Understanding the association between diet and nutrition in upper gastrointestinal cancer

Abstract

Human vulnerability to cancers of the upper gastrointestinal tract is strongly influenced by environmental factors. The esophagus, in particular, is highly vulnerable to the combined effects of exposure to environmental carcinogens and malnutrition, particularly in certain extreme environments of the developing world. Even in high-income countries, dietary carcinogens and nutrition play a major role in the etiology of oropharyngeal, esophageal and, to a lesser extent, gastric cancers, but the mechanisms are poorly understood. A thorough understanding of the biological mechanisms underlying the vulnerability of these organs to neoplasia would shed further light on the etiology of upper gastrointestinal cancers in all environments. In the meantime, the epidemiological evidence suggests that the risks can be minimized by dietary patterns that adhere closely to current public health recommendations, coupled with maintenance of body mass index within the healthy range.

Keywords:

• carcinoma
• diet
• esophagus
• food
• gastric
• oropharyngeal

The incidence rates for cancers of the oropharynx, esophagus and stomach vary dramatically, both between geographical regions and within populations undergoing environmental change. The epithelia of the upper alimentary tract are exposed to an extraordinary variety of environmental agents over an individual’s lifespan, including nutrients and other biologically active components of food, unavoidable environmental pollutants and toxic substances consumed for pleasure. There is growing evidence that many of these factors modulate the risk of neoplasia in the digestive organs, and that differing levels of exposure to food constituents account for much of the variation in incidence. This article aims to summarize the present state of knowledge, but it is emphasized that much of the evidence comes from observational studies which must be interpreted with caution and can rarely provide conclusive mechanistic explanations.

Following the full recognition of the links between cigarette smoke and lung cancer, there has been much interest in the role of environmental carcinogens in upper intestinal neoplasia. Ethanol in alcoholic beverages is classed as a carcinogen by the International Agency for Research on Cancer and, as is well documented in the World Cancer Research Fund report on Food, Nutrition and the Prevention of Cancer [1], alcohol consumption is a major dose-dependent risk factor for both oropharyngeal cancer (OC) and esophageal squamous cell carcinoma (ESCC). This is not, however, true for gastric cancer or for esophageal adenocarcinoma (EAC), which has recently undergone a dramatic and largely unexplained rise in incidence in many high-income countries. Indeed, there is limited observational evidence for a small protective effect of low-level alcohol consumption against EAC and adjacent cancers of the esophago-gastric junction [2]. The carcinogenicity of alcoholic beverages in relation to ESCC, which is probably due to both ethanol itself and acetaldehyde, a known carcinogen derived from ethanol metabolism, is amplified considerably by the concomitant use of tobacco products.

Despite the widespread use of alcohol, ESCC is not amongst the most common causes of death from cancer in high-income countries, whereas in certain rural regions of the developing world, notably Linxian in China, and the Caspian littoral zone of Iran (where alcohol consumption is negligible), it is the leading cause of death in adults. These regions often show very steep differences in incidence across short geographical distances, and although genetic factors appear to play some role [3], the prevalence of powerful environmental risk factors is obvious. In many cases, the risk of disease appears to be correlated with extreme climatic and agricultural conditions, prompting the hypothesis that poor soil quality and restricted diets, low in micronutrients and enriched perhaps with carcinogens, play a major role in disease etiology [4]. Observational studies have often confirmed the presence of highly carcinogenic food and water-borne N-nitrosamines, frequently combined with exposure to mucosal irritants such as salt, spices and excessively hot traditional foods and beverages. Thus, ESCC in these high-incidence regions is evidently a multifactorial disease driven by the combined effects of malnutrition and prolonged exposure to dietary insults. However, attempts to confirm the role of malnutrition and to reduce disease by correcting micronutrient deficiencies with dietary supplements have been generally disappointing. For example, in a placebo-controlled trial in Linxian, poorly nourished individuals with cytological evidence of esophageal squamous dysplasia underwent 6 years of supplementation with multivitamins, but showed no evidence of reduced mortality from esophageal or gastric cancer after 6 or 20 years of follow-up [5].

The tragically high levels of ESCC in localized geographical hotspots suggest that under extreme environmental conditions, the esophagus is vulnerable to multiple carcinogenic insults. To what extent do the combined effects of food-borne carcinogens and the very different nutritional characteristics of western diets drive cancers of the upper digestive tract in prosperous industrialized societies? Amongst the most commonly consumed conventional foods, red and processed meats have long been under suspicion as causes of gastrointestinal cancers. This is a plausible hypothesis because meats of various types are known to contain mutagens. These include heterocyclic amines and polycyclic aromatic hydrocarbons generated during cooking at high temperatures, and nitrosamines derived by gut bacterial metabolism of unabsorbed protein residues and nitrates from preserved meat products. The World Cancer Research Fund (WCRF) panel (2007) found no evidence of any association between meat consumption and OC, and only limited evidence of a causal role for meat in esophageal or gastric cancers [1]. Nevertheless, recent observational evidence suggests that red and processed meat do play some role. For example, a recent European cohort study (the most reliable type of epidemiological investigation) showed that the risk for EAC was positively associated with intake of heme from meat products, and varied more than twofold between the upper and lower quartiles of processed meat consumption [6]. However, though a recent meta-analysis showed an increased risk of OC associated with high consumption of processed meats, there were no apparent associations with total, red, carcass meat or poultry [7].

