Is Kambu (Pearl Millet, Bajra) Good for Diabetes? Koozh, Dosa and the Portion

IN MODERATION

In moderation — kambu is a better staple than white rice, not a free one. Pearl millet carries four times the fibre of milled rice and a pooled glycemic index of 56.6, which is intermediate rather than low. One tumbler of kambu koozh, or two to three kambu dosai, in place of that meal's rice.

Answered by Dr. Dante Ruskin · Diabetologist & Intensivist

The grain does the work, not the ritual. Kambu's fibre, fat and protein slow the meal down, but souring koozh overnight does not lower the sugar hit — measured, it slightly raises it. Swap your morning idli or rice kanji for savoury koozh with buttermilk; never add it beside the rice.

Is kambu really better than rice, or does it only sound better?

Better — and the tables say by how much. Per 100 g of raw grain, IFCT 2017 puts kambu at 61.8 g of available carbohydrate against 78.2 g for milled rice, and, more importantly, at 11.5 g of dietary fibre against 2.8 g. That is four times the fibre. Kambu also carries 6.4 mg of iron where rice gives 0.65 mg, which is why this was always the grain of dry-country field labour. Pooled across published trials, its glycemic index is 56.6 — intermediate, sitting between the low band and milled rice at 71.7 (Anitha et al., Frontiers in Nutrition, 2021). So the verdict is moderation, not an unqualified yes: kambu flattens the slope of the rise, it does not remove it. Substitute it for rice and most patients’ two-hour readings improve. Add it to a plate that already carries rice and nothing improves at all.

Raw grain, per 100 gCarbohydrateFibreProteinIronMean GI
Kambu (bajra)61.8 g11.5 g11.0 g6.4 mg56.6
Ragi (finger millet)66.8 g11.2 g7.2 g4.6 mg61.1
White rice, raw milled78.2 g2.8 g7.9 g0.65 mg71.7

Composition: IFCT 2017 (ICMR-NIN). Glycemic index: Anitha et al., 2021.

How much carbohydrate is in one tumbler of kambu koozh?

Published home recipes let us do this arithmetic in the open. One widely used Tamil recipe grinds half a cup of kambu — about 100 g of grain — into three cups of koozh (Subbu’s Kitchen); a thicker household version uses a full cup for three to four cups (Essence of Life). Applying IFCT’s 61.8 g of carbohydrate per 100 g, that is roughly 20 g of carbohydrate in a thin tumbler and about 35 g in a thick one — estimates, because no two houses mix it the same way. Set that beside a cup of cooked white rice at about 45 g and the point becomes obvious: koozh is a meal, not a drink you add to breakfast. The mistake I correct most often is the patient who drinks a tumbler at seven and still eats idli at eight.

What you actually takeKambu in it (est.)Carbohydrate (est.)
One tumbler thin koozh (½ cup grain → 3 glasses)~33 g grain~20 g
One tumbler thick koozh (1 cup grain → 3½ glasses)~57 g grain~35 g
Two instant kambu dosai (¼ cup kambu + ¼ cup wheat → 4 dosai)~18 g flour~22 g
One cup cooked white rice, for comparison—~45 g

My arithmetic from published recipes (Subbu’s Kitchen, Essence of Life, Raks Kitchen) applied to IFCT 2017 composition — estimates, not label values.

Does the overnight fermentation make koozh safer for a diabetic?

Here the folk claim and the laboratory disagree, and patients deserve the laboratory. The overnight souring is real and useful: it improves digestibility, drops the pH to roughly 4.3–4.9 and keeps the pot usable through a hot day (Down To Earth, 2021). What it does not do is soften the sugar hit. When one group measured it directly, the estimated glycemic index of pearl millet flour rose from 57.4 to 61.6 after lactic-acid fermentation, because fermentation consumes the resistant starch that was slowing digestion (Mohapatra et al., Frontiers in Bioscience-Elite, 2024). Read that as a correction, not a condemnation: ferment koozh for taste, gut comfort and keeping quality, and stop treating “it is fermented” as permission for a second tumbler. What genuinely flattens the curve in that tumbler is the buttermilk stirred through it, the raw shallot and green chilli beside it, and eating it sitting down as a meal.

Koozh, dosa, or the millet health-mix packet?

Ranked, and the ranking matters more than the grain. Savoury koozh with buttermilk is the best form — thick, sour, salted, eaten with onion and chilli, and it stands in for a rice breakfast without any negotiation. Kambu dosai and kambu adai come next: an instant recipe stretches a quarter-cup of kambu flour across four dosai (Raks Kitchen), so each piece is a modest load, and adding dal or onion to the batter slows it further — the same logic as on idli and dosa for diabetes. Kambu kali or roti is fine, in a normal-sized piece. Sweet koozh with jaggery drops to the bottom, because jaggery is sugar with minerals attached — the whole argument sits on is jaggery good for diabetes. Last come instant kambu malt sachets and “millet health mix” powders, which are often rice-flour-dominant with sugar high on the ingredient list. Read the back of the packet, not the front of it.

