Efficacy of herbal medicine Xiao-Feng-San combined with auricular acupuncture for atopic dermatitis: A randomized controlled trial

Abstract

Background

Atopic dermatitis (AD) is a common chronic inflammatory skin disease with a considerable burden. Xiao-Feng-San (XFS) has been widely used in traditional medicine for the wind-dampness-heat pattern. Auricular acupuncture (AA) may relieve AD symptoms, but its added benefit with XFS is unclear.

Methods

In this randomized, sham-controlled clinical trial, 156 adults with non-severe AD and wind-dampness-heat pattern, all of whom were prescribed XFS decoction, were additionally assigned to receive either AA (XFS plus AA group, n=78) or sham AA (XFS plus Sham AA group, n=78) for four weeks. The primary outcome was the change in the SCORAD index from baseline to week 4. Secondary outcomes included fexofenadine use, Dermatology Life Quality Index (DLQI), serum total IgE, and adverse events.

Results

Compared with the XFS plus sham AA, the XFS plus AA group showed a significantly greater reduction in SCORAD at week 4 (mean difference [MD] −9.1; p < 0.0001), exceeding the minimal clinically important difference. DLQI also improved more in the XFS plus AA group (MD −1.8; p = 0.0362). Fexofenadine use was lower in this group, though not significantly; no significant differences in serum total IgE were found. Adverse events were mild and transient.

Conclusions

Adding AA to XFS enhanced clinical outcomes and quality of life in patients with non-severe AD, with a favorable safety profile. These findings support AA as a promising adjunctive therapy in integrative approaches for AD. Further studies should validate these results across broader populations, with post-intervention follow-up, and diverse traditional medicine regimens.

Trial registration information

ClinicalTrials.gov (NCT06492902).

1. Introduction

Atopic dermatitis (AD), also known as eczema or atopic eczema, is the most prevalent chronic inflammatory skin condition in adults, affecting around 5 % of this population.¹ It is associated with significant physical discomfort, psychological stress, and socioeconomic impact. AD is commonly characterized by persistent itching, erythematous and scaly plaques, crusting, fissures, lichenification, and in some cases, oozing or exudation, particularly during acute flares.^2,3

Treatment of AD focuses on restoring the skin barrier, reducing inflammation, and preventing scratching. Basic care includes moisturizers, bath oils, and allergen avoidance. Common topical therapies are corticosteroids, calcineurin inhibitors, UV therapy, and wet wraps.^3,4 If topical treatments are insufficient, systemic immunosuppressants such as corticosteroids, cyclosporine A, methotrexate, mycophenolate mofetil, or azathioprine may be used.⁵ However, long-term use can cause side effects like skin atrophy, purpura, acne, and adrenal suppression.^5,6

In traditional medicine (TM), AD is commonly attributed to the accumulation of wind, dampness, and heat, which are believed to underlie its clinical features. The most frequent syndrome patterns include damp-heat, spleen deficiency with dampness accumulation, spleen deficiency and blood deficiency, and wind-dryness. Accordingly, treatment typically follows the principles of clearing heat, drying dampness, and dispersing wind.⁷ Based on these principles, Xiao-Feng-San (XFS) has been widely used as a core herbal prescription for managing AD with wind-dampness-heat patterns.⁸ Nevertheless, XFS monotherapy may provide incomplete control, and robust long-term safety data remain limited.

A nonpharmacologic adjunct may address these gaps. In addition, acupuncture has shown promising effects in alleviating both subjective and objective symptoms of AD, while improving quality of life.⁹ Among acupuncture modalities, auricular acupuncture (AA) offers practical advantages. It enables prolonged stimulation by fixing press needles or seeds on specific ear points for up to one week. This minimally invasive and easy-to-apply technique is well-tolerated and suitable for clinical use. Our previous trial demonstrated that AA was more effective than sham AA in reducing AD severity and improving quality of life, with a favorable safety profile.¹⁰

Given its potential, AA may serve as an effective adjunct to traditional herbal therapies. However, evidence on combining AA with herbal medicine remains limited. This study aimed to evaluate whether adding AA to XFS decoction, a standard treatment for AD with wind-dampness-heat patterns, could enhance treatment outcomes compared to XFS alone.

