Salivary Biomarkers for Diagnosis and Monitoring of Thyroid Diseases: A Systematic Review of Current Evidence

Introduction

The hypothalamic-pituitary-thyroid axis plays a crucial role in the regulation of numerous physiological processes. Thyroid hormones influence metabolism, growth and development, influencing the function of all tissues in the body, thereby affecting not only the physical well being but also mental and emotional health. ¹

These hormones regulate basal metabolic rate, energy expenditure, metabolism of carbohydrates, lipids, and proteins. ² They are also essential for normal somatic and neural development during childhood. They also regulate vital functions of the body including cardiovascular function and central nervous system activity. Therefore the diagnosis and management of thyroid disorders are important for the maintenance of healthy state of living. ³

Thyroid disorders are the most prevalent endocrine disorders worldwide and in India, it is estimated that more than 42 million people are affected, with the number still growing. The most common and treatable disorders among them is iodine deficiency goiter. However, other clinically important conditions include hypothyroidism, hyperthyroidism, Hashimoto’s thyroiditis and thyroid cancer. ⁴ Early diagnosis and monitoring are essential for effective management for these conditions.

Currently, the diagnostic approaches are centered on circulating thyroid hormones and related biomarkers in serum such as thyroxine (T4), triiodothyronine (T3), thyroid-stimulating hormone (TSH), and thyroid autoantibodies. Although well established and clinically reliable, these tests require invasive methods to collect samples, limiting the feasibility of frequent monitoring. ⁵ Considering the prevalence and need for frequent monitoring, currently the interest in identifying non-invasive diagnostic techniques has gained significant momentum.

In this aspect, recently, saliva has been considered as a promising biofluid of significant diagnostic value for several reasons. Ability for non-invasive collection, thereby facilitating repeated sampling has added more interest to this aspect. Saliva has been reported to reflect various systemic physiological and pathological changes by expressing various biochemicals such as hormones, proteins and inflammatory markers enabling detection and estimation for diagnosis. ⁶

Recent studies have reported presence of thyroid-related biomarkers in saliva, including thyroid hormones, metabolic signatures, glycan profiles, and immunological markers. However, the available evidence remains fragmented, and the diagnostic value of these salivary biomarkers in thyroid disorders has not been comprehensively synthesized. ⁷

Therefore, the present systematic review aims to synthesize and evaluate currently available evidence regarding salivary biomarkers that are associated with thyroid diseases. In addition, it also aims to assess their potential diagnostic applicability as non-invasive indicators of deviation in thyroid function.

Results

The search resulted in 207 articles and following PRISMA 2020 guidelines carefully, resulted in 20 articles, after screening fulltext, only 9 articles were included in the final stage. The flow diagram is shown in Figure 1. In current systematic review that included 9 observational studies, various molecules identified in saliva regarding thyroid diseases were summarized. There was no restriction on the type of thyroid disease where included studies analyzed thyroid-related antibodies, amino acids, electrolytes, metabolic profiles, glycoproteins, enzymes, and microbial signatures (Table 1). Techniques of analysis varied widely and significantly. For assessing the methodological quality, Newcastle–Ottawa Scale (NOS) was used and all articles had a score of about 7 to 8, which indicated that all studies were of good overall quality (Table 2). Further, only potential molecules or markers could be identified and these studies were not designed for assessing diagnostic accuracy.

OPEN IN VIEWER

Figure 1.

PRISMA flowchart.

OPEN IN VIEWER

Table 1.

PECOS Table (Data Extraction and Risk of BIAS Combined).

