Editorials
Consumption of hot spicy foods and mortality—is chilli good for your health?
BMJ 2015; 351 doi: http://dx.doi.org/10.1136/bmj.h4141 (Published 04 August 2015)Cite this as: BMJ 2015;351:h4141
Diet and nutrition have long been regarded as central to health and longevity. Given the vast variety and complexity of human diets, however, the ongoing challenge has been to identify the specific dietary components with a direct effect on health and mortality. The general consensus is that health gains for chronic disease are most likely from healthy dietary patterns that include adequate consumption of fruits, vegetables, whole grains, nuts, seeds, fibre, and fish and that are low in red and processed meats, sugary beverages, and salt.1 2 3 Yet there remains a parallel interest in other common dietary components that may serve as functional foods. Hot spices are one such example and are the subject of a linked paper by Lv and colleagues (doi:10.1136/bmj.h3942).4
Among 0.5 million adults in the China Kadoorie Biobank the authors examined the prospective association of self reported consumption of spicy foods with total and cause specific mortality. Over a median of 7.2 years of observation with 3.5 million person years, during which 20 224 deaths occurred, they report a 14% lower risk (95% confidence interval 10% to 18%) in total mortality when comparing those who reported frequent consumption of spicy foods (6 or 7 days a week) with those who reported little consumption of spicy foods (less than once a week). A similar reduction in mortality was apparent even among those who reported consuming spicy foods 3-5 or 1 or 2 days a week compared with those whose consumption was infrequent. Inverse associations were also observed for cause specific deaths due to cancer, ischaemic heart disease, and respiratory disease. How should we interpret these novel findings and what are their implications for nutritional advice?
This research has several strengths, including a large sample size, the inclusion of 10 geographical regions of China representing both urban and rural settings, a prospective design, and sound application of statistical methods. The authors acknowledge limitations of measurement error, possible bias, confounding, and reverse causality, which are common problems in epidemiology, and indeed their efforts to minimise some of these limitations are notable.
The use of spices is an integral part of the Chinese diet5 however, the authors only adjusted for three crudely measured dietary covariates (self reported consumption frequency of red meat, fresh vegetables, and fresh fruits) and were unable to account for energy intake or for other dietary habits that may be correlated with spicy foods. As the authors recognise, this could cause residual confounding. Their definition of spicy foods is synonymous with the frequency of consumption of types of chilli, fresh, dried, or as chilli oil or chilli sauce. It is unclear whether the observed associations are the direct result of chilli intake or whether chilli is simply a marker for other beneficial but unmeasured dietary components.
The effect of the quantity or strength (degree of hotness) of chilli consumed is also unknown, along with the effect of other behaviours such as alcohol consumption. The significant inverse association between chilli consumption and mortality only among those who did not consume alcohol (and a null association among those who did consume alcohol) remains unexplained. Future studies should explore if confounding or effect modification by other drinking habits might play a part, as it is highly likely that drinks such as water or different types of tea are consumed in greater amounts among those with a greater chilli intake. Concurrently there is evidence for an inverse association between tea consumption and mortality.6 Lv and colleagues may be able to deal with some of these questions, using data collected on tea intake and other variables in their study.
So, should we encourage people to eat more chilli? As the authors acknowledge, a cause and effect relation cannot be inferred from their work. In this prospective study, Lv and colleagues have shown temporality of association, but we need to evaluate additional criteria to judge the strength of evidence.7 Their findings should be considered hypothesis generating, not definitive, and will undoubtedly encourage further work.
The use of hot spices in food to enhance taste has captured the attention of the popular press as well as food outlets, including supermarkets, restaurants, and fast food shops, fuelling a worldwide trend towards greater consumption.8 9 In parallel, there is increasing scientific interest in spicy foods. Many potential benefits4 have been suggested for chilli or its bioactive compound capsaicin, including but not limited to antimicrobial, anti-oxidant, anti-inflammatory, and anti-cancer properties, a beneficial influence on gut microbiota, and anti-obesity effects through thermogenesis and appetite,10 energy balance,11 and weight management.12
Despite a large published literature on capsaicin (a search of PubMed on 23 July 2015 listed 12 571 articles), a systematic appraisal of potential beneficial and adverse impacts of spicy foods and their bioactive compounds is lacking and is warranted. Finally, although dietary modification trials are challenging for logistical reasons, adding or not adding spice to foods may be achievable, at least for short term trials reporting intermediate endpoints.13
Future research is needed to establish whether spicy food consumption has the potential to improve health and reduce mortality directly or if it is merely a marker of other dietary and lifestyle factors. The added contribution of spicy food intake to the benefits of a balanced healthy diet and healthy lifestyles also remains to be investigated. However, the current findings should certainly stimulate dialogue, debate, and further interest in research.
Should people eat spicy food? It is too early to say, but the debate and the research interest are certainly hotting up.
Notes
Cite this as: BMJ 2015;351:h4141
Footnotes
- Competing interests: I have read and understood the BMJ policy on declaration of interests and declare the following: none.
- Provenance and peer review: Commissioned; not externally peer reviewed.
