Association Between Scalp Microbiota Imbalance, Disease Severity, and Systemic Inflammatory Markers in Alopecia Areata

Abstract

Introduction

Alopecia areata (AA) is an autoimmune disease causing non-scarring hair loss, with both genetic and environmental factors implicated. Recent research highlights a possible role for scalp microbiota in influencing both local and systemic inflammatory responses, potentially impacting AA progression. This study examines the link among scalp microbiota imbalances, AA severity, and systemic inflammation.

Methods

We conducted a cross-sectional study with 24 participants, including patients with AA of varying severities and healthy controls. Scalp microbial communities were analyzed using swab samples and ion torrent sequencing of the 16S rRNA gene across multiple hypervariable regions. We explored correlations among bacterial abundance, microbiome metabolic pathways, and circulating inflammatory markers.

Results

Our findings reveal significant dysbiosis in the scalp microbiota of patients with AA compared to healthy controls. Severe AA cases had an increased presence of pro-inflammatory microbial taxa like Proteobacteria, whereas milder cases had higher levels of anti-inflammatory Actinobacteria. Notable species differences included abundant gram-negative bacteria such as Alistipes inops and Bacteroides pleibeius in severe AA, contrasted with Blautia faecis and Pyramydobacter piscolens predominantly in controls. Significantly, microbial imbalance correlated with AA severity (SALT scores) and systemic inflammatory markers, with elevated pro-inflammatory cytokines linked to more severe disease.

Conclusion

These results suggest that scalp microbiota may play a role in AA-related inflammation, although it is unclear whether the shifts are a cause or consequence of hair loss. Further research is needed to clarify the causal relationship and mechanisms involved.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13555-024-01281-2.

Key Summary Points

This study investigates the relationship among scalp microbiota imbalance, alopecia areata (AA) severity, and systemic inflammatory markers.

Significant dysbiosis was observed in the scalp microbiota of AA patients, with distinct microbial profiles associated with disease severity.

Patients with severe AA exhibited higher levels of pro-inflammatory microbial taxa, while milder cases showed more anti-inflammatory taxa.

The study suggests that scalp microbiota could serve as a biomarker for AA severity, potentially guiding more personalized therapeutic approaches.

Further research is necessary to clarify whether these microbiota shifts are a cause or consequence of AA and their role in systemic inflammation.

Introduction

Alopecia areata (AA) presents as an autoimmune condition that causes non-scarring hair loss and significantly impairs the quality of life for those affected, leading to emotional distress and social stigma due to its unpredictable and visible hair loss [1–3]. Severe forms, such as alopecia totalis (AT) and alopecia universalis (AU), are associated with higher relapse rates and poor outcomes, with mechanisms that remain unclear [4]. AA's link to other autoimmune diseases and potential increased risk of cardiovascular and thromboembolic events further affects patients' overall well-being [5, 6]. Treatment options are still limited, showing efficacy only in a portion of cases and often presenting safety concerns [7]. For all these reasons, intensive research into the pathophysiology of AA is crucial to develop more effective and safer therapies.

Its pathogenesis has not yet been fully elucidated, but it is speculated to be the result of a complex network of factors including stress, genetic predisposition, environmental influences, and immune dysregulation [8, 9]. Emerging evidence suggests bidirectional communication between these skin-gut ecosystems, whereby alterations in gut microbiota composition can impact skin health and vice versa [10]. Factors such as diet, antibiotics, and stress can influence both gut and skin microbiota, underscoring their interconnectedness [11].

Immune privilege collapse in AA, though not fully understood, may be triggered by bacteria, viruses, fungi, or stress in genetically predisposed individuals, activating CD8 ⁺ NKG2D ⁺ T lymphocytes [12]. These cells produce interferon gamma (IFN-γ) and upregulate MHC-I, leading to a perifollicular infiltrate of CD8⁺ and CD4 ⁺ T lymphocytes, mast cells, and NK cells [13]. This damages hair follicles in the anagen phase, promoting premature apoptosis.

The immune response is sustained by IL-15 production from keratinocytes, which activates CD8 ⁺ NKG2D ⁺ T lymphocytes via pathways such as MAP kinase, mTOR, and JAK/STAT. This results in upregulation of key genes like MICA, ULBP3, IL-15, and CXCR3, perpetuating the inflammatory process by producing NKG2D ligands and chemokines, such as CXCR3, that further escalate the immune attack on hair follicles.

