The Efficacy of Complementary and Alternative Medicines in Medical Dermatology: A Comprehensive Review

Abstract

Background

In recent years, there has been a widespread patient use of complementary and alternative medicines (CAMs) for dermatological application, despite few RCT-level studies on these supplements. This creates a barrier for dermatologists and others in counseling patients who may be using or might be tempted to use these CAM agents. This review investigates various CAM modalities used by patients for medical dermatology, exploring their efficacy and toxicity profiles.

Methods

A comprehensive review was performed on the effectiveness of several CAMs utilized in medical dermatology by patients. A literature search was conducted using PubMed, Embase, Google Scholar, Web of Science, and Cochrane.

Results

Most CAM modalities had statistically insignificant results, and for agents that had significant results in efficacy, these studies were questionable with flawed designs and methodologies.

Conclusion

These CAM supplements have promising potential in dermatologic use and are deserving of further investigation in well-crafted RCT-level studies. A more practical focus in future studies should involve a comparison of CAM agents to conventional therapies either alone or in an integrative fashion. This would accurately represent how these agents will be used clinically by actual patients and will be more helpful to clinicians. In the meantime, dermatologists should be aware of bias in published studies demonstrating the effectiveness of certain CAM modalities, and their corresponding toxicity. By doing so, physicians act as a valuable resource to patients who would like to explore various CAM products, better guiding patient interactions and treatment with improved patient outcomes.

The National Center for Complementary and Alternative Medicine (NCCAM) defines CAM as practices and products outside conventional medicine, with "Integrative Medicine" combining CAM and conventional therapies having demonstrated efficacy.^1,2 National surveys show 30 to 50 percent of adults use some CAM agents, with higher usage among women, the elderly, non-whites, immigrants, and lower/higher income groups.³ CAM usage is even greater (35–69%) among patients with dermatologic conditions such as atopic dermatitis or psoriasis, most commonly homeopathy, special diets, and herbs.⁴ Additionally, a National Health Interview Survey found that 85 percent of patients reporting skin problems use CAM products.⁵

Despite high patient usage, physicians of almost all disciplines lack CAM training, reporting unfamiliarity with efficacy and side effects.^6,7 While skeptical of efficacy, most physicians desire more CAM knowledge to counsel patients meaningfully.^7,8 Limited dermatologist data suggests similar attitudes. Nonetheless, greater CAM familiarity is quite important considering higher CAM usage among dermatological patients.^9,10 Awareness of effective CAMs allows counseling on integrative plans and watching for drug interactions and adverse effects. High dermatology CAM usage could also suggest unmet patient needs, so comprehensive CAM knowledge could improve physician-patient interactions.⁴

This review explores CAM therapies across dermatologic conditions with elucidated mechanisms, focusing on those with randomized, controlled trials (RCTs). The Jadad Scale, Table 1, will assess evidence quality to guide practitioners on potential CAM efficacy.¹¹ Adverse effects are also outlined for physician awareness of risks from unsupervised CAM use.

TABLE 1.

Jadad scale for assessing the quality of randomized controlled trials

YES	DESCRIPTION	NO
+1	The study was randomized	0
+1	The study’s method of randomization was appropriate (computer-generated, random numbers)	0
+1	The study was double-blinded	0
+1	The study included details of appropriate double-blinding (identical placebo, dummy)	0
+1	All patients within the study, including withdrawals and dropouts, were accounted for	0

ATOPIC DERMATITIS

Traditional Chinese medicine. Traditional Chinese herbal medicine (TCM), typically an extract of 10 herbs, was reported effective for atopic dermatitis (AD) in early 1990s RCTs.¹² In RCTs from Sheehan et al,¹³ children followed by adults participated in cross-over, four-week oral administrations of either TCM or placebo plants, and TCM improved skin damage and erythema compared to the placebo group;¹⁴ however, subsequent students had high withdrawal rates (only 40% of pediatric and 50% of adult patients continuing treatment) potentially due to lack of response.^15,16

Proposed TCM mechanisms for AD in non-human studies include inhibitory effects on both mononuclear and mast cells,^17,18 modulating effects on dendritic cells/Langerhans cells,^19,20 and reducing IgE complexes with CD23+ cells.^21,22 However, actual use in human-RCTs, such as Fung et al,²³ found no improvement with the TCM product Zemaphyte®. Though some studies found significant results with TCM therapy, most studies have found mixed or limited benefits for TCM use in AD management.^24–29

Serious adverse effects like hypersensitivity, organ toxicities, and contamination with heavy metals, steroids, or NSAIDs have been reported with TCM use.^12,30–33 Jadad scoring (Average 3.5) is complicated by herb heterogeneity, preparation variation, and use of differing subjective outcomes across studies. While most were well-designed RCTs suggesting evidence for improving visual AD severity, the quality is affected by small sizes, varied outcome measures, and potential subjectivity caveats.

Dietary: Fatty acids and probiotics. Alterations in fatty acid metabolism in AD patients may lead to deficiencies in anti-inflammatory metabolites like prostaglandin E1 and 15-hydroxy-eicosatrienoic acid.³⁴ RCTs examining supplementation with γ-linolenic acid (GLA) sources like borage, primrose, black currant seed, and fish oils (n-3 fatty acids) showed modest improvements in total area affected and pruritus with low incidence of adverse effects.^35–37 Unfortunately, such improvements were not statistically significant.

Probiotics, such as Lactobacillus, have been studied for managing AD in infants, children, and adolescents.^37–48 RCTs giving Lactobacilli alone or in combination to pregnant mothers and infants showed significantly reduced AD incidence (RR: 0.51);⁴¹ (OR: 0.51) for non-IgE-associated AD;⁴² (RR: 0.32) for breastfed infants.⁴³ Kalliomaki et al⁴⁴ follow-up studies demonstrated that positive benefits persisted up to age 7.⁴⁵ Proposed mechanisms include increased breast milk TGF-β2.⁴³ However, other RCTs utilizing probiotics have shown either mixed results or no effect at all.^46–48

These RCTs were generally well-designed (Jadad of 5), with low risk of bias, but some of the equally well-designed and executed trials showed no improvement in AD incidence. For treatment of previous diagnosed pediatric AD, RCTs, such as Boyle et al,⁴⁹ showed slight but non-significant symptom improvements with 6 to 12 week probiotic treatment; this is quite modest compared to current conventional therapy. Thus, while probiotics show some potential as a preventative intervention, their role as AD treatment has limited supportive evidence. Additional research is needed on specific strains of probiotics, dosing, and treatment duration for both preventative and therapeutic applications.

Other botanicals and non-botanical CAM modalities. Other botanicals studied for AD include St. John's Wort (Hypericum perforatum) due to known anti-inflammatory and anti-bacterial effects. A small RCT (n=18) showed topical St. John's Wort improved SCORAD scores versus vehicle placebo but lacked inter-group analysis.⁵⁰ Camomile cream demonstrated mild superiority over placebo.⁵¹ Licorice (Glycyrrhiza glabra) gel at 1 to 2 percent reduced subjective itch, edema, and erythema scores versus placebo in a reported RCT lacking report of precise methodological details (Jadad of 4).⁵² Arnica, Calendula, tea tree oil, and witch hazel propose anti-inflammatory mechanisms but lack RCT-quality evidence.⁵³ Some of these agents have been reported to cause allergic dermatitis from sesquiterpenes or GI toxicity with ingestion.^{12, 34, 54}

Non-botanical CAMs include homeopathy, mind-body interventions, and energy medicine (acupuncture and bio-resonance). There have unsurprisingly been a few reproducible RCT-level studies that have shown a positive effect of homeopathy in AD.^34,55 Mind-body RCTs showing efficacy had major design flaws.⁵⁶ More structured parent-adolescent educational mind-body programs demonstrated some efficacy.^56,57 Massage therapy reduced anxiety and AD symptoms in one study but lacked proper control group matching and blinding.^34,58 Acupuncture has proposed protocols but limited RCT evidence,^59–61 and bio-resonance has been promoted in case reports but lacks well-designed RCTs showing efficacy.⁶² Overall, most botanicals and non-botanical CAM therapies lack rigorous RCT support for AD treatment despite proposed anti-inflammatory mechanisms. More high-quality research is needed.

