Is MyNetDiary accurate? What four independent studies found

Sergey Oreshko - Co-founder and CEO of MyNetDiary

Four independent studies have tested MyNetDiary's accuracy and quality since 2013. In the latest, MyNetDiary earned the highest quality score among 17 apps, and its average energy estimate came within 5 kcal of a dietitian's reference. Here is what each study found and what the results mean.

Since 2013, four university research groups in the United States, Australia, the United Kingdom and Turkey have evaluated MyNetDiary alongside the other popular nutrition apps of their day, and we had no role in any of them. We publish this page because that independent record is a better answer to whether MyNetDiary is accurate than anything we could say ourselves. Our own 2026 study is here too, in its own section and labeled as company research, and a short section covers clinical trials in which MyNetDiary was the study tool.

Highest quality score among 17 apps. Energy within 5 kcal of the reference.
Independent 2026 validation study, Nutrition and Health. MyNetDiary's quality score was 4.39 of 5, the highest of the 17 apps evaluated, and its three-day mean energy estimate was 2,001 kcal against the dietitian-calculated 2,005 kcal reference.

Independent research evaluating MyNetDiary

Other research involving MyNetDiary

2026 validation | 2024 study | 2015 assessment | 2013 review | Our JMIR study | How to read the results | Clinical trials | FAQ

What did the 2026 Nutrition and Health study find about MyNetDiary?

In the 2026 Nutrition and Health study, MyNetDiary had the highest quality score of the 17 apps evaluated (4.39 of 5), and its three-day mean energy estimate for dietitian-designed menus came within 5 kcal of the reference (2,001 vs 2,005 kcal). Only MyNetDiary and FatSecret came that close; ten of 13 apps overestimated, by 327 kcal per day on average.

How the 2026 study was run

Researchers at Istanbul Gelisim University's Department of Nutrition and Dietetics identified 807 apps in Apple's App Store in Turkey in July 2025; 17 met their inclusion criteria. Two trained raters scored each app on the Mobile App Rating Scale (MARS) and counted features across seven domains. For accuracy, a registered dietitian designed three 2,000 kcal menus without Turkish dishes, so they would match English-language food databases. One researcher logged every menu in the 13 apps with manual entry. Each app's energy and macronutrient totals were then compared with reference values from BEBIS, a nutrition analysis system. Only free-to-download apps and free features were assessed, so Meal Scan and AI Coach were out of scope. The MyNetDiary version tested was 9.7.4.

MyNetDiary's 2026 results

The authors attribute the differences between apps mainly to database quality and estimation methods, and note that a high quality score did not, in general, predict accuracy; MyNetDiary was the only app among the five highest-rated whose energy deviation was under 1 percent.

What the 2026 authors caution

The apps were iOS only and drawn from one country's App Store, and only free-to-download apps were included. The menus were standardized and entered by one researcher, so the results describe how each app's database performs under controlled entry. The MARS involves some subjective judgment, and the 17 apps are the functionally eligible subset of the 807 identified. The authors say the accuracy findings should not be generalized directly to real-life use.

Citation: Bayram HM, Ozturkcan A. Quality and nutritional accuracy of mHealth nutrition apps in Türkiye: A validation study using food records. Nutrition and Health. Published online July 10, 2026. DOI 10.1177/02601060261469016. Abstract free on PubMed (42429442); full text at the publisher.

What did the 2024 University of Sydney study in Nutrients find?

In the 2024 University of Sydney study in Nutrients, MyNetDiary tied for the third-highest quality score of 18 apps (4.29 of 5) and for the highest engagement score. Its Western-menu energy estimate, 490 kJ (117 kcal) per day above the dietitian's record, was among the five closest of 16 apps; its Asian and Australian-guideline estimates fell further below the record.

