Determination of the Sensory Attributes of Dried Milk Powders and Dairy Ingredients

A standardized descriptive language for skim milk powder and dried dairy ingredients was developed. The lexicon was initially identified from a large sample set of dried dairy ingredients (138). A highly trained descriptive panel (n = 14) refined terms and identified references. Dried dairy ingredients (36) were then evaluated using the developed language. Twenty-one descriptors were identified for dried dairy ingredients. Seventeen flavors and tastes were identified in skim milk powders (27) with nine flavors/tastes observed in all skim milk powders. Dried dairy ingredients were differentiated using the language (P C 0.05). There were flavors common to all dried dairy ingredients while other flavors were specific to particular products. INTRODUCTION Worldwide very large amounts of dried milk products are manufactured. The primary dry milk product manufactured in the U.S. is skim milk powder (SMP) although whole milk powder is also produced (ADPI 2000). Over 663,000 tons of SMP and over 50,000 tons of whole milk powder (WMP) were produced domestically in 2000 (ADP! 2000). SMP is mainly produced by spray-drying of fluid pasteurized milk and the finished product composition is not more than 1.5% fat by wt (ADPI 1998), Three basic types of SMP are produced: low heat, medium heat, and high heat (ADPI 1998). Typically, SMP should have a shelf-life (at 25C) of 12-18 months (ADPI 1998). Currently, U.S. SMP production continues to increase while the U.S. has only a 17% share of the world SMP market (Lagrange 2001). The global market for SMP is expected to grow by 20% by the year 2005 (Lagrange 1998). SMP and other dried dairy ingredients should ideally have a clean, sweet and pleasant taste free of flavor defects (Bodyfelt er a!, 1988). For SMP, cooked flavors may be present and vary according to heat treatment (low, medium, high) of the milk prior to evaporation and spray-drying. Dry milk products must have good flavor characteristics for direct consumption as well as for product applications. The largest outlet for SMP is ingredient application in dairy products. Recommended applications of SMP classification based on preheat treatment vary. Low and medium heat-treated powders are used in fluid milk fortification, cottage cheese, yogurt and chocolate dairy drinks because of their high solubility and minimal cooked flavor. However, high heat powders are mostly used in bakery, confectionery, meat products for texture modification, water absorption/binding, emulsification and gelation (Varnam and Sutherland 1996; ADPI 1998). Off-flavors present in SMP and other dried dairy ingredients can carry through to the final end-product and result in poor end-product quality. Therefore, flavor and sensory properties are considered important criteria in grading dry milk powders and other dried dairy ingredients (Bodyfelt et al. 1988). Grading has been used by the industry for many years to evaluate overall quality of milk powders and other dried dairy ingredients but is defect-based, rather than descriptive in nature. These defects are not well-defined, and grades are based on one person's somewhat subjective opinion. Descriptive sensory analysis is traditionally used to identify and quantify flavors in products and can be a powerful tool for describing and differentiating product flavors for research, product development or marketing (Meilgaard et al. 1999; Drake and Civille 2003). Previous research that has addressed flavor of SMP has not used descriptive sensory analysis (Kurtz et al. 1971; Driscoll et al. 1985). Descriptive sensory analysis has been conducted widely on fluid milk (Lawless and Claassen 1993; Watson and McEwan 1995; Chapman et al. 2001; Born Frost et al. 2001), as well as ice cream (Prindiville et al. 1999, 2000) and defined sensory languages have been developed specifically for cheese (Heisserer and Chambers 1993; Murray and Delahunty 2000; Drake er al. 2001). Descriptive sensory flavor research has not been conducted on milk powders or other dried dairy ingredients. Since flavor consistency has reportedly been a problem in SMP and WMP (Dryer 1999), the development of a descriptive sensory language for SMP and WMP would provide a powerful research and quality control tool. The objective of this study was to identify and develop a descriptive sensory language for milk powders and other dried dairy ingredients. MATERIALS AND METHODS Milk Powders and Dried Dairy Ingredients SMP, WMP and other dried dairy ingredients were donated or purchased from producers nationally and internationally. Thirty eight SMP were obtained from U.S. producers (twenty low heat, five medium heat, thirteen high heat) and twenty-four SMP were obtained from international sources in eleven countries (seventeen low heat, four medium heat, three high heat). An additional fourteen SMP (all low heat) were obtained from U.S. universities that varied in age from 3 months to 20 years. Eleven WMP (eight U.S., three international) were obtained. Sixteen whey protein concentrates (WPC) or whey protein isolates (WP1), five milk protein concentrates (MPC), and ten caseinates were also obtained from U.S. and international producers. The other dairy ingredients were obtained to get as wide an array of dairy milk product flavors as possible. Samples were received as 1, 5. 10 or 50 lb samples and were stored at -20C in the dark upon receipt. For sensory analysis, skim milk powders were rehydrated on a 10% solids basis, 10 g SMP/100 mL water (ADPI 2001) and whole milk powders were rehydrated using the formula: 1000/100-dry %fat content of WMP = g of WMP dissolved in 90 g water (IDF 1997). Powders were rehydrated on the percent solids basis with deodorized water (prepared by boiling 4 L of distilled water until its volume was decreased by one-third) and blended with an electric hand-held mixer. Eight percent-solution of WPC, WPIs, MPCs and caseinates were rehydrated in warm water (60C) based on their protein amount (Harris 2000). Language Identification and Development Rehydrated powders were evaluated for flavor and initial language generation by a group of six individuals experienced ( >500 h experience each) with descriptive analysis of products. Rehydrated powders were dispensed into Styrofoam cups with lids (to prevent light oxidation) and allowed to equilibrate to 7C for 1 h prior to evaluation. The initial language is listed in Table 1. The panel also grouped the rehydrated powders based on general flavors for future evaluation and language refinement. [Table 1] Fine-tuning and Identification of References Following the roundtable and generation of the preliminary language, a descriptive panel was trained to refine the developed language and to identify references. Fourteen individuals, six male and eight female, 22 to 44 years of age were selected from university staff and students based on availability, interest, and a demonstrated liking of dairy products, including milk. Twelve panelists had prior experience with descriptive analysis. The panel received fifty hours of training using the Spectrum' method (Meilgaard et al. 1999) and the identified language. The rehydrated dairy ingredients initially evaluated and grouped were presented to the panel. Panelists were encouraged to volunteer new descriptive terms and to address existing terms. Panelists received food treats and monthly gift certificates at local restaurants for their participation. During training, panelists were presented with and evaluated potential references (food or chemical) and the panelists selected the best reference(s) to represent the identified descriptive terms [Table 2].