Sensory shelf life estimation Ana Giménez, Gastón Ares Sensometrics & Consumer Science, Instituto Polo Tecnológico de Pando, Universidad de la República, Canelones, Uruguay
Chapter outline 1 Introduction 333 2 Sensory shelf life estimation 334 2.1 Experimental designs for sensory shelf life experiments 335
3 Methodologies for sensory shelf life estimation 337 3.1 Sensory shelf life estimation based on analytic tests 338 3.2 Sensory shelf life estimation based on hedonic tests 344 3.3 Sensory shelf life estimation based on the combination of analytic and hedonic tests 349
4 Recommendations and challenges 351 References 352 Further reading 357
1 Introduction Shelf life of a food product can be defined as the time period acceptable eating quality is retained, from a safety, nutritional, and sensory standpoint (IFST, 1993). During this time period, the product should provide consumers its intended sensory experience, performance and benefits (ASTM E2454-05, 2011). Shelf life dates are compulsorily included on the labels of prepackaged food products in most countries as a means to provide consumers with an anchor point that provides guarantees on their quality and safety (Department for Environment, Food and Rural Affairs, 2011). Two typologies of shelf life dates exist: use by and best before. According to the EU label system, use by date is determined by safety aspects, whereas best before date is related to sensory and nutritional quality (European Commission, 2000). Food manufacturers are responsible for determining the type of shelf life date appropriate for a specific food product (European Commission, 2002). Therefore, accurately estimating shelf life dates poses a challenge to food scientists, since manufacturers, retailers, and consumers rely on this information along the food production and consumption chain. Manufacturers need to assure products meet food safety criteria, along with quality thresholds that ensure consumer satisfaction throughout the labeled shelf life. Brand and product loyalty might be compromised if consumer expectations are not met (Harcar and Karakaya, 2005). In an attempt to reduce this risk, manufacturers tend to become more conservative when establishing the maximum period of time a product can be stored, which leads to an excessively short shelf life (Newsome et al., 2014). In addition, in many instances, the need to introduce new products to the market within a Food Quality and Shelf Life. https://doi.org/10.1016/B978-0-12-817190-5.00011-2 © 2019 Elsevier Inc. All rights reserved.
Food Quality and Shelf Life
short time frame leads companies to establish shelf life based on prior experience with similar products. These practices might lead to an early retrieval of a food product from the marketplace that is still acceptable for consumption, from sensory, nutritional, and safety perspectives. In fact, empirical data shows that a considerable amount of products are discarded from the market if they are not sold within their shelf life (Spada et al., 2018). This implies both economic and environmental consequences, particularly given the need to reduce food waste to achieve a sustainable transformation to keep human development within safe boundaries (Rockström et al., 2009). Consumers look for shelf life dates on food labels motivated by their interest for fresh, natural, safe, and superior quality food products. Shelf life dates are a proxy for healthiness, nutritional value, freshness, and safety during their decision making (Alongi et al., 2018; Giménez et al., 2012; Milne, 2013; Newsome et al., 2014; Wansink and Wright, 2006; Wilson et al., 2017; Ragaert et al., 2004). In addition, shelf life dating can influence acceptability, taste perceptions, and perceived quality of food products (Priefer et al., 2016; Wansink and Wright, 2006). Research has shown that shelf life labeling is one of the reasons underlying food waste at the household level, as a proportion of consumers tend to discard products when they are past their shelf life, without even trying them (Gaiani et al., 2018). Strategic efforts to prolong shelf life and improve consumer understanding of shelf life dating typologies have been highlighted as potential contributors to reducing food waste worldwide and improving the sustainability of the food sector (Spada et al., 2018; Gaiani et al., 2018). In this sense, recent research has shown that shelf life dating focused on maximizing consumer acceptance and minimizing changes with respect to the fresh product are usually too conservative, which can lead to unnecessary food waste (Man, 2016). During storage, food products change as a consequence of biological, enzymatic, and physicochemical reactions that take place (Labuza and Szybist, 2001). These changes might impact the nutritional, microbiological, or sensory quality of a food product, limiting the time period it should be consumed. However, sensory changes are the ones limiting the shelf life of most food products (Hough, 2010). According to Hough and Garitta (2012), “once the sanitary and nutritional hurdles have been overcome, the remaining barrier depends on the sensory properties of the product”. For this reason, best before dates based on sensory shelf life are the most common in packaged food products. Although some sensory changes may go unnoticed by consumers, others would affect food quality perception. Thus, information on how the characteristics of the products change throughout storage is needed to determine when a food product no longer has the intended sensory characteristics and may be rejected by consumers. In the upcoming sections, methodologies for shelf life estimation are presented and the methodological considerations needed to obtain accurate results are discussed.
2 Sensory shelf life estimation Sensory shelf life can be defined as the period of time a product retains its intended sensory characteristics under specific storage conditions. Therefore, sensory shelf life
Sensory shelf life estimation335
estimation aims at determining the specific point in time when changes in the sensory characteristics are no longer acceptable. A typical shelf life study involves evaluating sensory or hedonic characteristics of food samples stored for different periods of time. In order to design this kind of study adequately, a number of issues should be carefully addressed in advance. The total length of the study should surpass the expected shelf life of the product. This is necessary to identify the time point at which the product reaches the minimum acceptable sensory quality or failure criterion. The frequency of sampling should be such that it allows estimates to have the desired level of precision. For this reason, a minimum of six sampling times is generally advised (Hough, 2010; Kilcast, 2011). Sampling can be equally distributed along the total storage period or an increased sampling frequency can be used as the product approaches the end of its expected shelf life. This last approach can provide more useful information than equally spaced sampling times in the case of products with long shelf lives or that deteriorate following an exponential kinetics (Gacula and Singh, 1984). As Robertson (2010) pointed out, a key concept when designing a shelf life study is to minimize the cost and time of the testing, yet provide reliable and statistically valid data.
2.1 Experimental designs for sensory shelf life experiments The samples included in the study should be selected taking into account production variability and study duration. Two different experimental designs can be used for obtaining samples with different storage times when conducting a sensory shelf life study: basic and reversed design.
2.1.1 Basic design The basic storage design requires a single batch of product to be stored under certain pre-established storage conditions. Samples are extracted at each of the selected sampling times and analyzed from a sensory and/or hedonic perspective, using the methodologies described later in the chapter. This design is particularly useful when dealing with products that largely vary across batches (e.g., fresh products). Examples of application include minimally processed vegetables (Ares et al., 2009; Paulsen et al., 2018), apple juice (Ferrario and Guerrero, 2016), yogurt (Salvador et al., 2005), and cod (Østli et al., 2013). Although basic design is the most widely used for sensory shelf life estimation, it is not efficient regarding the use of time and resources (Lawless and Heymann, 2010). Basic design can be particularly cumbersome in the case of sensory methods, as they require gathering a panel of trained assessors or consumers at each of the storage times. In addition, as stated by Hough (2010), participants might realize they are participating in a sensory shelf life study and provide biased results or change their criteria, particularly if long storage times are required or sampling periods are infrequent. These difficulties could lead to expectation errors, and inconsistent use of scales and
Food Quality and Shelf Life
learning effects. An alternative to minimize these biases is to include one or more fresh samples in each evaluation.
2.1.2 Reversed design This kind of testing design allows all samples with different storage times to be evaluated in one single instance, reducing the cost and resources needed for the sensory measurements. This can be achieved in three different ways: 1. Drawing samples from different production batches and storing them under the desired conditions at the preestablished sampling times so that all samples with different storage times are evaluated on the same day (Fig. 1A). In this case, consistent production quality is usually sought for assuring minimal batch variability. 2. Storing samples from one single batch under conditions that stop all deterioration processes and removing samples at each selected sampling time to store them under normal conditions until their evaluation (Fig. 1B). 3. Storing samples from one single batch under the desired conditions and once the elapsed time for sampling has been reached, keeping the sample under conditions where no further deterioration occurs (Fig. 1C).
