Introduction: When infants cannot consume breast milk, the most commonly available alternative milk formula is cow milk-based. Due to a rise in the prevalence of cow milk protein allergy (CMPA) among children, this study aimed to assess the biofilm formation and acidogenicity of cow milk-based formulas as well as milk formulas suggested for children with CMPA. Methods: Cow milk-based formulas with 0%, 10%, or 18% sucrose added, partially hydrolyzed formula (pHF), extensively hydrolyzed formula (eHF), amino acid-based formula (AAF), and soy-based formulas with 0%, or 11% sucrose added were evaluated. Streptococcus mutans was used as a representative microorganism associated with caries. The acidogenicity after 24-h incubation was assessed by the pH of the formed biofilm and lactic acid formation. Biofilm formation was quantified using crystal violet staining. Additionally, the biofilm characteristics were determined using confocal laser scanning microscopy (CLSM). Comparisons were made among formulas without added sucrose to observe protein-based differences. Furthermore, formulas with different sucrose percentages were compared to explore the impact of sucrose content. Results: When comparing the formulas without added sucrose, the biofilm formation in the cow milk-based formula and pHF were significantly greater than the soy-based formula, eHF, and AAF. In the presence of S. mutans, all formulas reduced the biofilm pH below the critical enamel pH. The cow milk-based formula and AAF showed a significantly lower biofilm pH than the pHF, soy-based, and eHF groups, while the lactic acid production was markedly higher in the cow milk-based formula, pHF and AAF, compared with the eHF and soy-based formula. Adding sucrose into the cow milk-based and soy-based formulas substantially increased biofilm mass. The biofilm pH of the cow milk-based formulas, with or without sucrose, was significantly lower than that of the soy-based formulas. The CLSM indicated distinct biofilm characteristics among the different protein-based formulas, with sucrose supplementation promoting S. mutans aggregation in cow milk-based formula biofilm and increased density and intact biofilm in the soy-based formula. Conclusion: All assessed milk formulas had caries-inducing factors, including those without supplemental sucrose. Among them, the eHF demonstrated the least caries-inducing factors, attributed to its minimal biofilm formation and the highest biofilm pH.

Human breast milk is the best nutrition source for infants. The World Health Organization recommends exclusive breastfeeding from birth to 6 months old and supplemental breastfeeding is recommended until 2 years old. If an infant cannot consume breast milk, there are commercially available milk formulas as substitutes [1, 2]. The most common formula is a cow milk-based formula; however, the prevalence of cow milk protein allergy (CMPA) in children has increased. CMPA is caused by a specific immune response to one or more milk proteins [3]. The current treatment for CMPA is avoiding cow milk and food containing cow milk while providing alternative meals, including milk substitute choices [1, 4]. Milk formulas commonly used for CMPA are protein hydrolysate and soy-based formulas. Protein hydrolysate formulas are milk formulas in which the proteins are broken down into peptides or smaller components. These can be classified as partially hydrolyzed formula (pHF), extensive hydrolyzed formula (eHF), and amino acid-based formula (AAF) [5]. The pHF has a higher molecular weight (>6,000 Daltons) than the eHF (<3,000 Daltons), whereas the AAF contains protein in the form of free amino acids and is therefore considered as a completely nonallergenic milk formula. Soy-based formula is a plant-based formula that is more palatable and less expensive than eHF and AAF [6].

Milk formula is the major source of nutrition for infants and dietary supplement for children. The sugar composition of milk formulas includes lactose as a natural sugar in milk, sucrose supplements in flavored milk, and industrial sugars, e.g., corn syrup, maltodextrin, and glucose syrup. Sucrose is the most cariogenic and lactose is the least cariogenic [7]. Although the plain milk-based formula is recommended for use as a breast milk substitute, some parents choose flavored milk formulas or high caloric density milk formulas that contain high sucrose supplementation for increasing the children’s appetite. Previous studies reported that the cariogenicity of cow milk with 2–10% sucrose added was significantly higher than plain milk containing only lactose [8‒10]. However, information regarding the cariogenicity of alternative milk formulas is limited and inconclusive [11]. Some alternative milk formulas for CMPA add extrinsic sugars to improve their palatability. It has been reported that the oral health of children with CMPA was worse than children without CMPA, suggesting the possible involvement of milk substitution with sweetened and low pH products, which enhance enamel demineralization and biofilm formation [12]. Similarly, de Carvalho et al. [13] found a correlation between CMPA and dental caries, including primary dentition treatment needs.