How the 2024 study was run

The University of Sydney team screened the 200 top free and 200 top paid Health and Fitness apps in each of Australia's Apple and Google stores. The 18 nutrition apps with a food diary were rated by three researchers on the Mobile App Rating Scale and the App Behavior Change Scale (ABACUS), which counts behavior-change features. For accuracy, dietitians created three-day food records for a Western diet, an Asian diet and a Recommended diet built to the Australian Dietary Guidelines. The researchers logged each record in 16 apps and compared the totals with the records analyzed in Foodworks.online. Photo recognition was tested separately in seven apps; Meal Scan, a Premium feature, was not part of that test. Only free features were evaluated. The MyNetDiary version tested was 9.11.

MyNetDiary's 2024 results

The authors attribute the underestimation of the Asian menu across apps to food databases built around United States foods, with limited Asian items and portion-size options, and note that every app underestimated the Recommended menu. Their practice recommendations include that dietitians prescribe the more accurate manual-logging apps for client self-monitoring and make sure clients are trained in thorough food record entry.

Limits of the 2024 study

Only free features were evaluated, so Premium functions and some nutrient outputs were not assessed. The menus were designed by dietitians and entered by trained researchers. Photo recognition was tested under controlled lighting and distance. The three raters were young and technology-literate, and the reference data were Australian. The authors also note that matching Asian foods and portions was difficult in most apps.

Citation: Li X, Yin A, Choi HY, Chan V, Allman-Farinelli M, Chen J. Evaluating the Quality and Comparative Validity of Manual Food Logging and Artificial Intelligence-Enabled Food Image Recognition in Apps for Nutrition Care. Nutrients. 2024;16(15):2573. DOI 10.3390/nu16152573. Open access.

What did the 2015 JMIR mHealth and uHealth quality assessment find?

The 2015 JMIR mHealth and uHealth assessment of 28 weight-loss apps ranked MyNetDiary's two apps second (tied) and fourth. Both scored 27 of 32 on the study's scientific coverage and accuracy scale, the second-highest score, and they were the only two apps in the study that provided individualized, tailored feedback.

What the 2015 assessment scored

University of Sydney and University of Leeds researchers screened the top free and paid Health and Fitness apps in each Australian store in August 2014. A dietitian assessor used the 28 weight-management apps with food tracking for five days each (30 percent double-rated) and scored them out of 100: scientific coverage and accuracy (32 points), behavior change techniques (26), usability (20), technology-enhanced features (14) and accountability (8). She also logged her own three-day weighed food record into the 23 apps that calculate energy intake.

MyNetDiary's 2015 results

The authors judged the category as a whole to be suboptimal in quality, with limited scientific content and few behavior change techniques. MyNetDiary's scores sat well above the averages on every criterion except accountability, which measured whether an app displayed author credits and credentials, information sources, sponsorship and recent updates. The energy comparison (table above) used 2014 versions of the apps, one assessor's Australian foods and predominantly United States food databases, and the authors note that matching foods and serving sizes was a general source of error across apps.

Limits of the 2015 assessment

A single dietitian assessor, whose training may have made food matching easier than it would be for the public; apps from one country's stores; and 2014 versions of every app. The authors conclude that recommending weight-loss apps in clinical practice should proceed with caution and that efficacy trials are needed.

Citation: Chen J, Cade JE, Allman-Farinelli M. The Most Popular Smartphone Apps for Weight Loss: A Quality Assessment. JMIR mHealth and uHealth. 2015;3(4):e104. DOI 10.2196/mhealth.4334. PubMed 26678569. Open access.

What did the 2013 American Journal of Preventive Medicine study find?

MyNetDiary's two iPhone apps led the 2013 American Journal of Preventive Medicine review of 30 weight-loss apps, including 13 of the 20 evidence-based strategies coded (65 percent) against 25 percent for the next-highest apps. The authors advised physicians who refer patients to weight-loss apps to select free apps such as MyNetDiary with the highest proportion of evidence-based strategies.

How the 2013 review worked

A University of Massachusetts Medical School team, with colleagues at Boston University and Rosalind Franklin University, took the top 100 paid and top 100 free health and fitness apps in the iTunes and Android stores in January 2012 and kept the 30 weight-loss apps that let users log weight, diet and physical activity. Two assessors coded each app for 20 behavioral strategies drawn from the Diabetes Prevention Program, such as goal setting, calorie balance, portion control and problem solving, and for seven technology-assisted features such as a barcode scanner and reminders. Agreement was 92 percent before discrepancies were resolved. The study did not rate accuracy, usability or weight-loss outcomes.

