The protocol for this analysis was registered in PROSPERO (International Prospective Register of Systematic Reviews; CRD42024564852, https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=564852). The systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines ( Supplementary Document ) (3). Electronic searches were conducted from the database’s inception to July 2024 using the PubMed, Embase, and Scopus databases. The search strategy was designed to retrieve all studies on PCA using keywords and a controlled vocabulary. There were no restrictions on the language or publication period in the searches. Only full-text articles were considered, and conference abstracts were excluded due to limited detail and lack of peer review. The search included a combination of terms: ‘primary scarring alopecia’, ‘primary cicatricial alopecia’, ‘lichen planopilaris’, ‘frontal fibrosing alopecia’, ‘Graham Little syndrome’, ‘fibrosing alopecia in a pattern distribution’, ‘pseudopelade of Brocq’, ‘central centrifugal cicatricial alopecia’, ‘discoid lupus erythematosus’, ‘alopecia mucinosa’, ‘keratosis follicularis spinulosa decalvans’, ‘folliculitis decalvans’, ‘dissecting cellulitis’, ‘acne keloidalis’, and ‘erosive pustular dermatosis.’ Supplementary Table 1 provides details about the search strategy. Grey literature and unpublished data were not considered. If multiple studies utilized patient data from the same database with an overlapping inclusion period, data from the more recent publication or from a case-control study is selected. The comorbid data were not quantitatively utilized if they were reported in a composite manner, such as LPP/FFA or PCA.

Two reviewers (TY and KT) independently evaluated each article at both the full-text and title/abstract levels. Full texts of potentially eligible studies were assessed for inclusion. Disagreements regarding the studies’ eligibility were resolved via discussion with a third reviewer (PS). Inclusion criteria were: (1) randomized controlled trials, cohort studies, cross-sectional studies, and case-control studies; (2) studies involving patients with confirmed diagnosis of PCA based on clinical and/or histopathological criteria; (3) studies reporting data allowing calculation of odds ratios (OR) or prevalences of specific comorbidities; (4) full-text articles in any language. Exclusion criteria were: (1) conference abstracts; (2) case reports or case series; (3) studies without clear diagnostic criteria for PCA; (4) studies reporting only composite PCA data without subtype specification. Comorbidities were defined based on clinical diagnosis using ICD codes, physician diagnosis, or validated diagnostic criteria as reported in each study.

Data were extracted from the included studies using a standardized form. The following data were collected: bibliographic data (authors, year of publication), study characteristics (type of study, single or multicenter, study duration, country), alopecia group characteristics (number, age, gender, ethnicity, body mass index (BMI), comorbidity), and treatment information), control group characteristics (number, age, gender, ethnicity, BMI, comorbidity, whether controls were matched for any relevant factors), and comorbidity data.

Corresponding investigators were contacted via email if there was missing data. Two independent reviewers (TY and KT) extracted data, and any discrepancies were discussed and resolved with input from a third reviewer (PS).

TY and KT independently assessed the quality of descriptive and case-control studies using the Newcastle-Ottawa Scale (NOS) (4). The NOS is a scoring tool comprised of seven items with nine scores that assess how well the investigators selected their participants (score ranges from 0 to 4), the comparability of their results (score ranges from 0 to 2), and the applicability of the outcomes (score ranges from 0 to 3). The higher the score, the higher the study’s quality and the lower the likelihood of bias. Therefore, we classified studies as having high quality if they received a total score of 7 or more, fair quality if they received a score of 4-6, and low quality if they received a score of less than 4. Sensitivity analyses were performed excluding fair-quality studies to assess the robustness of our findings. Meta-analyses were weighted by study quality, with higher-quality studies given greater influence in pooled estimates. Any discrepancies between reviewers regarding the risk of bias in specific studies were resolved through discussion with a third reviewer (PS).

A meta-analysis was performed, using an inverse variance method, to pool the ORs of a specific PCA and a specific comorbidity, as well as the prevalences of various comorbidities associated with a specific PCA. Each PCA disorder was analyzed separately. Quantitative analyses were conducted on comorbidity, either OR or prevalence, that were reported in at least two studies; otherwise, they were only analyzed qualitatively.