2. Methods

2.1. Study protocol

The protocol was registered at ClinicalTrials.gov (NCT06492902). The study adhered to the Declaration of Helsinki and International Conference on Harmonisation - Good Clinical Practice. The protocol was approved by the Ethics Committee of the University of Medicine and Pharmacy at Ho Chi Minh City (No. 270/HĐĐĐ-ĐHYD), and all participants provided written informed consent. The study design followed the Consolidated Standards of Reporting Trials (CONSORT) and the Standards for Reporting Interventions in Clinical Trials of Acupuncture (STRICTA).¹¹

2.2. Study design

This study was a participant- and assessor-blinded, randomized, sham-controlled clinical trial with two parallel groups in a 1:1 allocation ratio. It was conducted at two centers, including the University Medical Center Ho Chi Minh City - Campus 3 and the Traditional Medicine Hospital of Ho Chi Minh City, from July 15, 2024, to March 23, 2025. The aim was to evaluate the efficacy of adding AA to the XFS decoction regimen in patients with non-severe AD, comparing an XFS plus AA group with an XFS plus Sham AA group.

2.3. Participants

Patients diagnosed with AD according to the American Academy of Dermatology guidelines, classified as having mild to moderate disease (SCORAD ≤ 50), who were also diagnosed by TM practitioners with wind-dampness-heat patterns and prescribed XFS decoction, were included. Exclusion criteria included the presence of other active skin diseases or infections requiring systemic treatment; recent use of systemic therapies, biologics, phototherapy, or herbal medicine; prior experience with acupuncture; auricular lesions; or concurrent participation in other clinical trials. Women who were pregnant, breastfeeding, or planning pregnancy during the study period were also excluded. More details were presented in Supplement 1.

2.4. Randomization and blinding

Eligible participants were assigned to one of two groups in a 1:1 ratio using block randomization with a block size of four. A total of 39 blocks were generated, and an independent staff member ensured allocation concealment. Group assignments were determined using sequentially numbered, sealed envelopes containing randomization codes, which were opened in the order of participant enrollment.

Blinding was maintained through the use of sham acupuncture, ensuring that participants were unaware of their group assignments. Although the practitioners performing AA were aware of treatment allocation, they were not involved in outcome assessment or data analysis. All outcome assessors and data analysts remained blinded to group assignments throughout the study.

2.5. Interventions

Participants were prescribed the standard XFS decoction regimen of the respective hospital for four weeks and were additionally assigned to receive either AA or sham AA as a parallel adjunctive intervention. The XFS decoction protocol at each hospital was based on the Ministry of Health guidelines. The formula consisted of 12 herbs: Herba Schizonepetae (Jingjie) 12 g, Radix Saposhnikoviae (Fangfeng) 12 g, Radix Angelicae Sinensis (Danggui) 12 g, Radix Rehmanniae (Shengdi) 12 g, Radix Sophorae Flavescentis (Kushen) 12 g, Rhizoma Atractylodis (Cangzhu) 8 g, Periostracum Cicadae (Chantui) 4 g, Fructus Arctii (Niubangzi) 12 g, Rhizoma Anemarrhenae (Zhimu) 10 g, Gypsum Fibrosum (Shigao) 12 g, Rhizoma Alismatis (Zexie) 6 g, and Radix Glycyrrhizae (Gancao) 6 g. Those herbs were decocted and packaged into two 90 mL pouches. Participants took one pouch per dose, twice daily.

AA or sham AA was administered once per week, for a total of four sessions over the four-week intervention period. Two TM practitioners with over five years of AA experience performed all procedures. Both received stimulation at the same four auricular points: Lung (CO14), Shenmen (TF4), Endocrine (CO18), and Adrenal gland (TG2) (Supplement 2A). Two types of 10 mm square patches were used, one containing a sterilized needle (0.2  ×  2  ×  1.5 mm) for AA, inserted to a depth of 1.5 mm, and the other without a needle for sham AA (Supplement 2B). The patch remained in place for one week, with alternating ears used weekly. Participants were instructed to apply pressure to acupoints for approximately one minute, three times daily.

In cases of intolerable pruritus, participants were allowed to take fexofenadine 60 mg as a rescue therapy, up to three times per day, with the number of tablets recorded. No other treatments were permitted during the study period.