In the first case, differences among samples are expected to be due to deterioration processes undergone during storage and differences between batches assumed to be small. Gámbaro et al. (2005) followed this approach for estimating the sensory shelf life of “alfajor,” a chocolate-coated cake (individually wrapped), working with different industrial batches stored at 20.0 ± 0.5°C for 0, 28, 46, 60, 65, 70, 75, and 80 days. Fresh samples were stored at the different selected times, so that samples with eight different storage times were evaluated on the same day. The other two alternatives are only applicable to some product categories, for which deterioration processes can be stopped without modifying the sensory characteristics of the products. For some categories, freezing or storing at very low refrigeration temperatures might be a feasible option for stopping deterioration. The storage conditions selected should guarantee not only that product aging stops, but also that the freezing-defrosting cycle does not introduce sensory changes. An example of the application of a reversed design was reported by Jacobo-Velázquez and Hernández-Brenes (2011), when estimating the shelf life of high hydrostatic pressure processed avocado pulp at 4°C. These authors stored processed avocado pulp under refrigerated conditions for 45 days, collecting samples every 5 days and storing them at −80°C until they were all analyzed at the end of the storage period (45 days). The reversed design overcomes the main drawbacks of the basic storage design. All samples are tested on the same day, reducing time and resources necessary to conduct the testing. It is particularly useful when consumer studies are used to estimate sensory shelf life of food products. However, selecting storage conditions that halt all deterioration processes might not always be possible. Products such as fresh or minimally processed fruits and vegetables cannot be frozen to stop their aging process.
Sensory shelf life estimation337
Evaluation of samples
Evaluation of samples
No deterioration Storage times at condition (e.g. –18°C) 20°C
Storage times at 20°C
(A) Initial batch
Evaluation of samples
Storage times No deterioration at 20°C condition (e.g. –18°C))
Fig. 1 Shelf life test reversed designs: (A) samples from different batches are removed at each selected sampling time and stored until all samples are gathered for evaluation; (B) samples from one single batch are stored under a nondeterioration condition and subsets are removed to selected storage condition when sampling time is reached; (C) samples from one single batch are stored in selected conditions until each of the sampling times is reached and each subset is removed to a nondeterioration condition until evaluation.
3 Methodologies for sensory shelf life estimation Sensory evaluation offers a toolbox from which to draw when carrying out sensory shelf life testing. Sensory evaluation can be defined as a scientific discipline that “evokes, measures, analyzes, and interprets responses to the characteristics of products as perceived by the senses” (Stone and Sidel, 2004). This discipline can be
Food Quality and Shelf Life
divided into two areas: analytical tests that provide objective information about the sensory characteristics of products, and hedonic tests, which evaluate consumers’ hedonic reaction towards products (Ares and Varela, 2017). Sensory shelf life estimation has been based on both analytical and hedonic tests, as will be presented later in this section. Regardless of the methodological approach, standard sensory evaluation practices should be followed in order to gather reliable data from which accurate and reproducible sensory shelf life estimations can be drawn (Lawless and Heymann, 2010). Readers not familiar with sensory evaluation techniques should refer to textbooks for detailed information about sensory methodologies and recommendations for best practice (e.g., Stone and Sidel, 2004; Lawless and Heymann, 2010). Sensory shelf life estimation requires establishing a sensory endpoint or a failure criterion, that is, the maximum tolerable deterioration in sensory quality. Each methodology involves a different failure criterion, which means a different sensory quality at the end of a product’s shelf life. This has economic and environmental implications that researchers should be aware of. In the following sections, the implications of each methodology are discussed, together with their advantages and disadvantages.
3.1 Sensory shelf life estimation based on analytic tests These methods are based on the objective evaluation of changes over time in the sensory quality of products or in the intensity of the main sensory characteristics responsible for product deterioration. Analytic tests should be conducted with panels of trained assessors, who should be screened and selected based on their sensory acuity for basic sensory characteristics and their ability to detect differences among samples (Stone and Sidel, 2004). Then, assessors should be extensively trained to evaluate the sensory characteristics of products using specific methodological approaches. In the case of sensory shelf life estimation, three methodologies have been extensively used: difference from control, rating overall quality, and quantifying attribute intensity.
3.1.1 Difference from control test The difference from control test has been traditionally applied to sensory quality control of food products (Meilgaard et al., 1991; Yantis, 1992). An overall degree of difference from a control product is rated by a trained assessor panel, generally using a 10 cm scale, and a criterion for product acceptance/rejection is agreed. In the case of sensory shelf life studies, the degree of difference between stored samples and a fresh control sample is measured. In order to apply this methodology, several issues are to be taken into account. First is the need to keep a fresh sample to serve as control as storage time progresses. Depending on the kind of product under study, this might be a relatively simple task; in some product categories, storage conditions can be easily selected to guarantee no deterioration takes place. The selected condition to keep a control sample should be previously tested in order to assure no significant sensory changes occur. For example, Man (2015) suggested storing a control product at −18°C when studying the shelf life of a cooked chilled potato dish, whereas Chouliara et al. (2007) kept a control sample
Sensory shelf life estimation339
at −30°C when studying the effect of modified atmosphere packaging and oregano essential oil addition on the shelf life of refrigerated chicken breast meat. However, it should be stressed that keeping the same fresh control over the whole study period might not be possible for other product categories. In those cases, an alternative would be to have the control sample replaced periodically, but the replacement should be validated using discriminative tests to assure that no differences in the sensory characteristics of the control samples exist. An example of this approach can be found in Hough et al. (1999). These authors kept a control sample of ricotta cheese at 2°C–3°C for 7 days without significant sensory changes when conducting a sensory shelf life study of this product category. In order to replace the control sample, every 7 days a triangle test with a trained panel was performed and a new fresh sample replaced the previous control sample. However, this approach would be difficult to implement in the case of natural fresh products, such as fruit and vegetables. Second, assessors should undergo an adequately designed training program prior to engaging in this kind of task. It is crucial that they become familiar, not only with the characteristics of the control sample, but also with the sensory characteristics of samples representing different points along the scale. During the training sessions, samples with different storage times are evaluated and through open discussion with the panel leader, assessors agree on the scores to be assigned to the different samples when compared to the control sample. A blind control sample is to be included in every test session as part of the test set, serving as a baseline. During the shelf life study, assessors receive the control sample labeled as K, and the stored samples and the blind control sample labeled with 3-digit numbers. They are asked to rate the samples as to the degree of difference from the control sample, generally using a 10 cm line scale. Additional verbal descriptions of different degrees of difference are sometimes included along the scale. Data from this test can be analyzed by analysis of variance and stored samples significantly different from the control sample can be identified. However, these data do not provide any information on the reasons for the difference or whether those differences are relevant to the consumer. Additional questions exploring attributes perceived as different—even providing a scale to rate attribute difference—may be added. A cut-off score that defines the maximum tolerable difference between a stored sample and the fresh control should be set in advance to establish the end of the product’s shelf life. The average difference between the stored samples and the fresh sample could be regressed over time and the sensory shelf life estimated as the time the stored product reaches the cut-off point (Hough et al., 1999; Freitas and Costa, 2006). Practitioners are advised to avoid arbitrarily selection of this failure criterion and to rely on information about consumer perception for its definition. Other methodologies can be used to estimate differences between stored samples and the fresh control. In particular, discriminative tests such as paired comparisons, triangle, tetrad, or duo-trio are useful alternative approaches. These methodologies are frequently used in industrial environments, being easier to implement than difference from control scales (Kilcast, 2011; Bouillé and Beeren, 2016). An example of the application of this approach is the estimation of “high-quality life” (HQL) for chilled and frozen food products. HQL is defined as “the time elapsed between freezing of an
Food Quality and Shelf Life
initially high-quality product and the moment when, by sensory assessment, a statistically significant difference (p < .01) from the initial quality can be established” (IIR, 1999; Evans, 2016). In closing this section, it should be emphasized that sensory limits established when working with this kind of methodology usually lead to conservative estimations that might impact on an early withdrawal of food products from the shelf, leading to undesired financial and environmental consequences.