The cutaneous microbiome, comprising a diverse community of bacteria, fungi, and viruses, exists in a symbiotic relationship with epidermal cells and the immune system [14]. The microbial consortium, crucial for skin health, protects against pathogenic colonization, aids immune regulation, facilitates tissue repair, and maintains biological barrier integrity, while also adapting to diverse skin conditions to support keratinocyte development and systemic immunological balance [15]. The variability of the cutaneous microbiome among different individuals and skin regions is attributed to a combination of factors such as age, gender, seasons, and ethnicity [16–19]. For example, research has revealed that microbial colonization of the skin begins immediately after birth and continues to evolve during the first year of life, highlighting the importance of age in shaping the microbiome [20, 21].

Preliminary studies on the role of the cutaneous microbiome in various immune-mediated inflammatory skin diseases have attracted considerable attention [22, 23]. In the context of inflammatory skin diseases, the involvement of the microbiome has been investigated in conditions such as psoriasis [24], hidradenitis supurativa [25], atopic dermatitis [26–29], hand eczema [30], rosacea [31], acne [32, 33], wound healing [34], lichen striatus [35], and vitiligo [36].

Current research on AA is still in its nascent stages, with more in-depth studies required to better understand how variations in the microbiome and systemic inflammatory responses affect disease severity. Our study specifically aimed to analyze and quantify the bacterial populations on the scalps of patients with AA compared to healthy controls, identify the predominant bacterial phyla, genera, and species, explore the relationship between skin microbiota composition and AA severity, and investigate how the microbiome varies with disease severity. This comprehensive approach will help clarify the role of the cutaneous microbiome in the pathogenesis of AA and its potential impact on disease progression.

Material and Methods

Inclusion and Exclusion Criteria

Participants eligible for the study were required to meet specific criteria to ensure the integrity and validity of the research findings. Specifically, individuals with AA should not have undergone any systemic immunosuppressive therapy in the 3 months preceding enrollment. Additionally, topical immunomodulators, including corticosteroids, were prohibited for at least 1 month before participation. Individuals with other inflammatory skin or systemic diseases, except for those with stable thyroiditis in a euthyroid status, with or without medication, were excluded. Participants diagnosed with concomitant androgenic alopecia were also excluded. All participants were required to comply with the inclusion and exclusion criteria. This study was conducted according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [37]. The research was conducted following the standards of good clinical practice and adhering to the principles outlined in the by the Helsinki Declaration of 1975, as revised in 1983, and Belmont Report and adhered to Spanish and European laws regarding data privacy. Before participation, all patients provided informed consent after receiving and reviewing the patient information sheet.

Clinical and Sociodemographic Variables

Epidemiological variables such as sex and age were collected from all subjects, along with disease duration. For patients with AA, severity was assessed using the Severity of Alopecia Tool (SALT), with mild to moderate AA characterized by a SALT score < 50 and a disease duration < 1 year [38]. This classification, which includes individuals with single or multiple patchy areas of hair loss, provides a clear framework to distinguish between disease severities. Severe AA was defined as a SALT score ≥ 50 or a disease duration > 1 year, further categorized into single plaque, multiple plaque, alopecia totalis (AT), or alopecia universalis (AU).

Sample Collection and Processing

Swab samples were collected from lesional scalp areas of patients with AA and from the shaved heads of healthy controls. In both groups, participants were instructed not to wash their hair for at least 24 h prior to sample collection. The scalp samples from controls were obtained after head shaving in the hospital ward the day before surgery, without any head washing and well before the application of antiseptic products on the day of surgery. Samples were collected using 4N6 FLOQ Swabs Genetics® (Copan Italia, Italy) and stored at – 80 °C. Blood samples were anticoagulated with K2-EDTA, centrifuged at 1800 g for 10 min to obtain plasma, and then stored at – 80 °C for cytokine analysis.

Skin Microbiome Sequencing

DNA was extracted from the samples using the QIAamp Fast DNAStool mini kit (Qiagen®, Germany) and processed according to the manufacturer's protocol. The integrity of DNA was evaluated with a BioAnalyzer 2100 (Agilent®, USA) and Qubit 3.0 fluorometer using the dsDNA HS assay (Life Technologies®, USA). Six amplicons from seven hypervariable regions of the 16S rRNA gene were sequenced on Ion Torrent S5™ following the Ion 16S Metagenomics Workflow. Sample analysis for quality control, preprocessing, and database consultation was carried out using bioinformatic analyses. For more details, see Supplementary Methods.

