The human microbiome—microbes and their genes inhabiting the body—plays an essential role in regulating health and disease, and is influenced by host conditions and several environmental factors, especially the diet (1). Excess sugar consumption is characteristic of the modern diets of children and adolescents (2–4), and associated with several health risks such as dental caries, obesity, and cardiovascular diseases (5–7). Evidence links the aberrant composition of the gut microbiome to the development of conditions including metabolic diseases such as obesity and diabetes (8), possibly a consequence of sugar consumption that may affect the gut habitat leading to altered bacterial communities and influencing metabolism (9). The majority of the mono- and disaccharides ingested are absorbed in the small intestine (9), and thus do not reach the large intestine unless consumed in large quantities (10), whereas oral bacteria are exposed to easily metabolized ingested sugars for which metabolism leads to harmful end-products such as lactic acid (11). Dental caries is a disease resulting in a net mineral loss of teeth due to the acidity produced by bacterial sugar metabolism (12). Furthermore, the frequent consumption of sugar leads to an increase in acid-producing bacteria in the dental biofilm and further to caries.

The role of sugar in the development of caries is well-known, whereas studies on the relationship between oral microbiota profiles and sugar consumption remain scarce. Additionally, existing evidence on the association between oral microbiota profiles and sugar consumption is inconsistent. In adults, no association was found between sugar consumption and saliva microbiota (13, 14), but, in contrast, in subgingival microbiota frequent sucrose consumption associated with decreased species richness and differences in the beta diversity (15), and some cariogenic bacteria such as Streptococcus sobrinus were less abundant among those who consumed fewer free sugars (14). Among 17- to 21-year-olds, saliva microbiota profiles varied according to differences in sucrose intake (16). We found only one study examining associations between sugar consumption and oral microbiota among children. In that study among 11-year-old children, the daily consumption of sugar-sweetened beverages was linked to less diversity and richness in oral microbiota and differences in the bacterial abundance (17).

Oral microbiota harbors the largest microbial community after the gut, consisting of several different microbial communities—that is, the microbial composition varies in different locations in the mouth (18). Saliva microbiota consists of bacteria shed from oral surfaces (19), thus representing different communities without having its own resident microbiota (20). Saliva microbiota has been previously linked to metabolic, autoimmune, and immunodeficient conditions, rendering it an affordable and feasible source for biomarkers (21). Furthermore, saliva microbiota may mirror the caries and periodontitis status of the host, and remains relatively stable in orally and systematically healthy individuals (19). Interestingly, oral bacteria can pass through the gastrointestinal tract to a greater extent than previously thought, possibly colonizing the large intestine even in healthy individuals (18, 22). In some disease states, such as rheumatoid arthritis and colorectal cancer, the flux from the oral cavity to the gut may be more pronounced (22), suggesting that oral microbiota plays a role not only in oral diseases, but in systemic health as well.

This study aims to examine whether saliva microbiota profiles, including the diversity and composition, differ according to the consumption of sugary products (“sweet treats”) among school-aged children. In addition, we aim to explore the functional capacities of saliva microbiota. To our knowledge, this is the first study to examine the associations between sugary products and saliva microbiota and its functional capacities. Thus, our study provides novel evidence about the relationships between saliva microbiota and sugar consumption among school-aged children.

This study utilized material from the Finnish Health in Teens (Fin-HIT) study, a large, geographically diverse cohort consisting of over 11 000 children and adolescents aged 9–12 years old at enrolment. Data were collected in 2013 and 2014 in schools across Finland. A detailed description of the cohort appears elsewhere (23). We previously analyzed saliva samples from 1,000 randomly selected Fin-HIT participants (24). After exclusion based on antibiotic use 3 months prior to saliva sampling, missing values in sweet treat consumption as well as low sequencing depth, 700 children remained in the cohort for further investigation.

In this study of Finnish children aged approximately 12 years old, we found differences in the composition and functional capacities of saliva microbiota between those with low sweet treat consumption and those with high sweet treat consumption. As sweet treats, we considered a range of commonly consumed sugary drinks and foods, reflecting a relevant health-related eating behavior instead of individual food items. The groups did not differ in terms of their oral health. We identified no differences in the alpha diversity between groups, whereas different compositions of saliva microbiota were observed. We also found several genera differentially abundant in the group with a high sweet treat consumption. Therefore, we sought to identify their effects by predicting their functional capacities. We uncovered two metabolic pathways relating to nitrate and gondoate metabolism highlighted in the group with a high sweet treat consumption.