3.1.2 Overall quality scoring This method is based on direct measurement of sensory quality throughout storage. In order to provide valid and reliable information, researchers should define quality specifications and set tolerance limits in advance. The trained assessors’ task is to rate sensory quality by weighting different sensory characteristics of the products and integrating them into a single score. This requires trained assessors to have a mental reference of the sensory characteristics of the fresh product and to know which weight to assign to the sensory changes the product undergoes during storage when evaluating overall sensory quality. This task is particularly complex considering that an increase or decrease of characteristic sensory attributes or the onset of sensory defects can affect overall quality differently. Extensive training and a minimum number of experts or trained assessors are required (8 to 20) to get accurate and reliable results (Lawless and Heymann, 2010). Quality scores are averaged across assessors and regressed over time. Sensory shelf life can be estimated as the time point when sensory quality reaches a specific quality score or failure criterion, which corresponds to the minimum sensory quality of a commercial product. The failure criterion is usually arbitrarily determined (e.g., the mid-point of the quality scale), which is not a recommended practice. Quality-based methods have been used to estimate the sensory shelf life of a variety of products such as fresh fruits and vegetables, dairy, meat, and sea products. Allende et al. (2008) evaluated the sensory quality of fresh cut escarole and iceberg lettuce during storage with a trained panel of eight assessors, measuring the overall visual quality with a 10-point scale (0 = very bad, not characteristic, 10 = very good, very characteristic of the product). The end of the sensory shelf life was reached when the overall visual quality of the products reached a score of 5. Similarly, Medina et al. (2012) used this approach for sensory shelf life estimation of minimally processed baby spinach. One of the most popular applications of quality-based methods for sensory shelf life estimation is the Quality Index Method (QIM), widely used in the fish industry. This method measures the degree of change in selected sensory attributes using a scoring system that ranges from 0 to 3 (the lower the score, the fresher the fish). A qualitative description of each score for each sensory attribute is provided to simplify and standardize the evaluation. Demerit scores are summed up to a total to provide an overall evaluation. A QIM scheme has to be developed for each species, since it does not take into account the differences between species. Considering that this method does not take into account the differences between species, QIMs have been developed
Sensory shelf life estimation341
for a large number of commercially important species (Hyldig et al., 2010) and they continue to be developed as commercial interest in other species grows (e.g., Borges et al., 2013; Baixas-Nogueras et al., 2003; Ritter et al., 2016; Sant'Ana et al., 2011; Fogaça et al., 2017). Ritter et al. (2016) developed a QIM protocol to study the sensory shelf life of tambatinga fish, working with 12 panelists. As a first training step, panelists were asked to observe and describe the sensory changes perceived in fish samples with six different storage times for 12 sessions. A second training stage was carried out, adjusting the preliminary QIM protocol, working with 18 fish samples during eight sessions. This stage was followed by a validation stage where assessors had to evaluate nine samples with different storage times using the 13 quality parameters they had previously defined (Table 1). Although quality-based methods have been frequently applied for sensory shelf life estimations of different product categories, a large number of published studies do not comply with recommendations for best practice in sensory evaluation. This reduces the validity of the sensory shelf life estimation and signals the need for advancements in the application of standard practices when dealing with sensory trained panels, particularly for the evaluation of sensory quality. It is not unusual to find published research where three to five assessors have performed the quality scoring task with little or no training (Gómez-López et al., 2007; Lopez-Galvez et al., 2013; Medina et al., 2012; Mendes et al., 2011) and, in some cases, hedonic information gathered from trained assessors has been regarded as an indication of product quality (Costa et al., 2012; Gómez-López et al., 2007; Siripatrawan and Noipha, 2012). Such drawbacks could limit the validity of the data and the accuracy of sensory shelf life estimations.
3.1.3 Attribute intensity measurement Descriptive methodologies aim at providing a complete qualitative and quantitative description of the sensory characteristics of products (Lawless and Heymann, 2010). In the specific case of sensory shelf life estimation, descriptive methodologies can be used to obtained a detailed description of the sensory changes a product undergoes during storage. These methods provide intensity ratings for individual sensory attributes, enabling the estimation of sensory shelf life, as the storage time when the intensity of a critical attribute reaches a preestablished value (Giménez et al., 2012). In order to do so, specifications must be set identifying the sensory attributes considered critical for sensory quality and establishing intensity limits, that is, the maximum or minimum intensity of sensory attributes of a product that is still acceptable for consumption. The use of intensity scales demands 8–12 assessors to be selected and extensively trained in attribute identification and scaling. First, a set of descriptors are developed by the trained assessors, working with a wide array of samples with different storage times to identify those sensory characteristics that best describe differences among samples (ISO, 2016). Assessors individually select the descriptors they consider appropriate to describe the differences among samples and then, through an open discussion with the panel leader, agree on the terms to be used. Once descriptors are selected, a training stage should follow where the panelists are exposed to reference standards that anchor descriptors’ intensity. Unstructured or structured line scales
Food Quality and Shelf Life
Table 1 Quality Index Method (QIM) protocol to study sensory shelf life of gutted ice-stored tambatinga (Colassoma macrosporum x Piaractus brachypomum) (Ritter et al., 2016) Quality attributes
Intense Bright Opaque Firm/elastic Softened Convex Flat/irregular Crystalline Cloudy Opaque Black/well-defined Reddish/well-defined Black/not-defined Red blood Brownish red Brownish Fresh/neutral Rancid/sour Putrid Whitish Light salmon Grayish pink Fresh/neutral Rancid/sour Putrid Full/wet/with elasticity Full/resected tips/with elasticity Failed/resected tips/no elasticity Sum of all the scores
0 1 2 0 1 0 1 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1
Godet test Eyes
(10 or 15 cm) are generally used to rate attribute intensity. Panelists are trained to the use of the intensity scales for the selected attributes using intensity references and selected samples. Only when panel performance is considered reliable and consistent are the samples under study evaluated. Once the panel is trained, the average intensity of the sensory attributes at each sampling time is calculated and regressed over time (Fig. 2). Sensory shelf life can be determined as the time point when the intensity of a sensory attribute reaches the predetermined criterion. Researchers are strongly advised to avoid the selection of arbitrary failure criterion for attribute intensity. Several authors have used this methodology to estimate sensory shelf life of different food products (Cardenas Bonilla et al., 2007; Ares et al., 2009; Garitta et al.,
Sensory shelf life estimation343
Fig. 2 Average off-flavor intensity of orange juice with different storage time, evaluated by a trained assessors’ panel using 10 cm intensity scales. A reversed design based on the evaluation of different batches (c.f. Fig. 1A) was used.
2004). Jacobo-Velázquez and Hernández-Brenes (2011) followed this procedure when estimating shelf life of high hydrostatic pressure processed avocado paste. A trained panel developed a list of six descriptors related to changes during storage, working with samples with different storage times. A training phase followed, where assessors were trained to recognize those descriptors and their references in several sessions, using a 10 cm scale (Table 2). In a third phase, through open discussion with the panel leader, two of the descriptors, sour and rancid, were agreed to be critical. Assessors were then trained with samples that covered the whole range of intensities of the two Table 2 Reference samples to train assessors on avocado descriptors (Jacobo-Velázquez et al., 2010) Descriptor
Preparation of reference sample
Freshly HHPa-processed avocado paste (100 g) was allowed to oxidize at room temperature until the paste changed to brown color (approximately 30 min). Browning was stopped by adding citric acid (1.2 g) to the avocado paste. Fresh and brown avocado pastes were used as reference samples and were visually evaluated by the trained panel Caffeine (3000 ppm) was mixed into freshly HHP-processed avocado paste Citric acid (12,500 ppm) was mixed into freshly HHP-processed avocado paste Hexanal (200 ppm) was mixed into freshly HHP-processed avocado paste Sucrose (50,000 ppm) was mixed into freshly HHP-processed avocado paste Canola oil (50,000 ppm) was mixed into freshly HHP-processed avocado paste
Bitter Sour Rancid Sweet Oily a
HHP: High hydrostatic pressure.
