxiv.org/content/early/2020/04/20/2020.02.15.20023259/F1) Figure 1. The Development of Personalized Nutrition. Personalized nutrition was born in the context that a conventional ‘one-size-fits-all’ approach usually fail to meet individual’s need for the nutritional requirements. ‘n-of-1’ clinical trial is a novel study design for the research of personalized nutrition comparing with the traditional designs such as observational study and randomized controlled trial. Integration of multi-omics data including nutrigenomics, proteomics, metabolomics, microbiome and other phenotypes is key for the development of personalized nutrition. We will describe the trial protocol of the WE-MACNUTR: a n-of-1 feeding clinical trial among healthy adults. The feeding trial is composed of two dietary interventions, a HF-LC diet and a LF-HC diet. The primary objective is to investigate the response of postprandial blood glucose to the dietary interventions and the primary outcomes include postprandial maximum glucose (PMG), the area under the curve (AUC24) of postprandial glucose from 0.00 to 24.00 and the mean amplitude of glycemic excursions (MAGE) obtained by measuring the arithmetic mean of the differences between consecutive peaks and nadirs. Secondary objectives of this study include evaluating different phenotypic responses, such as circulating lipid profile changes to a specific diet among individuals and evaluating the impact of different dietary components on the composition and structure of gut microbiota. ## Methods ### Study Design The n-of-1 trial is a multiple crossover feeding trial conducted in a single participant, comparing her/his response to different interventions and assesses the variability in these responses25. In the WE-MACNUTR study, a series of n-of-1 trials are employed simultaneously and a common regimen of interventions is applied to all participants (**Figure 2**). Participants experience three successive 12-day intervention pairs including a 6-day wash-out period between each intervention. The diets are isocaloric and all provided by the researchers, and the primary distinguishing feature is their fat and carbohydrate contents, including a HF-LC diet and a LF-HC diet. Each 12-day intervention pair comprises 6 days of HF-LC (three meals daily) and 6 days of LF-HC (three meals daily) diets in a random order. The order of the diets in each pair is determined using a block randomization. Major investigators and laboratory personnel responsible for the measurements are masked to group allocation. Meal providers are aware of group and diet allocation, but they are not involved in the rest of the trial. Participants are required to complete a basic questionnaire on a daily basis after dinner to summarize their eating behaviors throughout the day. Biological samples including blood, saliva, urine and fecal samples of each participant are collected at baseline and every 6 days during the first 12-day intervention pairs (**Figure 3**). Starting from the second 12-day intervention pair, only saliva, urine and fecal samples are collected every six days. Prior to each intervention period, each participant is required to wear a subcutaneous glucose sensor (on the back of the upper arm) for continuous glucose monitoring (CGM) up to 14 days and an AX3 band that can capture physical activity data from the wrist of each participant. ![Figure 2.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2020/04/20/2020.02.15.20023259/F2.medium.gif) [Figure 2.](http://medrxiv.org/content/early/2020/04/20/2020.02.15.20023259/F2) Figure 2. Flow Diagram of the Westlake N-of-1 Trials for Macronutrient Intake (WE-MACNUTR) Trial. The flowchart summarizes the preparation phase and the first set of the trial. Set 2 and set 3 share the same trial design as set 1 except for no blood sample collection before each wash-out or intervention period. The sequence of two types of six-day dietary interventions in each set is randomized using block randomization as LF-HC and HF-LC diets in set 1; HF-LC and LF-HC diets in set 2 and HF-LC and LF-HC diets in set 3, respectively. HF-LC: high-fat, low-carbohydrate; LF-HC: low-fat, high-carbohydrate. ![Figure 3.