About

MaryAnne Drake is the William Neal Reynolds Distinguished Professor of Food Science and the Director of the Southeast Dairy Foods Research Center at North Carolina State University. She holds a Ph.D. and M.S. in Food Science from Washington State University and a B.S. in Biology from Central Washington University. Her research focuses on sensory analysis and flavor chemistry, primarily related to dairy products. Her work investigates how flavor varies with processing and storage and how these variations relate to consumer perception. She employs qualitative market research methods such as descriptive analysis, consumer testing, and preference mapping, alongside instrumental flavor analysis techniques including gas chromatography/olfactometry and gas chromatography/mass spectroscopy, to relate sensory properties to the chemical components of foods. Her research encompasses both fundamental methods development and the application of these techniques to solve industrially relevant problems.

Research topics

• Food science
• Chemistry
• Mathematics
• Computer Science
• Marketing
• Chromatography
• Business
• Statistics
• Biochemistry
• Organic chemistry
• Economics
• Agricultural science
• Environmental science
• Biology
• Advertising
• Biotechnology

Selected publications

• Impact of lactose and soluble milk salts on the heat stability of milk proteins

  Journal of Dairy Science · 2025-11-08

  articleOpen accessSenior author

  Our objective was to determine the separate effects of lactose, soluble mineral concentration, and added dipotassium phosphate (DKP) on the heat stability of milk protein beverages made from rehydrated milk protein concentrate (MPC) with 85% crude protein on a DM basis (MPC85). Seven batches of rehydrated (7.5% protein) liquid MPC were made from each of 2 lots of MPC85 powder. Treatment 1 was 7.5% MPC85 in deionized (DI) water. Treatment 2 was 7.5% MPC85 that was rehydrated with 4°C Jenness-Koops (JK) buffer, a mix of salts that simulate milk salts. Treatment 3 was 7.5% MPC85 with JK buffer with 4.8% (wt/wt) anhydrous lactose added. Treatment 4 was 7.5% MPC85 with a dilution of JK buffer to ∼14% of its original concentration. Treatment 5 was 7.5% MPC85 with diluted JK buffer with lactose added to 0.6% (wt/wt). Treatment 6 was 7.5% MPC85 in DI water with 0.15% (wt/wt) DKP. Treatment 7 was 7.5% MPC85 rehydrated with 4°C UF permeate. Heat stability testing was conducted using an oil bath with a 30-s hold time at 6 temperatures: 135°C, 140°C, 145°C, 150°C, 155°C, and 160°C. The composition of the solute phase of rehydrated and smoothed dried MPC85 had a large impact on the heat stability and aggregation of milk proteins caused by UHT thermal processing temperatures. When MPC85 was rehydrated in deionized water, it had the highest heat stability and the least protein aggregation, and the heat stability was not increased by the addition of DKP. The heat stability of MPC85 was decreased by rehydration in a typical concentration of milk salts (i.e., JK buffer) and was further decreased by the addition of 4.8% anhydrous lactose to the JK buffer. When the MPC85 was rehydrated in UF milk permeate (i.e., milks salts + lactose), the heat stability was comparable to the JK buffer with added lactose. Using diluted JK buffer (∼14% of the level of milk salts and lactose) in MPC85 resulted in increased heat stability that was similar to rehydration in deionized water.

  PublisherOA PDFDOI

• Effects of storage time and temperature on the chemical and sensory properties of aseptic milk

  Journal of Dairy Science · 2025-03-03 · 6 citations

  articleOpen accessSenior author

  Our objective was to determine the effects of storage temperature and storage time on the chemical, physical, and sensory properties of ultra-high-temperature-direct steam injection aseptic milk. Milk was collected on 2 different processing dates (2 replicates) at a commercial aseptic milk processing facility immediately as containers came off the processing line. Milk was heat treated by direct steam injection (142°C for 3 s) with a flash vacuum cooling step following the holding tube and packaged aseptically in 946-mL aseptic packages. Packages were randomly assigned to 1 of 2 treatment groups (storage temperature of 4°C or 21°C). Half of the packages of 1% aseptic milk were cooled immediately off the processing line in ice and held at 4°C, and the other half were cooled to 21°C and held at 21°C for 12 mo. An unopened package of 1% aseptic milk that had been stored at 4°C or 21°C was opened and analyzed each month (for 12 mo) by chemical and descriptive sensory analyses. Chemical analyses included volatile compounds, viscosity, furosine, and dissolved oxygen. At 2 wk, 6 mo, and 12 mo, milk samples stored at each temperature were evaluated by consumers along with commercial 1% HTST milk. By descriptive sensory analysis, sulfur-eggy flavor decreased faster and caramelized flavor increased faster in aseptic milk stored at 21°C than in that stored at 4°C. Similarly, relative abundance of sulfur volatiles was lower initially in milk stored at 21°C than in that cooled and stored at 4°C. Furosine concentration was higher in milk stored at 21°C than in milk stored at 4°C. There were differences in consumer liking between the aseptic milk samples stored at the 2 storage temperatures, but they did not translate to an advantage in flavor liking for aseptic milk. Consumers indicated higher liking scores for fresh, refrigerated HTST milk than aseptic milk stored at either temperature at all 3 storage time points. Fluid milk processors need to focus on developing technology to increase the sensory liking scores of aseptic milk to maintain and increase fluid milk consumption as consumer lifestyles increase the demand for shelf-stable beverages.