Author, year	Type of study, study design, sample size	Population (type of thyroid disease)	Type of salivary biomarkers	Analytical method	Key findings and conclusions	NEWCASTLE–OTTAWA QUALITY ASSESSMENT SCALE (Higher value better the quality is)
Al-Hindawi et al (2018) 8	Observational Study	Hypothyroid patients (20-53 y)	Thyroid antibodies (TPO-Ab and Tg- Ab)	ELISA	Salivary antibodies detected, but did not vary among control and test	8
Jiao et al (2022) 9	Observational Study	Thyroid cancer and thyroid nodule patients	Oral microbiome	NGS Sequencing	Alloprevotella, Anaeroglobus, Acinetobacter, unclassified Bacteroidales, and unclassified Cyanobacteriales were significantly enriched in the thyroid cancer group. In contrast, the microbiome of the healthy controls was mainly composed of the genera Haemophilus, Lautropia, Allorhizobium Neorhizobium Pararhizobium Rhizobium, Escherichia Shigella, and unclassified Rhodobacteraceae. The thyroid nodules group was dominated by genre uncultured Candidatus Saccharibacteria bacterium, unclassified Clostridiales bacterium feline oral taxon 148, Treponema, unclassified Prevotellaceae, Mobiluncus, and Acholeplasma. In contrast, the genera unclassified Rhodobacteraceae and Aggregatibacter dominated the healthy control group.	8
Kadhom and Radhi (2023) 10	Observational Study	Hyperthyroidism under medication. Control was healthy individuals	Salivary electrolytes and pH	Atomic absorption spectroscopy and colorimetry	Increase in K and Cl ion in saliva with hyperthyroidism. [Cl: Control − 0.266(0.011), Test − 0.327(0.008); P = .000; K: Control − 12.627(0.127), Test 15.660(0.182); P = .000]	8
Mykolaivna (2021) 11	Observational Study	Hypothyroidism in the young patient	Calcium and Phosphorus	Colorimetry and atomic absorption spectroscopy	Decreased calcium and increased phosphorus content in saliva, associated with periodontitis. With hypothyroidism, Ca/P ratio was doubled in periodontitis	7
Nosratzehi et al (2024) 12	Observational Study	Hashimotos thyroiditis	Glutathione peroxidase	ELISA	glutathione peroxidase was significantly lower in Hashimoto’s thyroiditis (13.26 ± 13.55) compared to control group (24.95 ± 18.06) (P = .002)	7
Ren et al (2022) 13	Observational Study	Thyroid cancer and thyroid nodule patients	Salivary Glycoproteins	Lectin Microarrays	Machine learning algorithms were able to identify specific glycan patterns that were significantly different between saliva samples from cancer patients and healthy individuals, potentially serving as biomarkers for diagnosis and prognosis.	7
Tang et al (2025) 14	Observational Study	Thyroid nodules	Metabolic Profiles	ELISA	Salivary metabolic profiles are useful to distinguish between benigh and malignant thyroid nodules	8
Zhang et al (2021) 15	Observational Study	Papillary thyroid cancer	Aminoacids	ultra-high performance liquid chromatography-high resolution mass spectrometry	10 amino acids showed consistent difference. alanine, valine, proline, phenylalanine were specifically predictive for thyroid cancer. (sensitivity: 91.2%; specificity: 85.2%)	8
Zhag et al (2022) 16	Observational Study	Thyroid cancer and thyroid nodule patients	Free-T3	Chemiluminescence method	salivary FT3 is an independent risk factor for PTC and may be utilized as a diagnostic biomarker.	8

OPEN IN VIEWER

Table 2.

Risk of Bias Assessment.

Author	Year	Selection				Comparability	Exposure			Total Score
		Adequate definition of patient cases	Representativeness of patient cases	Selection of controls	Definition of controls	Control for important or additional factors	Ascertainment of exposure	Same method of ascertainment for participants	Nonresponse rate
Al-Hindawi et al 8	2018	1	1	1	1	1	1	1	1	8/8
Jiao et al 9	2022	1	1	1	1	1	1	1	1	8/8
Kadhom and Radhi 10	2023	1	1	1	1	1	1	1	1	8/8
Mykolaivna 11	2021	1	1	1	1	0	1	1	1	7/8
Nosratzehi et al 12	2024	1	1	1	1	0	1	1	1	7/8
Ren et al 13	2022	1	1	1	1	0	1	1	1	7/8
Tang et al 14	2025	1	1	1	1	1	1	1	1	8/8
Zhang et al 15	2021	1	1	1	1	1	1	1	1	8/8
Zhao et al 16	2022	1	1	1	1	1	1	1	1	8/8

The narrative synthesis is described here. Of these studies, numerous studies explored thyroid cancer and thyroid nodules. Zhang et al ¹⁵ used ultra-high performance liquid chromatography to identify 10 amino acids in saliva that can be potentially used to predict the papillary thyroid carcinoma. In this study the sensitivity was 91.2% and specificity was 85.2%, implying high accuracy. In similar lines, Tang et al ¹⁴ have reported that salivary metabolic profiling could potentially discriminate between benign and malignant nodules. Likewise, Ren et al ¹³ combined lectin microarrays and machine learning to showcase unique glycan profiles present in saliva of thyroid based malignancies.

Another significant aspect of microbial differences that reflect in saliva were reported by Jiao et al ⁹ where they gave proof for specific genera enrichment (eg, Alloprevotella, Anaeroglobus) in thyroid cancer, and that is strikingly different from Haemophilus and Lautropia of normal individuals.

Various studies on hypothyroidism^8,11) could not find any significant variation of thyroid antibodies (TPO-Ab, Tg-Ab) in saliva samples, but could note reduced calcium and increased phosphorus. However, Kadhom and Radhi ¹⁰ have reported raised salivary potassium and chloride ion levels in hyperthyroidism. Nosratzehi et al ¹² could find significantly lower salivary glutathione peroxidase in Hashimoto’s thyroiditis.

In separate category, Zhao et al ¹⁶ showed that salivary free T3 (FT3) showed changes in thyroid diseases. and may be looked upon as an independent diagnostic marker for papillary thyroid carcinoma.