Microbiome Analyses

We quantitatively analyzed community profiles using well-established statistical methods such as abundance profiling, alpha and beta diversity analyses, clustering, and correlation network analysis to unveil inherent patterns and correlations within our data and identify significant features or potential biomarkers via statistical and machine learning methods. This involved using bar charts, box plots, or other visualizations to display the relative abundances.

Functional Prediction Analysis

Our analysis employed functional prediction to discern significant features or potential biomarkers through the application of statistical and supervised machine learning techniques. By utilizing MetaCyc [39] and Kyoto Encyclopedia of Genes and Genomes (KEGG) [40] pathway enrichment analyses, we explored microbial community samples from individuals with mild and severe AA and from healthy control subjects.

Inflammatory Plasma Markers

One microliter plasma from all individuals was analyzed using OLINK Proseek® multiplex high-throughput platform with inflammation. The list of 92 markers is detailed in Supplementary Table S1. Protein expression profiles were modeled by linear models using R’s limma package. Correlations were performed between the plasma levels of the proteins and the operational taxonomic unit (OTU) counts at different levels (phylum). Proteins with fold changes/FCHs > 1.3 and FDR < 0.05 were considered differentially expressed.

Statistical Analyses

Statistical significance was established through Kruskal-Wallis and Dunn's post hoc tests for multiple comparisons. Spearman's rank correlation coefficient assessed associations between bacterial abundance and disease severity and between bacterial taxa and plasma inflammatory protein concentrations. Linear discriminant analysis effect size (LEfSe) facilitated precise identification of differentially abundant pathways. Principal component analysis (PCA) explored variations in bacterial abundance relative to plasma protein concentrations and disease severity. P-values underwent adjustment using the Benjamini-Hochberg procedure for false discovery rates (FDRs), with two-sided FDR < 0.05 indicating statistical significance.

Statistical analyses were performed using the statistical language R [41] and packages available through Bioconductor [42], and Python libraries (QIIME2, skbio, NumPy, Pandas) in the Jupyter Notebook [43]. The web tool MicrobiomeAnalyst 2.0 functionalities were used for the rest of the comparative studies, plots, and regression models [44].

Results

Population Characteristics

The study included 19 patients with AA and 4 healthy controls (Table 1). The median age of patients with AA was 41 years (IQR 38–58), with 11 women and 8 men. Controls had a median age of 37 (IQR 33–79) years, with two women and two men. AA severity ranged from mild to severe, with 8 severe cases and 11 mild-to-moderate cases. Among the participants, 8 were diagnosed with severe AA, exhibiting a median SALT score of 72.5% ± 14.6% (including three with AU), while 12 individuals presented with mild AA, with a median SALT score of 7% ± 33.5%. Additionally, four participants were categorized as healthy controls. Among the participants, 8 were diagnosed with severe AA, exhibiting a median SALT score of 72.5% ± 14.6% (including three with AU), while 12 individuals presented with mild AA, with a median SALT score of 7% ± 33.5% (Photograph S1) (Table 1).

Table 1.

Sociodemographic characteristics of the study population

Subject code	Severity	SALT	Sex	Age (years)	Pattern of AA	Comorbidities
ARE001	Mild	2%	Woman	76	AA single patch	Hypercholesterolemia, biliary dyspepsia, osteoporosis, dry syndrome, subclinical hypothyroidism
ARE002	Mild	2%	Woman	70	AA single patch	–
ARE003	Mild	2%	Woman	62	AA single patch	–
ARE005	Severe	60%	Man	47	AA multiple patches	–
ARE006	Mild	4%	Woman	24	AA single patch	–
ARE007	Mild	7%	Woman	38	AA single patch	–
ARE008	Mild	35%	Man	32	AA multiple patches	–
ARE011	Severe	70%	Man	41	AA multiple patches	–
ARE012	Mild	38%	Man	64	AA multiple patches	–
ARE013	Mild	7%	Man	38	AA single patch	–
BIO021	Mild	4%	Woman	34	AA single patch	–
ARE016	Severe	100%	Woman	41	AU	–
ARE017	Severe	62%	Woman	38	AA multiple patches	–
ARE018	Severe	60%	Man	61	AA multiple patches	–
ARE019	Mild	36%	Man	58	AA multiple patches	–
ARE020	Mild	19%	Woman	30	AA multiple patches	–
ARE028	Mild	37%	Man	58	AA multiple patches	–
ARE029	Severe	75%	Woman	46	AA multiple patches	–
ARE030	Severe	100%	Woman	54	AU	–
ARE031	Severe	100%	Woman	39	AU	–
CON204	Control	–	Man	37	–	Mild rosacea
CON214	Control	–	Woman	33	–	–
CON215	Control	–	Woman	79	–	–
CON216	Control	–	Man	23	–	–