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2020/04/20/2020.02.15.20023259/F3.medium.gif) [Figure 3.](http://medrxiv.org/content/early/2020/04/20/2020.02.15.20023259/F3) Figure 3. The Timeline of the Westlake N-of-1 Trials for Macronutrient Intake (WE-MACNUTR) Trial. The timeline illustrates a preparation period for participants recruitment, a baseline data collection period, and three feeding trial periods. The first set of the trial consists of two wash-out periods (highlighted in green) and two randomized dietary intervention periods (highlighted in grey). In all three sets of the trial, both wash-out and intervention periods last for 6 days respectively. The study was approved by the Westlake University Internal Ethical Review Board in Hangzhou, China, and registered with [ClinicalTrials.gov](http://ClinicalTrials.gov) (identifier: [NCT04125602](http://medrxiv.org/lookup/external-ref?link_type=CLINTRIALGOV&access_num=NCT04125602&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom)). Written informed consent are obtained from all study participants. ### Dietary Intervention A dietitian designs the diet for the intervention and the washout period based on the Chinese Dietary Guidelines (2016) and Chinese Dietary Reference Intakes (2013) as well as the participants’ demographic information, eating habits and physical activity levels. Besides, factors such as local food availability, the cooking methods of the Westlake University canteen kitchen, and the current recipes are important factors in the meal planning. Participants are instructed to consume only the provided foods or beverages. #### Wash-out Diet (WD) Prior to each dietary intervention, participants are provided with a 6-day diet as the ‘wash-out diet’ to reduce potential sources of bias and eliminate any carry-over effect of previous intervention. The wash-out diet consists of 30% of total energy (%E) for fat, 15%E for protein and 55%E for carbohydrate based on the acceptable macronutrient distribution range (AMDR)26. #### HF-LC Diet Intervention Throughout the 6-day intervention, participants are provided with a HF-LC diet, including a three-day diet consisting of 60%E for fat, 15%E for protein and 25%E for carbohydrate while the other three-day diet consisting of 70%E for fat, 15%E for protein and 15%E for carbohydrate. #### LF-HC Diet Intervention Throughout the 6-day intervention, participants are provided with a LF-HC diet, including a 3-day diet consisting of 20%E for fat, 15%E for protein and 65%E for carbohydrate while the other 3-day diet consisting of 10%E for fat, 15%E for protein and 75%E for carbohydrate. ### Participants Participants are recruited among students and staffs from Westlake University, Hangzhou, China. Inclusion criteria are as follows: healthy Chinese adults aged between 18 and 65 years; and ability to complete the study. Exclusion criteria are inability or unwillingness to provide informed consent; neurological conditions that might affect the assessment of the study measurement; hospitalization or surgery planned within 3 months; gastrointestinal diseases; other serious medical conditions, such as liver, kidney, or systemic disease; pregnant or lactating women; tobacco, alcohol, or illicit drug abuse; antibiotics uses within two weeks prior to the trial; participants on a vegan diet; food allergy; no access to a smart phone or computer with an internet connection; concurrent intervention study. ### Measures The primary outcomes are (1) postprandial maximum glucose (PMG). PMG is the peak value of CGM within 3 hours after the first bite of a meal or the maximum value of CGM between two meals when the interval is less than 3 hours. (2) area under the curve (AUC24). AUC24 refers to total area under the CGM curve from 0.00 to 24.00.(3) mean amplitude of glycemic excursions (MAGE). MAGE is obtained by measuring the arithmetic mean of the differences between consecutive peaks and nadirs. Secondary outcomes include different phenotypic responses, such as circulating lipid profile and gut microbiome profile, to a specific diet among individuals (**Table 1**). View this table: [Table 1.](http://medrxiv.org/content/early/2020/04/20/2020.02.15.20023259/T1) Table 1. **Outcome and Assessment Points** ### Sample Size Calculation For an individual n-of-1 trial, sample size calculations are not recommended27. In the present study, a total of 30 participants will be enrolled, and Bayesian hierarchical model meta-analysis will be applied to combine the results from each n-of-1 trial generating estimates of intervention effect at population level. However, no formula-based methodology exists for sample size or power calculations for such designs28. Referring to the method reported by Stunnenberg et al., we performed a simulation-based statistical power calculation with modifications24. In brief, a prior distribution for the mean intervention effect (HF-LC vs. LF-HC) on the PMG was prespecified based on results of a previous RCT reported by Parr et al29. In their study, 1.6 mmol/L (0.62 mmol/L vs. 0.78 mmol/L) difference on the PMG was established among overweight/obese participants. From this prior distribution, we drew a random realization and simulated data (3 sets, 2 intervention periods per set and 18 observations per intervention period) accordingly for the 30 n-of-1 trials using R software. Thereafter, all data from these n-of-1 trials were used to perform a Bayesian meta-analysis with flat non-informative priors at population level. The Bayesian meta-analysis