*2.5. Statistical Analysis*

Statistical analysis of the results was performed using Statistica 10.0 software (StatSoft, Inc., Tulsa, OK, USA). The experimental data were submitted to one-way analysis of variance (ANOVA) followed by Fisher's least significant differences (LSD) to know the difference between each pair of means with 95% confidence. Pearson correlation analysis for relationships between some selected parameters was also used.

#### **3. Results and Discussion**

#### *3.1. Total Phenolic Content and Antioxidant Activity of OL and OMW Extracts*

As reported in Table 1, the total polyphenol content of OL and OMW samples was 134.7 ± 2.1 and 13.4 ± 0.2 mg GAE g−1, respectively, with the OLE showing a phenolic concentration at about 10-fold higher than that found in OMWE. These results are considerable different from those previously reported in the literature for aqueous OLE, as well as for extracts obtained from OMW produced according to a three-phase centrifugation process [31,32]. However, considering that the amount of bioactive compounds accumulated in olive oil by-products may vary widely depending on many factors, the most important of which are pedoclimatic conditions, olive varieties, degree of maturity, and processing and extraction conditions, making an effective comparison of the data very difficult [33].

**Table 1.** Total polyphenol content and antioxidant activity of olive leaf and oil mill wastewater extracts.


<sup>1</sup> Mean value ± standard deviation. Within columns, values (mean of three repetitions) with the same letter do not differ significantly from each other according to Least Significant Difference (LSD) test (*p* < 0.05). OLE: olive leaf extract; OMWE: olive mill wastewater extract. GAE: gallic acid equivalent.

The antioxidant activity of the investigated extracts evidenced a similar trend, with the OL sample exhibiting a free radical scavenging capacity higher than that of the OMWE (4.26 ± 0.08 vs. 0.32 ± 0.01).

Moreover, another important aspect that emerged from the comparison of the two experimental extracts and that deserves to be emphasized is related to the fact that OL rather than OMW seemed to be the best option, in terms of environmental resources, to obtain polyphenol-rich extracts. In fact, the aqueous extraction of phenolic compounds from OL, besides being the easiest, efficient, and solvent-free extraction procedure, allows for a higher yield with a low environmental impact.

#### *3.2. Moisture Content and Water Activity of GF Breadisticks*

As reported in Table 2, when comparing the control breadsticks to the fortified samples significant (*p* < 0.05) differences in terms of both moisture content and aw were observed. All the enriched breadsticks showed moisture values (ranging from 9.24 to 11.50 g 100 g<sup>−</sup>1) significantly higher than the control (7.66 ± 0.07 g 100 g−1), the extent of the change being more pronounced at increasing level of supplementation for both the investigated extracts. In particular, the highest moisture values were observed in the sample Leaf100 (11.50 ± 0.12), closely followed by the sample Leaf50 (10.87 ± 0.14). Such an increase was probably due to the presence of fiber in the OLE, enabling the absorption of more water than the basic GF ingredients, thus suggesting a potential effect on the water absorption capacity of the resulting breadsticks.

**Table 2.** Moisture content and aw of gluten-free (GF) breadsticks.


<sup>1</sup> Mean values ± standard deviation. Within rows, values (mean of five replicates) with the same letter do not differ significantly from each other according to LSD test (*p* < 0.05).

> A similar trend was observed for the aw, which significantly (*p* < 0.05) increased from a value of 0.41 observed in the control sample to a maximum value of 0.67 exhibited

by the breadsticks prepared with the high level of addition of the OLE (Table 2). As confirmation of this, values of correlation coefficients (r) revealed that higher moisture content corresponded to larger amount of available water (r = 0.985; *p* < 0.001). As expected, the data obtained in this study are considerably higher than those observed in wheat-based breadsticks by other authors, who reported moisture content ranging between 2.09 and 6.15 [23,34,35] and aw values of about 0.14–0.38 [23,36]. Even though such differences should be not surprising given the higher amount of water usually required to obtain machinable GF dough with a proper consistency, it is worth noting that an increase in both moisture and aw, could lead to undesirable stability issues of the GF breadsticks during storage. In fact, in starch-based system, the presence of higher amounts of available or weakly associated water, which is not able to bind to starch as strongly as it does with protein, could lead to a faster moisture migration both within and out of the products, thus shortening their shelf life [37].