Discussion

Current systematic review has synthesized the existing evidence on salivary components that change in the thyroid diseases. It was aimed to arrive at potential components that can be used to diagnose disease. The included studies have reported a wide range of biomolecules such as thyroid hormones, amino acids, metabolomic signatures, salivary microbiome alterations, and immunological markers. These findings clearly suggest that diagnosis of thyroid disorders through saliva may be approached through multiple aspects. Despite multiple contemplated advantages of using saliva for diagnosis of thyroid dysfunction, the identification of specific biomarkers related to thyroid physiology still remains a critical challenge.

Previous studies have shown functional changes in salivary glands in relation with thyroid dysfunction, like alterations in salivary flow rate, pH, and other physicochemical properties and such changes are not pathognomonic for thyroid disease, consequently were not considered relevant to the primary objective of the current review.

The studies included in this review have reported multiple molecules as being related to thyroid dysfunction, however, they belong to different biochemical categories. Of them, thyroid-related molecules, viz. salivary free triiodothyronine (FT3) and thyroid autoantibodies were the most direct indicators of thyroid gland function. Zhao et al have reported the diagnostic utility of salivary FT3 for differentiating thyroid cancer from thyroid nodules. Al-Hindawi et al have detected thyroid-specific antibodies (thyroid peroxidase antibodies (TPO-Ab) and thyroglobulin antibodies (Tg-Ab)) and suggested that thyroid-specific molecules can be detected in saliva.

In addition to these thyroid-specific molecules, several reports have investigated metabolomic and biochemical changes associated with thyroid disease. Studies by Kadhom and Radhi ¹⁰ and Mykolaivna ¹¹ have shown increased concentrations of salivary potassium, chlorine, and phosphorus and decreased calcium levels in thyroid dysfunction. Nosratzehi et al ¹² have reported relation with reduction in salivary glutathione peroxidase levels with Hashimoto’s thyroiditis. Zhang et al ¹⁵ have shown metabolomic changes in saliva like increased alanine, valine, proline, and phenylalanine levels associated with thyroid carcinoma. These are specific observations but are not specific to thyroid pathologies, and hence require further validation.

Recent investigations have also explored the alterations in salivary metabolomics and glycoproteomic profiles using machine learning algorithms. Associations were found to be associated with thyroid disease, but applicability is limited due to smaller population sizes used in the studies.

With regard to salivary microbiome, Jiao et al ⁹ have shown that certain changes in bacterial composition in thyroid cancer in contrast with thyroid nodules and healthy controls. While microbiome alterations are more probably due to biochemical alterations in saliva due to thyroid pathology. Consistent changes in microbiome can be used to identify thyroid pathology after appropriate clinical validation.

Strengths of the current systematic review is the comprehensive synthesis of the currently available evidence in an area that has been less consolidated in the existing literature. Further, the review had studies of diverse analytical approaches that provide a broad overview of potential biomarkers. The structured search strategy and systematic screening process had ensured the reliability of the findings and have highlighted key areas for future research.

Coming to the limitations of the review, one major limitation observed was the methodological heterogeneity. Especially, the variations were seen in in saliva collection techniques, and analytical techniques. Another limitation is that none of the studies aimed to assess these biomarkers for diagnostic use and they were mere findings in the disease. This systematic review worked carefully to use information from them to propose and assess their diagnostic use. Another consequence of this heterogeneity is the inability to perform quantitative synthesis like meta analysis.

Further, studies have not related salivary concentration to serum concentration creating a lacuna in consistent correlation. One study Al-Hindawi et al ⁸ did atttempt such correlation but could not succeed in it. Therefore future studies must evaluate saliva–serum correlations for the potential thyroid biomarkers. Since salivary composition can be heavily confounded by other physiological processes, future studies must be designed to eliminate such confounding.

Summarily, the available evidence suggests that relevant molecules found in saliva can be potential biomarkers as non-invasive indicators of thyroid disease. Nevertheless, the current evidence is limited by various factors including small sample sizes, methodological and heterogeneity, thereby absence of validated thyroid-specific biomarkers. Also, results point to multiple parameters and not one parameter making it difficult to compare the outcomes. This review is aimed at consolidation of evidence and may not be complete in comparing the outcome of included studies.

From the clinical perspective, presence of a reliable salivary biomarkers for thyroid diseases can offer a non-invasive diagnostic method, that may be specifically useful in large-scale population screening and longitudinal monitoring of thyroid function, especially in pediatric, geriatric, and community-based settings. In this regard longitudinal cohort or case control studies are essentially needed. Further many more avenues may have to be explored with regard to salivary biomarkers in thyroid disease like small RNA therapeutics based approaches, microarray based approaches and so on in future. ¹⁷

Conclusion

This systematic review is one of the pioneering attempts to search for biomarkers related to thyroid in saliva. Current evidence points to few molecules as putative markers of thyroid disease in saliva, but yet correlation with serum levels are not established. This systematic review has collected useful evidence on change in salivary molecules in thyroid disease using reports with good methodology. Conclusively, Thyroid antibodies and thyroid hormones in saliva, at current evidence level may be explored further to become biomarkers for thyroid pathology, after standardization and validation. Salivary biomarker based diagnosis of thyroid diseases can be extremely useful in low resource settings.