Association between consumption of ultra-processed foods and all cause mortality: SUN prospective cohort study
BMJ 2019; 365 doi: https://doi.org/10.1136/bmj.l1949 (Published 29 May 2019) Cite this as: BMJ 2019;365:l1949Linked Research
Ultra-processed food intake and risk of cardiovascular disease
Linked Editorial
Ultra-processed food and adverse health outcomes
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Dear Editor
We read with great interest the paper by Rico-Campà and colleagues (1) on the possible effects of ultra-processed foods (UPF) on all-cause mortality in the SUN cohort. We would appreciate if the authors would provide some additional clarification regarding the data presented and their interpretation.
In table 3, the crude number of fatal events is lowest, before any statistical adjustment, in the group with UPF intakes higher than 4 portions per day (n=73) and highest among persons consuming less than 2 portions per day (n=108), with relative risk (RR) = 0.68 for the fourth vs. the first UPF intake group. It is not clear which parameter, in the Cox model used by the authors, may be responsible for the change of this RR, which increases to 1.78.
One would speculate that the subjects in the highest UPF intake group might be, on average, younger (or even much younger) than those with the lowest UPF intake; unfortunately the mean ages of the four groups are not provided; however, the RR is not materially changed after adjustment for age and gender (before: 1.78; after: 1.72).
A significant difference in age, if confirmed, might also help to understand why in the highest UPF intake group only 61% of observed deaths is caused by a cardiovascular event or a cancer, while in the lowest UPF intake group the corresponding figure is 78%. Unfortunately, the precise causes of such non-cardiovascular and non-cancer deaths are not provided, and it is not possible to evaluate the biological plausibility of their association with UPF intake. Their contribute to the association between UPF intake and all-cause mortality, on the other hand, seems to be essential, since no association was identified between UPF intake and cardiovascular mortality or cancer mortality, taken separately, in the whole cohort.
So, the association between UPF consumption and all-cause mortality, in the SUN cohort, seems to be driven by a class of fatalities which could hardly be correlated with the diet, or with specific characteristics of the diet. As an example, a recent paper of the GBD research group stressed that the large majority of diet-associated deaths is attributable to cardiovascular diseases or cancers (2).
Therefore, the potentially causative role of UPF on all-cause mortality proposed by the authors in their conclusion does not seem to be supported by the reported data. A case-control study, allowing a better control of the subjects age, might perhaps provide more relevant information to the topic.
1. Rico-Campà A, Martínez-González MA, Alvarez-Alvarez I, Mendonça RD, de la Fuente-Arrillaga C, Gómez-Donoso C, Bes-Rastrollo M. Association between consumption of ultra-processed foods and all cause mortality: SUN prospective cohort study. BMJ. 2019;365:l1949. https://doi.org/ 10.1136/bmj.l1949.
2. GBD 2017 Diet Collaborators. Health effects of dietary risks in 195 countries, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2019;393(10184):1958-1972. https://doi.org/10.1016/S0140-6736(19)30041-8.
Competing interests: Andrea Poli and Franca Marangoni are respectively Chairman and Responsible for research of NFI-Nutrition Foundation of Italy, a non-profit organization partially supported by 19 food and beverage companies.
Re: Association between consumption of ultra-processed foods and all cause mortality: SUN prospective cohort study
The prospective cohort study by Rico-Campà et al addresses an important question on the risk of ultra processed foods (UPF) and all cause mortality1. The authors correctly state that the study could be subject to residual confounding as a number of confounders could not be controlled for. However, another important limitation of the study with respect to confounding is lack of control for time varying confounders which can affect previous exposure2,3.
In this study body mass index (BMI) was only measured at baseline but in reality a subject’s BMI can change over time which can affect that subject’s use of UPFs. For example, intake of UPFs can lead to a high BMI but later if that subject’s BMI is lowered (secondary to a weight loss diet) it can affect UPF use (the person might stop UPFs or take a smaller amount). Here BMI acts as a classic case of a time varying confounder affected by previous exposure3. Proper assessment of time varying confounders requires use of nuanced methods in causal inference that can account for the time varying nature of these variables2,3,4. We hope that future studies that examine the effect of certain diets and their relation to morbidity or mortality can appropriately control for time varying confounders.
References:
1. Rico-Campà A, Martínez-González MA, Alvarez-Alvarez I, Mendonça RD, de la Fuente-Arrillaga C, Gómez-Donoso C, Bes-Rastrollo M. Association between consumption of ultra-processed foods and all cause mortality: SUN prospective cohort study. BMJ. 2019;365:l1949.
2. Mansournia MA, Altman DG. Inverse probability weighting BMJ 2016; 352:i189.
3. Mansournia MA, Etminan M, Danaei G, et al. Handling time varying confounding in observational research. BMJ 2017;359:j4587.
4. Mansournia MA, Naimi AI, Greenland S. The implications of using lagged and baseline exposure terms in the longitudinal-causal and regression models. Am J Epidemiol 2019 1;188(4):753-759.
Competing interests: Dr Etminan is founder of unravelmedicine.com, an unfunded academic platform that reviews high impact observational studies where this study was recently reviewed