Although several methods can be used to examine the diet’s cariogenicity, biofilm formation and its pH are commonly used [6, 14, 15]. Dental biofilm supporting microbial adhesion on teeth and acid end products of these microbes are related to tooth decay. Streptococcus mutans is a well-known species implicated in early childhood caries (ECC) [16]. S. mutans’s ability to form dental biofilm, especially in the presence of sucrose, allows it to effectively bind to tooth surfaces, while its acidic metabolites decrease the biofilm pH. Sustained biofilm pH values below pH 5.5, the critical pH of enamel favor enamel demineralization and dental caries development [17]. Among the organic acids produced by S. mutans, lactic acid is the most crucial acid related to dental caries because this acid is the major end-product of glycolysis by S. mutans under conditions of sugar excess or low environmental pH [18].

Currently, there is no study comparing biofilm formation and acid production between cow milk-based formulas and various milk formulas used by CMPA children. Furthermore, there is no report about the characteristics of biofilm formed in these milk formulas. Therefore, the aim of the present study was to evaluate biofilm formation, biofilm pH, lactic acid formation, and biofilm characteristics of cow milk-based formulas and alternative milk formulas for CMPA children in the presence of S. mutans. Our objective was to compare these properties among milk formulas with different protein compositions and among the formulas with varying sucrose percentages. For each comparison, we carefully selected commercially available milk formulas with the most similar sucrose percentages.

Milk Formulas

This study used 8 types of milk formulas; plain cow milk-based formula, cow milk-based formulas with 10% sucrose, cow milk-based formula with 18% sucrose, partial hydrolyzed formula (pHF), extensive hydrolyzed formula (eHF), AAF, plain soy-based formula, and soy-based formula with 11% sucrose. The details of the milk formulas on the manufacture’s labels are presented in online supplementary Table S1 (for all online suppl. material, see https://doi.org/10.1159/000538882).

The milk formulas were prepared by dissolving the milk powder in sterile distilled water. The weight per volume for each formula followed the manufacturer’s instructions. The microorganisms present in the milk formulas were also verified before using them in the experiments. Each dissolved milk was transferred to a microcentrifuge tube (1 mL/tube) and incubated at 37°C in a 5% CO2 incubator for 24 h. Next, 100 μL of each solution was spread on brain heart infusion (BHI) agar (Difco Laboratories, Detroit, MI, USA) plate. The colonies that grew on the plate were Gram stained (online suppl. Fig. S1).

Buffer Capacity Determination

The buffering capacity of each milk formula was determined by pH titration using a pH meter (Thermo Fisher Scientific, Orion Star A211, China) with the ST230 pH probe (OHAUS, NJ, USA). The amount of hydrochloric acid needed for a pH change of one unit was recorded for the buffering capacity [19]. Three independent experiments were performed in triplicate.

S. mutans Suspension Preparation

A stock culture of S. mutans strain UA159 grown to late exponential phase in BHI broth (Difco) was kept at −80°C in a 50% BHI glycerol solution. BHI broth was used to grow and maintain the strain. The S. mutans suspension was cultured in a 37°C, 5% CO2 incubator overnight. Subsequently, 500 μL of the overnight culture was transferred to 4.5 mL fresh BHI broth and cultured until the mid-exponential phase (optical density at 600 nm ∼0.4; viable cells = 109 CFU/mL). A 100-fold dilution was performed to obtain 107 CFU/mL of S. mutans suspension in the BHI broth.

In the assays described below, two groups, i.e., (i) with S. mutans and (ii) without S. mutans groups were evaluated. For the group with S. mutans, each dissolved milk formula was mixed with the prepared S. mutans suspension (107 CFU/mL) to a final concentration of 105 CFU/ml S. mutans in each solution. For the group without S. mutans, BHI broth without S. mutans was added to each milk solution instead of the S. mutans suspension.

Biofilm Formation Assay

Biofilm formation was evaluated in 96-well flat bottom plates (Thermo Fisher Scientific, Jiangsu, China). For the group with S. mutans, 10 μL S. mutans suspension (107 CFU/mL) was mixed with 990 μL of each test solution. Then, 100 μL S. mutans solution was distributed to each well of the 96-well plate. For the group without S. mutans, 10 μL BHI broth was mixed with 990 μL of each solution and distributed in the same manner into the same plate. BHI broth with and without 10% sucrose were used as negative and positive controls, respectively. The plate was incubated at 37°C, 5% CO2 for 24 h.

After the 24-h incubation, biofilm detection was performed using a standard crystal violet staining protocol. The culture solutions in the 96-well plate were decanted and rinsed three times with distilled water to remove the non-adherent cells. The biofilms formed in the wells were fixed with a 25% formaldehyde solution (100 μL) for 10 min at room temperature. After washing with distilled water, biofilms were stained with 100 μL 0.05% (w/v) crystal violet in water for 3 min, washed with water and the absorbance at 590 nm (A590 nm) was determined, dissolved in 100 μL 7% acetic acid, and was measured using ELx800 Absorbance Reader (BioTek Instruments Inc., Winooski, VT, USA). Three independent experiments were performed in triplicate.