PAPULOSQUAMOUS DISEASES: PSORIASIS

Vitamins, minerals and other botanicals. CAM usage is prevalent (43–69%) among patients with psoriasis.⁶³ Limited RCTs exist for vitamins and mineral supplements studied in psoriasis. Well-designed RCTs found no benefit of zinc,^64,65 selenium,⁶⁶ vitamin D3,^67,68 or both vitamin B12 and avocado cream⁶⁹ on improving psoriasis. Some of these studies had questionable study designs, lack of blinding, manufacturer funding, and small sample sizes.⁶⁶

For botanicals, one RCT found 0.5% aloe vera cream improved undefined subjective endpoints versus placebo but lacked inter- and intra-group statistics.⁷⁰ Another well-designed left/right trial showed no significant topical aloe vera benefit for psoriasis.⁷¹ Neem supplementation significantly reduced psoriatic symptoms when compared to topical tar and salicylic acid therapy.⁷² The commercial 10% Mahonia aquifolium (berberine) product, Relieva, significantly decreased psoriatic symptoms with minor side effects (pruritus and burning) in a large RCT study.⁷³ While suggesting some efficacy, botanicals have only limited RCTs to support clinical use. Nonetheless, what trials do exist are generally of good quality (Jadad range of 3–5).

In summary, vitamins and mineral supplements lack therapeutic evidence for psoriasis treatment, while some botanicals like neem and berberine show promising potential.

Fish oil (omega-3 fatty acids). Fish oil supplements, containing omega-3 fatty acids, such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), are commonly used as complementary psoriasis treatment. Proposed mechanisms involve inhibiting inflammatory arachidonic acid metabolites.⁶³ Multiple RCTs examined fish oil with or without concomitant therapy, with mixed findings. Some studies have shown benefits, such as reduced itching, erythema, and scaling with 10 capsules of daily fish oil⁷⁴ and modest symptom reductions with an oily fish diet⁷⁵ or EPA supplementation to etretinate.⁷⁶ Few RCTs also found decreased psoriasis severity scores with intravenous omega-3 fatty acids, but have had methodological limitations.^77,78

Other well-designed RCTs such as Gupta et al,⁷⁹ found significant reduction in disease severity when comparing fish oil to olive oil controls;however, there were no significant improvements with fish oil use when compared to betamethasone controls,⁸⁰ or purified omega-6 (corn oil) fatty acids.⁸¹ A left/right comparison trial showed benefits only on plaque scale and thickness but not itching and erythema, with the latter being important and relevant clinical endpoints.⁸²

Despite some studies with significant positive results, the overall evidence quality for fatty acids in psoriasis management is limited (average Jadad 3.6) due to issues like lack of blinding, high supplement burden, and lack of comparison to conventional treatments.

Climatotherapy/balneotherapy. Climatotherapy (exposure to Dead Sea-like conditions) and balneotherapy have been studied as complementary psoriasis treatments. Dawe et al⁸³ performed a left/right comparison study with 12-week treatments of combined dead sea salt soaks and narrow-band UVB therapy compared to the phototherapy alone. Study findings showed mixed results; however, large well-designed multicenter RCTs showed significant improvement in achieving a 50 percent reduction of psoriasis severity score (PASI 50) both with low (5–10%) or high (25%) concentration saline baths complementing phototherapy over phototherapy-only.^84,85 Another study found phototherapy-complementary saline treatment significantly reduced disease severity score outcomes when compared to groups treated with tap water or psoralen-containing bath add-ons.⁸⁶

Alas, a host of other studies have not demonstrated as much success with climatotherapy: 30% dead sea salt lotion was ineffective;⁸⁷ saline baths provided no added benefit to phototherapy;⁸⁸ saline water and tap water baths had equivalent findings;⁸⁹ dead sea salt and common salt baths showed no significant difference between treatment groups.^90,91

Adverse effects like erythema, burning and pruritus occurred, increasing dropout rates. Lack of patient blinding to salinity introduces potential placebo effects. Despite mixed results, the large high-quality RCTs, such as Brockow et al,^84,85 suggest benefit for saline balneotherapy complementing phototherapy; however, an average Jadad of 3 was given due to lack of proper blinding with the added caution that such therapy may produce irritation.

Mind/body, meditation, and stress relaxation.

Psychological stress has been implicated as an exacerbating factor in psoriatic flare-ups. Several studies have examined stress alleviation techniques as complementary treatments for psoriasis, with ambiguous findings. A small pilot RCT (n=18) found a significant but modest improvement with daily home meditation for scalp psoriasis.⁹² Kabat-Zinn et al⁹³ studied stress relaxation audio tapes during phototherapy, concluding improved treatment outcomes, but with inconsistent and unblinded endpoint scoring. Zachariae et al⁹⁴ used psychotherapy and relaxation tapes, finding a modest positive effect, but no significant difference in severity index scores or correlation between disease activity measures and stress levels. A similar RCT comparing relaxation techniques with and without biofeedback found improved patient-reported scores but no difference in blinded physician assessments.⁹⁵ Tausk et al⁹⁶ used hypnosis with and without suggestion, finding no difference in disease severity score change between hypnosis types, and also lacked an untreated control group.

While these mind-body interventions show potential beneficial effects and no reported negative adverse effects, the overall quality of evidence for their effectiveness as complementary psoriasis therapy is minimal, primarily due to improper blinding, risk of bias, and mixed or insignificant results (average Jadad score of 3).

CONTACT DERMATITIS

While a substantial proportion of allergic contact dermatitis patients report use of a CAM agent for treatment, upwards of 40 percent in one survey,⁹⁷ relatively few have been tested in clinical studies. One notable CAM modality that has been studied, however, is jewelweed (Impatiens biflora), first studied in a report from the 1950s where investigators found it comparable to conventional therapy in the treatment of poison ivy contact dermatitis.⁵⁴ Abrams Motz et al⁹⁸ claimed to confirm this efficacy in a non-blinded study when they observed that jewelweed mashes resulted in a decreased rash score following poison ivy skin exposure. However, the control used, water-only, did not account for any potentially beneficial non-pharmacologic properties of the mash, and outcomes were selectively emphasized as most results were on balance negative with no effect found using a purified jewelweed extract (vs. mash), no relationship was found between concentration of active ingredient, lawsone, and clinical response, and no added benefit was found between groups exposed to jewelweed-containing soaps compared to regular dish soap. These on-balance negative findings seem to be confirmed in two other studies. Long et al⁹⁹ performed a small but well-designed left-right RCT that compared the treatment of rashes derived from urushiol patches (combined poison ivy and poison oak) with jewelweed extract or water-only control and found no difference in blinded rash scores over 9 days. This confirmed negative result from an earlier observational study.¹⁰⁰ Therefore, there seems to be inconsistency between study outcomes, and a possible risk of bias in the positive study [Abrams], even though the average quality of the evidence (i.e. Jadad score) was a 4.

ACNE

Acne vulgaris. Given the high prevalence of acne, numerous CAM modalities have been suggested as potential treatments, but only a few have been studied through randomized controlled trials.¹⁰¹ Tea tree oil (TTO) has shown anti-microbial and anti-inflammatory effects, with one industry-sponsored study finding significant improvement with 5% topical TTO gel over placebo.¹⁰² These significant findings were consistent in another RCT;¹⁰³ however, another study (n=124) reported greater effectiveness of 5% benzoyl peroxide compared to TTO gel, though accompanied with increased dryness, and lack of proper blinding due to TTO's distinct odor.¹⁰⁴ An average Jadad score of 3.7 was assigned to these studies.

RCTs on fruit-derived alpha-hydroxy acids (AHAs) like gluconolactone, glycolic acid, and azelaic acid demonstrated mixed results compared to conventional treatments (i.e. benzoyl peroxide and tretinoin), with concerns on study design and adverse effect assessment (average Jadad score of 4)^105–108 Needless to say, many practitioners employ glycolic acid peels to help with both active acne and post-inflammatory hyperpigmentation. A study on topical basil oil reported implausibly rapid improvement but lacked blinding details, and omitted key statistics.^109,110 RCTs on Japanese Kampo drugs and 2% green tea extract had major flaws like absent endpoint assessment and lack of controls.^111,112 However, a better-crafted study on green tea extract showed no significant results.¹¹³ Other agents with limited RCT-level evidence include pyridoxine, vitamin A, Indian Ayurvedic herbs, Nigerian toto, and shark liver extract, but their efficacy cannot be reliably evaluated due to insufficient RCT-quality evidence.

Acne rosacea. Given the frustration with conventional acne rosacea therapies and the potentially high cost associated with prescription medications, patients may turn to CAM treatments. Niacinamide (vitamin B3) has anti-inflammatory and sebum-reducing properties, along with the potential for improving keratin synthesis and keratinocyte differentiation.^114–116 An open-label study on a niacinamide-zinc product showed promise but lacked proper match controls.^114,117 Studies on zinc alone have demonstrated conflicting results, showing efficacy in Sharquie et al¹¹⁸ and no significant improvement in Bamform et al.¹¹⁹ Licorice (Glycyrrhiza) extracts demonstrated anti-inflammatory effects in vitro, but human trials to date have lacked controls and investigator/patient blinding.^120–122 Feverfew has anti-inflammatory and anti-oxidant activity but hasn't been studied in published RCTs for rosacea.^123,124 Colloidal oatmeal has anti-oxidant and anti-inflammatory properties in vitro and animal studies but no RCT-level human studies.^125,126 Green tea and coffeeberry (used for antioxidant polyphenols), aloe vera (for anti-pruritic effect), chamomile, turmeric, and mushroom extracts may be tried as alternative therapies but lack rosacea-specific clinical human studies.¹²⁷ A recent RCT examined tea tree oil (TTO) with permethrin, showing positive improvements, but the TTO-specific contribution could not be adequately assessed.¹²⁸ Despite the limited high-quality CAM studies for acne rosacea, the average Jadad score was 4.3, indicating reasonable study design and quality.

