Optimizing Pasteurized Fluid Milk Shelf-Life Through Microbial Spoilage Reduction

Psychrotolerant spore-forming bacteria, entering raw milk primarily on-farm, represent a major challenge for fluid milk processors due to the ability of these bacteria to survive heat treatments used for milk processing (e.g., pasteurization) and to cause premature spoilage. Importantly, fluid milk processors require tools to identify optimal strategies for reducing spore-forming bacteria, thereby extending product shelf-life by delaying spoilage. Potential strategies include (i) introducing farm-level premium payments (i.e., bonuses) based on spore-forming bacteria counts in raw milk and (ii) investing in spore reduction technologies at the processing level of the fluid milk supply chain. In this study, we apply an optimization methodology to the problem of milk spoilage due to psychrotolerant spore-forming bacteria and propose two novel mixed-integer linear programming models that assess improving milk shelf-life from the dairy processors' perspective. Our first model, imposed to a budgetary constraint, maximizes milk's shelf-life to cater to consumers who prefer milk with a long shelf-life. The second model minimizes the budget required to perform operations to produce milk with a shelf-life of a certain length geared to certain customers. We generate case studies based on real-world data from multiple sources and perform a comprehensive computational study to obtain optimal solutions for different processor sizes. Results demonstrate that optimal combinations of interventions are dependent on dairy processors' production volume and quality of raw milk from different producers. Thus, the developed models provide novel decision support tools that will aid individual processors in identifying the optimal approach to achieving a desired milk shelf-life given their specific production conditions and motivations for shelf-life extension.

No Associations Between Dairy Intake and Markers of Gastrointestinal Inflammation in Healthy Adult Cohort

Objectives We examined associations between reported dairy intake and markers of gastrointestinal (GI) inflammation in a healthy human adult population. We hypothesized there would be a negative association of yogurt intake with GI inflammation, suggesting a protective effect, and no associations of total dairy, fluid milk, and cheese intake with GI inflammation. Methods Participants completed up to 3 unannounced 24-hour dietary recalls using ASA24 and a Block 2014 Food Frequency Questionnaire (FFQ) to assess recent and habitual intake, respectively. Those who also provided a stool sample were included in this analysis (n = 342). Stool samples were stored on ice immediately after collection and homogenized within 24 hours. Inflammatory markers from stool, including calprotectin, neopterin, and myeloperoxidase were measured by ELISA along with LPS-binding protein (LBP) from plasma. Regression models tested associations between dairy intake variables and inflammatory markers with and without covariates: sex, age, and body mass index (BMI). As yogurt is episodically consumed, t-tests were used to examine differences in inflammatory marker levels between consumers (>1 cup per month reported in FFQ) and non-consumers. Each dairy variable was also expressed as a percentage of mean energy intake from 24-hour recalls and regressed with inflammatory markers. Results Without covariates, we found no associations between total dairy, fluid milk, and yogurt from both dietary assessments with any inflammatory markers. Cheese intake reported in the FFQ was positively associated with plasma LBP (P = 0.02). However, this association was not significant after covariate adjustment. There were no significant differences in GI inflammatory marker levels between yogurt consumers and non-consumers (P > 0.05). When expressed as a percentage of mean energy intake, cheese from ASA24 was associated with increased LBP (P = 0.03), but this was not significant after adjustment for covariates. There were no other associations of dairy variables as a percentage of energy intake with GI inflammatory markers. Conclusions We found no clinically relevant associations between dairy intake and markers of GI inflammation in a healthy human adult cohort. Funding Sources California Dairy Research Foundation, USDA ARS 2032–51,530-026–00D.

Using threshold co-integration to estimate asymmetric price transmission in the Turkish milk market

We investigate the price dynamics between retail milk price and raw milk price in the Turkish fluid milk market. The study uses monthly fluid milk prices for 14 years between January 2003 and December 2016. We analyze the price adjustment in the fluid milk market through an asymmetric error correction model with threshold co-integration. We find that the transmission between the two prices has been asymmetric in both the long term and short term period. Differences between the farm milk prices and retail milk prices may exist due to marketing costs across the supply chain and pricing policies associated with the market structure. Results of the long-run analysis indicate a significant market power in the fluid milk market. Therefore, in this asymmetric case, the deviations are likely to be the reason for the market power of the processors/retailers and the reason for the oligopolistic market structure in the sector.

Machine learning and advanced statistical modeling can identify key quality management practices that affect post-pasteurization contamination of fluid milk

Spoilage of HTST- (high-temperature, short-time) and vat- pasteurized fluid milk due to introduction of Gram-negative bacteria post-pasteurization remains a challenge for the dairy industry. While processing facility level practices (e.g., sanitation practices) are known to impact the frequency of post-pasteurization contamination (PPC), the relative importance of different practices is not well defined, affecting the ability of facilities to select intervention targets that reduce PPC and provide the greatest return on investment. Thus, the goal of this study was to use an existing longitudinal dataset of bacterial spoilage indicators obtained for pasteurized fluid milk samples collected from 23 processing facilities between July 2015 and November 2017 (with 3 to 5 samplings per facility) and data from a survey on fluid milk quality management practices, to identify factors associated with PPC and rank their relative importance, using two separate approaches: (i) multimodel inference and (ii) conditional random forest. Data pre-processing for multimodel inference analysis showed (i) nearly all factors were significantly associated with PPC when assessed individually using univariable logistic regression and (ii) numerous pairs of factors were strongly associated with each other (Cramer’s V ³0.80). Multimodel inference and conditional random forest analyses identified similar drivers associated with PPC; factors identified as most important based on these analyses included cleaning and sanitation practices, activities related to good manufacturing practices, container type (which is a proxy for different filling equipment), in-house finished product testing, and designation of a quality department, indicating potential targets for reducing PPC. In addition, this study illustrates how machine learning approaches can be used with highly correlated and unbalanced data, as typical for food safety and quality, to facilitate improved data analyses and decision-making.