In short: Two 2025 studies show where saliva testing is heading. One built a smart mouthguard with an electrochemical sensor for a periodontal inflammation marker. The other improved a lateral flow assay, the same format as a rapid test strip, to detect a periodontitis-related endotoxin at picogram levels in saliva. Both relied on carefully collected saliva, and both used Oasis collectors for it.

This article describes published research prototypes, not approved clinical tests. Oasis Diagnostics devices are for research use.

Key takeaways

  • A wearable mouthguard sensor detected MMP-8, a periodontal inflammation biomarker, and agreed with ELISA on clinical samples (r = 0.92).
  • A SERS-enhanced lateral flow assay detected P. gingivalis endotoxin in saliva down to 18 pg/mL, versus 0.3 ng/mL by ordinary colorimetric reading.
  • Both studies used Oasis collectors: Super•SAL was one of two collection systems evaluated in the mouthguard study, and Pure•SAL collected the saliva in the endotoxin study.
  • Saliva often needs pretreatment before a sensor can read it reliably.

Why saliva for point-of-care sensing?

Saliva can be collected without a needle, and sensors that read it could give quick answers outside a lab. The lateral flow assay (LFA) is described by the endotoxin study’s authors as the most popular rapid diagnostic tool because of its low cost, ease of use and fast response, but they note that conventional designs have relatively low sensitivity and cannot quantify.

Study 1: a smart mouthguard for periodontal inflammation

Researchers built a wearable electrochemical biosensor integrated into a smart mouthguard to detect salivary MMP-8, a marker of periodontal inflammation. The sensing layer is a molecularly imprinted polymer on an electrode, with graphene oxide to boost sensitivity. The system connects wirelessly and adds a deep-learning model to filter noise and predict periodontitis severity.

What they reported:

  • Detection limits. A limit of detection of about 82 ng/mL and limit of quantification of about 250 ng/mL, with linear response over a wide concentration range.
  • Clinical samples. Saliva from three cohorts, healthy individuals, people with periodontitis and recovered individuals, with 20 people in each.
  • Agreement with the lab method. Sensor readings on ELISA-validated clinical samples showed strong concordance with ELISA (r = 0.92). For standard solutions, more than 95% of measurements fell in clinically reliable zones A/B.

How saliva was collected. The authors evaluated two collection systems for retaining the MMP-8 biomarker: a super-absorbent pad (Super•SAL), which collected more than 1 mL of saliva in 1–3 minutes with a colour change from yellow to blue, and a centrifugation-dependent system (Salivette). They report that both were effective at retaining MMP-8 for the sensing step. Saliva was also pretreated to remove impurities that cause non-specific binding on the electrode. All participants gave written informed consent.

Study 2: picogram-level endotoxin detection in a lateral flow assay

This study targeted lipopolysaccharide (endotoxin) from Porphyromonas gingivalis, a bacterium that causes severe periodontitis. It used surface-enhanced Raman scattering (SERS) with silver-coated gold nanostars as the reporter, in an LFA cassette that slows the saliva flow by about 29 seconds compared with a normal strip.

What they reported:

  • Sensitivity. Detection limits of 18 pg/mL by SERS and 0.3 ng/mL by colorimetric reading.
  • An earlier version. The group’s previous colorimetric LFA in saliva reached 46.5 ng/mL, according to the paper.
  • Saliva pretreatment. Saliva from multiple individuals was collected and pretreated to remove components that interfere with endotoxin detection. Signal improved with pretreatment, which used a saliva filter and starch in a syringe.

How saliva was collected. Saliva was collected orally using a Pure•SAL saliva collector, in different volumes for analysis.

What this means for anyone building a saliva test

  1. Collection is part of the assay. Both teams treated collection and pretreatment as steps to validate, not afterthoughts.
  2. Know your volume. In the mouthguard study, the Super•SAL pad collected more than 1 mL in 1–3 minutes, with a visual indicator.
  3. Expect interferents. Saliva contains mucins and other components that can interfere. The endotoxin study found pretreatment improved the signal.
  4. Validate against a reference method. The mouthguard study compared against ELISA on clinical samples.

Related Oasis products

  • Super•SAL: high-volume collection with a yellow-to-blue volume indicator.
  • Pure•SAL: collection and purification, with a highly absorbent pad that removes mucinous material and a filtration medium that removes further interferants.
  • VerOFy: Oasis’s own rapid saliva point-of-care platform with immunochromatographic test strips and the LIAM scanner.
  • Research that used Oasis devices

The endotoxin study used Pure•SAL as a collector and then applied its own pretreatment. It did not test Pure•SAL’s filtration as the pretreatment.

Sources

  1. Jeon S, et al. A wearable electrochemical biosensor for salivary detection of periodontal inflammation biomarkers: molecularly imprinted polymer sensor with deep learning integration. Adv Sci. 2025. https://doi.org/10.1002/advs.202509658
  2. Picogram-level detection of endotoxin in human saliva using ultrasensitive surface enhanced Raman scattering in lateral flow assay. RSC Adv. 2025. https://doi.org/10.1039/d5ra04515k