journal article
LitStream Collection
The Ideal Thigh: A Crowdsourcing-Based Assessment of Ideal Thigh Aesthetic and Implications for Gluteal Fat Grafting
MD, Emma Vartanian,;PhD, Daniel J Gould, MD,;MD, Ziyad S Hammoudeh,;MS, Beina Azadgoli,;FACS, W Grant Stevens, MD,;FACS, Luis H Macias, MD,
2018 Aesthetic Surgery Journal
doi: 10.1093/asj/sjx191pmid: 29365056
Abstract Background As the popularity of aesthetic gluteoplasty continues to grow, there is renewed focus on defining the ideal buttocks. However, the literature lacks studies characterizing an ideal thigh, despite the impact of thigh contour on overall gluteal aesthetic. Objectives The authors performed the first population analysis of the characteristics of perception of attractive thighs, to identify a role for fat grafting of the thigh in gluteoplasty. Methods Survey images were digitally modified to create thighs of varying widths and angles relative to fixed buttocks. Thigh-to-buttock ratios and the buttock-thigh junction were studied. Data were stratified and analyzed according to age, gender, and ethnicity of the respondents. Amazon Mechanical Turk was used as a novel crowdsourcing platform for surveying aesthetic preferences. Results A total of 1034 responses were included of whom 54.4% were male, and 45.6% were female. All age groups and ethnicities were represented. Overall, 43.8% of respondents preferred the widest buttock-thick junction angle on posterior view. There was no clear preference between larger or smaller thigh-to-hip ratios on lateral view. Conclusions Characteristics of the ideal thigh include wider thighs with greater horizontal projection, creating a more natural contour from the augmented buttock. These findings represent a paradigm shift from the traditionally assumed preference for slender thighs. Plastic surgeons should carefully consider thigh anatomy in their gluteal augmentation patients, as simultaneous thigh augmentation may lead to a more aesthetically pleasing outcome. Further research is needed into best practices and techniques to attain ideal thigh proportions. The female buttocks has been a ubiquitous symbol of femininity throughout history, and as modern trends shift towards more dramatic curves, gluteal augmentation surgery continues to grow in popularity. Over 20,000 buttock augmentation procedures, with implants or fat transfer, were performed in 2016, compared to less than 8000 performed in 2011, a 180% increase in just five years.1,2 Since the first successful augmentation in 1973 by Cocke and Ricketson, using a round implant designed to have more projection than a breast implant,3 there has been ongoing reinvention of the ideal gluteal shape. Free fat grafting4 has become prevalent in gluteal augmentation due to the success seen in breast grafting,5-8 and though questions exist about technique and safety, the procedure is increasing in popularity.9,10 Recently, the efforts of many surgeons in the arena of fat grafting have contributed to a surge in the number of patients undergoing lipogluteoplasty, and to the clinical understanding of successful aesthetic outcomes.11-17 However, while many studies have looked at goals and results in gluteal augmentation,18,19 there is no literature that discusses the role of the thigh contour in overall buttock aesthetic. Enhancing the butt and lateral hip, whether with implants or lipoinjection, adds volume to the buttocks without altering the base of the thigh, producing what some cosmetic surgeons colloquially refer to as “the lollipop deformity” or “marshmallow on a stick.” The result is a disproportionate figure that is viewed as unattractive, and thereby counterproductive to patients’ desires. We hypothesize a broader thigh base would allow a more natural curvature and aesthetically pleasing outcome after gluteal augmentation. Shaping such a base would require augmenting the thigh as well, however such an intervention is likely to be rejected in an era where thigh liposuction and lifts are the status quo. We therefore set out to objectively define characteristics of the ideal thigh. Several authors have defined gluteal subunits and universal attractive features, such as a V-shaped gluteal crease and the presence of lateral depressions.18 Mendieta divides key buttock components into volume, shape, and skin laxity.20,21 He further defines ten aesthetic units of the posterior region, each with its own aesthetic needs and pitfalls.22 To date, no such analysis has been conducted on the thigh. Wong et al in their recent publication updating standards for the ideal buttock found that that the traditional 0.7 waist-to-hip ratio, anthropologically correlated with the greatest fertility, had been replaced by new ideal ratios of 0.6 and 0.65.19,23 Building on this framework, we designed a survey to quantify key ratios for the ideal thigh, while keeping buttock features the same. Importantly, we distributed this survey through a novel crowdsourcing website operated by Amazon, called Mechanical Turk. Amazon