  PublisherOA PDFDOI

• Interaction of fatty acid composition and temperature cycling on melting properties of milk fat

  Journal of Dairy Science · 2025-06-11 · 2 citations

  articleOpen accessSenior author

  Fat was extracted from milk produced by 40 individual cows that were specifically selected over time in lactation to produce a graduation of natural samples with a wider range of milk fatty acid composition (i.e., ratio of de novo to preformed fatty acids) than would be produced by changes in dairy cow feeding used under commercial dairy herd management conditions. The melting properties of the milk fats were determined using differential scanning calorimetry. Higher levels of de novo fatty acids (i.e., shorter chain) and higher levels of unsaturated fatty acids both caused softening of milk fat, but by different mechanisms. Unsaturation directly lowered the melting point by disrupting the bilayer (2L) and trilayer (3L) triglyceride crystallization process (polytypism) on first cooling, whereas differences in de novo fatty acid concentration likely caused a shift in the proportion of 2L and 3L crystal forms (polytypism) and affected polytrophic behavior. Milk fat with more de novo fatty acids was more responsive to softening by temperature cycling than milk fat with more unsaturation. The impact of de novo fatty acid concentration in milk fat and in the total milk supply should be considered in controlling the hardness of butter.

  PublisherOA PDFDOI

• Upcycled vs. Sustainable: Identifying Consumer Segments and Recognition of Sustainable and Upcycled Foods Within the United States

  Foods · 2025-10-15 · 2 citations

  articleOpen access

  Upcycled foods are a rising trend as the issue of food waste and sustainability becomes an increasing concern. The objective of this research was to evaluate the perception of upcycled foods through the characterization of consumer segments. An online questionnaire was developed to evaluate food neophobia, lifestyle, behavior, beliefs, awareness, and familiarity or recognition of upcycled and sustainable food products using a pre- and post-infographic intervention. The survey was created using Red Jade SaaS and distributed to participants through the Cal Poly Sensory database, North Carolina State University Sensory Service Center database, social media (LinkedIn and Nextdoor), and personal communication. Participants (n = 947) were segmented using a k-means clustering algorithm on lifestyle, neophobia, and beliefs questions. Four clusters were identified: Greenthusiasts (n = 306)—environmentally conscious and open-minded to new products, Balanced Believers (n = 347)—supportive of new products with reasonable doubt, Healthy Hesitants (n = 208)—willing to make health-focused changes but hesitant towards new products and technologies, and Eco-Skeptics (n = 86)—doubtful and resistant to change, most food and technologically neophobic. Based on pre- and post-intervention, Eco-Skeptics had the lowest initial awareness and recognition of sustainable and upcycled food products, while Greenthusiasts had the highest. All four segments trended towards improved recognition of the food products post-intervention.

  PublisherDOI

• Consumer perception of organic peanuts and peanut products: Identifying acceptance and purchase interest

  Applied Food Research · 2025-10-17

  articleOpen access

  • Consumer interest in organic food products is increasing • Organic peanuts are not readily available in the US • Consumers do not associate whole peanuts with branded products • Consumers at present are not willing to pay a premium for organic peanuts An online survey was conducted by the North Carolina State University Sensory Service Center to determine the purchase intent of consumers of peanuts and peanut-containing products that have been grown under organic conditions. From the surveys completed (313), it was determined that the highest number of consumers purchased peanuts either as peanut butter or as snack or energy bars. They considered organic foods to be more nutritious and safer than conventionally grown foods. The importance of organic peanuts and peanut products being available for purchase did not rate above no preference on a sliding scale. When asked about paying a premium price for organically produced peanut products, an increased price over that of conventionally produced products was small. The data indicates that the increased cost to growers in the North Carolina region to produce organically produced peanuts would not be justified at this time.

