Methodology

Last updated: August 2, 2026

About this page

Here we openly document where INVY gets its numbers, what formulas it uses, what reference ranges it displays, and which studies underpin its recommendations.

Important to understand upfront:

• INVY does not diagnose any condition
• INVY does not prescribe treatment
• INVY does not replace consultation with an endocrinologist or primary care physician
• INVY is an educational tool for tracking your own metabolism and building sustainable habits

If your numbers concern you, discuss them with your doctor. INVY exports your data as a PDF for your visit.

Data Sources

Nutrition data

• Open Food Facts — open database (openfoodfacts.org), over 3M products, ODbL license
• USDA FoodData Central — official USDA nutrition database (fdc.nal.usda.gov)
• Ukrainian dishes — INVY's own curated database

Glycemic Index (GI) and Glycemic Load (GL)

• International Tables of Glycemic Index and Glycemic Load Values — Sydney University (glycemicindex.com)
• Primary publication: Atkinson FS, Foster-Powell K, Brand-Miller JC. Diabetes Care. 2008;31(12):2281–2283 [1]
• 2021 update: Atkinson FS et al. Am J Clin Nutr. 2021;114(5):1625–1632 [2]
• Over 4000 foods in current version

Insulin Index (II)

• Original Insulin Index of Foods — Holt SH, Brand Miller JC, Petocz P. Am J Clin Nutr. 1997;66(5):1264–1276 [3]
• Extension for composite meals — Bao J et al. Am J Clin Nutr. 2009;90(4):986–992 [4]
• Protein-containing foods — Bell KJ et al. Eur J Clin Nutr. 2014;68(9):1055–1059 [5]
• Clinical validation — Bell KJ et al. Diabetes Technol Ther. 2016;18(4):218–225 [6]

For foods without measured II values, INVY uses its own predictive algorithm based on food category, macronutrient composition, and published coefficients from Bao 2009 and Bell 2014. Every II value in the app is tagged with a reliability marker:

• measured — measured in published studies
• approximated — calculated using the composite-meal formula
• category — estimated by food category
• ai_estimate — AI-based estimate (lowest confidence)

Formulas

HOMA-IR (Homeostatic Model Assessment of Insulin Resistance)

Basis: Matthews DR et al. Diabetologia. 1985;28(7):412–419 [7]

When fasting glucose is in mmol/L:

HOMA-IR = (glucose × insulin) / 22.5

When fasting glucose is in mg/dL:

HOMA-IR = (glucose × insulin) / 405

Insulin values are in μIU/mL (mIU/L).

Limitations of HOMA-IR: the calculated value is a directional indicator for trend-tracking, not a diagnostic tool. What constitutes "normal" varies across labs and populations. INVY shows widely-cited ranges from ADA and Endocrine Society guidelines, but any value should be discussed with a physician.

Glycemic Load (GL)

Basis: Salmerón J et al. JAMA. 1997;277(6):472–477 [8]

GL = (GI × carbohydrates in grams per serving) / 100

Thresholds per Harvard T.H. Chan School of Public Health [9]:

Per meal:

• Low: ≤10
• Moderate: 11–19
• High: ≥20

Per day:

• Low: ≤80
• Moderate: 81–119
• High: ≥120

Predictive Insulin Index for uncatalogued meals

For meals where II hasn't been directly measured, INVY applies the composite-meal approach from Bao 2009 [4]:

II_meal ≈ (∑ II_component × energy fraction of component) / 100

Protein-containing foods carry the Bell 2014 correction [5]. Reported accuracy of this approach in clinical studies is approximately ±20% of measured values for mixed meals [4].

Reference Ranges

Important: reference ranges vary between labs. The values below are widely-cited indicative ranges based on American Diabetes Association [10] and Endocrine Society guidelines. Compare your results to your own lab's reference ranges, and discuss any borderline or elevated values with your doctor.

HOMA-IR (indicative)

HbA1c (per ADA 2025) [10]

Fasting glucose (per ADA 2025) [10]

Prediabetes and diabetes categories are established by a physician based on repeat measurements and clinical picture. INVY does not establish these categories — the app only shows where your number falls on a widely-cited scale.