Mechanical Turk is an internet service that provides an on-demand human workface from around the globe, allowing diverse and rapid feedback. This tool was previously validated in the plastic surgery literature by a study using conjoint analysis with crowdsourcing to characterize the most important attributes of aesthetic surgeons.24 It has been applied in behavioral science and psychiatric research,25,26 but our study is the first of its kind to apply Mechanical Turk to cosmetic preferences. The aim of this study was thus twofold. First, to highlight the importance of thigh contour in overall buttock aesthetic, and secondly, to illustrate the untapped potential of Mechanical Turk to classify broad preferences and guide aesthetic surgery decision making. To meet patients’ expectations, a plastic surgeon must understand their requests within the context of their personal identity, and given the changing demographic makeup of the United States, population surveys are an invaluable tool to reveal cultural ideals. METHODS The authors developed a study survey by digitally altering images of thighs using Adobe Photoshop CC (Adobe Systems, Inc., San Jose, CA) to create thighs of varying proportions. Images were obtained from the operative photos of a 27-year-old female patient who signed an informed consent allowing photographs to be used for research purposes. Surveys were distributed electronically through a new polling platform, Amazon Mechanical Turk, in which anonymous users were offered $0.05 for their responses during February to May 2017. Two thigh panels were presented and respondents were asked to rank each set of images from most to least attractive. A blank copy of the survey is available as Appendix A (available online as Supplementary Material at www.aestheticsurgeryjournal.com). Within each panel, images were arranged in a random order to avoid bias. Randomization was performed by the Research Electronic Data Capture (REDCap, Vanderbilt University, Nashville, TN) software. Users also answered demographic questions regarding age, gender, ethnicity, and nationality, with age range divisions based on standard demographic delineated categories. Institutional Review Board approval was not required as none of the respondents were patients; we conducted the survey in accordance with the ethical standards set out by the Declaration of Helsinki. To capture the spatial relationship between thigh and buttock contour in our figures, we specifically analyzed the buttock-thigh junction. All proportions were measured with the Adobe Photoshop CC ruler tool for confirmation of accuracy. We defined a lateral buttock-thigh ratio as the width of the thigh at the infragluteal crease over the width the point of maximal projection of the buttocks, which was set at a vertical level halfway between the infragluteal crease and the iliac crest (Figure 1). This ratio varied from 0.5 to 0.9 in five adjacent images presented on one panel (Figure 2). We then looked at the obtuse angle between an anatomical vertical meridian from the ASIS to the intertrochanteric crest, transposed laterally in the diagram to intersect the thigh-buttock convexity, and an oblique line connecting the widest point of horizontal buttock projection to the superolateral thigh-buttock junction (Figure 1). This “junction angle” allowed to us to adjust the severity of transition from buttock to thigh while preserving the overall shape of the thigh. The angle was set between 170 and 110 degrees in increments of 15 degrees between each image. Given that the ideal waist-hip ratio (defined as the width of the narrowest portion between the ribs and iliac crest compared to the width of the widest point of the buttocks) had previously been found to be 0.65, we used this value and kept the ratio uniform across all images. Figure 1. View largeDownload slide (A, B) Photographs of a 27-year-old female patient. Lateral thigh-to-buttock ratio represented by a/b, where a is the horizontal distance across the buttock-thigh junction (at the level of the gluteal crease) and b is the horizontal distance to the point of maximal buttock projection. Posterior view with thigh-buttock junction angle represented by angle θ, the angle between an anatomical vertical meridian from the anterior superior illiac spine (ASIS) to trochanteric crest, transposed laterally in the diagram to intersect the thigh-buttock convexity, and an oblique line from the widest point of buttock projection to the thigh-buttock junction. Figure 1. View largeDownload slide (A, B) Photographs of a 27-year-old female patient. Lateral thigh-to-buttock ratio represented by a/b, where a is the horizontal distance across the buttock-thigh junction (at the level of the gluteal crease) and b is the horizontal distance to the point of maximal buttock projection. Posterior view with thigh-buttock junction angle represented by angle θ, the angle between an anatomical vertical meridian from the anterior superior illiac spine (ASIS) to trochanteric