resulted in a (marginal) posterior probability for the mean intervention effect. Chan et al. reported a difference of 19 mg/dL (1.05 mmol/L) in daytime PMG between nondiabetic youth and prediabetic counterparts, whereas another study found this difference between youth and middle-aged ones was 10mg/dL (0.56 mmol/L)30,31. Therefore, we determined the posterior probability on both intervention effect of at least 1.05mmol/L and 0.56 mmol/L from the simulation-based Bayesian meta-analysis, and a ≥ 80% posterior probability was treated as positive result. We repeated the above procedures for 1000 times (corresponding with data from 1000 aggregated N-of-1 trials) and the fraction of meta-analyses that returned a positive conclusion (i.e. a >80% posterior probability on a prespecified treatment effect) was treated as measure for the power. Consequently, the power was estimated to be 99.7 % and 100% for detecting the intervention effect of at least 1.05 mmol/L and 0.56 mmol/L respectively. Given the present study focus on the postprandial glucose response to different diet patterns among healthy youths, we set the meaningful difference as 0.56 mmol/L, which may be still of significance for reminding healthy subjects to shift the structure of diet to improve postprandial glucose when this meaningful difference is detected. On the other hand, in order to determine the type I error, we performed another set of simulation under the assumption of no intervention effect. For each virtual participant, we simulated data that allowed the participant to have a fixed mean value and expected variability during both intervention periods. All data from these 30 virtual individual N-of-1 trials was used to perform the Bayesian meta-analysis at population level and a strict ≥ 20% posterior probability on a meaningful intervention effect (i.e., 0.56 mmol/L) was treated as false positive result. After repeating the procedures for 1000 times, the fraction of meta-analyses that returned a false positive conclusion was determined as measure for the type I error rate, which was found to be 0.001. Thus, this simulation-based sample calculation indicates that with 30 participants completing the trial (3 sets, 2 intervention periods per set and 18 observations per intervention period), it will have satisfactory type I error rate and enough power to detect the prespecified meaningful intervention effect. ## Statistical Analysis Plan ### Data Management Even though all participants are students/staffs who routinely have meals in campus, we anticipate that some participants will skip some meals that we provide at dining room and eat other foods instead, which adds uncertainty to the effects on postprandial blood glucose. On this occasion, we will ask the volunteers to record all the other foods they have eaten not provided by the researchers, and we will then perform a sensitivity analysis based on the data. Other major violations, such as failure to complete at least one set of intervention, will prevent statistical analysis at individual level and lead to exclusion of the participants from meta-analysis at population level. ### Primary Analysis The primary analysis of intervention effect is the comparison of the effect of HF-LC diet on postprandial blood glucose level compared with LF-HC diet. At individual level, we use individual intervention effects to guide dietary decisions for each participant. On the other hand, we generate estimates of intervention effect at population level by combing the n-of-1 results with meta-analysis. ### Analysis of Baseline Data Descriptive statistics with demographics and baseline characteristics will be presented for each participant. ### Analysis of Individual N-of-1 Trials Statistical analysis will be performed separately for each n-of-1 trial to estimate intervention effect at individual level. Bayesian models will be applied for estimating the intervention effects (**Figure 4**). Posterior probabilities of outcomes will be calculated using an interface that incorporates open-source R software (3.6.1) and open-source WinBUGS (1.4.3). The results will be reported for primary variables (e.g., AUC of postprandial glucose concentration over 24 hours and peak concentration of postprandial blood glucose), and participants will be provided with an estimate of differences in the variables and the probabilities that the differences are induced by the different interventions (HF-LC vs. LF-HC). ![Figure 4.