Food Quality and Shelf Life
critical descriptors, until they provided reproducible and consistent results. Finally, the trained panel was requested to evaluate the samples under study. The failure criterion for sensory shelf life estimation was selected based on consumer data, following the approach that will be discussed later in the chapter. The focus when using this methodology is on the sensory profile of the test samples and the insights it provides regarding a product’s deterioration processes. Unlike other applications of descriptive analysis, when applied to shelf life studies, the number of descriptors is usually limited to those that are critical, generally ranging in number from 3 to 6 (Hough, 2010). A relatively small number of descriptors helps reduce the time and resources needed to maintain a well-trained and calibrated panel. The reader could refer to other sensory methodologies when time or resources hinder the proper development of a highly trained panel. Although the method provides rich, valid, and reliable information, several published studies do not follow recommendations for best practice in sensory science. It is not uncommon to find published research where a reduced number of assessors is used (Medina et al., 2012; Costa et al., 2011, 2012) or where hedonic expressions are used to quantify attribute intensities (Chouliara et al., 2007; Costa et al., 2011). These practices compromise the accuracy and reliability of sensory shelf life estimations. In addition, several studies have asked trained assessors for hedonic information—asking them to decide whether a product was acceptable or not for consumption—and established a shelf life limit based on the panel’s average acceptability (Allende et al., 2008; Chouliara et al., 2007; Costa et al., 2017; El-Obeid et al., 2018; Lopez-Galvez et al., 2013; Vandekinderen et al., 2009).
3.2 Sensory shelf life estimation based on hedonic tests Hedonic tests gather information directly from the consumer: how much a product is liked or disliked, how it is perceived, whether it is accepted or rejected for consumption, whether it would be bought or not. The consumer perspective becomes an invaluable asset when estimating how long a food product should remain on the shelf (retail and home). The sensory changes that take place during storage do not necessarily modify how much a product is liked. In addition, trained assessors are expected to be more sensitive than consumers (Lawless and Heymann, 2010). Therefore, the sensory changes a trained panel detects and quantifies could go unnoticed to consumers or even if noticed, consumers might not care about them. For this reason, establishing a shelf life limit based solely on differences perceived by a trained panel could become too conservative and reduce the commercial life of a food product, which has negative economic and environmental consequences. One of the important concepts that researchers should acknowledge is that sensory shelf life does not depend only on how the sensory characteristics of the product change over time, but rather on how consumers react to these changes, as they are the ones who decide whether a product after a certain storage time is still acceptable or not (Hough et al., 2003). Consumers should determine whether a product is still acceptable for consumption from a sensory perspective or how much it is liked after a certain period of storage time has elapsed. Consumer perception has been recognized as a
Sensory shelf life estimation345
key input for the selection of failure criteria for shelf life estimation, as they are “the ultimate arbiters of food quality” (Robertson, 2010). Two consumer methodologies can be used for sensory shelf life estimation: acceptability limit and survival analysis. Both methodologies require evaluations performed by large groups of naïve consumers, who should evaluate samples based on their subjective perception.
3.2.1 Acceptability limit Consumers’ hedonic reaction towards a food product can be measured by asking their degree of liking using a hedonic scale (Lawless and Heymann, 2010). The most widely used hedonic scale is the traditional 9-point category scale, with phrases ranging from dislike extremely to like extremely and a neutral point at the center, developed at the U.S. Army Food and Container Institute (Peryam and Girardot, 1952). Alternatively, an anchored 9-box scale (Fig. 3) could also be used (Curia et al., 2001; Giménez et al., 2007). Consumers should be chosen to represent the target consumer population. A minimum of 100 consumers is generally recommended to obtain reliable information (Hough et al., 2006). Participants are presented with a set of samples with different storage times and are asked to score their overall liking using the above-mentioned scale. Depending on the sensory shelf life design, the product, and the time frame, it might be possible to have the same consumers taste all test samples. Consumers may also be asked to score their liking of different attributes, or even provide additional information on how different sensory characteristics are perceived, by means of additional tasks, for example, answering a check-all-that-apply (CATA) question. Overall acceptability scores for each sample are averaged, and a linear or nonlinear regression of overall liking against time is usually carried out (Giménez et al., 2007; Montes Villanueva and Trindade, 2010). The shelf life of the product is defined as the storage time at which the product’s overall acceptability falls below a previously set value. In order to set this overall acceptability limit, authors have used different criteria. Some authors have used an acceptability value of 6.5 on the 9-point scale, a limit initially proposed by Muñoz et al. (1992) for quality control specifications, whereas a number of authors have considered an overall liking score below 6 (like slightly) (Tomac et al., 2017; Gámbaro et al., 2005; Giménez et al., 2008). Montes Villanueva and Trindade (2010) and Latou et al. (2014) chose an average score of 5.0 (neither like nor dislike) as the acceptability limit. Fig. 4 provides an example of shelf life estimation using the acceptability limit methodology. The selection of a single critical acceptability limit as cut-off point across product categories has certain drawbacks. For some food categories, overall acceptability of the fresh product might be in a range close to the cutoff point, which in turn could lead to too conservative shelf life estimations that limit their commercial life. On the other
Fig. 3 Anchored hedonic box scale for determining consumer overall liking of food products.
Food Quality and Shelf Life
Fig. 4 Average overall liking scores of toast bread samples with different storage time (following a basic design), evaluated by consumers using 9-point hedonic scores. Sensory shelf life was estimated as the storage time when overall liking reached a score of 6 (9 days).
hand, consumers would probably reject products that are usually well-liked and have high overall liking scores when fresh before a low acceptability score is reached. Hough et al. (2002) proposed a different criterion to establish an overall acceptability limit. The first significant difference in overall acceptability from that of the fresh product is calculated and the sensory shelf life is defined as the time at which this acceptability score is reached. The acceptability limit (S) is calculated using the following equation: 2.MSE (1) n where S is the minimum tolerable acceptability of stored sample; F is acceptability of fresh sample; Zα is one-tailed coordinate of the normal curve for α significance level; MSE is the mean square of the error derived from the analysis of variance of the consumer data; and n is the number of consumers. This criterion reflects the time consumers perceive a difference in the stored product with respect to the fresh one that relates to changes in its sensory characteristics during storage. It should be noted that considering this criterion can lead to overly conservative estimations. A slight decrease in acceptability does not necessarily relate to consumers rejecting or discarding the product for consumption. Giménez et al. (2007) compared different failure criteria to establish the shelf life of brown pan bread and concluded that by the time the acceptability limit was reached a relatively low percentage of consumers rejected the product, providing an estimation that was not compatible with commercial practices. Other authors have also reported similar conclusions (Garitta et al., 2015) and recommended other bases to calculate cut-off points, as will be discussed later. S F Z
3.2.2 Survival analysis Overall liking scores provide no insight as to what a consumer would do when confronting a food product that has been stored for some time, whether he/she would
Sensory shelf life estimation347
d ecide to eat it or discard it. Alternatively, sensory shelf life can be estimated considering consumers’ rejection to consume or purchase a product. Hough et al. (2003) were the first to apply survival analysis statistics to sensory shelf life, shifting the focus from the product deteriorating to the consumer decision of accepting or rejecting the product for consumption. Survival analysis statistics, extensively used in clinical studies, epidemiology, biology, sociology, and reliability studies (Klein and Moeschberger, 1997; Meeker and Escobar, 1998) could be applied to sensory shelf life estimations based on consumers’ acceptance or rejection of samples with different storage times. Participants are asked whether they would accept or reject the different samples for consumption. They are usually instructed to answer considering they have already bought the product or the product is being served at home (Hough et al., 2003). Early publications using survival analysis to estimate sensory shelf life of food products worked with 50–80 consumers. However, Hough et al. (2007) performed simulation studies on the ideal number of consumers to recruit for this kind of study, and they concluded a number close to 120 consumers is recommended. In those cases where a reversed storage design is used, each consumer evaluates the whole sample set in a single session. If a basic storage design is followed, then the consumer panel should be assembled on each sampling occasion. Another alternative is to have different consumers evaluate a single sample at each of the storage times. Libertino et al. (2011) performed simulation studies to estimate the number of consumers to obtain valid estimations and recommended a minimum of 50 for each of storage time, considering a total of six storage times. This approach would be also useful in those studies where a fresh control cannot be kept (Araneda et al., 2008) or for certain types of sample sets where a significant carry-over effect could be expected. Survival analysis has been extensively used in the literature to estimate sensory shelf life of a wide variety of food products such as dairy (Garitta et al., 2004; Curia et al., 2005; Salvador et al., 2005), bakery (Giménez et al., 2008, 2007, 2017; Gámbaro et al., 2005), minimally processed fruits and vegetables (Lareo et al., 2009; Ares et al., 2009; Jacobo-Velázquez and Hernández-Brenes, 2011; Garitta et al., 2018), and fish (Østli et al., 2013). The test setup is simple, participants are presented with a randomized set of samples with different storage times and asked to answer the question “would you normally consume this product” with a yes or no answer. The data gathered from these studies are summarized in a table like Table 3, where the acceptance/rejection data for each consumer is included in each row (acceptance = Yes, rejection = No). The exact moment at which a consumer rejects a product is not actually observed, considering a consumer evaluates a fixed number of samples with different storage times providing, as a result, censored data (Hough et al., 2003). Data collected is classified in three categories: left, interval, and right censored. If a consumer rejects the sample at the first storage time considered, then rejection was not observed and happened before that first observation time (T ≤ t1), and that data is classified as left censored. If a consumer accepts a sample at t2 and rejects a sample at t3, the exact time at which he/she rejects the product occurs between t2 and t3 (t2 < T ≤ t3) and his/her data is interval censored. If rejection happens after the last storage time (T > t6), then the data is classified as right censored (Table 3).