MyNetDiary's 2013 results

The authors attribute MyNetDiary's score largely to its in-app library, which they describe as the source of many of the strategies they found. They report that paid apps were no more likely than free apps to include evidence-based strategies. Seven of the 20 strategies, including stress reduction, relapse prevention and nutrition label reading, appeared in no app at all, MyNetDiary included. The study also included an early Android version of MyNetDiary, which scored well below the two iPhone apps.

What the 2013 review did not cover

The apps were January 2012 versions drawn from the top 100 lists only. The coding scheme was derived from the Diabetes Prevention Program curriculum instead of an established taxonomy, and whether every strategy is necessary for an intervention to work is unknown. Features available at extra cost, companion websites and coaching were not evaluated. The study made no rating of the reliability or validity of any app's nutrition database, of usability, or of outcomes.

Citation: Pagoto S, Schneider K, Jojic M, DeBiasse M, Mann D. Evidence-Based Strategies in Weight-Loss Mobile Apps. American Journal of Preventive Medicine. 2013;45(5):576-582. DOI 10.1016/j.amepre.2013.04.025. PubMed 24139770. Abstract free on PubMed; full text at the publisher. The findings were reported by NBC News on October 8, 2013.

What has MyNetDiary's own research found?

In a 2026 Journal of Medical Internet Research cohort study of 5,132 MyNetDiary users, people who tracked food more days per week lost more weight over six months, with 35 percent higher odds of losing at least 5 percent of body weight per additional tracking day. The authors are MyNetDiary's founder and lead dietitian, and the design cannot establish causation.

Who was in the JMIR study

The cohort was drawn from 412,193 people who signed up in May 2025 and kept those with a weight-loss goal, at least 21 food-logging days and a recorded weight every month through November: 5,132 users, 1.2 percent of new signups. Every result below describes that selected, engaged group. Mean age was 42.1; 72.8 percent were women; mean starting weight was 91.8 kg (about 202 lb).

What the JMIR study found

Across the whole cohort, mean loss was 9.3 kg (about 20.5 lb, or 9.5 percent of body weight); 70.1 percent lost at least 5 percent and 44.8 percent lost at least 10 percent, while 9.0 percent gained weight. Users whose longest consecutive streak exceeded 90 days lost 11.55 percent on average, against 4.21 percent for those whose longest streak was a week or less. The adjusted odds ratio per additional tracking day was 1.35 (95 percent CI 1.31 to 1.40), adjusted for age, sex and starting BMI. An anonymized user-level replication dataset was published with the paper. A plain-language walkthrough is in our article on the study.

Limits of the JMIR study

The design shows association only: more motivated users may both track more and lose more. A selected cohort: participants were older and heavier at the start than the users the filters excluded. Diet quality, physical activity, sleep and medication use, including GLP-1 receptor agonists, were not measured, so some of the loss may reflect concurrent treatment. And it is a company study: both authors are MyNetDiary employees, which the paper discloses, and which is why it has its own section on this page.

Citation: Oreshko S, Heikkinen S. Association Between Daily Food-Tracking Frequency and Clinically Significant Weight Loss: Retrospective Cohort Study of 5132 Mobile App Users. Journal of Medical Internet Research. 2026;28:e98144. DOI 10.2196/98144. Open access.

Is MyNetDiary accurate? How to read these study results

MyNetDiary's results in four independent studies. Details and limitations are in the sections above.

MyNetDiary's energy estimate came within 5 kcal of the dietitian's reference in the 2026 test, and its scores were at or near the top of each sample across four studies. Those results record app quality and database accuracy under controlled conditions; they do not show that any app is the most accurate, and none measured nutrient depth or real-world logging.

Quality scales measure user experience and information quality. The accuracy tests measure how an app's database and portion logic perform when trained researchers enter set menus, which is a different thing from a week of real-world logging.

