Mingau de Amaranto with Banana and Cinnamon

What it is

Whole-grain amaranth simmered with the lid on in water, a pinch of salt and some cinnamon, until the water is nearly gone and the grains are soft. Then in goes the milk, stirred on low until the pot turns thick and creamy. A little brown sugar, two minutes off the heat, banana and a last pinch of cinnamon on top. Two warm bowls.

In Brazil, mingau is a soft porridge, the warm kind you eat with a spoon, and this one is made with milk. I won't tell you amaranth mingau is anybody's tradition, because no source I found says so. I'm borrowing the format for a grain that deserves more of your mornings.

Amaranth is Protein Champion number seventeen because it breaks a rule most grains follow: its protein carries a generous share of lysine, the amino acid grains usually run short on. Its teammate is milk. Rice and beans, back in number one, traded. This pair tops up.

Why you should eat it

The short version, with the whole label on the table. USDA FoodData Central lists 13.6 g of protein and 6.7 g of fiber per 100 g of uncooked amaranth grain (record 170682), and this pot starts from 145 g of it. By my arithmetic the amaranth brings about 54% of the pot's protein and the whole milk about 42%, at 3.15 g of protein per 100 g of milk (record 171265). The amaranth also carries about 72% of the fiber, and the banana most of the rest. Now the full picture: this is a carbohydrate-rich bowl. Most of its sugars come from the milk and the banana (12.2 g per 100 g of raw banana, record 173944), and the brown sugar adds about 4.5 g a bowl. Most of the saturated fat comes from the whole milk, and skim is in the swaps. The salt is a single pinch. Why amaranth and milk make such a good team comes while the pot works.

Tips

• Milk catches on the bottom fast. Once it's in, keep the heat low and scrape the bottom every time you stir.
• Don't expect fluffy. Cooked this way, amaranth turns creamy and thick, and it keeps thickening as it stands (my kitchen note, no source behind it). Like it looser? A splash more milk off the heat, which the card doesn't count.
• Grains still firm when the 20 minutes are up? The Whole Grains Council's advice for any grain: add a little more water and keep cooking. Cooking time shifts with the grain's age, its variety and your pan.
• Taste before the sugar goes in. The banana brings its own: USDA lists 12.2 g of sugars per 100 g of raw banana (record 173944), about 14.4 g in a medium one.
• Forty-five minutes is a slow-morning breakfast, so save it for a rest day or the weekend. Twenty of those minutes are the lid-on simmer: that's when the coffee gets made.

Substitutions

• Whole milk: Skim milk USDA lists 3.37 g of protein and 0.08 g of fat per 100 g of skim milk (record 171269), against 3.15 g and 3.25 g for whole (record 171265). The bowl keeps its protein and loses most of the milk fat. The card is built on whole milk, so it won't match.
• Brown sugar (açúcar mascavo): Leave it out About 4.5 g less added sugar per bowl, and the banana still sweetens it. The card is built on 2 tsp of brown sugar.
• Banana: Any fruit you like on top, or none USDA lists 2.6 g of fiber per 100 g of raw banana (record 173944), about 3.1 g in a medium one. Swap it and the card won't match.

Storage

Let any leftovers cool, get them into the fridge within one to two hours of cooking, and eat them within 48 hours. That's the UK Food Standards Agency's line for leftovers, and your fridge should sit between 0 and 5 °C. The Whole Grains Council gives plain cooked grains 3 to 4 days, but this one is cooked in milk, so my own call is the shorter clock. It firms up in the fridge: warm it in a pan with a splash of milk or water, stirring, until hot right through.

While it simmers: the grain that broke the lysine rule

Lid on, heat low, twenty minutes with nothing to do. Pull up a chair.

Protein is built from amino acids, and the essential ones have to come from food. Grains are famous for running short on one of them, lysine. In USDA FoodData Central's records, lysine makes up 3.62 g of every 100 g of protein in raw white rice and 4.83 g in dry oats.

Amaranth didn't get the memo. Same database, same arithmetic: 5.49 g of lysine per 100 g of its protein. The review in case four says the same of the whole family, the pseudocereals (amaranth, quinoa, buckwheat): botanically apart from wheat, rice and maize, and high in the lysine those cereals run low on.

So if lysine isn't amaranth's gap, what is? The records point at leucine: 6.46 g per 100 g of amaranth protein, against 8.26 g in rice and 7.42 g in oats. A team in Lima saw it in people: in young children eating amaranth as their only protein, the share of leucine in their blood fell after the meal, then threonine, which the researchers read as the first two to run out. Lysine's dip was borderline. I've read only their summary, so hold that lightly.

Enter the teammate. Whole milk protein carries 9.49 g of leucine and 8.38 g of lysine per 100 g of protein (USDA again). Mix this pot's amaranth and milk and, by my arithmetic, the blend comes to about 7.8 g of leucine and 6.8 g of lysine per 100 g of protein, with amaranth still supplying 56% of it. Number one was a trade. This one is a top-up: the sulfur pair sits almost level in both proteins (3.07 g in amaranth, 3.24 g in milk).

Here's the picture I'd draw for a class: a school choir. Amaranth can hit the lysine note most grains can't reach, but its leucine voice is soft. Milk walks in with a big leucine voice, and the song holds.

The choir stops being true in three places. These are shares of the protein in the food, before digestion has its say, and in cases one and two the dairy protein was digested more completely. The leucine finding comes from children's blood, not anyone's muscles. And nobody here tested amaranth with milk for building muscle, so I won't promise you a bigger squat.

So, in one line: amaranth brings more lysine than a grain usually does, and milk turns up the leucine it's quiet on.