We found no difference in the diversity and richness of saliva microbiota between low and high sweet treat consumption frequencies. In a study among 11-year-old Chinese children, 30 children reporting daily consumption of sugar-sweetened beverages presented with less rich and diverse oral microbiota when compared with 150 children who consumed such beverages less than six times a week (17). By contrast, a study measuring the intake of free sugars among young Danish adults found no difference in the alpha diversity between low (<5% of total energy intake [E%]) and high intake (≥5E%) groups (14). In that study, they measured the amount rather than the frequency of intake, finding that even the high consumption group exhibited a mean intake below the recommendation of no more than 10E% (36). The different measurements of sugar consumption in these studies may partly explain the differing results.

In contrast to the diversity, we found that the composition of microbiota differed according to sweet treat consumption. A relative abundance analysis revealed 37 differentially abundant OTUs, indicating that sugar consumption (measured as a use frequency of sweet treats) is a relevant predictor of the composition of saliva microbiota. We observed OTUs belonging to the core taxa including Veillonella, Prevotella, and Streptococcus (37) enriched in the group with a high sweet treat consumption. Although part of the core microbiota, these genera may associate with oral diseases (38–40). Veillonella species are non-motile anaerobic bacteria in the phylum Firmicutes, abundantly found in the oral cavity (41). Some oral Veillonella species have been associated with the development of caries, periodontitis, peri-implantitis, and other oral diseases (38, 42), and, for example, V. parvula associates with severe early childhood caries (43). However, some Veillonella species. convert lactic acid to weaker acids, possibly diminishing caries (41). Moreover, they can convert nitrate into nitrite, which is favorable to oral health. As an early colonizer in the oral cavity, Veillonella play a critical role in guiding the development of polymicrobial biofilm communities in the oral microenvironment (42). Prevotella are among the most dominant and abundant bacteria in the oral cavity with moderately saccharolytic abilities (44). Furthermore, Prevotella have been connected to early childhood caries (40), and P. intermedia was shown to be characteristic in patients with periodontitis (45).

Streptococci form a part of the core microbiota in the oral cavity, with the majority of Streptococcus species being acidogenic and/or acid-tolerant (46). Such bacteria have the ability to metabolize carbohydrates through fermentation and produce acids as well as efficiently colonize oral tissues (39). S. mutans is a well-known cariogenic species (47), and, while present in a healthy mouth, its abundance increases with the growing availability of sucrose (15). Sucrose increases the biomass of S. mutans in the biofilm through an acidic environment and intra- and extracellular polysaccharide synthesis (48), with frequent exposure to dietary sugar acidifying the microenvironment on the teeth leading to enamel demineralization (39). In addition to S. mutans, S. sobrinus is another species linked to caries (39). By contrast, some Streptococcus species can neutralize acidity by producing alkali, as well as producing hydrogen peroxide and antimicrobial compounds (39), thus inhibiting the growth of S. mutans. In a healthy mouth, a balance between cariogenic bacteria and non-cariogenic commensal bacteria exists. Haemophilus is another common inhabitant of the oral cavity, including H. parainfluenzae, which appears to carry some beneficial immunomodulatory effects (49). Then again, H. influenzae is a well-known pathogen (50). Here, we noticed a lower abundance of Haemophilus in the high sweet treat consumption group compared with the low sweet treat consumption group.

We found Treponema and Staphylococcus were less abundant in the high sweet treat consumption group compared with the low sweet treat consumption group. Treponema has been associated with periodontitis as well as correlating positively with inflammatory cytokines in these patients (51). Although T. denticola belongs to the red complex, which is considered the most destructive group of periodontal pathogens (52), and is commonly found in patients with periodontal diseases, other Treponema species associate with periodontal diseases as well (53). By contrast, S. epidermidis was found enriched in caries-free children (54), and some antibiotic-resistant S. epidermis have been identified in the dental plaque of healthy individuals (55). Then, again, pathogenic species S. aureus has been linked to several oral diseases such as periodontitis (56). The clinical relevance of these findings remains unclear.

