Purpose: The purpose of this study was to assess the risk of salmonellosis associated with the consumption of dry roasted sunflower seeds using both laboratory data and QMRA mathematical models.
Methods: Sunflower seed samples were individually inoculated with cocktails of Salmonella spp. to achieve ~108 CFU/g. Dry roasting parameters included: Oven settings of 225°F or 275°F and roasting times of 5 to 45 min. The dry roasting study was repeated 3 times and Salmonella spp. were enumerated using scientifically valid methods. Prediction intervals (95%) were used to estimate the range of expected log reduction during dry roasting. Data on sunflower seed consumption and Salmonella prevalence and concentration on sunflower seeds were based on data from published sources, internal company data or expert opinion. A QMRA model was developed using SAS software. A variety of what-if scenarios were evaluated.
Results: Applying minimum 4-log reduction to sunflower seeds reduced the average number of Salmonella positive servings significantly (P < 0.05) compared to consumption of raw sunflower seeds. Predicted salmonellosis cases from dry roasted sunflower seeds were less than one per year, and based on a worse-case scenario (99th percentile), given the assumed parameters, the model estimates that a single salmonellosis case would occur about every 5 years. When CDC under-reporting factors are considered, the model estimates that one salmonellosis case from dry roasted sunflower seeds would be detected every 100 years.
Significance: The quantitative risk analysis estimations indicate that the risk of salmonellosis based from properly dry roasted sunflower seeds is extremely low. The techniques used in this study can provide a scientific basis for food safety plans in the sunflower dry roasting industry.