Ultra-Processed Foods and Colorectal Neoplasia: Is There a Link?

Amanda J. Cross, PhD; Marc J. Gunter, PhD


J Natl Cancer Inst. 2023;115(2):117-119. 

Modern lifestyles and the concomitant need for more easily prepared and convenient meals have driven the production of food products that are low cost with a long shelf life, do not necessarily need refrigeration, and are simple and quick to prepare. This has led to the development and growth in the use of numerous food processing techniques and heavy reliance on additives and preservatives. Such ultra-processed foods (UPFs) tend to be high in salt, sugar, fat, and calories and to have poor overall nutritive value; examples include soft drinks, cakes, biscuits, confectionary, and ready meals.[1] The global consumption of UPFs has been rising in recent decades with the largest increases seen in upper-middle income countries and lower-middle income countries.[2] Several surveys assessing individual food intake, household food expenses, or supermarket sales have suggested that UPF products can contribute from 10% of total daily energy intake in some countries, such as Italy,[3] to 48% in Canada[4] and 60% in the United States,[5,6] with some estimates as high as almost 80% in the Netherlands and Germany.[7]

There has been growing interest in the potential health effects of UPFs, and several categorization systems have been proposed to classify foods according to their degree of processing; the most extensively used being the NOVA classification.[1] NOVA assigns foods to 1 of 4 groups: 1) unprocessed or minimally processed foods; 2) culinary ingredients, such as salt, sugar, and oil; 3) processed foods, such as freshly baked or canned, which are produced from NOVA groups 1 and 2 items; 4) UPFs.[1]

Diet is known to have strong links with a myriad of chronic diseases, including cardiovascular disease and cancer. To date, UPFs have been associated with an increased risk of obesity,[8] type 2 diabetes,[9–11] hypertension,[12] cardiovascular and cerebrovascular diseases, depression, and all-cause mortality[9,12–15] in multiple studies from various countries, including France,[11] Spain,[15] and the United Kingdom.[10] Among cancers, colorectal cancer risk is arguably the most affected by diet, with an estimated 54% of cases thought to be associated with modifiable risk factors, such as diet and lifestyle.[16] Obesity, alcohol, and diets high in red and processed meats have been consistently associated with a higher risk of colorectal cancer, whereas diets high in fiber, particularly whole grains, have been associated with a lower risk.[17] In addition, studies have examined dietary patterns, such as a Mediterranean diet, in relation to colorectal cancer, but fewer studies have examined the degree of food processing.

The paper presented in this issue of the Journal by Hang and colleagues[18] is a companion piece to a similar analysis, which examined the association of UPFs with colorectal cancer risk in 3 US cohorts:[19] the Nurses' Health Study I, the Nurses' Health Study II, and the Health Professionals Follow-up Study. Wang et al.[19] reported a positive association between UPFs and colorectal cancer risk that was largely confined to men and was stronger for distal than proximal colon cancer and null for rectal cancer in both sexes.[19] The observed associations remained after adjustment for body mass index and other indicators of the nutritional quality of the diet.[19] These results corroborated similar findings from a multicentric, population-based case-control study in Spain.[20] However, a previous prospective study from the NutriNet-Sante cohort in France failed to find an association between UPFs and colorectal cancer, though this analysis was based on only 153 colorectal cancer cases.[21]

As an extension of this work and to try to better understand where in the colorectal carcinogenesis pathway UPFs might be acting, the new study by Hang et al.[18] examined the association between UPFs and colorectal adenomas in the same 3 US-based cohorts previously examined.[19] Such a focus on colorectal cancer precursor lesions may help determine whether lowering exposure to UPFs may be a viable target for colorectal cancer prevention. Indeed, they showed that UPFs were positively associated with colorectal adenomas and serrated lesions, as well as high-risk polyps. In contrast to the earlier publication on colorectal cancer, the findings presented here show a positive association for all subsites of proximal colon, distal colon, and rectal cancer, as well as within both sexes.

From a mechanistic viewpoint, it is conceivable that consumption of UPFs could raise colorectal cancer risk via obesogenic pathways or those related to poor overall nutrition. However, although it may be challenging to fully control for the effects of obesity and diet, that this study reports statistically significant positive associations between UPFs and colorectal polyps following adjustment for body mass index and indicators of diet quality suggests that other aspects of UPFs may be related to colorectal neoplasia. Emulsifiers and artificial sweeteners, for example, have the potential to increase inflammation in the gut, which is known to be a key driver of colorectal cancer.[22] In addition to pro-inflammatory effects, UPFs may also have detrimental effects on the gut microbiota, which could hypothetically influence colorectal cancer development through the promotion of inflammation and metabolic dysfunction.[23–25] The potential direct carcinogenic properties of specific additives or compounds formed during food processing can be challenging to evaluate; nevertheless, there are specific examples suggesting a link with cancer risk is possible. For example, sodium nitrite is used in the production of processed meats, and animal models and human studies have shown that this additive can promote the formation of carcinogenic N-nitroso compounds.[26] Industrially derived trans fats, which are often found in UPFs, have also been linked to cancer in some studies,[27–29] and acrylamide is another example of a potentially carcinogenic compound formed during heat processing of certain foods.[26] Overall, more experimental research is needed to ascertain the potential impact of UPFs on colorectal neoplasia and disentangle the underlying mechanisms.

Although the study presented here[18] and its companion paper published earlier this year[19] provide complementary findings that warrant further investigation of the role of UPFs in colorectal neoplasia, there are a number of limitations to consider. Estimating intake of UPFs in large cohort studies usually relies on quantifying diet via food frequency questionnaires administered at the beginning of the study, although in this analysis the cohorts had also collected dietary information during the follow-up period. Such questionnaires were generally not designed or validated to accurately measure UPF intake and nor do they capture the expanding range of UPFs available over time. The difficulties of correctly assigning all foods into 1 of the 4 NOVA categories and the difficulties of capturing precise characteristics of various UPFs have previously been addressed.[30] The current study presents findings for the association of UPFs (NOVA group 4) with colorectal neoplasia, however, it does not report what was found for the other 3 NOVA groups. It would be of interest to understand whether an association with processed foods (NOVA group 3) was observed or whether diets rich in unprocessed, whole foods (NOVA group 1) are potentially protective against colorectal adenoma development. Finally, confounding is a major concern as UPF intake is associated with other aspects of the diet as well as other lifestyle and socioeconomic factors that may have an important bearing on cancer risk. Collaborative initiatives, such as pooled analyses across multiple, large-scale cohorts with harmonized data that can enable careful control as well as stratification by important confounding factors, could be one approach to begin to disentangle these effects. Other methodological approaches to characterize the type of food processing as well as to identify the etiologically relevant components of UPFs should also be considered.

In summary, this investigation, along with others, suggests that consumption of UPFs may have long-term detrimental effects on health. Specifically, this study reports provocative and possibly important findings relating to the consumption of UPF with colorectal neoplasia risk; however, the findings need to be replicated in other large-scale prospective cohorts with detailed dietary data that enable estimation of UPF consumption. In addition, there is a need for studies examining the potential underlying mechanisms linking UPFs to cancer. If the association between UPFs and colorectal neoplasia is found to be causal, then reducing exposure to these foods in the diet could be one strategy by which colorectal cancer risk could be lowered. Given the high incidence of colorectal cancer and its growing burden in low- to middle-income countries where the consumption of UPFs has been rising, such interventions may represent an important public health priority.