crest, transposed laterally in the diagram to intersect the thigh-buttock convexity, and an oblique line from the widest point of buttock projection to the thigh-buttock junction. Figure 2. View largeDownload slide Image panels used in the survey, prior to randomization and with numeric labels added for ease of description. (A) Depiction of decreasing thigh-buttock junction angle in 15 degree increments, from an angle of 170 degrees in image 1 to 110 degrees in image 5. (B) Depiction of increasing thigh-to-buttock ratio, from a ratio of 0.5 to 0.9. Figure 2. View largeDownload slide Image panels used in the survey, prior to randomization and with numeric labels added for ease of description. (A) Depiction of decreasing thigh-buttock junction angle in 15 degree increments, from an angle of 170 degrees in image 1 to 110 degrees in image 5. (B) Depiction of increasing thigh-to-buttock ratio, from a ratio of 0.5 to 0.9. Data were stratified according to the age range, gender, ethnicity, and nationality of the respondents. Statistical analysis of the data was performed with Microsoft Excel 2016 (Microsoft Corp., Redmond, Wash.), with values of P < 0.05 designating statistical significance. RESULTS A total of 1052 responses were collected. There were 18 respondents who did not complete the survey in its entirety and whose responses were thus excluded from data analysis. Of the total 1034 respondents included, 562 (54.4%) were male and 472 (45.6%) were female (Table 1). Ages ranged from 18 to 72 years old, with a mean age of 39. The most represented groups were the 25 to 34 year-old (49.7%) and 35 to 44 year-old (25.4%) demographics. Most of our population was located in either the United States (44.6%) or Asia (41.7%), however we received responses from every geographic region. Table 1. Respondent Demographic Results Demographic characteristic No. of respondents (%) Total no. of respondents 1034 Gender Male 562 (54.4) Female 472 (45.6) Age group (years) 18-24 92 (8.9) 25-34 514 (49.7) 35-44 263 (25.4) 45-54 90 (8.7) 55-64 54 (5.2) >65 21 (2) Location of Origin Africa 19 (1.8) Asia 431 (41.7) Australia 2 (0.2) Canada 12 (1.2) Caribbean/Pacific Islands 7 (0.7) Europe 62 (6.0) Mexico 3 (0.3) Middle East 7 (0.7) South America 30 (2.9) United States 461 (44.6) Ethnicity Black/African 50 (4.8) Chinese 21 (2.0) Hispanic/Latino 42 (4.1) Indian subcontinent 365 (35.3) Japanese 3 (0.3) Korean 5 (0.5) Middle Eastern 23 (2.2) Native American 20 (1.9) Pacific Islander 29 (2.8) White/Caucasian 439 (42.5) Other/multiracial 37 (3.6) Demographic characteristic No. of respondents (%) Total no. of respondents 1034 Gender Male 562 (54.4) Female 472 (45.6) Age group (years) 18-24 92 (8.9) 25-34 514 (49.7) 35-44 263 (25.4) 45-54 90 (8.7) 55-64 54 (5.2) >65 21 (2) Location of Origin Africa 19 (1.8) Asia 431 (41.7) Australia 2 (0.2) Canada 12 (1.2) Caribbean/Pacific Islands 7 (0.7) Europe 62 (6.0) Mexico 3 (0.3) Middle East 7 (0.7) South America 30 (2.9) United States 461 (44.6) Ethnicity Black/African 50 (4.8) Chinese 21 (2.0) Hispanic/Latino 42 (4.1) Indian subcontinent 365 (35.3) Japanese 3 (0.3) Korean 5 (0.5) Middle Eastern 23 (2.2) Native American 20 (1.9) Pacific Islander 29 (2.8) White/Caucasian 439 (42.5) Other/multiracial 37 (3.6) View Large Table 1. Respondent Demographic Results Demographic characteristic No. of respondents (%) Total no. of respondents 1034 Gender Male 562 (54.4) Female 472 (45.6) Age group (years) 18-24 92 (8.9) 25-34 514 (49.7) 35-44 263 (25.4) 45-54 90 (8.7) 55-64 54 (5.2) >65 21 (2) Location of Origin Africa 19 (1.8) Asia 431 (41.7) Australia 2 (0.2) Canada 12 (1.2) Caribbean/Pacific Islands 7 (0.7) Europe 62 (6.0) Mexico 3 (0.3) Middle East 7 (0.7) South America 30 (2.9) United States 461 (44.6) Ethnicity Black/African 50 (4.8) Chinese 21 (2.0) Hispanic/Latino 42 (4.1) Indian subcontinent 365 (35.3) Japanese 3 (0.3) Korean 5 (0.5) Middle Eastern 23 (2.2) Native American 20 (1.9) Pacific Islander 29 (2.8) White/Caucasian 439 (42.5) Other/multiracial 37 (3.6) Demographic characteristic No. of respondents (%) Total no. of respondents 1034 Gender Male 562 (54.4) Female 472 (45.6) Age group (years) 18-24 92 (8.9) 25-34 514 (49.7) 35-44 263 (25.4) 45-54 90 (8.7) 55-64 54 (5.2) >65 21 (2) Location of Origin Africa 19 (1.8) Asia 431 (41.7) Australia 2 (0.2) Canada 12 (1.2) Caribbean/Pacific Islands 7 (0.7) Europe 62 (6.0) Mexico 3 (0.3) Middle East 7 (0.7) South America 30 (2.9) United States 461 (44.6) Ethnicity Black/African 50 (4.8) Chinese 21 (2.0) Hispanic/Latino 42 (4.1) Indian subcontinent 365 (35.3) Japanese 3 (0.3) Korean 5 (0.5) Middle Eastern 23 (2.2) Native American 20 (1.9) Pacific Islander 29 (2.8) White/Caucasian 439 (42.5) Other/multiracial 37 (3.6) View Large Thigh Preferences Overall From a population perspective, the most attractive posterior thigh option was image 1 (43.8% of respondents), with a thigh-buttock junction angle of 170 degrees and thus the widest thigh base. Chi-square fit analysis determined there was a