No evidence that spice consumption is a cancer prevention mechanism in human populations

Abstract

Background: Why humans historically began to incorporate spices into their diets is still a matter of unresolved debate. For example, a recent study (Bromham et al. There is little evidence that spicy food in hot countries is an adaptation to reducing infection risk. Nat Hum Behav 2021;5:878-91.) did not support the most popular hypothesis that spice consumption was a practice favoured by selection in certain environments to reduce food poisoning, parasitic infections, and foodborne diseases.

Methods: Because several spices are known to have anticancer effects, we explored the hypothesis that natural selection and/or cultural evolution may have favoured spice consumption as an adaptive prophylactic response to reduce the burden of cancer pathology. We used linear models to investigate the potential relationship between age-standardized gastrointestinal cancer rates and spice consumption in 36 countries.

Results: Patterns of spice are not consistent with a cancer mitigation mechanism: the age-standardized rate of almost all gastrointestinal cancers was not related to spice consumption.

Conclusions: Direction other than foodborne pathogens and cancers should be explored to understand the health reasons, if any, why our ancestors developed a taste for spices.

Keywords: behaviour; oncogenes; prophylaxy.

Figure 1.

World map with, in blue, the 36 countries included in this study with both spice consumption and cancer data available. Spice data were expressed as the average number of spice ingredients included in a recipe, while the cancer data were expressed in age-standardized incidence rates.

Figure 2.

Effect of the mean number of spices included in recipes on the age-standardized rate of gastrointestinal cancers using univariate models (i.e. not accounting for confounding variables). Only cancer types with at least one significant model are presented (all other cancer types are presented in Supplementary data 1). Statistically significant linear regression lines (in red) are showed with their 95% confidence intervals bands (in grey). After controlling for confounding variables in multivariate models, only the relationship between mean number of spices and age-standardized pancreatic cancer rate remained significant.

Figure 3.

(A) Effect of the mean number of spice ingredients included in meat recipes on the age-standardized incidence rate of pancreas in 36 countries (black dots). (B) Marginal effects of the mean number of spice ingredients included in meat recipes on the age-standardized incidence rates of pancreas cancers when accounting for confounding factors using the 2010, 2015 and 2018 alcohol, tobacco or human development index data. In (A) and (B), the solid lines represent the average predicted by the models and the error bands their associated 95% confidence intervals. Little effect of the year is observed.