#### *3.3. Color and Textural Properties of GF Breadisticks*

Visual and textural characteristics play an important role in determining the final quality of baked products and, in turn, in influencing consumer choice. This is particularly true in GF baked products, which, unlike their gluten-containing counterparts, often exhibit an overly white coloration and a too hard, dry, and grainy texture that consumers find unappealing [27].

In terms of color features, all the enriched GF breadsticks exhibited the same lightness *L\** and yellowness (*b\** positive), but lower red (*a\** positive) values when compared to the control breadsticks, with no significant differences between samples prepared with the same supplementation level (Table 3). The only exception was observed in the sample WasteW100, which significantly differed from the control also in terms of *b\** values. In particular, all the fortified breadsticks showed a significant tendency (*p* < 0.05) to change from a red to a more greenish coloration, especially at the highest level of addition, probably as a consequence of the typical green and green to yellow color of the added OLE and OMWE, respectively. However, as confirmed by the total color difference, which values were <1 for all the analyzed samples (Table 3), these observed differences were not obvious for the human eye, suggesting only a slight influence of both extracts on the color of the resulting breadsticks.


**Table 3.** Textural and color properties of GF breadsticks.

<sup>1</sup> Mean values ± standard deviation. Within rows, values (means of 10 repetitions for color measurements and 20 repetitions for textural properties) with the same letter do not differ significantly from each other according to LSD test (*p* < 0.05).

> From a textural point of view, baked snacks like breadsticks are characterized by a rigid, stiff structure with a little tendency to deform before fracture when subjected to small forces [22]. In the present study, the different mechanical behavior of the experimental breadsticks was measured on the day of baking to assess their quality in terms of hardness and brittleness.

> As reported in Table 3, both OLE and OMWE significantly (*p* < 0.05) lowered the maximum force needed to break the experimental breadsticks, irrespective of whether they

have been added at low or high level. In particular, while the control sample showed the highest values of force at break (or the maximum resistance when broken) (51.57 ± 3.83 N), the most pronounced decrease in hardness values was observed in the samples enriched with the OLE (at about 45 N for both samples). Since in low moisture food systems, water mainly acts as plasticizer [38], a possible explanation for this softening effect could be related to the higher moisture content and aw observed in the samples enriched with OLE, closely followed by those enriched with the OMWE.

Brittleness, which is a textural parameter describing the distance traveled by the blade through the sample before its breaking and, thus, how far a sample can be deformed before fracture, did not show significant differences (*p* < 0.05) among the experimental samples. However, a slight increase of this parameter was observed in the sample Leaf50, suggesting a more leathery or rubbery behavior. The authors of [38] when measuring the mechanical properties equilibrated at different aw of a particular baked snack, called dried bread, demonstrated that when values of aw are higher than 0.56 the stress is released on rupture in an increasingly gradual manner, making prominent the ductile behavior and, thus, deformation over brittleness. However, it is noteworthy that the effect of hydration on the textural properties of baked cereal-based snacks is quite complex and varies depending on the basic formulation of the product, so that the critical values of both aw and water content, corresponding to changes from a crispy to a more deformable behavior, may be different [39].

#### *3.4. Polyphenol Fractions and Antioxidant Activity of GF Breadisticks*

In the present study, the polyphenol fractions and the antioxidant activity of GF breadsticks enriched with natural phenolic-rich extracts from both OL and OMW were compared with a conventional unfortified GF breadstick sample. Results are summarized in Table 4 and Figure 1.