statistically significant skew in responses (P < 0.05). We applied a binomial test to compare preference for image 1 to the expected distribution, and determined this outcome was significant (P < 0.05). Image 2 was the second most attractive (31.3%), image 3 was the third most (36.9%), and image 4 was fourth most (36.9%). Image 5, with the sharpest thigh-buttock junction angle at 110 degrees and the slimmest lateral thigh, was found to be the least attractive (43.9%). Each of these ratings was statistically significant (P < 0.05), supporting a direct decrease in attractiveness as posterior thigh width grew narrower. From the lateral view, there was a broad distribution of preferences among groups. The preferred thigh-buttock ratio was 0.8, with 27.6% of respondents rating this as the most attractive image. The second-most attractive ratio was 0.6 (26.9%), then 0.5 (27.6%) and 0.6 (27.4%) as third- and fourth-most attractive, respectively. The least attractive thigh was image 5 (43.6%) with a thigh-buttock ratio at 0.9, the widest displayed. These preferences were nonsignificant, except for the finding that image 5 was least attractive (P < 0.05). These results are tabulated in Table 2. Given the subtle shift in thigh shape between images on the lateral panel, we performed a subanalysis dividing responses into three groups, to determine if broader themes in preferences were present. Choice of narrow (0.5-0.6), medium (0.7), and wide (0.8-0.9) thigh ratios as most attractive were compared, and respondents demonstrated no significant predilection towards any category (P > 0.05). Table 2. Thigh Aesthetic Preferences Most attractive (percentage of respondents) Second-most attractive (percentage of respondents) Third-most attractive (percentage of respondents) Fourth-most attractive (percentage of respondents) Least attractive (percentage of respondents) Posterior image 1 (43.8) 2 (31.3) 3 (31.4) 4 (36.9) 5 (43.9) Thigh-buttock junction angle 170° 155° 140° 125° 110° Lateral image 4 (27.6) 3 (26.9) 1 (27.6) 2 (27.4) 5 (43.6) Thigh-hip ratio 0.8 0.7 0.5 0.6 0.9 Most attractive (percentage of respondents) Second-most attractive (percentage of respondents) Third-most attractive (percentage of respondents) Fourth-most attractive (percentage of respondents) Least attractive (percentage of respondents) Posterior image 1 (43.8) 2 (31.3) 3 (31.4) 4 (36.9) 5 (43.9) Thigh-buttock junction angle 170° 155° 140° 125° 110° Lateral image 4 (27.6) 3 (26.9) 1 (27.6) 2 (27.4) 5 (43.6) Thigh-hip ratio 0.8 0.7 0.5 0.6 0.9 View Large Table 2. Thigh Aesthetic Preferences Most attractive (percentage of respondents) Second-most attractive (percentage of respondents) Third-most attractive (percentage of respondents) Fourth-most attractive (percentage of respondents) Least attractive (percentage of respondents) Posterior image 1 (43.8) 2 (31.3) 3 (31.4) 4 (36.9) 5 (43.9) Thigh-buttock junction angle 170° 155° 140° 125° 110° Lateral image 4 (27.6) 3 (26.9) 1 (27.6) 2 (27.4) 5 (43.6) Thigh-hip ratio 0.8 0.7 0.5 0.6 0.9 Most attractive (percentage of respondents) Second-most attractive (percentage of respondents) Third-most attractive (percentage of respondents) Fourth-most attractive (percentage of respondents) Least attractive (percentage of respondents) Posterior image 1 (43.8) 2 (31.3) 3 (31.4) 4 (36.9) 5 (43.9) Thigh-buttock junction angle 170° 155° 140° 125° 110° Lateral image 4 (27.6) 3 (26.9) 1 (27.6) 2 (27.4) 5 (43.6) Thigh-hip ratio 0.8 0.7 0.5 0.6 0.9 View Large Thigh Preferences Categorized by Age Ranges of Respondents When ratings were categorized by age demographics, all groups except those 65 years or older chose image 1 as the most attractive posterior view (P < 0.05). The numbers of respondents who were aged 65 years or older were few (21 in total) in comparison with those of other groups studied, making it difficult to draw any conclusions regarding buttock preferences in this age range. Preferences on lateral view were mixed, however respondents aged 25 to 34, 35 to 44, and 65 or older rated a thigh-hip ratio of 0.8 as the most attractive (Table 3). These findings were not significant. Table 3. Most Attractive Thigh by Age Ranges Age range (years) Posterior view image (% of respondents) Lateral view image (% of respondents) 18-24 1 (37.0) 1 (27.2) 25-34 1 (40.0) 4 (26.3) 35-44 1 (52.1) 4 (30.8) 45-54 1 (51.1) 3 (26.7) 55-64 1 (28.1) 3 (36.6) >65 2 (28.6) 4 (23.8) Age range (years) Posterior view image (% of respondents) Lateral view image (% of respondents) 18-24 1 (37.0) 1 (27.2) 25-34 1 (40.0) 4 (26.3) 35-44 1 (52.1) 4 (30.8) 45-54 1 (51.1) 3 (26.7) 55-64 1 (28.1) 3 (36.6) >65 2 (28.6) 4 (23.8) View Large Table 3. Most Attractive Thigh by Age Ranges Age range (years) Posterior view image (% of respondents) Lateral view image (% of respondents) 18-24 1 (37.0) 1 (27.2) 25-34 1 (40.0) 4 (26.3) 35-44 1 (52.1) 4 (30.8) 45-54 1 (51.1) 3 (26.7) 55-64 1 (28.1) 3 (36.6) >65 2 (28.6) 4 (23.8) Age range (years) Posterior view image (% of respondents) Lateral view image (% of respondents) 18-24 1 (37.0) 1 (27.2) 25-34 1 (40.0) 4 (26.3) 35-44 1 (52.1) 4 (30.8) 45-54 1 (51.1) 3 (26.7) 55-64 1 (28.1) 3 (36.6) >65 2 (28.6) 4 (23.8) View Large We next analyzed the data by stratifying posterior view ratings by both age and gender, to see if males and females within each group demonstrated varying preferences. For ages 25 to 34, 35 to 44, and 45 to 54, both males and females rated image 1 as the most attractive (Figure 3). In the 18 to 24 demographic, females found image 1 most attractive (49.7%), whereas males found image 2 most attractive (32.1%). Similarly, for those aged 55 to 64, the preferred thigh for females was image 1 (60.4%), and for males was image 2 (29.3%). Figure 3. View largeDownload slide Choice for most attractive thigh from posterior view categorized by gender and age of respondents. Percentages of female respondents are shown in blue and those of male respondents are shown in orange. (A) Preferences of respondents aged 18 to 24 years. (B) Preferences of respondents aged 25 to 34 years. (C) Preferences of respondents aged 35 to 44 years. (D) Preferences of respondents aged 45 to 54 years. (E) Preferences of respondents aged 54 to 65 years. (F) Preferences of respondents over aged 65 years. Figure 3. View largeDownload slide Choice for most attractive thigh from posterior view categorized by gender and age of respondents. Percentages of female respondents are shown in blue and those of male respondents are shown in orange. (A) Preferences of respondents aged 18 to 24 years. (B) Preferences of respondents aged 25 to 34 years. (C) Preferences of respondents aged 35 to 44 years. (D) Preferences of respondents aged 45 to 54 years. (E) Preferences of respondents aged 54 to 65 years. (F) Preferences of respondents over aged 65 years. Thigh Preference by Ethnicity When evaluating aesthetic appeal from the posterior thigh view, all ethnicities were unanimous in preferring image 1, except those who identified as Middle Eastern (Table 4). Of these 23 respondents, 34.8% found image 3 to be the most attractive (P > 0.05). Responses for lateral view preferences were again broadly distributed. Those of Chinese, Indian, Middle Eastern, and Caucasian descent selected a thigh-buttock ratio of 0.8 as most attractive, whereas Black, Hispanic, and multi-racial ethnicities preferred a ratio of 0.6. Native Americans and Pacific Islanders leaned toward a ratio of 0.6, and only Koreans chose 0.9 as the most attractive proportion. Small sample sizes made it challenging to draw statistically significant conclusions. Table 4. Preferences for Most Attractive Thigh by Ethnicity % of respondents Black/ African Chinese Hispanic/ Latino Indian sub-continent Japanese Korean Middle East Native American Other/ Multi-racial Pacific Islander White/ Caucasian Posterior image 1 (50.0) 1 (38.1) 1 (52.3) 1 (35.6) 1 (66.7) 1 (100) 3 (34.8) 1 (35.0) 1 (48.6) 1 (27.8) 1 (50.3) Lateral image 3 (32.0) 4 (33.3) 3 (33.3) 4 (25.8) 4 (100) 5 (60.0) 4 (34.8) 2 (35.0) 3 (27.0) 2 (41.4) 4 (29.2) % of respondents Black/ African Chinese Hispanic/ Latino Indian sub-continent Japanese Korean Middle East Native American Other/ Multi-racial Pacific Islander White/ Caucasian Posterior image 1 (50.0) 1 (38.1) 1 (52.3) 1 (35.6) 1 (66.7) 1 (100) 3 (34.8) 1 (35.0) 1 (48.6) 1 (27.8) 1 (50.3) Lateral image 3 (32.0) 4 (33.3) 3 (33.3) 4 (25.8) 4 (100) 5 (60.0) 4 (34.8) 2 (35.0) 3 (27.0) 2 (41.4) 4 (29.2) View Large Table 4. Preferences for Most Attractive Thigh by Ethnicity % of respondents Black/ African Chinese Hispanic/ Latino Indian sub-continent Japanese Korean Middle East Native American Other/ Multi-racial Pacific Islander White/ Caucasian Posterior image 1 (50.0) 1 (38.1) 1 (52.3) 1 (35.6) 1 (66.7) 1 (100) 3 (34.8) 1 (35.0) 1 (48.6) 1 (27.8) 1 (50.3) Lateral image 3 (32.0) 4 (33.3) 3 (33.3) 4 (25.8) 4 (100) 5 (60.0) 4 (34.8) 2 (35.0) 3 (27.0) 2 (41.4) 4 (29.2) % of respondents Black/ African Chinese Hispanic/ Latino Indian sub-continent Japanese Korean Middle East Native American Other/ Multi-racial Pacific Islander White/ Caucasian Posterior image 1 (50.0) 1 (38.1) 1 (52.3) 1 (35.6) 1 (66.7) 1 (100) 3 (34.8) 1 (35.0) 1 (48.6) 1 (27.8) 1 (50.3) Lateral image 3 (32.0) 4 (33.3) 3 (33.3) 4 (25.8) 4 (100) 5 (60.0) 4 (34.8) 2 (35.0) 3 (27.0) 2 (41.4) 4 (29.2) View Large Thigh Preference by Location Given the wide distribution of locations of origin, we looked at responses from the four most represented regions—United States, Asia, Europe, and South America (Table 5). Respondents from all areas listed posterior image 1 as the most attractive, with over half of surveyed Europeans and South Americans preferring this widest thigh (P < 0.05). Lateral view responses were evenly distributed. Respondents from the United States, Asia, and South America selected image 4 as the most attractive by slim margins, whereas those from Europe preferred image 3. These findings were nonsignificant (P > 0.05). A small minority of respondents resided in areas not mentioned