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2020/04/20/2020.02.15.20023259/F4.medium.gif) [Figure 4.](http://medrxiv.org/content/early/2020/04/20/2020.02.15.20023259/F4) Figure 4. Representation of the Hierarchical Bayesian Estimation for the Primary Outcomes at Both Individual and Population Level (combination of single patient studies). The observed repeated measurements of the postprandial peak glucose on a given patient are combined into a sample mean and sample variance. The model assumes that the patients’ measurements follow a normal distribution centered about that patient’s true mean effect (μi) with variance σ12. At the population level, the various patients’ true mean (μi) are assumed to follow a normal distribution centered about an overall population mean (μ) with between-patient variance τ2. For the Bayesian specification, prior distributions are assigned for β, μ, σ12, and τ2. In the present study, these prior distributions are standard non-informative prior distributions. X means the independent variable: dietary pattern (high-fat and low-carbohydrate vs. low-fat and high-carbohydrate); PMG: the peak value of CGM within 3 hours after the meal or the peak value of CGM between two meals when the interval is less than 3 hours. Secondary and exploratory outcomes will be analyzed similarly. ### Meta-analysis of N-of-1 Trials To combine results of the multiple n-of-1 trials, a Bayesian multi-level model will be used24,32,33. Participant will be treated as a random effect and a common within-participant residual variance will be assumed. Non-informative priors will be applied for all model parameters, with mean parameters using normal prior distributions with very large standard deviations and variation parameters using inverse gamma distributions with both shape and scale parameters equal to 0.01. Using the interface that incorporates R software (3.6.1) and WinBUGS (1.4.3), combining the data from the individual n-of-1 trials will obtain posterior distributions for the mean intervention effect at the population level and for the between-participant variation. Secondary and exploratory outcomes will be analyzed similarly. ## Discussion To advance the field of personalized nutrition, n-of-1 clinical trial appears to be a promising study design to advocate, although the real-world example is rare. We will use the WE-MACNUTR trial as an exemplar to showcase the study design of n-of-1 trial as to test the individualized response to macronutrient intake among healthy adults. The study, if successful, will provide insight into the feasibility of n-of-1 approach to personalize or tailor dietary intervention to individuals. Previous study suggested that the magnitude of postprandial responses to mixed meals depended largely on the total amount of fat and carbohydrates intakes1,5,6.The American Diabetes Association (ADA) recommends monitoring carbohydrate intake to achieve better glycemic control in patients with type 2 diabetes, which is based on studies showing reduced postprandial glucose and triglyceride responses in individuals consuming HF-LC diets34-36. However, previous systematic reviews discussed the effects of HF-LC and LF-HC diets on metabolic risk factors and showed inconsistent results5,7,37. Hsu et al suggested LF-HC diet consisting of high fiber contents showed beneficial effects on glucose and insulin sensitivity among both Asian Americans and Caucasian Americans38. Several studies have reported that both HF-LC and LF-HC diets reduced HbA1c and fasting glucose in obese adults with T2D, while HF-LC diet achieved better improvements in glycemic control1,2. Kamada et al suggested that changes in HbA1c and fasting plasma glucose did not differ significantly between HF-LC and LF-HC diets among Japanese diabetic patients39. The benefits and drawbacks of different dietary patterns on health are under intensive study these days despite the lack of a standardized definition regarding the macronutrient contents40-42. Therefore, n-of-1 trial has a huge potential to explore the main effects of a specific dietary intervention, and identify the factors that influence individual response to nutritional factors. In the present study, it is expected that the trial will provide information on the response of postprandial blood glucose level among individuals to different dietary interventions, namely HF-LC and LF-HC diets, enabling better understanding of intra-individual differences in absorption, distribution and metabolism of macronutrients. Individual human beings are not only unique in the aspect of host genome, but also in the aspect of the gut microbiome, which represents the combined influence of the diet and lifestyles, as well as host genetics43,44. Both animal and human studies have demonstrated that the composition of the gut microbiome can be rapidly affected by a specific dietary component exposure within four days45,46. Furthermore, integration of machine learning algorithms with gut microbiome features have shown powerful potentials to predict one’s response to different dietary patterns in term of postprandial glucose response47. Zeevi *et al*. monitored the postprandial glucose response in a cohort of 800 participants in Israel in response