Among the experimental breadsticks, the control formulation exhibited amounts of total polyphenols significantly lower (162.87 ± 1.15 mg of GAE 100−<sup>1</sup> d.m.) than those observed when the OLE and OMWE were individually added to the basic formulation (from 168.40 ± 1.86 to 189.28 ± 3.55 mg of GAE 100−<sup>1</sup> d.m.) (Table 4). Only the sample WasteW50, which was prepared by adding the low percentages of OMWE, showed a total polyphenols content similar to that observed in the control, indicating that the most efficient increase was achieved by supplementing the basic recipe with the OLE at both supplementation levels. In particular, the increment in the total phenolic content of the breadsticks enriched with low and high levels of OLE was four-fold and two-fold higher than that observed by adding low and high levels of OMWE, respectively (14–16% vs. 3–8%). These results were somewhat expected considering the higher polyphenol content observed in the OLE, which was at about 10-fold higher compared to that of the OMWE (Table 1).

**Table 4.** Polyphenol fractions and antioxidant activity of GF control and fortified breadsticks.


<sup>1</sup> Mean values ± standard deviation. Within rows, values (mean of three repetitions) with the same letter do not differ significantly from each other according to LSD test (*p* < 0.05). <sup>2</sup> The total polyphenol content was calculated as the sum of the soluble and insoluble fractions. <sup>3</sup> Corresponding to 36 mg of breadsticks, which consumed these percentages when 0.17 μmol of 2,2-diphenyl-1-picrylhydrazyl (DPPH) are available for reaction. GAE: gallic acid equivalent; IP/PS: insoluble polyphenols/soluble polyphenols ratio.

**Figure 1.** Time evolution of the DPPH curves in methanol of organic extracts from GF control and fortified breadsticks.

However, as it is well known, polyphenols can exist in the plant kingdom in both free and bound form. Therefore, to better evaluate the composition of the phenolic compounds in both control and fortified breadsticks, soluble and insoluble polyphenol fractions were also determined. As reported in Table 4, while the incorporation of both OLE and OMWE was significant (*p* < 0.05) in enhancing the soluble phenolic content of the resulting breadsticks, neither of the two affected the insoluble polyphenol fraction, which did not show significant differences among the investigated samples. In particular, the most significant changes were observed in the samples containing the OLE–with an increment with respect to the control ranging from 76% (Leaf50) to 126% (Leaf100)–followed by those prepared with the OMWE (+32% and +56% for WasteW50 and WasteW100, respectively). Thus, the incorporation of the investigated phenolic-rich extracts led to a significant decrease (*p* < 0.05) in the insoluble/soluble polyphenols average ratio of the resulting breadsticks, an effect more prominent at increasing levels of addition, especially in the case of OLE supplementation (Table 4). To better understand such an improvement, it must be born in mind that fruits and vegetables, compared with cereal grains like rice and corn–in which at about 70% of the total polyphenols exists in the bound forms–have most of their phenolic compounds in the free or soluble conjugate forms [40,41]. Therefore, the addition of extracts from plant wastes can favor the accumulation of the phenolic fraction more rapidly absorbed in the grastrointestinal tract, thus leading to an effective enrichment of the final products [42].

Since there is no direct relationship between the amount of polyphenols in foods and their bioavailability, to evaluate how many of the ingested phenolic compounds could be effectively absorbed and utilized by the human body, thus exerting their biological effects [43], the bio-accessible polyphenol fraction was also determined. As reported in Table 4, although all the fortified breadsticks showed amounts of bioaccessible polyphenols significantly higher than those observed in the control, the most significant increment in polyphenol bioavailability was achieved in the sample prepared with the highest level of OLE (+23% with respect to the control), followed by the samples Leaf50 and WasteW100, wich behaved in a similar way (+14.5% and +15.1%, respectively). This enhancement effect was in line with the same effect previously described for the soluble polyphenol fraction, as also confirmed by the highly significant (*p* < 0.001) linear correlations observed between the soluble and bioaccessible fractions (r = 0.919). In fact, the contribution of the insoluble polyphenols to the bioavailability of the final product is usually lower than that of the