above, thus preferences were not evaluated according to every possible location. Table 5. Most Attractive Thigh by Location of Origin Location of origin Posterior view image (% of respondents) Lateral view image (% of respondents) United States 1 (47.3) 4 (27.8) Asia 1 (35.7) 4 (25.5) Europe 1 (67.7) 3 (33.9) South America 1 (60.0) 4 (33.3) Location of origin Posterior view image (% of respondents) Lateral view image (% of respondents) United States 1 (47.3) 4 (27.8) Asia 1 (35.7) 4 (25.5) Europe 1 (67.7) 3 (33.9) South America 1 (60.0) 4 (33.3) View Large Table 5. Most Attractive Thigh by Location of Origin Location of origin Posterior view image (% of respondents) Lateral view image (% of respondents) United States 1 (47.3) 4 (27.8) Asia 1 (35.7) 4 (25.5) Europe 1 (67.7) 3 (33.9) South America 1 (60.0) 4 (33.3) Location of origin Posterior view image (% of respondents) Lateral view image (% of respondents) United States 1 (47.3) 4 (27.8) Asia 1 (35.7) 4 (25.5) Europe 1 (67.7) 3 (33.9) South America 1 (60.0) 4 (33.3) View Large Characteristics of Respondents Categorized by Lateral Thigh Preference Lastly, to confirm the seemingly even distribution of lateral thigh preferences, we analyzed demographic characteristics of the respondents who had picked each of the five lateral thighs as the most attractive (Supplemental Table 1, available online as Supplementary Material at www.aestheticsurgeryjournal.com). Among those who preferred image 2, there were a larger number of females than males (P < 0.05). Conversely, significantly more males were represented in the group that preferred image 5, depicting the widest lateral thigh. Groups who selected the other images did not have significant differences in gender. Age distribution was similar between groups, with 25 to 34 year olds comprising the majority of respondents for each image, followed by 35 to 44 year olds. Respondents who identified as Caucasian or Indian made up the majority for all groups. No differences in ethnic make up between those who selected each image reached significance. DISCUSSION This study aimed to define the general population’s preference for ideal thigh shape. As the popularity and prevalence of gluteal augmentation has grown in recent years, interest in buttock aesthetic has increased accordingly. Mendieta and others have described detailed classification systems for buttock features and landmarks to guide surgical approach to gluteoplasty.20 However, the thigh, which in its immediate proximity to the buttocks plays a key role in posterior region appearance, has been neglected by researchers and surgeons alike. By surveying diverse individuals through an anonymous internet survey, powered by Amazon Mechanical Turk, we sought to establish population-based guidelines for the ideal thigh. Our study found that a wider lateral thigh width, as measured by the buttock-thigh junction angle, was preferred in comparison with a slimmer thigh. Almost half of respondents chose the widest displayed thigh as the most attractive, and this preference was similar across ages, ethnicities, and countries. Notably, as the lateral thigh narrowed, respondents tended to rate it as increasingly less attractive, demonstrating the consistency of this preference. Image five, with the narrowest thigh, was overwhelmingly rated as the least attractive image, by 43.9% of respondents. This supports the conclusion that it was relative thigh width, rather than any other feature, that guided ratings. Males and females voted similarly, refuting the widespread notion that men prefer slimmer frames. In fact, significantly more men preferred the thicker thigh on lateral view compared to female respondents. When approaching the thigh from a lateral view, there was no clear consensus on ideal width. Thigh-buttock ratios from 0.5 to 0.8 were found to be similarly attractive, and there were no notable differences in the demographic make up of populations selecting each image. However, a ratio of 0.9 was found to be least attractive by a statistically significant margin. Our results suggest that additional breadth is more appreciated in the coronal plane of the thigh than in the sagittal plane. Extending thigh width to the extreme on side view produced a displeasing result, whereas a spectrum of thinner thighs was equally acceptable. These findings are consistent with the modern predilection for a curvy, feminine figure. Augmenting the lateral aspects of the thighs, adjacent to the gluteal crease, creates a smoother transition from buttock to leg and avoids unnatural mismatch. Conversely, on lateral view, respondents chose to preserve buttock projection by limiting relative thigh extension. This makes intuitive sense as the aim of gluteoplasty is often to enhance and distinctly define the buttock. This study points to the need for a major paradigm shift in the field of gluteoplasty. While many women would balk at the idea of thigh augmentation, and surgeons would be unlikely to suggest it, our data indicate most