Optimization of Quinoa and Soy Milk for Steamed Pumpkin Cake Using Response Surface Methodology

Abstract

This present study aimed to enhance nutrients and reduce fat content in steamed pumpkin cakes. Two independent variables, including quinoa (30−80% w/w by pumpkin weight) and soymilk (15−60% w/w by coconut milk weight), were studied using response surface methodology. The statistical results showed that the quinoa inclusion had significant linear and quadratic effects (p<0.05). The soymilk showed a negligible effect, and there was no interaction effect between the two variables. Increasing quinoa has decreased all sensory scores of the samples, mainly when replacing pumpkin flesh with more than 50% quinoa inclusion. The pumpkin cake with 30% quinoa and 46% soymilk was the optimal formulation with a 0.936 desirability value. The optimal sample lowered all sensory attributes than the regular sample, particularly the appropriate qualities, such as unpleasant flavor and softer texture. About 54 percent of the consumers were willing to buy the product. The product is healthy with a 14.6% protein increment and 27% fat reduction.