ALOPECIA

Alopecia areata. A few RCTs have been performed assessing CAM agents in the treatment of alopecia areata. Hay et al¹²⁹ studied a seven-month aromatherapy treatment with a combination of essential oils from thyme, rosemary, lavender, and cedarwood in a carrier with jojoba and grapeseed oil compared to a carrier oil-only control. The treatment group showed significant improvement in the primary outcome (subjective disease severity judging of photographs by two blinded physicians) and in the secondary outcome of affected area reduction (measured by tracings). It is unclear, however, if groups were well-randomized before the trial; no data is provided on baseline disease severity or area affected. Furthermore, primary outcome measures were intra-group with limited information provided as to how scores were “improved” and secondary outcomes of area affected were only measured in a small fraction of study participants with no explanation offered. Two other studies have been performed assessing topical treatment with onion and garlic for alopecia areata due to purported therapeutic effects of their constituent sulfur amino acid allicin.^130,131 While both studies report effectiveness, both suffer fatal design and methodological flaws. The onion study does not assess baseline randomization, does not describe the determination of “response” endpoint or its blinding, and critically, does not perform an intention-to-treat analysis to account for the substantial drop-out rate (~50%) in the treatment group.¹³⁰ The garlic study reports significantly better baseline data on patch sizes, terminal hair number, and total hair number in the treatment group, but does not address this randomization problem and proceeds with inter-group comparisons after three months of treatment.¹³¹ Several other CAM modalities have shown promise as alopecia areata therapies in isolated case reports or in small patient series, including hypnotherapy, massage and relaxation therapy, energy healing therapies (magnetic stimulation, acupuncture), and other biologics (vitamin A, TCM), but these have yet to be examined in an RCT-level study to properly analyze efficacy.^132,133

Androgenic alopecia. A few CAMs have also been examined in RCTs for effectiveness in androgenic alopecia. Kessels et al¹³⁴ examined a commercial product called Dabao composed of multiple different Chinese herbal extracts including saffron, Mulberry, sesame, ginger, and peppers. The authors conclude a modest, but statistically significant improvement of the product lotion over placebo in blinded hair counts and cosmetic evaluation by subjective panel evaluation. However, while the study was well-randomized and blinded, the statistics used to assess differences were inappropriate (one-sided t-test) and unclear (somehow analyzed the difference between inter-group means). Further, it is difficult to parse out the exact medicinal and mechanistic effect of a product containing dozens of botanicals. Takahashi et al¹³⁵ tested a topical application of the apple-derived procyanidin B-2 in a small study (n=29) due to purported polyphenol-related antioxidant effects.¹³⁵ This investigator group concluded strong effectiveness based upon improvements in total hair number per 0.25cm² area, increased ratio of large hairs, and subjective disease improvement. However, there was no change in one of their primary outcomes, overall hair diameter, and methodology on the areas from which hair counts were derived were not described. Furthermore, overall design was in doubt as there was no description of placebo-matching to ensure blinding and no baseline comparisons made between groups in the outcomes measured, a critical oversight as most changes were evaluated in an intra-group fashion.

Many other CAMs have been tested in vitro, in animal studies, and in non-RCT level human studies for potential application in androgenic alopecia. Topical application of raspberry ketones derived from red raspberries (Rubus idaeus) may activate sensory neurons to promote insulin-like growth factor-I release and showed promising results in hair regrowth in mice and non-placebo-matched humans.¹³⁶ Several other botanical agents, mostly postulated to act through 5α-reductase inhibition-related mechanisms, have shown positive effects in animal studies including botanicals green tea-derived EGCG,¹³⁷ red ginseng (Panax ginseng C),¹³⁸ Cuscuta reflexa,¹³⁹ a polyherbal mixture of E. officinalis, C. asiatica and aloe vera,¹⁴⁰ and other herbal systems including traditional Thai herbal plants¹⁴¹ and Indian Ayurvedic herbs.¹⁴² Finally, Nioxin, a popular commercial system of shampoos, conditioners, and other topical products with multiple constituents including botanicals (tea extracts, peppermint oil, many others), vitamins, and amino acids is purported to help with a variety of causes of thinning hair.¹⁴³ However, no scientific studies (in vitro, animal, or human) solely investigating this agent have been published, assessing its efficacy in improving thinning hair or in treating any other form of alopecia.

PIGMENTATION DISORDERS: MELASMA

Azelaic acid. Azelaic acid, a dicarboxylic acid, acts in melasma by competitive inhibition of tyrosinase, inhibiting melanocyte proliferation and decreasing reactive oxygen species formation.¹⁴⁴ Two RCTs tested a 20% topical formulation, with one study comparing it to a vehicle-only control¹⁴⁵ and the other combining it with glycolic acid and comparing it to 4% hydroquinone (with glycolic acid lotion vehicle).¹⁴⁶ While these studies reported efficacy, they lacked proper randomization, baseline comparisons, and assessor blinding descriptions, resulting in an average Jadad score of 3. Baliña et al¹⁴⁷(n=329) found no significant differences between 20% azelaic acid and 4% hydroquinone but higher local irritation in the azelaic acid group, while Sivayathorn (n=340) reported significantly greater improvements with azelaic acid compared to 2% hydroquinone.¹⁴⁸ However, both studies had serious methodological flaws, including lack of randomization and blinding descriptions, no intention-to-treat analysis, and potential conflicts of interest. Other RCTs on this agent reported mixed findings.^149,150 Overall, there were also opportunities for serious risk of bias issues, (average Jadad of 3.2). It is worth noting that the risk of application site irritation is greater with azelaic acid compared to conventional treatments, despite at-best equivalent efficacy.

Kojic acid. Kojic acid is a compound derived from Aspergilline oryzae and is postulated to act in melasma by inhibiting tyrosinase, and chelating copper at its active site.¹⁵¹ Lim et al¹⁵² compared a 2% kojic acid gel with 2% hydroquinone in a left-right study with both treatments used in combination with 10% glycolic acid, assessing subjective disease clearance percentage and overall effectiveness. The authors conclude that kojic acid is effective with higher 50 percent clearance at 12 weeks than hydroquinone and better or equivalent patient-reported overall effectiveness, but they offered no description of how outcomes were tabulated, do not indicate that assessment was blinded, and offered no statistics to properly compare the treatments. Thus, this study has diminished weight in determining kojic acid efficacy; the study only rates a Jadad score of 2. Garcia et al¹⁵³ performed a similar left-right comparison with the same treatment groups, concluding equivalent efficacy based on similar overall reductions in pigment. The methodology for measuring this outcome was not described, no statistics were provided, and there is no indication that the study was blinded. Greater irritation side effects were described in the kojic acid group but not elaborated upon. Even a recent study by Al-Dhalimi et al¹⁵⁴ which explored combination therapy of kojic acid and Yttrium Aluminum Garnet Laser vs. kojic acid alone found using only kojic acid as treatment to have less statistically improved clinical outcomes compared to the combination therapy. As a result, we give these studies an overall average quality of evidence of 2 per the Jadad scale, for risk of bias and note that the serious methodological problems with these studies make the kojic acid treatment especially difficult to recommend without more rigorous study.