Heterogeneity was assessed and considered present if a Cochrane Q test p-value was < 0.1 or Higgins I² ≥ 25% (5). The sources of heterogeneity were explored by fitting each covariate (e.g., age, female gender, and BMI) at a time in a meta-regression model. If the τ² was decreased by ≥ 50% or statistically significant β was revealed, a subgroup analysis was performed based on that covariate if possible (6).

To evaluate publication bias, Deeks funnel plots of the primary outcomes were generated. The Egger linear regression test was applied when a funnel plot suggested possible asymmetry (7). If Egger’s test for a regression intercept gave a p-value < 0.05, a trim and fill method was used to adjust the OR (7). Comprehensive Meta-Analysis software (version 3.3.070, Biostat, Englewood, NJ) was used for all statistical analysis.

After removing duplicates, 4,797 references were screened by title and abstract. At the full-text stage, 308 full articles met our predefined selection criteria and were sought. We further excluded 192 references for the following reasons: conference abstract or review article (n = 60), not population of interest (i.e., non-alopecia or non-scarring alopecia diagnosis, n = 15), not outcome of interest (i.e., no documented comorbidity prevalence of the patients, n = 115), and duplicate patient data (n = 2) ( Figure 1 ). The review included 116 studies, enrolling a total of 33,494 patients with PCA [8,871 LPP patients (8–31), 6,595 FFA patients (12, 16, 19, 21, 23, 25, 27, 30, 32–83), 3,539 CCCA patients (24, 84–99), 157 PPB patients (20), 24 patients with fibrosing alopecia in a pattern distribution (FAPD) (55), 6,158 FD patients (20, 100–104), 4,752 DC patients (20, 105–108), 3,218 AKN patients (109–120), and 180 EPD patients (121–124)] between 2009 and 2024, were included in the review. Characteristic features of the included studies are provided in Supplementary Tables 2 - 6 .

Among 24 FAPD patients, we identified AA [prevalence 12.5%], celiac disease [prevalence 8.3%], rheumatoid arthritis [prevalence 8.3%], frontal fibrosing alopecia [prevalence 8.3%], pernicious anemia [prevalence 4.2%], scalp psoriasis [prevalence 4.2%], and traction alopecia [prevalence 4.2%]. For AKN, common comorbidities include acne [prevalence 29%], acne conglobata [prevalence 14%], FD [prevalence 21%], and pseudofolliculitis barbae [prevalence 26%]. Acne conglobata [prevalence 24%] is also prevalent in DC, and a higher prevalence of hidradenitis suppurativa was found for DC [prevalence 26%], compared to FD [prevalence 4%] and AKN [prevalence 3%]. And for EPD, patients often report actinic keratosis [prevalence 21%], basal cell carcinoma [prevalence 7%], and squamous cell carcinoma [prevalence 13%] as their comorbidities. Supplementary Figures 4 - 7 show forest plots for the pooled prevalence of the comorbidities in patients with FD, DC, AKN, and EPD, respectively.

Supplementary Table 7 provides a summary of the quality assessment scores for comparative and descriptive studies included in the review. The average quality assessment score was 7.78 (range: 4–9), with 101 high-quality and 15 fair-quality studies. No studies were classified as low-quality. The most common quality concerns in fair-quality studies were related to the comparability of cohorts and adequacy of follow-up. Sensitivity analyses excluding the 15 fair-quality studies did not substantially alter our main findings, with all significant associations remaining statistically significant.

Meta-regression was performed for all analyses that were found to have significant heterogeneity and did not find any potential source of heterogeneity, except for one analysis in which we found that female proportion seems to play a role in the odds of depression in the LPP cohorts ( Supplementary Figure 8 ). However, we chose not to perform a subgroup analysis due to the limited number of studies that included a narrow range of female proportions, specifically from 72.4 to 91.5%.