AA alopecia areata, AU alopecia universalis

Microbiome Diversity in Scalp of Patients with AA

In our study, no statistically significant differences were found in alpha diversity across groups, suggesting that there is neither a higher nor a lower number of microorganisms on the scalp of A patients compared to control subjects (Fig. 1a–c). However, significant differences were found in beta diversity between controls and patients with AA, but not by severity (Fig. 1d–f). This indicates that the changes in the composition of microbial communities differ between healthy and diseased states, highlighting the impact of microbial diversity in distinguishing between health and disease conditions.

Fig. 1.

Comparative analysis of scalp cutaneous microbial abundance and diversity across control, mild, and severe conditions. This figure employs a combination of alpha and beta diversity metrics, pie charts, bar charts, and analyses of relative abundance to present a comprehensive view of microbial community dynamics. On the left side, panels (a–c) depict alpha and beta diversity metrics, providing quantitative comparisons of microbial richness and similarity across control, mild, and severe conditions, respectively. In the center, panels (g–i) feature pie charts that illustrate the percentage distribution of different microbial taxa at phylum, class, and order levels within each group. Directly below, panel (k) presents a stacked bar chart that details the relative abundance of microbial taxa across individual subjects in each group. On the right, panels (h–j) focus on the comparative prevalence of key phyla—Actinobacteria, Proteobacteria, Bacteroidota, and Firmicutes

The microbial ecosystem of the scalp is predominantly composed of four major phyla (Fig. 1g-k). A shift in the proportion of Actinobacteriota (from 56.1% in controls to 48.4% in mild AA and 37.9% in severe AA) and Proteobacteria (from 7% in controls to 6.1% in mild AA and 17.4% in severe AA) was observed, indicating a significant alteration in the microbial ecosystem associated with the severity of AA. Together, these phyla encompass a diverse array of 18 different groups, forming a complex and balanced microbial ecosystem essential for maintaining scalp health. There is a lower representation of other bacterial phyla on the scalp such as Nitrospirota, Bdellovibrionota, Desulfobacterota, Verrumicrobiota, Fusobacteriota, Patescibacteria, and Campylobacterota.

The analysis of relative abundance changes across the health gradient from control to severe skin conditions helps identify which microbes are consistently increasing or decreasing as the severity of the condition escalates (Fig. 1j).

Analysis of Microbial Dynamics and Ecological Niches in AA

The correlation analysis reveals that certain types of bacteria tend to coexist (Fig. 2a).

Fig. 2.

Correlation of Top 20 phyla with skin condition severity across control, mild, and severe states. This figure elucidates the relationships between the top 20 microbial phyla and varying severities of skin conditions. The graph uses correlation coefficients ranging from − 1.0 to 1.0 to illustrate these relationships. Panel (a) features a heatmap that visually represents the strength and direction of correlations between pairs of phyla. This detailed visualization helps identify which bacteria are commonly found coexisting within the same samples, illustrating both synergistic and antagonistic interactions among the microbial communities. The heatmap format enhances understanding of complex microbial dynamics by clearly showing positive and negative associations. Panel (b) offers a color-coded gradient reflecting the severity of skin conditions from control to severe. This gradient is divided into regions showing high and low correlations. In this panel, colors closer to red indicate positive correlations, suggesting that the presence of certain phyla increases with the severity of the skin condition. Conversely, colors closer to blue represent negative correlations, where phyla are more prevalent in less severe or control conditions. This visual aid is crucial for quickly pinpointing which microbial taxa are most strongly associated with changes in skin condition severity, providing valuable insights into potential targets for therapeutic interventions or diagnostics. Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001

The data show a significant negative correlation between Actinobacteriota and several other bacterial phyla, including Proteobacteria (ρ = − 0.646, p < 0.001), Firmicutes (ρ = − 0.627, p < 0.001), and Bacteroidota (ρ = − 0.667, p < 0.001). These strong inverse relationships suggest a competitive or antagonistic interaction, possibly indicative of resource competition or differing environmental adaptations within the scalp microbiome.