Food Quality and Shelf Life
Table 3 Typical data matrix obtained from survival analysis Storage time Consumer
1 2 3 … n
No Yes Yes … Yes
No Yes Yes … Yes
No No No … Yes
No No Yes … Yes
No No No … Yes
No No No … Yes
Left Interval Interval … Right
Acceptance = Yes; rejection = No.
Survival analysis allows estimation of the rejection function, and modeling of the proportion of consumers that reject a food product as a function of storage time. As storage time increases, the proportion of consumers that reject the product increases as well. If a random variable T is defined as the storage time a consumer rejects the sample, the survival function S(t) can be defined as the probability of a consumer accepting a product stored for a time period longer than t, that is S(t) = P(T > t). To estimate the parameters of the survival function, the likelihood function is used. The likelihood function describes the joint probability of obtaining the given observations for the n consumers (Klein and Moeschberger, 1997): L S ri 1 S li S li S ri iR
where:R is the set of right-censored observations; L is the set of left-censored observations; and I is the set of interval-censored observations. Each type of censoring makes a different contribution to the likelihood function. The distribution of the survival times could be assumed to follow a Weibull or log-normal distribution and estimates of the of the survival function S(t) could be obtained (Klein and Moeschberger, 1997). If the Weibull distribution is chosen, the survival function is: ln t S t Ssev
where:Ssev (•) is the survival function of the smallest extreme value distribution: Ssev (w) = exp (−ew); and μ and σ are the model’s parameters. When the lognormal distribution is chosen for T, then the survival function is: ln t S t 1
where: Φ(.) is the standard normal cumulative distribution function; and μ (location parameter) and σ (shape parameter) are the model’s parameters. For a given set of experimental data, the model’s parameters are estimated by maximizing the likelihood function and substituting S(t) in Eq. (2) by the expression
Sensory shelf life estimation349
Fig. 5 Consumers’ rejection to consume as a function of storage time for samples of dulce de leche in plastic packages (basic design).
given in Eqs. (3) or (4), depending on the model chosen, using specialized software. Calculations using R software have been provided by Hough (2010). These parameters are used to graph the proportion of consumers rejecting the product as a function of storage time, as shown in Fig. 5. In order to estimate the shelf life of the product, a probability level of consumers’ rejection should be set. A shelf life value considering 50% rejection has been recommended in the literature (Gacula and Singh, 1984; Cardelli and Labuza, 2001; Hough et al., 2003). Other authors (Ares et al., 2006; Giménez et al., 2007; Jacobo-Velázquez et al., 2010; Giménez et al., 2017) have used a 25% rejection to make shelf life estimations; a more conservative criterion, but still considered reasonable from a practical point of view.
3.3 Sensory shelf life estimation based on the combination of analytic and hedonic tests This approach builds on the combination of the methodologies described above. It aims at obtaining less arbitrary failure criterion for sensory attributes or sensory quality using consumer perception as input. Sensory changes during storage can be evaluated with a panel of assessors, provided that attributes that limit the product’s shelf life are identified. The intensity of these critical attributes may increase, as would the occurrence of off-flavors or offodors, or decrease during storage, as would the loss of aroma or desirable texture attributes. Hough and Garitta (2012) proposed, as a first step, to obtain samples with different intensity levels of the critical descriptors. These samples are to be evaluated by a trained panel for attribute intensity and by a consumer panel for acceptability. Analytical or instrumental measurements may be gathered as well, since this kind of data is simpler and easier to collect. However, it is not always possible to rely on
Food Quality and Shelf Life
analytical and instrumental measurements to monitor changes during storage. A regression of the acceptability data versus the trained sensory panel data is performed and the sensory cut-off point is obtained by introducing a failure criterion for overall liking (6 or the minimum difference in overall liking scores) in the regression equation (Fig. 6A). This cut-off point indicates a limit on the intensity of a sensory or instrumental parameter above which (or below, for a positive characteristic) there is a significant decrease in consumer acceptability. Finally, a regression of the sensory or instrumental data versus storage time allows prediction of the sensory shelf life as the storage time necessary for the product to reach that cut-off point intensity of the measured parameter (Fig. 6B). A limitation of this approach is that the estimation would depend on the critical descriptor selected. Considering all the sensory characteristics of the product change simultaneously, and that their impact altogether on consumers’ reaction can be different
(B) Fig. 6 (A) Correlation between overall liking and firmness of hamburger buns with different storage time (reversed design based on frozen samples, Fig. 1B). The cut-off corresponding to an overall liking score of 6 is shown; and (B) Firmness of hamburger bans with different storage time. The shelf life corresponding to the firmness score determined using the cut-off point is shown.
Sensory shelf life estimation351
from the impact of one high intensity single sensory defect, selecting the most relevant attribute might be cumbersome. A single measurement of the sensory deterioration of products could provide a single failure criterion. Ares et al. (2009) developed a sensory quality index for strawberries based on the correlation between trained assessors’ data and consumer acceptability scores. Instead of using one single critical descriptor to establish a sensory cut-off limit, the authors performed a principal c omponent analysis (PCA) on the mean scores of the trained assessors’ data followed by a multiple regression analysis considering acceptability as a dependent variable, and sample loadings for the first two principal components of the sensory data as explanatory variables. In this case, the authors selected a score of 6 as acceptability limit and were able to predict a sensory quality index value that could, in turn, be correlated to storage conditions to predict the product’s shelf life. Similarly, Derossi et al. (2016) applied PCA to sensory data to define a cut off criteria for shelf life estimation of fresh cut lettuce. This same approach can be used by combining other analytic and hedonic tests. For example, survival analysis based on consumer acceptance/rejection data can be used to estimate cut-off points for sensory attributes (Garitta et al., 2015). Similarly, in order to relate sensory quality data as perceived by trained assessors to consumer perception, some authors have proposed to integrate consumer data with quality indexes. For instance, Djekic et al. (2018) developed a total quality index (TQI) to evaluate the quality of sliced apples dried with different technologies during storage. Their approach took into account instrumental and sensory parameters, selecting quality parameters according to consumers’ previous ranking of attribute importance to the sensory quality of the product. A panel of eight trained assessors used a 5-level quality scale to score the four sensory parameters selected (appearance, odor, oral texture, and flavor). A mathematical model was developed for calculating a single TQI using quality scores. Earlier, Ares et al. (2009) developed a sensory quality index based for strawberries correlating consumer and trained panel data.
