HMF is a yellow solid compound with great solubility in water, and its structure is a six-carbon heterocyclic structure with two functional groups: aldehyde and alcohol (hydroxymethyl). The amount of HMF present in honey is indicative of its freshness, and reflects its and storage length and conditions (15). The honey was heated in a water bath to 40°C to remove impurities and filtered. Five grams of sample was homogenized in 20 mL of distilled water, and HMF was measured at 25°C with a reflectometer (HANNA, RQflex 10, Italy).

The amount of honey used for the extraction of bioactive compounds varied depending on the analysis to be performed. Two grams were used for the determination of antioxidant activity, and 1 g was used for the quantification of total phenolic compounds. Ten milliliters of acidified water (adjusted to pH 2 with 2 N HCl) was added to the sample (16), homogenized for 1 min using a vortex device and subjected to an ultrasonic bath (Auto science, Serial Ultrasonic Cleaner, 110 V/60 Hz, China) for 30 min at 25°C. Subsequently, the sample was left to macerate for 24 h in the dark at a temperature of 25°C. Afterward, the extract was centrifuged at 18.0 x g for 15 min at a temperature of 27°C, the supernatant was filtered, and the resulting product was preserved in amber tubes.

Antioxidant activity was determined using the 2,2-diphenyl-1-picrylhydrazyl (DPPH) synthetic free radical method (16). Absorbance was measured at a wavelength of 517 nm in a UV spectrophotometer (Thermo Fisher Scientific, Genesys 10-S, United States). A calibration curve was performed using Trolox at concentrations of 0–100 μg/mL in methanol. From the absorbance readings obtained, the antioxidant activity was calculated, and the results were expressed as mg Trolox/kg honey.

The concentration of total phenolic compounds in honey was determined by the Folin–Ciocalteu method (16). Absorbance was measured in a UV spectrophotometer (Thermo Fisher Scientific, Genesys 10-S, United States) at a wavelength of 760 nm. A calibration curve was performed using gallic acid as a standard at concentrations of 0–100 mg/mL, and the results were expressed as mg gallic acid/kg honey.

The density of the marshmallows was determined in a cylindrical container with a known volume and weight. The foam was uniformly poured into the container and weighed to calculate its mass by weight difference and subsequently its density in g/mL.

For the determination of phenolic compounds and antioxidant capacity, extracts of the different marshmallow treatments were prepared. For this, freeze-dried samples were used at a vacuum pressure of 9.5 × 10 mm Hg for 24 h (10). Each extract was prepared by mixing 2 g of freeze-dried sample with acidified water (pH 2) at a 1:2 w/v ratio. This solution was homogenized in a vortex and subjected to ultrasonication for 30 min. It was then macerated in the dark for 24 h at a temperature of 25°C. After this time, the solution was centrifuged at 18.0 x g at 27°C for 20 min and finally filtered to remove the supernatant. The extracts were kept in amber tubes prior to use in the corresponding analyses.

To determine the antioxidant capacity in marshmallows, the same methodology used for honey was implemented (16), with modification of the resting conditions; in this case, they were in a water bath at 37°C for 30 min.

The determination of total phenolic compounds was carried out using the Folin Ciocalteu method, according to the methodology of Cedeño-Pinos et al. (21), which was slightly modified. First, 0.25 mL of marshmallow extract was mixed with 7.75 mL of distilled water and 0.8 mL of 10% Folin–Ciocalteu reagent. This solution was left to stand for 8 min, and then 7.5% (w/v) Na₂CO₃ was added and homogenized. This mixture was left to stand in the dark at 25°C for 2 h. Finally, absorbance was determined at 760 nm in a spectrophotometer (Thermo Fisher Scientific, Genesys 10-S, USA).

To determine the viability of the microorganism, 1 g of marshmallow was added to 9 mL of 5% (w/v) sodium citrate, and then the mixture was stirred for 15 min to disintegrate the capsules and release the cells. The first dilution was performed with 1 mL of the solution in 9 mL of 0.1% peptone water. It was sown in Petri dishes by the pour plate technique from the 10⁻⁵ to 10⁻⁸ dilution in triplicate using MRS agar as culture medium. The dishes were incubated at 32°C for 48 h, and then cell counts were performed. The viability of the microorganisms was expressed as CFU/g marshmallow.