Conversely, Bacteroidota exhibits a strong positive correlation with a cluster of phyla, including Campylobacteriota (ρ = 0.758, p < 0.001), Desulfobacterota (ρ = 0.831, p < 0.001), Verrucomicrobiota (ρ = 0.816, p < 0.001), and Fibrobacterota (ρ = 0.683, p < 0.001), indicating a co-occurrence that might suggest a ecological niche with potential synergistic relationships. This pattern of positive correlations extends to Cyanobacteria (ρ = 0.631, p < 0.001), which also shows a positive correlation with Patescibacteria (ρ = 0.634, p < 0.001), further supporting the notion of a collaborative network or consortium within this ecological niche.

Figure 2b illustrates the correlations between abundance of microbial phyla and condition severity. Phyla such as Proteobacteria and Firmicutes show positive correlations with increasing condition severity, indicating that their abundance increases as the condition worsens. Conversely, phyla like Actinobacteriota and Synergistota exhibit negative correlations, decreasing in abundance as condition severity increases. This suggests that these microbes are more prevalent in less severe or control conditions, potentially playing a role in maintaining healthier states or being displaced by other microbes in more severe conditions.

Analysis of Bacterial Species Related to the Severity of AA

In most cases, no significant differences were found among the different types of AA. Most of the overrepresented bacterial species on the scalp of patients with severe AA were gram negative, including Alistipes inops, Bacteroides plebeius, Parabacteroides johnsonii, Helicobacter pullorum, and Anaerobiospirillum succiniproducens (Fig. 3). In controls, the overrepresented species included Blautia faecis, Pyramydobacter piscolens, Enterobacteriaceae bacterium, and Anoxybacillus flavithermus.

Fig. 3.

Analysis of bacterial species in the scalp microbiome related to the subtypes of AA vs controls. The figure represents multiple panels of boxplots comparing the log₁₀-transformed count of bacteria species among different subtypes of AA (single patch, multiple patch, AU) and controls

Metabolic Pathway Adaptations across AA Severity Levels

Across both mild and severe AA, there is a consistent reduction in primary energy metabolism pathways like glycolysis, pointing to a universal metabolic challenge imposed by AA on the scalp microbiome (Fig. 4). However, pathways such as glycosphingolipid biosynthesis remain unchanged, signifying these as fundamental microbial functions that are robust to the effects of AA.

Fig. 4.

Comparative analysis of MetaCyc and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways across skin condition severity. This figure illustrates the functional potential of microbial communities in control, mild, and severe skin conditions using MetaCyc and KEGG pathway analyses. Each bar, informed by LEfSe (linear discriminant analysis effect size) analysis, represents the significance and relative activity of pathways, expressed as an LDA score on a logarithmic scale. Panel (a) showcases a bar graph of MetaCyc pathways, highlighting variations in metabolic activity across the three conditions. Panel (b) displays a similar analysis for KEGG pathways, comparing functional capabilities of microbial communities in different skin states. Both panels provide key insights into how metabolic functions shift with skin condition severity, aiding in the identification of potential metabolic markers or therapeutic targets

In mild AA, the microbiota shifts towards nucleotide-related metabolic pathways, such as purine metabolism and nucleotide sugar biosynthesis, enhancing nucleotide turnover and DNA repair to maintain stability under less severe disease conditions. Enhanced UTP and CTP dephosphorylation activities suggest adaptations to manage energy more efficiently and respond to mild disease stress. Additionally, pathways involved in managing immune responses and breaking down aromatic compounds are uniquely activated, reflecting distinct microbial adaptations to milder disease manifestations.