The results of this biofilm assay are shown as the biofilm formed by S. mutans in each solution after subtracting the background, which was calculated by the following formula: A590 nm of (i) S. mutans added – A590 nm of (ii) no S. mutans added. Details regarding the A590 nm readings of both groups are provided in online supplementary Table S2.

Biofilm pH Measurement

The measurement of acid production and the formed biofilm pH was performed in 24-well flat bottom plates (Thermo Fisher Scientific). For the group with S. mutans, each dissolved milk (9.9 mL) was mixed with 0.1 mL BHI broth containing S. mutans (107 CFU/mL), whereas sterilized BHI broth (0.1 mL) was added to each milk solution instead of the S. mutans suspension in the group without S. mutans. The prepared solutions were then transferred into 24-well plates (2 mL/well) and incubated at 37°C, 5% CO2 for 24 h. For controls, BHI broth with and without 10% sucrose were the negative and positive controls, respectively.

The pH of the non-incubated solutions in both groups, (i) with and (ii) without S. mutans, was also determined from the leftover solutions. A pH meter (Thermo Fisher Scientific, Orion Star A211) with the ST230 pH probe (OHAUS) was used to determine the pH. The starting pH of the milk solutions are shown in online supplementary Table S3.

To determine the biofilm pH after 24-h, the incubated 24-well plates were decanted to remove the culture solutions, and the formed biofilm pH was measured by a STSURF pH probe (OHAUS). Three independent experiments in triplicate were conducted for each solution.

Lactic Acid Formation Assay

The preparation of each milk solution with or without S. mutans was performed in a 1.5 mL microcentrifuge tube. For the group with S. mutans, 10 μL S. mutans suspension (107 CFU/mL) was mixed with 990 μL of each solution. For the group without S. mutans, 10 μL of the sterilized BHI broth was mixed with 990 μL of the milk solution. The negative and positive controls were BHI broth with and without 10% sucrose, respectively. The lactic acid concentration measurement was performed after incubation at 37°C, 5% CO2 for 24 h.

After the 24-h incubation, the solutions in the 1.5 mL microcentrifuge tubes were centrifuged twice at 14,000 rpm, 4°C for 10 min to remove the milk precipitates and bacterial cells. The supernatant was transferred to a new tube. The concentration of lactic acid in the supernatant was measured using the colorimetric L-Lactate Assay Kit (Sigma-Aldrich, Saint Louis, USA) per the manufacturer’s instructions. Briefly, 20 μL of the supernatant sample was transferred into a well of 96-well flat bottom plates (Thermo Fisher Scientific). The reaction mixture from the Kit (80 μL) containing Assay Buffer, Enzyme A, Enzyme B, NAD, and MTT was added and mixed thoroughly. The plate was immediately measured for its initial absorbance at 570 nm (A570 nm). After incubation at room temperature for 20 min, the measurement was performed again for the final A570 nm. The initial A570 nm was subtracted from the final A570 nm to obtain the ∆A570 nm, which was used to determine the lactic acid concentration according to the ∆A570 nm of a standard curve made from 2 mM L-Lactate as recommended by the manufacturer’s instructions. If the supernatant samples showed a ∆A570 nm that was higher than the ∆A570 nm of the 2 mM L-Lactate standard, they were diluted in ultrapure water before measuring the lactic acid concentration. Three independent experiments were performed for all milk solutions.

Confocal Laser Scanning Microscopy

The S. mutans biofilm structure that formed in each milk formula and BHI broth with 10% sucrose (control) was evaluated using a confocal laser scanning microscope (CLSM). Biofilm was formed on a round cover glass (Thermo Fisher Scientific) immersed in 1 mL milk solution containing S. mutans (105 CFU/mL) in a 24-well plate well (Thermo Fisher Scientific). The plate was incubated at 37°C, 5% CO2 for 24 h.

The wells were rinsed three times with sterile normal saline solution to remove the non-adherent cells. The round cover glasses were moved to new wells for the staining procedure. The bacterial cells were labeled by adding normal saline solution containing 0.334 μM SYTO™ 9 (500 μL/well; Invitrogen™, Thermo Fisher Scientific) and incubated at room temperature in the dark for 15 min. Imaging was performed using a Leica stellaris 5 confocal microscope (Leica, Germany) with a 488/522 nm laser (SYTO9) and ×63 objective. Biofilm images were captured from three random positions in two independent experiments to validate the reproducibility of the biofilm characteristics.

Statistical Analysis

The data were analyzed using SPSS Version 18 (SPSS Inc., Chicago, IL, USA). One-way ANOVA, Turkey and Dunnett’s T3 were used for statistical analysis. p values <0.05 were considered to be statistically significant in all tests.