Overall, the carcinogenic effects of dietary components other than alcohol in the upper gastrointestinal tract seem relatively weak in high-income populations, but there is evidence for protective effects of diets rich in fruits, vegetables and whole-grain cereals. The first WCRF panel report published in 1997 [8] concluded that there was convincing evidence that non-starchy vegetables, fruits, and foods containing carotenoids and vitamin C were all protective against OC, esophageal and gastric cancers. In the 2007 report [1], the strength of this evidence was judged to have declined from ‘convincing’ to ‘probable’, probably because the earlier report was more reliant on case–control studies, which are prone to bias. However, subsequent systematic reviews continue to show a protective role of fruits against gastric cancer [9], and both fruits and vegetables against ESCC [10] and EAC [11]. Importantly, given the growing importance of EAC, there is some evidence that the risk of its precursor lesion, Barrett’s esophagus, is also inversely related to fruit and vegetable intake [12,13]. A recent study of patients referred for gastroesophageal endoscopy showed that adherence to WCRF guidelines on diet and exercise, previously shown to be protective against gastric and esophageal cancers amongst others [14], was also associated with a lower risk of both Barrett’s esophagus and its progression to EAC [15].

Given the inescapable complexity of human diets and lifestyles, future epidemiological studies on their multiple pro- and anti-carcinogenic effects should ideally adopt multivariate statistical techniques applied to datasets encompassing as many of the relevant exposures as possible. The data can then be simplified to describe a few dietary patterns against which to compare the relative risks of disease. Several groups have already used this approach to explore the impact of supposedly high- and low-risk patterns, defined a priori, on cancers of the upper alimentary tract. For example, Bosetti et al. used data from three case–control studies conducted in northern Italy to show that close adherence to a traditional Mediterranean dietary pattern was associated with a lower risk of OC, laryngeal and esophageal cancers [16]. As might be expected for diseases of multifactorial origin, the relative risks associated with different dietary patterns (ca. two- to threefold) exceeded those previously reported for meat and plant foods separately. Li et al. used dietary data from questionnaires completed by nearly half a million Americans enrolled in the (NIH)-AARP Diet and Health cohort study [17]. During 11 years of follow-up, subjects scoring highly for a dietary pattern close to that defined by the 2005 Dietary Guidelines for Americans had a significantly lower risk of both ESCC and EAC. There was an inverse association between the risk of ESCC and adherence to a standardized Mediterranean dietary pattern adapted to the American population. However, neither dietary pattern showed any relationship with gastric cardia or non-cardia adenocarcinoma.

Cancers of the upper gastrointestinal tract are clearly multifactorial diseases driven by very different combinations of pro-and anti-carcinogenic factors in widely disparate environments. Even in high-income countries, the welcome decline in gastric cancers over the last 50 years has been accompanied by a dramatic and puzzling rise in the incidence of EAC. At present, we lack a deep understanding of these phenomena. There is clearly a need for further mechanistic research on epithelial carcinogenesis combined with robust epidemiological studies conducted both in high-income countries and in the puzzling disease hotspots that undoubtedly have much more to tell us about upper intestinal neoplasia. Certain specific dietary components, including, for example, the dietary fiber and fatty acid constituents of Mediterranean diets, warrant particular attention. In the meantime, the risk of oropharyngeal and esophageal neoplasia can best be minimized by adopting lifestyles that adhere closely to current public health recommendations. Dietary advice should include an emphasis on plant foods, particularly non-starchy vegetables, fruits and whole grain cereals, coupled with limited consumption of red and processed meat. Finally, whereas overweight and obesity are clear risk factors for EAC and cancer of the gastric cardia [18], leanness appears unexpectedly to be associated with increased risk of both OA [19] and ESCC [20]. The reasons for these contrasting relationships are unknown, although further research to investigate a possible confounding effect of dysphagia and weight loss due to advancing esophageal disease seems called for. In any case, maintenance of body weight within a healthy range (BMI >18–<25) is clearly prudent.

Financial & competing interests disclosure

IT Johnson is an Emeritus Fellow at The Institute of Food Research and is an external member of the UK SACN Working Group on Carbohydrates and Health and the Scientific Advisory Committee of the British Nutrition Foundation. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.