When we predicted the functional capacities of the microbiota, we found two metabolic pathways—nitrate reduction and gondoate biosynthesis—enriched in the group with a high sweet treat consumption. Nitrate metabolism has been associated with a lower occurrence of caries and gingival inflammation, and appears to participate in the maintenance of host oral and systemic health (57). Nitrate (NO₃^–) appears to inhibit the salivary acidification resulting from glucose ingestion. Denitrifying oral bacteria reduce nitrate to nitrite (NO₂^–) and further to nitric oxide (NO), which has antimicrobial properties and can possibly limit the growth of certain bacteria affecting the composition of the biofilm (57). Nitrite can be reduced to ammonium as well and the presence of nitrate in the oral cavity and is associated with a decrease in the production of lactate and an increase in the production of ammonium. This may result in neutralizing pH and further protect from caries. Moreover, the enterosalivary nitrate–nitrite–nitric oxide pathway depends on oral nitrate-reducing bacteria, and these bacteria contribute to the storage pool of nitrite and nitric oxide in the blood and tissues, consequently positively influencing the host’s physiological status such as blood pressure (58). Dietary sugars, especially in the form of sugar-sweetened beverages, may impact blood pressure (59). Taken together, the nitrate reduction pathway may be upregulated to neutralize acidity resulting from frequent sugar consumption. This may in part explain the similar caries status we found in low and high sweet treat consumption groups. Gondoate is a long-chain unsaturated fatty acid, and its biosynthesis pathway was enriched in the saliva microbiota of adult patients with oral squamous cell carcinoma (60). In the gut microbiome, gondoate biosynthesis decreased during the exacerbation of Crohn’s disease (61). The clinical relevance of our findings that the gondoate pathway upregulates in the high sweet treat consumption group remains unclear.

We adjusted the alpha and beta diversity analyses for age, sex, waist–height ratio, and maternal SES. These potential confounders did not influence the alpha and beta diversities, indicating that sweet treat consumption independently impacts microbial diversity and composition at least in the saliva. In addition, we ran a sensitivity analysis among those children for whom information on their oral health was available (the DMFT index to indicate a history of cavitated caries lesions and CPITN to indicate gingival health), and adjusted the alpha and beta diversity analyses further for these variables. In this study, these were utilized as dichotomous variables, and adjusting for them changed neither the alpha nor the beta diversity results. Given the nature of the dichotomous variables, we were unable to identify those with more severe oral health problems. However, we found some OTUs related to oral infection-associated genera at a higher abundance in the high sweet treat consumption group. Oral health was examined on average 4.3 (±2.2) months prior to or after saliva sampling, but this was unlikely to have impacted these results.

The strengths of the study include the large number of children in the Fin-HIT cohort study conducted across Finland as well as the anthropometric measurements and saliva sampling completed by trained fieldworkers. Moreover, data relating to oral health were obtained from objective and reliable national health care registers. Another strength of our study lies in our use of a summed variable, indicating the consumption of several different types of sugary drinks and foods. Yet, we acknowledge that this measurement was not capable of capturing all possible sources of added sugar. Nonetheless, this study carries some potential limitations as well. There was a lack of standardization in saliva sampling in relation to the time of day or meal, although sampling was carried out primarily in the morning after breakfast, but before lunch. The timing of food intake may influence the diurnal variation of microbiota (62), although microbial profiles were found to remain stable for 24 hours and even for a week (63). We had no information on the putative cases with a low saliva flow rate, possibly impacting microbiota (20). Furthermore, our data do not include information on dental hygiene habits. To overcome this, we considered CPITN as a reflection of such habits. Not assessing smoking represents a minor limitation, given that smoking is extremely rare among Finnish 12-year-old children (64). Food consumption information was assessed through a short, self-administered food frequency questionnaire, and while it is feasible in a large cohort of school-aged children, it does not provide detailed information on the components of the diet. Moreover, misreporting is possible when measuring food consumption. Then, again, food frequency questionnaires rank participants according to their food consumption (65), which we achieved here. Data on saliva microbiota were based on 16S rRNA gene amplicon sequencing and the Silva reference database,¹ which provide species-level identification of bacteria for only some OTUs. Moreover, we relied on sequencing data from previous Fin-HIT studies (24), and thus, were limited to a genus-level analysis. Species-level identification of the bacteria would have been beneficial. We had no information on total bacterial loads, possibly relevant for this topic (66, 67). In addition, the functionality of microbiota was predicted using PICRUSt2, which provides >80% reliability to pathway discovery (34, 68). Some methodological challenges may be overcome using shot-gun metagenomic approaches instead (69).

To conclude, this study revealed that sweet treat consumption shapes saliva microbiota and its functions in school-aged children. The frequent consumption of sweet treats associated with a differentially abundant microbiota as well as with differentially expressed metabolic pathways. Our findings improve our understanding of the impact of sugary diets on the oral microbiome and can be used to target future studies.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: EGA; EGAS00001003039.

This study adhered to the Declaration of Helsinki, and the children and one guardian of each child provided their written informed consent before participating. The Ethics Committee of the Hospital District of Helsinki and Uusimaa approved the study protocol (169/13/03/00/10).

SL performed the data analysis with assistance from MM and NA, and drafted the manuscript, tables, and figures. All authors contributed to the study design, participated in the data interpretation, revised the manuscript, and accepted the final manuscript for submission.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.