soluble phenolic fraction, since they have to be released from the cell structure before being absorbed [41]. Interestingly, the authors of [44] demonstrated that the absorption and, consequently, the bioavailability of hydroxytyrosol in the gastrointestinal tract could be maximize after a supplementation of the diet with its naturally precursor oleuropein–which is the most abundant bio-phenol in the OLE–rather than with its free form (mainly present in the OMWE) or aglycone forms. However, to the best of our knowledge, data on soluble, insoluble and bioaccessible polyphenols in GF baked products are limited to only one previous study from the same authors, who found similar results in GF breads fortified with a natural apicultural product like bee pollen [26]. For this reason, a comprehensive comparison of the obtained data with the literature is difficult.

Very often, the main source of antioxidants in food products are represented by phenolic compounds, which are able to exert several biological functions, including antioxidant and free radical scavenging activity [4]. In the present study, the antioxidant activity of both control and enriched breadsticks was evaluated by using the DPPH radical scavenging assay, which is a method based on electron donation of antioxidants to neutralize the free radicals. As reported in Figure 1, the time evolution of the DPPH concentration curves in methanol of organic extracts from both control and GF breadsticks evidenced that all the fortified samples exhibited an antioxidant activity higher than the control. In particular, the OLE seemed to be more effective in enhancing the scavenging activity against the stable radical DPPH of the resulting breadsticks, especially at the highest supplementation level (at about 44%). These findings are similar to those reported by other authors in wheat-based baked snacks enriched with OLE [25]. As in the case of OLE, the OMWE, irrespective of whether it had been added at low or high level, also increased the antioxidant activity of the enriched samples, but the extent of this increase was significantly lower than that observed for the OLE, with the sample WasteW100 showing values similar to those observed in the sample Leaf50 (at about 38% and 39%, respectively) (Table 4). These results were in line with those previously observed for phenolic compounds. As a confirmation of this, significant positive correlations (*p* < 0.001) have been found between antiradical activity and soluble (r = 0.958), bioaccessible (r = 0.899), and total (r = 0.868) polyphenols.

#### *3.5. Oxidation Stability (Oxitest) and Estimated Shelf-Life of GF Breadisticks*

Lipid oxidation is one of the major causes of quality deterioration of dehydrated or low moisture bakery products–such as breadsticks and other bread substitutes–which usually require the addition of non-negligible amounts of fatty substances (from 5–15%, or even more, depending on their nature) to obtain desirable texture, appearance and flavor attributes [45]. Lipids, in fact, are susceptible to complex chemical changes that proceed through free-radical propagated chain reactions, which are triggered by unsaturated fatty acids reacting with oxygen. Other oxidation initiators, such as light or heat, certain enzymes, and metal ions are also involved in the process, by enhancing the lipid oxidation during storage [46]. The formation of free radicals and primary oxidation products, such as hydroperoxides, and their decomposition into secondary oxidation products, such as aldheydes, ketones, and hydrocarbons, are directly responsible for the formation of undesirable flavors in rancid foods. Therefore, the use of natural antioxidants–such as OLE and OMWE–well known for their ability to protect against free radicals, may be an effective way to promote the oxidation stability of functional GF breadsticks, thus exending their shelf-life [23]. In the present study, the effect of the investigated extracts on the oxidation stability of both GF control and fortified breadsticks was evaluated by using accelerated oxidation tests performed in the OXITEST reactor at four different working temperatures (60, 70, 80, and 90 ◦C) and at a costant oxygen overpressure (6 bar) to allow the estimation, in a short period of time, of the potential shelf-life of the experimental samples. Monitoring the drop in the oxygen pressure inside the oxidation chambers–which correspond to a certain level of detectable rancidity or a rapid change in the oxidation rate–evidenced that, while the control samples exhibited the lowest lipid stability to oxidation (at about 3 and 52 h at 90 and 60 ◦C, respectively), the addition of both extracts significantly increased the