In severe AA, the microbiota primarily focuses on lipid metabolism, with significant increases in secondary bile acid and fatty acid biosynthesis. These pathways likely contribute to the inflammatory nature of severe AA by modulating inflammatory responses. Additionally, there is a pronounced emphasis on amino acid metabolism to potentially alleviate stress and inflammation. The condition is further characterized by disrupted cholesterol metabolism and altered responses to xenobiotics, marking these metabolic shifts as potential indicators of disease progression.

Interplay between Scalp Microbiota and Systemic Inflammatory Status in AA

The scalp microbiota significantly influences systemic inflammatory states in AA, as depicted in Fig. 5a. This interaction is primarily governed by a dynamic equilibrium between pro-inflammatory and anti-inflammatory cytokines, which not only mitigates tissue damage but could also slow AA's progression.

Fig. 5.

Comparative analysis of scalp microbial communities and inflammatory markers across skin conditions. This figure illustrates the interplay between microbial communities and circulating inflammatory markers in varying skin conditions through two analytical approaches: Panel (a) features a clustered heatmap that illustrates the correlation coefficients between the abundance of bacterial phyla and concentrations of plasma inflammatory proteins. Warm colors (e.g., reds) indicate strong positive correlations, while cool colors (e.g., blues) suggest negative correlations. This arrangement offers a clear visual representation of how microbial populations are associated with inflammation markers. Panel (b) presents a principal component analysis (PCA) based on bacterial abundance and plasma protein concentrations related to inflammation and disease severity. Each point represents a bacterial phylum plotted against the first two principal components, reflecting major variations in protein profiles. The proximity of points indicates similarity, with color gradations from blue (negative correlation) to red (positive correlation) showing their relationship to disease severity. Together, these panels provide a succinct overview of microbial dynamics and their correlation with inflammatory responses in different skin conditions

Actinobacteriota's abundance correlates positively with higher levels of anti-inflammatory cytokines (IL-10, IL-33, IL-1 alpha, IL-13, TGF-alpha) and negatively with pro-inflammatory cytokines (CCL23, CCL19, TRANCE, TSLP). Similarly, Patescibacteria shows associations with both essential cytokines and growth factors (e.g., FGF-19, HGF) and specific chemokines and inflammation mediators (e.g., CXCL9, CXCL11), suggesting a role in modulating immune responses.

Conversely, phyla like Acidobacteriota and Gemmatimonadota are linked to increased pro-inflammatory cytokines (IL-8, IL-6) and a decrease in anti-inflammatory markers, indicating potential exacerbation of inflammation. In contrast, Bacteroidota, Campylobacteriota, Cyanobacteria, and Firmicutes exhibit a mixed cytokine response, reducing certain pro-inflammatory markers (e.g., IL-17C, IFN-gamma), which may suggest a nuanced regulatory role in inflammation.

PCA further identifies the inflammatory nature of these bacterial groups (Fig. 5b). Proteobacteriota, associated with high levels of pro-inflammatory proteins (IL-8, IL-6, IL-17A/C, IFN_ɣ, TNF_β), indicates a severe inflammatory profile, while Actinobacteriota shows an inverse relationship, prevalent in controls and linked to higher anti-inflammatory and lower pro-inflammatory protein levels.

Overall, this comprehensive analysis underscores the complex and varied impacts of different bacterial phyla on the inflammatory landscape of AA, highlighting potential targets for therapeutic intervention based on microbial influence on systemic inflammation.

Discussion

Our study sheds light on the intricate relationship between the scalp microbiome, their metabolic functions, and systemic inflammation in adults with AA, highlighting a consistent microbial diversity across varying disease severities and discernible differences in microbial compositions compared to healthy individuals.

Notably, we observed significant variability in the abundance of specific microbial phyla, such as Actinobacteriaceae and Proteobacteriaceae, which exhibited correlations with disease severity, suggesting a pivotal role for the microbiota in both the onset and progression of AA. Furthermore, our findings indicate that the scalp microbiota in patients with AA are actively engaged in the metabolism of environmental substances and xenobiotics and in pathways crucial to immune responses and energy metabolism, distinguishing them from those of healthy controls. Moreover, the correlation between microbial composition and circulating inflammatory markers suggests a bidirectional relationship wherein the microbiota may significantly influence, or be influenced by, systemic inflammatory responses.

Moving forward, these insights pave the way for future investigations aimed at elucidating the mechanistic underpinnings of the scalp microbiota in AA pathogenesis and exploring novel therapeutic strategies targeting microbial dysbiosis to mitigate disease progression.