2.6. Outcomes measures

The primary outcome was the between-group difference in the mean change in SCORAD index from baseline (week 0) to week 4. The SCORAD index was a validated clinical tool used to assess AD severity, integrating both objective and subjective parameters. It was calculated using the formula: SCORAD = A/5 + 7B/2 + C, where A represented the extent of skin lesions (0-100), B was the sum of six objective signs (erythema, edema, lichenification, excoriation, oozing/crusting, and xerosis), each scored from 0 to 3 (total 0-18), and C included two patient-reported symptoms (pruritus and sleep disturbance), each rated on a 10-point visual analog scale (total 0-20). The overall SCORAD score ranged from 0 to 103, with higher scores indicating more severe disease.¹² The minimal clinically important difference (MCID) had been established at 8.7 points.¹³

Secondary outcomes included antihistamine consumption (fexofenadine 60 mg), quality of life, and serum total immunoglobulin E (IgE) levels. The use of fexofenadine was recorded based on participant self-report. Quality of life was assessed using the Dermatology Life Quality Index (DLQI), a patient-reported questionnaire designed to evaluate the impact of skin disease over the previous week. It included 10 items across six domains (symptoms and feelings, daily activities, leisure, work and school, personal relationships, and treatment), each scored from 0 to 3, with a total score ranging from 0 to 30. Higher scores reflected greater impairment in quality of life.¹⁴

SCORAD and DLQI were assessed at baseline, week 1, week 2, week 3, and week 4; weekly fexofenadine use was recorded for week 1 to week 4, while serum total IgE was measured at baseline and week 4.

2.7. Sample size calculation

The sample size was calculated to detect differences in the primary outcome, the change in SCORAD index from baseline, with a type I error (α) of 0.01 and a type II error (β) of 0.2. A conservative two-sided α = 0.01 was used at the planning stage to inflate the required sample size for adequate power; the inferential threshold for analyses was prespecified as two-sided α = 0.05. Our previous studies showed that XFS reduced SCORAD scores by 7.4 points with a standard deviation (SD) of 8.3, while AA and sham AA reduced scores by 23.5 and 18.2 points, respectively.^10,15 Based on this, the combination of XFS with AA was estimated to reduce SCORAD by 30.9 points, compared to 25.6 points with XFS and Sham AA. This corresponds to a planned between-group mean difference of 5.3 points in change from baseline, with a common SD of 8.3 derived from prior data, yielding a standardized effect size of Cohen’s d = 0.64.^10,15 Using a two-sample independent-means comparison formula, the required sample size was estimated to be 58 per group; allowing for a 25 % dropout, 78 participants were required per group.

2.8. Adverse events

Expected adverse events (AEs) related to AA were prespecified and assessed, including local pain, discomfort, skin irritation (itching and redness), inflammation or bleeding, chondritis, dizziness, nausea, and hypersensitivity reactions. In addition, any unexpected AEs associated with the procedure were documented and monitored throughout the study.

2.9. Statistical analysis

All analyses of treatment efficacy were performed on the intention-to-treat (ITT) population. Missing data were handled using multiple imputation. Continuous variables were summarized as means ± SDs. Comparisons were performed using Student’s t-test with results expressed as mean differences (MDs) and 95 % confidence intervals (CIs). Between-group differences at week 4 were additionally estimated using ANCOVA with the baseline values, age, and sex as covariates. AEs were reported as n (%), with between-group comparisons conducted using Fisher’s exact test, and results expressed as risk ratios (RRs) with 95 % CIs. Blinding was evaluated using James’ and Bang’s blinding index (BI). After their last AA or sham AA session, participants were asked to guess their group allocation. This included those who withdrew early but remained contactable. James’ BI assessed overall blinding, while Bang’s BI evaluated group-specific directional bias. All statistical analyses were performed using R software (version 4.3.3), with a two-sided p-value < 0.05 considered statistically significant (equivalent to α = 0.05). In addition, motivated by the conservative planning α described, the primary endpoint at week 4 was evaluated as a sensitivity analysis against a two-sided α = 0.01.

3. Results

3.1. Participants

From July 15, 2024, to March 23, 2025, a total of 231 patients were assessed for eligibility. Of these, 156 met the inclusion criteria and were randomly assigned to either the XFS plus AA group or the XFS plus Sham AA group. All analyses followed the ITT population. The recruitment, allocation, and follow-up process was presented in Fig. 1.

Fig. 1.

CONSORT flow diagram.

AA, auricular acupuncture; XFS, Xiao-Feng-San.

Participant characteristics were generally balanced between the two groups. The mean age was 38.6 years in the XFS plus AA group and 40.7 years in the Sham AA group, with a predominance of female participants. Most patients were in the subacute stage of the disease, with moderate severity and significantly impaired quality of life (Table 1).

Table 1.

Participant characteristics.