The results of this study were divided into two main parts, the comparison among milk formulas with different protein compositions and the comparison among milk formulas with different sucrose amounts. Each part comprised experiments regarding biofilm formation, biofilm acidity, lactic acid formation, and structural analysis of biofilm by CLSM.

Part 1: The Comparison among Milk Formulas with Different Protein Compositions

This part comprised five types of milk formulas with different protein compositions: cow milk-based formula without added sucrose, pHF, eHF, AAF, and soy-based formula without added sucrose.

Biofilm Formation

Visual observation of the formed biofilm before crystal violet staining (online suppl. Fig. S2) presented that S. mutans biofilm formation in eHF and soy-based formula were relatively similar to the negative control (BHI broth). However, crystal violet staining resulted in greater dark staining in soy-based formula than eHF as presented by A590 nm in Table 1. The pHF group showed thick biofilm with a chalky appearance, while the biofilm formed in the AAF group appeared as chalky clusters in the center of the wells that disappeared after formaldehyde fixation and crystal violet staining. By crystal violet staining and measuring the A590 nm of the dye, the biofilm formation of cow milk-based formula and pHF was approximately double that of soy-based formula (Table 1). In contrast, the eHF and AAF showed the least biofilm formation, significantly lower than the other formulas.

Biofilm pH

After 24 h, the pH of the biofilm formed by S. mutans in each type of milk formula was below 5.5, the critical pH of enamel; however, all milk types possessed buffering capacity (Table 1). Both AAF and cow milk-based formula exhibited significantly lower biofilm pH levels compared with the pHF, soy-based, and eHF groups. Notably, the correlation between buffering capacity and biofilm pH was not direct. Despite similarities in pH levels among the pHF, soy-based, and eHF groups, the soy-based formula exhibited nearly half the buffering capacity compared with pHF and eHF.

Lactic Acid Formation

Lactic acid production in the presence of S. mutans was significantly higher for cow milk-based formula, pHF, and AAF compared with eHF and soy-based formula (Table 1). Among the milk formulas, the soy-based formula generated the lowest concentration of lactic acid, two-fold lower than that of cow milk-based formula, pHF, and AAF.

Biofilm Characteristics

A clear distinction was observed in the biofilm characteristics based on their different protein compositions, as illustrated in Figure 1. The biofilm formed in the cow milk-based formula was spread with a porous structure (Fig. 1a). In contrast, the biofilm of the pHF group manifested as a nonporous sheet covering the entire surface area (Fig. 1b). For the eHF group, although bacterial cells were present, no detectable biofilm formation was observed (Fig. 1c). The AAF’s biofilm was an irregular pattern, comprising bacterial clusters (Fig. 1d). The biofilm formation in the soy-based formula presented as clustered formations with a honeycomb structure (Fig. 1e).

Part 2: The Comparison among Milk Formulas with Different Sucrose Compositions

Five types of milk formulas were evaluated according to the different sucrose compositions: cow milk-based formula, cow milk-based formula with 10% sucrose, cow milk-based formula with 18% sucrose, soy-based formula, soy-based formula with 11% sucrose.

Biofilm Formation

Visual observation revealed that cow milk-based formula with added sucrose had a thick and chalky appearance (online suppl. Fig. S3), whereas cow milk-based formula without added sucrose and soy based-formulas exhibited a more translucent biofilm. After staining with crystal violet, among formulas with similar protein compositions, those with added sucrose – both cow milk-based and soy-based – displayed significantly higher A590 nm values compared with their counterparts lacking added sucrose (Table 2).

Biofilm pH

In the presence of S. mutans, the biofilm pH of all milk formulas with different sucrose compositions dropped below the critical pH of enamel after a 24-h incubation (Table 2). Both cow milk-based formulas, with and without added sucrose, showed significantly lower biofilm pH values compared with the soy-based formulas. The lowest biofilm pH was in the cow milk-based formula with 18% sucrose. This type of milk formula also had the lowest pH, including in the condition without S. mutans. For the soy-based formulas, biofilm pH values were similar regardless of the added sucrose.

Lactic Acid Formation

With S. mutans, the highest concentration of lactic acid was found in the cow milk-based formula with 18% sucrose, which was significantly greater than the other milk formulas (Table 2). Moreover, the production of lactic acid from cow milk-based formulas with 0% and 10% sucrose was significantly higher compared with soy-based formulas with 0% and 11% sucrose. There was no significant difference in lactic acid production between soy-based formula with and without sucrose.