Basis for Recommendations

Post-meal walk

INVY recommendation: walk for 10–20 minutes within 60–90 minutes after a main meal. Duration scales with the meal's glycemic load (GL ≥20 → 20 min, GL 11–19 → 15 min, GL ≤10 → 10 min).

Scientific basis:

• Reynolds AN et al. Diabetologia. 2016;59(12):2572–2578 — walking immediately after meals lowers postprandial glycemia more effectively than walking at other times of day [11]
• Buffey AJ et al. Sports Med. 2022;52(8):1765–1787 — even short interruptions of sitting with light-intensity walking improve cardiometabolic biomarkers [12]

4-hour spacing between meals

INVY recommendation: aim for at least 4 hours between main meals (without caloric snacking).

Scientific basis: insulin typically returns to baseline levels approximately 3–4 hours after a mixed meal in metabolically healthy individuals. Continuous grazing keeps insulin elevated, which is one factor contributing to insulin resistance. This recommendation balances physiology with practical adherence.

Sources:

• Paoli A et al. Nutrients. 2019;11(4):719 [13]
• Sutton EF et al. Cell Metab. 2018;27(6):1212–1221 [14]

14-hour daily fasting window

INVY recommendation: aim for at least 14 hours without food (typically overnight). This aligns with a broad definition of 14:10 intermittent fasting.

Scientific basis:

• Cienfuegos S et al. Cell Metab. 2020;32(3):366–378 — time-restricted eating improves insulin sensitivity [15]
• Sutton EF et al. Cell Metab. 2018 — early time-restricted feeding improves metabolic markers even without weight loss [14]

Individual variability

INVY builds recommendations from population-average study data. Individual responses to the same food can differ significantly.

Basis for this caveat:

• PREDICT Study (ZOE, King's College London) — Berry SE et al. Nat Med. 2020;26(6):964–973 [16]. A study of over 1000 twin and non-twin pairs showed postprandial responses to identical foods can vary 5–10× between individuals
• Stanford CGM studies — individual variability in glycemic response is substantially higher than assumed in standard GI tables

That's why INVY asks you to log both food and how you feel — so over time you see your personal patterns, not just population averages.

What INVY does NOT do

• ❌ We don't diagnose. We won't tell you "you have insulin resistance" or "you have prediabetes." A physician makes that diagnosis based on clinical context.
• ❌ We don't prescribe treatment. We don't recommend medications, doses, or changes to your doctor's prescriptions.
• ❌ We're not a medical device. INVY is not certified as a medical device and has not undergone clinical trials as a diagnostic tool.
• ❌ We don't replace your endocrinologist, primary care physician, or dietitian.
• ❌ We don't guarantee outcomes. We don't promise weight loss, "reversed" insulin sensitivity, or any specific clinical result.
• ❌ We don't analyze images for medical decisions. AI food and lab recognition is directional, not diagnostic.

Limitations

• Estimated Insulin Index — for most foods, II is calculated algorithmically, not measured. Accuracy is approximately ±20% of measured values.
• Reference ranges are population-based. Your lab may have different thresholds. Individual "normal" varies.
• Individual response varies. Two people can eat the same meal and have different responses.
• Accuracy depends on logging quality. If you log portions inaccurately or skip meals, trends will be less reliable.
• HOMA-IR has limitations in advanced type 2 diabetes with insulin deficiency. In later stages of impaired insulin secretion, HOMA-IR may be lower than actual insulin resistance.
• INVY doesn't account for genetics, hormonal status, or some medical conditions (e.g. thyroid function, sleep quality) that affect metabolism.

Updates

Methodology is reviewed quarterly. Significant changes are logged below:

• August 2, 2026 — initial publication

Questions

Reach out at [email protected].

If you're a researcher or clinician and spot an inaccuracy — we welcome the correction. We're open to constructive criticism and update promptly.