IP of the fortified breadsticks at all the working temperatures, with the samples WasteW100 being more stable (at about 5 and 99 h at 90 and 60 ◦C, respectively) than the samples WasteW50 (3 and 57 h at 90 and 60 ◦C), Leaf100 (3 and 60 h at 90 and 60 ◦C), and Leaf50 (2 and 58 h at 90 and 60 ◦C), respectively. Starting from this point and considering that with increasing working temperature the IP of the oxidative reaction decreased, thus suggesting a linear dependence of oxidation stability and temperature, the potential shelf-life of the experimental samples at the storage temperature of 25 ◦C was estimated by using linear regression equations on a semi-log scale. The estimated shelf-life of the GF breadsticks based on lipid oxidation data is shown in Table 5.


**Table 5.** Estimated shelf-life of GF breadsticks based on lipid oxidation data (day at 25 ◦C).

<sup>1</sup> Mean values ± standard deviation. Within rows, values (mean of two repetitions) with the same letter do not differ significantly from each other according to LSD test (*p* < 0.05).

As can be seen, the incorporation of increasing percentages of both OLE and OMWE was followed by a concurrent increase in the estimated shelf-life of the resulting breadsticks (Table 5), indicating an effective role of antioxidants in preserving the final products. In particular, the greatest values were observed in the sample prepared with the high level of addition of OMWE which nearly doubled the shelf-life of the control, closely followed by the sample WasteW50 (+48% increment with respect to the control). A significant shelf-life extension was also registered in those breadsticks prepared with the addition of OLE, but the extent of this increase was significantly lower than that observed for the OMWE extracts at both supplementation levels (+23 and 32% for Leaf50 and Leaf100, respectively). These findings, however, were somewhat unexpected considering the opposite results observed in terms of antioxidant activity among the fortified samples. In fact, in spite of a lower (or similar) antioxidant activity (Table 4), the OMWE-enriched breadsticks seemed to be kept fresh for longer, at least in terms of oxidation stability, than those prepared with the OLE, irrespective of the level of substitution used. A possible explanation of this contrasting behavior may be related to the significant differences recorded in the aw values among OLE and OMWE breadsticks (Table 2). In fact, it has been demonstrated that water can play both pro-oxidant and antioxidant roles in lipid oxidation, depending on whether its content in the food is within or above the monolayer moisture content [47,48]. At low levels of aw–near to the monolayer range–water can exhibit an antioxidant role by forming a barrier that protects the sensitive sites from reactions with oxigen, but also by lowering metal catalytic activity, increasing hydration of hydroperoxides (and consequently decreasing the rate of free radicals formation), as well as by promoting recombination of free radicals. In contrast, at higher aw values, as is the case with the experimental OLE-enriched breadsticks, water can show a pro-oxidant role by acting as a plasticizing agent, thus promoting mobility and solubilization of catalysts, as well as by inducing matrix swelling, thus exposing new reactive sites [47,48]. Therefore, the lower oxidative stability observed in the breadsticks fortified with the OLE in comparison to that observed in those enriched with the OMWE might suggest that, in such a complex matrix. the effect of aw is greater than the conservative role exerted by the added antioxidants.

However, considering that in low moisture foods the causes of lipid oxidation are still not completely understood and that both monolayer and glass transition theories have led to contrasting results when used to predict lipid oxidation rates as a function of aw [48], further studies are needed to give consistency to the obtained results.

#### *3.6. Sensory Evaluation of GF Breadisticks*

In the first two sessions of the sensory evaluation, participants were asked to rank the overall preferences for the freshly prepared breadsticks, by comparing the control to the samples enriched with both OLE and OMWE separately. As reported in Table 6, the obtained results evidenced that no sample was significantly preferred to another in both comparisons.

**Table 6.** Results of the ranking preference tests on the freshly prepared GF breadsticks.


<sup>1</sup> (a) comparisons among the control sample and the OLE-enriched breadsticks; (b) comparisons among the control sample and the OMWE-enriched breadsticks.