to identical meals. A multidimensional data including gut microbiome features, anthropometrics, blood parameters, and physical activity were integrated into a machine learning algorithm that was capable of predicting personalized postprandial glucose response with the gut microbiota48. These new progresses have stimulated more research on the application and integration of gut microbiome into the personalized nutrition field. Results from this study may provide further evidence suggesting that n-of-1 trial is feasible in characterizing individual microbiome profile. With the aggregated data from isocaloric meal but different carbohydrate to fat ratio, our study makes deeper investigations of the underlying interactions between specific food components and microbiota species possible. Therefore, future methodological studies on developing and implementing effective evaluation of personalized dietary intervention could assist individuals in promoting healthy gut microbiota profile and preventing cardiometabolic diseases. Another strength of the present study with n-of-1 method is its flexibility which enables the study design to be personalized to individuals’ interests and requirements, and its high level of evidence for making clinical decisions for individuals alongside systematic reviews of RCTs. ## Limitations The n-of-1 study does have limitations. Participation in feeding trial such as WE-MACNUTR requires time and effort so the trial cannot be conducted in an ideally controlled setting. Participants will face consciousness, reliability, and persistence over time as they are required to follow restricted dietary patterns with no extra food intake throughout the trial. Any extra snack or beverage intake could affect individual blood glucose level. Besides, slight changes in cooking method, food groups or food ingredients under inevitable circumstances would also jeopardize the final results. ## Conclusions In summary, WE-MACNUTR trial, as an exemplar of nutritional n-of-1 trial, will address the call for new method to advance the field of personalized nutrition. WE-MACNUTR will potentially help clarify the individual postprandial glucose response to diets with diverse macronutrient proportions, and help design and optimize macronutrient composition in long-term dietary intervention studies. The results of WE-MACNUTR will also be helpful in terms of understanding the individual response in gut microbiome to macronutrients. ## Data Availability The corresponding author will provide access to individual de-identified participant data from the present study, upon request by qualified researchers whose proposals are approved. Requests for access to the individual-level data from this study can be submitted via email to zhengjusheng{at}westlake.edu.cn and proposals are required to be attached for approval. ## Funding This study was funded by National Natural Science Foundation of China (81903316) and Westlake University (101396021801) ## Contributors JSZ designed the study and was the principal investigator of WE-MACNUTR; YT, YM, YF and JSZ drafted the manuscript and approved the final version. ## Acknowledgements We thank all the volunteers participating in the Westlake N-of-1 Trials for Macronutrient Intake (WE-MACNUTR). * Received February 15, 2020. * Revision received April 16, 2020. * Accepted April 20, 2020. * © 2020, Posted by Cold Spring Harbor Laboratory This pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at [http://creativecommons.org/licenses/by-nc-nd/4.0/](http://creativecommons.org/licenses/by-nc-nd/4.0/) ## References 1. 1.Tay J, Luscombe-Marsh ND, Thompson CH, et al. Comparison of low- and high-carbohydrate diets for type 2 diabetes management: a randomized trial. Am J Clin Nutr. 2015;102(4):780–790. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiYWpjbiI7czo1OiJyZXNpZCI7czo5OiIxMDIvNC83ODAiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyMC8wNC8yMC8yMDIwLjAyLjE1LjIwMDIzMjU5LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 2. 2.Tay J, Thompson CH, Luscombe-Marsh ND, et al. Effects of an energy-restricted low-carbohydrate, high unsaturated fat/low saturated fat diet versus a high-carbohydrate, low-fat diet in type 2 diabetes: A 2-year randomized clinical trial. Diabetes Obes Metab. 2018;20(4):858–871. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/dom.13164&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29178536&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 3. 3.Kdekian A, Alssema M, Van Der Beek EM, et al. Impact of isocaloric exchanges of carbohydrate for fat on postprandial glucose, insulin, triglycerides, and free fatty acid responses-a systematic review and meta-analysis. Eur J Clin Nutr. 2020;74(1):1–8. 4. 4.Accurso A, Bernstein RK, Dahlqvist A, et al. Dietary carbohydrate restriction in type 2 diabetes mellitus and metabolic syndrome: time for a critical appraisal. Nutr Metab (Lond). 