4 Recommendations and challenges For most food products, shelf life is determined by changes in their sensory characteristics throughout storage. Sensory shelf life estimation has traditionally focused on assuring that minimum changes in the sensory characteristics of products occur during storage. This approach can lead to an unnecessarily high amount of food waste, since consumers may still consume a stored product that is different from the fresh one. In this sense, shelf life extension has been identified as a potential strategy to reduce food waste worldwide, thus contributing to an improvement of the sustainability of the food sector (Spada et al., 2018; Gaiani et al., 2018). Therefore, sensory shelf life estimation should aim at maximizing the storage period, while assuring that consumers would not find an unacceptable product within its stated shelf life. For this reason, practitioners are advised to use consumer input to determine the maximum tolerable deterioration for consumers. In this sense, consumer-based methodologies or methodologies that combine analytic and hedonic tests are recommended over those based exclusively on analytic tests.
Food Quality and Shelf Life
One of the main challenges of consumer-based methods for sensory shelf life estimation is the definition of the failure criterion, that is, the minimum overall liking score or the maximum percentage of consumers rejecting the sample. Research is necessary to understand better the implications of setting different failure criteria from both an economic and environmental perspective. In this sense, information about the amount of food discarded at the retailer and household levels for different failure criteria is necessary to refine the methodologies used for estimating sensory shelf life. In addition, another challenge for improving the accuracy of sensory-shelf life estimations is to apply more ecologically valid approaches to measure consumer perception of products with different degrees of deterioration. This requires acknowledging that a wide range of variables beyond sensory characteristics influence consumer perception and choice, including personal characteristics (e.g., socio-economic level, gender, age), product variables (e.g., brand, price, packaging), and context (e.g., where, when, and how the product is consumed). In line with that, virtual reality and evoked context studies attempt to provide industry a better prediction of consumers’ behavior (Galiñanes Plaza et al., 2018). The recent use of evoked contexts in sensory shelf life studies (Giménez et al., 2015) has shown that providing context to currently used methodologies could lead to more reliable estimations. Finally, it is important to highlight that accurate sensory shelf life estimation requires the application of methodologies that comply with recommendations and best practices in sensory and consumer science. However, there are a large number of examples of studies published in peer-reviewed journals reporting invalid estimations due to methodological flaws. Improving the methodological approaches used for sensory shelf life estimation will bring benefits at the academic and industrial level, as well as for society as a whole.
References Allende, A., Selma, M.V., Lopez-Galvez, F., Villaescusa, R., Gil, M.I., 2008. Role of commercial sanitizers and washing systems on epiphytic microorganisms and sensory quality of fresh-cut escarole and lettuce. Postharvest Biol. Technol. 49, 155–163. Alongi, M., Sillani, S., Lagazio, C., Manzocco, L., 2018. Effect of expiry date communication on acceptability and waste of fresh-cut lettuce during storage at different temperatures. Food Res. Int. https://doi.org/10.1016/j.foodres.2018.09.056. in press. Araneda, M., Hough, G., Wittig de Penna, E., 2008. Current-status survival analysis methodology applied to estimating sensory shelf life of ready-to-eat lettuce (Lactuca sativa). J. Sens. Stud. 23, 162–170. Ares, G., Varela, P., 2017. Trained vs. consumer panels for analytical testing: fueling a long lasting debate in the field. Food Qual. Prefer. 61, 79–86. Ares, G., Parentelli, C., Gámbaro, A., Lareo, C., Lema, P., 2006. Sensory shelf life of shiitake mushrooms stored under passive modified atmosphere. Postharvest Biol. Technol. 41, 191–197. Ares, G., Barrios, S., Lareo, C., Lema, P., 2009. Development of a sensory quality index for strawberries based on correlation between sensory data and consumer perception. Postharvest Biol. Technol. 52, 97–102.
Sensory shelf life estimation353
ASTM E2454–05, 2011. Standard Guide for Sensory Evaluation Methods to Determine the Sensory Shelf Life of Consumer Products. ASTM International, West Conshohocken, PA. 2011, http://www.astm.org. Baixas-Nogueras, S., Bover-Cid, S., Veciana-Nogués, T., Nunes, M.L., Vidal-Carou, M.C., 2003. Development of a quality index method to evaluate freshness in Mediterranean hake (Merluccius merluccius). J. Food Sci. 68, 1067–1071. Borges, A., Conte-Júnior, C.A., Franco, R.M., Freitas, M.Q., 2013. Quality index method (QIM) developed for pacu Piaractus mesopotamicus and determination of its shelf life. Food Res. Int. 54, 311–317. Bouillé, A.G., Beeren, C.J.M., 2016. Sensory evaluation methods for food and beverage shelf life assessment. In: Subramaniam, P. (Ed.), The Stability and Shelf Life of Food. second ed. Woodhead Publishing Series in Food Science, Technology and Nutrition: Number 297, Elsevier Ltd., Kidlington, UK, pp. 199–224. Cardelli, C., Labuza, T.P., 2001. Application of Weibull hazard analysis to the determination of the shelf life of roasted and ground coffee. Lebensm.-Wiss. Technol. 34, 273–278. Cardenas Bonilla, A.C., Sveinsdottir, K., Martinsdottir, E., 2007. Development of Quality Index Method (QIM) scheme for fresh cod (Gadus morhua) fillets and application in shelf life study. Food Control 18, 352–358. Chouliara, E., Karatapanis, A., Savvaidis, I., Kontominas, M.G., 2007. Combined effect of oregano oil and modified atmosphere packaging on shelf-life extension of fresh chicken breast meat, stored at 4°C. Food Microbiol. 24 (6), 607–617. Costa, C., Lucera, A., Conte, A., Mastromatteo, M., Speranza, B., Antonacci, A., Del Nobile, M.A., 2011. Effects of passive and active modified atmosphere packaging conditions on ready-to-eat table grape. J. Food Eng. 102, 115–121. Costa, C., Conte, A., Buonocore, G.G., Lavorgna, M., Del Nobile, M.A., 2012. Calciumalginate coating loaded with silver–montmorillonite nanoparticles to prolong the shelf-life of fresh-cut carrots. Food Res. Int. 48, 164–169. Costa, C., Lucera, A., Licciardello, F., Conte, A., Del Nobile, M.A., 2017. Application of preservation strategies to improve the shelf life of spreadable cheese. Food Packaging Shelf Life 11, 16–20. Curia, A.V., Hough, G., Martínez, M.C., Margalef, M.I., 2001. How argentine consumers understand the Spanish translation of the 9-point hedonic scale. Food Qual. Prefer. 12, 217–221. Curia, A., Aguerrido, M., Langohr, K., Hough, G., 2005. Survival analysis applied to sensory shelf life of yogurts I: argentine formulations. J. Food Sci. 70, s442–s445. Department for Environment, Food and Rural Affairs, 2011. Guidance on the Application of Date Labels to Food. DEFRA, London, pp. 1–26. Derossi, A., Mastrandrea, L., Amodio, M.L., de Chiara, M.L.V., Colelli, G., 2016. Application of multivariate accelerated test for the shelf life estimation of fresh-cut lettuce. J. Food Eng. 169, 122–130. Djekic, I., Tomic, N., Bourdoux, S., Spilimbergo, S., Smigic, N., Udovicki, B., Hofland, G., Devlieghere, F., Rajkovic, A., 2018. Comparison of three types of drying (supercritical CO2, air and freeze) on the quality of dried apple—quality index approach. LWT- Food Sci. Technol. 94, 64–72. El-Obeid, T., Yehia, H., Sakkas, H., Lambrianidi, L., Tsiraki, M., Savvaidis, I., 2018. Shelf-life of smoked eel fillets treated with chitosan or thyme oil. Int. J. Biol. Macromol. 114, 578–583. European Commission, 2000. European Commission. Directive 2000/13/EC of the European Parliament and of the council of 20 march 2000 on the approximation of the laws of the member states relating to the labelling, presentation and advertising of foodstuffs. Off. J. Eur. Union 109, 29–41.
