A diary is useful when it fits into your day.
In dietary self-monitoring, you write down what you ate, and then you can look back at the patterns. In a review of 59 studies, most lifestyle programmes that used detailed or simplified logging showed a more favourable weight change than the control group. The methods were very different. So the review cannot identify one logging form that is best for everyone.
A 2021 review summarised 39 studies of adults with overweight or obesity. They tracked eating, physical activity or weight with digital tools. In 50 of the 68 analysed associations (74%), more frequent digital self-monitoring was associated with greater weight loss. This does not prove that more frequent logging alone caused the weight loss.
That is why Oregano Prime accepts photos, text and dictation. These options are for convenience. How much time the app saves, and how much it helps you keep logging, must be measured separately.
Reducing the effort of use needs separate attention.
In a 2025 review of 39 studies, the time needed and technical difficulties were recurring barriers. Simplicity, personalised feedback, regular contact and support appeared as helpful factors. The review mainly examined live, remote weight management programmes, so the result applies only partly to a standalone app.
The photo is a starting point. Your correction is valuable data.
A 2025 study compared three language models on 52 food photos. Even for the two better models, the mean absolute percentage error of the energy estimate was about 36%. They regularly underestimated larger portions. This is the result of the tested models, not the error rate of Oregano Prime.
Another study evaluated 195 foods with different amounts of background data. The mean absolute energy error of the image-only estimate was 123 kcal. With detailed ingredient and quantity data, it decreased to 53 kcal. The quality of the reference data varied, and the study did not test the correction process of real users. But it shows that information that the image does not show is important.
In Oregano Prime, you can see what the app understood, and you can correct the result in words. Your approval accepts the estimate that is shown. It does not make it a laboratory measurement.
[2] Fridolfsson et al. (2025) [3] Rodríguez-Jiménez et al. (2025)
Recognition and calculation are separate tasks.
Recognising the food, determining the portion and calculating the nutrients are separate sources of error. In the DietAI24 research, image recognition was connected to an external food database. On the tested data sets, this gave a smaller error than several compared methods. The research was based mainly on American foods, and it did not evaluate Oregano Prime.
Oregano Prime uses the available label and database data. The system calculates the totals from the quantities and the nutrient data. When there is no suitable data source, the value stays an estimate. A traceable calculation shows the origin, but alone it does not correct a wrong food or portion.
Personalisation can give more than general advice.
In the European Food4Me study, advice adapted to the diet gave a more favourable change than general advice in several dietary indicators over six months. 1269 people completed the study. Adding genetic data and some other personal data did not clearly improve the result further. This was a personalised nutrition programme, not a study of a generative AI coach*.
Oregano Prime uses your recorded meals and your own goal for its feedback. The plan that you choose can include an AI review or advice from a human dietitian. We label the two roles separately.
More messages alone are not enough.
In the SMARTER study of 502 participants, personalised daily messages added to logging did not give significantly greater weight loss over 12 months than in the comparison group that also logged. All participants got initial counselling and digital tools.
A 2024 review evaluated the data of 19 studies with 3261 participants in total. Feedback showed a small advantage in the pooled comparisons of physical activity. Because the methods were different, the weight and diet results could not be pooled in the same way. Comparisons of human and algorithm-generated feedback gave mixed results. These studies do not prove the effectiveness of a generative AI coach* or of Oregano Prime.
That is why, for the coach, the content and a next step that you can take are important too. The three-day frequency of Oregano Prime is a product decision. Without a separate study, we cannot say that it is better than another schedule. The advice also depends on diary activity, on day closing and on recent dietitian guidance.
For useful AI advice, the background is important too.
A 2026 JMIR study compared two ways of using the same GPT-5.1 model in 30 simulated cases of recurrent kidney stones. Three experts, who did not know the origin of the answers, evaluated the recommendations of StoneAgent, which was built on professional guidelines and structured case data. These recommendations followed the guidelines better and were more useful in practice than the answers of general model use. The advantage remained when the cases were given as everyday questions.