individuals feel a wide upper thigh is more attractive. The goal of cosmetic surgeons is to increase buttock depth and projection, simulating an hourglass shape. Yet if the soft tissue is augmented in these dimensions, whether with implants or with fat grafting, and the thigh is left unchanged, the result is disproportionate and displeasing. Gluteal augmentation procedures should therefore address thigh contour with as much care as they do buttock shape. Previous studies have concluded that the most attractive waist-to-hip ratio in the modern era is closer to 0.65 than the traditional 0.7, reflecting preference for a more dramatic appearance of a small waist and voluptuous buttocks.27 Our study builds on this finding to demonstrate that a curvier buttock is best complemented by a broader based thigh. Our data also demonstrates the utility and potential of Amazon Mechanical Turk for determining population preferences. This platform allowed our survey to be distributed to a heterogenous group of individuals across the globe. Plastic surgeons continually debate questions of ideal aesthetic; however, Turk offers a window into real-time beauty standards. Knowing what a specific ethnic population finds attractive, or what females of a certain age are looking for, will allow the surgeon to adjust his techniques to provide patients with well-received outcomes. Mechanical Turk has already been successfully applied to behavioral science, psychology research, and consumer data mining. This study represents the one of the first incorporations of this tool into aesthetic surgery research, a field with highly relevant applications. This study had several limitations. It was a largely two-ethnicity study, and findings were limited by the skewed number of Caucasians and Southeast Indians in comparison with other ethnic groups. Consequently, we were unable to meaningfully comment on preferences of other populations prevalent among aesthetic patients, Asians, Latinos, or African Americans. We anticipate results more tailored to each culture in future studies, with a more even number of respondents from each location and ethnic group. There were morphometric limitations, as our digitally altered images oversimplify the thigh-buttock relationship and do not account for changes in leg dimensions. We designed the study in this manner to isolate single variables, despite the small risk of nonanatomical depictions. Furthermore, since the widest displayed thigh on posterior view was found to be the most attractive, and there was no comparison to an even wider thigh, we are unable to determine whether our maximal width is truly the ideal and whether further increases would be less favorable. Another consequence of isolating thigh measurements is that a wider thigh-buttock ratio could alternately be interpreted as favoring a narrower buttock rather than an augmented thigh. Certainly the relationship between anatomical parts is a key take-away, but the aim of our study was to highlight the specific role of thigh dimension. By keeping the buttocks neutral and constant across our images, the salience and therefore impact of a wide or narrow buttock was minimized. In practice, an integrated approach to buttock and thigh shaped is necessary for individualized patient planning. Finally, due to the survey-based nature of our study, our results are dependent on the honesty of our respondents in answering demographic questions. The 18 individuals who did not complete the survey in its entirety were excluded, to mitigate the effect of inaccurate information. Through Mechanical Turk, each user was required to submit a unique worker identification code to receive payment upon survey completion, and codes were not allowed to be repeated. This ensured that each response was a novel data point, rather than a reiteration of a previous respondent. CONCLUSIONS The goal of gluteoplasty procedures is to make the posterior region more pleasing to the patient. The thigh plays a major role in final appearance, yet there are no guidelines for thigh aesthetic. As various authors have repeatedly discussed, in order to evaluate the success of surgery there must be an ideal, or “yardstick,” against which to compare results.28 Quantification of ideal relationships not only provides a more accurate understanding of anatomy, but serves to guide surgical techniques towards optimal aesthetic outcomes.29 In seeking to elucidate modern preferences, this study highlights the importance of thigh contour in creating an attractive result. A broad upper thigh, with seamless transition from buttock to thigh, was found to be most attractive shape. This finding has important implications for simultaneous thigh augmentation during gluteoplasty. Crowdsourcing, with the aid of modern photography and software, is thus a powerful tool to define aesthetic goals by turning patients into engaged consultants. Supplementary Material This article contains supplementary material located online at www.aestheticsurgeryjournal.com. Disclosures The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article. Funding The authors received no financial support for the research, authorship, and publication of this article. REFERENCES 1. Cosmetic Surgery National Data Bank Statistics . Aesthet Surg J . 