Statistically significantly higher odds of SLE were found among those with lymphocytic PCA, particularly LPP, FFA, and CCCA, compared with controls. Peroxisome proliferator-activated receptor-γ (PPAR-γ), a nuclear hormone receptor, is involved in the regulation of inflammation and tissue repair (126). It has been proposed to be the link between LPP and SLE (12), as PPAR-γ has been implicated in the pathogenesis of both PCA and SLE (126, 127). Specifically, PPAR-γ expression is altered in both conditions, potentially contributing to dysregulated immune responses and tissue damages (126, 128–130). PPAR-γ is also involved in the pathogenesis of rheumatoid arthritis (131), which we found to have a significant association with FFA. Therefore, further studies exploring the link between PCA and various autoimmune diseases are needed.

The association between PCA and multiple autoimmune conditions aligns with the concept of Multiple Autoimmune Syndrome (MAS), where at least three autoimmune diseases coexist in the same patient. Tatu and Ionescu described a case of MAS type 3, characterized by the coexistence of autoimmune thyroiditis, vitiligo, and AA (132). Although their report focused on non-scarring alopecia, the principle of shared autoimmune pathophysiology may extend to primary cicatricial alopecias, suggesting that PCA could represent a cutaneous manifestation within a broader autoimmune context.

Similar to a previous meta-analysis, we also found a statistically significant association between LPP and hypothyroidism (133). In addition, we also found FFA to be associated with hypothyroidism, which can be observed in nearly one fifth of patients with either LPP or FFA. The association between thyroid disorders, especially hypothyroidism, and PCA could be explained by the prevalent immune-mediated hypothyroidism, such as Hashimoto thyroiditis. We believe hypothyroidism should be considered when evaluating LPP or FFA patients, given the well-established link between thyroid diseases and autoimmune skin diseases (134). This association underscores the importance of routine thyroid function screening in patients diagnosed with these forms of PCA, especially in cases with clinical suspicion. However, the need for thyroid investigation in asymptomatic patients remains controversial.

Compared to previous meta-analysis, we found a similar association and prevalence of rosacea in FFA (135). We also find that rosacea is strongly associated with LPP, although the prevalence is lower than in FFA. Apart from rosacea, LPP is also strongly linked to lichen planus and other immune-mediated skin diseases, including atopic dermatitis and psoriasis.

Although the evidence is limited to two studies, allergic contact dermatitis is also significantly associated with FFA. Given that FFA association with skin care products has been shown previously (136), it may be sensible to inform patients of the risk.

In contrast to other autoimmune skin diseases such as vitiligo and AA (137–140), we found a higher risk of skin cancer and melanoma among LPP patients. Actinic keratosis and non-melanocytic skin cancer are also found to be prevalent among EPD. Given the increased risk of skin cancer and dermatologic diseases, we believe a detailed full-body dermatological examination for patients with PCA is necessary.

We have confirmed the findings of a previous meta-analysis that dyslipidemia is not statistically associated with LPP (141). Additionally, we also did not identify any association between LPP and other metabolic diseases, including diabetes mellitus, hypertension, and obesity. However, we did find an association between FFA and dyslipidemia, while CCCA is associated with both dyslipidemia and diabetes mellitus. These associations suggest a potential role of metabolic inflammation in their pathogenesis (126). Further studies exploring the association between PCA and metabolic disorders are needed.

Kim et al. have previously found an association between PCA and incident cardiovascular disease, particularly coronary heart disease (125), which is also evident in our LPP findings. Certain interleukins (IL), especially IL-18 (142), are associated with cardiovascular risk, which makes autoimmune patients more prone to coronary diseases. We believe this issue requires further investigation and physicians should be on high alert for coronary events.

Similar to a previous meta-analysis (143), evidence is limited for vitamin D deficiency and PCA, and evidence so far suggests that vitamin D deficiency is not statistically associated with LPP. In addition, the lack of statistical association between uterine leiomyoma and CCCA is noted in our study, while strong associations between allergic rhinitis, anxiety, and CCCA were found. We suspect that the high prevalence of uterine leiomyoma may not be directly related to the disease. Regular psychosocial assessment and support is crucial for patients with CCCA as well as other PCAs. Further studies exploring the relationship between allergic rhinitis and CCCA are needed.

Recent histopathological studies have identified variations in inflammatory patterns across PCA subtypes that may explain their different comorbidity profiles. The predominantly CD8+ T-cell-mediated inflammation in LPP and FFA contrasts with the CD4+ T-cell-predominant mixed inflammatory infiltrate seen in CCCA, potentially explaining their distinct associations with autoimmune and metabolic conditions, respectively (144–146).

