REPFEDs (e.g., cooked–chilled, vacuum-packaged, sous vide, modified atmosphere-packaged foods) rely on mild heat treatments and refrigerated storage to preserve product quality and extend shelf life. However, non-proteolytic C. botulinum spores can survive mild thermal processing, and the pathogen is capable of growth and botulinum neurotoxin production at low temperatures.

The researchers sought to explore process and formulation conditions with the intent of clarifying a “gray area” of existing industry recommendations for controlling non-proteolytic C. botulinum; specifically, recommendations related to heat treatments below 90 °C for 10 minutes under slightly acidic conditions (i.e., pH values above 5).

Modeling C. botulinum Outgrowth Under Realistic Storage Conditions

The researchers developed a probabilistic model integrating previously published data and predictive models with distributions representing realistic domestic refrigerator temperatures and product consumption times among the European population. Variability and uncertainty were considered separately.

Under modeled domestic refrigerated storage conditions, for a product with a pH of 6.5 and water activity (a_w) of 0.999, a heat treatment of 85 °C for 18 minutes resulted in an estimated median time before outgrowth of 104 days (with a 95 percent uncertainty interval of 76–156 days).

For a product with a 28-day “Use By” date, the probability that outgrowth would occur before consumption was estimated at 0.06 percent (with a 95 percent uncertainty interval of 0.00–0.4 percent). This corresponded to an estimated inhibition level of 3.2 log colony forming units per gram (CFU/g).

At a constant storage temperature of 8 °C following the same heat treatment, the estimated median time before outgrowth was 65 days at pH 6.4 and 58 days at pH 6.5. The researchers noted that the underlying predictive model appeared conservative under the evaluated conditions, although they lacked sufficient data to properly validate its outputs.

Mild Heat and pH Hurdles Could Provide Significant Inhibition

The researchers also modeled combinations of heat treatments from 82–88 °C for 10 minutes and pH values from 6.0–7.0, with a_w held at 0.999.

A heat treatment of 84 °C for 10 minutes combined with a pH of 6.4 resulted in a median estimated inhibition of 3.1 log CFU/g. At pH 6.6, treatments of 87 °C and 88 °C for 10 minutes resulted in estimated inhibition levels of at least 3 log CFU/g.

The researchers also found that pH had an important effect on predicted inhibition. They noted that inhibition associated with formulation and processing hurdles could be considered alongside spore inactivation from thermal treatment when evaluating an overall degree of protection.

The modeling approach could therefore be used to estimate C. botulinum inhibition for existing REPFED formulations or to explore combinations of processing and formulation parameters when developing new products.

Study Limitations

The researchers cautioned that the study focused on consumer storage under European conditions and did not explicitly model proteolytic C. botulinum. Additionally, the probabilistic model was based on observational data rather than mechanistic elements and lacked direct validation against a sufficiently large dataset of REPFED products at the point of consumption.

Still, the researchers concluded that probabilistic modeling could help food businesses capitalize on existing scientific data to map processing and formulation conditions for controlling non-proteolytic C. botulinum in chilled foods.