In Oregano Prime, feedback is based on your logged meals, your goal and the available dietitian guidance. So the AI coach* can give next steps that relate to your own situation.
Choose a diet and a schedule for your own goal.
In the DIETFITS study, 609 adults without diabetes got a programme that supported a healthy diet with reduced fat or reduced carbohydrate. After 12 months, there was no significant difference in weight change between the two groups. Both groups could take part in 22 education sessions. This does not prove that every diet is equally suitable.
The 2026 Cochrane review evaluated the data of 22 studies and 1995 people. Compared with usual dietary advice, intermittent fasting may make little or no difference to weight loss. The certainty of several results is low, and follow-up was at most 12 months.
In Oregano Prime, you can record your goal and the fasting plan that you choose. The timer follows the plan that you set. It does not measure the physiological state of your body, and it does not promise a separate result from fasting.
Tracking gives feedback.
According to a review of 18 randomised studies, changing water intake can have advantages in some situations. The number of studies of weight was small, and they used different groups and interventions. This does not give a water amount that applies to everyone, or a weight-loss effect of a water diary on its own.
Oregano Prime shows how much water you recorded. This is not a measurement of your total fluid intake or of your hydration.
Regular weighing can be useful as part of a behaviour change programme with several elements. In a meta-analysis, only one study evaluated the effect of self-weighing alone; it did not show a clear advantage. With enough measurements, Oregano Prime also shows a smoothed trend. This helps you review the change; it is not a treatment on its own.
Cook once, plan for several meals.
In a study of 40,554 French adults, planning meals in advance was associated with better diet quality and more variety. The study was observational: it did not prove that planning caused the difference. It did not measure the effect of batch cooking or of portion handling in an app.
In Oregano Prime, you can already reuse earlier and saved foods. Saving a cooked quantity on its own and logging a proportional part of it later is a planned feature. This is a different task from copying a meal.
Sources
- Raber et al. (2021). A systematic review of the use of dietary self-monitoring in behavioural weight loss interventions
- Fridolfsson et al. (2025). Performance Evaluation of 3 Large Language Models for Nutritional Content Estimation from Food Images
- Rodríguez-Jiménez et al. (2025). Image-Based Dietary Energy and Macronutrients Estimation with ChatGPT-5
- Yan et al. (2025). DietAI24 as a framework for comprehensive nutrition estimation using multimodal large language models
- Celis-Morales et al. (2017). Effect of personalized nutrition on health-related behaviour change: the Food4Me European randomized controlled trial
- Burke et al. (2022). The Effect of Tailored, Daily, Smartphone Feedback to Lifestyle Self-Monitoring on Weight Loss at 12 Months: the SMARTER Randomized Clinical Trial
- Wang et al. (2026). Evaluating a Guideline-Integrated Clinical Interaction Framework Vs a Standard Large Language Model Interaction for Dietary Recommendations in Recurrent Urolithiasis: In Silico Study
- Gardner et al. (2018). The DIETFITS Randomized Clinical Trial
- Garegnani et al. (2026). Intermittent fasting for adults with overweight or obesity
- Hakam et al. (2024). Outcomes in Randomized Clinical Trials Testing Changes in Daily Water Intake
- Madigan et al. (2015). Is self-weighing an effective tool for weight loss: a systematic literature review and meta-analysis
- Ducrot et al. (2017). Meal planning is associated with food variety, diet quality and body weight status in a large sample of French adults
- Whitehall et al. (2025). A systematic review of influences on engagement with remote health interventions targeting weight management for individuals living with excess weight
- Patel et al. (2021). Self-Monitoring via Digital Health in Weight Loss Interventions: A Systematic Review Among Adults with Overweight or Obesity
- Krukowski et al. (2024). Impact of feedback generation and presentation on self-monitoring behaviors, dietary intake, physical activity, and weight: a systematic review and meta-analysis