Biofilm Characteristics

When considering the biofilm structures of cow milk-based formulas, adding sucrose resulted in the aggregation of S. mutans biofilm as clusters, similar to the pattern observed in the control group, BHI with 10% sucrose. We observed a direct correlation: the higher the percentage of added sucrose, the larger the size of these clusters (Fig. 1f, g). In contrast, the biofilm of soy-based formula without added sucrose appeared as clusters with a honeycomb structure (Fig. 1e), while the formula with added sucrose exhibited a denser and more intact biofilm (Fig. 1h).

This study evaluated cow milk-based formulas and the formulas for CMPA children. We investigated all types of milk by in vitro experiments: biofilm formation, determination of biofilm pH, and measurement of lactic acid production. S. mutans was selected as the representative cariogenic pathogen. Because having a salivary S. mutans count above 105 CFU/mL is an indicator of high caries risk [20, 21], the concentration of S. mutans in the assays was adjusted to 105 CFU/mL to represent the moderate to high caries risk status. However, it is important to recognize that human dental caries involves a complex interplay of plaque microorganisms, where S. mutans represents only a fraction of this diverse microbial community. Although our model, using S. mutans as a representative species, could contribute to understanding certain aspects of the cariogenic process, it was limited by focusing solely on one species.

To compare milk formulas with different protein compositions, cow milk-based formula, pHF, eHF, AAF, and soy-based formula were evaluated (Table 1). No sucrose was added in these five formulas; thus, we could observe the effect of the protein composition in each milk. The biofilm formation assay indicated that the protein compositions of the milk formulas influenced the quantity and characteristics of the formed biofilm. The quantification of the biofilm mass by crystal violet staining demonstrated that S. mutans formed more biofilm in the cow milk-based formula compared with o the soy-based one (Table 1). These results were consistent with previous reports [14, 15]. In comparing protein hydrolysate formulas with varying molecular weights of proteins, the pHF displayed biofilm amounts similar to cow milk-based formula. In contrast, both eHF and AAF resulted in significantly lower biofilm amounts compared with cow milk-based formula, pHF, including the soy-based formula. To our knowledge, there are no reports on the biofilm formation of S. mutans in either pHF or eHF. Additionally, only one study, conducted by Sadan et al. in 2020, examined the biofilm formed in AAF. Their findings revealed that AAF exhibited the lowest biofilm formation compared with cow milk-based and soy-based formulas. In addition, our study is the first to use CLSM to demonstrate S. mutans biofilm in various protein-based milk formulas. Although crystal violet staining showed no significant difference in biofilm amounts between pHF and cow milk-based formula, CLSM revealed distinct biofilm structures (Fig. 1a, b). Furthermore, CLSM images for eHF and AAF supported the biofilm mass results, displaying only bacterial cells and small bacterial clusters, respectively (Fig. 1c, d).

After a 24-h incubation, the biofilm pH with S. mutans in all formulas with varying protein compositions dropped below the critical pH level (Table 1). This aligns with findings from similar in vitro studies using the same incubation period [15, 19]. Interestingly, we observed a pH decrease below 5.5 in the eHF without S. mutans (Table 1). Because the eHF used in our study contains Lactobacillus rhamnosus as a probiotic (online suppl. Table S1; Fig. S1e), the acidic pH observed in this formula with no S. mutans may arise from the fermentation products of this probiotic. However, the pH of non-incubated eHF solution was the lowest among all formulas (online suppl. Table S3). In the presence of S. mutans, biofilm pH values were significantly lower in cow milk-based and AAF formulas compared with the pHF, eHF, and soy-based groups (Table 1). The lactic acid concentrations in cow milk-based, pHF, and AAF were significantly higher than those in eHF and soy-based formulas when S. mutans was present (Table 1). These differences in lactic acid levels explained why cow milk-based and AAF had distinct biofilm pH values compared with eHF and soy-based formulas. However, factors other than lactic acid might contribute to the biofilm pH because pHF had higher lactic acid concentration, but no significant pH difference from eHF and soy-based formulas. Furthermore, the buffering capacity of milk based on protein types could not fully explain the biofilm pH values (Table 1).

Previous studies on protein hydrolysate formulas primarily relied on plaque pH after rinsing the oral cavity with a milk solution [6, 22]. The present study, however, conducted in vitro experiments. We found that among milk formulas with different protein compositions, eHF had the lowest number of factors contributing to caries due to minimal biofilm formation and the highest biofilm pH. These results align with Danchaivijitr et al. [6], who noted the eHF’s minimal pH drop in dental plaque compared with cow- and soy-based formulas. In contrast, a separate study found that protein hydrolysate and soy-based formulas caused a pH drop below the critical level, while cow milk-based ones did not [22]. Regarding AAF, only one study assessed S. mutans biofilm in this formula without reporting on its biofilm acidogenicity [14]. This study noted AAF’s low biofilm-forming ability, consistent with our findings, but we additionally observed a low biofilm pH alongside high lactic acid production. Hence, we strongly recommend further evaluation using both in vitro and in vivo studies for pHF, eHF, and AAF formulas. When comparing caries risk factors associated with biofilm formation between cow milk-based and soy-based formulas, our findings indicated a higher tendency in the cow milk-based group due to higher biofilm formation, lower biofilm pH, and increased lactic acid production. These results contradict prior studies that found a higher tendency with soy-based formulas [6, 23], and some studies found no significant difference in acidogenicity between soy-based and milk-based formulas [24]. However, these previous studies mostly used soy-based formulas with added sucrose, while our comparison involved the formulas without sucrose added to highlight the protein-based differences.