The association between PCA and various systemic conditions reflects complex shared pathogenic mechanisms involving inflammatory and immune-mediated pathways. The histopathological hallmark of lymphocytic PCA, destruction of follicular stem cells in the bulge region with perifollicular fibrosis, mirrors inflammatory patterns seen in associated autoimmune conditions. This parallel extends to the molecular level, where PPAR-γ signaling emerges as a crucial link between PCA and systemic disorders.

In LPP and FFA, decreased PPAR-γ expression in hair follicle stem cells triggers proinflammatory cytokine production and subsequent stem cell loss. This mechanism shows striking similarity to PPAR-γ dysfunction in SLE, where impaired PPAR-γ signaling contributes to immune dysregulation and tissue damage (126). The strong association between these conditions likely reflects this shared pathogenic pathway. For CCCA, the significant association with metabolic disorders may similarly reflect common inflammatory mechanisms. PPAR-γ’s role in regulating both immune responses and metabolic homeostasis provides a potential molecular explanation for these associations (147). The presence of metabolic inflammation in conditions such as diabetes and dyslipidemia may exacerbate the follicular inflammation characteristic of CCCA.

FAPD has been recently recognized as a subtype of LPP that presents with a pattern mimicking androgenetic alopecia (148). Histologically, FAPD exhibits a lymphocytic scarring process similar to classic LPP, although detailed characterization of the lymphocyte subtypes in its inflammatory infiltrate remains lacking. While its recognition has expanded our understanding of how scarring alopecia can clinically manifest, data regarding its associated systemic conditions remain limited.

The magnitude of associations found in this meta-analysis has important clinical implications. A 3-fold increased risk of SLE in LPP patients translates to an absolute risk of approximately 2%, compared to 0.65% in the general population. While this remains a relatively low absolute risk, it warrants clinical vigilance given SLE’s potential severity. More practically significant is the 17-19% prevalence of hypothyroidism in LPP and FFA, a frequency that justifies routine screening.

The 8% prevalence of coronary artery disease in LPP patients (OR 1.63) represents a clinically meaningful increase that should prompt cardiovascular risk assessment and aggressive management of modifiable risk factors. Similarly, the high prevalence of dyslipidemia in CCCA (54%) and diabetes mellitus (26%) necessitates proactive metabolic screening and intervention in this population.

Based on our findings, we recommend a comprehensive screening approach for PCA patients. Initial assessment should include thyroid function tests and thyroid antibody screening for all PCA patients. Autoimmune screening with antinuclear antibody testing should be considered in patients presenting with systemic symptoms. Metabolic screening including lipid profile and fasting glucose is particularly important in FFA and CCCA patients. Annual full-body skin examinations are warranted in LPP patients due to their increased skin cancer risk.

The complexity of comorbidities in PCA necessitates multidisciplinary care coordination. Establishing collaborative relationships with endocrinology for thyroid management, rheumatology for suspected autoimmune conditions, and maintaining regular dermatologic surveillance are essential components of comprehensive care. Mental health support should be integrated into the management plan, given the significant psychological burden associated with PCA. This coordinated approach ensures that the multiple systemic manifestations of PCA are adequately addressed, ultimately improving patient outcomes and quality of life.

This study has some limitations. Firstly, some comorbidities were infrequently reported and thus were not able to be quantitatively analyzed. Additionally, there is an inadequate number of cohort studies that document incident cases of various comorbidities for quantitative analysis. Furthermore, significant statistical heterogeneity was found in some of our analyses, and despite meta-regression analysis, we were unable to identify the source of heterogeneity, except for one analysis. This is partly due to the limited number of available studies. Potential sources of heterogeneity include variations in diagnostic criteria for PCA subtypes, differences in study populations (such as ethnicity, age distribution, geographic location, disease severity, and comorbidity burden), methodological differences across studies, differences in healthcare systems and access to specialist care, and temporal changes in awareness and screening practices for both PCA and associated conditions. These factors highlight the need for cautious interpretation of pooled estimates for these specific associations. While we employed appropriate statistical methods, the adjusted results may still be influenced by unreported negative findings.