132 districts

Last week: two numbers. Survey wasting ~19%. Dashboard wasting 3%. Stunting is close. Acute malnutrition is not.

Part 1 → Two nutrition numbers

This week: the map.

West Singhbhum. Stunting 66%. Underweight 44%. Measured 97% of the children on the roll.

Nandurbar. 63% stunted. 45% underweight. Measured 99%.

These are not secret blocks. They are at the top of a list the public dashboard can produce and almost never does: 132 districts that, in August 2026, sit in very high risk on the Anganwadi file.

Not a survey. Not NFHS. What the machine wrote down, this month.

How the list is made

Four numbers on children 0–5: stunting, underweight, wasting, overweight. Combined into a score. Twenty or more = very high risk.

August 2026:

Forty-three percent of districts are high or very high. The file is not a wall of red. It is also not a success slide.

This is relative to the dashboard. Part 1 still holds: do not read 132 as “NFHS says 132.” Read it as: these are the places the programme itself cannot call fine.

The top of the page

Coverage = share of registered 0–5 children measured that month. Figures are stunting / underweight / wasting.

Frontline walked Madhya Pradesh this August. Free Press Journal put the state Tracker rates in September. Neither published 38 of MP’s 55 districts in this band. That is the new sentence.

Where the 132 sit

Madhya Pradesh is 29% of the entire very-high-risk list. Add Bihar, Jharkhand, Tripura: 68 of 132. Half the red, four states. That pile is Part 4.

Telangana’s six will look different next week. They are on a state that moved the wrong way.

What stays on the list and does not lead it

Thirteen very-high-risk rows have measurement coverage under 85%. Eleven are in Arunachal (Anjaw 46%, Tirap 47%…). Two in Nagaland. A 56% stunting rate on half the children measured is not Nandurbar’s 63% on 99%. They stay. They do not open the piece.

Six August rows have no public name — dummy ids, including a Gujarat 990 with zero children. Plumbing. Not a place.

Mahe, Puducherry, 57% stunted. In the file. Not a national headline.

This week’s only job

A targeting list. Current. From the dashboard itself.

A national average can fall and these 132 can still be here.

Next — Part 3: 634 down. 94 up. Who moved since 2023. Some of these names have been falling for three years and remain. Some — Delhi, Telangana — have been rising.

Data: Part 2 of 5. Poshan Tracker, August 2026. Very high risk = score ≥ 20 on stunting, underweight, wasting, overweight (0–5). Coverage from November 2024. Unnamed district-ids dropped from any map you publish. Programme data, not NFHS. Also on LinkedIn.