	XFS plus AA (n = 78)	XFS plus Sham AA (n = 78)
Sex (female), n (%)	41 (52.6)	44 (56.4)
Age (years), mean ± SD	38.6 ± 14.1	40.7 ± 13.5
Occupation, n (%)
No labor	12 (15.4)	14 (17.9)
Manual labor	19 (24.4)	15 (19.2)
Mental labor	47 (60.3)	49 (62.8)
History of allergic rhinitis, n (%)	21 (26.9)	16 (20.5)
History of asthma, n(%)	0 (0)	3 (3.9)
Disease stage, n (%)
Acute	6 (7.7)	5 (6.4)
Subacute	60 (76.9)	58 (74.4)
Chronic	12 (15.4)	15 (19.2)
Disease severity, n (%)
Mild	2 (2.6)	2 (2.6)
Moderate	76 (97.4)	76 (97.4)

AA, auricular acupuncture; IQR, interquartile range; SD, standard deviation; XFS, Xiao-Feng-San.

3.2. Primary outcomes

Both groups showed a reduction in the SCORAD index over time. The XFS plus AA group demonstrated a significantly greater improvement from week 2 (MD [95 % CI]: −2.1 [−3.8, −0.4], p = 0.017; not significant at α = 0.01) (Supplement 3). After four weeks of treatment, the XFS plus AA group showed a significantly better reduction, compared to the XFS plus Sham AA group (MD [95 % CI]: −9.1 [−10.9, −7.3], p < 0.0001) (Table 2). This difference reached the MCID and remained significant at two-sided α = 0.01. Notably, SCORAD-A did not change significantly in either group. Significant differences were observed in SCORAD-B and SCORAD-C, favoring the XFS plus AA group (Table 2, Supplement 3). ANCOVA adjusted for SCORAD baseline values, age, and sex confirmed these effects and was consistent with the unadjusted analyses (Table 2).

Table 2.

Outcomes at baseline and week 4.

		XFS plus AA (n = 78)	XFS plus Sham AA (n = 78)	MD (95 % CI)	p-value
SCORAD	Baseline	43.1 ± 7.8	41.9 ± 7.7	-	-
	Week 4	20.8 ± 6.0	28.7 ± 7.2	−7.9 (−10.0, −5.8)	< 0.0001
	Change*	−22.3 ± 6.0	−13.2 ± 5.4	−9.1 (−10.9, −7.3)	< 0.0001
	ANCOVA†	-	-	−8.5 (−10.0, −7.0)	< 0.0001
SCORAD – A	Baseline	23.0 ± 17.5	18.9 ± 15.5	-	-
	Week 4	22.1 ± 17.3	18.2 ± 14.6	3.8 (−1.2, 8.9)	0.1330
	Change	−0.9 ± 3.3	−0.7 ± 4.0	−0.2 (−1.4, 1.0)	0.6920
	ANCOVA†	-	-	−0.1 (−1.1, 1.1)	0.9150
SCORAD – B	Baseline	7.7 ± 1.7	7.6 ± 1.9	-	-
	Week 4	3.5 ± 1.2	4.9 ± 1.4	−1.4 (−1.9, −1.0)	< 0.0001
	Change	−4.3 ± 1.3	−2.7 ± 1.6	−1.6 (−2.1, −1.1)	< 0.0001
	ANCOVA†	-	-	−1.4 (−1.7, −1.1)	< 0.0001
SCORAD – C	Baseline	11.3 ± 3.4	11.4 ± 3.2	-	-
	Week 4	4.2 ± 2.3	7.9 ± 2.5	−3.7 (−4.4, −2.9)	< 0.0001
	Change	−7.2 ± 2.7	−3.6 ± 1.9	−3.6 (−4.3, −2.9)	< 0.0001
	ANCOVA†	-	-	−3.5 (−4.1, −3.0)	< 0.0001
Fexofenadine (tablets)	Week 4	0	0	-	-
	Total	0.2 ± 0.6	0.5 ± 1.1	−0.3 (−0.5, 0.0)	0.0800
	ANCOVA†	-	-	−0.3 (−0.6, 0.0)	0.0590
DLQI	Baseline	17.2 ± 6.5	16.5 ± 5.6	-	-
	Week 4	9.3 ± 5.1	11.0 ± 5.4	−1.8 (−3.4, −0.1)	0.0362
	ANCOVA†	-	-	−2.3 (−3.3, −1.2)	< 0.0001
Total IgE (IU/mL)	Baseline	553.1 ± 784.5	424.8 ± 661.6	-	-
	Week 4	518.8 ± 753.1	388.0 ± 610.7	130.7 (−86.1, 347.6)	0.2355
	Change	−34.3 ± 105.9	−36.7 ± 52.9	2.3 (−24.0, 28.8)	0.8588
	ANCOVA†	-	-	9.9 (−13.0, 32.9)	0.3930

Data are presented as mean ± standard deviation.

