Root caries is prevalent in the population, with a global prevalence rate of 41 % [1]. As natural teeth last longer and gingival recession from periodontal diseases increases, older adults are more vulnerable to root caries [2]. With the ageing population, untreated root caries contributed significantly to the global burden of oral diseases [3].

Early prevention is essential to reduce the burden of untreated root caries, primarily because of the difficulties in managing advanced root carious lesions. Root caries tend to spread laterally, creating shallow yet extensive lesions that challenge the retention of restorative materials, especially in hard-to-access proximal regions. Restorative treatments for these lesions often have a poor prognosis due to their proximity to the gingival margin, which compromise effective moisture control [4]. Therefore, effective strategies are demanding to prevent and manage root caries, ideally in its early stage.

Understanding the aetiology of root caries is key to developing effective preventive strategies. Root caries is a multifactorial disease, with microbial factors playing a pivotal role [5]. Early culture-dependent studies identified various acid-producing and protein-degrading bacteria predominant in root carious lesions, including species from the groups Mutans streptococci, Lactobacilli, Actinomyces, Propionibacterium and Bifidobacterium, revealing a polymicrobial nature of root caries [6]. Culture-independent approaches later uncovered greater oral microbial diversity [7], yet 16S rRNA gene sequencing studies fail to reach a consensus on root caries-associated microbial changes partly due to its inherent limitations [8]. A significant constraint of 16S rRNA amplicon sequencing is its inability to deliver accurate species-level information [9], which is problematic since different species within the same genus can exhibit site-specific preferences and distinct roles in dental health, either promoting or preventing caries [10].

Furthermore, previous person-level case-control studies struggled to discriminate disease-associated microbial shifts from interindividual variation. Research indicated that microbiota in caries-affected patients demonstrated even greater interindividual variation compared to that of caries-free adults [11], highlighting the need to control for host-related variations in the microbiome. Moreover, Yang et al., identified tooth-specific spatial shift in microbial composition using single-tooth resolution sampling [12], raising concerns about potential bias associated with pooled sampling strategy.

To overcome these limitations, we employed Type IIB Restriction-site Associated DNA for Microbiome (2bRAD-M)—a method validated for accurate species resolution even from low DNA yields and host-contaminated samples. By utilizing Type IIB restriction enzymes to generate isolength DNA fragments for sequencing, 2bRAD-M effectively reduces PCR bias, ensuring high fidelity in profiling low biomass clinical samples that required extensive PCR amplification [13]. Benchmarking studies confirm its high sensitivity and comparable robustness to whole genome sequencing in detecting microbial signatures [13,14]. This study aims to assess both person-level and tooth-level changes of species-resolved microbial and functional profile in root caries biofilms of older adults.