What to say at the koozh stall, and on an Aadi temple day

Street koozh is a real summer trade in Tamil Nadu — hawkers work with pots strapped to a cycle, and a tumbler goes for around ten rupees (Down To Earth, 2021). In our own Nanjil belt kambu is not the grandmother grain it is in Madurai or Ramnad; this is rice country, and kambu arrives here as flour packets and as summer koozh. Either way, these are the words that get you the version you can drink:

  • “கம்பங்கூழ், மோர் விட்டு” (kambankoozh, mor vittu) — with buttermilk, meaning the savoury one, not the sweet pot.
  • “வெல்லம் வேண்டாம்” (vellam vendaam) — no jaggery.
  • “கெட்டியா, ஒரு டம்ளர் மட்டும்” (kettiya, oru tumbler mattum) — thick, one tumbler, and no refill when he offers.
  • “வெங்காயம், பச்சை மிளகாய் ஜாஸ்தி” (vengayam, pachai milagai jaasti) — extra shallot and green chilli.
  • “உப்பு கொறச்சு” (uppu korachu) — less salt, and say it every time if you carry blood pressure, because koozh is salted generously.

During Aadi, unfermented koozh is cooked in quantity at Amman temples and handed out as prasadam. Take the small tumbler — it is prasadam, and I do not ask anyone to refuse it. Then let it stand in for that meal’s rice, and walk twenty minutes afterwards.

Who should take care with kambu

Three groups, and none of them is told to avoid it outright. Thyroid: pearl millet carries C-glycosylflavones that inhibit thyroid peroxidase, the reason millet-dependent, iodine-poor regions carry a goitre burden (Gaitan et al., J Clin Endocrinol Metab, 1989). With iodised salt in the kitchen and kambu a few times a week this is a footnote — but if you already take thyroxine, keep it occasional and your tests on schedule. Kidney: kambu carries 365 mg of potassium and 289 mg of phosphorus per 100 g against rice’s much smaller load (IFCT 2017), so ask the doctor managing your kidneys before any millet becomes daily. Medication: if you take insulin or a sulfonylurea and you swap rice for kambu at a meal, the slower rise can leave the dose too large — watch for the shaky, sweaty low sugar in the first week, write the readings down, and bring them to your review. Your doctor adjusts the dose; this page only sets the direction.

Is ragi good for diabetes? · Varagu (kodo millet) for diabetes · Thinai (foxtail millet) for diabetes · Samai (little millet) for diabetes · Millets and the glycemic index · Summer drinks for diabetics

Frequently asked questions

The savoury version, yes, in one tumbler. Koozh made from pearl millet and soured overnight, stirred with buttermilk and eaten with raw shallot and green chilli, is a reasonable breakfast that replaces rice. The sweet version made with jaggery is not — that is the same carbohydrate with fast sugar poured on top.

They are close, and the difference is which mineral you need. Pooled trial data put pearl millet at a mean glycemic index of 56.6 and finger millet at 61.1 (Anitha 2021), and IFCT 2017 gives kambu far more iron, ragi far more calcium. Alternate them rather than choosing once and for all.

Most people can, provided it replaces part of the day's rice instead of joining it. Daily kambu is a fair habit for someone with normal thyroid and kidney function. If you have hypothyroidism, keep it to a few times a week, and if your kidney function is reduced, ask your own doctor before making any millet a daily staple.

Pearl millet contains C-glycosylflavones that inhibit thyroid peroxidase, which is why goitre clusters in iodine-poor regions living on millet (Gaitan, J Clin Endocrinol Metab, 1989). In a household using iodised salt and eating kambu a few times a week this is a footnote. If you already take thyroxine, keep kambu occasional and your tests on schedule.

Usually yes, because the flour brings fibre a rice-and-urad batter does not, and each instant kambu dosai carries a small carbohydrate load — roughly 11 g by my arithmetic from a published recipe. The gain is lost if you eat six of them, or if the accompanying chutney arrives with a sweet sambar and a second serving.

No — measured in the laboratory it moved the other way, with the estimated glycemic index of pearl millet rising from 57.4 to 61.6 after lactic-acid fermentation as resistant starch was consumed (Mohapatra, 2024). Ferment koozh for taste, digestibility and keeping quality. Do not treat the souring as permission to drink a second tumbler.