Vitamin C (ascorbic acid). Vitamin C is also thought to inhibit tyrosinase, interacting with copper at the active site to decrease activity; it may also potentially act by reducing dopaquinone through inhibition of dihydrochinindol-2-carboxyl acid oxidation.^151,155 A few studies have examined its utility as an alternative or complementary treatment for melasma. One small left-right comparison study from Espinal-Perez et al¹⁵⁶ examined 16-week topical application of a 5% L-ascorbic acid formulation compared to a 4% hydroquinone emulsion and measured colorimetry and subjective patient-reported lightening. Patients reported significantly greater “very good” and “excellent” responses on the hydroquinone side with colorimetric results favoring hydroquinone, but that difference was not significant. The authors conclude that vitamin C may play a role in melasma therapy despite poorer results as it featured fewer irritation side effects. The actual significance of these results is questionable. However, as blinding for colometry assessment was not described and the methodology for patient-reported treatment efficacy was not explained. Soliman et al¹⁵⁷ compared a group receiving 5% ascorbic acid treatment with 20% TCA peels to a group receiving the peels alone and concluded that complementary Vitamin C helped improve and maintain treatment results based on a greater and more sustained intra-group decrease in melasma area and severity index (MASI) score. There was, however, no difference in the other measured study outcome (patient-reported global assessment), and the implications of the MASI outcome are uncertain given study methodology problems; there was no matched complementary placebo gel used in the control group and no indication that score evaluation was done by blinded assessors. Finally, in a left-right study, Huh et al¹⁵⁸ examined application of a more readily ionizing Vitamin C preparation compared to water using an iontophoresis machine with authors claiming efficacy based on a statistically significant intragroup decrease in colorimeter-measured luminance in a representative spot on the treatment side compared to control.¹⁵⁸ However, blinded participants rated the treatment and control responses as equally efficacious. While the study was described as double-blind, investigator blinding was not detailed and calorimetry measurement methodology poorly described with the possibility of variability in the measurement of treatment spots, reference comparison spots or both causing this mis-match between investigator- and patient-reported outcomes. Overall, when factoring in the risk of bias, flawed randomization, and design, the average Jadad score for the quality of the evidence of the aforementioned articles is a 2.7.

Other botanicals. RCTs have been performed using a few other topical botanical CAM agents to assess efficacy in melasma therapy. Licorice extract has been evaluated due to reported effects inhibiting the rate-limiting step of tyrosinase activity and other purported anti-inflammatory mechanisms.¹⁵¹ One small left-right comparison study by Amer et al¹⁵⁹ tested a cream with 2% liquiritin, the active ingredient in licorice, compared to a vehicle control with authors concluding efficacy after four weeks based on an ill-defined pigmentation scale. This study contained no essential methodological information in the methods with no information provided on blinding, baseline scores, placebo matching, or outcome measurements, lending little credibility to its results. Costa et al¹⁶⁰ examined a 60-day application of a commercial combination product of licorice, emblica (Phyllanthus emblica), and belides (Bellis perennis) compared to 2% hydroquinone cream and measured subjective clinical response and photograph-derived UV spot count and score (size and tone). Authors found statistically significant intra-group improvement over the study course in all measures, no inter-group differences between any measure at study endpoint, and fewer adverse events (mild erythema, burning), concluding that the combination product is an effective alternative to hydroquinone. The study featured appropriate inter-group comparisons and mostly well-detailed methodology, but suffered from a few limitations; it was only single-blind (patients knew group assignment), the description of a key outcome of UV spot score was poorly described and no intention-to-treat analysis was performed despite some dropouts. Furthermore, while authors claim no conflict of interest, the study funding was provided by the product manufacturer.

Other agents with potential application for melasma include arbutin, derived from berry plants with potential mechanism on tyrosinase inhibition and melanosome maturation, and soy, with its constituent soybean trypsin inhibitor acting on the protease-activated receptor-2 pathway important in keratinocyte malanosome phagocytosis.¹⁵¹ While in vitro studies suggest a role for these agents in hyperpigmentation, limited clinical studies have been performed in patients with solar lentigines and melasma,¹⁶¹ with no properly designed and blinded RCT-level studies performed on patients with melasma.

CUTANEOUS SIDE EFFECTS OF DERMATOLOGIC AND NON-DERMATOLOGIC CAM

Practicing dermatologists should be aware of potential dermatologic side effects that may result from patients using CAM therapies to treat both their dermatologic and non-dermatologic conditions. Unfortunately, many of the above studies and other CAM studies on other non-dermatologic conditions fail to do a complete accounting of adverse events occurring during trials, and even if adverse events are recorded, the study size is often so small that some low-frequency events may be missed. However, myriad case reports have been published on dermatologic diseases resulting from CAM use which will be briefly discussed here and summarized in Table 9.^31,162

Contact dermatitis and exacerbation of other dermatides. Inclusion of various metals in CAM preparations may lead to hypersensitivity reactions and both systemic and local contact dermatitis. Nickel, for instance, is found in many homeopathic remedies in addition to both Nigerian and Chinese herbal preparations in sufficient amounts to cause systemic or local contact dermatitis.¹⁶³ Allergic contact dermatitis can result from metals contained in acupuncture needles, with a patient presenting in one case report with papulovesicular dermatitis resulting from zinc exposure at acupuncture sites following treatment for dizziness.¹⁶⁴ Other non-metal CAM sensitizers have been implicated in contact dermatitis. For example, several case reports exist of occupational contact dermatitis resulting from exposure to essential oils and fragrances derived from CAM agents and used in aromatherapy settings (masseuse, reflexologist.), typically resulting in multiple different sensitizations.^165,166 Mercury contained in a homeopathic coughing remedy caused Baboon Syndrome in a pediatric patient in another case report,¹⁶⁷ and mercury contamination in TCM or Indian Ayurvedic medicines can have dermatologic effects including dryness, ulceration, erythroderma and hyperkeratotic palmar eczema.¹⁶² Notable botanical CAM modalities reported to cause allergic contact dermatitis include echinacea used as an immune stimulant,¹⁶⁸ chamomile and curcumin/turmeric used for general anti-inflammatory purposes,^169–171 TCM and other Chinese herbal preparations,¹⁷² TTO used for a variety of applications.^31,173,174 Some reports describe exacerbations of previously diagnosed disorders with CAM therapy. For example, reports documenting atopic dermatitis exacerbation with homeopathic therapies containing metals or other substances³¹ and an erythrodermic psoriasis flare-up in a patient with chronic plaque psoriasis resulting from high concentration dead sea salt exposure.¹⁷⁵ As such, CAM treatments should be considered as potential triggers of dermatitis flare-ups.

TABLE 3.

CAM therapies for atopic dermatitis

REFERENCE	N	JADAD SCORE	CAM	DOSAGE	ROA	STUDY RESULTS	SIDE EFFECTS
Sheehan et al14	40	4	TCM	200ml	Oral	The geometric mean score for surface damage was 11.3 (5.8 to 21.8) for the treatment phase and 111.0 (68 to 182) for the placebo phase. There was a subjective improvement in itching (p<0.001) and sleep (p<0.078) for the treatment phase.	No side effects were reported.
Funget et al23	40	4	TCM	600ml; 100ml	Oral	Zemaphyte, offered no statistically significant treatment effect over placebo for all four clinical parameters, except for lichenification at Week 4.	No side effects were reported.
Hon et al24	85	4	TCM	9g capsule	Oral	Over 12 weeks, the mean SCORAD score fell from 58.3 to 49.7 in the TCHM group (n=42; P=0.003) and from 56.9 to 46.9 in the placebo group (n=43; P=0.001). The CDLQI in TCHM-treated patients was significantly improved compared with patients receiving placebo at the end of the 3-month treatment and 4 weeks after stopping therapy (P=0.008 and 0.059, respectively). The total amount of topical corticosteroid used was also significantly reduced by one-third in the TCHM group (P=0.024)	No significant adverse effects were observed between the groups.
Shapira et al25	49	4	TCM	N/A	Oral	Total SCORAD rapidly decreased from 47.9 ± 3.6 to 31.8 ± 2.8 (P<0.0001) after 2 weeks of the study medication treatment (Fig 1). However, patients in the placebo group had a similar decrease [47.2 ± 2.7 to 29.0 ± 3.5 (P<0.0001)] without a statistical difference between the groups.	No noted side effects.
Ma et al26	176	3	TCM	30ml	Oral	The EASI score (χ2=14.181, P=0.003), recurrence rate (χ2=7.398, P=0.007), and itching score (F=−3.427, P=0.001) were lower in the QZLXD group than the RZZYC group.	No significant adverse effects were observed between the groups.
Ahn et al27	34	4	Jaungo	1 finger unit dose	Topical	There was a significant decline of SCORAD scores in treatment Group 1 and placebo group (p<0.05). Patients in all groups showed decreased TEWL and DLQI scores with no significant difference. There was a significant decline of IL-17 in all groups (p<0.05).	A total of 4 subjects (14.3%) had adverse events after application of the experimental medications. The adverse events were pruritus (n=4) and wheal formation (n=1).
Liu et al28	275	3	TCM	150 mL/2x daily	Oral and topical	At Week 36, the difference in QOL scores showed a significantly greater improvement in both Chinese herbal medicine-treated groups than in the control group (P<0.001 and P<0.001, respectively).	Feeling cold (n=18) and diarrhea (n=5) were adverse events found in the treatment group. There were no severe adverse events.
Cheng et al29	56	5	TCM	4g of Xiao-Feng-San/pack	Oral	The decrease in the total lesion score in the treatment group at 8 weeks was significantly greater than that of the placebo group (79.7 ± 5.8% vs. 13.5 ± 7.64%; p<0.001). There was also a statistically significant difference between the treatment and placebo groups with regard to erythema, surface damage, pruritus, and sleep scores. The difference between the 2 groups was still significant for all outcome measures except the erythema score at the 12-week follow-up.	Patients reported no side effects from treatment.
Komulainen et al36	439	5	Fish oil and probiotics	2 fish oil capsules/day; 1 probiotic capsule/day	Oral	No difference between the infants in the four intervention groups were found regarding physician-diagnosed food allergy, atopic eczema, or atopy at the age of 12 or 24 months (all p >0.05). The probiotic intervention was associated with lower odds of recurrent wheezing at 24 months (OR 0.39, 95% CI 0.18-0.84, p=0.017), but not at 12 months.	No noted side effects.
Palmer et al37	706	3	Fish oil	500mg/daily	Oral	No differences were seen in the overall percentage of infants with IgE-associated allergic disease between the n-3 LCPUFA and control groups (32/368 (9%) v 43/338 (13%)	No noted side effects.
Feíto-Rodríguez38	70	5	Probiotics	109 colony forming unit per capsule/daily	Oral	Mean SCORAD index at 12 weeks showed a statistically significant difference of -5.43 (95% confidence interval -10.65 to -0.21) between probiotic (13.52) and placebo groups (18.96); P = 0.04.	No adverse events were reported during the study period.
Michelott et al39	80	5	Probiotics	1 capsule/day	Oral	Subjects receiving the probiotic mixture showed an improvement in skin smoothness, skin moisturization, self-perception, and a decrease in SCORAD index as well as in the levels of inflammatory markers associated with AD at T28d, with a positive trend up to T56d which was maintained at T84d.	No adverse events were reported during the study period.

