An ordinal logistic regression approach to predict the variability on biofilm formation stages by five Salmonella enterica strains on polypropylene and glass surfaces as affected by pH, temperature and NaCl

Abstract

This study assessed the adhesion and formation of biofilm by five Salmonella enterica strains (S. Enteritidis 132, S. Infantis 176, S. Typhimurium 177, S. Heidelberg 281 and S. Corvallis 297) on polypropylene (PP) and glass (G) surfaces as affected by pH (4-7), NaCl concentration (0-10% w/v) and temperature (8-35 °C). Sessile counts <3 log CFU/cm² were considered lack of adhesion (category 1), while counts ≥ 3 and < 5 log CFU/cm² corresponded to adhesion (category 2) and counts ≥ 5 log CFU/cm² corresponded biofilm formation (category 3). The obtained results categorized in these three responses were used to develop ordinal regression models to predict the probability of biofilm stages on PP- and G-surfaces. The experimental outcomes for lack of adhesion were >90% on PP- and G-surfaces. Generally, adhesion outcomes corresponded to approximately 36% of the total, whereas biofilm outcomes were close to 65% in both PP- and G-surfaces. The biofilm stages varied among the strains studied and with the material surface under the same experimental conditions. According to the generated ordinal models, the probability of adhesion and biofilm formation on PP-surface by the five S. enterica strains tested decreased at pH 4 or 5 in NaCl concentrations >4% and at a temperature <20 °C. On G-surface, the probability of adhesion increased pH 6 or 7, in the absence of NaCl and temperatures <20 °C, while, the probability of biofilm formation increased in the same pH, NaCl concentration up to 4% and temperatures ≥20 °C. This is the first study assessing the biofilm formation through categorical, ordinal responses and it shows that ordinal regression models can be useful to predict biofilm stages of S. enterica as a function of pH, NaCl, and temperature or their interactions.

Keywords: Attachment; Cleaning; Cross contamination; Disinfection; Food safety; Predictive microbiology.