Evidence That Really Matters

Shapely centrefolds? Temporal change in body measures: trend analysis

BMJ 2002; 325 doi: http://dx.doi.org/10.1136/bmj.325.7378.1447 (Published 21 December 2002) Cite this as: BMJ 2002;325:1447
  1. Martin Voracek, research resident (martin.voracek{at}akh-wien.ac.at)a,
  2. Maryanne L Fisher, PhD candidateb
  1. a Department of Psychoanalysis and Psychotherapy, Statistics and Documentation Branch, University of Vienna Medical School, AKH/Währinger Gürtel 18-20, A-1090 Wien, Austria,
  2. b Department of Psychology, York University, 4700 Keele Street, Toronto, ON M3J 1P3, Canada
  1. Correspondence to: M Voracek

    Body mass index (weight (kg)/(height (m)2) and waist:hip ratio in women are linked to fertility, endocrine status, risk of major diseases, and longevity.13 Health related optimums for body mass index (20 or slightly lower2) and waist:hip ratio (0.7 or slightly lower3) are also maximally sexually attractive to men. 1 3 According to evolutionary research, these attractiveness optimums reflect evolved optimal design and thus should not be subject to temporal change.3

    This assumption is not consonant with the decline in the optimally attractive body mass index that has occurred in the past few decades, as exemplified by fashion models depicted in the media. With increases in the incidence of eating disorders in the general population of women, this decline is a cause for concern. 4 5 In contrast, Singh has reported evidence for the temporal stability of the maximally attractive waist:hip ratio, on the basis of analysis of the waist:hip ratios of centrefold models inPlayboy.3 However, Singh based this conclusion, as is the case for other studies pertaining to body measurements of Playboycentrefolds,4 on a partial sample.

    Subjects, methods, and results

    We looked at the trends in Playboy centrefold models' body measurements by analysing 577 consecutive monthly issues, from the magazine's inception in December 1953 to December 2001. We extracted centrefolds' anthropometric data: height, weight, and measurements for bust, waist, and hip (n=532-543, owing to missing data). We calculated composite measures from these data: body mass index, waist:hip ratio, waist:bust ratio, bust:hip ratio, and an androgyny index (suggested by a reviewer)—waist/((hip*bust)**0.5). We correlated individual measures with magazine issue number (1 to 577).

    All measures except weight, which was nearly stable (r=−0.02) and hence may indicate a stable attractiveness cue, showed significant temporal change (if not specified, P<0.001). Whereas the increase in height (r=0.36) merely reflects the well known secular acceleration trend, and an increase in the age of models (r=0.22) was not relevant to this investigation, all other changes call for attention. Over time, bust size (r=−0.36) and hip size (r=−0.29) decreased, while waist size increased (r=0.27). Composite measures of body shape captured the same trends: body mass index (r=−0.46) and bust:hip ratio (r=−0.13; P=0.002) decreased, while waist:hip ratio (r=0.47), waist:bust ratio (r=0.48), and androgyny index (r=0.50) increased.

    Embedded Image

    Changing trends in body shape: Rubens's wife (1636-8); Marilyn Monroe (1952); Eva Herzigova (2001)


    Embedded Image

    Embedded Image


    The data suggest notable temporal trends in measures of body shape in Playboy centrefold models (figure). The typical body mass index of Playboy centrefolds has further descended below corresponding population levels, whereas their typical waist:hip ratio now approaches population levels. In sum, centrefold models' shapely body characteristics have given way to more androgynous ones. These temporal trends are at odds with claims that centrefolds' body shapes are still more “hourglasses” than “stick insects”4 and that the maximally sexually attractive female waist:hip ratio is stable.3


    Temporal change in body measures of Playboy centrefold models. Circles show individual data points (slight random jitter added to increase visibility). Linear regression (middle line) and 99% confidence intervals (upper and lower curves) are superimposed


    Contributors: MV and MLF conceived the study, analysed the data, and wrote the article. MV is the guarantor.


    • Funding None.

    • Competing interests None declared.