Eur Food Res Technol (2006) 224: 91–100 DOI 10.1007/s00217-006-0293-1 ORIGINAL PAPER A. Valero · E. Carrasco · F. P´ erez-Rodriguez · R. M. Garc´ ıa-Gimeno · G. Zurera Growth/no growth model of Listeria monocytogenes as a function of temperature, pH, citric acid and ascorbic acid Received: 14 July 2005 / Revised: 31 January 2006 / Accepted: 13 February 2006 / Published online: 10 March 2006 C  Springer-Verlag 2006 Abstract A linear logistic regression model was built to describe the growth/no growth boundaries of Listeria monocytogenes as a function of temperature (4–30 ◦ C), pH (4.5–6), citric acid (0–0.4%) and ascorbic acid (0–0.4%). A fractional factorial design was followed among the fac- tors considered and an inoculum size of 10 5 CFU/ml was used. Evaluation of growth was performed by optical den- sity measurements in Bioscreen C (Labsystems, Finland), during 21 days. Data of optical density were transformed to log CFU/ml by using a calibration curve. Among the 232 combinations of factors tested, growth was observed in 133 and no growth in 99. The degree of agreement be- tween predictions and observations was 97.8% concordant and 2.2% discordant. An internal validation with additional data within the interpolation region of the model was per- formed. The predictions were concordant in 94% of the cases, and all the wrong cases failed to the safe side of the boundary region. The probability of growth was strongly influenced at low temperatures (<15 ◦ C). The effect of pH was more notorious below 5.3, since the minimum temper- ature that inhibited growth was higher. Citric acid was more effective than ascorbic acid when the analysis was based on the undisssociated acid concentration (u.a.c.). However, when dealing with the undissociated fraction (u.a.f.), ascor- bic acid presented more inhibitory effect. Organic acids also accentuate the temperature and pH inhibition of bacte- rial growth limits, and increased the minimum pH at which growth was detected. These results have a practical implica- tion for stakeholders and risk managers in order to identify the treatments that can be applied to food and ensure that no growth of this pathogenic microorganism will occur. A. Valero · E. Carrasco · F. P´ erez-Rodriguez · R. M. Garc´ ıa-Gimeno · G. Zurera () Department of Food Science and Technology, University of C´ ordoba, Campus de Rabanales, C-1, 14014 Cordoba, Spain e-mail: bt1zucog@uco.es Fax: +34-957-212000 Keywords Probability model . Predictive microbiology . Listeria monocytogenes . Food safety Introduction Listeria monocytogenes is a microorganism of concern for food industries due to its ubiquity in the natural environ- ment [1] and its capability to grow in a wide variety of con- ditions, which explains its high prevalence in different kinds of food products. Raw meat, soft cheese or minimally pro- cessed vegetables are frequently implicated in foodborne outbreaks [2]. The long shelf-life and the physico-chemical characteristics of these foods allow growth and survival of L. monocytogenes, because it can grow at refrigeration tem- peratures, reaching dangerous levels for consumer’s health [3]. Due to its high pathogenicity, the level of L. mono- cytogenes present in food should remain low in order to guarantee safe food products. In the 1990s some models defined absolute limits to mi- crobial growth in a multifactorial space. The cases of liste- riosis worldwide posed the development of these kinds of models [4], and different strategies were proposed to limit levels of contamination at the time of consumption at less than 100 CFU/g [5]. Predictive microbiology focuses on determining the be- haviour of a given microorganism, combining mathemat- ical modelling with experimental data under some envi- ronmental factors. Predictive microbiology can be divided into two kinds of models in this respect [4]: Kinetic models and probabilistic models. The former intends to determine the microbiological life of a food product. The latter de- termines whether the microbial growth might occur or not in a specific range of conditions. This is because micro- bial growth is restricted to finite limits of factors, and even growth sometimes declines abruptly at a very small incre- ment in the level of each factor. This concept is known as ‘hurdles technologies’ and it was quantified by McMeekin et al. [6]. The effect of combining hurdles may be additive or synergistic on preventing the growth of microorganisms