Two flatbreads with about the same calories. Pregnant women given the amaranth one gained 2 kg more

Most trials of pregnancy nutrition test a tablet or a sachet. This one tested bread. In Hawela Lida, a farming district in the Sidama region of southern Ethiopia where maize is the everyday staple, researchers from Hawassa University wanted to know what would happen if part of that staple were swapped for amaranth, a tiny-seeded grain that grows well in dry conditions and carries more protein, iron and calcium than maize.

The trial, published in PLOS One on 8 October 2026, was randomised by community, not by person. Ten kebeles, the smallest local administrative units, were drawn by lot: five got the amaranth bread and five got the comparison bread. In them, 306 women who were less than three months pregnant and had mild to moderate anaemia took part. Every day for six months a distributor brought each woman a flatbread of about 200 g and watched her eat it. One was 70% amaranth and 30% chickpea, the other all maize, and the grains had been soaked, sprouted, roasted and fermented first. The two breads were made to look alike, and the lab analysis found they had about the same energy; the amaranth one had more protein, iron and calcium. Both groups had the same nutrition counselling.

By the end, 293 women were still in the study. Those on the amaranth bread had gained 8.1 kg on average since enrolment, against 5.9 kg on maize. Their upper arms, a quick field measure of how well-nourished someone is, averaged 25.3 cm against 23.7 cm. At the start about 28% of the women in both groups had an arm measurement under 23 cm, the cut-off used for undernutrition. At the end it was 12% in the amaranth group and still 29% in the maize group. And 64% of the amaranth group had gained weight in the recommended range, against 41%.

Those are big differences for one food, so the cautions matter. With only ten communities, a chance difference between a few of them can sway the result, and wealth, media use and diet variety weren't evenly balanced at the start (the analysis adjusted for them). Nobody measured what else the women ate, so we can't tell whether the bread added to their meals or changed them. The trial didn't follow up birth weight or the babies' health, which is the reason weight gain in pregnancy matters at all. The figures also wobble inside the paper: the abstract gives slightly different averages from the results table, and the drop in undernutrition is reported as 59% in one place and 65 to 68% in another. The Nestlé Foundation paid for the study and, the authors say, had no part in running or reporting it. A reasonable reading is that a better staple helped undernourished women in a place where food is short. It says nothing about amaranth for a well-fed pregnancy.