  PublisherDOI

• Comparison of Gins Using Temporal Dominance of Sensations ( <scp>TDS</scp> ), Temporal Check‐All‐That‐Apply ( <scp>TCATA</scp> ), and Temporal Ranking ( <scp>TR</scp> )

  Journal of Sensory Studies · 2025-03-04 · 1 citations

  articleOpen accessSenior authorCorresponding

  ABSTRACT Gin is a distilled spirit characterized by juniper flavor, often accompanied by other botanical flavors. The objective of this research was to use gin as a case study to compare temporal dominance of sensations (TDS), temporal check‐all‐that‐apply (TCATA), and temporal ranking (TR). Trained panelists used each of these methods to evaluate the temporal profile of six gins (standardized to 30% ABV). Differences in temporality were visualized by difference graphs for pairwise gin comparisons across methods. TDS differentiated gins based on dominant flavor attributes but did not fully describe the complexity of the gins. TCATA was effective in capturing differences in subtle flavor attributes but did not discriminate samples based on juniper flavor. TR differentiated gins based on differences in juniper flavor intensity and documented additional differences in complex gins but did not discriminate subtle flavor attributes. Researchers should consider these tradeoffs when selecting a temporal method.

  PublisherOA PDFDOI

• The Impact of Usage Method and Coffee Variety on Consumer Preferences of Coffee Creamers

  Journal of Sensory Studies · 2025-02-01

  articleOpen accessSenior authorCorresponding

  ABSTRACT Consumer evaluation of coffee creamers necessitates evaluation in a cup of coffee. Two phases of research were used to determine if method of serving creamer or coffee variety impacted the acceptance of four creamers. In Phase I, consumers evaluated creamers in medium roast coffee using one of three methods: (1) Fixed amount ( n = 127), (2) free pour ( n = 120), and (3) warm‐up ( n = 122). In Phase II, the serving method was constant (fixed amount), but consumers ( n = 134) evaluated creamers in two coffees (light and dark roast). Data were analyzed using univariate and multivariate statistics. No significant interactions ( p &gt; 0.05) were observed for overall liking (key metric) for either creamer × method (Phase I) or creamer × coffee (Phase II). Thus, the choice of serving method will not impact creamer preference. Further, the choice of coffee roast will not impact creamer preference, but creamers served in light roast coffee had higher liking scores than creamers served in dark roast coffee.

  PublisherOA PDFDOI

• A sedimentation test to measure heat stability of milk protein beverages

  Journal of Dairy Science · 2025-07-18 · 3 citations

  articleOpen accessSenior author

  A small-scale oil bath immersion-heating system using sealed stainless-steel process tubes was developed that has the flexibility to evaluate a wide range of holding temperatures and times. After heat treatment and rapid cooling, the liquid product was removed from each stainless-steel tube and a centrifugal protein sedimentation test and laser light scattering particle size analysis were done to determine differences among milk protein beverages in heat stability and casein micelle aggregation before and after thermal processing. The sedimentation and particle size analysis method was able to determine differences in protein sedimentation and casein micelle aggregation of milk protein concentrate (MPC) solutions with 85% protein on a DM basis (MPC85) that were rehydrated at 4°C overnight and heated to 55°C and smoothed by homogenization or not smoothed. The cold rehydrated homogenized MPC85 solutions had a casein micelle size particle distribution similar to skim milk. Smoothing reduced particle size before heat treatment and reduced heat-induced protein aggregation versus no smoothing. The oil bath method was applied to cold rehydrated and smoothed MPC85 high-protein beverages over a range of temperatures typical of retorting and UHT-direct steam injection processing with and without added dipotassium phosphate (DKP). The oil bath heat stability method was able to detect differences in sedimentation and casein micelle aggregation with increasing temperature and the large effect of added DKP in preventing protein sedimentation and casein micelle aggregation. Eight different lots of rehydrated commercial MPC85 powders were rehydrated and smoothed; the method was able to identify 2 out of 8 MPC85 powders that had lower heat stability and more casein micelle aggregation than the others. The heat stability method could be a practical tool to screen heat stability of small volumes of different protein beverage formulations and help avoid pilot plant or preliminary production runs on products that will not be heat stable.