Food Quality and Shelf Life
European Commission, 2002. Commission Regulation (EC) No 178/2002 of 28 January 2002 laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety. Off. J. Eur. Communities 31, 1–24. Evans, J., 2016. Effects of food and beverage storage, distribution, display, and consumer handling on shelf life. In: Subramaniam, P. (Ed.), The Stability and Shelf Life of Foods, second ed. Woodhead Publishing Series in Food Science, Technology and Nutrition: Number 297, Elsevier ltd, Kidlington, UK, pp. 107–157. Ferrario, M., Guerrero, S., 2016. Effect of a continuous flow-through pulsed light system combined with ultrasound on microbial survivability, color and sensory shelf life of apple juice. Innov. Food Sci. Emerg. Technol. 34, 210–224. Fogaça, F.H.D.S., Gonzaga Junior, M.A., Vieira, S.G.A., Farias, E., Ferreira-Bravo, I., Silva, T., Calvet, R., Pereira, A.L.M., Prentice-Hernández, C., 2017. Appraising the shelf life of farmed cobia, Rachycentron canadum, by application of a quality index method. J. World Aquacult. Soc. 48 (1), 70–82. Freitas, M.A., Costa, J.C., 2006. Shelf life determination using sensory evaluation scores: a general Weibull modeling approach. Comput. Ind. Eng. 51, 652–670. Gacula, M.C., Singh, J., 1984. Statistical Methods in Food and Consumer Research. Academic Press, New York, NY. Gaiani, S., Caldeira, S., Adorno, V., Segrè, A., Vittuari, M., 2018. Food wasters: profiling consumers’ attitude to waste food in Italy. Waste Manag. 72, 17–24. Galiñanes Plaza, A., Delarue, J., Saulais, L., 2018. The pursuit of ecological validity through contextual methodologies. Food Qual. Prefer. https://doi.org/10.1016/j.foodqual.2018.11.004. in press. Gámbaro, A., Garitta, L., Giménez, A., Varela, P., Hough, G., 2005. Sensory shelf-life estimation of Alfajor by survival analysis. J. Sens. Stud. 19, 500–509. Garitta, L., Hough, L., Sánchez, R., 2004. Sensory shelf life of dulce de leche. J. Dairy Sci. 87, 1601–1607. Garitta, L., Langhor, K., Gómez, G., Hough, G., Beeren, C., 2015. Sensory cut-off point obtained from survival analysis statistics. Food Qual. Prefer. 43, 135–140. Garitta, L., Langhor, K., Elizagoyen, E., Ottaviano, F., Gómez, G., Hough, G., 2018. Survival analysis model to estimate sensory shelf life with temperature and illumination as accelerating factors. Food Qual. Prefer. 68, 371–376. Giménez, A., Varela, P., Salvador, A., Ares, G., Fiszman, S., Garitta, L., 2007. Shelf life estimation of brown pan bread: a consumer approach. Food Qual. Prefer. 18, 196–204. Giménez, A., Ares, G., Gámbaro, A., 2008. Survival analysis to estimate sensory shelf life using acceptability scores. J. Sens. Stud. 23, 571–582. Giménez, A., Ares, F., Ares, G., 2012. Sensory shelf-life estimation: a review of current methodological approaches. Food Res. Int. 49, 311–325. Giménez, A., Gagliardi, A., Ares, G., 2015. Influence of evoked contexts on consumers’ rejection of two products: implications for shelf life estimations. Food Res. Int. 76 (3), 527–531. Giménez, A., Gagliardi, A., Ares, G., 2017. Estimation of failure criteria in multivariate sensory shelf life testing using survival analysis. Food Res. Int. 99, 542–549. Gómez-López, V., Ragaert, P., Debevere, J., Devlieghere, F., 2007. Pulsed light for food decontamination: a review. Trends Food Sci. Technol. 18, 464–473. Harcar, T., Karakaya, F., 2005. A cross-cultural exploration of attitudes toward product expiration dates. Psychol. Mark. 22, 353–371. Hough, G., 2010. Sensory Shelf Life Estimation of Food Products. Taylor & Francis Group, Boca Raton, FL.
Sensory shelf life estimation355
Hough, G., Garitta, L., 2012. Methodology for sensory shelf-life estimation: a review. J. Sens. Stud. 27, 137–147. Hough, G., Puglieso, M.L., Sánchez, R., Mendes Da Silva, O., 1999. Sensory and microbiological shelf-life of a commercial ricotta cheese. J. Dairy Sci. 82, 454–459. Hough, G., Sánchez, R.H., Garbarini De Pablo, G., Sánchez, R.G., Calderón Villaplana, S., et al., 2002. Consumer acceptability versus trained sensory panel scores of powdered milk shelf-life defects. J. Dairy Sci. 85, 1–6. Hough, G., Langohr, K., Gómez, G., Curia, A., 2003. Survival analysis applied to sensory shelf life of foods. J. Food Sci. 68, 359–362. Hough, G., Wakeling, I., Mucci, A., Chambers IV, E., Méndez-Gallardo, I., Alves, L., 2006. Number of consumers for sensory acceptability tests. Food Qual. Prefer. 17, 522–526. Hough, G., Calle, M.L., Serrat, C., Curia, A., 2007. Number of consumers necessary for shelf life estimations based on survival analysis statistics. Food Qual. Prefer. 18, 771–775. Hyldig, G., Martinsdóttir, E., Sveinsdóttir, K., Schelvis, R., Bremner, A., 2010. Quality index methods. In: Nollet, L.M.L., Toldrá, F. (Eds.), Handbook of Seafood and Seafood Products Analysis. CRC Press, New Jersey, pp. 463–481. IFST, 1993. Shelf Life of Foods: Guidelines for Its Determination and Prediction. Institute of Food Science and Technology, London. IIR, 1999. Recommendations for the Processing and Handling of Frozen Foods. IIF-IIR, France. ISO, 2016. Sensory Analysis—Methodology—General Guidance for Establishing a Sensory Profile. European committee for standardization, Brussels, Belgium. ISO 13299: 2016. Jacobo-Velázquez, D.A., Hernández-Brenes, C., 2011. Sensory shelf-life limiting factor of high hydrostatic pressure processed avocado paste. J. Food Sci. 76, s388–s395. Jacobo-Velázquez, D.A., Ramos-Parra, P.A., Hernández-Brenes, C., 2010. Survival analysis applied to the sensory shelf-life dating of high hydrostatic pressure processed avocado and mango pulps. J. Food Sci. 75, s286–s291. Kilcast, D., 2011. Sensory evaluation methods for food shelf life assessment. In: Kilcast, D., Subramaniam, P. (Eds.), Food and Beverage Stability and Shelf Life. Woodhead Publishing, pp. 350–380. Klein, J.P., Moeschberger, M.L., 1997. Survival Analysis, a Self-Learning Text. SpringerVerlag, New York. Labuza, T.P., Szybist, L.M., 2001. Open Dating of Foods. Food & Nutrition Press, Trumbull, CT. Lareo, C., Ares, G., Ferrando, L., Lema, P., Gámbaro, A., Soubes, M., 2009. Influence of temperature on shelf life of butterhead lettuce leaves under passive modified atmosphere packaging. J. Food Qual. 32, 240–261. Latou, E., Mexis, S.F., Badeka, A.V., Kontakos, S., Kontominas, M.G., 2014. Combined effect of chitosan and mofdified atmosphere packaging for shelf life extension of chicken breast fillets. LWT- Food Sci. Technol. 55, 263–268. Lawless, H.T., Heymann, H., 2010. Sensory Evaluation of Food. Principles and Practices, second ed. Springer, New York. Libertino, L.M., López Osornio, M.M., Hough, G., 2011. Number of consumers necessary for survival analysis estimations based on each consumer evaluating a single sample. Food Qual. Prefer. 22, 24–30. Lopez-Galvez, F., Ragaert, P., Palermo, L., Eriksson, M., Devlieghere, F., 2013. Effect of new sanitizing formulations on quality of fresh-cut iceberg lettuce. Postharvest Biol. Technol. 85, 102–108. Man, D., 2015. Shelf Life, second ed. John Wiley & Sons, Ltd, West Sussex, UK, p. 81. Man, D., 2016. Food storage trials. In: Subramaniam, P. (Ed.), The Stability and Shelf Life of Food, second ed. Woodhead Publishing Series in Food Science, Technology and Nutrition: Number 297, Elsevier Ltd., Kidlington, UK, pp. 171–198.