In the part evaluating the effect of sucrose amounts, we compared various milk formulas with different sucrose compositions: cow milk-based formula, cow milk-based formula with 10% sucrose, cow milk-based formula with 18% sucrose, soy-based formula, and soy-based formula with 11% sucrose. We specifically selected commercially available milk formulas with the most similar sucrose percentages for comparison, i.e., the cow milk-based formula with 10% sucrose and the soy-based formula with 11% sucrose. In addition, the cow milk-based formula with 18% sucrose was included because it is found as the commercially available formula with the highest percentage of sucrose. This is a high caloric density formula recommended for children who lack an appetite and those who have a nutritional risk. Notably, all three formulas containing sucrose are products of the same company, implying that the sucrose percentages stated on their labels can be compared.

The effect of sucrose supplementation on enhancing S. mutans biofilm formation was demonstrated when comparing cow milk-based and soy-based formulas without sucrose to the same protein-based formulas with sucrose added (Table 2). Similar to our results, most studies reported that the cow milk formula containing sucrose led to higher biofilm formation than the formula containing only lactose [10, 15]. A similar observation was also found for plant-based milk, such as almond milk [25]. Sucrose is generally accepted as the most cariogenic sugar [7], and glucan synthesis and binding in the process of sucrose-dependent adhesion enable S. mutans accumulation to a critical density at which point the plaque becomes pathogenic [17, 26]. Moreover, the effect of sucrose supplement on the biofilm characteristics by CLSM was different among different types of milk formulas. Cow milk-based formulas with added sucrose demonstrated aggregation of S. mutans in the biofilm (Fig. 1f, g), and soy-based formula with added sucrose demonstrated a denser intact biofilm (Fig. 1h). The involvement of these different biofilm characteristics in cariogenic potential is of interest for further study.

Unlike the observed enhanced biofilm formation due to added sucrose, adding sucrose to cow milk-based and soy-based formulas showed no significant impact on acid production, as seen in biofilm pH and lactic acid levels (Table 2). However, a notable decrease in biofilm pH and increased lactic acid formation occurred only in cow milk-based formula with 18% sucrose compared with other cow milk-based formulas. Fermentation of other industrial sugars present in milk formulas, aside from lactose and sucrose, could affect biofilm pH and lactic acid levels. A previous study compared the properties of S. mutans in high fructose corn syrup versus sucrose, revealing higher acidogenicity in high fructose corn syrup, but thicker biofilm formation in sucrose [27]. Another study on maltodextrin-containing products found that they decreased biofilm pH, albeit less than sucrose [28]. Moreover, maltodextrin, while less acidogenic than sucrose, still significantly lowered plaque pH, potentially leading to enamel demineralization [29]. Based on our findings, the other sugars in milk formulas, except for sucrose, might be more important to acid production than adding 10−11% sucrose. Therefore, the sucrose composition in milk formulas and the other industrial sugars contained in milk formulas should be considered.

The limitation of this study was that the experiments were performed using in vitro models; thus, the results may differ from the conditions inside the oral cavity, such as the lack of cleansing and buffering effects provided by saliva. However, the methods utilized may reflect baby bottle-feeding, which is one of the major causes of ECC. Many organizations recommend bottle weaning between 12 and 24 months of age [30]. Given that primary teeth start to erupt at 6 months of age, prolonged and frequent naptimes or nighttime bottle-feeding after tooth eruption is associated with the development of ECC, particularly due to the sugar content in the bottle [31]. Moreover, our study evaluated the milk formulas prepared based on the manufacturer’s recommendation. An autoclave was not used to sterilize the milk solutions in this study. Because milk formulas are not sterile, we detected some bacteria in some types of milk formulas (online suppl. Fig. S1). According to the product labels, some of the bacteria such as Lactobacillus spp. were supplemented as probiotics to promote gut health (online suppl. Table S1). Therefore, the presence of these microorganisms may influence the results of either biofilm formation or acidogenicity in this study. Additionally, there are only limited sources of milk formulas, especially those for the cow milk allergy children, in Thailand. Therefore, it was not possible to compare milk formulas with same percentage of sucrose.