Case one: twelve men, amaranth against cheese

Twelve adult men in Guatemala, studied by the Institute of Nutrition of Central America and Panama, who kept up their usual physical activity throughout. In three nine-day periods, protein started very low and climbed every two days, from 0.2 to 0.4 to 0.6 g per kg of body weight a day, with energy, vitamins and minerals held steady. In each period all of it came from one food: extruded amaranth, popped amaranth (both as a sweet purée) or processed cheese, the reference. Every man ate all three, in random order, and everything in and out was collected to follow the nitrogen, which is how scientists track protein.

What happened: cheese protein scored 101.4% on their digestibility measure, extruded amaranth 89.8% and popped 85.5%. Of the protein absorbed, the men kept about the same share whatever the source, a statistical tie (0.97 for cheese, 0.98 and 0.96 for the amaranths). Against everything eaten, cheese stayed ahead: 0.97 against 0.86 and 0.79. The authors still ranked amaranth protein among the best of plant origin, close to animal foods.

For your plate: amaranth protein was digested a little less completely than cheese, and what got absorbed was kept about as well. Here the dairy is a teammate, not a rival.

Where it stops: twelve men, two days per protein level, at the low intakes this method needs, eating processed amaranth. The full paper is paywalled, so I've read the summary only; the men's ages, the authors' own limits and the funding stay unseen. An old, careful study, and a small one.

Case two: young children, amaranth against milk protein

Young children in Lima, Peru, studied by the Instituto de Investigación Nutricional: nine in a first round, seven in a second. For nine days at a time, amaranth was their only source of protein and fat and half their energy, as toasted flour, popped grain or flakes. Before and after each stretch came a diet built on casein, the main protein in milk, so each child was compared with themselves. Every diet kept protein low, at 6.4 to 6.7% of energy.

What happened: the children absorbed about 84% as much nitrogen from amaranth as from casein, whatever the form. They kept less: 70.9% of the casein figure for toasted flour, 65.9% for popped grain and 59.0% for flakes. Their stools carried more energy, fat and carbohydrate on amaranth, so more of the food left undigested.

In the second round, toasted amaranth flour supplied 20% or 30% of the protein in maize meal: absorption came to 70.4% and 72.1% of intake, against 83.8% on casein. The authors still judged that maize with 20% of its protein from amaranth should be able to cover young children's protein and fat needs.

For your plate: how amaranth was processed changed how much protein the children kept, and milk protein was the stronger performer. This bowl puts milk protein in with it.

Where it stops: a handful of children in 1988, on a deliberately low-protein test diet, eating processed amaranth, never whole grain simmered in milk. The same team later reported that their earlier mixes, with 20, 30 or even 50% of the maize protein from toasted amaranth, had shown no gains over maize or amaranth alone, and suspected heat had damaged some lysine. Summaries only: the children's ages, the funding and the authors' own limits stay unseen.

Case three: popped amaranth with milk, and blood sugar

Adults living with type 2 diabetes in Hyderabad, India, studied by a university food and nutrition department with Osmania General Hospital. On separate occasions they ate portions with 50 g of carbohydrate each: two amaranth-and-wheat flour mixes, wheat, rice, or popped amaranth eaten with milk. Each rise in blood sugar was scored against glucose, which counts as 100. The paper's summary gives no head count; the International Tables of Glycemic Index, which list the result, record six people for the amaranth meal, sweetened with a non-nutritive sweetener, and four blood samples over three hours.

What happened: popped amaranth with milk scored 97.3, the highest of every meal and close to glucose itself. The 25:75 amaranth-wheat flour scored 65.6, wheat 65.7, rice 69.2 and the 50:50 flour 75.5. One label check of my own: the authors called the 65.6 to 69.2 group 'low', and on the Tables' cut-offs (55 or less is low, 70 or more is high) those land in the middle.

For your plate: the review in case four notes that processing can substantially change a pseudocereal's starch and its effect on blood sugar, and popping is processing. This bowl is whole grain simmered soft, and nobody has tested it. It's still a carbohydrate-rich breakfast, and the card says so plainly.

Where it stops: six people, four samples, a wide spread around that 97 (plus or minus 19), and the Tables file it in their table for small groups, impaired glucose tolerance or wide variability. Summary only, so the funding and the authors' limits stay unseen.

If you're managing blood sugar, take this to your doctor. A recipe can't make that call for you.

Case four: the review that says the evidence is thin

Two nutrition researchers at Hacettepe University in Ankara searched for human studies, from 1980 to 2025, of quinoa, buckwheat and amaranth on blood fats, blood sugar, blood pressure and body weight, and found 33.

Amaranth's share was small: five studies on blood fats, most of them testing amaranth oil, not the grain, with mixed results. In one 3-week trial of 125 people with heart disease, high blood pressure and obesity, amaranth oil came with falls in triglycerides and in total and LDL cholesterol. In another 3-week trial, in 44 people with overweight or obesity, it raised total and LDL cholesterol compared with rapeseed oil. On blood sugar, one study: in 38 healthy adults, a snack bar of 90% amaranth had the lowest glycemic index of the bars tested.

Their verdict: too few human studies, especially of amaranth, to recommend it for heart and metabolic health. Across all three pseudocereals, they call the benefits for blood fats possibly mild and not clinically significant.

For your plate: eat amaranth for its unusual lysine and its fiber, not as a cholesterol treatment. The word 'superfood' on a packet is running well ahead of the data.

Where it stops: a narrative review, so nothing was pooled or scored for quality, and its studies were mostly small and short, a median of four weeks, with side effects rarely reported. No external funding, no conflicts declared. For once, the review is the one asking everyone to calm down. If you're managing your cholesterol, your doctor is the right person to weigh this.