AA, auricular acupuncture; CI, confidence interval; DLQI, dermatology life quality index; IgE, immunoglobulin E; MD, mean difference; SCORAD, scoring atopic dermatitis; XFS, Xiao-Feng-San.

^⁎Primary outcome.

^†For SCORAD, DLQI, and total IgE, adjusted analyses (ANCOVA) included the corresponding baseline value, age, and sex as covariates; for total rescue fexofenadine tablets (with no baseline by design), models were adjusted for age and sex.

3.3. Secondary outcomes

The need for additional use of fexofenadine was low in both groups. The XFS plus AA group used significantly fewer tablets at weeks 2 and 3 (Supplement 4). After four weeks of treatment, neither group required any additional tablets. The total number of tablets used during the treatment period was lower in the XFS plus AA group, although the difference was not statistically significant (MD [95 % CI] of −0.3 [−0.5, 0.0], p = 0.08) (Table 2). DLQI decreased steadily in both groups, indicating an improvement in quality of life (Supplement 4). A significant difference favoring the XFS plus AA group was observed from week 3, with an MD (95 % CI) of −1.8 (−3.4, −0.1), p = 0.0362 at week 4 (Table 2, Supplement 4). There was no significant difference in serum total IgE between the two groups after four weeks of treatment. Adjusted analyses yielded concordant findings, indicating greater improvement in DLQI with XFS plus AA, a small tendency toward lower rescue fexofenadine use, and no between-group difference in change in serum total IgE (Table 2).

3.4. Adverse events

A total of 15 participants (19.1 %) in AA experienced AEs, primarily transient sharp pain at the site, which resolved immediately after needle removal. In sham AA, a total of nine participants (11.5 %) mainly experienced skin irritation caused by the adhesive patch, which subsided spontaneously after removal. Bruising occurred only in AA and resolved within a few days after needle removal (Supplement 5). All AEs were mild and left no sequelae.

3.5. Blinding assessment

Among participants, 68.8 % in AA and 52.6 % in sham AA believed they had received the real intervention. James’ BI indicated acceptable overall blinding (0.537; 95 % CI: 0.466, 0.610). While Bang’s BI suggested a tendency for correct guesses in AA (0.532; 95 % CI: 0.365, 0.699) and incorrect guesses in sham AA (−0.276; 95 % CI: −0.464, −0.088), both groups predominantly believed they were receiving the active treatment. One-sided CIs supported the directional interpretation (Supplement 6).

4. Discussion

4.1. Summary of main findings

The current study suggests that adding AA to XFS decoction offers additional benefits for non-severe AD. Greater improvements were observed in SCORAD total, SCORAD-B (objective signs), SCORAD-C (subjective symptoms), and quality of life compared to sham AA. Rescue antihistamine use was low in both groups, with a non-significant trend favoring the combination group. No differences were found in SCORAD-A or serum total IgE. AEs were mild and transient, supporting the safety of the intervention.

4.2. Comparison with previous studies

Evidence from high-quality randomized placebo-controlled trials has demonstrated that TM formulations can improve the severity and extent of skin lesions, as well as sleep quality, in patients with AD.¹⁶ Cheng et al. (2011) reported that oral administration of XFS significantly reduced total lesion scores, erythema, pruritus, and sleep disturbance in individuals with refractory AD.¹⁷ A systematic review and meta-analysis by Liang et al. (2024) confirmed that various acupuncture methods, including AA, significantly reduced SCORAD scores, alleviated pruritus, and improved quality of life.⁹ Our previous trial (2023) similarly found that AA, using the same auricular points as in the current study, was more effective than sham AA in reducing SCORAD and enhancing quality of life among patients with AD.

Building on this foundation, the current study contributes new evidence by demonstrating that the addition of AA to XFS yields superior clinical outcomes compared to XFS combined with sham AA. Furthermore, another of our previous trials evaluating thread-embedding acupuncture showed that its combination with XFS was more effective than XFS monotherapy in improving SCORAD scores and quality of life in patients with non-severe AD.¹⁵ Collectively, these findings support the notion that diverse acupuncture techniques may enhance the therapeutic efficacy of XFS in the management of AD.

It is noteworthy that previous studies have shown that although TM formulations and acupuncture methods can reduce both subjective and objective symptoms of AD, improvements in IgE levels were not observed.^9,16 Similarly, in the current study, the addition of AA to the XFS decoction did not result in a significant improvement in serum total IgE levels after four weeks of intervention.

