Introduction
Methods
Inclusions
Results
Reason for exclusions | Number |
---|---|
Dentofacial surgical correction/Le fort osteotomy orthognathic | 14 |
Psychological effects of beauty/personality and beauty/brain effects on beauty | 12 |
Inappropriate for other reasons | 8 |
Orbital surgery/ear placement in reconstruction | 6 |
Skeletal analysis | 5 |
Cleft lip palate and surgery | 3 |
Adolescent or child after manual records reviewed | 3 |
Cancer surgery | 3 |
Cosmetic or cosmeceuticals | 3 |
Comparison of different fillers | 2 |
Endoscopic lifting surgery | 2 |
Burns victims/trauma victims | 2 |
Qualitative measurements of facial aesthetic outcomes | 2 |
DNA forensic analysis | 2 |
Portrait painting theories | 2 |
Cadaver | 1 |
Total | 70 |
Article | Year of study | Sample size | Measures rating beauty | Type of study/level of evidence | Outcome | Comments |
---|---|---|---|---|---|---|
Liu et al. [12] | 2017 | 360 | Distances and angles | Computation based on photographs level III | Measurements do not have a normal distribution, no constant relationship of proportionality | An in-depth mathematical analysis of distances and angles |
Heidekrueger [13] | 2017 | 1011 | Lip ratio preference | Survey level IV | Lip ratio of 1.0:1.0 was most attractive | Survey of surgeons’ preference |
Koidou et al. [14] | 2017 | 193 | Angulation of smile | Case control level III | Smaller mean angulation of smile more aesthetically pleasing | |
Jang et al. [61] | 2017 | 93 | Measurements from three-dimensional sampling | Case control level III | Longer face smaller lower lip and chin preferred. deviation from golden ratio | Korean population |
Popenko et al. [15] | 2017 | 20 digital images altered to create 100 faces | Lip surface area and lower/upper lip ratio | Survey level IV | 53.5% increase in surface area and 2:1 ratio of lower to upper lip more attractive | Age 18–25 white female faces |
Benslimane et al. [16] | 2017 | 450 photos 1000 portraits 339 patient photos | Eye fissure frame ratio or ‘Frame concept’ | Cross-sectional level IV | Frame height is inversely proportional to attractiveness and narrower eye fissure frame more attractive | Novel idea of ‘Frame concept’ |
Melo et al. [17] | 2017 | 30 | Harmony of features | Cross-sectional level IV | Subjective influence on assessment of attractiveness | Subjective facial analysis criteria used. photographs rated by 50 evaluators |
Kaipainen et al. [18] | 2016 | 59 | Effect of regional facial asymmetry on attractiveness | Observational level IV | Attractiveness not influenced by asymmetry | Age group 16–25 |
Hwang et al. [19] | 2016 | 120 | Relative eyebrow width/relative medial midpupilary and lateral heights of eyebrows to length of palpebral fissure measure over last century from photographs in Vogue magazine | Observational level IV | REW unchanged RLH greater than REW over time | Cross cultural difficult to compare |
Galantucci et al. [20] | 2016 | 66 | 25 anatomical landmarks total of 5610 data items | Cross-sectional level IV | Greatest influences on attractiveness are facial width, upper facial convexity; distance between nasion and midpoint of tragi; nasolabial angles and mouth width | Three-dimensional anthropometric analysis to set up a database statistically significant differences only in some measurements. |
Heidekrueger et al. [13] | 2016 | 1011 | Lip shape preference | Survey level IV | Non-caucasian surgeon prefer larger lips and caucasian surgeons prefer smaller lips | 14% response rate |
Murakami et al. [21] | 2016 | 9 morphed facial types | Lip position | Observational level IV | Favoured lip position differed between lay person and clinician | Japanese population—limited to specific ethnicity |
Bagheri et al. [62] | 2016 | 200 | Lip morphology | Case control level III | Medium and full lip preference in males and medium and thin preference in females | Anatolian females computer-assisted redesign solution for lip augmentation |
Tauk et al. [22] | 2016 | 18 | Visual Analogue Scale | Cross-sectional level IV | Entire face profile used to assess beauty | |
Forte et al. [29] | 2015 | 66 | Attractiveness and tiredness on a 0–10 scale with digital alteration of facial subunits | Survey level IV | Neck ptosis, jowels, vertical lip rhytids, crows’ feet lower lid herniation influenced perception of age | Perception of tiredness and attractiveness extrapolated from impact on age |
Alam et al. [60] | 2015 | 286 | Comparison to golden ration | Cross-sectional level IV | Only 17.1% conform to the ratio. 54% have shorter face. No association between golden ratio and facial evaluation scores | Malaysian population |
Gibelli et al. [23] | 2015 | 40 | Lip measurements and differences in gender and age | Cross-sectional level IV | Male lips larger than female. Younger people have larger lips than older. Lower lip thickness highest percentage if correct for age | Three-dimensional technology used for morphological and metrical analysis |
Penna et al. [24] | 2015 | 176 | Lip morphology | Cross-sectional level IV | High ratio of upper vermillion height to mouth–nose distance and chin–nose distance in and wider vermillion height/chin–mouth distance in attractive females | 250 voluntary judges through an Internet presentation |