There have been various studies on the scalp microbiota in patients with AA, but the findings have been inconclusive. For example, Won et al. explored severe cases of AA and noted a significant increase in bacteria from the Actinobacteria group, such as Corynebacterium and Cutibacterium, along with a decrease in Staphylococcus caprae from the Firmicutes phylum [45]. In contrast, Juhasz et al. observed an increase in Clostridia, another class within the Firmicutes, in the scalps of patients with AA [46]. This finding appears to contradict the decrease in other members of the Firmicutes reported by Rinaldi et al., suggesting that AA's impact on Firmicutes subpopulations may be more complex and species-specific [47]. Moreover, Pinto et al. provided additional details, indicating that there was an increase in Neisseria (Proteobacteria) and Anaerococcus (Firmicutes) in the epidermis of patients with AA, while in the dermis, there was a decrease in Candidatus aquiluna (Actinobacteria) and Staphylococcus epidermidis (Firmicutes), with the genus SMB53 (family Clostridiaceae, Firmicutes) being completely absent [48]. Collectively, these findings reinforce our idea that there is a microbial imbalance in the pro- vs anti-inflammatory bacterial communities associated with the severity of AA rather than a predominance of one or two specific species involved.

On the other hand, it has recently been confirmed that there is a systemic inflammatory state in AA, and many of the markers of tissue inflammation are also dysregulated in plasma, albeit to a lesser extent than in the tissue. In this context, Glickman et al. found that in the scalp affected by AA, 608 inflammatory genes were differentially expressed [49]. This included genes associated with Th1 (IFNG/IL12B/CXCL11), Th2 (IL13/CCL18), and T-cell activation (ICOS). These plasma markers were significantly correlated with the clinical severity of AA in both lesional and non-lesional tissues. In our study, we also observed that the pattern of bacterial imbalance not only reflected the differences in state (disease vs. healthy) and severity of the subjects but also mirrored the pattern of systemic inflammation in the subjects.

To clarify whether scalp follicle dysbiosis initiates the immunological failure resulting in inflammatory hair disorders or if it results from a preexisting inflammatory, endocrine, or metabolic disturbance, Pinto et al.'s study stands out as it delves into functional profiles of the scalp microbiome through PICRUSt and KEGG analyses, revealing enriched profiles connected to environmental information processing and cellular antigens in patients with AA [50]. The enriched profiles related to environmental information processing might suggest that the scalp microbiome in patients with AA is adapted to respond to external stimuli more robustly or differently than in healthy individuals. This adaptation could trigger an immune response that targets the hair follicles, leading to the characteristic hair loss seen in AA. The balance of cooperation and competition among bacterial communities is essential for understanding the interactions of the microbiome and its relationship with skin health and the progression of AA.

Our study stands out for its high-quality sequencing data across multiple hypervariable regions of the 16S rRNA gene, providing a more comprehensive overview of the scalp microbiome in AA compared to previous studies that analyzed fewer regions [45, 46, 48]. This method, utilizing ion torrent sequencing, enhances species identification, phylogenetic analysis, and understanding of microbial community composition, significantly improving the precision of our results [47]. Unlike prior research focusing on severe AA, our study expands to include mild cases, offering a broader understanding of the microbiological variations across different disease severities and filling a crucial gap in existing literature [51]. This approach ensures robust representation of microbial diversity and strengthens the reliability of our findings, contributing to a holistic understanding of the microbiota's role in AA [52].

Limitations

However, our study does face several limitations. One is the difficulty in finding appropriate controls matched by age and sex. This challenge arises from the fact that it is socially uncommon for women to have the shaved heads necessary for obtaining scalp microbiota samples. Therefore, we opted to use patients scheduled for neurosurgery as controls, as their heads were shaved as part of the surgical preparation, allowing us to collect the required samples.

Our analysis identified and addressed outliers that could affect the interpretation of miRNA expression profiles. These outliers might arise from difficulties in matching controls or variations in clinical characteristics that influence microbiome distribution. While peripheral blood inflammation markers are relevant for assessing systemic inflammation in severe AA, they may not fully represent local scalp inflammation. More precise data could be obtained using non-invasive methods like tape strips or invasive procedures such as skin biopsies. Additionally, swab sampling captures microbiota only from the upper stratum corneum, limiting insights into deeper skin follicle areas.