Table 2 shows the results of the physicochemical analyses of the different marshmallow treatments. In general, the marshmallows had ash contents of 0.43–0.57% and pH values of 3.50–5.05 and 60.00–65.50 °Brix. In particular, the honey concentration had no effect on the ash content of the marshmallows, so it is likely that the ash came from the probiotic capsules added to the marshmallows, which were made with alginate, gelatin and mucilage, in addition to having been subjected to a CaCl₂ solution. The pH was lower at higher honey concentrations (p < 0.0001), with pH values ranging from 4.67 to 5.05 for marshmallows without honey, 3.75 to 3.8 for marshmallows with 50% honey, and 3.5 to 3.7 for marshmallows with 75% honey. These pH variations may be related to the content of organic acids present in the honey, so that the higher the honey concentration is, the higher the organic acid content. On the other hand, the bloom degrees of gelatin had an inversely proportional effect on the °Brix of the marshmallows, while the addition of honey increased the °Brix proportionally.

The marshmallows had a density of 0.35–0.44 g/mL. The density decreased as the percentage of honey increased and that of sugar decreased and was higher in marshmallows with 265 °bloom gelatin. Sugar likely limits the amount of air entering the foam during the whipping process, so honey may be useful in providing desirable characteristics in marshmallows. In general, this property is important for foamy foods because it can define softness, lightness or porosity (39). Other studies have shown a similar density in marshmallows: 0.474 g/cm³ (40) and 0.41 g/mL (41).

Figure 2 shows the moisture content of the marshmallow treatments on Days 1, 7 and 14 of storage. In general, moisture loss during storage was minimal and decreased with storage time in all samples, indicating the stability of the confection. Both physical and chemical aspects of freshness and stability during long-term storage can be affected by the moisture content. Even minor deviations from defined standards can have adverse effects on the physical properties of food (42). After 14 days of storage in a polypropylene wrapper, the moisture content of the marshmallows without honey was 25 to 35%, and those with the highest amount of honey (75%) had 32 to 36% moisture. Other authors have reported different moisture values in confectionery products with functional properties: 19 to 21.5% in marshmallows with anthocyanins (11); 16.7 to 22.9% in marshmallows with Stevia rebaudiana, oligofructose and isomaltose as sugar substitutes (10); 22.06 to 22.55% in jelly candies with rosemary extract (21); and 20.05 to 25.3% in jelly candies with honey and fruit juices (43). The marshmallows with probiotics had a higher moisture content than the aforementioned confections, probably due to the content of the probiotic beads.

Figure 3 shows the AC of the marshmallow treatments during storage. Antioxidant activity was not affected by the gelatin bloom degree; however, honey concentration had a significant effect on the treatments (p < 0.0001), as well as the interaction between the two factors (p < 0.0001). The marshmallows presented AC values from 61.87 to 100.81 mg Trolox/kg marshmallow. In general, it can be observed that marshmallows with honey presented more AC than those without honey, being less noticeable on Day 14, which may be due to a slight survival of probiotics, since strains of lactobacilli and some bididobacteria possess AC and are able to decrease the risk of free radical accumulation (44). Furthermore, Sharma et al. (45) found that L. rhamnosus GG had an inhibition percentage of 59%. Therefore, AC in marshmallows is likely influenced by probiotic content (21).

At the end of the experiment (Day 14), treatment M7 (0% honey-315 bloom) presented the lowest AC with 58.58 mg Trolox/kg, while M3 (75% honey-265 bloom) had the highest AC with 82.02 mg Trolox/kg marshmallow. Periche et al. (10) made marshmallows with an aqueous extract of Stevia that had an AC of 117 mg Trolox/100 g aqueous extract, while Santoso et al. (12) prepared marshmallows with Kinang extract, and the treatment with 80% extract presented 2.78 mg/mL antioxidant activity.

The honey concentration in the marshmallows presented a significant effect (p < 0.0001) on the total phenol content. Although bloom degrees did not present a significant effect, the interaction of the two factors had a significant influence (p < 0.0001). Figure 4 shows that the honey treatments had a higher phenol content, which is explained by the TPC of the honey added to the marshmallows (386 GA/kg). In general, the concentration of phenolic compounds increased with storage time.