In conclusion, the biofilm formation and acidogenicity of milk formulas in this study suggested that all cow milk-based formulas and milk formulas used in CMPA may contribute to caries risks, including milk formulas that stated that they did not contain supplemental sucrose. Therefore, parents and caregivers should be informed about the need to practice good oral hygiene after consuming milk, particularly for infants who consume high caries risk formulas.

This study did not involve experiments on humans or animals, and no approval was required from an Ethics Committee.

The authors have no conflicts of interest to declare.

The study was supported by the Faculty of Dentistry, Mahidol University (DTRS-EG-2022-06). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

N.R. and A.S. initiated and conceptualized hypothesis. N.R., J.L., O.T., and R.S. designed and performed experiments. N.R., J.L., O.T., and A.S. collected and analyzed data. N.R. and J.L. wrote the paper. J.L. and A.S. supervised the project. All authors have read, edited, and approved the final manuscript.

All data generated or analyzed during this study are included in the article and its supplementary files. Further inquiries can be directed to the corresponding author.

1.
Luyt
D
,
Ball
H
,
Makwana
N
,
Green
MR
,
Bravin
K
,
Nasser
SM
, et al
.
BSACI guideline for the diagnosis and management of cow’s milk allergy
.
Clin Exp Allergy
.
2014
;
44
(
5
):
642
72
.
2.
Vandenplas
Y
.
Prevention and management of cow’s milk allergy in non-exclusively breastfed infants
.
Nutrients
.
2017
;
9
(
7
):
731
.
3.
Vandenplas
Y
,
Koletzko
S
,
Isolauri
E
,
Hill
D
,
Oranje
AP
,
Brueton
M
, et al
.
Guidelines for the diagnosis and management of cow’s milk protein allergy in infants
.
Arch Dis Child
.
2007
;
92
(
10
):
902
8
.
4.
Sicherer
SH
,
Sampson
HA
.
Food allergy: epidemiology, pathogenesis, diagnosis, and treatment
.
J Allergy Clin Immunol
.
2014
;
133
(
2
):
291
308
.
5.
Sauser
J
,
Nutten
S
,
de Groot
N
,
Pecquet
S
,
Simon
D
,
Simon
HU
, et al
.
Partially hydrolyzed whey infant formula: literature review on effects on growth and the risk of developing atopic dermatitis in infants from the general population
.
Int Arch Allergy Immunol
.
2018
;
177
(
2
):
123
34
.
6.
Danchaivijitr
A
,
Nakornchai
S
,
Thaweeboon
B
,
Leelataweewud
P
,
Phonghanyudh
A
,
Kiatprajak
C
, et al
.
The effect of different milk formulas on dental plaque pH
.
Int J Paediatr Dent
.
2006
;
16
(
3
):
192
8
.
7.
Koulourides
T
,
Bodden
R
,
Keller
S
,
Manson-Hing
L
,
Lastra
J
,
Housch
T
.
Cariogenicity of nine sugars tested with an intraoral device in man
.
Caries Res
.
1976
;
10
(
6
):
427
41
.
8.
Bowen
WH
,
Pearson
SK
.
Effect of milk on cariogenesis
.
Caries Res
.
1993
;
27
(
6
):
461
6
.
9.
de Mazer Papa
AM
,
Tabchoury
CP
,
Del Bel Cury
AA
,
Tenuta
LM
,
Arthur
RA
,
Cury
JA
.
Effect of milk and soy-based infant formulas on in situ demineralization of human primary enamel
.
Pediatr Dent
.
2010
;
32
(
1
):
35
40
.
10.
Giacaman
RA
,
Munoz-Sandoval
C
.
Cariogenicity of different commercially available bovine milk types in a biofilm caries model
.
Pediatr Dent
.
2014
;
36
(
1
):
1E
6E
.
11.
Tan
SF
,
Tong
HJ
,
Lin
XY
,
Mok
B
,
Hong
CH
.
The cariogenicity of commercial infant formulas: a systematic review
.
Eur Arch Paediatr Dent
.
2016
;
17
(
3
):
145
56
.
12.
Moimaz
SAS
,
Amaral
MA
,
Garbin
CAS
,
Saliba
TA
,
Saliba
O
.
Caries in children with lactose intolerance and cow’s milk protein allergy
.
Braz Oral Res
.
2018
;
32
:
e91
.
13.
de Carvalho
CA
,
Zanlorenzi Nicodemo
CA
,
Ferreira Mercadante
DC
,
de Carvalho
FS
,
Buzalaf
MA
,