2008;5:9. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/1743-7075-5-9&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=18397522&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 5. 5.Kodama S, Saito K, Tanaka S, et al. Influence of fat and carbohydrate proportions on the metabolic profile in patients with type 2 diabetes: a meta-analysis. Diabetes Care. 2009;32(5):959–965. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoiZGlhY2FyZSI7czo1OiJyZXNpZCI7czo4OiIzMi81Lzk1OSI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDIwLzA0LzIwLzIwMjAuMDIuMTUuMjAwMjMyNTkuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 6. 6.Noakes TD, Windt J. Evidence that supports the prescription oflow-carbohydrate high-fat diets: a narrative review. Br J Sports Med.2017;51(2):133–139. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoiYmpzcG9ydHMiO3M6NToicmVzaWQiO3M6ODoiNTEvMi8xMzMiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyMC8wNC8yMC8yMDIwLjAyLjE1LjIwMDIzMjU5LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 7. 7.Brand-Miller J, Hayne S, Petocz P, Colagiuri S. Low–glycemic index diets in the management of diabetes: a meta-analysis of randomized controlled trials. Diabetes Care. 2003;26(8):2261–2267. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoiZGlhY2FyZSI7czo1OiJyZXNpZCI7czo5OiIyNi84LzIyNjEiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyMC8wNC8yMC8yMDIwLjAyLjE1LjIwMDIzMjU5LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 8. 8.Saslow LR, Mason AE, Kim S, et al. An Online Intervention Comparing a Very Low-Carbohydrate Ketogenic Diet and Lifestyle Recommendations Versus a Plate Method Diet in Overweight Individuals With Type 2 Diabetes: A Randomized Controlled Trial. J Med Internet Res. 2017;19(2):e36. [PubMed](http://medrxiv.org/lookup/external-ref?access_num=28193599&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 9. 9.Bhanpuri NH, Hallberg SJ, Williams PT, et al. Cardiovascular disease risk factor responses to a type 2 diabetes care model including nutritional ketosis induced by sustained carbohydrate restriction at 1 year: an open label, non-randomized, controlled study. Cardiovasc Diabetol. 2018;17(1):56. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/s12933-018-0698-8&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29712560&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 10. 10.Wylie-Rosett J, Aebersold K, Conlon B, Isasi CR, Ostrovsky NW. Health effects of low-carbohydrate diets: where should new research go? Curr Diab Rep. 2013;13(2):271–278. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s11892-012-0357-5&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23266565&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 11. 11.Braconi D, Cicaloni V, Spiga O, Santucci A. An Introduction to Personalized Nutrition. In: Trends in Personalized Nutrition.2019:3–32. 12. 12.Ordovas JM, Ferguson LR, Tai ES, Mathers JC. Personalised nutrition and health. BMJ. 2018;361:k2173. [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29898881&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 13. 13.German JB, Zivkovic AM, Dallas DC, Smilowitz JT. Nutrigenomics and personalized diets: What will they mean for food? Annu Rev Food Sci Technol. 2011;2:97–123. [PubMed](http://medrxiv.org/lookup/external-ref?access_num=22129377&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 14. 14.Valdes AM, Walter J, Segal E, Spector TD. Role of the gut microbiota in nutrition and health. BMJ. 2018;361:36–44. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1136/bmj.k2179&link_type=DOI) 15. 15.Fito M, Melander O, Martinez JA, Toledo E, Carpene C, Corella D. Advances in Integrating Traditional and Omic Biomarkers When Analyzing the Effects of the Mediterranean Diet Intervention in Cardiovascular Prevention. Int J Mol Sci. 2016;17(9). 16. 16.Schork NJ, Goetz LH. Single-Subject Studies in Translational Nutrition Research. Annu Rev Nutr. 2017;37:395–422. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1146/annurev-nutr-071816-064717&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=28715990&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 17. 17.McDonald S, Quinn F, Vieira R, et al. The state of the art and future opportunities for using longitudinal n-of-1 methods in health behaviour research: a systematic literature overview. Health Psychol Rev. 2017;11(4):307–323. 18. 18.Duan N, Kravitz RL, Schmid CH. Single-patient (n-of-1) trials: a pragmatic clinical decision methodology for patient-centered comparative effectiveness research. J Clin Epidemiol. 2013;66(8 Suppl):S21–28. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jclinepi.2013.04.006&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23849149&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 19. 