Wu et al. [25] | 2015 | 80 patients 50 landmarks | Facial characteristics | Case control level III | Attractive men had large forehead reduced mandible round baby face characteristics | Consider individual faces—Chinese population |
Farrera et al. [26] | 2015 | 565 patients | Asymmetry | Cross-sectional level IV | Attractiveness and asymmetry are not correlated | Use two-dimensional digital photographs and geometric morphometric methods Mexican population |
Bronfman et al. [41] | 2015 | 13 studies | Facial distances, angles and features | Systematic review of level III trials level III | Japanese adults have less bilabial protrusion, less prominent nose. Japanese adults prefer a more retruded profile | Used some skeletal measurements |
Hwang et al. [19] | 2015 | 37 | Eye measurements | Cross-sectional level IV | Beautiful women and femme fatales have same inter-pupillary distance | Western society |
Hwang et al. [27] | 2014 | 31 | 43 distances and angles in young and old Leonardo’s profile drawings | Cross-sectional level IV | 39 anthropometric items did not differ. Upper lip height, upper face height and nasolabial angle greater in young. | Comparing old and ‘ugly’ with young and beautiful |
Park et al. [28] | 2013 | 52 | 17 anthropometric ratios | Observational level IV | Femme fatales had narrow noses and attractive midface | Comparison of portrait paintings |
Rosetti et al. [30] | 2013 | 400 | Facial distances | Observational level IV | Eye–mouth distance/height of mandible ratio influenced by attractiveness. Most facial ratios differ from golden ratio | Three-dimensional facial distances used |
Wong et al. [31] | 2010 | 197 | Lip measurements and subjective assessment of attractiveness in different ethnicities | Observational level IV | Smaller than average in midline upper lip rated more attractive. Ethnic differences | Three-dimensional facial distances used. Lips did not contribute to attractiveness as much as previously thought |
Pancherz et al. [32] | 2010 | 158 | 5 transverse and 7 vertical measures compared with PHI | Observational level IV | Attractive individuals have proportions close to PHI | Testing Ricketts’ hypothesis |
Pallett et al. [33] | 2010 | 122 raters | Eye mouth distance intraocular distance | Survey level IV | Vertical distance between eyes and mouth = 36% of length horizontal distance between eyes is = 46% of width | Attempt to redefine ‘new’ golden ratio |
Komori et al. [34] | 2009 | 114 | Averageness and symmetry | Observational level IV | Males and females both averageness and symmetry rate positive, whereas in female only averageness does | |
Jahanbin et al. [35] | 2008 | 50 | 5 landmarks 5 ratios | Cross-sectional level IV | Only some measures conform to the divine proportion | Use two-diemensional digital photographs |
Holland [36] | 2008 | 0 | Analysis of the Marquardt’s mask | Observational level IV | Marquardt’s mask described as ‘not ideal’ | |
Medici et al. [37] | 2007 | 20 digital images | Ratios of facial features rated by 12 individuals | Survey level IV | A relationship exists between divine proportion and aesthetic face | Manipulation of ratios by morphing from 2.0 to the divine ratio |
Danel et al. [49] | 2007 | 77 | Eye mouth angle | Observational level IV | Attractiveness negative correlation to EME | |
Kim et al. [38] | 2007 | 40 | Rating of pre and post-operative photographs with the Marquardt mask | Observational level IV | Results not statistically significant but mask a ‘useful’ tool | |
Costa et al. [39] | 2006 | 1065 | Eye lip size and roundness | Case Control level III | Eye and lip roundness, eye height and width and lip height are enhanced in artistic portraits compared to photographic | One part of three studies |
Milutinovic et al. [40] | 2014 | 107 | Facial distances and proportions | Observational level IV | Smaller face/uniformity of thirds and fifths and most parameters meet the ‘ideal proportions’ in aesthetically pleasing faces | |
Gan et al. [54] | 2014 | 307 | Self-taught learning computer based | Cohort level III | Facial beauty can be recognised at a rate 87.3% of face | |
Xie et al. [55] | 2015 | 500 | Benchmarking the SCUT-FBP dataset | Case control level III | Confirming the SCUT-FBP dataset is reliable for predicting attractiveness |
Discussion
Upon reviewing the data, it became evident that the diverse measurement criteria, methodologies used and population types in trials made it difficult to compare data. For example, different measurements from different types of photographic techniques would introduce photographic bias [43]. The trials have at best been of Level III or less for evidence, mainly being cross-sectional studies or observational studies. Despite these difficulties, some common themes were discovered and are highlighted below. These were related to lip analyses, eye measurements, symmetry, ethnicity, automation of analysis and the golden ratio.In the treatment of adults requesting facial aesthetic improvement, is there an evidence-based approach in quantitatively assessing beauty that is useful in everyday aesthetic practice?