References

1. Atkinson FS, Foster-Powell K, Brand-Miller JC. International tables of glycemic index and glycemic load values: 2008. Diabetes Care. 2008;31(12):2281–2283. DOI:10.2337/dc08-1239
2. Atkinson FS, Brand-Miller JC, Foster-Powell K, Buyken AE, Goletzke J. International tables of glycemic index and glycemic load values 2021: a systematic review. Am J Clin Nutr. 2021;114(5):1625–1632. DOI:10.1093/ajcn/nqab233
3. Holt SH, Miller JC, Petocz P. An insulin index of foods: the insulin demand generated by 1000-kJ portions of common foods. Am J Clin Nutr. 1997;66(5):1264–1276. DOI:10.1093/ajcn/66.5.1264
4. Bao J, de Jong V, Atkinson F, Petocz P, Brand-Miller JC. Food insulin index: physiologic basis for predicting insulin demand evoked by composite meals. Am J Clin Nutr. 2009;90(4):986–992. DOI:10.3945/ajcn.2009.27720
5. Bell KJ, Gray R, Munns D, Petocz P, Howard G, Colagiuri S, Brand-Miller JC. Estimating insulin demand for protein-containing foods using the food insulin index. Eur J Clin Nutr. 2014;68(9):1055–1059. DOI:10.1038/ejcn.2014.126
6. Bell KJ, Gray R, Munns D, Petocz P, Steil G, Howard G, Colagiuri S, Brand-Miller JC. Clinical Application of the Food Insulin Index for Mealtime Insulin Dosing in Adults with Type 1 Diabetes: A Randomized Controlled Trial. Diabetes Technol Ther. 2016;18(4):218–225. DOI:10.1089/dia.2015.0254
7. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412–419. DOI:10.1007/BF00280883
8. Salmerón J, Manson JE, Stampfer MJ, Colditz GA, Wing AL, Willett WC. Dietary fiber, glycemic load, and risk of non-insulin-dependent diabetes mellitus in women. JAMA. 1997;277(6):472–477. DOI:10.1001/jama.1997.03540300040031
9. Harvard T.H. Chan School of Public Health. Carbohydrates and Blood Sugar. hsph.harvard.edu
10. American Diabetes Association Professional Practice Committee. Standards of Care in Diabetes—2025. Diabetes Care. 2025;48(Suppl 1). diabetesjournals.org
11. Reynolds AN, Mann JI, Williams S, Venn BJ. Advice to walk after meals is more effective for lowering postprandial glycaemia in type 2 diabetes mellitus than advice that does not specify timing: a randomised crossover study. Diabetologia. 2016;59(12):2572–2578. DOI:10.1007/s00125-016-4085-2
12. Buffey AJ, Herring MP, Langley CK, Donnelly AE, Carson BP. The Acute Effects of Interrupting Prolonged Sitting Time in Adults with Standing and Light-Intensity Walking on Biomarkers of Cardiometabolic Health in Adults: A Systematic Review and Meta-analysis. Sports Med. 2022;52(8):1765–1787. DOI:10.1007/s40279-022-01649-4
13. Paoli A, Tinsley G, Bianco A, Moro T. The Influence of Meal Frequency and Timing on Health in Humans: The Role of Fasting. Nutrients. 2019;11(4):719. DOI:10.3390/nu11040719
14. Sutton EF, Beyl R, Early KS, Cefalu WT, Ravussin E, Peterson CM. Early Time-Restricted Feeding Improves Insulin Sensitivity, Blood Pressure, and Oxidative Stress Even without Weight Loss in Men with Prediabetes. Cell Metab. 2018;27(6):1212–1221.e3. DOI:10.1016/j.cmet.2018.04.010
15. Cienfuegos S, Gabel K, Kalam F, Ezpeleta M, Wiseman E, Pavlou V, et al. Effects of 4- and 6-h Time-Restricted Feeding on Weight and Cardiometabolic Health: A Randomized Controlled Trial in Adults with Obesity. Cell Metab. 2020;32(3):366–378.e3. DOI:10.1016/j.cmet.2020.06.018
16. Berry SE, Valdes AM, Drew DA, Asnicar F, Mazidi M, Wolf J, et al. Human postprandial responses to food and potential for precision nutrition. Nat Med. 2020;26(6):964–973. DOI:10.1038/s41591-020-0934-0
17. Watson KT, Simard JF, Henderson VW, Nutkiewicz L, Lamers F, Nasca C, Rasgon N, Penninx BWJH. Incident Major Depressive Disorder Predicted by Three Measures of Insulin Resistance: A Dutch Cohort Study. Am J Psychiatry. 2021;178(10):914–920. DOI:10.1176/appi.ajp.2021.20101479