The media and fashion industries have recently showcased a wider range of body types and ethnicities, leading to a perception of growing inclusivity. But a new analysis published in the Proceedings of the National Academy of Sciences reveals that the typical female fashion model’s body has remained exceptionally thin and largely unchanged over the past twenty-five years. While representation has broadened on the margins, the central ideal of beauty remains entirely stable.
The media and fashion industries function as central arbiters of body ideals, consistently emphasizing thin female figures and lean male physiques. Despite growing calls for diversity, quantitative evidence tracking long-term shifts in representation has remained limited. Past attempts to measure these trends often relied on small samples or focused on specific geographic regions. This made it difficult to assess systemic changes across the entire industry.
To address this gap, researchers set out to track exactly how representation has evolved over a long period. The study was led by Louis Boucherie, a researcher at the Technical University of Denmark and the University of Copenhagen. Boucherie worked alongside scientists from Northeastern University, Harvard Medical School, the Complexity Science Hub in Vienna, and other institutions in Europe and the United States.
Boucherie and his colleagues compiled a dataset of 793,199 professional modeling records spanning from 2000 to 2024. They gathered this information from major industry portfolio platforms where modeling agencies submit verified data. The records captured a wide variety of work, including runway shows, magazine covers, editorials, and advertising campaigns.
The data contained detailed physical measurements and visible traits, such as hair color, eye color, and national origin. When self-reported demographic information was missing, the team used computer vision software to estimate gender and ethnicity from profile photographs. To assess body fat levels accurately, the scientists calculated a metric called Relative Fat Mass for the models.
Relative Fat Mass is a body composition index that estimates a person’s body fat percentage using their height and waist circumference. Health experts consider it a more accurate predictor of whole-body fat percentage than the traditional Body Mass Index. This allows researchers to accurately gauge physical thinness without relying solely on weight data.
The analysis revealed a striking contradiction in how beauty standards have evolved. On the surface, the variety of body sizes represented in fashion media has increased over the past two decades. However, the average measurements of female models have remained incredibly stable.
During this twenty-four-year period, the average female model stood at approximately 177 to 178 centimeters tall, or about 5 feet 10 inches. Her waist hovered constantly around 60 to 61 centimeters, while her hips remained stable at roughly 88 to 89 centimeters. Only the average bust circumference showed a consistent decline, dropping from roughly 85 centimeters to 82 centimeters.
“On the mean, nothing happens. Everything is super stable,” Boucherie said. “When we then look at the change in variation, we find what you’d expect: body size diversity has grown. But when we look at how that variation is distributed, we can see that the middle stays stable. So, the change is happening at the outliers.”
Visualizing this data highlights the evolution of representational diversity for female models between 2000 and 2024. Anthropometric means for height, bust, waist, hips, and Relative Fat Mass remain stable across the period for fashion shows, advertisements, magazine covers, and editorials. Only the bust shows a modest decline. At the same time, standard deviations of these same measures increase across all work types, indicating growing variability in represented body sizes despite stable central tendencies.
Other visual traits have diversified alongside these shifting extremes. Hair color distribution shifts away from lighter phenotypes, as blonde declines while darker shades gain share. Eye color distribution shows similar diversification, with blue eyes declining and brown increasing. National origins by world region have also shifted, as Eastern European representation peaks in the early 2000s then declines, while contributions from Sub-Saharan Africa, East Asia, and South Asia increase.
Despite this outward phenotypic diversification, the core standard of beauty has not become more inclusive in terms of body composition. The industry has selectively hired a small number of plus-size models to appear alongside the traditional, exceptionally thin models. To understand how these models compare to everyday people, the researchers matched their data against a massive health survey of American women aged seventeen to thirty. The two groups showed almost no overlap in body fat distribution.
“When we compare the US models to the general US population, there is almost no overlap between the two. And if you look carefully, you see that even the plus size models are still below the average US body size. So, what the fashion industry calls plus size corresponds much more closely to the average American woman,” Boucherie said.
The scientists also looked at how different marginalized identities overlap, a concept known as intersectionality. They found that racial representation in fashion imagery has broadened notably. The share of models identified as non-white grew from roughly thirteen percent in 2011 to more than forty percent in recent years. However, the burden of size diversity falls heavily on these same non-white models.
The data indicates that a plus-size model is 4.5 times more likely to be non-white. This highlights a symbolic diversification where the industry satisfies representational demands without abandoning its core aesthetic.
“What these patterns of representation end up meaning is that the burden of representing diversity often falls on a relatively small group of non white models,” Boucherie said. By hiring non-white, plus-size individuals, fashion institutions can project a diverse image without structurally broadening what counts as aspirational.
The authors also mapped out a data-driven prestige hierarchy based on how often top brands shared the same elite models. Elite fashion houses featured the thinnest models in the entire dataset. Interestingly, these top-tier brands also hired a higher share of visibly plus-size models compared to mid-level brands. This creates a polarized environment at the very top of the fashion world, where extreme thinness is the norm, peppered with occasional extreme outliers.
Finally, the researchers examined whether government or industry regulations have successfully altered beauty standards. They compared a 2006 rule at Milan Fashion Week, which established a strict numerical minimum for a model’s Body Mass Index, with a 2017 French law that required a doctor’s medical certificate without a specific numerical cutoff. The strict numerical threshold in Milan immediately reduced the number of extremely thin models on the runway. The more flexible French law produced no detectable shift in body representation.
“What we see is that in Milan, where there was a hard numerical threshold, there is a clear reduction in the number of extremely thin models after the regulation was introduced. In France, however, where the regulation was much softer and based on doctor certification, we don’t see the same kind of effect. We’re very careful not to claim causality here, but descriptively the difference between a hard threshold and a flexible system is quite striking,” Boucherie said.
The authors noted some limitations in their approach. To process hundreds of thousands of images, the computer algorithms relied on simplified categories. Gender was reduced to a binary classification, and ethnicity was sorted into broad white and non-white groups.
“We don’t have very fine grained racial categories in the data. We essentially must work with a white versus non white distinction, which is obviously a coarse way of doing it. But it’s the only way to do the analysis consistently across the full dataset and over time,” Boucherie said.
This broad categorization risks hiding the unique experiences of specific racialized groups and nonbinary individuals. The study also focused primarily on women’s fashion imagery. While male models face similarly narrow ideals for muscular and lean physiques, the data for men showed less change over time, making it difficult to draw the same detailed statistical conclusions.
The scientists caution that the comparison between the Milan and Paris regulations highlights correlations rather than guaranteed causal effects. Still, the differing outcomes suggest that policy design matters greatly when attempting to change industry behavior. Future research might explore these dynamics in user-generated social media content or monitor how artificial intelligence reproduces these same narrow standards when generating synthetic media.
“I think people already knew there was a problem, which has been debated repeatedly. What we’ve done is to quantify it. And I think that’s the new part. We’re just here to say that there is this problem, and then it’s the responsibility of the advertisers and the people organizing fashion shows and editing magazines to decide what to do with that information,” Boucherie said.
The study, “Cultural evolution of beauty standards,” was authored by Louis Boucherie, Sagar Kumar, Katharina Ledebur, August Lohse, and Karolina Sliwa.