Prakoeswa et al40	30	5	Probiotics	2 x 1010 CFU/day	Oral	The SCORAD score was significantly lower in the probiotic than the placebo group on the 8th week. The IL-4 and IL-17 levels were significantly lower in the probiotic than in the placebo group. However, the IgE levels remained significantly unchanged.	No noted side effects.
Dotterud et al42	415	4	Probiotics	Probiotic milk	Oral	The odds ratio (OR) for the cumulative incidence of AD was 0.51 in the probiotic group compared with the placebo [95% confidence interval (CI) 0.30–0.87; P=0.013]. There were no significant effects on asthma (OR 0.68, 95% CI 0.26–1.80; P=0.437) or atopic sensitization (OR 1.52, 95% CI 0.74–3.14; P=0.254).	No noted side effects.
Niers et al46	102	4	Probiotics	N/A	Oral	Parental-reported eczema during the first 3 months of life was significantly lower in the intervention group compared with placebo, 6/50 vs 15/52 (P= 0.035). After 3 months, the incidence of eczema was similar in both groups.	No noted side effects.
Boyle et al47	250	3	Probiotics	1.8 × 101- cfu/day	Oral	Prenatal probiotic treatment was not associated with reduced risk of eczema (34% probiotic, 39% placebo; RR 0.88; 95% CI 0.63, 1.22) or IgE-associated eczema (18% probiotic, 19% placebo; RR 0.94; 95% CI 0.53, 1.68).	No noted side effects.
Kopp et al48	105	5	Probiotics	5x109 colony forming units of Lactobacillus GG 2x daily	Oral	Severity of atopic dermatitis was comparable between the 2 groups. Notably, children with recurrent (> or = 5) episodes of wheezing bronchitis were more frequent in the Lactobacillus GG group (26%; n=13), as compared with the placebo group (9.1%; n=4). No difference was observed between both groups in total immunoglobulin E concentrations or numbers of specific sensitization to inhalant allergens.	No noted side effects.
Schempp et al50	21	5	St John's Wart	2x daily	Topical	Hypericum-cream was significantly superior to the vehicle at all clinical visits (Days 7, 14, 28) (p<0.05). Skin colonization with Staphylococcus aureus was reduced by both verum and placebo, showing a trend to better antibacterial activity of the hypericum-cream (p=0.064)	4 adverse events (AEs) like contact dermatitis, were recorded in 3 patients. None of the AEs was classified as serious.
Saeedi et al52	108	4	Glycyrrhiza	N/A	Topical	2% licorice topical gel was more effective than 1% in reducing the scores for erythema, edema and itching over two weeks (p<0.05).	None noted.
Staab et al57	1010	3	AD education	6 sessions/weekly	N/A	Significant improvements in the severity of eczema and subjective severity were seen in all intervention groups compared with control groups. Parents of affected children aged less than 7 years experienced significantly better improvement in all five quality-of-life subscales, whereas parents of affected children aged 8–12 years experienced significantly better improvement in three of five quality-of-life subscales.	No noted side effects.
Schachner et al58	20	3	Massage	20 mins/daily	N/A	Over 1 month, parents of massaged children reported lower anxiety levels in their children, and the children improved significantly on all clinical measures including redness, scaling, lichenification, excoriation, and pruritus. The control group only improved significantly on the scaling measure.	No noted side effects.
Rotter et al59	121	4	Acupuncture	30 min sessions/biweekly	N/A	The adjusted mean SCORAD was similar in all groups. The study found a clinically relevant difference in visual analog scale (VAS) itching between the Acupuncture (ACU) group and the control group (CG). The response measured by VAS itching analysis favored ACU and Osteopathic Medicine over CG.	No noted side effects.
Park et al60	36	5	Acupuncture	2x weekly	N/A	The mean SCORAD score substantially decreased 2 weeks after starting the acupuncture treatment and continued to improve for at least 4 weeks after the end of the treatment compared to the sham group (each P <.0001).	No serious adverse events were noted.
Kang et al61	30	3	Acupuncture	3x weekly	N/A	The SCORAD, VAS (Pruritus), VAS (Insomnia), POEM, DLQI, and EASI were significantly improved in the acupuncture groups. There was no significant difference between acupuncture Groups 1 and 2 in all the main evaluation indices.	8 cases of side effects: contusion (n=3), diarrhea (n=2), cold (n=2), and incised wound of the right index finger (n=1).
Schöni et al62	32	4	Bioresonancee	N/A	N/A	BIT had no significant additive measurable effect on the outcome variables determined in this study.	No noted side effects.

TABLE 4.

CAM therapies for psoriasis

REFERENCE	N	JADAD SCORE	CAM	DOSAGE	ROA	STUDY RESULTS	SIDE EFFECTS
Yu et al65	211	3	Zinc	N/A	Topical	The margin of superiority between the experimental and placebo groups was >0 [96% confidence interval (CI) 13.22% (0.43% to ∞)] in the FAS. The experimental group was superior to the placebo group. The noninferiority margin between the experiment and control groups was >-15% [96% CI –1.43% (–14.91% to ∞)] in the FAS. The experimental group was not inferior to the control group.	No adverse effects were reported.
Serwin et al66	22	5	Selenium	200 micro g/2x daily	Oral	In the treatment group, in which the PASI score throughout the study was higher than that in the placebo group, the pretreatment concentration of sTNF-R1 was significantly higher than that in controls. The study found that sTNF-R1 declines with an improvement in skin lesions. The dynamic of the decline was different in both groups: in patients receiving SeMet, the concentration of sTNF-R1 after 4 weeks remained higher than that in controls.	No adverse effects were reported.
Siddiqui et al67	50	5	Vitamin D	1 micro g/daily	Oral	Slight improvement (+1) was observed in 9 (45%) of the 20 patients taking Alpha-calcidol and in 8 (38.2%) of the 21 patients taking placebo. The difference is not statistically significant (chi-squared test). Treatment was ineffective in 11 (55%) of the 20 patients taking Alpha-calcidol and in 13 of the 21 patients (61.1%) receiving placebo.	No adverse effects were reported.
Jenssen et al68	122	5	Vitamin D	100,000 IU, loading dose, followed by 20,000 IU/week	Oral	There was no significant difference in change in PASI score between the groups. There was no significant difference in change in Physician Global Assessment score, self-administered PASI, or Dermatology Life Quality Index between the groups.	No adverse effects were reported.
Stucker et al69	13	3	Vitamin B12	N/A	Topical	There was a more rapid development of beneficial effects with the use of calcipotriol in the initial 8 weeks, although differences in effects were significant only at the time point of therapy Week 8 (p<0.05). After 12 weeks, neither the PASI score nor 20-MHz sonography showed significant differences between the two treatments.	No adverse effects were reported.
Syed et al70	60	4	Aloe vera	100g/3x daily/5 days/weekly	Topical	The Aloe vera extract cream had cured 25/30 patients (83.3%) compared to the placebo cure rate of 2/30 (6.6%) (P<0.001) resulting in significant clearing of the psoriatic plaques (3 28/396 (82.8%) vs placebo 28/366 (7.7%), P<0.001) and a decreased PASI score to a mean of 2.2.	Patients complained of no adverse symptoms or other side effects.
Paulsen et al71	41	5	Aloe vera	2x daily	Topical	The score sum of erythema, infiltration, and desquamation decreased in 72.5% of the Aloe vera-treated sites compared with 82.5% of the placebo-treated areas from Week 0 to Week 4, which was statistically significant in favor of the placebo treatment (P=0.0197).	55% of the patients reported local side effects, mainly skin dryness in test areas.
Pandey et al73	200	5	Mahonia aquifolium	2x daily	Topical	The results indicate statistically significant (P<0.05) improvements in PASI and QLI in the Mahonia-treated group, compared to control group.	The side effects reported were infrequent, <1%, and minor; the most frequent side effects were rash, a burning sensation.
Bittiner et al74	28	3	Fish oil	10 capsules/daily	Oral	After 8 weeks' treatment there was a significant lessening of itching, erythema, and scaling in the active treatment group, with a trend towards an overall decrease in body surface area affected.	No adverse effects were reported.
Collier et al75	18	3	Fish oil	170 g/daily or white fish/daily	Oral	The oily fish but not the white fish diet led to a modest clinical improvement (11% and 15%, P<0.01) which was accompanied by a rise in plasma eicosapentaenoic acid (20:5n-3) concentrations.	No adverse effects were reported.
Danno et al76	40	3	Fish oil	1800mg/daily	Oral	Better and more rapid improvement was obtained with the combination therapy for 12 weeks than with low-dose etretinate monotherapy.	Adverse reactions were noted in 7 patients with ReEPA therapy and in 8 patients with etretinate monotherapy. They included cheilitis, dry mouth, dry eyes, desquamation of palms, folliculitis, gastric symptoms, and increased liver enzyme levels.

