Intended for healthcare professionals

Feature Christmas 2016: Food for Thought

Is caviar a risk factor for being a millionaire?

BMJ 2016; 355 doi: (Published 09 December 2016) Cite this as: BMJ 2016;355:i6536

Re: Is caviar a risk factor for being a millionaire? Definitions of 'risk factor' and related concepts

Unable as I am, unfortunately, to invite Dr Huitfeldt to caviar, I nevertheless wish to sincerely thank him for his thoughtful reference to ‘A dictionary of epidemiology’ –which is not exactly “Miquel Porta’s”, by the way, but sponsored by the International Epidemiological Association [] and coedited with Sander Greenland, Miguel Hernán, Isabel dos Santos Silva, and John Last. Although Dr Huitfeldt cites the previous edition of the dictionary (reference 17), the part of the definition of ‘risk factor’ that he reproduces does appropriately correspond to the latest, 6th. edition (Oxford University Press, 2014).
I think the interesting issues Dr Huitfeldt raises in the article might be further enlighted by the complete text that the dictionary offers on ‘risk factor’, which includes related terms [here, in capital letters] and numerous references [here omitted]:
“RISK FACTOR (Syn: determinant) A factor that is causally related with a change in the RISK of a relevant health process, outcome or condition. The causal nature of the relationship is established on the basis of scientific evidence (including evidence from EPIDEMIOLOGICAL RESEARCH) and CAUSAL INFERENCE. The causal relationship is inherently probabilistic, as it happens in many other spheres of nature and human life. If the relationship is noncausal the factor is just a RISK MARKER. Examples of types of risk factors are offered throughout this book; they may be a socioeconomic characteristic, personal behaviour or lifestyle, environmental exposure, inherited characteristic or another TRAIT. Risk factors for human health often have individual and social components; even when individual and social risk factors can be separated, they often interact. A DETERMINANT that can be modified by intervention is a modifiable risk factor. (...)
To prevent MEDICALIZATION of life and IATROGENESIS, the RELEVANCE and SIGNIFICANCE of the factor-outcome risk relationship must be cautiously assessed; so must uncertainties and ambiguities in risk-related concepts, as well as different legitimate meanings of risk across and within cultures.”
On a minor note of divergence with Dr Huitfeldt, may I note that he does not justify his belief that the definition above “implicitly assumes that epidemiologists are only interested in causality, to the exclusion of other worthy research objectives such as reducing diagnostic or prognostic uncertainty.” The definition above does not assume that epidemiologists are only interested in causality. Fundamental as causal inference is in SCIENCE, it is not the only aim for epidemiological research and epidemiologists. Many scientific articles and textbooks (including the dictionary) show that epidemiologists and EPIDEMIOLOGY are not only interested in causality.
As also mentioned elsewhere in the dictionary, in human health –and, specifically, in DISEASES OF COMPLEX ETIOLOGY– sets of risk factors often act jointly in relatively complex and long-term processes. They commonly operate both at aggregate and distal levels (e.g., social, regional, global), as well as at the individual, personal level; i.e., across macro- and micro-levels, SYSTEMICALLY.
Current knowledge on RESEARCH DESIGNS and causal inference (for instance, on ways to assess and prevent REVERSE CAUSATION) might also shed further light on the questions addressed in the article. I, of course, tend to agree that “semantic ambiguity may lead to methodological confusion; and vice versa.
Best wishes for the season, above all, and malgré tout.

Competing interests: No competing interests

09 December 2016
Miquel Porta
Professor of epidemiology
IMIM UAB PRBB, Carrer Dr Aiguader 88, E-08003 Barcelona, Catalonia, Spain