Modeling undertaken in 112 pregnant women (65.5%) consuming >1 serve of beef per week. AI, Adequate Intake; EAR, Estimated Average Requirement; NRV, Nutrient Reference Value.

> Further analyses of replacing reference foods with environmentally sustainable alternatives can be found in the Supplementary Material. For example, replacing one serve of beef with one isoenergetic serve of mixed beans reduces protein, saturated fat, and zinc content as expected, while increasing dietary fiber, iron, calcium, and folate content, as well as lowering all measures of environmental sustainability (Table S2). In addition, replacing one serve of milk with an isoenergetic serve of soy milk does not negatively impact on calcium content (+68.4 mg/serve) but reduces the impact on all measures of environmental sustainability (Table S5). The net effect of this replacement per day over the course of an entire pregnancy would reduce GHG emissions by 138.9 kg CO2 eq, equivalent to the emissions produced by a typical passenger vehicle driven for 559 km (Table S5).

### 3.2.2. Protein-Matching Environmentally Sustainable Alternatives

The nutrient analysis and measures of environmental sustainability for the reference foods and protein-matched serves of more environmentally sustainable alternatives are shown in Table S6. The net differences in nutrients and measures of environmental sustainability of replacing one serve of a reference food for a protein-matched serve of a more environmentally sustainable alternative are shown in Table S7 (beef), Table S8 (chicken), Table S9 (white fish), and Table S10 (milk).

The weight of protein-matched portion sizes for all of the more environmentally sustainable alternatives was markedly greater than that of beef, with some plant-based alternatives weighing more than four times as much as one serve of beef (Table S4). Similar to the isoenergetic serves, animal-derived foods were rich in iodine, whilst plant-based alternatives were rich sources of calcium, folate, and dietary fiber.

In general, plant-based alternatives were high in iron when matching protein content, with the richest sources from silken tofu, mixed beans, and lentils (Table S4). The estimated

iron absorption of these three foods was three times as much as beef in one protein-matched serve. The average zinc content provided by plant-based alternatives was lower than that of beef but higher than other animal-derived alternatives.

Using protein-matched serves did not markedly alter the results regarding the net benefits to environmental sustainability, particularly GHG emissions. For example, if pregnant women substitute one serve of beef with a protein-matched serve of firm tofu each week throughout the course of pregnancy, GHG emissions would be reduced by 363 kg CO2 eq, equivalent to 1461 km of typical driving distance by an average passenger vehicle. The net results of this replacement would decrease the proportion of women who meet NRVs for zinc by 5% and increase the proportion of pregnant women meeting NRVs for calcium and fiber by 10% and 2%, respectively (Table 5). For other measures of environmental sustainability, all food alternatives have land use less than 6.4 m2, except for lamb (36.3 m2), per protein-matched serve (Table S6). Most food alternatives have acidifying emissions of lower than 15 g SO2 eq per protein-matched serve. White fish produced the highest eutrophying emissions (20.8 g PO4 <sup>3</sup><sup>−</sup> eq). Lastly, protein-matched serves of more environmentally sustainable alternatives have relatively high stress-weighted water use on average, although this varied greatly, ranging from 861L per serve for firm tofu through to over 13,000 L per serve for mixed nuts and lamb.

**Table 5.** Modeling the nutrient intake of pregnant women replacing one serve of beef per week with a protein-matched serve of firm tofu #.