2017 ; 37 ( suppl_2 ): 1 - 29 . 2. American Society for Aesthetic Plast Surg . 2011 Cosmetic Surgery National Data Bank Statistics . Available at: https://www.surgery.org/sites/default/files/ASAPS-Stats2011.pdf. Accessed September 10, 2017 . 3. Mofid MM , Gonzalez R , de la Peña JA , Mendieta CG , Senderoff DM , Jorjani S . Buttock augmentation with silicone implants: a multicenter survey review of 2226 patients . Plast Reconstr Surg . 2013 ; 131 ( 4 ): 897 - 901 . 4. Peer LA . The neglected free fat graft . Plast Reconstr Surg . 1956 ; 18 ( 4 ): 233 - 250 . 5. Khouri RK , Eisenmann-Klein M , Cardoso E , et al. Brava and autologous fat transfer is a safe and effective breast augmentation alternative: results of a 6-year, 81-patient, prospective multicenter study . Plast Reconstr Surg . 2012 ; 129 ( 5 ): 1173 - 1187 . 6. Coleman SR , Saboeiro AP . Fat grafting to the breast revisited: safety and efficacy . Plast Reconstr Surg . 2007 ; 119 ( 3 ): 775 - 785 . 7. Clauser L , Polito J , Mandrioli S , Tieghi R , Denes SA , Galiè M . Structural fat grafting in complex reconstructive surgery . J Craniofac Surg . 2008 ; 19 ( 1 ): 187 - 191 . 8. Coleman SR . Structural fat grafting: more than a permanent filler . Plast Reconstr Surg . 2006 ; 118 ( 3 Suppl ): 108S - 120S . 9. Del Vecchio D , Rohrich RJ . A classification of clinical fat grafting: different problems, different solutions . Plast Reconstr Surg . 2012 ; 130 ( 3 ): 511 - 522 . 10. Cárdenas-Camarena L , Arenas-Quintana R , Robles-Cervantes JA . Buttocks fat grafting: 14 years of evolution and experience . Plast Reconstr Surg . 2011 ; 128 ( 2 ): 545 - 555 . 11. Toledo LS . Gluteal augmentation with fat grafting: the Brazilian buttock technique: 30 years’ experience . Clin Plast Surg . 2015 ; 42 ( 2 ): 253 - 261 . 12. Roberts TL III , Weinfeld AB , Bruner TW , Nguyen K . “Universal” and ethnic ideals of beautiful buttocks are best obtained by autologous micro fat grafting and liposuction . Clin Plast Surg . 2006 ; 33 ( 3 ): 371 - 394 . 13. Perén PA , Gómez JB , Guerrerosantos J , Salazar CA . Gluteus augmentation with fat grafting . Aesthetic Plast Surg . 2000 ; 24 ( 6 ): 412 - 417 . 14. Pereira LH , Radwanski HN . Fat grafting of the buttocks and lower limbs . Aesthetic Plast Surg . 2004 ; 20 ( 5 ): 409 - 416 . 15. Pedroza D , Valero L . Fat Transplantation to the Buttocks and Legs for Aesthetic Enhancement or Correction of Deformities: Long‐Term Results of Large Volumes of Fat Transplant . Dermatol Surg . 2000 ; 26 ( 12 ): 1145 - 1149 . 16. Murillo WL . Buttock augmentation: case studies of fat injection monitored by magnetic resonance imaging . Plast Reconstr Surg . 2004 ; 114 ( 6 ): 1606 - 1614 . 17. Gutowski KA , Force AFGT . Current applications and safety of autologous fat grafts: a report of the ASPS fat graft task force . Plast Reconstr Surg . 2009 ; 124 ( 1 ): 272 - 280 . 18. Cuenca-Guerra R , Lugo-Beltran I . Beautiful buttocks: characteristics and surgical techniques . Clin Plast Surg . 2006 ; 33 ( 3 ): 321 - 332 . 19. Wong WW , Motakef S , Lin Y , Gupta SC . Redefining the Ideal Buttocks: A Population Analysis . Plast Reconstr Surg . 2016 ; 137 ( 6 ): 1739 - 1747 . 20. Mendieta CG . Classification system for gluteal evaluation . Clin Plast Surg . 2006 ; 33 ( 3 ): 333 - 346 . 21. Mendieta CG . Gluteoplasty . Aesthet Surg J . 2003 ; 23 ( 6 ): 441 - 455 . 22. Mendieta CG . Gluteal reshaping . Aesthet Surg J . 2007 ; 27 ( 6 ): 641 - 655 . 23. Singh D . Adaptive significance of female physical attractiveness: role of waist-to-hip ratio . J Pers Soc Psychol . 1993 ; 65 ( 2 ): 293 - 307 . 24. Wu C , Scott HC , Diegidio P . What do our patients truly want? Conjoint analysis of an Aesthetic Plast Surg practice using internet crowdsourcing . Aesthet Surg J . 2017 ; 37 ( 1 ): 105 - 118 . 25. Bates JA , Lanza BA . Conducting psychology student research via the Mechanical Turk crowdsourcing service . N Am J Psychol . 2013 ; 15 ( 2 ): 385 . 26. Crump MJ , McDonnell JV , Gureckis TM . Evaluating Amazon’s Mechanical Turk as a tool for experimental behavioral research . PLoS One . 2013 ; 8 ( 3 ): e57410 . 27. Freese J , Meland S . Seven tenths incorrect: heterogeneity and change in the waist-to-hip ratios of Playboy centerfold models and Miss America pageant winners . J Sex Res . 2002 ; 39 ( 2 ): 133 - 138 . 28. Millard DR Jr . Principlization of Plastic Surgery . Boston : Little, Brown ; 1986 . 29. Aly A , Tolazzi A , Soliman S , Cram A . Quantitative analysis of aesthetic results: introducing a new paradigm . Aesthet Surg J . 2012 ; 32 ( 1 ): 120 - 124 . © 2018 The American Society for Aesthetic Plastic Surgery, Inc. Reprints and permission: [email protected] This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)