On Day 1, treatment M7 (0% honey-315°bloom) showed the lowest phenol content, with 54.45 mg GA/kg marshmallow, while M6 (75% honey-300°bloom) had a content of 100.55 mg GA/kg marshmallow. At the end of the experiment (Day 14), M7 (0% honey-315°bloom) again showed the lowest phenol content, with 95.33 mg GA/kg marshmallow, and the treatment with the highest concentration was M5 (50% honey-300°bloom) with 130.4 mg GA/kg marshmallow. Cedeño-Pinos et al. (21) analyzed jelly candies enriched with a rosemary extract with 73.9 mg polyphenols/g; however, the candies had polyphenol contents from 0.19 to 0.41 mg/g; this indicates that the phenol content can be easily lost over storage time, so it is important that the phenol source has a significant amount for the best results.

Figure 5 presents the results of the viability analysis of the probiotic L. rhamnosus incorporated in the marshmallows. On the first day of storage, the viability of the lactobacilli remained above the minimum recommended limit (10⁶ CFU/g) (46) in all formulations. However, by Day 7, viability was reduced from 4.8 to 6.0% in marshmallows without honey, which presented 1.23 × 10⁷, 1.58 × 10⁷ and 1.76 × 10⁷ CFU/g marshmallow for treatments with 265, 300 and 315°bloom, respectively, while marshmallows with 50 and 75% honey had a drastic reduction in CFU/g marshmallow on this same day. Finally, by Day 14, none of the treatments reached minimum viability.

Lactobacillus survival can be influenced by several factors, such as storage temperature, water activity (a_w), titratable acidity, presence of sugars, and processing conditions (heat treatment, cooling rate, packaging material, etc.) (47). On the other hand, pH limits the growth and stability of probiotic bacteria (47). The acidification of the medium due to lactic-acid bacteria growth poses a challenge for industrial production. The accumulation of lactic acid may affect cell physiology, leading to growth interruption or reduction. Therefore, maintaining a pH close to neutrality can promote a higher growth rate (48); however, the pH of the marshmallows varied from 3.50 to 5.05, being lower in the treatments with honey, since it showed a pH of 4.34, so that on Day 7, the viability of the marshmallows with honey was lower than that of the treatments without it.

On the other hand, another important factor in the viability of probiotics is acidity. This characteristic is due to the presence of organic acids, among which gluconic acid (~0.5% w/v) is important in honey; it is generated from the oxidation of glucose by the endogenous glucose oxidase enzyme and is a very powerful antibacterial agent. In addition, when honey is diluted, glucose oxidase is activated and acts on endogenous glucose to produce hydrogen peroxide (H₂O₂), which has antibacterial properties (49, 50). The antimicrobial activity of honey derived from low water activity (0.56 to 0.62) and its high viscosity, among other characteristics, can also be considered (51).

High concentrations of sugars in marshmallows could trigger the osmosis process in probiotics, causing water loss from bacterial cells and affecting bacterial growth (52). Additionally, the survival of probiotics can be affected by the characteristics of the packaging material, such as thickness and gas permeability (53).

According to the viability results of the marshmallows, the applied encapsulation was not able to protect the microorganisms from the mentioned factors. However, these probiotics fall within the classification of paraprobiotics, also called inactivated probiotics or ghost probiotics, which are inactivated cells or fractions of microbial cells that provide a benefit to the host at appropriate doses (54). Therefore, they could be used in more vulnerable patients without risk (55).

Shigwedha et al. (56) showed that the administration of inactivated L. rhamnosus V. probiotics is useful in the prevention and treatment of infectious diseases, allergies, fatigue and fibromyalgia, whereas Good et al. (57) demonstrated that oral administration of UV-inactivated L. rhamnosus HN001 attenuates the severity of necrotizing enterocolitis in neonatal mice and preterm piglets. Although the project aimed to preserve the viability of probiotics in marshmallows, more time is required to test other encapsulation methods and materials, so there are opportunities for future research. However, based on the above, it is possible that the marshmallows produced in this study could be studied by researchers in the area of health as a supplement with potential for the prevention of cancer and other gastrointestinal system conditions. On the other hand, the project made it possible to obtain marshmallows with antioxidant activity and phenolic compounds derived from Apis mellifera honey, thus obtaining a product with functional properties.