Conclusions

In conclusion, while current evidence indicates a differential microbial profile in the AA-affected scalp, the functional consequences of these alterations, their interplay with host immune responses, and their impact on therapy responsiveness remain to be fully elucidated. The interplay among the skin microbiota, immune responses, and onset of skin conditions underscores the potential for microbiome-targeted interventions to manage these conditions.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

P.J. Gómez-Arias, J. Gay-Mimbrera, D. Gómez-Garre, B. Isla-Tejera, and J. Ruano contributed to the conceptualization of the study. Data curation was carried out by J. Gay-Mimbrera. Formal analysis was performed by S. Sánchez-González, A. Ortega-Hernández, and D. Gómez-Garre. Funding for the study was acquired by J. Ruano. F. Gómez-García conducted the investigation. D. Gómez-Garre and J. Ruano developed the methodology. Project administration was managed by J. Gay-Mimbrera and F. Gómez-García. Resources were provided by J. Gay-Mimbrera, I. Rivera-Ruiz, E. Parra-Peralbo, and M. Juan-Cencerrado. S. Sánchez-González, A. Ortega-Hernández, and J. Ruano handled the software. F. Gómez-García and B. Isla-Tejera supervised the study. M. Aguilar-Luque validated the findings. S. Sánchez-González, A. Ortega-Hernández, and J. Ruano were responsible for visualization. The original draft was written by P.J. Gómez-Arias, E. Parra-Peralbo, J. Gay-Mimbrera, and J. Ruano. P.J. Gómez-Arias, S. Sánchez-González, A. Ortega-Hernández, E. Parra-Peralbo, B. Isla-Tejera, and J. Ruano contributed to the review and editing of the manuscript.

Funding

This research has been exclusively supported by public foundations, with no involvement of private funds. This work received partial support from Project PI23/01590 (awarded to JR), funded by Instituto de Salud Carlos III (ISCIII) and co-funded by the European Union. Additionally, support was provided by the Plan Propio de Investigación del Instituto Maimónides de Investigación Clínica de Córdoba (IMIBIC), awarded to JG-M and MA-L, and the International Eczema Council through their 2002 Research Fellowship Program, which funded PG-A. EP-P acknowledges support from Universidad Europea de Madrid for conducting this research.The foundation did not influence the study design, data collection and analysis, publication decision, or manuscript preparation. It is important to note that no funding was received from any pharmaceutical company.

Data Availability

The datasets generated and/or analyzed during this study are provided in the supplementary information files accompanying this article. Additionally, raw data, processed data, and metadata have been deposited in the Gene Expression Omnibus (GEO) repository at the National Center for Biotechnology Information (NCBI) under accession number BioProject: PRJNA1115970, accessible at https://www.ncbi.nlm.nih.gov/geo/. All data and code will be made available to researchers upon reasonable request, with appropriate justification for further exploration, replication, or collaboration.

Declarations

Conflict of Interest

Juan Ruano is a member of the Editorial Board of Dermatology and Therapy but was not involved in the peer review or editorial decisions for this manuscript. All other (Pedro J. Gómez-Arias, Jesús Gay-Mimbrera, Irene Rivera-Ruiz, Macarena Aguilar-Luque, Miguel Juan-Cencerrado, Carmen Mochón-Jiménez, Francisco Gómez-García, Silvia Sánchez-González, Adriana Ortega-Hernández, Dulcenombre Gómez-Garre, Esmeralda Parra-Peralbo and Beatriz Isla-Tejera) co-authors declare no conflicts of interest.

Ethical Approval

Reviewed and approved by Provincial Research Ethics Committee of Córdoba, Spain; approval # HRP-503, August 2015. The legislation in effect in 2015 has not changed in a way that would impact the validity of the approval, and the original approved protocol remains relevant to the current study. No changes have been made to the study, such as modifications in methodology, inclusion of new participants, or alterations in study conditions, that would invalidate the initial approval. The research was conducted following the standards of good clinical practice and adhering to the principles outlined in the by the Helsinki Declaration of 1975, as revised in 1983, and Belmont Report and adhered to Spanish and European laws regarding data privacy. Before participation, all patients provided informed consent after receiving and reviewing the patient information sheet.