An Indian food database is useful only when you understand what its entries represent. IFCT 2017 by ICMR-NIN is a national food-composition reference built from analysed foods; the Indian Nutrient Databank, or INDB, focuses on common cooked recipes and documents how its data were assembled. They answer related but different questions. Neither tells you exactly what was in a particular client’s katori of dal, how much oil was used or how large a home serving was. Treat database outputs as inputs to an estimate, not as facts about an unmeasured plate.
What a food-composition table is for
The Indian Food Composition Tables 2017, published by the National Institute of Nutrition under ICMR, provide nutrient composition information for selected foods. The table is a reference for the foods and components it includes. It is not an exhaustive catalogue of every branded food, regional recipe or household variation eaten across India.
That distinction matters in practice. A raw ingredient entry and a cooked dish are not interchangeable. A food may change during preparation, and a mixed recipe includes ingredients in proportions that a table cannot know unless those proportions are recorded. A food name that sounds close to an entry does not prove that the client’s recipe matches it.
Use the table to understand a selected food entry and its documentation. Check whether the entry is raw or prepared, what unit it uses and which components are available. If you need to use the data in a client discussion, explain how you selected the entry and what assumptions remain. Do not cite the table as if its analysts tested the person’s meal.
What INDB adds
The Indian Nutrient Databank is an open-access resource developed by Anuvaad Solutions. Its public description focuses on common Indian cooked recipes. The methodology paper, “Development of an Indian Food Composition Database,” appeared in Current Developments in Nutrition. The database team describes drawing on IFCT and other food-composition sources while documenting gaps and recipe methods.
A recipe database can make it easier to find an entry resembling a prepared dish. That can be useful when a client names a familiar recipe rather than separate raw ingredients. But a reference recipe is still one representation. A home version may use a different pulse, oil, ingredient ratio or serving size. A restaurant preparation may differ again. Similar names do not guarantee identical composition.
When you select an INDB recipe, read its description and serving basis. Ask whether it resembles the client’s dish. If it does not, choose a more suitable approach or leave the numeric estimate out. Do not treat a database’s recipe entry as an authority on what the client ate.
| Resource | Useful for | What it does not establish |
|---|---|---|
| IFCT 2017 by ICMR-NIN | Looking up composition for listed food entries | Every cooked recipe or a client’s exact portion |
| INDB by Anuvaad Solutions | Reviewing documented common cooked-recipe entries | The ingredients and quantities in a specific home preparation |
| A client’s food diary | Understanding what the client reports eating | A laboratory measurement of the meal |
Ask what decision the data need to support
Before opening a database, name the question. Are you trying to discuss a food choice, compare a portion description or calculate an approximate total? A reference may help with the last question, but it may not be necessary for the others. Avoid looking up every food simply because the software makes that possible.
If the client’s note says “two roti, dal fry, curd,” first clarify what they mean where it matters. Which rotis? What kind of dal? How much curd? Did they use a familiar katori or a canteen serving? The purpose is not to force exact weights. It is to understand whether a selected database entry is a reasonable approximation for the conversation.
Keep the client’s words in the record. If you map “dal fry” to an entry with a different recipe, record that as your choice. If you approximate the serving, say so. These details make it possible to revisit the calculation later and explain why it might change when the client adds information.
Read the unit and food state carefully
Food references can present values by weight or by a defined serving. Check the unit before using an entry. Per-weight values do not describe a person’s portion until the amount is known. A serving measure supplied by a database may not match the katori or serving spoon used by the client.
Also check whether an item is raw, cooked, dried or prepared. A raw ingredient entry cannot be applied directly to a cooked dish without considering the preparation and amount used. Water changes weight during cooking; ingredients may be discarded or added. For a mixed dish, any calculation depends on the recipe assumptions.
If a unit or state is unclear, do not guess. Find the documentation or choose a more transparent path. A clear statement such as “this entry is only an approximate match” is more useful than an exact-looking total that hides a mismatch.
Keep gaps visible
No food database contains every food as prepared in every home. Regional dishes, local names, seasonal ingredients and family recipes create variation. A database may not have the exact item you need, or a nutrient field may be unavailable. The absence of an entry is not evidence that the food has no nutritional value.
When the food is missing, you can discuss the uncertainty with the client, use an appropriate ingredient-based estimate if you have enough information, or avoid a numeric answer. If you combine several reference entries, record that method. Do not fill gaps with invented values or imply that an estimate is comprehensive simply because it came from software.
Explain estimates in plain language
Tell the client what the number is based on and what could change it. “I used the closest cooked-recipe entry and your approximate serving description” is more informative than giving a number alone. If they correct the recipe or portion, revise the estimate. Make clear that the change reflects better inputs, not necessarily an error in the client’s memory.
Nutrition values should be labelled as estimates when the meal or portion is approximate. Do not attribute a specific value to ICMR-NIN or Anuvaad as though the institution measured the client’s food. Name the reference you used, explain how you applied it and distinguish the source data from your calculation.
How Shellel helps
Shellel estimates Indian meals from the client’s own food description and everyday portion language, with the result presented as an estimate. The nutritionist can review the diary and clarify the preparation with the client. Explore the nutritionist workspace and read roti, rice and dal portions in meal diaries for a practical guide to household measures.
Quick answers
Asked and answered
Is IFCT 2017 an Indian recipe book?
It is a food-composition reference, not a record of every household recipe. It can inform work with listed foods, but a cooked mixed dish may need additional recipe information or a different reference.
Is INDB more accurate than IFCT for every meal?
They serve different data needs, and neither is automatically a match for every client’s preparation. Check the food description, recipe basis and serving assumptions before using an entry.
Can I quote a database value as the exact nutrition in a client’s plate?
No, not unless the particular food and portion were measured in a way that supports that claim. A selected reference entry can inform an estimate, but it does not measure an individual meal.