  PublisherDOI

• Sources of soapy off‐flavor in clear acidic whey protein beverages

  Journal of Food Science · 2025-03-01 · 2 citations

  articleOpen accessSenior authorCorresponding

  Clear acidic protein beverages are made with whey protein isolate (WPI) due to its solubility at pH 3.3. These beverages can display a soapy off-flavor, which is not detected at neutral pH. Our hypothesis was that medium chain fatty acids are the source of this off-flavor. WPI (14 duplicate lots) were rehydrated to 10% solids (w/v), subjected to descriptive sensory analysis, and evaluated at their original pH, as well as pH adjusted to 6.3 (pre-acidif) or to 3.3 (neutral) at 21°C. Solvent-assisted flavor evaporation with gas chromatography mass spectrometry and gas chromatography olfactometry were performed on six selected WPIs (three neutral and three pre-acidified), followed by quantitation of octanoic, decanoic, and dodecanoic acids using stable isotope dilution assays. Retronasal and orthonasal thresholds of these acids were determined at pH 6.3 and 3.3. Model systems of WPI with the three acids were generated to confirm soapy flavor. The average concentrations of the three acids in selected WPIs were as follows: C8 = 14.4 ± 2.0 ppm; C10 = 27.8 ± 2.39 ppm; C12 = 24.7 ± 2.20 ppm. Sensory detection thresholds for all three acids at pH 3.3 were lower than at pH 6.3 (p < 0.05). Trained panel profiling of model systems demonstrated that a combination of octanoic and decanoic acids within the concentration range documented in WPI at pH 3.3 caused a soapy flavor, which was not detected at pH 6.3. Identification of the sources of soapy off-flavor in clear, acidic protein beverages provides the baseline to reduce or eliminate this off-flavor. PRACTICAL APPLICATION: Clear acidic protein beverages are one category of protein beverages within the expanding world of protein beverages. Identification of the source of soapy off-flavor in clear, acidic whey protein beverages is useful for consistent delivery of protein beverages that are desirable to consumers.

  PublisherOA PDFDOI

• Effects of storage time and temperature on the protein fraction of aseptic milk

  Journal of Dairy Science · 2025-06-11 · 2 citations

  articleOpen access

  <h2>ABSTRACT</h2> Milk was collected on 2 different processing dates (2 replicates) at a commercial aseptic milk processing facility immediately as containers came off the processing line. Milk was heat treated by direct steam injection (142°C for 3 s) with flash vacuum cooling. Half of the packages of 1% fat aseptic milk were cooled immediately in ice to 4°C, and half were cooled to 21°C; both were stored at these respective temperatures for 12 mo, and a new package was opened and analyzed monthly for 12 mo by Kjeldahl analysis for nitrogen fractions, particle size analysis for protein aggregation, visual observation of gelation, and SDS-PAGE to determine proteolytic damage to casein. Differences were found in rates of gelation at the 2 storage temperatures, but the end results of milk gelation were the same. Protein settling in the aseptic milk formed a gel layer that stuck to the bottom of the packages and caused the remaining liquid product poured from the containers to decrease in protein concentration with time of storage. Quantitative SDS-PAGE analysis did not show evidence of proteolysis from native milk proteases or heat-stable microbial proteases during product storage. Thus, nonproteolytic changes in the milk were responsible for age gelation of this product. Extensive heat-induced disulfide covalent binding of whey proteins to κ-casein occurred at the surface of the casein micelles, forming a larger, more hydrophilic covalently bonded κ-casein–whey protein complex on the surface of the casein micelles. Further increase in the hydrophilic character of this complex probably occurred due to thermally induced lactosylation of the whey proteins caused by UHT and the browning process. With storage time, the concentration of the highly hydrophilic κ-casein–whey protein complex increased in the serum phase of the milk, whereas the gel was enriched in the more hydrophobic α_s-casein and β-caseins.

  PublisherOA PDFDOI

Frequent coauthors

• Ronald Richter

  Rutgers, The State University of New Jersey

  72 shared

• Keith R. Cadwallader

  55 shared

• Patrick D. Gerard

  42 shared

• William R. Aimutis

  North Carolina State University

  37 shared

• Joseph O'donnell

  Cornell University

  36 shared

• A.F. Kertz

  36 shared

• Harold E. Swaisgood

  36 shared

• Adsa Represen- Tative

  Cornell University

  36 shared

Education

• Ph.D., Food, Bioprocessing and Nutrition Sciences

  North Carolina State University

  2005

• M.S., Food Science

  North Carolina State University

  1999

• B.S., Food Science

  North Carolina State University

  1997

Awards & honors

• William Neal Reynolds Distinguished Professor of Food Scienc…