Food Quality and Shelf Life
Medina, M.S., Tudela, J.A., Marín, A., Allende, A., Gil, M.I., 2012. Short postharvest storage under low relative humidity improves quality and shelf life of minimally processed baby spinach (Spinacia oleracea L.). Postharvest Biol. Technol. 67, 1–9. Meeker, W.Q., Escobar, L.A., 1998. Statistical Methods for Reliability Data. John Wiley & Sons, New York. Meilgaard, M., Civille, G.V., Carr, B.T., 1991. Sensory Evaluation Techniques. CRC Press, Inc., Boca Raton, Florida. Mendes, R., Alves Silva, H., Anacleto, P., Cardoso, C., 2011. Effect of CO2 dissolution on the shelf life of ready-to-eat Octopus vulgaris. Innov. Food Sci. Emerg. Technol. 12, 551–561. Milne, R., 2013. Arbiters of waste. Date labels, the consumer and knowing good, safe food. Sociol. Rev. 60 (2), 84–101. Montes Villanueva, N.D., Trindade, M.A., 2010. Estimating sensory shelf life of chocolate and carrot cupcakes using acceptance tests. J. Sens. Stud. 25, 260–279. Muñoz, A.M., Civille, G.V., Carr, B.T., 1992. Sensory Evaluation in Quality Control. Van Nostrand Reinhold, New York. Newsome, R., Balestrini, C.G., Baum, M.D., Corby, J., Fisher, W., Goodburn, K., et al., 2014. Applications and perceptions of date labeling of food. Compr. Rev. Food Sci. Food Saf. 13 (4), 745–769. Østli, J., Esaissen, M., Garitta, L., Nøstvold, B., Hough, G., 2013. How fresh is fresh? Perceptions and experience when buying and consuming fresh cod fillets. Food Qual. Prefer. 27 (1), 26–34. Paulsen, E., Barrios, S., Baenas, N., Morenoc, D., Heinzen, H., Lema, P., 2018. Effect of temperature on glucosinolate content and shelf life of ready-to-eat broccoli florets packaged in passive modified atmosphere. Postharvest Biol. Technol. 138, 125–133. Peryam, D.R., Girardot, N.F., 1952. Advanced taste-test method. Food Eng. 24 (58–61), 194. Priefer, C., Jörissen, J., Bräutigam, K.R., 2016. Food waste prevention in Europe—a causedriven approach to identify the most relevant leverage points for action. Resour. Conserv. Recycl. 109, 155–165. Ragaert, P., Verbeke, W., Devlieghere, F., Debevere, J., 2004. Consumer perception and choice of minimally processed vegetables and packaged fruits. Food Qual. Prefer. 15, 259–270. Ritter, D.O., Lanzarin, M., Novaes, S.F., Monteiro, M.L.G., Almeida Filho, E.S., Mársico, E.T., Franco, R.M., Conte-Junior, C.A., Freitas, M.Q., 2016. Quality index method (QIM) for gutted ice-stored hybrid tambatinga (Colossoma macropomum Piaractus brachypomum) and study of shelf life. LWT Food Sci. Technol. 67, 55–61. Robertson, G., 2010. Food Packaging and Shelf Life: A Practical Guide. CRC Press, Inc., Boca Raton, Florida. Rockström, J., Steffen, W., Noone, K., Persson, A., Chapin, F.S., Lambin, E.F., … Foley, J.A., 2009. A safe operating space for humanity. Nature 461 (7263), 472–475. Salvador, A., Fiszman, S., Curia, A., Hough, G., 2005. Survival analysis applied to sensory shelf life of yogurts-II: spanish formulations. J. Food Sci. 70 (7), s446–s449. Sant’Ana, L.S., Soares, S., Vaz-Pires, P., 2011. Development of a quality index method (QIM) sensory schee and study of shelf-life of ice-stored blackspot seabream (Pangellus bogaraveo). LWT- Food Sci. Technol. 44, 2253–2259. Siripatrawan, U., Noipha, S., 2012. Active film from chitosan incorporating green tea extract for shelf life extension of pork sausages. Food Hydrocoll. 27, 102–108. Spada, A., Conte, A., Del Nobile, M.A., 2018. The influence of shelf life on food waste: a model-based approach by empirical market evidence. J. Clean. Prod. 172, 3410–3414. Stone, H., Sidel, J.L., 2004. Sensory Evaluation Practices. Elsevier Academic Press, London. Tomac, A., Cova, M., Narvaiz, P., Yeannes, M., 2017. Sensory acceptability of squid rings gamma irradiated for shelf-life extension. Radiat. Phys. Chem. 130, 359–361.
Sensory shelf life estimation357
Vandekinderen, I., Devlieghere, F., Van Camp, J., Denon, Q., Sanchez Alarcon, S., Ragaert, P., De Meulenaer, B., 2009. Impact of decontamination step with peroxyacetic acid on the shelf life, sensory quality and nutrient content of grated carrots packed under equilibrium modified atmosphere and stored at 7°C. Postharvest Biol. Technol. 54, 141–152. Wansink, B., Wright, A.O., 2006. ‘Best if used by …’ how freshness dating influences food acceptance. J. Food Sci. 71 (4), s354–s357. Wilson, N., Rickard, B., Saputo, R., Ho, S.T., 2017. Food waste. The role of date labels, package size, and product category. Food Qual. Prefer. 55, 35–44. Yantis, J., 1992. The Role of Sensory Analysis in Quality Control. ASTM Manual Series: MNL 14, Baltimore, USA.
Further reading Aschemann-Witzel, J., Jensen, J.H., Jensen, M.H., Kulikovskaja, V., 2017. Consumer behaviour towards price-reduced suboptimal foods in the supermarket and the relation to food waste in households. Appetite 116, 246–258. Cruz, A.G., Walter, E.H.M., Silva Cadena, R., Faria, J.A.F., Bolini, H.M.A., Pinheiro, H.P., et al., 2010. Survival analysis methodology to predict the shelf-life of probiotic flavored yogurt. Food Res. Int. 43, 1444–1448. Dethmers, A.E., 1979. Utilizing sensory evaluation to determine product shelf life. Food Technol. 33 (9), 40–43. Djekic, I., Vunduk, J., Tomasevíc, I., Kozarski, M., Petrovic, P., Niksic, M., Pudja, P., Klaus, A., 2017. Application of quality function deployment on shelf-life analysis of Agaricus bisporus Portobello. LWT-Food Sci. Technol. 78, 82–89. Fernández-León, M.F., Fernández-León, A.M., Lozano, M., Ayuso, M.C., Amodio, M.L., Colelli, G., 2013. Retention of quality and functional values of broccoli Parthenon stored in modified atmosphere packaging. Food Control 31 (2), 302–313. Gacula, M.C., Kubala, J.J., 1975. Statistical models for shelf life failures. J. Food Sci. 40, 404–409. Heldman, D.R., Hartel, R.W., 1997. Principles of Food Processing. Aspen Publishers, Inc, Gaithersburg, Maryland. Kilcast, D., 2000. Sensory evaluation methods for shelf-life assessment. In: Kilcast, D., Subramaniam, P. (Eds.), The Stability and Shelf-Life of Food. CRC/Woodhead, Boca Raton, LF, pp. 79–105. Meiselman, H., 2013. The future in sensory/consumer research: …evolving to a better science. Food Qual. Prefer. 27, 208–214. Mexis, S.F., Chouliara, E., Kontominas, M.G., 2012. Shelf life extensión of ground chicken meat using an oxygen absorber and a citrus extract. LWT- Food Sci. Technol. 49, 21–27. Pedro, A.M.K., Ferreira, M.M.C., 2006. Multivariate accelerated shelf life testing: a novel approach for determining the shelf life of foods. J. Chemom. 20, 76–83. NO VA.