More specifically, when comparing the control sample to the OLE-enriched breadsticks, the lowest ranking score was assigned to the sample Leaf100, followed by the Leaf50 and the control breadsticks (Table 6(a)), suggesting a clear tendency for consumers to recognize the fortified samples as the most palatable choice. However, as evidenced by the obtained Ftest value (4.03), which was lower than the Fcritical value reported in in the χ<sup>2</sup> table (5.99), the recorded difference in the preference degree was not significant. A similar trend was also observed when comparing the control sample to the two breadsticks fortified with OMW*<sup>E</sup>* (Table 6(b)). In this case, although the differences in the preference degree assigned to the three samples were less pronounced (Ftest value: 1.13 and Fcritical value: 5.99), a consumer's tendency to prefer the breadsticks prepared with the high level of OMWE could also be observed.

Based on these results, Leaf100 and WasteW100 samples were then subjected to a paired comparison test to assess if a significant difference exists between them in terms of preference (Figure 2).

**Figure 2.** Differences in the number of responses accorded by consumer to the Leaf100 and WasteW100 samples in the paired comparison test.

According to the statistical table for paired difference test (two tailed), 48 is the minimum number of responses needed to conclude that a preference exists between two samples at the selected significance level (5%) and with a total number of assessors of 76. This value was not reached by either of the two experimental breadsticks, even if the sample Leaf100 came very close to this minimum number (Figure 2). However, it should be noted that, although the number of responses did not differ significantly between the two enriched breadsticks, the sample Leaf100 was preferred by the 61% of the assessors, indicating a marked, but not significant, tendency for consumers to consider it as the most appetizing choice.

#### **4. Conclusions**

Data obtained in the present study evidenced, for the first time, that the incorporation of phenolic-rich extracts from olive oil by-products in GF formulations can be considered a successful strategy in the preparation of technologically viable functional breadsticks with extended shelf-life. Although all fortified samples, especially those enriched with the OLE, were softer, they also exhibited a similar crumbly texture and minimal color changes compared to the control, indicating an only small impairment in their technological feasibility. This was also confirmed by the sensory evaluation, which showed a marked (but not significant) tendency of consumers to consider the enriched breadsticks, especially those prepared with the high percentage of OLE, as the most preferred. The incorporation of both extracts also resulted in improved nutritional and functional properties of the final breadsticks, as evidenced by the changes observed in the insoluble/soluble polyphenol ratio in favor of the soluble fraction, by the enhanced bioavailability of polyphenols, as well as by the higher antioxidant activity, especially in those samples prepared with the higher percentage of OLE. Furthermore, all the fortified breadsticks exhibited higher stability against the lipid oxidation and, in turn, an extended estimated shelf-life, even if, in this case, better behavior was observed in the samples prepared with both levels of OMWE.

In conclusion, the best results were achieved in those samples fortified with the high percentage of OLE in which, in spite of a slightly shorter shelf-life, a more proper balance among the technological, sensory and functional properties was observed. Further experiments are needed to give consistency to these preliminary findings, especially to better clarify the influence of both antioxidant activity and aw in the estimation of the shelf-life of GF low moisture baked snacks.

**Author Contributions:** Conceptualization, P.C., A.D.C., C.F., P.P.U. and A.P.; methodology, P.C., S.P., A.D.C., P.P.U.; formal analysis, P.C. and S.P.; investigation, P.C., S.P., A.D.B., G.D. and R.R.; data curation, P.C., S.P., A.D.C., P.P.U.; writing—original draft preparation, P.C., S.P., A.D.C., C.F. and A.P.; writing—review and editing, P.C., S.P., A.D.C., C.F., P.P.U., A.D.B., G.D., F.C., R.R. and A.P.; supervision, F.C. and A.P.; funding acquisition, P.C., F.C. and A.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the AGER 2 Project, grant n. 2016-0105. This work was also funded by the University of Sassari (Fondo di Ateneo per la Ricerca 2020): FAR2020CONTEP.

**Institutional Review Board Statement:** Ethical review and approval were waived for this study since the participation was voluntary. All data were anonymous.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

