EFFICACY AND SAFETY OF CENTELLA ASIATICA 451 DISCUSSION This is the fi rst S.R. an d N.M.R. study on cosmetics that investigated the effect of C. asiatica or its components as the sole active ingredient on wrinkles. The study suggests that C. asiatica or its pentacyclic triterpenoids reduce periorbital wrinkling. Furthermore, C. asiatica may be more effective than P. M. but possibly less than tretinoin. C. asiatica increased skin hydration, which might account for the anti-wrinkle action, whereas tret- inoin had no such action. Were the actions clinical ly useful? For cosmetics, the change in appearance is the most important endpoint. When this was assessed at 4-w intervals by independent blinded experts (dermatologists), or machine reading, wrinkling was consistently reduced in three studies (18,19,22). However, the effect sizes were small, including those by dermatolo- gist assessment in Tangsumroengwong (20) and Tongsrikeaw (21), although the Visioscan has a narrow dynamic range when measuring wrinkles. Ultimately, participants and users need to assess whether they would feel the improved cosmetic appearance is enough to justify the cost of the preparation and the time devoted to twice daily applications, per- haps indefi nitely. Most of the studies provided too little detail about participant self- assessment for users to draw any conclusions about their perceived effi cacy of the treatments. We surmise that all studies were conducted on Asians (Korean, Thais, and Figure 4. Lateral angle (cr o w’s feet) wrinkles measured by Visioscan.

JOURNAL OF COSMETIC SCIENCE 452 Chinese). Even among Asians, wrinkling has some ethnic specifi city (23) thus, the re- sults have limited generalizability. Detailed assessment of sa fety information derived from the studies was patchy. Thus, one study (20) found 10 AEs with C. asiatica extract, whereas another study (21) using the exact same preparation run in the same laboratory reported no AEs. Topical retinoids are well known for AEs a greater safety for C. asiatica was suggested when tretinoin was the comparator (20), although information on AE duration and severity was not reported. Although cellular actions of retinoids are understood, information is less clear for C. asiatica terpenoids. The triterpenes asiatic acid, asiaticoside, and madecassic acid are thought to mediate actions of C. asiatica. Asiatic acid has a bewildering list of molecular actions (24), and transforming growth factor β-receptor 1 kinase-independent Smad activation path- way involving collagen synthesis is a favored candidate for de-wrinkling skin (25). Study quality using the Co chrane Risk of Bias tool suggested minimal bias. However, the depth of reporting was lamentable. All the studies fell well short of the CONsolidated Standards of Reporting Trials 2010 checklist. Inadequate reporting of methodologies made it diffi cult to assess protocol rigor, and in some cases, over-summarizing and pool- ing of data compromised interpretation. How adherence was encouraged and monitored was unclear, and thus would infl uence apparent effectiveness. Reporting of AEs was either nonexistent or inadequate. No study reported severity of AEs and extent, for example, Figure 5. Under eye wrinkles m easured by Visioscan.