de Carvalho Sales-Peres
SH
.
Dental fluorosis in the primary dentition and intake of manufactured soy-based foods with fluoride
.
Clin Nutr
.
2013
;
32
(
3
):
432
7
.
14.
Sadan
H
,
Shanthala
BM
,
Zareena
MA
,
Babu
G
,
Vijayan
V
.
In vitro evaluation of milk-based, soy-based, and amino acid-based infant formulas on Streptococcus mutans biofilm formation
.
J Indian Soc Pedod Prev Dent
.
2020
;
38
(
4
):
350
4
.
15.
Trongsilsat
S
,
Lapirattanakul
J
,
Surarit
R
,
Smutkeeree
A
.
In vitro comparison of biofilm formation and acidogenicity between human breast milk and other milk formulas
.
Pediatr Dent J
.
2020
;
30
(
2
):
57
63
.
16.
Bhaumik
D
,
Manikandan
D
,
Foxman
B
.
Cariogenic and oral health taxa in the oral cavity among children and adults: a scoping review
.
Arch Oral Biol
.
2021
;
129
:
105204
.
17.
Banas
JA
.
Virulence properties of Streptococcus mutans
.
Front Biosci
.
2004
;
9
(
9
):
1267
77
.
18.
Dashper
SG
,
Reynolds
EC
.
Lactic acid excretion by Streptococcus mutans
.
Microbiology
.
1996
;
142
(
1
):
33
9
.
19.
Huang
Y
,
Thompson
T
,
Wang
Y
,
Yu
Q
,
Zhu
L
,
Xu
X
, et al
.
Analysis of cariogenic potential of alternative milk beverages by in vitro Streptococcus mutans biofilm model and ex vivo caries model
.
Arch Oral Biol
.
2019
;
105
:
52
8
.
20.
Petti
S
,
Hausen
HW
.
Caries prediction by multiple salivary mutans streptococcal counts in caries-free children with different levels of fluoride exposure, oral hygiene and sucrose intake
.
Caries Res
.
2000
;
34
(
5
):
380
7
.
21.
Leal
SC
,
Mickenautsh
S
.
Salivary Streptococcus mutans count and caries outcome: a systematic review
.
J Minim Interv Dent
.
2010
;
3
(
4
):
137
47
.
22.
Sheikh
C
,
Erickson
PR
.
Evaluation of plaque pH changes following oral rinse with eight infant formulas
.
Pediatr Dent
.
1996
;
18
(
3
):
200
4
.
23.
Dashper
SG
,
Saion
BN
,
Stacey
MA
,
Manton
DJ
,
Cochrane
NJ
,
Stanton
DP
, et al
.
Acidogenic potential of soy and bovine milk beverages
.
J Dent
.
2012
;
40
(
9
):
736
41
.
24.
Moynihan
PJ
,
Wright
WG
,
Walton
AG
.
A comparison of the relative acidogenic potential of infant milk and soya infant formula: a plaque pH study
.
Int J Paediatr Dent
.
1996
;
6
(
3
):
177
81
.
25.
Lee
J
,
Townsend
JA
,
Thompson
T
,
Garitty
T
,
De
A
,
Yu
Q
, et al
.
Analysis of the cariogenic potential of various almond milk beverages using a Streptococcus mutans biofilm model in vitro
.
Caries Res
.
2018
;
52
(
1–2
):
51
7
.
26.
Hajishengallis
G
,
Michalek
SM
.
Current status of a mucosal vaccine against dental caries
.
Oral Microbiol Immunol
.
1999
;
14
(
1
):
1
20
.
27.
Ma
R
,
Sun
M
,
Wang
S
,
Kang
Q
,
Huang
L
,
Li
T
, et al
.
Effect of high-fructose corn syrup on the acidogenicity, adherence and biofilm formation of Streptococcus mutans
.
Aust Dent
.
2013
;
58
(
2
):
213
8
.
28.
Bhat
SS
,
Dubey
A
.
Acidogenic potential of soya infant formula in comparison with regular infant formula and bovine milk: a plaque pH study
.
J Indian Soc Pedod Prev Dent
.
2003
;
21
(
1
):
30
4
.
29.
Al-Khatib
GR
,
Duggal
MS
,
Toumba
KJ
.
An evaluation of the acidogenic potential of maltodextrins in vivo
.
J Dent
.
2001
;
29
(
6
):
409
14
.
30.
Bably
MB
,
Laditka
SB
,
Paul
R
,
Racine
EF
.
Age of bottle cessation and BMI-for-age percentile among children aged thirty-six months participating in WIC
.
Child Obes
.
2022
;
18
(
3
):
197
205
.
31.
Amores-Esparza
JM
,
Altamirano-Mora
V
,
Villacís-Altamirano
I
,
Montesinos-Guevara
C
.
Breastfeeding and bottle-feeding as risk factors for dental caries and malocclusions in children with deciduous dentition: a scoping review
.
J Int Oral Health
.
2022
;
14
(
5
):
447
53
.