We used a multivariate dimension reduction technique, called principal component analysis,  to derive the body shapes from six anthropometric traits: height, weight, body mass index, waist circumference, hip circumference, and waist-to-hip ratio. This resulted in four distinct body shapes that may better capture the heterogeneous expression of adiposity and its health consequences compared with single anthropometric traits because of the way they combine.

The study included 340 152 men and women from nine European countries, mostly aged 35–65 years at recruitment (in 1990–2000) into the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Cox proportional hazards regression was used to estimate multivariable-adjusted hazard ratios (HRs) and 95% confidence intervals (CIs).

A body shape that characterizes overall adiposity was positively associated with overall cancer risk, with a HR per 1 standard deviation increment of 1.07 (95% CI, 1.05–1.08), and with risk of 10 cancer types, with HRs (per 1 standard deviation) ranging from 1.36 (95% CI, 1.30–1.42) for endometrial cancer to 1.08 (95% CI, 1.03–1.13) for rectal cancer. A body shape that characterizes tall stature with a low waist-to-hip ratio was positively associated with overall cancer risk (HR, 1.03; 95% CI, 1.02–1.04) and with risk of five cancer types. A body shape that characterizes tall stature with a high waist-to-hip ratio was positively associated with overall cancer risk (HR, 1.04; 95% CI, 1.03–1.05) and with risk of 12 cancer types. An “athletic” body shape was not associated with overall cancer risk (HR, 1.00; 95% CI, 0.99–1.01).

These findings suggest that the current cancer burden associated with adiposity and body size based on classic anthropometric traits is probably underestimated.

This exciting research journey started in 2016, when we came across a great article in Nature Communications by Ried et al. proposing that body shape phenotypes represents information that is not fully captured by single anthropometric traits (e.g., body mass index, BMI). This notion was not completely new, because despite the numerous advantages of BMI, for example, as a population-level measure of obesity prevalence or as a risk factor for at least 13 different cancers, it also received criticism. BMI neither differentiates between muscle and fat mass nor does it capture body fat distribution, which may affect its specificity in predicting cancer risk. We started from there, successfully replicating these body shape phenotypes in different large cohorts and investigated its associations with the incidence of 24 cancers in the EPIC study. 

The results summarized above and published in the British Journal of Cancer are promising and may help to refine cancer risk prediction, but the journey is far from over and a lot of work is still ahead of us. To make such endeavours a success, a great team of researchers is needed, which we certainly have, and funding, which we gratefully received from the French National Cancer Institute (INCa) and the Deutsche Forschungsgemeinschaft (DFG).