What none of the studies measured

None of the studies measured nutrient depth, which is where MyNetDiary differs most from other trackers: the quality scales record only whether an app tracks nutrients, the accuracy tests compared energy and macronutrients only, and the 2013 authors state that they did not rate any app's nutrition database. MyNetDiary's staff-verified food database, built on USDA and NCC research-grade sources, holds more than 2 million foods and tracks up to 108 nutrients per food entry. How we measure nutrient completeness, the number of nutrient fields populated per entry, and how we verify food data are explained in our guide to calorie tracking app accuracy and database quality and in the account of how the database is built. That is our own claim, stated separately from the study findings above.

How the findings have shaped MyNetDiary

We treat the studies as a work list. The 2013 and 2015 papers describe app versions from 2012 and 2014; since then the food database has grown to more than 2 million staff-verified foods, with 2,500 to 3,500 reviewed and updated every day, and the app has added Meal Scan, AI Coach, GLP-1 Companion and Advanced Autopilot, a personalized metabolism adjustment algorithm. The site now discloses the dietitian team and data sources that the 2015 accountability criteria found missing. Our own 2026 study found longer logging streaks associated with greater weight loss, and streak tracking was added in summer 2026; the 2024 results on Asian and Australian menus are guiding our work on international foods and portion sizes. Every new independent study will be added to this page, favorable or not.

Which clinical trials have used MyNetDiary as a study tool?

Two university trials have participants log food, activity, and weight in MyNetDiary: an NIH-funded type 2 diabetes trial led by Virginia Tech (2026 preprint) and an NIH-funded rural obesity trial at the University of Kansas Medical Center (in progress). Three more university studies are under way. MyNetDiary is one part of each program, so the results are not MyNetDiary outcomes.

The Virginia Tech type 2 diabetes trial (2026 preprint)

MyNetDiary was the self-monitoring tool in a 24-week, fully remote randomized trial funded by the National Institute of Diabetes and Digestive and Kidney Diseases: 90 adults with type 2 diabetes and obesity received Premium accounts and were asked to log diet and activity daily and weight weekly, alongside diabetes education and one-on-one coaching. Participants rated the app 4.2 to 4.5 out of 5 for helpfulness, on par with or just behind the live coaching sessions. The manuscript is a preprint and has not yet been peer-reviewed.

Citation: Stein JS, Brown JM, Tegge AN, Snyder HS, Stuart MN, Bickel WK, Rockwell MS, Epling JW, Epstein LH. Episodic Future Thinking in a Fully Remote, Multi-Component Behavioral Intervention for Type 2 Diabetes: A Pilot/Feasibility Trial in Rural and Non-Rural Adults. PsyArXiv preprint, 2026, not yet peer-reviewed. DOI 10.31234/osf.io/9ebhn_v1. ClinicalTrials.gov NCT05280925.

The University of Kansas Medical Center rural obesity trial (in progress)

RE-TOOL, a five-year cluster randomized trial funded by the National Cancer Institute, compares team-based telemedicine care with standard physician visits for 560 patients with obesity across 16 rural Kansas primary care practices. Participants in both arms track their food in MyNetDiary, and coaches in the telemedicine arm review the logs and give feedback in the app. The protocol was published in 2025; the registry lists an estimated completion date of October 2026, and no results have been posted yet.

Citation: Brown AR, Ellerbeck EF, Sullivan DK, Nelson EL, Klemp JR, Gajewski BJ, Saint Onge JM, Befort CA. Protocol for the rural engagement in TelemedTeam for options in obesity treatment solutions (RE-TOOL): Cluster randomized trial investigating team-based telemedicine in rural primary care. Contemporary Clinical Trials Communications. 2025. DOI 10.1016/j.conctc.2025.101559. Open access. ClinicalTrials.gov NCT05710510.

Three more studies under way

MyNetDiary is also the self-monitoring tool in three more university studies still in progress; each will be added here once it is public. MyNetDiary had no role in designing, running, or analyzing any of these trials.

Start with free,
accurate tracking today

Also available as www.plateai.com