19.Kravitz RL, Schmid CH, Marois M, et al. Effect of Mobile Device-Supported Single-Patient Multi-crossover Trials on Treatment of Chronic Musculoskeletal Pain: A Randomized Clinical Trial. JAMA Intern Med. 2018;178(10):1368–1377. 20. 20.Lillie EO, Patay B, Diamant J, Issell B, Topol EJ, Schork NJ. The n-of-1 clinical trial: the ultimate strategy for individualizing medicine? Per Med. 2011;8(2):161–173. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.2217/pme.11.7&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=21695041&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 21. 21.Senior HE, Mitchell GK, Nikles J, et al. Using aggregated single patient (N-of-1) trials to determine the effectiveness of psychostimulants to reduce fatigue in advanced cancer patients: a rationale and protocol. BMC Palliat Care. 2013;12(1):17–17. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/1472-684X-12-17&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23617794&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 22. 22.Moeller JD, Dattilo J, Rusch F. Applying Quality Indicators to Single-Case Research Designs Used in Special Education: A Systematic Review. Psychol Sch. 2015;52(2):139–153. 23. 23.Norcross JC, Wampold BE. What works for whom: Tailoring psychotherapy to the person. J Clin Psychol. 2011;67(2):127–132. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/jclp.20764&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=21108312&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 24. 24.Stunnenberg BC, Raaphorst J, Groenewoud HM, et al. Effect of Mexiletine on Muscle Stiffness in Patients With Nondystrophic Myotonia Evaluated Using Aggregated N-of-1 Trials. JAMA. 2018;320(22):2344–2353. 25. 25.Zucker DR, Ruthazer R, Schmid CH, et al. Lessons learned combining N-of-1 trials to assess fibromyalgia therapies. The Journal of Rheumatology. 2006;33(10):2069–2077. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NjoianJoZXVtIjtzOjU6InJlc2lkIjtzOjEwOiIzMy8xMC8yMDY5IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjAvMDQvMjAvMjAyMC4wMi4xNS4yMDAyMzI1OS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 26. 26.Trumbo P, Schlicker S, Yates AA, Poos M. Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein and Amino Acids. J Am Diet Assoc. 2002;102(11):1621–1630. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0002-8223(02)90346-9&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=12449285&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000179175200014&link_type=ISI) 27. 27.Design and Impkementation of N-of-1 Trials: A User’s Guide. Rockville, MD: Agency for Healthcare Research and Quality; 2014. 28. 28.Stunnenberg BC, Woertman W, Raaphorst J, et al. Combined N-of-1 trials to investigate mexiletine in non-dystrophic myotonia using a Bayesian approach; study rationale and protocol. BMC Neurol. 2015;15:43. 29. 29.Parr EB, Devlin BL, Callahan MJ, et al. Effects of providing high-fat versus high-carbohydrate meals on daily and postprandial physical activity and glucose patterns: a randomised controlled trial. Nutrients. 2018;10(5):557. 30. 30.Chan CL, Pyle L, Newnes L, Nadeau KJ, Zeitler PS, Kelsey MM. Continuous glucose monitoring and its relationship to hemoglobin A1c and oral glucose tolerance testing in obese and prediabetic youth. The Journal of Clinical Endocrinology & Metabolism. 2015;100(3):902–910. 31. 31.Group JDRFCGMS. Variation of interstitial glucose measurements assessed by continuous glucose monitors in healthy, nondiabetic individuals. Diabetes Care. 2010;33(6):1297–1299. 32. 32.Zucker DR, Schmid CH, McIntosh MW, D’Agostino RB, Selker HP, Lau J. Combining single patient (N-of-1) trials to estimate population treatment effects and to evaluate individual patient responses to treatment. J Clin Epidemiol. 1997;50(4):401–410. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0895-4356(96)00429-5&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=9179098&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1997WX07600005&link_type=ISI) 33. 33.Quintana M, Viele K, Lewis RJ. Bayesian Analysis: Using Prior Information to Interpret the Results of Clinical Trials. JAMA. 2017;318(16):1605–1606. [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29067406&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 34. 34.Jung CH, Choi KM. Impact of High-Carbohydrate Diet on Metabolic Parameters in Patients with Type 2 Diabetes. Nutrients. 2017;9(4). 35. 35.American Diabetes A, Bantle JP, Wylie-Rosett J, et al. Nutrition recommendations and interventions for diabetes: a position statement of the American Diabetes Association. Diabetes Care. 2008;31 Suppl 1:S61–78. [FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6NzoiZGlhY2FyZSI7czo1OiJyZXNpZCI7czoxOToiMzEvU3VwcGxlbWVudF8xL1M2MSI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDIwLzA0LzIwLzIwMjAuMDIuMTUuMjAwMjMyNTkuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 36. 