Grimminger et al77	20	4	Fish oil	2.1 mL/daily (n-3)	IV	The severity of disease markedly decreased in all patients of the n-3 group, with improvements in all score systems ranging between 45% and 76% within 10 days (P<0.05 for each variable). The difference in response to the n-3 and n-6 regimens was evident within 4–7 days after the onset of lipid infusion.	No adverse effects were reported.
Mayser et al78	83	4	Fish oil	200 mL/daily	IV	The total PASI score decreased by 11.2 +/– 9.8 in the omega-3 group and by 7.5 +/–8.8 in the omega-6 group (p=0.048). In addition, the omega-3 group was superior to the omega-6 group with respect to change in severity of psoriasis per body area, change in overall erythema, scaling, infiltration, overall assessment by the investigator, and self-assessment by the patients.	In the ω-3 group, the most reported adverse event was superficial thrombophlebitis (n=2).
Soyland et al81	145	5	Fish oil	6g/daily	Oral	The PASI score, as evaluated by the physicians, did not change significantly during the trial in either group. This was also true of the total subjective score reported by the patients. Scaling was reduced from baseline in both groups (P <0.01).	No adverse effects were reported.
Escobar et al82	25	3	Fish oil	6 hrs/daily	Topical	The results showed statistically significant improvement in erythema and scaling for both treatments compared to basal values; significant differences between treatments were achieved in scaling but not in erythema.	Irritation and a burning sensation were reported in the treatment group (n=1). This adverse effect reverted after completing the treatment.
Dawe et al83	60	3	Balneophototherapy	Weekly	N/A	The mean area under the psoriasis severity-time curves during treatment was not detectably lower with Dead Sea salt balneophototherapy than with NB-UVB monotherapy (P=0.099). There was no detectable difference in times to relapse.	No adverse were effects reported.
Brockow et al84	143	3	Balneophototherapy	3x weekly	N/A	Patients allocated to the treatment group attained a statistically significantly higher rate of PASI-50 at the end of the intervention period than patients allocated to UVB (control) [58/79 (73%) vs. 32/64 (50%); P=0.01; NNT, 4.3, 95% CI, 2.4–18.1]. The benefit persisted until 3 months only for one of the two secondary outcomes considered.	No adverse effects were reported.
Brockow et al85	160	3	Balneophototherapy	3x weekly	N/A	Participants allocated to HC-SSW-UVB attained a statistically significantly higher rate of PASI-50 than patients allocated to UVB only (68/79 [86%] versus 38/71 [54%]; p<0.001; number needed to treat, 3.1; 95% CI, 2.1– 6.0). Postintervention analysis did not yield a clear hint of a persisting effect.	No adverse effects were reported.
Schiener et al86	1241	3	Balneophototherapy	4x weekly	N/A	Patients who received SW UV-B or bath PUVA had a significantly higher incidence of therapeutic success than did patients treated with TW UV-B (74.9% vs 60.7%)	No adverse effects were reported.
Leaute-Labreze et al88	90	3	Balneophototherapy	Daily	N/A	For spa water therapy, change in PASI was −29%, thus showing a minor therapeutic effect of saline spa water alone and poor efficacy compared with uv-B therapy and combined spa water and uv-B therapy (P<0.001).	Skin irritation with SSW (n=3), pulmonary infection (n=1), itching, and skin burning.
Gambichler et al89	10	3	Balneophototherapy	4x weekly	N/A	A highly significant (P<0.001) decrease in the clinical baseline score was observed after 30 treatments; however, there was no significant (P >0.5) difference in clearance of the psoriatic lesions between the sites soaked in salt water and tap water.	No adverse events were observed.
Beylot-Barry et al91	260	3	Balneotherapy	3-week spa therapy	N/A	VQ-Dermato scores and pruritus VAS significantly improved. Outcomes at the 12-month follow-up of the immediate spa therapy group showed persistent improvement of DLQI, VQ-Dermato, and pruritus.	No adverse effects were reported.
Gaston et al92	18	3	Meditation	N/A	N/A	A time-series multivariate regression analysis, revealed a significant difference between the mean psoriasis ratings of treatment and control groups after treatment (partial r=0.30, p less than 0.01). This also indicated an average positive correlation between the severity of psoriasis symptoms and psychological distress (partial r=0.31, p less than 0.01) and the impact of adverse life events (partial r = 0.23, p less than 0.05).	No adverse effects were reported.

Kabat-Zinn et al93	37	3	Meditation	N/A	N/A	Cox-proportional hazards regression analysis showed that subjects in the tape groups reached the Halfway Point (p=0.013) and the Clearing Point (p=0.033) significantly more rapidly than those in the no-tape condition, for both UVB and PUVA treatments.	No adverse effects were reported.
Zachariae et al94	25	3	Psychotherapy	7 sessions	N/A	When analyses were performed for both groups separately, the treatment group displayed significant reductions for all three psoriasis activity measures, whereas no changes were seen in the control group.	No adverse effects were reported.
Keinan et al95	32	3	Biofeedback	N/A	N/A	No significant differences were found between the groups regarding the degree of improvement relative to the start of the study.	No adverse effects were reported.
Tausk et al96	11	3	Hypnosis	N/A	N/A	Highly hypnotizable subjects showed significantly greater improvement than did moderately hypnotizable subjects, independent of treatment group assignment (active suggestion or neutral hypnosis)	No adverse effects were reported.

TABLE 5.

CAM therapies for contact dermatitis

REFERENCE	N	JADAD SCORE	CAM	DOSAGE	ROA	STUDY RESULTS	SIDE EFFECTS
Abrams-Motz et al98	40	3	Jewelweed	150ml of the plant-infused soap	Topical	Jewelweed mash was effective in reducing poison ivy dermatitis, supporting ethnobotanical use. However, jewelweed extracts were not effective; and soaps made of these extracts were effective but no more so than jewelweed-free soaps.	No adverse effects were reported.
Long et al99	10	5	Jewelweed	N/A	Topical	There was no statistically significant difference in the objective scores at the sites treated with jewelweed extract versus the distilled water control sites.	All subjects developed dermatitis at each patch test site.

TABLE 6.