4.3. Potential mechanisms of the intervention

AD is a chronic inflammatory skin disease driven by genetic susceptibility, environmental factors, barrier dysfunction, and immune dysregulation. Its pathophysiology involves Th2/Th17-skewed responses and elevated cytokines, leading to pruritus, barrier damage, and persistent inflammation.^18,19

While the mechanisms of body acupuncture in AD have been explored, evidence regarding AA in AD remains limited.20, 21, 22, 23 AA may alleviate symptoms of AD through a range of potential immunomodulatory mechanisms. Auricular stimulation has been shown to modulate inflammatory mediators, including serum C-reactive protein, tumor necrosis factor-α, interferon-γ, interleukin (IL)-1β, IL-6, and IL-10, thereby promoting a shift toward an anti-inflammatory immune profile.^24,25 These effects are mediated via activation of the auricular branch of the vagus nerve, which engages the parasympathetic system and initiates the cholinergic anti-inflammatory pathway (ChAIP). In addition, AA modulates the neuroendocrine-immune axis, particularly through the hypothalamic-pituitary-adrenal axis and ChAIP, contributing to immune homeostasis.^26,27 At the cellular level, AA upregulates α7 nicotinic acetylcholine receptor expression and promotes regulatory T cell (Treg) activity, which helps suppress chronic inflammation.²⁶ Auricular areas outside vagal innervation may also exert effects via other sensory nerves, such as the great auricular and auriculotemporal nerves, which influence autonomic and immune pathways. This may help explain the efficacy of AA in non-vagal regions.27, 28, 29, 30 These mechanisms may explain the clinical benefits observed despite the absence of significant changes in serum total IgE.

4.4. Clinical practice and future research implications

The current study supports the integration of AA with XFS therapy for the management of non-severe AD, demonstrating greater improvements in both objective signs and subjective symptoms. AEs were mild, transient, and primarily related to skin penetration, aligning with existing safety data.³¹ To reduce AEs, non-invasive alternatives like Vaccaria seeds may be considered.

Although Bang’s BI indicated partial deviation, most participants believed they received active treatment, suggesting a balanced expectancy effect. The overall James index remained within acceptable limits, supporting blinding validity. Future trials should consider sham AA applied to non-therapeutic points rather than omitting stimulation entirely, to enhance blinding and control for placebo effects.

Further research should evaluate AA combined with a wider range of TM formulations reflective of real-world practice, and adopt longer follow-up durations to assess sustained efficacy. Mechanistic studies using molecular and neuroimmunological markers are also needed to elucidate AA’s therapeutic pathways.

4.5. Limitations

Several limitations should be acknowledged. First, despite acceptable overall blinding, participants tended to believe they had received active treatment, which may have amplified treatment effects through expectancy or placebo-related mechanisms. Second, the lack of post-intervention follow-up leaves it unclear whether the observed effects will be maintained or quickly diminish after the intervention. Third, concerns about longer-term safety or cost-effectiveness arising from extended intervention could not be addressed by the current study. Finally, because the sample comprised only individuals with mild-to-moderate AD diagnosed with the wind-dampness-heat pattern and only one herbal formula was evaluated, the generalizability of the findings to broader clinical contexts may be limited.

4.6. Conclusions

The addition of AA to XFS therapy demonstrated superior clinical benefits in managing non-severe AD, with improvements in both objective signs and patient-reported outcomes, and a safety profile. These findings support the potential of AA as a safe and effective adjunctive treatment in integrative approaches for AD. Further studies are warranted to validate these results across broader populations, with post-intervention follow-up, and diverse TM regimens.

Funding

This research was supported by the University of Medicine and Pharmacy at Ho Chi Minh City under the research project contract No. 270/2025/HĐ-ĐHYD.

Ethics statement

This research was reviewed and approved by the institutional review board of the University of Medicine and Pharmacy at Ho Chi Minh City (Approval No. 270/HĐĐĐ-ĐHYD). Informed consent was obtained from all participants.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

CRediT authorship contribution statement

Nga Thu Tran: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Project administration, Funding acquisition. An Hoa Tran: Software, Formal analysis, Writing – original draft, Writing – review & editing, Visualization. Dieu-Thuong Thi Trinh: Conceptualization, Methodology, Resources, Writing – review & editing, Supervision, Project administration, Funding acquisition.

Declaration of competing interest

The authors declare that they have no conflicts of interest.