36.Feinman RD, Pogozelski WK, Astrup A, et al. Dietary carbohydrate restriction as the first approach in diabetes management: critical review and evidence base. Nutrition. 2015;31(1):1–13. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.nut.2014.06.011&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=25287761&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 37. 37.Santos FL, Esteves SS, da Costa Pereira A, Yancy WS, Jr.., Nunes JP. Systematic review and meta-analysis of clinical trials of the effects of low carbohydrate diets on cardiovascular risk factors. Obes Rev. 2012;13(11):1048–1066. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/j.1467-789X.2012.01021.x&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=22905670&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 38. 38.Hsu WC, Lau KH, Matsumoto M, Moghazy D, Keenan H, King GL. Improvement of insulin sensitivity by isoenergy high carbohydrate traditional Asian diet: a randomized controlled pilot feasibility study. PLoS One. 2014;9(9):e106851. 39. 39.Kamada C, Yoshimura H, Okumura R, et al. Optimal energy distribution of carbohydrate intake for Japanese elderly patients with type 2 diabetes: the Japanese Elderly Intervention Trial. Geriatr Gerontol Int. 2012;12 Suppl 1:41–49. 40. 40.Naude CE, Schoonees A, Senekal M, Young T, Garner P, Volmink J. Low carbohydrate versus isoenergetic balanced diets for reducing weight and cardiovascular risk: a systematic review and meta-analysis. PLoS One. 2014;9(7):e100652. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1371/journal.pone.0100652&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=25007189&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 41. 41.Hockaday TD, Hockaday JM, Mann JI, Turner RC. Prospective comparison of modified fat-high-carbohydrate with standard low-carbohydrate dietary advice in the treatment of diabetes: one year follow-up study. Br J Nutr. 1978;39(2):357–362. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1079/BJN19780045&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=629924&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1978ER74600012&link_type=ISI) 42. 42.Schulze MB, Martinez-Gonzalez MA, Fung TT, Lichtenstein AH, Forouhi NG. Food based dietary patterns and chronic disease prevention. BMJ. 2018;361:k2396. [FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6MzoiYm1qIjtzOjU6InJlc2lkIjtzOjE3OiIzNjEvanVuMTJfOS9rMjM5NiI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDIwLzA0LzIwLzIwMjAuMDIuMTUuMjAwMjMyNTkuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 43. 43.Zhernakova A, Kurilshikov A, Bonder MJ, et al. Population-based metagenomics analysis reveals markers for gut microbiome composition and diversity. Science. 2016;352(6285):565–569. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjEyOiIzNTIvNjI4NS81NjUiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyMC8wNC8yMC8yMDIwLjAyLjE1LjIwMDIzMjU5LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 44. 44.Celis-Morales C, Livingstone KM, Marsaux CF, et al. Effect of personalized nutrition on health-related behaviour change: evidence from the Food4Me European randomized controlled trial. Int J Epidemiol. 2017;46(2):578–588. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/ije/dyw186&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=27524815&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) 45. 45.David LA, Maurice CF, Carmody RN, et al. Diet rapidly and reproducibly alters the human gut microbiome. Nature. 2014;505(7484):559–563. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/nature12820&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=24336217&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000329995000042&link_type=ISI) 46. 46.Tremaroli V, Backhed F. Functional interactions between the gut microbiota and host metabolism. Nature. 2012;489(7415):242–249. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/nature11552&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=22972297&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000308635900032&link_type=ISI) 47. 47.Cani PD. Human gut microbiome: hopes, threats and promises. Gut.2018;67(9):1716–1725. [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NjoiZ3V0am5sIjtzOjU6InJlc2lkIjtzOjk6IjY3LzkvMTcxNiI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDIwLzA0LzIwLzIwMjAuMDIuMTUuMjAwMjMyNTkuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 48. 48.Zeevi D, Korem T, Zmora N, et al. Personalized nutrition by prediction of glycemic responses. Cell. 2015;163(5):1079–1094. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.cell.2015.11.001&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26590418&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2020%2F04%2F20%2F2020.02.15.20023259.atom)