CAM therapies for acne vulgaris and rosacea

REFERENCE	N	JADAD SCORE	CAM	DOSAGE	ROA	STUDY RESULTS	SIDE EFFECTS
Acne vulgaris
Enshaieh et al102	60	5	Tea tree oil	N/A	Topical	There was a significant difference between tea tree oil gel and placebo in the improvement of the total lesions counted (TLC) and also regarding improvement of the Acne Severity Index (ASI). In terms of TLC and ASI, tea tree oil gel was 3.55 times and 5.75 times more effective than placebo respectively.	Side effects with both groups were relatively similar and tolerable.
Lubtikulthum et al103	77	3	Tea tree oil	1g/2x daily	Topical	At the 12-week point, the acne lesion counts decreased, with statistically significant differences from the baseline values in both groups and for all types of acne (P-value <0.001). There was no statistically significant difference in terms of porphyrin counts, spot scores, the Dermatology Life Quality Index, or satisfaction with efficacy between the groups.	The most common side effect was skin irritation.
Bassett et al104	124	3	Tea tree oil	N/A	Topical	Both 5% tea-tree oil and 5% benzoyl peroxide had a significant effect in ameliorating the patients' acne by reducing the number of inflamed and non-inflamed lesions (open and closed comedones), although the onset of action in the case of tea-tree oil was slower.	Fewer side effects were experienced by patients treated with tea tree oil.
Hunt et al105	150	4	Alpha hydroxy acids	N/A	Topical	Both gluconolactone and benzoyl peroxide had a significant effect in improving patients' acne by reducing the number of lesions (inflamed and non-inflamed).	Fewer side effects (like dryness) were experienced with gluconolactone treatment.
Kim et al106	26	3	Glycolic acid	70% solution biweekly	Topical	Acne grading of both treatments improved after 3 treatment sessions.However, there were no significant differences in treatment effects between the 2 methods.	No significant adverse events were reported.
Spellman et al107	70	4	Glycolic acid	2x daily	Topical	Patients treated with azelaic/glycolic acid experienced a significantly greater reduction in the number of papules, as well as a greater reduction in the number of inflammatory lesions, than those treated with tretinoin.	There were no reported adverse events.
Abels et al108	120	5	Glycolic acid	Daily	Topical	Acne improved significantly in the verum group up to Day 90. Already at Day 45, there was a statistically significant (5% level) difference against placebo.	There's no statistically significant difference between verum and placebo regarding adverse events.

Orafidiya et al109	126	3	Basil (Ocimum) oil	0.25cm3/2x daily	Topical	Ocimum oil preparations containing 2% oil in alcohol (ethanol) and in a cetomacrogol blend base were more effective and reduced lesions faster than 10% benzoyl peroxide.	Burning and other mild skin irritation side effect events reported in some of the preparations.
Orafidiya et al110	84	2	Basil oil and aloe vera	2% Ocimum oil with graded (0–100%) of aloe gel/2x daily	Topical	The efficacy of the Ocimum oil increased with increasing aloe gel contents. Products formulated with the undiluted or 50% aloe gels were most active and resolved inflammatory lesions faster than the standard product. The aloe gel alone showed minimal activity.	Mild skin irritation.
Lu et al113	80	5	Green tea extract	1,500mg of decaf GTE daily	Oral	Statistically significant differences were noted in inflammatory lesion counts distributed on the nose, periorally, and on the chin between the two groups. However, there were no significant differences between groups for total lesion counts.	Mild constipation (n=1) and abdominal discomfort (n=2) after GTE treatment.
Acne Rosacea
Sharquie et al118	25	4	Zinc sulfate	100mg/3x daily	Oral	The treatment group was effective in controlling rosacea lesions. The effect of zinc sulfate on the papules and pustules became statistically significant 1 month after starting therapy, which continued for 3 months after shifting to placebo in Group A.	Mild gastric upset in 3 (12%) patients on zinc sulfate.
Bamford et al119	44	5	Zinc sulfate	220mg/2x daily	Oral	Rosacea improved in both groups. There were no differences in magnitude of improvement based on rosacea severity scores between subjects receiving zinc sulfate and subjects receiving placebo (P=0.284)	Stomach upset, including nausea, discomfort, gas, diarrhea, constipation, and increased bowel movements [n=12 (44%) zinc vs. n=8 (32%) placebo]
Ebneyamin et al128	47	4	Tea tree oil	2x daily	Topical	Clinical features and global assessments showed papules, pustules and nontransient erythema had improvement in the drug group after 12 weeks (P values <0.05). The improvement of burning and stinging and dry appearance was greater than the placebo gel (P value <0.05).	Itching in the placebo group was significantly more than the treatment group (P=0.002).

TABLE 7.

CAM therapies for alopecia areata and androgenetic alopecia

REFERENCE	N	JADAD SCORE	CAM	DOSAGE	ROA	STUDY RESULTS	SIDE EFFECTS
Alopecia Areata
Hay et al129	86	4	Aromatherapy	Daily	Topical	19 (44%) of 43 patients in the active group showed improvement compared with 6 (15%) of 41 patients in the control group (P=0.008).	No adverse effects were reported.
Hajheydari et al131	40	4	Garlic gel	2x daily	Topical	At the end of the treatment, good and moderate responses were observed in 19 (95%) and one (5%) patients of the case group respectively, which was significantly better than the control group (P=0.001).	No adverse effects were reported.
Androgenetic Alopecia
Kessels et al134	396	5	Dabao	1.7mL/2x daily	Topical	In both the Dabao and placebo groups an increase in the amount of hair was observed; Regarding cosmetic effect, 42% of the participants in the Dabao group and 37% in the placebo group reported positive results.	n=1 developed a contact dermatitis for Dabao after 3 months. Some others reported isolated inflamed follicles in both groups.
Takahashi et al135	29	3	Apple	2 mins/daily	Topical	In the treatment group, 78.9% showed an increased mean value of hair diameter, whereas only 30.0% in the placebo group showed any increase. The increased ratio of hairs measuring more than 40 microm in diameter after 4 months of treatment was significantly higher than the placebo controls.	No adverse side effects were observed in either group.

TABLE 8.

CAM therapies for melasma

REFERENCE	N	JADAD SCORE	CAM	DOSAGE	ROA	STUDY RESULTS	SIDE EFFECTS
Lowe et al145	52	3	Azelaic acid	2x daily	Topical	Following a 24-week treatment period, azelaic acid produced significantly greater decreases in pigmentary intensity than the control as measured by both an investigator's subjective scale (P=0.021) and a chromometer analysis (P=0.039). There was a significantly greater global improvement with azelaic acid than control at Week 24 (P=0.008).	Burning and stinging.
Kakita et al146	65	3	Azelaic acid and glycolic acid	2x daily	Topical	The present study demonstrated that the combination of azelaic acid 20% cream and glycolic acid 15% or 20% lotion was as effective as hydroquinone 4% cream in the treatment of hyperpigmentation in darker-skinned patients.	Higher dryness and burning scores.
Baliña et al147	329	3	Azelaic acid	N/A	Topical	Over the treatment period the azelaic acid cream yielded 65% good or excellent results. No significant treatment differences were observed with regard to overall rating, reduction in lesion size, and pigmentary intensity.	A higher incidence of local irritation (burning, itching, scaling) in the azelaic acid group. Severe side effects such as allergic sensitization or exogenous ochronosis were not observed with azelaic acid.
Dayal et al149	60	3	Azelaic acid and glycolic acid	2x daily	Topical	The improvement in MASI and percentage decrease in MASI scoring were statistically significant 12 weeks onwards in the study group as compared to the control group. There was also a significant reduction in MELASQOL scores in the study group as compared to the control group after treatment.	For treatment group: post-peel erythema (n=4); pruritis (n=4); Burning (n=2); postinflammatory hyperpigmentation (n=2); Scaling (n=5). All events were reversible.
Farshi et al150	29	4	Azelaic acid	2x daily	Topical	One month after treatment, the mean MASI score decreased to 6.7 ± 3.4 with hydroquinone and 6.3 ± 3.4 with azelaic acid with no significant difference between them (t-test, CI 95% = –2.2 to 3). After 2 months' treatment, the MASI score was 6.2 ± 3.6 with hydroquinone and 3.8 ± 2.8 with azelaic acid, a significant statistical difference (t-test, CI 95% = 0.03–4.9).	For treatment group: Erythema (n=1); irritation (n=5); pruritus (n=1).
Lim et al152	40	2	Kojic acid and glycolic acid	N/A	Topical	All patients showed improvement in melasma on both sides of the face. The side receiving the kojic acid did better. More than half of the melasma cleared in 24/40 patients receiving kojic acid compared to 19/40 patients receiving placebo gel.	No adverse events were reported.
Garcia et al153	39	2	Kojic acid and glycolic acid	N/A	Topical	51% of the patients responded equally to hydroquinone and kojic acid. 28% had a more dramatic reduction in pigment on the kojic acid side; whereas 21% had a more dramatic improvement with the hydroquinone formulation. These results were not statistically different.	The kojic acid treatment was found to be more irritating.
Espinal-Perez et al156	16	3	Vitamin C	Daily	Topical	The best subjective improvement was observed on the hydroquinone side with 93% good and excellent results, compared with 62.5% on the ascorbic acid side (P<0.05); however, colorimetric measures showed no statistical differences.	Side effects, like irritation, were present in 68.7% (11/16) with hydroquinone vs. 6.2% (1/16) with ascorbic acid.
Soliman et al157	30	2	Vitamin C and trichloroacetic acid (TCA) peel	Daily	Topical	The treatment group compared with the control group showed a significant decrease in MASI score at the end of the 16-week follow-up period (P<0.003). Global evaluation showed that 13 patients (87%) in the treatment group improved or maintained their improvement compared to n=10 from the control group.	No adverse events were reported.
Huh et al158	29	3	Vitamin C	N/A	Topical	12 weeks after iontophoresis, the colorimeter of the treated site showed a significant decrease in the L value (from 4.60 to 2.78, p=0.002), compared to that of the control site (from 4.45 to 3.87, p=0.142).	No adverse events were reported.
Amer et al159	20	2	Liquiritin (licorice)	Daily	Topical	At Week 4, a reduction of pigmentary intensity by one to three levels was achieved on the side of the face treated with liquiritin cream; 16/20 (80%) showed an excellent response with no difference between the previously pigmented area and the normal skin surrounding the area; 2 patients (10%) showed a good response, and 2 patients (10%) showed a fair response with moderate pigmentation.	Mild skin irritation in the form of erythema and a burning sensation was observed in two patients (20%) treated with liquiritin cream.

Costa et la160	56	3	Emblica, licorice, and belides	Daily	Topical	Depigmentation was observed through medical evaluation (Treatment group (A): 78.3%; Control group (B): 88.9%) and volunteers' self-evaluation (Group A: 91.3%; Group B: 92.6%); these results were statistically significant (p <0.001), with no differences between groups (p >0.05).	2 volunteers in the treatment group and 7 in the control group had mild skin adverse events (mild erythema, burning.).
Morag et al161	102	4	β-arbutin	2x daily	Topical	The extract causes decreased levels of melanin in the skin pigmentation spot. Clinical effect in the form of lightening and evening skin tone on the discolored side was observed in 75.86% of the female patients with melasma.	No adverse events were reported.

Stevens-Johnson’s Syndrome. A few case reports describe cases of Stevens-Johnson’s Syndrome (SJS) thought to result from CAM exposure. One patient suffered SJS with diffuse bullae and erosions over her entire body with erythema, swelling, fever, and liver dysfunction following consumption of a commercial “health drink” containing Ophiopogonis tuber, an herb used in TCM for a variety of maladies.¹⁷⁶ Another case report described erosions and blistering over mucus membranes throughout the body following consumption of a multi-component herbal tablet taken as a general health tonic with recurrence upon re-exposure.¹⁷⁵ Both patients recovered with discontinuation of CAM treatment and systemic steroid therapy.

Photosensitivity. Potential photosensitivity may be a risk for a few CAM therapies. Psoralens contained in some aromatherapy preparations may lead to photosensitivity and burns. A case report detailed a patient who suffered 70 percent partial thickness burns after sunbed exposure preceded by a bath with bergamot oil (constituent 5-methoxypsoralen).¹⁷⁷ Another report describes extensive rash and blister formation resulting from sun exposure following application of juice from kaffir limes (Citrus hystrix) as an insect bite treatment.¹⁷⁸ Psoralens are found in many members of the citrus family (Rutaceae) as well as members of the carrot/parsley, fig, and legume/pea families (Umbelliferae, Moraceae and Leguminosae, respectively), and phytophototoxicity should be a noted risk with use of CAM therapies containing these agents.¹⁷⁸ Finally, several reports have documented the photosensitivity risk of St. John’s Wort (Hypericum perforatum) used for depression and other psychological and non-psychological diseases. Cases document pigment changes, itching, erythema, and neuropathy (made evident by stinging pain) in sun-exposed areas following ingestion of St. John’s Wort, likely caused by photoactivity of its constituent hypericins.^162,179

Bruising, lesion formation, cupping injuries. Many of the more physically oriented CAM therapies can result in traumatic skin injuries, bruising, and lesion formation. Coin rubbing, a Chinese and Vietnamese practice of firmly rubbing spoons or worn coins over lubricated skin to treat a variety of maladies, results in parallel bands of petechial bruising on patient skin.¹⁸⁰ There is also the possibility of systemic intoxication effects from absorption of toxic constituents like camphor in the lubricant.¹⁸¹ Cupping is another Asian therapy used to treat a wide variety of conditions including pain disorders, cervical spondylosis, herpes zoster, and acne; this involves the application of suction to the skin with a cup, leaving circular patches of petechiae and ecchymosis.¹⁸² The technique sometimes involves burning cotton to create a vacuum or incision of the skin (“wet cupping”). Complications have been reported with cupping, mostly attributed to improper or overzealous application of the technique and include crusting bullae formation,¹⁸³ panniculitis,¹⁸⁴ and even second-degree burns,¹⁸⁵ but it is generally considered a safe procedure. Finally, several reports describe a variety of adverse dermatologic presentations resulting from acupuncture therapy including blue nevus-like macules diagnosed as localized argyria, waxy orange-brown papules indicative of lymphocytoma cutis, amongst others, all at acupuncture sites and attributed to the procedure.³¹

Other diseases, relevance to peri-operative period and steroid contamination. Other dermatologic conditions of note resulting from CAM include alopecia with associated polyneuropathy from thallium contained in a Chinese herbal remedy¹⁸⁶ and erythema nodosum resulting from oral echinacea supplements.³¹ A few reports have been published concerning Sweet’s syndrome resulting from CAM exposure in patients. A leukemia patient with progressing necrotic lesions on the face and legs and fever was ultimately diagnosed with Sweet’s syndrome attributed to the use of a cream with 1.5% arnica, a much higher than homeopathic dose.¹⁸⁷ A separate report describes Sweet’s syndrome as likely resulting from extensive exposure to jalapeno peppers and their constituent capsaicin,¹⁸⁸ a CAM modality used primarily for neuropathic pain (Table 2).

TABLE 2.

Summary of CAM cutaneous side effect profiles on dermatologic and non-dermatologic conditions

CAM AGENT	ORIGINAL CAM INDICATION
Contact dermatitis (select agents of note)
Metals (e.g. nickel) in TCM	AD, others
Metals (e.g. zinc) in acupuncture needles	Multiple
Aromatherapy essential oils	Multiple
Mercury contamination (homeopathy, TCM, Indian ayurvedic)	Coughing, others
Echinacea	Immune stimulant
Chamomile	Anti-inflammatory
Turmeric	Anti-inflammatory
Tea tree oil	Acne, infections, others
Other Dermatitides
Metal contamination in homeopathy—AD exacerbation	Multiple
Dead sea salt—Erythrodermic psoriasis flare-up	Psoriasis
Stevens-Johnson Syndrome
Ophiopogonis tuber (TCM)	“Health drink”
References (Mochitomi, Monk)
Multi-component herbal tablet	“Health tonic”
Photosensitivity
Aromatherapy oils (bergamot; constituent psoralens)	Multiple
Citrus (Rutaceae), carrot/parsley, fig, legume/pea	Insect bites, multiple others
St. John’s Wort (constituent hypericins)	Depression, other psychological
Bruising, lesions, skin injuries
Coin rubbing–Petechiae bruising	Multiple
Cupping–petechiae/ecchymosis, bullae, panniculitis, burns	Pain, spondylosis, herpes zoster, others
Acupuncture—Argyria, lymphocytoma cutis, other	Multiple
Alopecia (with polyneuropathy)
Thallium in Chinese herbal remedy	Multiple
Erythema nodosum
Echinacea	Immunostimulator, others
Sweet’s Syndrome
Arnica; capsaicin (jalapenos)	Multiple; pain, others

Finally, as already mentioned, contamination of some topical and oral TCM agents with corticosteroids may be widespread. Analyses have reported the percentage of TCM agents contaminated with steroids to be as high as ~25 percent with steroid concentrations as high as 1500mg/g (mean 456mg/g).^32,189 This presents obvious potential for interference with treatment for many dermatologic disorders and may necessitate adjustments of the current treatment regimen.

CONCLUSION

The use of CAM agents has become widespread in dermatology. Driven by positive anecdotal evidence and promising in-vitro or animal findings, investigators have aimed to transition these CAM agents from folk medicine to scientific medicine through placebo-controlled, double-blind human RCTs. However, this effort has not been widely successful, as most studies concluding CAM benefit suffer from major methodological problems that impact the credibility of their claims. As documented throughout this revie, CAM studies often suffer from: poor or no blinding, subjective outcomes, inadequate randomization, and selective reporting of positive outcomes while ignoring negative or null findings. Despite these reservations, some study results and plethora of in-vitro and animal evidence suggest that many CAM agents deserve further investigation via well-crafted RCTs. Moving forward, a more useful focus would be comparing CAM treatments to common or effective conventional therapies, either alone or in an integrative approach, as this would accurately mimic how patients typically use these agents. In the meantime, given the prevalence of CAM use, practicing dermatologists should be aware of purported or demonstrated efficacy from available studies and be vigilant for adverse effects of popular CAM agents.